UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Polarization-sensitive optical coherence tomography imaging of articular cartilage Zhou, Xin 2017

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2017_september_zhou_xin.pdf [ 9.39MB ]
Metadata
JSON: 24-1.0354542.json
JSON-LD: 24-1.0354542-ld.json
RDF/XML (Pretty): 24-1.0354542-rdf.xml
RDF/JSON: 24-1.0354542-rdf.json
Turtle: 24-1.0354542-turtle.txt
N-Triples: 24-1.0354542-rdf-ntriples.txt
Original Record: 24-1.0354542-source.json
Full Text
24-1.0354542-fulltext.txt
Citation
24-1.0354542.ris

Full Text

 POLARIZATION-SENSITIVE OPTICAL COHERENCE TOMOGRAPHY IMAGING OF ARTICULAR CARTILAGE by  Xin Zhou  B. Eng., Zhejiang University, 2014   A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES  (Electrical and Computer Engineering)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  August 2017  © Xin Zhou, 2017   ii Abstract  Articular cartilage is the connective tissue protecting the joints, which can be divided into four structural zones from shallow to deep: superficial zone, transitional zone, deep zone and calcified zone. Osteoarthritis, a common joint disease, is associated with the progressive degeneration of the cartilage structure. The destruction progressively develops from the superficial zone towards the deep zone as the disease progresses. Thus, visualization of articular cartilage structural zones would significantly facilitate cartilage disease diagnosis, repair, regeneration, and transplantation.  Polarization sensitive optical coherence tomography (PS-OCT) is a powerful non-invasive imaging modality capable of evaluating the birefringent properties in biological tissue such as collagen. The second harmonic generation (SHG) imaging in multiphoton microscopy (MPM) can provide complementary high-resolution imaging of the microscopic structure of collagen fibers. In this thesis, we apply both PS-OCT and SHG imaging on articular cartilage.    An automated 3-D segmentation method based on PS-OCT phase retardation measurement is developed to differentiate the structural zones of articular cartilage. Since the collagen fiber organization varies from the tissue surface to deep regions, the depth-resolved phase retardation measured by PS-OCT is utilized to distinguish and segment the depth-related structural zones of cartilage tissue. The segmentation results are validated by the high-resolution SHG imaging. This method offers a novel quantification and tissue segmentation approach based on the phase retardation measurement by PS-OCT.    iii  A comparison between PS-OCT and SHG imaging on articular cartilage is also carried out. Based on the multilayer architecture of articular cartilage, various features extracted from PS-OCT and SHG are compared along the tissue depth. The segmentation method is implemented to distinguish the tissue layers based on the birefringence property. The segmentation results match well with the different quantification results achieved from the top-view and side-view illumination PS-OCT, and some features extracted from SHG, such as the SHG intensity and the collagen fiber orientation. The results show reasonable association between the tissue birefringence detected by PS-OCT and the fiber organization detected by SHG microscopy. PS-OCT and SHG are capable to analyze both the macro and micro characteristics of collagen fibers in articular cartilage, showing great potential in detecting related disease progression.   iv Lay Summary  Articular cartilage protects us during daily motion and sports. It has different structural zones along its depth, which is related to the progression of cartilage disease. Thus, visualization of articular cartilage depth-dependent structure is very important.  PS-OCT and MPM are two promising modalities to image the depth-dependent structure of articular cartilage. PS-OCT is powerful, non-invasive, and more suitable for clinic application, while MPM provides complementary high-resolution imaging of the microscopic structures in cartilage. In this thesis, we apply both of them to do the image.    A method based on PS-OCT is developed to differentiate articular cartilage structural zones. A comparison between PS-OCT and MPM imaging on a same cartilage sample is also carried out, where various features extracted from these two methods are compared along the tissue depth. The results show reasonable association. Thus, PS-OCT is potential to analyze the microscopic characteristics in articular cartilage, showing great clinic value.    v Preface  In Chapter 2, the PS-OCT and MPM systems have been developed in the UBC Biophotonics laboratory by previous students. I did improvements on the hardware and the performance of the PS-OCT system.   Chapter 3.  A preliminary version has been published in a conference paper:  Xin Zhou, Myeong Jin Ju, Lin Huang, and Shuo Tang, “Correlation between polarization sensitive optical coherence tomography and SHG microscopy in articular cartilage.” In Proc. of SPIE Vol, vol. 10053, pp. 1005319-1 (2017).  A more complete version has been prepared as a journal paper:  Xin Zhou, Myeong Jin Ju, Lin Huang, and Shuo Tang, “Phase-retardation-oriented slope-based segmentation by polarization sensitive optical coherence tomography: a structural guidance in articular cartilage” (in preparation).  I was responsible for the system operation, experiments, method development, data analysis, as well as manuscript composition. Meyong Jin Ju, a senior student in our lab, built the PS-OCT system and provided a lot of knowledge and experience about the system operation. Lin Huang, my colleague, offered some critical help during my experiments. My supervisor, Dr. Shuo Tang, involved throughout the whole project in experiments design, results discussion, and manuscript composition.    vi Chapter 4. A version of Chapter 4 is under preparation as a journal paper:  Xin Zhou, Lin Huang, and Shuo Tang, “Correlation between polarization sensitive optical coherence tomography and SHG microscopy in articular cartilage imaging” (in preparation).  I was responsible for the experiments, data analysis, as well as manuscript composition. Lin Huang, my colleague, offered some help during my experiments. My supervisor, Dr. Shuo Tang, first generated the basic idea for this project, and offered great help in the results analysis and manuscript composition.               vii Table of Contents  Abstract .......................................................................................................................................... ii	Lay Summary ............................................................................................................................... iv	Preface .............................................................................................................................................v	Table of Contents ........................................................................................................................ vii	List of Tables ..................................................................................................................................x	List of Figures ............................................................................................................................... xi	List of Abbreviations ...................................................................................................................xv	Acknowledgements .................................................................................................................... xvi	Dedication .................................................................................................................................. xvii	Chapter 1: Introduction ............................................................................................................... 1	1.1	 Articular cartilage ....................................................................................................... 1	1.2	 Optical Coherence Tomography ................................................................................. 3	1.3	 Polarization sensitive optical coherence tomography ................................................. 4	1.4	 Multiphoton microscopy ............................................................................................. 5	1.5	 Objective & Contribution ........................................................................................... 6	1.6	 Outline ......................................................................................................................... 7	Chapter 2: PS-OCT and MPM background ................................................................................ 9	2.1	 Principle of PS-OCT ................................................................................................... 9	2.2	 PS-OCT system ......................................................................................................... 11	2.3	 Principe of MPM....................................................................................................... 14	2.4	 MPM system ............................................................................................................. 16	  viii Chapter 3: PS-OCT phase retardation slope based segmentation ............................................. 19	3.1	 Introduction ............................................................................................................... 19	3.2	 Methods and materials .............................................................................................. 23	3.2.1	 Sample preparation ........................................................................................... 23	3.2.2	 Experiment scheme ........................................................................................... 23	3.2.3	 PS-OCT phase retardation multi-stage slope analysis ...................................... 25	3.3	 Results ....................................................................................................................... 28	3.3.1	 PS-OCT analysis ............................................................................................... 28	3.3.2	 SHG imaging with the guidance of PS-OCT analysis ...................................... 30	3.3.3	 PS-OCT multiple illumination angle results ..................................................... 33	3.4	 Summary ................................................................................................................... 36	Chapter 4: Comparison between PS-OCT and MPM in articular cartilage imaging ................ 38	4.1	 Introduction ............................................................................................................... 38	4.2	 Methods and materials .............................................................................................. 40	4.2.1	 Experiment design ............................................................................................ 40	4.2.2	 PS-OCT slope-based analysis ........................................................................... 43	4.2.3	 Quantitative analysis methods for MPM images .............................................. 44	4.3	 Results ....................................................................................................................... 50	4.3.1	 PS-OCT side-view measurement results .......................................................... 50	4.3.2	 MPM side-view measurement results ............................................................... 53	4.3.3	 Quantification results comparison between PS-OCT and MPM ...................... 56	4.4	 Summary ................................................................................................................... 59	Chapter 5: Conclusion and future work .................................................................................... 61	  ix 5.1	 Conclusion ................................................................................................................ 61	5.2	 Future directions ....................................................................................................... 62	Bibliography .................................................................................................................................64	   x List of Tables  Table 2.1: Specifications of the JMT system used in this study. .................................................. 14	Table 2.2: Specifications of the MPM system (with the 60× objective) used in this study. ........ 18	Table 3.1 Segmentation results summary ..................................................................................... 30	Table 4.1 The slope of the selected phase retardation B-scans by single-stage slope estimation 52	   xi List of Figures  Fig. 1.1 Schematic, cross-sectional diagram of healthy articular cartilage. A, cellular organization in the zones of articular cartilage; B, collagen fiber architecture [39]. ........................................... 2	Fig. 1.2 Comparison of resolution and imaging depth for OCT and other imaging modalities. .... 4	Fig. 2.1 Schematic diagram of advanced JMT system [51]. LP: linear polarizer, PC: polarization controller, FC: fiber collimator, M: mirror, PBS: polarizing beam splitter, BS: beam splitter, H- and V-BPD: balanced photo-detector for horizontally and vertically polarized signals, respectively. .................................................................................................................................. 13	Fig. 2.2 Physical principle of MPM [53]. ..................................................................................... 15	Fig. 2.3 Schematic diagram of the MPM system. ......................................................................... 17	Fig. 3.1 Experiment design. (a) Photograph of the sample with a blue rectangle indicating the region where the tissue cube was sectioned; (b) model of tissue with the scanning protocols of the two imaging methods indicated: the red arrow is the illumination direction for PS-OCT, the black dashed line is the normal direction to surface and the large rectangle represents a cross-sectional B-scan. The small rectangles in green are the approximate areas where SHG images are measured from the top illumination of the tissue. ......................................................................... 24	Fig. 3.2 Segmentation results from Multi-stage slope-based analysis. (a-b): PS-OCT images measured from the top-view, (a) is intensity OCT, (b) is accumulated phase retardation with the segmentation results indicated as black dashed lines; (c): Modified accumulated phase retardation A-line from the three positions marked by red arrows in (b); (d) 3-D rendering of the segmentation results. ..................................................................................................................... 29	  xii Fig. 3.3 SHG images measured with the guidance of PS-OCT analysis. (a-d): SHG images measured from the top of tissue with around 10~40 µm, from the surface, roughly at the position ‘1’ marked in Fig. 3.2 (b) (from natural surface), frame gap: 10 µm; (e-h): SHG images measured at the depth of around 80 ~ 140 µm off the surface, indicated by the position ‘2’ in Fig. 3.2 (b), gap: 20 µm; (i-l): SHG images measured the depth of around 600±100 µm off the surface, corresponding to the position ‘3’ in Fig. 3.2 (b), frame gap: 20 µm. Image size: 130 * 130 µm.  Scale bar is 30 µm. ........................................................................................................................ 32	Fig. 3.4 PS-OCT measurement results at different illumination angles. (a-i) PS-OCT B-scan images measured at illumination angles as indicated. The illumination angle varies in X-Z plane. In each B-scan image penal shown, the accumulated phase retardation image is on the top, and the Intensity OCT image is at the button. ..................................................................................... 33	Fig. 4.1 Experiment design. (a) Photograph of the sample with a blue rectangle indicating the region where the tissue cube was sectioned; (b) Morphological diagram of the depth-dependent ECM changing in cartilage tissue; (c) Tissue cube model indicating the scanning protocol in PS-OCT imaging: red arrows are the illumination beams, the gray rectangle represents one of the B-scans from the cross-section view (in X-Z plane), and the blue one represents one of the B-scans from the en-face view (in X-Y plane). (d) Tissue cube model indicating the scanning protocol in MPM imaging: each green square (in X-Z plane) represents a region of interest (ROI) from the tissue surface to deep region. The coordinate system is marked on the upper right corner of the tissue cube model in (d). ............................................................................................................... 42	Fig. 4.2 Fiber orientation measurement by Hough transform with intermediate results of the procedures. (a): SHG images measured from the side-view, with 16 sub-images divided by the   xiii red lines; (b): pre-process results of Sub-image 𝐈𝟐𝟒, marked as yellow shadow in (a); (c): Hough transform results of (b); (d): Angle distribution achieved form the Hough transform results. ..... 47	Fig. 4.3 Collagen fiber organization solved from SHG images. (a-c) SHG images from side-view (XZ plane), at the depths of around 20 µm, 100 µm, and 600 µm from the tissue surface, respectively. Image size: 130 * 130 µm. (d) the angle distributions for the collagen fibers in the (a)-(c). The coordinate system is marked in (c). ........................................................................... 48	Fig. 4.4 (a-i) PS-OCT images measured from the side-view of the cartilage tissue at different depths. The top ones represent the intensity B-scans, while the bottom ones are the accumulated phase retardation B-scans. The depth is labeled in each image. All the B-scans shown are in XY plane. The averaging region to calculate the slope is marked in (a). Scale bar is 500 µm. .......... 51	Fig. 4.5 Summing projections of the resliced image stack shown in Fig. 4.4. (a) Summing projection in YZ plane; (b) Summing projection in XZ plane. Surface and Bone boundary are indicated by red arrows. Sample size is ~3 mm (X) by 2 mm (Z). Scale bar is 500 µm. ............. 53	Fig. 4.6 MPM images measured from the side-cut of the cartilage cube (parallel to the plane X-Z); image (a) to (l) are captured from the natural surface to the deep region. All the rendered images are taken from the same depth (30 µm off) from the cutting surface. The depth of each image is indicted. All the images are oriented in a same fashion. green: SHG signal, red: TPEF signal. Image size: 130 * 130 µm. Scale bar: 30 µm. ................................................................... 55	Fig. 4.7 Comparison of the depth-related information obtained from various analyzing methods. The black and light-blue curve has the unit of degree. The red curve describes the Intensity varying within a range from 0 to 255. The dark-blue curve is the dominating collagen fiber direction in the X-Z plane, solved by the Hough-transform based method. The green curve is the   xiv slope (unit: rad/µm) from the PS-OCT side-view measurement, but is multiplied by 𝟏𝟎𝟑 times to match the magnitude. .................................................................................................................... 56	   xv List of Abbreviations  BPD Photo-detectors BS Beam splitter CT Computed tomography ECM Extracellular matrix ESNR Effective signal-to-noise ratio  FOV Field of view FWHM Full width at half maximum JMT Jones matrix tomography MPM Multiphoton microscopy MRI Magnetic resonance imaging NA Numerical aperture OCT Optical coherence tomography PD Polarization diversity PS-OCT Polarization sensitive optical coherence tomography SHG Second harmonic generation SNR Signal to noise ratio 3D             Three dimensional     xvi Acknowledgements  I would like to express my gratitude to my supervisor, Dr. Shuo Tang, for her kindly encouragement and guidance. With her kindness, instruction, and support, I learn a lot during my graduate study. I am grateful to have such a great mentor.  I would also like to thank the past and present members of the Bio-photonics Lab for their expertise, assistances and support.   Finally, I would like to thank my family and friends for their unconditional support through the course of my studies.     Xin Zhou The University of British Columbia   xvii Dedication    Time forks perpetually toward innumerable futures     1 Chapter 1: Introduction  1.1 Articular cartilage  Articular cartilage is the smooth, white tissue that covers the ends of bones in joints and plays an integral role in facilitating the transmission of loads within bones and protecting from injuries during daily motion and sporting activities. Articular cartilage is composed of a dense extracellular matrix (ECM) of collagen with a sparse distribution of chondrocytes [1]. Collagen is the most abundant type of protein in ECM.   Fig. 1.1 shows the tissue structure of articular cartilage.  Articular cartilage is composed of various zones from the joint surface to the bone, which are the superficial zone, transitional zone (or middle zone), deep zone, and calcified zone [2]. The superficial zone is the only zone that contains type I collagen, which forms large fiber bundles aligned almost parallel to the articular surface. The superficial zone also contains a relatively high number of flattened chondrocytes. The integrity of this thin layer is imperative in the protection and maintenance of deeper zones from shear stresses; The transitional zone provides an anatomic and functional bridge between the superficial zone and deep zone. The transitional zone contains mostly fibrillar type II collagen and the collagen fibrils are organized obliquely. The chondrocytes are more spherical and at lower density in this layer than in the superficial zone. Functionally, the transitional zone is the first line of resistance to compressive forces; The deep zone is composed of mostly fibrillar type II collagen where the fibrils are aligned perpendicular to the bone surface. The fibril diameter increases with depth from the transitional zone to the deep zone [3]. The deep zone is responsible for providing resistance to compressive forces. The chondrocytes in deep zone are   2 typically arranged in columnar orientation, parallel to the collagen fibers and perpendicular to the joint surface; The calcified zone secures the cartilage to bone by anchoring the collagen fibrils of the deep zone to subchondral bone.   Osteoarthritis, as the most common form of joint disease, is associated with the progressive degeneration of the cartilage ECM [1]. As the disease progresses, the ECM destruction and structural damage progressively develops along the depth into tissue, from the superficial zone towards the deep zone. Thus, the visualization and quantification of the depth-dependent structures and features in articular cartilage would significantly facilitate cartilage disease diagnosis, repair, regeneration, and transplantation. Traditional methods for clinical imaging of articular cartilage include radiography, magnetic resonance imaging (MRI), and arthroscopy. However, those techniques lack the resolution and sensitivity to visualize the changes of the ECM.   Fig. 1.1 Schematic, cross-sectional diagram of healthy articular cartilage. A, cellular organization in the zones of articular cartilage; B, collagen fiber architecture [39].   3 1.2 Optical Coherence Tomography  Compared to X-ray computed tomography (CT), MRI, and ultrasound, optical imaging can provide much higher resolution. Optical Coherence Tomography (OCT) [4] is a noninvasive optical imaging technique that is based on the principle of low coherence interferometry. It measures the depth-resolved reflections of light to obtain the cross-sectional images of tissue with micrometer resolution. With its capability of accessing tissue features at the micrometer scale noninvasively, OCT is considered as an “optical biopsy” approach, in which the information of tissue can be obtained noninvasively through the back-reflected light, compared to traditional biopsy approach which requires physical excision of tissue biopsies for histopathology examination.   OCT has several advantages compared to other imaging modalities [5]: 1) OCT uses a low power near-infrared (NIR) light as excitation, which is non-radiative. It keeps the sample free of photochemical or photothermal damage; 2) OCT has high resolution of around 2 to 15 µm ; 3) OCT imaging is fast at a speed of 100k-200k A-lines per second, which enables 3-D imaging of biological tissues within a realistic measurement time. Therefore, OCT is suitable for in vivo imaging.  Fig. 1.2 shows the resolution and penetration depth of several imaging modalities. The penetration depth of OCT is usually limited to a few millimeters due to tissue scattering. Compared to microscopy techniques, such as confocal microscopy, OCT has lower resolution but higher penetration depth. Thus OCT fills the gap between ultrasound and microscopy in terms of the imaging resolution and depth. OCT has been utilized in various clinical applications, such as ophthalmology [6-11], dermatology [12-19], dentistry [20-23], and cardiology [24-27], showing significant clinical value and potential.   4  Fig. 1.2 Comparison of resolution and imaging depth for OCT and other imaging modalities.  1.3 Polarization sensitive optical coherence tomography  Several functional OCT extensions have been developed, extending the clinical application of OCT by providing functional information about live, intact tissue beyond just structural characteristics. Polarization-sensitive OCT (PS-OCT) is a representative functional extension of OCT techniques. PS-OCT utilizes both the scattering and polarization properties of light, and extends the capability of traditional OCT by allowing the characterization of tissue birefringence. It is known that several types of biological tissues possess microscopic fibrous structures such as collagen fibers and nerve fibers. Those microscopic fibrous structures usually have a smaller diameter than the resolution of OCT and it’s hard to differentiate the subtle changes in the fibrous structure by standard OCT. However, tissue birefringence occurs in those fibrous structures where the refractive index along or perpendicular to the fiber direction is different, which defines a fast and a slow axis. When the fibers are highly organized over a sufficient range,   5 the phase retardation generated between the fast and slow axis can be measured by PS-OCT with very high sensitivity. PS-OCT has been applied to study birefringence properties in tissues such as tendon [28], muscle [29] and myocardium [30], skin [31], retinal nerve fiber layer [32-35] and some other ophthalmological areas [36-38]. Although PS-OCT is capable of characterizing the birefringence property of tissues, its resolution is yet insufficient to visualize individual collagen fibers in cartilage tissues.  1.4 Multiphoton microscopy In order to visualize collagen fibers, a microscopy technique with higher resolution than OCT is necessary. Multiphoton Microscopy (MPM) is a nonlinear microscopy technique which can image collagen fibers and cells in tissues with high resolution. By using a short-pulse laser such as a femtosecond laser to excite nonlinear signals, MPM can achieve high axial and lateral resolution comparable to that of confocal microscopy without having to use pinholes. The MPM imaging used in this study includes two-photon excitation fluorescence (TPEF) [39] and second harmonic generation (SHG) [40]. SHG microscopy is particularly suitable to study collagen, which is found in most connective load-bearing tissues such as cornea, skin, and cartilage. TPEF detects the auto-fluorescence signal from cells, such as the chondrocytes in cartilage tissue. MPM can achieve 3-D microscopic imaging of cells and collagen fibers with deep penetration depth, and less photodamage and photobleaching, compared to confocal microscopy. Thus MPM is suitable to study live cells and tissues [41]. Compared to PS-OCT, MPM can provide high resolution imaging of the collagen fiber organization and orientation in articular cartilage.    6 1.5 Objective & Contribution Optical imaging techniques have much higher resolution than the traditional clinical imaging modalities such as CT, MRI and ultrasound. Among the optical techniques, OCT shows clinical potential to differentiate healthy and degenerative cartilage tissues by its capability as “optical biopsy”. However, the contrast in standard OCT is the backscattered light, which is difficult to differentiate the subtle changes of ECM in the early stage of cartilage degeneration. In comparison, PS-OCT measures the tissue birefringence, which is very sensitive to the organization and orientation of the collagen fibers. Therefore, in this thesis, we will explore the potential of PS-OCT in imaging the articular cartilage tissue. Meanwhile, the resolution of PS-OCT is still not high enough to resolve individual collagen fibers in the ECM. MPM with its high resolution and the specific SHG contrast for collagen fibers, can provide complementary information about the ECM structure and organization at the micrometer scale. In this thesis, we will apply both PS-OCT and MPM to study the cartilage tissue. With PS-OCT and MPM, more detailed information about the ECM of articular cartilage can be obtained.  Articular cartilage has a zone structure, which consists the superficial, transitional, deep, and calcified zones from the surface to the subchondral bone. The different zones show different collagen structures and properties. Cartilage degeneration also advances from the superficial zone gradually towards the deeper zones as the disease progresses. Therefore, it has significant meaning to differentiate and visualize the different tissue properties over the different zones. It is the objective of this thesis to study the various tissue properties in the different zones using PS-OCT and MPM.    7 A segmentation method based on the PS-OCT phase retardation imaging is developed to differentiate the different zones based on the birefringence properties.  The MPM imaging is also applied to investigate the collagen fiber organization at the different depths of cartilage tissue. The PS-OCT measured tissue birefringence properties are compared with the depth dependent fiber organization properties measured by the MPM imaging. It is shown that highly aligned collagen fibers would generate higher tissue birefringence than more randomly orientated collagen fibers. Quantitative information such as the slope of the phase retardation, SHG intensity, and collagen fiber orientation angle are extracted and characterized in the cartilage tissue over the different zones.    Through our study, the capability of PS-OCT in imaging and characterizing the depth dependent ECM properties of articular cartilage is investigated. Good relationship is observed between the tissue level properties quantified by PS-OCT with the microscopic fiber structures measured by MPM. Our results show that PS-OCT has great potential in quantifying tissue birefringence, which is directly related to the changes of ECM in articular cartilage. Thus, PS-OCT can be a great clinic tool in investigating cartilage diseases such as osteoarthritis.  1.6 Outline In Chapter 1, a brief introduction on articular cartilage, PS-OCT, and MPM is presented.    In Chapter 2, the general principles of PS-OCT and MPM are introduced, with an overall description about the imaging systems.    8 In Chapter 3, an automated 3-D segmentation method based on PS-OCT in detecting the structural zones of the articular cartilage is presented and applied on swine articular cartilage tissue. The segmented layers are validated by high-resolution MPM imaging.   In Chapter 4, a comparison between PS-OCT and SHG microscopy results on articular cartilage imaging is conducted. Both the imaging modalities are utilized to resolve the depth-dependent features from the cartilage natural surface to deep region. The results show reasonable association between the tissue birefringence detected by PS-OCT and the fiber organization detected by SHG microscopy.   Chapter 6 concludes the dissertation and describes some future directions.   9 Chapter 2: PS-OCT and MPM background  The general principles of PS-OCT and MPM are presented in this chapter. The corresponding imaging systems used in our experiments are also introduced.     2.1 Principle of PS-OCT PS-OCT is a functional extension of OCT techniques. There are several different approaches to achieve PS-OCT, such as the circular-polarization-based method [42], Stokes-parameter-based method [43], Mueller-matrix-based method [44], and Jones-matrix-based method [45]. Our system is Jones-matrix-based method.  Jones-matrix-based OCT, or Jones matrix tomography (JMT) in short, is a method that determines the polarization properties of a sample by measuring the round-trip Jones matrix of the sample [46] or its similar matrix [45, 47-49]. This method can be implemented both with a bulk interferometer or a single-mode-fiber based interferometer. With the bulk interferometer, this method provides a round-trip Jones matrix of the sample directly. As for the single-mode fiber interferometer, a similar matrix of the round-trip Jones matrix of the sample is obtained. From the round-trip Jones matrix or its similar matrix, the phase retardation, di-attenuation, and relative optic-axis orientation can be obtained.  Jones matrices are operators that describe the polarization properties of optical medium, such as lenses, optical fibers, other optical elements, as well as biological tissues. In the general detection of JMT, at least two polarization states of incident lights are required to get the polarization   10 information of the sample. Assuming the two polarization states of the incident lights are  E()* and E()+. The incident lights enter the sample and then get backscattered. The corresponding polarization states of the output backscattered lights are E,-.*  and E,-.+ . The output light polarization state is related to the input light polarization state through the Jones matrix as follows [65]:  E,-.* 𝑧 = 𝐻,-.*(𝑧)𝑉,-.*(𝑧) = J 𝑧 E()* = J 𝑧 𝐻()*𝑉()*                                (2.1) E,-.+ 𝑧 = 𝐻,-.+(𝑧)𝑉,-.+(𝑧) = J 𝑧 E()+𝑒(7 = J 𝑧 𝐻()+𝑉()+ 𝑒(7                   (2.2)  where H and V represent the horizontal and vertical components of each polarization state, and  Ψ is the phase difference between the two polarization states of the incident lights. Here J 𝑧  is the Jones matrix, a 2×2 matrix, which describes the polarization property of the sample. Eqs. (2.1) and (2.2) can be combined as follows:  H:;<*(z) H:;<+(z)V:;<*(z) V:;<+(z) = J z H?@* e?7H?@+V?@* e?7V?@+                               (2.3)  In the JMT system, based on the two polarization states of the input and output lights, the Jones matrix J 𝑧  can be measured. The measured Jones matrix can be simplified as:  J 𝑧 = A λ* 00 λ+ AE*                                                 (2.4)   11  Here λ*,+ and A are the complex eigenvalues and the eigenvector matrix of J 𝑧 . The eigenvalues of a matrix can be obtained through matrix diagonalization [45] or the following equation [50]:  𝜆*,+ = 𝑇/2 ± 𝑇+/4 − 𝐷                                                 (2.5)  where T and D are the trace and the determinant of J 𝑧 , respectively.  The phase retardation of the sample, δ z , is then obtained as the phase difference between the two complex eigenvalues as:  δ z = Arg 𝜆*𝜆+∗ ∶ 0 ≤ 	Arg 𝜆*𝜆+∗ 	≤ 	𝜋Arg 𝜆*∗𝜆+ ∶ 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒	                                        (2.6)   Note that δ(z) is defined to be aliased into the range of [0, π] because the assignment of λ* and λ+	is unspecified.  2.2 PS-OCT system Fig. 2.1 shows the schematic of our JMT system [51]. The basic system has been developed by a pervious student and I have made improvements on the system. An MEMS-based swept-source (Axsun Technology Inc., MA) with a center wavelength of 1.06 µm, full width at half maximum (FWHM) of 111 nm, and scanning width of 123 nm is used as the light source. The scanning rate of the light source is 100 kHz, and the average output power is 30 mW. The interferometer is   12 built up with single-mode optical fibers. The light is split by a 90:10 coupler after passing through an isolator used for the protection of the source from back-reflected light. The 90 % port of the coupler is connected to a passive polarization delay unit, which is utilized to separate two polarization states in time domain (one polarization state of light is delayed more than the other polarization state of light). The 10 % port of the coupler is coupled to a reference arm. The light from the polarization delay unit then passes through an 80:20 fiber coupler. The 80% port is directed to a calibration reflector composed of a fiber collimator, a lens, and a mirror. The 20% port is connected to a sample arm which consists of a fiber collimator (F280 APC-C, Thorlabs Inc., NJ), a two-axis galvanometer scanner, and an objective lens (f= 60 mm). The optical power on the sample is around 2 mW.      13 Fig. 2.1 Schematic diagram of advanced JMT system [51]. LP: linear polarizer, PC: polarization controller, FC: fiber collimator, M: mirror, PBS: polarizing beam splitter, BS: beam splitter, H- and V-BPD: balanced photo-detector for horizontally and vertically polarized signals, respectively.   The backscattered light from the sample is recoupled back to the 80:20 coupler, and 80% of the back-scattered light is directed to a polarization diversity (PD) detection unit. The PD detection unit consists of a linear polarizer, a non-polarizing beam splitter (BS), two polarizing beam splitters (PBSs), and two 350 MHz balanced photo-detectors (BPDs, PDB430C, Thorlabs Inc.). The reference light is also directed to the PD detection unit, in which a linear polarizer is used to align the polarization state of the light to 45º angle. In the PD detection unit, the reference light and the backscattered light from the sample are combined at the BS, then split into horizontal and vertical polarization components by the two PBSs, and then detected by the BPDs.   OCT interference signals are generated on the photodetectors. The interference signals from the BPDs are sampled by a digitizer (ATS9350, AlazarTech Inc., Pointe Claire, QC, Canada) with 12-bit resolution and a sampling rate of 500 MHz, after passing through a high-pass (1.5 MHz) and a low-pass (250 MHz) filter (HP1CH3-0S and LP250Ch3-0S, R&K Co. Ltd., Shizuoka, Japan). The interference signal is sampled with 2560 sampling points for each A-line and the effective wavelength range being sampled is approximately 110 nm. The sampled interference signals are rescaled to the linear frequency domain using pre-defined rescaling parameters determined by a time-frequency calibration method [33]. The rescaling algorithm also cancels the spectral shift among the A-lines and stabilizes the phase of the OCT signal. After applying a Gaussian window, the interference signal is Fourier transformed to yield one line profile in Z   14 direction of the OCT image. The laser beam in the sample arm is scanned in X and Y directions by a two-axis galvanometer scanner to acquire 2-D or 3-D images. The specifications of the JMT system used in our study are listed in Table 2.1. As limited by tissue scattering, the actual measureable depth in cartilage tissue is around 1.5 mm.  Table 2.1: Specifications of the JMT system used in this study. Lateral resolution Axial resolution  Measurable depth-range Field of view 19.2 µm (in air) 8.1 µm (in air) 1.5 mm 3 mm × 3 mm   2.3 Principe of MPM  In light-tissue interaction, the induced polarization of material exposed to light can be expanded as [52]:  	𝑃 = 𝜒(*)𝐸 + 𝜒(+)𝐸+ + 𝜒(c)𝐸c + ⋯                                     (2.1)  where c(1) is the linear susceptibility, χ(n) is the nth order nonlinear susceptibility (for n>1) and 𝐸 is the electrical field of the incident light. Specifically, χ(1) leads to normal linear interactions such as absorption and reflection of light, χ(2) corresponds to SHG, and χ(3) contributes to TPEF and third-harmonic generation. MPM employs two or more photons to create multiphoton excitation of nonlinear signals such as TPEF and SHG.    15 TPEF and SHG have different fundamental origins compared with single-photon microscopy such as confocal microscopy [53, 54]. Fig. 2.1 shows the energy diagrams of TPEF and SHG [55]. In TPEF, two photons with a lower energy are absorbed simultaneously to excite a molecule from the ground state to the excited state. In comparison, a single higher energy photon is absorbed to excite the molecule in single-photon excitation. The emission photon carries slightly less energy than the sum of the two excitation photons due to energy loss in the transitions. TPEF is an incoherent process where the emitted photons radiate independently and isotropically. In the SHG process, instead of being absorbed, two excitation photons are simultaneously scattered and a SHG photon at exactly twice the energy of the excitation photon is created [40]. Compared to TPEF, SHG is a coherent process where the SHG photons experience constructive or destructive interference. Moreover, SHG requires non-centrosymmetric materials [52]. Due to the extremely low probability of TPEF and SHG in tissues, high flux of excitation photons is necessary in order to generate sufficient signals [56], which can be achieved by using a femtosecond laser and a tightly focused objective lens.  Fig. 2.2 Physical principle of MPM [53].    16 2.4 MPM system  Fig. 2.3 shows the schematic of the MPM system used in our study. The MPM system has been previously developed in our laboratory [104]. A Ti:sapphire laser (Fusion Pro 400, Femtolasers) is used as the light source in our MPM system. It has a center wavelength of 800 nm, a spectral bandwidth of 120 nm, and pulse width of ~10 fs. The beam from the laser source passes through a dispersion pre-compensation unit, which basically consists of two Brewster prisms. Afterwards the beam is guided to a two-axis galvanometer scanner (XY) (Cambridge Technology), a beam expander and then an upright commercial microscope (BX51W1, Olympus). The XY scanning of the laser beam is achieved by the two galvanometer mirrors.  The beam expander, consisted of two lenses, is used to couple the light to the microscope so that the light can fill the back aperture of the objective. The objective lens is mounted on a piezo scanner (MIPOS 500, Piezosystem Jena) for depth (Z) scanning. Two objective lenses are used for the MPM imaging, one is a water immersion 40× objective lens (LUMPlanFL N, Olympus) with a numerical aperture (NA) of 0.8, the other is a water immersion 60× objective lens (UPLSAP060XW, Olympus) with a NA of 1.2.    17  Fig. 2.3 Schematic diagram of the MPM system.  The TPEF and SHG signals are detected in backward direction. A dichroic mirror (FF670-SDi01, Semrock, dichroic mirror A in Fig. 2.3) is used to separate the TPEF and SHG signals from the residual excitation light. The SHG signal centered at 400 nm is reflected by a second dichroic mirror (450DCXRU, Chroma, dichroic mirror B in Fig. 2.3) and further filtered by a band pass filter (Semrock). The TPEF signal with longer wavelength (~500-600 nm) propagates through the dichroic mirror and is selected by a band pass filter (Semrock). Two photomultiplier tubes (PMT) (Hamamatsu) are used as the photon detectors. The specifications of the MPM dispersion	pre-compensation Ti:Sapphire	Laser stage galvanometer	scanner	(XY) objective	 lens	(Z) sample lens dichroic	mirror	A lens SHG	PMT TPEF	PMT filters dichroic	mirror	B   18 system used in this study are shown in Table 2.2. Limited by tissue scattering, the penetration depth of the MPM imaging in cartilage tissue is typical ~100-200 µm.   Table 2.2: Specifications of the MPM system (with the 60× objective) used in this study. Lateral resolution Axial resolution Penetration depth Field of view 0.4 µm 1.1 µm  100~200 µm  130 µm × 130 µm  As we can see, the PS-OCT and MPM systems have different resolutions, penetration depths, and field of views. Furthermore, PS-OCT measures the backscattered light and tissue birefringence, while MPM measures the TPEF and SHG contrasts. Thus, PS-OCT and MPM can provide complementary information about the cartilage tissue.    19 Chapter 3: PS-OCT phase retardation slope based segmentation  In this chapter, an automated 3-D segmentation method based on PS-OCT in detecting the structural zones of articular cartilage is presented and applied on swine articular cartilage tissue. The segmented layers are validated by high-resolution MPM imaging.  3.1 Introduction  As mentioned in Section 1.1, articular cartilage can be divided into four zones with different properties of collagen from the joint surface to the bone, which are the superficial zone, transitional zone, deep zone, and calcified zone [1]. Osteoarthritis is the most common form of joint disease and is associated with the progressive degeneration of the cartilage ECM. The ECM destruction and structural damage progressively develops from the superficial zone towards the deep zone as the disease progress. Visualization of articular cartilage structures would significantly facilitate cartilage disease diagnosis, repair, regeneration, and transplantation.   Clinical imaging modalities available today including routinely used radiography, MRI, and arthroscopy have been demonstrated to lack diagnostic accuracy because they either lack resolution or are unable to assess subsurface [57]. Microscopy has high resolution and can assess the microscopic structure of cartilage. Especially, SHG microscopy [58] is capable of imaging the orientation and organization of collagen fibers by nonlinear contrast [59]. However, microscopy is usually not suitable for in-vivo imaging as limited by its small field of view (FOV), shallow imaging depth and long imaging time. As a non-invasive imaging modality of clinical potential, OCT allows capturing cross-sectional images in real time and at high spatial resolution   20 (2-15 µm) over 1-2 millimeter penetration depth. OCT has been applied to characterize the difference between healthy and degenerative cartilage tissues [57, 60-63]. Quantitative information such as surface irregularity, tissue homogeneity, and signal penetration can be extracted from the OCT image. However, these studies only utilize the optical reflectivity (intensity) information, which does not directly represent the collagen organization in cartilage tissues. Thus, the detection of early stage of cartilage diseases remains as a challenge.  PS-OCT utilizes both the scattering and polarization properties of light [64], and extends the capability of standard OCT by allowing the characterization of depth-resolved tissue birefringence [51, 65-67]. Tissue birefringence occurs in fibrous structures such as collagen fibers and nerve fibers. PS-OCT has been applied to study birefringence properties in tissues such as tendon [28], muscle [29] and myocardium [30], skin [31], and retinal nerve fiber layer [32-35]. PS-OCT has also been used to access the collagen organization in articular cartilage [57, 68-73]. Drexler et al. [61] demonstrated that phase retardation was related to collagen organization and reduced birefringence was associated with collagen organization alterations in early cartilage degeneration. Similar findings were reported by several other groups [57, 72-73]. Xie et al. found that the tissue birefringence varied significantly depending on the direction [69] and topographical location [71] in articular cartilage imaging by PS-OCT. In their study, it was found that the severity of degenerative joint disease (DJD) was difficult to be distinguished until major matrix alterations at later disease stages, i.e., fibrous tissue formation, were observed [70]. Those studies were qualitative where no quantitative information has been extracted.   Nicolai et al. performed a quantitative analysis on the banding pattern of phase retardation   21 images, such as the width and height of the bands and total number of bands [74]. The banding pattern is caused by the accumulation of phase retardation and phase wrapping as it exceeds every π phase shift.  The band pattern approach is limited to tissues with strong birefringence where tissues with relatively weak birefringence won’t show the banding pattern. Other groups have detected the collagen fiber orientation by illuminating the sample with multiple incident angles and searching for the angle corresponding to the minimum birefringence [68, 75-76]. However, these methods did not consider the variation of collagen fiber orientation over the different zones along the depth. The slope of phase retardation from the PS-OCT images has been studied by several groups and a single slope from the linear region was extracted to represent the average tissue birefringence [77, 78]. This approach has been widely applied in dermatology [60, 79-83] and ophthalmology [33, 64, 84]. Shyu et al. [85] utilized the slope of phase retardation to access the birefringence of cartilage and compared normal and abnormal tissues. Their results showed only slight different birefringence in tissues at different disease stages. As the single slope approach only obtains the average tissue birefringence, it is not sensitive enough to show the gradual change of the collagen degeneration which progresses from the surface to the deeper zones in osteoarthritis.   Tissue birefringence depends on the properties of collagen matrix, such as fiber thickness, orientation, alignment and organization. The four zones of articular cartilage have different properties of collagen matrix. Thus the birefringence is expected to vary from zone to zone. In cartilage degeneration, the collagen matrix is altered progressively, starting from the superficial zone and gradually progressing to the deep zone [1, 86-87]. Therefore, it is necessary to quantify the tissue birefringence over different cartilage zones in order to better differentiate the layered   22 tissue structure and also to detect the progression of cartilage degeneration. A segmentation method will be very useful to differentiate and characterize the collagen properties in the different zones.   An intensity-based segmentation method in standard OCT was able to detect the cartilage surface and the boundary between cartilage and bone [88].  However, it was not able to differentiate the four zones. Duan et al. demonstrated a segmentation method based on PS-OCT phase retardation to distinguish the choroid and sclera layers in retina, using a two-segment linear fitting [89]. In articular cartilage, a theoretical model with and extended Jones matrix calculus (EJMC) was developed to consider a three-layered cartilage framework with superficial, transitional and deep zones [90]. To validate this model, a study with Monte Carlo simulation was carried out by another group, with comparable experiments [91], and their results showed that the three-layered cartilage framework could be extracted from the depth-resolved phase retardation imaging. However, the model assumed ideal conditions where the collagen fibers in the superficial zone were exactly parallel to tissue surface and all the fibers in deep zone were exactly perpendicular to the surface, which would not be practical in real situations.  In this Chapter, a slope based segmentation method in PS-OCT is developed to differentiate the different zones in articular cartilage. From the accumulated phase retardation images, we detect and distinguish the difference in birefringent features of the different zones using a multi-stage slope analysis. This method is capable to solve the depth (thickness) and the mean birefringence of each zone. As limited by the penetration of the PS-OCT, the top three zones (superficial, transitional, and deep zones) are analyzed. To visualize the variation of the collagen matrix in the   23 three zones, a series of SHG images are taken from the same sample by MPM [92]. Significant differences in collagen organizations are observed over the different zones from the SHG images, which match with the changes in the PS-OCT result. This method can have great potential in characterizing cartilage denegation by detecting and differentiating the different zones based on tissue birefringence.   3.2 Methods and materials   3.2.1 Sample preparation  All experiments involved in this chapter are performed on articular bone from the knee of adult swines obtained from a local farm. The articular bone is washed in distilled water and used for imaging within 24 hours. Cartilage tissue blocks are extracted from the anterior articular bone (shown in Fig. 3.1(a)), and sectioned into cubic shape with edge-length of around 2~3 mm, where the top of the cubic tissue is the intact natural surface. The sample is marked by 2 pins on the natural surface, to help locate a similar region when switching between the PS-OCT and MPM imaging.   3.2.2 Experiment scheme  The PS-OCT and MPM imaging systems used in this study have been described in Chapter 2.  The imaging procedure is described below.  PS-OCT imaging: In Fig. 3.1(b), the laser beam shines on the sample from the top view illumination. The incident angle θ is the angle between the laser beam and the normal of the   24 tissue surface. The sample is attached on a tilting mount so that the incident angle θ	can be adjusted in a user-defined azimuth plane (XZ plane). In normal illumination, θ is close to zero, where the beam is almost perpendicular to the tissue surface plane (XY plane). In tilted illumination, θ is rotated around the Y axis. The OCT scanning is performed by scanning the laser beam in X (fast scan) and Y (slow scan) directions, and 3-D volume can be acquired. Each OCT B-mode image shows a cross-section of the XZ plane.    Fig. 3.1 Experiment design. (a) Photograph of the sample with a blue rectangle indicating the region where the tissue cube was sectioned; (b) model of tissue with the scanning protocols of the two imaging methods indicated: the red arrow is the illumination direction for PS-OCT, the black dashed line is the normal direction to surface and the large rectangle represents a cross-sectional B-scan. The small rectangles in green are the approximate areas where SHG images are measured from the top illumination of the tissue.  MPM imaging: The MPM images are acquired with the laser beam shining on the sample from the top view illumination at normal direction. The laser beam is scanned in the X and Y Tissue	surface Illumination	beam Deeper	region (a) (b) x y z 𝜃 1 2 3 Bone Tide	mark   25 directions. Each MPM image shows a frame in the XY plane. An image stack is acquired by moving the objective lens in Z direction from the surface to a few hundred microns into the tissue. In order to reach the deep zone, a top layer of ~0.5 mm of the tissue is removed and another image stake is acquired.    3.2.3 PS-OCT phase retardation multi-stage slope analysis   In PS-OCT imaging, to analyze the birefringent property of cartilage tissue, the slope of the accumulated phase retardation can be utilized as a quantitative criterion [79-85].  To detect the birefringent property change along depth, a multi-stage slope analysis is developed.   With our PS-OCT system, co-registered images from the intensity of the backscattered light and the phase retardation can be obtained simultaneously. Fig. 3.2 shows an example of PS-OCT images, where Figs. 3.2(a) and 3.2(b) are the co-registered intensity and phase retardation images, respectively. The algorithm is described below, which is implemented on the example shown in Fig. 3.2.  Pre-process: The tissue surface can be identified clearly in the intensity image because there is strong backscattering from the tissue surface, as shown in Fig. 3.2(a). The tissue surface is segmented by a custom-built surface extraction algorithm based on the intensity image. The surface segmentation is marked as “S” in the co-registered phase retardation image, as shown in Fig. 3.2(b). With the surface segmentation, each A-line in the phase retardation image is shifted so that the surface is flattened and the surface of the tissue is set as the origin of that A-line. Each A-line is then modified by averaging it with its surrounding A-lines within a window of user-  26 defined size (typically 11×11).  Fig. 3.2(c) shows a representative modified A-line, where the curve shows a feature of multiple linear regions along the depth, a relatively flat slope region <200 µm, a slow slope region between 200-400 µm, and a fast slop region >400 µm. Thus a multi-stage linear fitting should suit the feature of the A-line.   Reliability thresholding: The signal-to-noise ratio (SNR) of the images drop significantly at deep depth due to light attenuation. To ensure reliable calculation in our algorithm, a thresholding is performed to remove the pixels below certain SNR. The pixels with the SNR< 13 dB in the intensity image are removed. The pixels with the effective SNR (ESNR) < 10 dB in the phase retardation image are also removed. ESNR is specific for phase retardation images measured by JMT [93], which balances the SNR in the 4 channels acquired by the two input and two output polarization states (2×2=4).    Slope-based segmentation: The depth locations and the slopes of the different linear regions are extracted using the following algorithm. It is implemented on the A-line shown in Fig. 3.2(c) as an example.  1) The tissue surface is considered as the origin of the A-line, labeled as S.  The end point is determined by the above thresholding, labeled as T. From the depth of T, we determine whether the penetration depth has reached the deep zone or the calcified zone, according to a prior knowledge of articular cartilage. Thus, three regions or four regions will be selected for the multi-stage linear fitting.     27 2) For the case of three regions (as shown in Fig. 3.2 (c)), the boundary B1 between the superficial zone and transitional zone, and the boundary B2 between the transitional zone and deep zone, need to be determined.   3) The first flat region B1<200 µm corresponds to the superficial zone that has low birefringence. The phase retardation in the first a few pixels of the A-line fluctuates dramatically due to surface effect, which should be excluded in the linear fitting of the first region, as labeled by B0. The A-line between B0 and B1 is fitted by a linear least squares (LLS) regression. The exact positions of B0 and B1 are determined by a traversing search algorithm, with the criterion to minimize the mean squared error (MSE) between the A-line and the linear fit, and the slope of the linear fit simultaneously.   4) On the A-line, three regions are defined by B0-B1 (flat slope, superficial zone), B1-B2 (slow slope, transitional zone), and B2-T (fast slope, deep zone). LLS regression is conducted on the three regions, respectively. B2 is searched between B1 and T, and B1 is fine-tuned around the positon found in the above step.  The exact positions of B1 and B2 are determined by a traversing search algorithm, with the criterion to minimize to the total absolute error, equivalent to the least absolute deviation (LAD), over the three fitted regions.  Quantification of birefringence: With the algorithm mentioned above, three or four regions (zones) with different phase retardation slopes can be distinguished: R-1, R-2, R-3, and R-4 (optionally). These regions are defined by any two consecutive boundaries solved by the segmentation method. The depth position and the thickness of each region can be obtained. In   28 birefringent tissue, the phase retardation accumulation can be calculated by δ z = 2π∆𝑛 ∙ z/𝜆j, where z is the thickness of the tissue, ∆𝑛 = 𝑛k − 𝑛,  is the difference between the refractive index along the fast and slow axis, and  𝜆j is the central wavelength. From the linear fitting, the slope of the phase retardation can be extracted. The slope is related to the birefringence parameter Δn as Slope = p qr = +s∆)tu  . Thus the slope is directly proportional to the birefringence parameter.    3.3 Results  3.3.1 PS-OCT analysis A cartilage tissue with near normal illumination on the tissue surface is imaged by PS-OCT.  The multi-stage slope analysis results are rendered in Fig. 3.2. The intensity OCT and phase retardation B-scans are shown in Fig. 3.2 (a) and (b) respectively. The segmentation results are rendered in 2-D and 3-D respectively in Figs. 3.2 (b) and (d). A representative phase retardation A-line (marked by a red arrow in Fig. 3.2(b)) is shown in Fig. 3.2(c) with the multi-stage fitting results as the red dashed segments. The boundaries S (surface) and T (reliable phase ending point) and the turning points B0, B1, and B2 are marked, with a blue dashed line indicating the thresholding depth.   The thickness and slope of the phase retardation of each region are shown in Table 3.1. The thickness is converted from optical pathlength by assuming n = 1.444 in articular cartilage. The slope increases from R-1 to R-3. With near normal illumination, the slope of R-3 is relatively   29 small because the fibers are nearly parallel to the illumination direction. Later, we will show that the slope can be much larger when the illumination angle is tilted. The slope in region R-1 has a larger standard variance than its mean value. This indicates that the phase in that region fluctuates dramatically so that it cannot offer enough accumulation. The thickness of the regions will guide the SHG measurement in Section 1.3.2.   Fig. 3.2 Segmentation results from Multi-stage slope-based analysis. (a-b): PS-OCT images measured from the top-view, (a) is intensity OCT, (b) is accumulated phase retardation with the segmentation results indicated as black dashed lines; (c): Modified accumulated phase retardation A-line from the three positions marked by red arrows in (b); (d) 3-D rendering of the segmentation results.    S BT 0 π (d) S B1 B2 T (a) (b) (c) S B1 B2 T B0 2 1 3   30 Table 3.1 Segmentation results summary Region Thickness (µm) Phase retardation slope (degree/µm) Corresponding Structural zone R-1 125.67 ± 5.78 0.009±0.052 Superficial zone R-2 396.48 ±64.21 0.046±0.021 Transitional zone R-3 873.27 ± 200.62 0.157±0.045 Deep zone   3.3.2 SHG imaging with the guidance of PS-OCT analysis  To investigate the microscopic structures of the collagen fibers in each of the regions determined by PS-OCT, SHG imaging is carried out on the same cartilage tissue in accordance with the segmented depth information. One stack of SHG images are taken from the natural tissue surface into ~200 µm deep. A second stack of SHG images are taken after a ~0.5 mm of the top layer of the tissue is removed. Fig. 3.3 shows the SHG images from approximately the three regions (marked by numbers 1, 2, 3 in Fig. 3.2(b)). The first row is for depth range 10~40 µm, near the tissue surface in region R-1; The second row is for depth range 80~140 µm, in the transition from R-1 to R-2; The third row is for depth range around 600 µm, inside region R-3.  The field of view of each image is ~130 µm × 130 µm.  Fig. 3.3 (a)–(d) show the collagen fibers in region R-1. Large fiber bundles are clearly distinguished and the fibers are almost parallel to the tissue surface. At different depths, the orientation of the fibers changes from frame to frame. Those features are consistent with the   31 collagen properties in the superficial zone. As the fibers change orientation at different depths, the phase retardation cannot accumulate and the birefringence in this layer would be low.  This is consistent with the small slope in region R-1 in the PS-OCT analysis.   Fig. 3.3 (i)-(l) show the collagen fibers in region R-3. The fibrils have very small diameter and they are densely packed together. The dark circular holes correspond to chondrocytes. Those features are consistent with the collagen properties in the deep zone, where the type II collagen forms fibrils and the collagen density is high.  The orientation of the fibrils is hard to distinguish as limited by the resolution of the MPM system.   Fig. 3.3 (e)-(h) show the collagen fibers in the transition area from R-1 to R-2. The fibers are much smaller than that in the superficial zone, and show some oblique orientation. The collagen density seems to be lower than that in the region R-3.  Those features are consistent with the properties of the transitional zone.    32  Fig. 3.3 SHG images measured with the guidance of PS-OCT analysis. (a-d): SHG images measured from the top of tissue with around 10~40 µm, from the surface, roughly at the position ‘1’ marked in Fig. 3.2 (b) (from natural surface), frame gap: 10 µm; (e-h): SHG images measured at the depth of around 80 ~ 140 µm off the surface, indicated by the position ‘2’ in Fig. 3.2 (b), gap: 20 µm; (i-l): SHG images measured the depth of around 600±100 µm off the surface, corresponding to the position ‘3’ in Fig. 3.2 (b), frame gap: 20 µm. Image size: 130 * 130 µm.  Scale bar is 30 µm.  Thus, with the high-resolution SHG imaging, the segmentations from the PS-OCT analysis approximately match with the superficial, transitional, and deep zones of articular cartilage, respectively.  (i) (j) (k) (l) 30	µm	 (a) (b) (c) (d) (e) (f) (g) (h)   33 3.3.3 PS-OCT multiple illumination angle results  PS-OCT with multiple incident angles of illumination has been used to evaluate the collagen fiber orientation [25-26, 32-33]. Here, a multiple illumination angle study is performed to further verify the multi-stage slope-based segmentation in PS-OCT.  Fig. 3.4 shows the PS-OCT B-scan images at illumination angles of −60° , 	−45° , −30° , 	−15° , 0° , 15° , 30° , 45° , and 60° . The birefringence is found to be higher (more banding structure) at larger illumination angles.    Fig. 3.4 PS-OCT measurement results at different illumination angles. (a-i) PS-OCT B-scan images measured at illumination angles as indicated. The illumination angle varies in X-Z plane. In each B-scan image penal shown, the accumulated phase retardation image is on the top, and the Intensity OCT image is at the button.  (a) (b) (c) (d) (e) (f) (g) (h) (i) High Low 500 𝝁m π 0 -60° -30° 60° 30° 0° -45° -15° 45° 15°   34   Fig. 3.5 PS-OCT multi-stage slope analysis results at different illumination angles; (a) averaged phase retardation A-line measured at 9 selected illumination angles; (b) Segmented boundary depths at all the different illumination angles corresponding to the images shown in Fig. 3.4; (c) Phase retardation linear regression slopes measured at different regions and different illumination angles.  The quantification results are shown in Fig. 3.5. Fig. 3.5 (a) shows the averaged A-lines over 5 regions of interest (ROI); each ROI is a window of 42 B-scans by 84 A-lines (0.5 mm by 0.5 mm). Noticeable difference is observed among the A-lines with different illumination angles. In (a) (b) (c)   35 the deep zone, the slope of the A-line increases monotonically as the illumination angle increases from 0º, ±30º, and ±60º. In the transitional zone, the slope does not change as much as in the deep zone.   The segmentation results are shown in Fig. 3.5(b) and Fig. 3.5(c).  The segmented layer depth and the slope of each layer are plotted as a function of the illumination angle. The tissue surface is set as zero depth. The green, red, blue, and black curves represent the segmentation boundaries B0, B1, B2 and T, respectively. Their average locations are 19.75±3.06 µm, 100.53±21.28 µm, 465.67±15.55 µm, and 1052.90±24.39 µm, respectively, off the surface. The layer depths are consistent with a prior knowledge of the typical depths of cartilage zones.   The mean slopes of the different regions are shown in Fig. 3.5(c). The R-3 region has larger slope, meaning higher birefringence, than the R-1 and R-2 regions. It is consistent with the anatomical fact that the collagen fibers in the deep zone are densely packed and highly aligned. The slope of R-3 region shows high sensitivity to the illumination angle. The minimum slope occurs at the illumination angle of 0º and 15º with the slope around 0.15 degree/µm. The maximum slope appears at the ±60º with the slope of around 0.4 degree/µm. The difference between the maximum and minimum slopes is more than 2 times. In articular cartilage, the collagen fiber in the deep zone is aligned in perpendicular direction to the tissue surface. When the illumination light is in parallel with the fiber orientation, no birefringence can be observed. When the fiber orientation is in perpendicular with the illumination light, birefringence is the highest. Therefore, a large slope is expected for the larger illumination angles. Nevertheless, as the collagen fiber is not exactly parallel with the illumination light even at near normal   36 illumination, some birefringence is still observed in our experiments. The slopes in the R-1 and R2 regions are not sensitive to the illumination angles. This is because the collagen fibers in the superficial zone change orientation at different depths, and the collagen fibers in the transitional zone are oblique to the surface.    In this study, the measured depths of the regions are not sensitive to the illumination angle.  This shows that the segmentation algorithm can reliably detect the zone thickness even when the illumination angle varies. The sensitivity of the slope in R-3 with illumination angle shows that this approach can also detect the orientation of the collagen fiber alignment with high sensitivity.   3.4 Summary A segmentation method in PS-OCT has been developed in this chapter. It analyzes the different slopes from different depth regions in the phase retardation image. The phase retardation image can indicate the tissue birefringence which is highly related to the collagen fiber organization and orientation. The segmentation algorithm is utilized to distinguish tissue layers based on the birefringence property. The different layers identified by the PS-OCT are further examined with the high resolution SHG imaging, which shows the collagen fiber organization at the microcopy level. Using both PS-OCT and SHG to study articular cartilage can bridge the relationship between the global and local collagen properties. A multiple angle illumination experiment is conducted to demonstrate the directional sensitivity of the segmentation approach with PS-OCT on the collagen fibers at different zones.  It is shown that PS-OCT can detect the variations in tissue birefringence with high sensitivity. Thus PS-OCT is a promising tool for studying collagen fiber organizations in the articular cartilage.    37  We have investigated the variation of tissue birefringence in the different zones of articular cartilage by PS-OCT. Various physical and biological factors can affect the tissue birefringence measurement, including the incident angle of illumination, the sample dimension, orientation, topography, and localization within the joint, and also the healthy state [68-71]. Those factors may combine together and affect the PS-OCT measurement. More investigations are needed to study the normal and diseased articular cartilage by PS-OCT.   To further explore the relationship between the macroscopic features obtained from PS-OCT and the microscopic characteristics achieved by MPM, a comparison work between those two modalities in imaging articular cartilage will be carried out in Chapter 4, which will show more detailed quantifications of the depth-dependent features in articular cartilage. With more knowledge about the association between the features achieved by PS-OCT and by MPM, it may be possible to use PS-OCT independently to access the depth-dependent collagen fiber organizations in articular cartilage, which is of significant potential.   38 Chapter 4: Comparison between PS-OCT and MPM in articular cartilage imaging  A comparison study between PS-OCT and MPM imaging is conducted on articular cartilage sample. The depth-dependent features extracted by PS-OCT and by MPM are compared and a good association is obtained.   4.1 Introduction As discussed in Section 1.1, articular cartilage has four distinct structural zones with different ECM properties. The cartilage disease, e.g. osteoarthritis, is also related to the depth-dependent ECM degeneration. PS-OCT [68] and MPM [86] are two promising methods for studying the articular cartilage structure and its diseases associated with the change of collagen organization. PS-OCT has been utilized to study the articular cartilage, showing great potential to detect the major collagen fiber orientation [68, 75-76], and to differentiate the diseased articular cartilage [57, 61, 70-74].  However, those studies did not investigate the depth-dependent features along the cartilage depth. Moreover, the resolution of PS-OCT is not high enough to resolve individual collagen fibers in the ECM.   MPM can assess the microscopic structure of cartilage with a high resolution around hundreds of nanometers [58]. MPM can detect TPEF and SHG simultaneously. Especially SHG microscopy can assess the orientation and organization of individual collagen fibers [97-99]. Several detailed features of collagen fiber organization have been quantified from the SHG images of different   39 tissues, such as skin and cornea [100-106].  PS-OCT and SHG microscopy are two promising methods for studying the articular cartilage structure related to collagen organization changes. They can provide complementary information about tissues due to their respective features. PS-OCT detects the tissue birefringence on the millimeter scale while SHG resolves individual fibers and detects the collagen fiber orientation on the micrometer scale. The comparison between PS-OCT and SHG imaging should be able to provide more comprehensive information about the tissues and provide a better understanding of the birefringence properties detected by PS-OCT. A previous study of the correlation between PS-OCT and a SHG microscopy on collagen has been conducted on skin samples [81]. The phase retardation measured by PS-OCT and the fiber organization measured by SHG imaging were quantitatively compared. The results showed that the skin with more random collagen fiber organization from the SHG image had a smaller slope of the accumulated phase retardation in PS-OCT. However, the collagen fibers in skin tissue are not well aligned, usually irregular and randomly oriented [31], which results in relatively low birefringence in skin. In contrast, articular cartilage contains abundant structural collagen fibers with highly organized orientations, and its multi-zoned architecture shows different collagen fiber organizations in the different zones [1, 86]. Thus, articular cartilage is especially suitable for using PS-OCT and SHG to assess its birefringence properties and the relationship with microscopic structures.  In this chapter, we explore the combined utilization of PS-OCT and SHG to investigate the ECM structural variation along the cartilage depth.  A comparison work is conducted by using PS-OCT and SHG to image a same articular cartilage tissue. Quantitative information such as the   40 slope of the phase retardation, SHG intensity, and collagen fiber orientation angle are extracted and characterized in the cartilage tissue over the different zones. The structural zones of the cartilage sample are determined by using the phase retardation slope based segmentation method discussed in Chapter 3. The slope of the phase retardation image measured from the side-view illumination of the sample is quantified to show the depth-dependent birefringent property variation over different structural zones of articular cartilage. The microscopic features of the different zones are characterized by SHG imaging of the same cartilage tissue. Properties such as the intensity of SHG signal, the collagen fiber orientation and randomness are investigated. By comparing the PS-OCT and SHG results in both qualitative and quantitative analysis, we demonstrate the relationship between the microscopic characteristics from SHG imaging and the tissue birefringence features from PS-OCT imaging, and explore how the different features along the depth match with the structural zones of articular cartilage.  4.2 Methods and materials   4.2.1 Experiment design   The sample preparation used in this study has been described in Section 3.2.1. The PS-OCT and MPM imaging systems used in this study have been described in Chapter 2.  The imaging procedure is described below.  The experiment design is shown in Fig. 4.1. Fig. 4.1(a) shows the photograph of the sample. A tissue cube is extracted from the region marked by the blue rectangle and then imaged by PS-OCT and MPM. Fig. 4.1(b) is an illustration of the different zones in articular cartilage. Fig.   41 4.1(c) and Fig. 4.1(d) indicate the scanning protocols of PS-OCT and MPM imaging, respectively. The direction of the cartilage depth is defined as the Z axis and the coordinates are marked in Fig. 4.1(d).  The collagen organization and the chondrocyte morphology vary from the tissue surface to deep region (along Z). To explore the various tissue properties along the depth, the imaging procedures and related scanning protocols are designed as follows. PS-OCT imaging: The sample orientation is adjusted so that the PS-OCT imaging can be acquired from two illumination directions as shown in Fig. 4.1(c). In the top-view illumination, the laser beam illuminates the specimen from the natural tissue surface along the Z direction. Fast scan is along X and slow scan along Y. Each B-scan shows the cross-section of an XZ plane, which covers from the natural tissue surface to the deep zone. However, due to light attenuation in the tissue, the imaging depth may not cover the full depth of the deep regions such as the deep zone and calcified zone. In the side-view illumination, the laser beam illuminates from the side-cut surface of the tissue block along the Y direction. Fast scan is along X and slow scan along Z. Each B-scan shows the cross-section of an XY plane. Multiple B-scans are acquired along the Z direction. With the side-view illumination, the different zones can be imaged separately and a wider range in the Z direction can be reached than the top-view illumination. In both illuminations, the incident laser beam is at normal direction that is approximately perpendicular to the respective surface.     42  Fig. 4.1 Experiment design. (a) Photograph of the sample with a blue rectangle indicating the region where the tissue cube was sectioned; (b) Morphological diagram of the depth-dependent ECM changing in cartilage tissue; (c) Tissue cube model indicating the scanning protocol in PS-OCT imaging: red arrows are the illumination beams, the gray rectangle represents one of the B-scans from the cross-section view (in X-Z plane), and the blue one represents one of the B-scans from the en-face view (in X-Y plane). (d) Tissue cube model indicating the scanning protocol in MPM imaging: each green square (in X-Z plane) represents a region of interest (ROI) from the tissue surface to deep region. The coordinate system is marked on the upper right corner of the tissue cube model in (d).  (a) (b) Tissue	surface Illumination	beam Deeper	region Tissue	surface Deeper	region MPM	ROI (c) (d) Tissue	surface Deeper	region x y z   43  MPM imaging: MPM images can also be acquired with the top-view or side-view illumination. In top-view illumination, each frame is in XY plane and a stack of frames can be acquired by moving the objective lens in Z direction. The imaging depth is limited to ~200 µm. To reach a deeper depth, a layer of tissue can be physically removed to exposure the deeper region. In side-view illumination, each frame is an XZ plane and an image stack can be acquired along the Y direction. After acquiring a stack at one location, the sample is moved in the Z direction to shift to a deeper Z location by a translation stage. Multiple stacks are acquired at different Z locations from its natural tissue surface to deep regions. With the side-view illumination, the MPM imaging can cover a wider range in the Z direction than the top-view illumination. Furthermore, the intensities of the SHG images at different Z locations can be compared at the similar light penetration depth.  4.2.2 PS-OCT slope-based analysis    In PS-OCT imaging, the slope of the accumulated phase retardation can be utilized as a quantitative criterion for the birefringent property of cartilage tissue [85]. In side-view illumination, each B-scan (XY plane) is in one specific zone of the cartilage tissue. A single slope of the accumulated phase retardation is calculated to indicate the birefringence of that layer.  In top-view illumination, each B-scan (XZ plane) covers multiple zones of the cartilage tissue. A multi-stage slope analysis is applied, where the slopes at different linear regions can be distinguished along the phase retardation A-line [107]. Details about the segmentation method have been demonstrated in Chapter 3. Due to the different collagen organizations in each zone, different zones will show different birefringence properties, so as to form different slope regions   44 along the depth. With the slope-based segmentation method, the depth-position for each zone can be distinguished, as well as the corresponding slope. The segmentation results can be regarded as a guidance to differentiate the structural zones in articular cartilage tissue.  4.2.3 Quantitative analysis methods for MPM images    Articular cartilage contains abundant collagen fibers that provide relatively strong SHG signal. To quantitatively analyze the SHG images, four steps are taken: 1) Image division; 2) Differentiate the collagen fibers and chondrocytes by thresholding; 3) Calculate the angle distribution of the collagen fiber orientation by Hough transform [102, 103]; 4) Quantification of the randomness of the fiber organizations. The implementation details are shown in Fig. 4.2 as an example.  4.2.3.1 Image division Within a SHG image (e.g. Fig. 4.2(a)), the signal brightness and the orientation of collagen fiber can vary at different regions. To achieve more specific quantification, the SHG image is divided into tiles (sub-images), where each sub-image will be more likely to consist of collagen fibers with similar features [102, 103]. The number of tiles is determined based on the different image contents, varying from 2 × 2 to 16 × 16. In the example in Fig. 4.2, a division of 4	× 4 is used. All the following analysis procedures are carried out on each sub-image. There are 16 sub-images and each has 128×128 pixels (~33×33 µm2).  The 16 sub-images are marked as I**~I~~ from top to button and left to right as shown in Fig. 4.2(a).    45 4.2.3.2 Signal thresholding and quantification    In articular cartilage, the SHG contrast comes from collagen fiber. Chondrocytes lack SHG contrast and are shown as round shaped dark regions in the SHG image. Chondrocytes can also be visualized by the TPEF contrast due to auto-fluorescence. To eliminate the effect from noise and errors from weak signal, a thresholding procedure is applied on the raw SHG images to only select the regions with sufficient signal for quantitative analysis. In each frame, all the pixels with signal intensity above 30% of the maximum intensity of that frame are considered as the effective region. Isolated pixels that do not belong to a group of 20 interconnected pixels are also eliminated. This procedure results in elimination of most of the background noise and the blurred collagen signal with a low SNR. An average SHG intensity is computed from the pixels in the effective region. The dark regions, where the SHG intensity is lower than the threshold, can be roughly regarded as the region of chondrocytes. Chondrocyte density can be calculated by the ratio of the pixel size of the dark regions versus the full image size.  4.2.3.3 Angle distribution and local orientation    To solve the collagen organization in SHG images, there are generally two quantification approaches. One is the Fourier-transform-based approach [100] that characterizes the global fiber organization by the autocorrelation of the intensity fluctuations in an image from a statistical perspective. The other is the Hough-transform-based approach [102] that accesses the local fiber orientation and obtains the angular distribution of the collagen fiber directions from the interconnections of the pixels in an image. In this study, we implement the algorithm based on Hough transform [102]. Hough transform is an algorithm that assesses simple line shape in a 2-D image. The implement procedure is explained with the example shown in Fig. 4.2.   46  1. Pre-process. De-noising and contrast enhancement [102] are applied on each sub-image. Fig. 4.2 (b) shows the pre-processed sub-image I+~. The signal of the fiber-like features becomes more significant compared to the background.  2. Hough Transform. In an XY plane, each straight line can be described by a pair of parameters (ρ, θ), where ρ is the distance from the origin to the closest point on the straight line, and θ is the angle between the X axis and the line connecting the origin with that closest point. Hough transform coverts the image from the XY plane to the (ρ, θ) plane. The ρ-θ mapping of Fig. 4.2(b) is shown in Fig. 4.2(c).  Two regions with high counts are marked by the circles. The count value in the ρ-θ mapping indicates how many times a specific line, represented by (ρ, θ), occur in the image.   3. From the ρ-θ mapping, the angle distribution p(θ) of the fiber orientations can be obtained.  The fiber orientation angle θ is referenced to the vertical direction of the image, where positive angle means clock wise and negative angle means counter clock wise from the reference direction. Fig. 4.2(d) shows the angle distribution obtained from Fig. 4.2(c). Two peaks are observed at 42.5° and −74.5°, indicating that the fibers are mostly align in those two orientations.   47  Fig. 4.2 Fiber orientation measurement by Hough transform with intermediate results of the procedures. (a): SHG images measured from the side-view, with 16 sub-images divided by the red lines; (b): pre-process results of Sub-image 𝐈𝟐𝟒, marked as yellow shadow in (a); (c): Hough transform results of (b); (d): Angle distribution achieved form the Hough transform results.  Fig. 4.3 shows the quantification of the collagen fiber orientation from the side-view SHG images measured at three depth locations, region-1, region-2, and region-3, roughly corresponding to the superficial zone, transitional zone, and deep zone, respectively. The side-view SHG images (XZ plane) are shown in Fig. 4.3(a-c) for regions 1, 2 and 3, respectively.  (b) (c) (d) (a)   48  Fig. 4.3 Collagen fiber organization solved from SHG images. (a-c) SHG images from side-view (XZ plane), at the depths of around 20 µm, 100 µm, and 600 µm from the tissue surface, respectively. Image size: 130 * 130 µm. (d) the angle distributions for the collagen fibers in the (a)-(c). The coordinate system is marked in (c).   In Fig. 4.3(a), the dominating fiber orientation in region-1 is almost horizontal, parallel to the tissue surface, which is consistent with the characteristics in superficial zone. The dominating fiber orientation obtained by Hough transform is marked by a blue arrow, and the reference vertical direction (Z direction) is marked by a red dashed arrow. Fig. 4.3(c) shows that the fibers in region-3 are highly aligned, almost perpendicular to the tissue surface, which corresponds to the deep zone. Fig. 4.3(b) shows fibers in oblique direction, showing transition from the horizontal to vertical orientation, which corresponds to the transitional zone.   (a) (b) (c) (d) x z Angle (respect to z axis) Probability distribution   49 The distribution of the fiber orientation angle obtained from the Hough transform is shown in Fig. 4.3 (d). The superficial zone has a sharp peak at around -80º, meaning that the fibers are almost horizontal to the surface. The transitional zone shows two bumps around ±45º, meaning that the fibers are in oblique direction. The deep zone shows a peak around 40º, meaning the fibers are highly aligned. As the fibrils are very small in the transitional and deep zones, the accuracy of the fiber orientation analysis is limited by the resolution of the MPM system and some errors may occur.   4.2.3.4 Orientation Index and statistical entropy    From the angle distribution of the collagen fiber ordinations in SHG image, the orientation index (OI) can be calculated using the following equation [26]:                                             OI θ‚ = 2 ƒ „ …:†‡ „E„ˆ‰Š ‹„ŒuŒu ƒ „ ‹„ŒuŒu − 1                                      (4.1)  where p θ  is the angle distribution. 莏  is the dominating angle, which indicates the angle with the maximum probability in the angle distribution. When OI equals to 1, it represents that all the fibers align in exactly the same direction. When OI equals to 0, it corresponds to a situation of total random distribution of fiber orientation.    The statistical entropy of the angle distribution can also be obtained as     50  ε = −∑p(θ){log	[p(θ)]}.                                                    (4.2)  where p(θ) is the probability of the angle distribution.  4.3 Results  4.3.1 PS-OCT side-view measurement results Side-view illumination PS-OCT imaging is performed to investigate the tissue birefringence properties at different depths, with the results rendered in Fig. 4.4. Figs. 4.4(a) - 4.4(i) show the B-scans (XY plane) of articular cartilage at different Z depths, where the intensity OCT image is shown at the top, and the corresponding accumulated phase retardation image is shown at the bottom. From the phase retardation images, we can see the birefringence varies at different depths. More banding structure represents higher birefringence.   In Fig. 4.4(a), the birefringence is strong near the surface; In Fig. 4.4 (c), it gradually drops to a lower value; In Fig. 4.4 (e-f), the birefringence increases back to a high level; In Fig. 4.4 (h-i), it drops again and disappears at the deep regions.     51   Fig. 4.4 (a-i) PS-OCT images measured from the side-view of the cartilage tissue at different depths. The top ones represent the intensity B-scans, while the bottom ones are the accumulated phase retardation B-scans. The depth is labeled in each image. All the B-scans shown are in XY plane. The averaging region to calculate the slope is marked in (a). Scale bar is 500 µm.  A single-stage slope analysis is conducted on each selected B-scan shown in Fig. 4.4. The slope is calculated from an averaged A-line, from the transitional 200 A-lines in the B-scan image. The averaging region is marked as a yellow rectangle in Fig. 4.4 (a). Table 4.1 summarizes the depth locations of each B-scan and the corresponding slope of the accumulated phase retardation. In Table 4.1, the quantitative slope values are consistent with the birefringence characteristics observed from the images in Fig. 4.4. Two regions with fast slope, indicating high birefringence, are observed. One region is at the depth around 35 µm from the surface within the superficial zone. The other region is near the depth of 670 µm, corresponding to the deep zone. The strong (a) (b) (c) (d) (e) (f) (g) (h) (i) High Low π 0 35𝝁𝒎 	 155	𝝁𝒎 	 280	𝝁𝒎 	 435	𝝁𝒎 	 555	𝝁𝒎 	 670	𝝁𝒎 	 825	𝝁𝒎 	 1000	𝝁𝒎 	 1210	𝝁𝒎 	 X Y   52 birefringence observed in the deep zone is consistent with the results in Chapter 3, because the fibers in the deep zone are well aligned and almost perpendicular to the surface. In Chapter 3, the superficial zone shows relatively low birefringence in the top-view illumination. However, in the side-view illumination, relatively high birefringence is observed in the superficial zone. This is because in the side-view illumination direction, the fibers mostly fall inside the same plane, which generates strong birefringence. This also shows that tissue birefringence measurement is sensitive to the light illumination direction.  Table 4.1 The slope of the selected phase retardation B-scans by single-stage slope estimation   The multiple B-scans can be reconstructed into a 3D volume and other image views can also be obtained. Fig. 4.5 (a) shows the YZ view and Fig. 4.5 (b) shows the XZ view of the volume. In both Figs. 4.5 (a) and 4.5 (b), the natural tissue surface of the articular cartilage can be identified and is labeled as “Surface”. This region has very tilted illumination due to the curved shape of the sample surface. Another boundary around Z=1210 µm can also be identified which is likely to be the boundary between calcified zone and subchondral bone and is labeled as “Bone boundary”. The region below this boundary shows relatively uneven surface (in Fig. 4.5 (a)), and lower signal intensity (in Fig. 4.5 (b)). The uneven surface is likely caused by the loose bone Figure No. (a) (b) (c) (d) (e) (f) (g) (h) (i) Depth(µm) 35 155 280 435 555 670 825 1000 1210 Slope (rad/µm) 0.0282 0.0170 0.0088 0.0181 0.0257 0.0298 0.0171 0.0179 0.0112   53 structure and the low intensity may be caused by the lower collagen concentration and the existence of light attenuating blood in the bone region.    Fig. 4.5 Summing projections of the resliced image stack shown in Fig. 4.4. (a) Summing projection in YZ plane; (b) Summing projection in XZ plane. Surface and Bone boundary are indicated by red arrows. Sample size is ~3 mm (X) by 2 mm (Z). Scale bar is 500 µm.  4.3.2 MPM side-view measurement results Fig. 4.6 shows a selected panel of MPM images acquired at different depths with Z ranging from 0 to 1.4 mm, and at approximately 30 µm below the illumination surface. In Fig. 4.6 (a), large collagen fiber bundles are observed and the fiber bundles are aligned mostly parallel to the natural tissue surface, which is consistent with the collagen fiber alignment in the superficial zone. In this region, the chondrocytes are also aligned mostly parallel to the natural tissue surface, and in consistent with the fiber bundle orientation. From Fig. 4.6 (b)-(d), the collagen fiber orientation gradually changes from parallel to vertical to the natural tissue surface. The chondrocytes also show an oblique orientation. In the deep regions, the collagen fiber has much smaller diameter than in the superficial zone. This is because the deep region mainly contains type II collagen which usually forms small fibrils. The superficial zone contains type I collagen Surface Bone	boundary Surface Bone	boundary (a) (b) X Z Y Z   54 which forms large fiber bundles. For the small fibrils, the Hough transform based approach can still identify the fiber orientation.  In Fig. 4.6 (e)-(h), the collagen fibrils are almost vertical to the natural surface. In this region, the chondrocytes are elongated and aligned vertically to the tissue surface. In Figs. 4.6 (i)-(l), the collagen fiber becomes vague and hard to be distinguished and the chondrocytes have more round shape. In Figs. 4.6 (k)-(l), the SHG intensity significantly drops. The SHG intensity also varies from the different zones. The SHG intensity is relatively low at the superficial zone. It gradually increases from the transitional zone to the deep zone until it reaches the depth at around 1.1 mm. Deep zone seems to have the highest SHG intensity. Afterwards, SHG intensity drops rapidly in the even deeper regions. SHG intensity is related to the density and diameter of collagen fibrils. The observed SHG intensity variation is in consistent with other report that the deep zone contains fibrils with larger diameter than the transitional zone [1]. TPEF signal (red color) can be observed from chondrocytes. The TPEF signal monotonically decreases from the superficial zone to the deeper regions. The intensity of TPEF signal could be related to the status of the chondrocytes. Near the surface the chondrocytes are more active than in the deeper region. In the clarified zone, empty lacunae are observed which contain mostly dead chondrocytes [94].  	  55 	Fig. 4.6 MPM images measured from the side-cut of the cartilage cube (parallel to the plane X-Z); image (a) to (l) are captured from the natural surface to the deep region. All the rendered images are taken from the same depth (30 µm off) from the cutting surface. The depth of each image is indicted. All the images are oriented in a same fashion. green: SHG signal, red: TPEF signal. Image size: 130 * 130 µm. Scale bar: 30 µm.  (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) (k) (l) 70	𝝁𝒎 170	𝝁𝒎 290	𝝁𝒎 435	𝝁𝒎 560	𝝁𝒎	 680	𝝁𝒎 770	𝝁𝒎 900	𝝁𝒎 1010	𝝁𝒎 1130	𝝁𝒎 1255	𝝁𝒎 1380	𝝁𝒎   56 4.3.3 Quantification results comparison between PS-OCT and MPM In Fig. 4.4 and Fig. 4.6, we can see that different features along the depth are observed in both PS-OCT and SHG imaging. Such depth dependent properties are further quantified and compared in this section. And the comparison of the different properties obtained from PS-OCT and SHG is shown Fig. 4.7 shows.     Fig. 4.7 Comparison of the depth-related information obtained from various analyzing methods. The black and light-blue curve has the unit of degree. The red curve describes the Intensity varying within a range from 0 to 255. The dark-blue curve is the dominating collagen fiber direction in the X-Z plane, solved by the Hough-transform based method. The green curve is the slope (unit: rad/µm) from the PS-OCT side-view measurement, but is multiplied by 𝟏𝟎𝟑 times to match the magnitude.  R-1 R-2 R-3 R-4 R-5   57 The cyan curve is the averaged phase retardation A-line from the top-view PS-OCT. Since the top-view PS-OCT covers the different zones, multi-stage slope based segmentation is applied to differentiate the different zones. The black segments are the corresponding slope-based segmentation result. Four regions can be identified as R-1, R-2, R-3, and R-4, which may respectively represent the superficial zone, transitional zone, deep zone and calcified zone. Comparing the cyan and the black curves, the transitional zone has a smaller slope than the deep zone. In the calcified zone, the slope is almost near zero, indicating very low birefringence.   The green curve shows the phase retardation slope versus depth, obtained from the side-view illumination PS-OCT, which is consistent with the observation from the selected images in Fig. 4.4 and the corresponding slopes in Table 4.1. In side-view illumination, each B-scan is from a specific depth, so a more detailed quantification can be carried out. The mean value and the standard deviation of the slope are plotted as a function of the depth. From the green curve, the mean value of the slope shows two high regions. The first region is for the depth <100 µm, showing relatively large slope and high standard variation. This region matches approximately with the R-1 zone identified above, which is the superficial zone. In side-view illumination, the superficial zone shows relatively high birefringence because the collagen fiber bundles are aligned inside the plane that is parallel to the nature tissue surface. This characteristic is also observed in the SHG images measured in the superficial zone. The other high slope region is near the depth from 500~800 µm, which falls within the above R-3 deep zone. The high slope value and low standard variation mean that the collagen fibers are well aligned in certain direction. In the deep zone, the collagen fibrils are supposed to be highly aligned in a perpendicular direction to the bone surface. Between the two high slope regions, there is a low   58 slope valley, which matches with the R-2 transitional zone. In the transitional zone, the collagen fiber is oblique and not aligned, resulting in reduce birefringence. In the deep region, with depth >900 µm, the slope is also very low, indicating that the deep region has low birefringence. This region approximately falls within the R-4 zone. In the side-view illumination, the PS-OCT and MPM imaging can reach an even deeper depth beyond R-4. This region also shows specific characteristics and thus is marked as R-5, which may likely be the bone. In R-5, the slope value is very low, which may be due to the low content of collagen.   The red curve shows the averaged SHG signal intensity as a function of depth. The SHG intensity is relatively low in R-1. It increases gradually in R-2. A rapid increase in the SHG intensity is observed in R-3. The SHG intensity peaks in the deep zone, about doubling the value compared to that in the transitional zone. The SHG intensity drops rapidly in R-5. Similar characteristics are also shown in the images in Fig. 4.4. There is a regular oscillation in the red curve, which may be affected by the illumination non-uniformity in the MPM imaging. From the images in Fig. 4.4, we can also notice that in most images the bottom area shows slightly higher intensity than the top area, even with very similar collagen structures. A calibration work may help solve this issue.  The blue curve shows the collagen fiber major orientation (in XZ plane) along the depth. The orientation is 80º, almost parallel to the tissue surface in R-1. It gradually changes from 60º to 30º in R-2. It keeps around 20º in most part of R-3. These characteristics match well with the typical collagen fiber orientation variation from superficial zone, to transitional zone and deep zone.    59  Other features are also extracted from the MPM images, such as the chondrocyte density and collagen fiber randomness. The chondrocyte density shows low variation over the different zones along the depth. The collagen randomness is judged by OI and the entropy calculated from the angle distribution. However, neither of them shows significant depth-related characteristics in the current results. More investigations are needed to study those features.   A reasonable relationship between the features extracted from the PS-OCT and SHG is demonstrated. The features also match with a prior knowledge about the typical collagen fiber organizations in the corresponding zones of articular cartilage.   4.4 Summary A comparison between PS-OCT and SHG imaging on articular cartilage has been carried out. The different features extracted from the PS-OCT and SHG are compared, including the phase retardation slopes from top-view and side-view PS-OCT, SHG intensity, and the dominating collagen fiber orientation. The segmentation regions identified by PS-OCT top-view measurement match well with the other characteristics solved by the side-view PS-OCT and MPM, which is also consistent with a prior knowledge about collagen fiber organizations along the tissue depth. PS-OCT and MPM measurements can associate with each other, showing significant potential to study tissue properties related to collagen fiber organization.   In comparing the PS-OCT results with the MPM results, a major challenge is how to match the locations in the two imaging modalities. A combined system which can acquire MPM and PS-  60 OCT imaging from the same integrated system can address this challenge. Combined MPM/OCT systems have been reported [92, 109-111]. In the future, a combined MPM/PS-OCT can be developed to study tissues with collagen fibers and birefringence, such as articular cartilage.   The collagen fibrils in the transitional and deep zones have very small diameters. The MPM system has limited resolution, which makes it challenge to resolve the individual fibrils and fiber orientation in the deep regions.  This can potentially be addressed by improving the resolution of the MPM system and developing algorithms to identify small fibrils.     61 Chapter 5: Conclusion and future work  5.1 Conclusion PS-OCT extends the capability of traditional OCT by allowing the characterization of depth-resolved tissue birefringence. Tissue birefringence occurs in articular cartilage, and varies over different structural zones along its depth, which can be evaluated by PS-OCT. With the multi-stage slope based segmentation method, PS-OCT phase retardation imaging can differentiate the different zones in the articular cartilage. The segmentation results match well with the SHG images taken from the depths located in the three zones in cartilage: superficial zone, transitional zone, and deep zone. The experiment with multiple illumination angles further demonstrates the directional sensitivity of the segmentation approach with PS-OCT on the collagen fibers at different zones. This method is promising to detect the grade of the depth-related progression of cartilage diseases, which is of significant clinic values.  A comparison between PS-OCT and SHG imaging on articular cartilage has been carried out both qualitatively and quantitatively. The phase retardation image in PS-OCT can indicate the tissue birefringence which is highly related to the collagen fiber organization and orientation. The slope-based multi-stage segmentation algorithm has been implemented to distinguish tissue layers based on the birefringence property. The segmented zones solved by PS-OCT top-view measurement match well with the slopes solved by PS-OCT side-view measurement and some features extracted from MPM, such as the SHG intensity and the collagen fiber orientation. PS-OCT and MPM measurements can associate with each other, showing significant potential to work together to explain certain detailed organization features of collagen.   62  By comparing the PS-OCT and SHG imaging results, the relationship between tissue birefringence and collagen fiber organization can be visualized and revealed by the direct imaging approaches. Moreover, the relationship between the macroscopic and microscopic properties of collagen tissues should be useful in further studies of articular cartilage or other biological tissues with abundant collagen fibers.   Compared with MPM, PS-OCT is a more suitable tool for in vivo clinic studies, due to its large FOV, fast-speed, and non-invasive property. MPM is a supportive method in this study, which is utilized to offer validation and explanation to the features solved by PS-OCT. In this study, the structural zones segmented by PS-OCT are validated by the collagen fiber organizations solved by MPM. The depth-dependent features provided by PS-OCT are proven to be associated with the collagen fiber characteristics observed by MPM. Thus, the contrast mechanism of PS-OCT and its association with collagen fiber organizations are demonstrated through our study.  It is shown that PS-OCT has great potential to evaluate the collagen fiber organizations in clinical applications.  5.2 Future directions Further improvement in the imaging modalities is of importance, as well as the quantification analysis methods. As a more enhanced type of PS-OCT, Optical polarization tractography (OPT), with capability of accessing depth-resolved optic axis and local birefringence, has been developed and applied in imaging articular cartilage [96]. In the future, quantification of local   63 birefringence and optic axis will be developed in our system, which may provide more detailed information and enhance the tissue segmentation in the PS-OCT phase retardation imaging.   We will also explore potential clinical applications with our PS-OCT. Several studies have been reported to explore the utility of PS-OCT to differentiate health and diseased cartilage tissues, or even to grade the severity of disease.  Our method, with reliable performance in differentiating the structural zones in articular cartilage tissue, may offer a more detailed evaluation about the depth-dependent features from the cartilage surface to deep region. Such features may be promising to distinguish the tissues suffered from different severity of cartilage diseases. More related experiments will be carried out in the future to test the utilization of our method on both healthy and disease cartilage tissues.    64 Bibliography [1] Sophia Fox AJ, Bedi A, Rodeo SA, “The Basic Science of Articular Cartilage: Structure, Composition, and Function,” Sports Health 1(6):461-468 (2009). [2] Camarero-Espinosa, S., Rothen-Rutishauser, B., Foster, E. J., & Weder, C, “Articular cartilage: from formation to tissue engineering,” Biomaterials science 4(5), 734-767 (2016). [3] Changoor, A., Nelea, M., Methot, S., Tran-Khanh, N., Chevrier, A., Restrepo, A., ... & Buschmann, M. D, “Structural characteristics of the collagen network in human normal, degraded and repair articular cartilages observed in polarized light and scanning electron microscopies,” Osteoarthritis and cartilage 19(12), 1458-1468 (2011). [4] Drexler, Wolfgang, and James G. Fujimoto. Optical coherence tomography: technology and applications. Springer (2015). [5] A. Rollins, S. Yazdanfar, M. Kulkarni, R. Ung-Arunyawee, and J. A. Izatt, “In vivo video rate optical coherence tomography,” Opt. Express 3, 219-229 (1998). [6] Alam, S., Zawadzki, R. J., Choi, S., Gerth, C., Park, S. S., Morse, L., & Werner, J. S., “Clinical application of rapid serial Fourier-domain optical coherence tomography for macular imaging,” Ophthalmology 113(8), 1425-1431 (2006). [7] Srinivasan, V. J., Wojtkowski, M., Witkin, A. J., Duker, J. S., Ko, T. H., Carvalho, M., ... & Fujimoto, J. G, “High-definition and 3-dimensional imaging of macular pathologies with high-speed ultrahigh-resolution optical coherence tomography,” Ophthalmology 113(11), 2054-2065 (2006).   65 [8] M. Hangai, Y. Ojima, N. Gotoh, R. Inoue, Y. Yasuno, S. Makita, M. Yamanari, T. Yatagai, M. Kita, and N. Yoshimura, Three-dimensional imaging of macular holes with high-speed optical coherence tomography, Ophthalmology 114, 763-773 (2007). [9] T. C. Chen, “Spectral domain optical coherence tomography in glaucoma: qualitative and quantitative analysis of the optic nerve head and retinal nerve fiber layer (an AOS thesis),” Transactions of the American Ophthalmological Society 107, 254 (2009). [10] J. A. Izatt, M. R. Hee, E. A. Swanson, C. P. Lin, D. Huang, J. S. Schuman, C. A. Puliato, and J. G. Fujimoto, “Micrometer-scale resolution imaging of the anterior eye in vivo with optical coherence tomography,” Arch. Ophthalmol. 112, 1584-1589 (1994). [11] A. R. S. Radhakrishnan, “Real-time optical coherence tomography of the anterior segment at 1310 nm,” Arch. Ophthalmol 119, 1179-1185 (2001). [12]   Welzel, J., Lankenau, E., Birngruber, R., & Engelhardt, R., “Optical coherence tomography of the human skin,” Journal of the American Academy of Dermatology 37(6), 958-963 (1997). [13] Y. Pan and D. L. Farkas, “Noninvasive imaging of living human skin with dual-wavelength optical coherence tomography in two and three dimensions,” J. Biomed. Opt. 3, 446-455 (1998). [14] P. J. Tadrous, “Methods for imaging the structure and function of living tissues and cells: 1. optical coherence tomography,” J. Pathol. 191, 115119 (2000). [15] J. Welzel, “Optical coherence tomography in dermatology: a review,” Skin Res. Technol. 7, 19 (2001).   66 [16] T. Gambichler, G. Moussa, M. Sand, D. Sand, P. Altmeyer, and K. Homann, “Applications of optical coherence tomography in dermatology,” J. Dermatol. Sci. 40, 85-94 (2005). [17] T. Gambichler, R. Matip, G. Moussa, P. Altmeyer, and K. Homann, “In vivo data of epidermal thickness evaluated by optical coherence tomography: Eects of age, gender, skin type, and anatomic site,” J. Dermatol. Sci. 44, 145-152 (2006). [18] V. R. Korde, G. T. Bonnema, W. Xu, C. Krishnamurthy, J. Ranger-Moore, K. Saboda, L. D. Slayton, S. J. Salasche, J. A. Warneke, D. S. Alberts, and J. K. Barton, “Using optical coherence tomography to evaluate skin sun damage and precancer,” Lasers Surg. Med. 39, 687695 (2007). [19] J. Lademann, N. Otberg, H. Richter, L. Meyer, H. Audring, A. Teichmann, S. Thomas, A. Knttel, and W. Sterry, “Application of optical non-invasive methods in skin physiology: a comparison of laser scanning microscopy and optical coherent tomography with histological analysis,” Skin Res. Technol 13, 119132 (2007). [20] B. Colston, U. Sathyam, L. DaSilva, M. Everett, P. Stroeve, and L. Otis, “Dental OCT,” Opt. Express 3, 230-238 (1998). [21] F. Feldchtein, V. Gelikonov, R. Iksanov, G. Gelikonov, R. Kuranov, A. Sergeev, N. Gladkova, M. Ourutina, D. Reitze, and J. Warren, “In vivo OCT imaging of hard and soft tissue of the oral cavity,” Opt. Express 3, 239-250 (1998). [22] B. T. Amaechi, S. M. Higham, A. G. Podoleanu, J. A. Rogers, and D. A. Jackson, “Use of optical coherence tomography for assessment of dental caries: quantitative procedure,” J. Oral Rehabil. 28, 10921093 (2001).   67 [23] R. Brandenburg, B. Haller, and C. Hauger, “Real-time in vivo imaging of dental tissue by means of optical coherence tomography (OCT),” Opt. Commun. 227, 203-211 (2003). [24] I.-K. Jang, G. J. Tearney, B. MacNeill, M. Takano, F. Moselewski, N. Iftima, M. Shishkov, S. Houser, H. T. Aretz, E. F. Halpern, and B. E. Bouma, “In vivo characterization of coronary atherosclerotic plaque by use of optical coherence tomography,” Circulation 111, 1551-1555 (2005). [25] N. Gonzalo, P. W. Serruys, T. Okamura, Z. J. Shen, Y. Onuma, H. M. Garcia-Garcia, G. Sarno, C. Schultz, R. J. v. Geuns, J. Ligthart, and E. Regar, “Optical coherence tomography assessment of the acute eects of stent implantation on the vessel wall: a systematic quantitative approach,” Heart 95, 1913-1919 (2009). [26] F. Prati, E. Regar, G. S. Mintz, E. Arbustini, C. D. Mario, I.-K. Jang, T. Akasaka, M. Costa, G. Guagliumi, E. Grube, Y. Ozaki, F. Pinto, and P.W. J. Serruys, “Expert review document on methodology, terminology, and clinical applications of optical coherence tomography: physical principles, methodology of image acquisition, and clinical application for assessment of coronary arteries and atherosclerosis,” Eur. Heart J. 31, 401-415 (2010). [27] Y. Ozaki, H. Kitabata, H. Tsujioka, S. Hosokawa, M. Kashiwagi, K. Ishibashi, K. Komukai, T. Tanimoto, Y. Ino, S. Takarada, T. Kubo, K. Kimura, A. Tanaka, K. Hirata, M. Mizukoshi, T. Imanishi, and T. Akasaka, “Comparison of contrast media and low-molecular-weight dextran for frequency-domain optical coherence tomography,” Circ. J. 76, 922-927 (2012).   68 [28] De Boer, J. F., Milner, T. E., van Gemert, M. J., & Nelson, J. S., “Two-dimensional birefringence imaging in biological tissue by polarization-sensitive optical coherence tomography,” Optics letters 22(12), 934-936 (1997). [29] De Boer, Johannes F., Thomas E. Milner, and J. Stuart Nelson. “Determination of the depth-resolved Stokes parameters of light backscattered from turbid media by use of polarization-sensitive optical coherence tomography.” Optics letters 24.5: 300-302 (1999). [30] Hitzenberger, C. K., Götzinger, E., Sticker, M., Pircher, M., & Fercher, A. F., “Measurement and imaging of birefringence and optic axis orientation by phase resolved polarization sensitive optical coherence tomography,” Optics Express, 9(13), 780-790 (2001). [31] Pierce, M. C., Strasswimmer, J., Park, B. H., Cense, B., & de Boer, J. F., “Birefringence measurements in human skin using polarization-sensitive optical coherence tomography,” Journal of biomedical optics, 9(2), 287-291 (1997). [32] Pircher, M., Götzinger, E., Leitgeb, R., Sattmann, H., Findl, O., & Hitzenberger, C. K., “Imaging of polarization properties of human retina in vivo with phase resolved transversal PS-OCT,” Optics Express, 12(24), 5940-5951 (2004). [33] Cense, B., Chen, T. C., Park, B. H., Pierce, M. C., & De Boer, J. F., “Thickness and birefringence of healthy retinal nerve fiber layer tissue measured with polarization-sensitive optical coherence tomography,” Investigative ophthalmology & visual science, 45(8), 2606-2612 (2004). [34] Yamanari, M., Miura, M., Makita, S., Yatagai, T., & Yasuno, Y., “Phase retardation measurement of retinal nerve fiber layer by polarization-sensitive spectral-domain optical   69 coherence tomography and scanning laser polarimetry,” Journal of biomedical optics, 13(1), 014013-014013 (2008). [35] E. Götzinger, M. Pircher, B. Baumann, C. Hirn, C. Vass, and C. K. Hitzenberger, “Retinal nerve fiber layer birefringence evaluated with polarization sensitive spectral domain OCT and scanning laser polarimetry: A comparison,” J. Biophoton. 1, 129-139 (2008). [36] E. Götzinger, M. Pircher, M. Sticker, A. F. Fercher, and C. K. Hitzenberger, “Measurement and imaging of birefringent properties of the human cornea with phase-resolved, polarization-sensitive optical coherence tomography,” J. Biomed. Opt. 9, 94-102 (2004). [37] M. Pircher, E. Götzinger, O. Findl, S. Michels, W. Geitzenauer, C. Leydolt, U. Schmidt-Erfurth, and C. K. Hitzenberger, “Human macula investigated in vivo with polarization-sensitive optical coherence tomography,” Invest. Ophthalmol. Vis. Sci. 47, 5487-5494 (2006). [38] Y. Yasuno, M. Yamanari, K. Kawana, T. Oshika, and M. Miura, “Investigation of post-glaucoma-surgery structures by three-dimensional and polarization sensitive anterior eye segment optical coherence tomography,” Opt. Express 17, 3980-3996 (2009). [39] G.M. Maria, "Über Elementarakte mit zwei Quantensprüngen", Annalen der Physik, 401, pp.273-294 (1931). [40] P.A. Franken, A.E. Hill, C.W. Peters, and G. Weinreich, "Generation of Optical Harmonics", Physical Review Letters, 7, pp. 118-119 (1961).   70 [41] N. Prent, R. Cisek, C. Greenhalgh, A. Major, B. Stewart, and V. Barzda, "Real-time studies of muscle cell contractions with second harmonic generation microscopy", Conference on Lasers and Electro-Optics (2009).  [42] M. R. Hee, D. Huang, E. A. Swanson, and J. G. Fujimoto, “Polarization-sensitive low-coherence reflectometer for birefringence characterization and ranging,” J. Opt. Soc. Am. B 9, 903-908 (1992). [43]    J. F. de Boer, T. E. Milner, and J. S. Nelson, “Determination of the depthresolved Stokes parameters of light backscattered from turbid media by use of polarization-sensitive optical coherence tomography,” Opt. Lett. 24, 300-302 (1999). [44] G. Yao and L. V. Wang, “Two-dimensional depth-resolved Mueller matrix characterization of biological tissue by optical coherence tomography,” Opt. Lett. 24, 537-539 (1999). [45]   M. Yamanari, S. Makita, V. D. Madjarova, T. Yatagai, and Y. Yasuno, “Fiberbased polarization-sensitive fourier domain optical coherence tomography using b-scan-oriented polarization modulation method,” Opt. Express 14, 6502-6515 (2006). [46] T. Gambichler, R. Matip, G. Moussa, P. Altmeyer, and K. Homann, “In vivo data of epidermal thickness evaluated by optical coherence tomography: Eects of age, gender, skin type, and anatomic site,” J. Dermatol. Sci. 44, 145-152 (2006). [47]   M. Yamanari, S. Makita, and Y. Yasuno, “Polarization-sensitive swept-source optical coherence tomography with continuous source polarization modulation,” Opt. Express 16, 5892-5906 (2008). [48] S. Jiao, W. Yu, G. Stoica, and L. Wang, “Optical-ber-based mueller optical coherence tomography,” Opt. Lett. 28, 1206-1208 (2003).   71 [49] B. H. Park, M. C. Pierce, B. Cense, and J. F. de Boer, “Jones matrix analysis for a polarization-sensitive optical coherence tomography system using fiber-optic components,” Opt. Lett. 29, 2512-2514 (2004). [50] M. Yamanari, S. Makita, Y. Lim, and Y. Yasuno, “Full-range polarizationsensitive swept-source optical coherence tomography by simultaneous transversal and spectral modulation,” Opt. Express 18, 13964-13980 (2010). [51] M. J. Ju, Y.-J. Hong, S. Makita, Y. Lim, K. Kurokawa, L. Duan, M. Miura, S. Tang, and Y. Yasuno, “Advanced multi-contrast Jones matrix optical coherence tomography for Doppler and polarization sensitive imaging,” Opt. Express 21, 19412-19436 (2013). [52] P. J. Campagnola and L. M. Loew, "Second-harmonic imaging microscopy for visualizing biomolecular arrays in cells, tissues and organisms," Nature biotechnology, vol. 21, pp. 1356-1360 (2003). [53] N. S. Claxton, T. J. Fellers, and M. W. Davidson, "Laser scanning confocal microscopy," Olympus. Available online at http://www. olympusconfocal. com/theory/LSCMIntro. pdf, (2006). [54] J. B. Pawley, Handbook of biological confocal microscopy: Kluwer Academic Publishers, (1995). [55] F. Helmchen and W. Denk, “Deep tissue two-photon microscopy,” Nature methods, vol. 2, pp. 932-940, (2005). [56] S. J. Mulligan, B. A. MacVicar, A. Méndez-Vilas, and J. Díaz, "Two-photon fluorescence microscopy: basic principles, advantages and risks," Modern Research and Educational Topics in Microscopy, A. Méndez-Vilas, and J. Díaz, eds.(FORMATEX, Badajoz, Spain, 2007), pp. 881-889 (2007).   72 [57] Li, X., Martin, S., Pitris, C., Ghanta, R., Stamper, D.L., Harman, M., Fujimoto, J.G. and Brezinski, M.E., “High-resolution optical coherence tomographic imaging of osteoarthritic cartilage during open knee surgery,” Arthritis Res Ther, 7(2), p.R318 (2005). [58] Cicchi, R., Vogler, N., Kapsokalyvas, D., Dietzek, B., Popp, J., & Pavone, F. S., “From molecular structure to tissue architecture: collagen organization probed by SHG microscopy,” Journal of biophotonics, 6(2), 129-142 (2013). [59] Yeh, A. T., Hammer-Wilson, M. J., Van Sickle, D. C., Benton, H. P., Zoumi, A., Tromberg, B. J., & Peavy, G. M., “Nonlinear optical microscopy of articular cartilage.,” Osteoarthritis and cartilage, 13(4), 345-352 (2005). [60] Herrmann, J. M., C. Pitris, B. E. Bouma, S. A. Boppart, C. A. Jesser, D. L. Stamper, J. G. Fujimoto, and M. E. Brezinski. "High resolution imaging of normal and osteoarthritic cartilage with optical coherence tomography," The Journal of rheumatology 26, no. 3: 627-635 (1999). [61] Drexler, W. O. L. F. G. A. N. G., D. E. B. R. A. Stamper, C. H. R. I. S. T. I. N. E. Jesser, X. I. N. G. D. E. Li, C. O. S. T. A. S. Pitris, K. A. T. H. L. E. E. N. Saunders, S. C. O. T. T. Martin, MARY BRIGHID Lodge, JAMES G. Fujimoto, and MARK E. Brezinski, "Correlation of collagen organization with polarization sensitive imaging of in vitro cartilage: implications for osteoarthritis." The Journal of rheumatology 28, no. 6: 1311-1318 (2001). [62] Han, C. W., C. R. Chu, N. Adachi, A. Usas, F. H. Fu, J. Huard, and Y. Pan, "Analysis of rabbit articular cartilage repair after chondrocyte implantation using optical coherence tomography," Osteoarthritis and cartilage 11, no. 2: 111-121(2003).   73 [63] Roberts, Mark Joseph, S. B. Adams, N. A. Patel, D. L. Stamper, M. S. Westmore, S. D. Martin, J. G. Fujimoto, and M. E. Brezinski, "A new approach for assessing early osteoarthritis in the rat." Analytical and bioanalytical chemistry 377, no. 6: 1003-1006 (2003). [64] De Boer, Johannes F., and Thomas E. Milner, "Review of polarization sensitive optical coherence tomography and Stokes vector determination," Journal of biomedical optics 7.3: 359-371 (2002). [65] Makita, Shuichi, Masahiro Yamanari, and Yoshiaki Yasuno, "Generalized Jones matrix optical coherence tomography: performance and local birefringence imaging," Optics express 18: 854-876 (2010). [66] Lim, Y., Hong, Y. J., Duan, L., Yamanari, M., & Yasuno, Y., “Passive component based multifunctional Jones matrix swept source optical coherence tomography for Doppler and polarization imaging,” Optics letters, 37(11), 1958-1960 (2012). [67] B. Baumann, W. Choi, B. Potsaid, D. Huang, J. S. Duker, and J. G. Fujimoto, “Swept source / fourier domain polarization sensitive optical coherence tomography with a passive polarization delay unit,” Opt. Express 20, 10229-10241 (2012).  [68] Ugryumova N, Attenburrow DP, Winlove CP, Matcher SJ, “The collagen structure of equine articular cartilage, characterized using polarization-sensitive optical coherence tomography,” J Phys D: Appl Phys 38: 2612–2619 (2005). [69] Xie, T., Guo, S., Zhang, J., Chen, Z., & Peavy, G. M., “Use of polarization-sensitive optical coherence tomography to determine the directional polarization sensitivity of articular cartilage and meniscus. Journal of biomedical optics,” 11(6), 064001-064001 (2006).   74 [70] Xie, T., Guo, S., Zhang, J., Chen, Z., & Peavy, G. M., “Determination of characteristics of degenerative joint disease using optical coherence tomography and polarization sensitive optical coherence tomography,” Lasers in surgery and medicine, 38(9), 852-865 (2006). [71] Xie, T., Xia, Y., Guo, S., Hoover, P., Chen, Z., & Peavy, G. M., “Topographical variations in the polarization sensitivity of articular cartilage as determined by polarization-sensitive optical coherence tomography and polarized light microscopy,” Journal of biomedical optics 13(5), 054034-054034 (2008). [72] Chu, C. R., Williams, A., Tolliver, D., Kwoh, C. K., Bruno, S., & Irrgang, J. J., “Clinical optical coherence tomography of early articular cartilage degeneration in patients with degenerative meniscal tears,” Arthritis & Rheumatology, 62(5), 1412-1420 (2010). [73] Chu, C. R., Izzo, N. J., Irrgang, J. J., Ferretti, M., & Studer, R. K., “Clinical diagnosis of potentially treatable early articular cartilage degeneration using optical coherence tomography,” Journal of biomedical optics, 12(5), 051703-051703 (2007). [74] Brill, Nicolai, Mathias Wirtz, Dorit Merhof, Markus Tingart, Holger Jahr, Daniel Truhn, Robert Schmitt, and Sven Nebelung, "Polarization-sensitive optical coherence tomography-based imaging, parameterization, and quantification of human cartilage degeneration," Journal of biomedical optics 21, no. 7: 076013-076013 (2016). [75] Ugryumova, N., Gangnus, S. V., & Matcher, S. J., “Three-dimensional optic axis determination using variable-incidence-angle polarization-optical coherence tomography,” Optics letters, 31(15), 2305-2307 (2006). [76] Ugryumova, N., Jacobs, J., Bonesi, M., & Matcher, S. J., “Novel optical imaging technique to determine the 3-D orientation of collagen fibers in cartilage: variable-  75 incidence angle polarization-sensitive optical coherence tomography,” Osteoarthritis and cartilage, 17(1), 33-42 (2009). [77] K. H. Kim, B. H. Park, Y. Tu, T. Hasan, B. Lee, J. Li, and J. F. de Boer, “Polarization-sensitive optical frequency domain imaging based on un-polarized light,” Opt. Express 19(2), 552–561 (2011). [78] J. F. de Boer, T. E. Milner, M. J. C. van Gemert, and J. S. Nelson, “Two-dimensional birefringence imaging in biological tissue by polarization-sensitive optical coherence tomography,” Opt. Lett. 22(12), 934–936 (1997). [79] C. E. Saxer, J. F. de Boer, B. H. Park, Y. Zhao, Z. Chen, and J. S. Nelson, ‘‘High-speed fiber-based polarization-sensitive optical coher- ence tomography of in vivo human skin,’’ Opt. Lett. 25(18), 1355– 1357 (2000).  [80] B. H. Park, C. E. Saxer, S. M. Srinivas, J. S. Nelson, and J. F. de Boer, ‘‘In vivo burn depth determination by high-speed fiber-based polarization sensitive optical coherence tomography,’’ J. Biomed. Opt. 6(4), 474–479 (2001).  [81] Pierce MC, Sheridan RL, Hyle Park B, Cense B, de Boer JF., “Collagen denaturation can be quantified in burned human skin using polarization-sensitive optical coherence tomography,” Burns 30: 511–7 (2004). [82] K. H. Kim, M. C. Pierce, G. Maguluri, B. H. Park, S. J. Yoon, M. Lydon, R. Sheridan, and J. F. de Boer, “In vivo imaging of human burn injuries with polarization-sensitive optical coherence tomography,” J. Biomed. Opt. 17(6), 066012 (2012). [83] Le, V. H., Lee, S., Kim, B., Yoon, Y., Yoon, C. J., Chung, W. K., & Kim, K. H., “Correlation between polarization sensitive optical coherence tomography and second   76 harmonic generation microscopy in skin,” Biomedical optics express, 6(7), 2542-2551 (2015). [84] Cense, B., Chen, T. C., Park, B. H., Pierce, M. C., & de Boer, J. F., “In vivo birefringence and thickness measurements of the human retinal nerve fiber layer using polarization-sensitive optical coherence tomography,” Journal of biomedical optics 9(1), 121-125 (2014). [85] Shyu, J. J., Chan, C. H., Hsiung, M. W., Yang, P. N., Chen, H. W., & Kuo, W. C., “Diagnosis of articular cartilage damage by polarization sensitive optical coherence tomography and the extracted optical properties,” Progress In Electromagnetics Research 91, 365-376 (2009). [86] He, B., Wu, J. P., Kirk, T. B., Carrino, J. A., Xiang, C., & Xu, J., “High-resolution measurements of the multilayer ultra-structure of articular cartilage and their translational potential,” Arthritis research & therapy 16(2), 205 (2014). [87] Y. Xia, J. B. Moody, H. A. Alhadlaq, N. Burton-Wurstert, and G. Lust, “Characteristics of topographical heterogeneity of articular cartilage over the joint surface of a humeral head,” Osteoarthritis Cartilage 10, 370–380 (2002). [88] Rogowska J, Bryant CM, Brezinski ME., “Cartilage thickness measurements from optical coherence tomography.,” Opt. Soc. Am. A 20: 357–367 (2003). [89] Lian Duan, Masahiro Yamanari, and Yoshiaki Yasuno, "Automated phase retardation oriented segmentation of chorio-scleral interface by polarization sensitive optical coherence tomography," Opt. Express 20, 3353-3366 (2012)   77 [90] Félix Fanjul-Vélez and José Luis Arce-Diego, "Polarimetry of birefringent biological tissues with arbitrary fibril orientation and variable incidence angle," Opt. Lett. 35, 1163-1165 (2010). [91] Kasaragod, D. K., Lu, Z., Jacobs, J., & Matcher, S. J., “Experimental validation of an extended Jones matrix calculus model to study the 3D structural orientation of the collagen fibers in articular cartilage using polarization-sensitive optical coherence tomography,” Biomedical optics express 3(3), 378-387 (2012). [92] Tang, S., Sun, C. H., Krasieva, T. B., Chen, Z., & Tromberg, B. J., “Imaging subcellular scattering contrast by using combined optical coherence and multiphoton microscopy,” Optics letters 32(5), 503-505 (2007). [93] Duan, L., Makita, S., Yamanari, M., Lim, Y., & Yasuno, Y., “Monte-Carlo-based phase retardation estimator for polarization sensitive optical coherence tomography,” Optics express 19(17), 16330-16345 (2011). [94] Mansfield, Jessica C., and C. Peter Winlove, "A multi-modal multiphoton investigation of microstructure in the deep zone and calcified cartilage," Journal of anatomy 220.4: 405-416 (2012). [95] Yasuno, Y., Sugiyama, S., Hong, Y. J., Kasaragod, D., Uematsu, S., Miura, M., & Ikuno, Y., “Local birefringence imaging of ocular tissue by multifunctional Jones matrix OCT,” Investigative Ophthalmology & Visual Science 56(7), 1311-1311 (2015).  [96] Yao, X., Wang, Y., Ravanfar, M., Pfeiffer, F. M., Duan, D., & Yao, G., “Nondestructive imaging of fiber structure in articular cartilage using optical polarization tractography,” Journal of biomedical optics 21(11), 116004-116004 (2016).   78 [97] Su PJ, Chen WL, Li TH, Chou CK, Chen TH, Ho YY, Huang CH, Chang SJ, Huang YY, Lee HS, Dong CY. “The discrimination of type I and type II collagen and the label-free imaging of engineered cartilage tissue,” Biomaterials 31: 9415-21 (2010). [98] He, B., Wu, J. P., Kirk, T. B., Carrino, J. A., Xiang, C., & Xu, J., “High-resolution measurements of the multilayer ultra-structure of articular cartilage and their translational potential,” Arthritis research & therapy 16(2), 205 (2014). [99] Bell JS, Christmas J, Mansfield JC, Everson RM, Winlove CP, “Micromechanical response of articular cartilage to tensile load measured using nonlinear microscopy,” Acta. Biomater. 10: 2574-81 (2014). [100] H. J. C. de Vries, D. N. Enomoto, J. van Marle, P. P. van Zuijlen, J. R. Mekkes, and J. D. Bos, “Dermal organization in Scleroderma: The Fast Fourier Transform and the laser scatter method objectify fibrosis in nonlesional as well as lesional skin,” Lab. Invest. 80(8), 1281–1289 (2000).  [101] Kim, A., Lakshman, N., & Petroll, W. M., “Quantitative assessment of local collagen matrix remodeling in 3-D culture: the role of Rho kinase,” Experimental cell research 312(18), 3683-3692 (2006). [102] Bayan, C., Levitt, J. M., Miller, E., Kaplan, D., & Georgakoudi, I., “Fully automated, quantitative, noninvasive assessment of collagen fiber content and organization in thick collagen gels,” Journal of applied physics 105(10), 102042 (2009). [103] P. Matteini, F. Ratto, F. Rossi, R. Cicchi, C. Stringari, D. Kapsokalyvas, F. S. Pavone, and R. Pini, “Photothermally-induced disordered patterns of corneal collagen revealed by SHG imaging,” Opt. Express 17(6), 4868–4878 (2009).    79 [104] S. Wu, H. Li, H. Yang, X. Zhang, Z. Li, and S. Xu, “Quantitative analysis on collagen morphology in aging skin based on multiphoton microscopy,” J. Biomed. Opt. 16(4), 040502 (2011).  [105] S. L. Wu, H. Li, X. M. Zhang, W. R. Chen, and Y. X. Wang, “Character of skin on photo-thermal response and its regeneration process using second-harmonic generation microscopy,” Lasers Med. Sci. 29(1), 141– 146 (2014).  [106] Freeman, S. A., Christian, S., Austin, P., Iu, I., Graves, M. L., Huang, L., ... & Roskelley, C. D., “Applied stretch initiates directional invasion through the action of Rap1 GTPase as a tension sensor,” J Cell Sci 130(1), 152-163 (2017). [107] Zhou, X., Ju, M. J., Huang, L., & Tang, S., “Correlation between polarization sensitive optical coherence tomography and SHG microscopy in articular cartilage,” In Proc. of SPIE Vol. 10053, pp. 1005319-1 (2017) [108] Guo, Shuguang, et al, "Depth-resolved birefringence and differential optical axis orientation measurements with fiber-based polarization-sensitive optical coherence tomography," Optics Letters 29(17): 2025-2027 (2004). [109] Jeong, B., Lee, B., Jang, M. S., Nam, H., Yoon, S. J., Wang, T., ... & Kim, K. H., “Combined two-photon microscopy and optical coherence tomography using individually optimized sources,” Optics express 19(14), 13089-13096 (2011). [110] Chong, S. P., Lai, T., Zhou, Y., & Tang, S., “Tri-modal microscopy with multiphoton and optical coherence microscopy/tomography for multi-scale and multi-contrast imaging,” Biomedical optics express 4(9), 1584-1594 (2013).   80 [111] S. Tang, Y. Zhou, K. K. H. Chan, and T. Lai, “Multiscale multimodal imaging with multiphoton 
microscopy and optical coherence tomography,” Opt. Lett. 36(24), 4800–4802 (2011). 
  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0354542/manifest

Comment

Related Items