UBC Publications

UBC Publications

UBC Publications

Innovations 2013

Item Metadata


JSON: focus-1.0115136.json
JSON-LD: focus-1.0115136-ld.json
RDF/XML (Pretty): focus-1.0115136-rdf.xml
RDF/JSON: focus-1.0115136-rdf.json
Turtle: focus-1.0115136-turtle.txt
N-Triples: focus-1.0115136-rdf-ntriples.txt
Original Record: focus-1.0115136-source.json
Full Text

Full Text

Fall/Winter 2013
"7447   0"93 74 0"
I UBC      a place of
Lasers as an imaging tool
for cancer detection
An intelligent, cloud-based platform will
enable anytime, anywhere video gaming
fall/winter 2013
Sharon Cavalier
ICICS Administrator
Craig Wilson
ICICS Communication Writer
Industry Design
University of British Columbia
289-2366 Main Mall
Vancouver, BC, Canada V6T 1Z4
Tel: 604-822-6894
Debugging complex systems
director's desk
2    Fall/Winter 2013
Bringing back the audience HUMAN-IN-THE-LOOP
Designing robots that can safely
interact with humans
Portable device allows mining
engineers to analyze rock
fragments in the field
I would like to welcome you to another edition of ICICS'
Innovations magazine. We now span 18 departments at UBC,
which allows us to undertake large-scale research projects that help
us build bridges with industry. A case in point is the ICICS People
& Planet Friendly Home sustainability initiative, which has gained
considerable momentum since we wrote about it in our Fall/Winter
2012 edition. NSERC has doubled our industry funding from a TELUS-
led consortium, and major new industry players wish to participate.
On another front, UBC was awarded a $ 10 million Canada Excellence
Research Chair in Digital Media in late 2012, and is now establishing an
associated Centre for Innovation in Digital Media at ICICS. The Chair,
Centre, and a wide range of ICICS researchers will work closely with
gaming, animation, film and TV post-production houses and other
digital media organizations to produce innovations that individual
companies or researchers cannot realize on their own. Our conversations
with industry are helping us shape the Centre, and we look forward to
collectively propelling British Columbia toward global competiveness in
digital media.
In this issue, we provide a sampling of recent research advances made
by ICICS members in biomedical engineering, microsystems, robotics,
complex systems verification, and other areas. This diversity forms the
backbone of ICICS initiatives, and makes us unique in Canada. Read on!
Panos Nasiopoulos, ICICS Director
Unique sports helmet
design minimizes damage
from head-first impacts
A new, state-of-the-art lab will accelerate microsystems research,
commercialization, and teaching
Adding "sight" to a blind
innovations magazine
As mobile computing devices from phones to
tablets to ultrabooks proliferate, consumers
want their traditionally tethered content to follow. The video game industry, in which Canada
is ranked third in the world, has responded by shifting
from console-based games to online games. With online
games, scenes in a game may be rendered locally at mobile devices, or rendered remotely in the cloud and sent
to the mobile devices for display as video frames. Neither
is an ideal solution: the former method imposes a heavy
computational load on the devices with diverse capacities,
while the latter approach imposes a heavy demand on network communication resources.
A team of ICICS researchers from communications, signal and video processing, and cognitive psychology are
tackling the problem by developing a cloud-based gaming platform that dynamically adjusts to the nature and
capabilities of the communications networks, cloud in
frastructure, and devices being used by the gamers. Led
by Electrical and Computer Engineering Professor Victor
Leung, their goal is to develop play-anywhere-on-any-
device video gaming. This would enable a player to pause
a game they are playing through a set-top box at home,
resume playing it on a portable device while riding the
bus to a friend s house, and pick up where they left off on
a smart TV in the friends living-room, all while enjoying
a consistent quality of experience.
Quality of Experience Testing
Artificial intelligence and quality of experience testing
are key to realizing this new architectural framework.
Subjective user testing will help the researchers understand how a players quality of experience is influenced
by the varying capabilities of different devices, networks,
and cloud infrastructures, across a range of game genres.
4    Fall/Winter 2013 Device parameters they will look at include screen size,
sensor availability (e.g., accelerometers), input characteristics, processing power, etc. Network variables such
as bandwidth, latency, and loss rate will be considered
for a range of networks, including 3G/4G cellular, WiFi,
cable, DSL, etc. The processing power of the cloud servers being used also needs to be considered, as well as
their geographical distribution, since they maybe spread
across a continent, and subject to data-packet delays.
Once an acceptable quality of experience has been determined for these various combinations, it can be incorporated into data compression and resource-management
algorithms that maintain the user's requisite quality of
experience as much as possible, no matter where they
are playing a game or on what device.
"The key to making all of this work," says Leung, "is
gaining knowledge of the resources and characteristics
of the cloud, the access network, and the end-user device. We can then dynamically utilize these resources to
enable the best possible experience for the player."
Intelligence Gathering
Mobile agents, or "gamelets", dispatched by the cloud
when the user logs on to play a cloud-based game will
provide the required information to the game's architectural framework. The framework will then intelligently
adapt according to variables such as processing load,
bandwidth and memory use, number of players and
their geographical distribution, and game-state delays.
Compression schemes will be adjusted to favour, for example, a steady frame rate over video quality, in order to
avoid jerkiness in the game. Players will be connected to
the nearest cloud servers to avoid delays. Game versions
with varying components will be selected that best suit
device and network characteristics.
These and other adjustments will be made continuously
as conditions change. The system will learn over time
what works best for a given mix of variables and game
types, to make appropriate decisions faster and more efficiently.
With a multitude of sensors increasingly embedded in
the fabric of our lives, we are moving toward the Internet of Everything. The information being harvested will
need to be intelligently processed and accessed, using
whatever digital device we have at our disposal. The researchers' cloud-based cognitive platform may also be a
game-changer in this area.
For more information, contact Victor Leung
at vleung@ece.ubc.ca
innovations magazine
Fall/Winter 2013    5 /
for Cancer Detection
Fall/Winter 2013 In biophotonics, light is used to study biological material. Lasers, for example, can be used to provide both
structural and functional information about tissue, and
identify potential problems. In optical coherence tomography (OCT), a long-wavelength laser is directed at the tissue
in question. Reflected photons are detected by photodetectors
and used to build up an image of the tissue. Since different tissue layers, such as the epidermis and dermis in skin, reflect
laser light differently, cross-sectional images can be rendered,
to a depth up to about 2 millimetres.
Surgical Robot Guidance
In related research, Tang has joined a team of UBC biomedical engineers led by Tim Salcudean who are working
on improving the guidance system of the da Vinci surgical
robot. Targeting prostate and kidney cancer treatment, the
team is combining MRI, ultrasound, and X-ray imaging
with the robot's camera-based system (see Innovations,
Fall/Winter 2011). The goal is to enable the surgeon operating the robot to work with greater precision, for minimal
tissue and nerve damage.
OCT can provide a quick, preliminary scan
of a fairly wide area and identify abnormalities to explore more thoroughly. Electrical
and Computer Engineering professor Shuo
Tang couples OCT with microscopy to perform both functions on the same platform.
Multiphoton microscopy (MPM) also uses
a laser as light source, but to create a high-
resolution structural picture showing differentiation, at the cellular level, of the tissue constituents of the area being scanned. In MPM, cells and
extracellular material are excited by the laser, to emit characteristic fluorescence and harmonic patterns at different wavelengths. Since MPM can image the morphology of single cells
and their association with surrounding cells and extracellular
material, abnormal tissue areas stand out.
Optical Biopsies
Tang's innovation is to use the same laser source and scanners
for OCT and MPM, allowing the respective structural and
functional images to be readily combined. A fast scan can be
performed by OCT to identify suspicious regions, and then
these can be zoomed in on and potentially diagnosed using
"We're hoping this will eventually replace the need to do biopsies," Tang says. "We call it optical biopsy' because you can see
similar structures in vivo and noninvasively that you would in
traditional biopsy'
If successful, optical biopsies would eliminate the need to take
numerous tissue samples, a procedure that is painful, alters the
tissue, and may not see test results for weeks. Tang and colleagues are currently targeting a skin cancer application, as the
skin is easy to access and does not require an endoscope probe.
They are also looking at miniaturizing the system, using fibre-
based lasers for scanning the lung, colon, and other internal
"We're hoping
this will eventually replace
the need to do
Tang is incorporating photo-acoustic
imaging by adding a fibre-optical probe
to the robot's ultrasound probe. In the
case of prostate cancer surgery, the laser
will heat up the prostate tissue, causing
it to expand and generate waves that are
picked up by the ultrasound receiver. Because the laser's light penetrates into the
prostate, a 3D "optical absorption map" of
the prostate and surrounding tissue can be generated and
added as a real-time overlay to the other obtained images.
Photo-acoustic imaging can detect blood oxygenation, as
hemoglobin in blood is a major optical absorber. Tang is
also exploiting this capability by tuning the laser's wavelength to locate areas of oxygen depletion, an indicator of
tumour growth since tumours induce excess microvascu-
latures to supply nutrition—the surgeon will be able to see
the actual tumour. Blood vessels around the tumour can
also be detected, and avoided for better surgical outcomes.
Tang plans to validate the technique on excised, cancerous
prostates, and then in patient studies, to see if the identified cancerous regions match those located by pathologists. "We are currently aiming at surgical guidance," she
says. "But if you could get an image of a region that is suspicious because of its microvasculature and oxygen level,
you could also do a targeted biopsy."
Shuo Tang's discipline-spanning work shows great promise for the detection and treatment of cancer. Her laser focus is a welcome addition to the field.
For more information, contact Shuo Tang at
innovations magazine
Fall/Winter 2013    7 FORMAL
8    Fall/Winter 2013 H   ^ft MICROCHIPS BECOME MORE AND MORE COM-
These can be very costly, delaying timely product launches
or even resulting in massive recalls. Companies are obviously reticent to disclose costs, but as an example, a bug
in an early Pentium processor and the resulting recall cost
chip manufacturer Intel $475 million in the mid-1990s.
Similar degrees of complexity are now also seen in software systems. It is no longer enough to simply test whether these systems work over a given range of values, since
the "state space", or total possible states of these systems,
has become so huge. Testing a system for a given set of
values in a conventional simulation proves only that it
works for those values and may miss critical bugs. A 2002
study by the US National Institute of Standards and Technology estimated that software bugs cost the US economy
alone nearly $60 billion per year.
UBC Verification Pioneers
Verification is the process of debugging a system and
showing that it meets its requirements. Computer Science
professor Alan Hu is a pioneer in automated system verification, but he acknowledges a long tradition of verification research at UBC: "Even back in the
1970s, the second head of the Computer
Science Department, Paul Gilmore, had
done foundational work on mathematical logic underlying formal verification." By the early 1990s, UBC had three
professors, Mark Greenstreet, Jeff Joyce,
and Carl Seger, whose research focused on verification.
Within a few years, however, two had been lured away by
industry. Hu jokes, "I always tell Carl how grateful I am
that Intel stole him from UBC, because his leaving created
the opening to hire me."
"The more automatic I can make
formal verification, the faster and
cheaper it is."
ic, and then on carrying out a labour-intensive mathematical proof. Greenstreet, Joyce, and Seger developed newer,
highly automated formal verification techniques, and
pioneered the combination of these techniques with the
older ones. Hu has devoted his career to improving purely
automatic approaches. "It's really an economic question,"
he says. "Can I help you find bugs faster and cheaper than
you otherwise could? The more automatic I can make formal verification, the faster and cheaper it is." By the end of
the 1990s, Hu, having made significant advances in hardware verification, was applying similar techniques to the
formal verification of software.
Industrial Recognition
Hu's work has drawn the attention of Microsoft, where it
has been used in verification software to find a number of
bugs in Windows drivers. The hardware verification work,
having started earlier, has found even greater recognition.
It is used in the verification products of the electronic
design automation (EDA) company Jasper Design Automation, where Hu has served on the Technical Advisory
Board since the company's inception. Jasper is now one of
the biggest EDA companies in the world, focusing on the
design and verification of semiconductors, particularly
system-on-a-chip designs. The company's stature in the industry was recognized this spring when it received a
Red Herring Top 100 Award as one of
the most promising technology companies in the United States and Canada.
Red Herring is a media company that
produces a number of technology publications. Its Top 100 Awards are widely
recognized as one of the more prestigious recognitions in
the technology and life sciences industries. Hu and UBC's
Department of Computer Science can be justifiably proud.
For more information, contact Alan Hu at
UBC research has advanced the techniques of "formal
verification," which are ways to mathematically prove that
a certain property of a system model—for example, that
a computer program doesn't crash or that a circuit computes a correct result within a time limit—holds for every
possible execution. Earlier researchers had concentrated
mainly on how to formalize systems in mathematical log-
innovations magazine
Portable Device Allows Mining Engineers to Analyze
Rock Fragments in the Field
By Bahram Sameti, Motion Metrics International Corp.
In mining, knowing the size of the rock fragments
following a blast is valuable, so that loading, hauling, and crushing equipment can be fine-tuned accordingly, and blast parameters can be adjusted.
Very small rock fragments, for example, indicate over-
blasting, and large ones under-blasting. Sieve analysis has
traditionally been used to determine the size distribution
of fragments, in an exhaustive procedure. More recently,
image-based rock-fragmentation analysis has provided a
feasible and more sophisticated alternative.
In most commercially available systems, rock delineation results are converted to physical rock sizes by introducing one or more objects with known geometry (such
as a basketball) to the scene as a size-scaling reference.
two cameras is then augmented with depth information
and fed to the rock delineation algorithm to identify the
boundaries of each fragment in the image. For each set of
boundary pixels, the system identifies the corresponding
points in the 3D point cloud to determine the real-world
coordinates of the individual rock boundaries, which are
then used to calculate the rock's size and volume.
"This is an ideal application of stereo imaging," Tafazoli
says, "since the disparity of the rock scene lends itself well
to stereo vision. The device is portable, so blast engineers
can easily do rock fragmentation sensing in the field,
without needing a reference object in the scene."
Adding a 3D orientation sensor to the device makes it
Portable device captures the scene in 3D and identifies
rock-fragment boundaries without needing a reference object.
A group of engineers at Motion Metrics led by Bahram
Sameti has developed a patent-pending technology that
eliminates the need for a reference object in the scene.
Motion Metrics is a UBC spin-off company founded by
ICICS member and Electrical and Computer Engineering
adjunct professor, Shahram Tafazoli (see ICICS FOCUS
magazine, Spring 2010).
In this new approach, stereo imaging is used to measure a
3D point cloud of the scene. The 2D image from one of the
possible to perform other useful calculations, such as remote sensing of the slope of the bench face in an open-pit
mine. Automatic sensing of both the slope and the distance of the device to the face gives the device enough
information to warn the user if they're too close to be safe.
For more information, contact Motion Metrics at
10    Fall/Winter 2013 Human-in-
This  year's   2013   Motion   Metrics/
ICICS Graduate Scholarship has
been  awarded to  Mechanical
Engineering  graduate  student
Matthew Pan. The annual award, created by ICICS and Motion Metrics International (see adjacent story) in 2011,
is granted to an ICICS-affiliated graduate student whose research spans disciplines and is geared
toward developing applications.
Matthew's work clearly fits the bill. Supervised by Professor Elizabeth Croft in the Collaborative Advanced Robotics and Intelligent Systems Laboratory (CARIS) at UBC,
Matthew is focusing on controlling interactions between
humans and robots for industrial applications.
Robots are already widely used in manufacturing, to carry out repetitive, assembly-line tasks such as welding car
parts. These robots work in isolation, performing pre-programmed chores that do not involve humans. Designing
robots that can safely and effectively interact with humans,
however, is a pressing problem faced by today's robotics researchers that must be overcome if robots are to play a bigger role in our daily lives.
Pan is tackling this problem head-on, by
developing motion-control models to govern interactions between mobile robots
and workers who have no specialized
robotics training. He is framing his
work around cooperative tasks that
are commonly seen in the workplace, including object handover
and shared lifting. The research
will produce innovations in real-time robotic gesture recognition, role negotiation, and
safety systems.
Pan   intends   to   test
and  refine  his  prototype       strategies
using     state-of-the-
art  robotics  platforms
available in the CARIS lab, and
through user evaluations. "My
goal," he says, "is to develop a
library of motion-control strategies for mobile, manipulator-
type robots that are safe, intuitive, and ergonomic. They will
enable workers and their robotic
assistants to make full use of their
respective skills."
Although the initial focus of his work is on industrial applications, in the long-term Pan's research could have an impact on rehabilitation, homecare
for the elderly and those with disabilities, and education.
For more information, contact Matthew Pan at
innovations magazine
Fall/Winter 2013   11 HELPING
Unique Sports Helmet Design
Minimizes Damage from
Head-First Impacts
risk-injury curves are then developed to guide the design of
protective devices. Mechanical engineering professor Peter
Cripton is an expert in this area, known as injury biomechanics. Understanding and preventing spinal cord injury
(SCI) is a major focus of his work.
Cripton is Co-Director of the UBC Orthopaedics and Injury Biomechanics Research Group, with Canada Research
Chair in Spine Biomechanics Tom Oxland. He and Oxland
are also Principal Investigators with the International Collaboration on Repair Discoveries (ICORD), a multidisciplinary spinal-cord injury research centre based at Vancouver General Hospital. "Spinal cord injury is a mechanical
impact that results in a biological response," he says. "Our
interactions with the neuroscientists at ICORD provide us
with the clinical input we need to guide our work." An example of this might be determining the cellular responses to
injury of the neurons that make up the brain or spinal cord
tissue. Criptons clinical colleagues at ICORD can also give
him feedback based on firsthand knowledge of the potential
benefits of a proposed protective device.
Designing a Better Sports Helmet
Head-first impacts are common in sports such as football,
hockey, and mountain biking, with football players slamming
into each other, hockey players crashing into the boards, or
mountain bikers going over the handlebars. In these impacts,
the head comes to an abrupt halt while the
torsos momentum continues, deforming
the cervical spine (neck) and potentially
causing a spinal cord injury. A new type of
helmet designed by Cripton and PhD student Tim Nelson may help prevent some of
these injuries.
The Pro-Neck-Tor™ consists of inner and
outer shells separated by about 20 millimetres and joined together at the sides by pivot mechanisms. In
most impacts, the helmet absorbs loads as a normal helmet
would. In impacts exceeding a certain injury threshold, however, a component of the pivot mechanism breaks, allowing
the inner shell to slide forward or backward within the outer
shell. The momentum of the following torso is absorbed by the
helmet, rather than the neck, reducing SCI by up to 56% in
experimental tests.
"It's important
that we study as
many different
impacts as we can
before certifying
the helmet."
The double-shell design of the Pro-Neck-Tor™ has an added benefit "It's important that we study as many different impacts as we
can before certifying the helmet," Cripton stresses. "We wouldn't
want to cause a brain injury, for example, while trying to prevent
a neck injury' In the course of testing the helmet in the sort of
scenario that might cause a concussion, Cripton found that it decreased head acceleration and the resulting shear stress on the brain
to a greater extent than a conventional football helmet would, due
to deformation of the outer shell and cushioning from the interior
Cripton and colleagues have been approached by several companies interested in licensing the technology from UBC, and are now
designing sport-specific versions of the helmet.
A Lifelike Surrogate Neck
A major problem faced by injury biomechanics researchers is that
they cannot conduct tests on live human subjects beyond the injury
threshold. Instead, they have to choose from a variety of surrogates,
including crash-test dummies, computational models, and cadavers. Each has advantages and drawbacks, but none is "biofidelic"
or truly lifelike. Cripton has addressed this deficit by developing an
aluminum neck that approximates the human neck's response to
head-first impacts, specifically a diving accident. Using a combination of pulleys, springs, articulating vertebrae, and rubber "discs" he
has created a surrogate neck that, according to Cripton, "behaves
like a human neck, especially with respect to the reaction forces
that develop over time" in a head-first impact.
To test the Pro-Neck-Tor™ using the neck, Cripton and colleagues
attached an aluminum head with a simulated scalp and dropped it
from a drop-tower. Using high-frame-rate video, high-speed Xray,
and various accelerometers, they were able to
capture the metrics they needed to validate the
helmet against risk-injury curves.
Cripton, along with Tom Oxland and ICORD
Principal Investigator John Street (Orthopaedics), is now looking at the effects of lateral
bending of the neck in head-first impacts. The
results of this Canadian Institutes of Health
Research-funded study will help guide modifications to the Pro-Neck-Tor™ to further reduce SCI. The study
may also lead to other preventative measures, and improve treatment for those suffering from SCI.
For more information, contact Peter Cripton
at cripton@mech.ubc.ca
innovations magazine
Fall/Winter 2013   13 ADVANCED
A New, State-of-The-Art Lab
Will Accelerate Microsystems
Research, Commercialization,
and Teaching
By lulia Nairn Schriver
In Canada, there are currently four micro-electromechanical systems (MEMS) research centres focused on designing and testing new microdevices,
including UBC's new Adaptive Microsystems Lab,
or AdaMist, which opened in July 2013. Led by Electrical and Computer Engineering (ECE) professor Edmond
Cretu, AdaMist is part of the Embedded Systems Canada
(emSYSCAN) national network. The 37 institutions comprising the emSYSCAN network provide platform-based
microsystems design and prototyping facilities to advance
microsystems R&D. AdaMist, however, is unique in Canada. As Cretu explains, "there are no other centres in the
country that have this grouping, this set of equipment.
Our integration among 3D printing, aerosol jet deposition, photo-lithography, and laser removal is unique."
Cretu's efforts to bring the new AdaMist lab to UBC began
in 2006 when he initiated contact with CMC Microsystems, an umbrella organization based at Queen's University that supports integrated microsystems research and
commercialization. Relationships with CMC developed
over time. "We gradually developed a kind of bilateral
communication," Cretu says. "They invited me to dinner
and we discussed the vision I had for setting up a lab at
UBC. They apparently liked my vision and were thinking
along the same lines."
Cretu proposed developing complex heterogeneous microsystems by integrating conventional technologies in
UBC's Advanced Materials and Process Engineering
Laboratory (AMPEL) with new technologies now available in the AdaMist lab. His proposal helped inform the
overall vision that led to the award of a $48.26 million
Canadian Foundation for Innovation (CFI) Leading-
Edge Fund national grant in 2010, and the establishment
of the emSYSCAN network.
State-of-the-Art Equipment and
AdaMist is equipped with state-of-the art equipment
such as an OPTOMEC aerosol jet printer, valued at over
$400,000 USD. This printer enables the size of electronic
systems to be greatly reduced, by incorporating nano-
materials to produce fine-featured circuitry and embedded components. Innovative MEMS designs can be
adhered to a flexible substrate; surgical instruments, for
example, can be upgraded with MEMS components for
greater precision. Aerosol jet printing is useful in low-
14    Fall/Winter 2013 volume fabrication of electronic circuitry and components, ideal for life sciences research and prototyping,
and ultimately for the development and fabrication of
next-generation microelectronic devices.
"Together with our graduate students," Cretu remarks,
"we have the potential to
create novel types of fabrications steps. We want to get
from the design level to the
application level and not just
play with isolated pieces."
Cretu stresses that "it is very
important for ECE students
to realize that the world of
engineering has changed,
that most of the projects
they will be dealing with
outside of the lab will be interdisciplinary in nature. Students studying nanotechnology also need to know about quantum mechanics,
chemistry, optics, and other disciplines, as the borders
between fields are becoming fuzzier."
Connecting with Industry
Opening up opportunities for ECE students to design
and fabricate MEMS devices by providing access to
state-of-the-art equipment, is key to attracting young
minds to careers in MEMS research. Cretu, who spent
"We want to get from
the design level to the
application level and
not just play with isolated pieces."
five years in Belgium working with teams to create integrated circuits primarily used in the automotive industry, wants AdaMist to be not only a place of research but
also a place where researchers and companies connect
and make use of both the technology and intellectual resources available in the lab. BC companies, such as Assex
Technology and Ultrasonix Medical
Corporation, specializing in imaging systems, took notice in 2010
and collaborated with Cretu's team
by integrating ultrasound transducer heads into their product line.
Incorporating MEMS devices to
upgrade biomedical imaging equipment, such as ultrasound machines,
continues to be commercially lucrative, as ultrasound imaging occupies more than 25% of the biomedical imaging market. It is little wonder that microsystems
designs emerging from the Adaptive Microsystems Lab
are garnering the attention of some very large biomedical
companies in the United States. Cretu's efforts to establish
AdaMist on the national level are also beginning to unfold
on a much wider stage.
For more information, contact Edmond Cretu at
Fall/Winter 2013   15
ight" to a Blind Proce
In this procedure, the anesthesiologist selects the injection
site by feel, then guides the needle into the space just above
the dura, or membrane, enveloping the spinal cord. The target site is identified by a loss of resistance to injected saline
solution. A catheter is then fed through the needle to supply
the anesthetic.
The success of this essentially blind procedure depends
highly on the skill of the practitioner, and can take up to
100 attempts to perfect. It also becomes much more difficult
in obese patients or those with scoliosis or a previous back
surgery. The most common complication is overshoot of the
epidural space and puncture of the dura, causing leakage of
cerebrospinal fluid and severe headaches. Inadequate anesthesia is also a common problem. In extreme cases, respiratory failure or paralysis may occur. With as many as 175,000
epidurals performed annually in Canada alone and a failure
rate of 6-20 percent, there is a clear need for an improved
guidance technique.
Approaching the Problem from
the Side
Advances in ultrasound technology have enabled adding
visual guidance to help the anesthesiologist select the puncture site. This is useful only before the procedure, however,
as the ultrasound transducer sits over and obstructs the injection site. UBC biomedical engineers Rob Rohling (ECE/
MECH) and Purang Abolmaesumi (ECE), along with Drs.
Allaudin Kamani and Vit Gunka at BC Women's Hospital
and Dr. Jill Osborn at St. Paul's Hospital, have harnessed
new 3D ultrasound technology to develop a technique for
real-time imaging of epidural needle insertion through to
the target space.
Rohling and Abolmaesumi have considerable experience
in exploiting ultrasound to improve surgical outcomes (see
Innovations, Fall/Winter 2011 and this issue, p. 7). In their
epidural work, they are making use of a plastic needle guide
16    Fall/Winter 2013 invented by Rohling that clips on to the side of the ultrasound transducer. Knowing the angle of the needle
relative to the coordinates of the 3D ultrasound volume,
and using a feature-selection technique based on training
data from a number of scanned patients, they can automatically identify the midline of the spine and choose an
appropriate injection site.
Targeting Widespread Use
Echoes of the ultrasound beam off the needle in this technique are not ideal, however, given the angle of the needle
relative to the transducer. One way to get around this is to
use expensive "echogenic" needles, which are pitted with
tiny divots to reflect ultrasound beams in different directions, keeping the needle visible at all angles. But as Rohling
says, "We wanted to keep costs down so that our innovation
will have the widest possible impact." To realize this goal,
the team is employing an image-rendering method they developed previously, based on "compound imaging," where
multiple ultrasound images taken at slightly different angles
are combined to generate a coherent image of the needle.
Variations and extensions of this method have been adopted by several different ultrasound manufacturers, and it is
in clinical use. In the current application, it enables the team
to use a much more economical standard epidural needle. The
3D ultrasound machine they are using is also widely available.
With the input of many of the staff anesthesiologists at the collaborating hospitals, the team will also design a unique user
interface, intended to satisfy the requirements of the end user.
The interface will be refined as the system goes through a series
of tests, initially on volunteers to test the ultrasound imaging
component before needle insertion and then as a supplementary tool to guide needle insertion in patient studies.
The team's end goal is not to replace current epidural anesthetic
procedures but to complement them. "The loss-of-resistance
technique for detecting the epidural space will still be used,"
Abolmaesumi stresses, "but SURE (Smart Ultrasound Rendering for Epidurals) will help make epidurals safer, easier to master, and more widely practiced."
In the long-term, SURE may also provide visual guidance for
other medical procedures such as needle biopsies.
For more information, contact Rob Rohling at
rohling@ece.ubc.ca or Purang Abolmaesumi at
innovations magazine
Viewers accustomed to social media wish to be more actively engaged with the media they are consuming than
is possible with traditional TV programming. While the
production values and editorial capabilities of broadcast
TV remain far superior to those of the digital media competition, advertising revenues are waning. The long-term
viability of the medium many of us, for better or worse,
grew up on is in question.
Mahsa Pourazad, Research Director of ICICS' Digital
Media Lab, is leading a group of researchers who are developing a comprehensive suite of technologies aimed at
wooing back TV audiences. Their overall goal is to create
a broadcast platform that enlists viewers as providers of
additional content, with personalized interactivity options
tied to program elements. Transforming broadcast television into a form of social media, the researchers believe,
will bring back audiences and help guarantee its survival.
18    Fall/Winter 2013 Trustworthy Crowdsourcing
"A major focus of our work," Pourazad says, "is on developing technologies that will allow the public to contribute to programming." A broadcaster, for instance, might
request a selection of these individuals to create and upload local content for a documentary about the Canadian
Arctic. Contributor guidelines will ensure it meets minimum industry requirements for video quality, compatibility, etc. The broadcaster will then use tools developed
by Pourazad and her colleagues to analyze and select submitted content for inclusion in the program.
For broadcasters to adopt this new approach to producing
programs, however, they must be able to trust the contributors. Pourazad plans to develop a trustworthiness
measure for assessing the reliability and accountability of
contributors, and to detect fraudulent content. Individual
trustworthiness profiles will be refined over time, improving the robustness and accuracy of the measure.
Before additional content can be added to programming, the main video stream must be prepared appropriately. The researchers are identifying data that must
be captured and stored while the program is being shot,
including camera parameters and position, scene depth
maps, and objects of interest. These data will later be used
to integrate video streams, as well as Web content and
augmented-reality enhancements (e.g., information overlays). Pourazad is also developing video analysis tools that
will allow broadcasters and post-production houses to
incorporate these enhancements into existing programs,
tied to distinct program elements such as landmarks and
segments of historical interest.
"Our focus on content-specific interactivity," Pourazad
emphasizes, "distinguishes our work from existing approaches, where additional program information is not
tied to program components." Machine-learning techniques will be developed to personalize options for viewers. Interactivity cues will be presented to the viewer based
on their interests, defined initially by questionnaire, and
refined over time based on viewing patterns. Viewers will
always have the option to customize interactivity choices,
on either a global or program-specific basis. They will also
be able to suggest links to additional content based on personal experience or Web browsing, by providing feedback
to a broadcaster's social media site. The trustworthiness
tools developed to assess incoming video streams will be
customized to measure the reliability of viewer feedback.
Broadcasters will be in a position to enrich programming,
and, if desired, target it at a specific demographic, in a
much more comprehensive manner than is possible with
the current focus-group based approach.
For more information, contact Mahsa Pourazad at
innovations magazine
Fall/Winter 2013   19 Heavyweight plan.
Featherweight commitment.
Introducing TELUS SharePlus Plans,
now on 2 year terms.
Unlimited nationwide talk & text*
■ Add an extra line from $35/mo,
■ Share your data*
Get it all on Canada's most reliable 4G network.'
Samsung GALAXY S4~
Learn more at
the future is friendly"
^Premium and subscription messages are not included. Customers with devices not able to display picture or video messages will receive a text message that includes a web address for viewing. tOnly one SharePlus subscriber on the account reguires a data option. That data can be shared with
up to four additional devices. The first device on each TPLUS SharePlus Plan must be a Smartphone, Smartphone Lite or other mobile phone. Compatible phone reguired for BYOD option, tBased on testing of voice-call success rates, data-session completion rates and industry-standard call-guality
measures against other national wireless service providers in metropolitan areas across Canada. TELUS, the TELUS logo and the future is friendly a re trademarks of TELUS Corporation, used under licence. All other trademarks are the property of their respective owners. ©2013 TELUS
UBC's Institute for Computing, Information and Cognitive Systems (ICICS) is an
umbrella organization that promotes collaboration among researchers from the faculties of
Applied Science, Arts, Commerce, Education, Forestry, Medicine, and Science, and with
industry. ICICS facilitates the collaborative multidisciplinary research of approximately
200 faculty members and 1,000 graduate students in these faculties.
Our members attract approximately $18 million annually in grants and contracts. Their
work strengthens Canada's strategic Science and Technology research priority areas,
benefiting all Canadians.
((.(CD   hj|
fAUUESTiu** vuAiuitncr ^^^^^
a place of mind
ICICS, University of British Columbia
289-2366 Main Mall
Vancouver, BC V6T 1Z4


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items