Open Collections

Abstract

Spinal surgery, particularly pedicle screw placement, demands exceptional precision to avoid severe complications such as neurological injury, vascular damage, and mechanical instability. Current navigation techniques—including freehand methods, fluoroscopy, computed tomography (CT)-based navigation, and robotic systems—offer varying degrees of accuracy but are constrained by challenges such as high costs, radiation exposure, and workflow disruptions. To address these limitations, this thesis proposes and evaluates an innovative, ultrasound-augmented, self-localizing drill guide that integrates polymer-based capacitive micromachined ultrasonic transducers (poly-CMUTs) for real-time surgical navigation. The research methodology comprised computational simulations, experimental evaluations, and machine learning-based optimizations. Initial benchmarking of existing technologies (Chapter 2) highlighted persistent challenges in achieving a balance between accuracy, cost, and safety in pedicle screw navigation. Computational modeling (Chapter 3) then demonstrated that strategically placed ultrasound transducers could achieve sub-millimetric localization accuracy, providing a foundation for the tool's design. Experimental studies (Chapter 4) validated the accuracy of poly-CMUT transducers in measuring bone surfaces, while machine learning techniques (Chapter 5) significantly improved bone segmentation and registration accuracy. Finally, prototype emulation studies (Chapter 6) confirmed the clinical feasibility of the ultrasound-augmented drill guide, demonstrating its potential to match or exceed the performance of existing systems without the need for ionizing radiation. The findings underscore the clinical and practical implications of integrating ultrasound into surgical tools. The proposed approach offers a radiation-free, cost-effective, and highly accurate alternative to CT-based navigation systems, with broad applicability in spinal surgery and potential extensions to other surgical domains. Future work will focus on translating these promising experimental results into clinical practice, exploring regulatory pathways, and refining the system for real-world applications. The thesis advances the field of spinal navigation by presenting a novel approach that prioritizes patient safety, surgeon convenience, and healthcare accessibility.