Safe anatomy-driven tracking for navigation and robotics in medical and surgical procedures

Medical imaging enables clinicians to visualize human anatomy and guide procedures with greater precision. However, existing navigation systems are prohibitively expensive, infrastructure-heavy, and difficult to operate, especially in resource-limited settings. This thesis explores these challenges by introducing two novel robotic platforms that leverage state-of-the-art computation, machine learning, and computer vision to deliver accurate image guidance for diagnostic and interventional medical procedures. The first system is a low-cost, articulated-arm neurosurgical navigation system for external ventricular drain (EVD) placement. Using a novel skull-mounting mechanism, the device provides forward-kinematic tracking of neurosurgical instruments, eliminating the need for traditional infrastructure-intensive tracking systems and enabling use in bedside emergency surgery settings. The second platform is an autonomous eye examination robot called EYESIGHT, designed to perform reliable, high-quality retinal photography on freestanding patients. By unifying robust human-tracking algorithms, closed-loop robotic control strategies, and user-friendly clinical workflows, these systems demonstrate how safe, anatomy-driven robotics can be developed to enable safe anatomy-driven system control and workflows in medical and surgical procedures. Phantom studies, ex-vivo validations, and preliminary user evaluations confirm that both platforms attain sub-millimeter accuracy while substantially mitigating workflow or resource requirements for both procedures described in this work. These findings open new frontiers for image-guided interventions in underserved clinics, emergency departments, and telemedicine environments. By bridging validation methodology advances with clinical demands, this thesis lays fundamental groundwork for broader, more equitable adoption of image-guided medical diagnostics and interventions globally. It proposes a framework for pre-clinical evaluation of novel systems through rapid prototyping, phantom testing, and iteration toward clinical evaluation. This testing framework encourages researchers in the intersection of medicine and technology to rapidly conceptualize and evaluate new systems.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Computer Science

Grantor

University of Illinois Urbana-Champaign

Year dc:date

2025

Author dc:creator

• Smith, Alexander Daniel

Contributors dc:contributor

• Hauser, Kris
• Arnold, Paul
• Driggs-Campbell, Katie
• Wang, Shenlong
• Draelos, Mark

• Medical Robotics, Engineering and Medicine, Neurosurgery, Surgical Navigation, Image Guided Therapy, Ophthalmology, Optometry, Medical Imaging, Medical Devices, Medical Technology, Computer Science, Systems Engineering

Statement dc:rights

• © 2025 Alexander Daniel Smith

Language dc:language

en

Handle dc:identifier

https://hdl.handle.net/2142/132518

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/132518