Perception Augmentation and Assistance for Improved Surgical Awareness

Abstract

Robotic surgery and surgery in deep and confined spaces present challenges in situational awareness due to limits on miniaturization and disposability of sensing equipment, limited or occluded sensory information, and lack of visualization of the entire surgical scene. To address these perception deficits, this work explores augmentation of situational awareness in minimally invasive deep and confined spaces, such as the inner ear, and the use of robotic technology in a range of surgical scenarios.

Methods are evaluated using both commercially-available robotic systems and research prototype robotic platforms with new capabilities for force-sensing of continuum robots explored in this research. I present methods for assistance in middle ear and cochlear implant surgery, where limited intraoperative sensory information requires external information sources and tools to guide surgical execution. In single-port surgical domains, I present novel methods for obtaining force information with continuum robots and new control and feedback methods for controlling forceful interaction with soft organs. This is used for updating surgical registration and augmenting task completion strategies in simulated standard surgical procedures.

This work found that expanding the toolset of sensing tools and manipulation methods could assist surgeons in the completion of surgical tasks and in providing assistance in surgical awareness. In the middle ear, a new robotic platform allowed access to previously inaccessible structures that could be visualized but not reached. In cochlear implant insertion, novel visual guidance methods showed the ability to increase repeatability in certain surgical performance metrics. Novel methods for compensating for uncertainty in intrinsic force sensing with continuum robots were developed and used for palpation of soft organs. Multi-modality feedback methods were developed and evaluated for telemanipulated forceful surgical task execution.

The methods presented in this thesis will help form the basis for future exploration and approaches in surgical assistance during a variety of procedures. As more safety and autonomous features are integrated into surgical practice, these methods will be useful as potential approaches to limit surgeon cognitive workload, expand surgeon capacities, and increase surgical performance.