PhD Thesis Proposal: Michael A. Kokko

Monday, May 20, 2019, 9:00–11:00am

Jackson Conference Room, Cummings Hall

“Toward Autonomous Navigation in the Abdominal Cavity: A Framework for Navigated Organ Exposure in Robotic Laparoscopy”


The field of surgical robotics has demonstrated marked success over the past two decades in augmenting the dexterity and reach of the minimally invasive surgeon, while maintaining (or improving upon) patient outcomes associated with traditional laparoscopic and open approaches. Barriers to wider adoption of robotics include high capital and operating expenses, the potential for increased operative time relative to open surgery, and attenuation of the visual and tactile sensory information that has guided the hands of surgeons for centuries. Sensory cues are of particular importance during the exposure phase of abdominal surgery, when surgeons navigate with limited visibility through webs of fascia and adipose tissue by incrementally exposing anatomical landmarks, consulting pre-operative imaging, and drawing from their own surgical experience. While the image-guided surgery community has developed a number of navigational solutions for visualizing tumors within exposed organs, relatively little attention has been paid to assisting surgeons during the process of gaining initial exposure. A successful guidance system for abdominal exposure will reduce operating time (lower cost), surgeon cognitive load (higher quality output), and anesthesia time (quicker recovery) without increasing the risk of iatrogenic damage.

This proposal introduces a novel method of guidance for surgical exposure based on mobile robot navigation. Under this framework, a probabilistic geometrical model of patient anatomy will be constructed from pre-operative imaging, registered to the surgical field, and updated based on intraoperative measurements. Camera-centric estimates of the location/orientation of the target anatomy and associated uncertainty metrics will be available to surgeons as selectable overlays on the endoscopic video feed. System performance will be evaluated using both phantom and animal models of Robot-Assisted Laparoscopic Partial Nephrectomy (RALPN), setting the stage for undertaking a larger animal study and first-in-human trials.

Thesis Committee

For more information, contact Daryl Laware at