VISE Fall Seminar Jie Ying Wu, PhD 11.3.22
VISE Fall Seminar
to be led by
Jie Ying Wu, PhD
Assistant Professor, Computer Science
Date: Thursday, November 3, 2022
Time: 11:45 a.m. Lunch, 12:00 p.m. start
Location: Stevenson 5326
Building better tools for surgeons through machine learning
Training for surgeons has long relied on the paradigm “see one, do one, teach one”. This view of training does not incorporate much feedback, nor does it target training to each individual. This comes at a societal cost as surgical skill has been shown to be one of the biggest predictors of surgical outcome. We know individualized learning results in better skill acquisition yet feedback for surgery has generally been expensive: requiring an expert surgeon to watch trainees and then assess them for each procedure. With the variety of tools available these days, we can develop automated ways to give feedback and train surgeons so that they acquire skills faster and retain them better. My research has looked at two aspects: 1) how can we better measure surgeon skills and 2) how can we target training to areas that need improvement. I look at the eye-gaze of surgeons and measure their pupillometry as indicators of cognitive load. Once we have identified the parts of surgery that are difficult to learn, we can also build tools that make those parts easier. Towards this goal, I use machine learning to improve surgical scene modeling and do sensorless force estimation for robot-assisted surgeries.
Her work on leveraging vision and kinematics data to improve realism of biomechanic soft tissue simulation for robotic surgery received the audience award at the 2020 International Conference on Information Processing in Computer-Assisted Interventions. Wu has published in nine full-length publications at top-tier venues, and she has won multiple awards. In addition, Wu was part of a Johns Hopkins team that received a Kaggle COVID-19 Dataset Award for informing the United States’ response to COVID-19.