How Wearable Technology Can Facilitate AI Analysis of Surgical Videos

Figures

• 

  FIGURE 1.. Images and data from the American College of Surgeons Surgical Metrics Project. A, Surgeon instrumented with multiple sensors (forehead EEG, lapel audio, hand motion) during simulated bowel repair. B, EEG spectrograms for 2 different surgeons. One surgeon was visually and verbally engaged with their operative assistant and the other one was not hence mostly blue signal. C, Close up of motion tracking wires beneath surgical gloves and motion diagram showing hand positions near the bowel and tool tray as well as distance away from the bowel when pulling suture through. D, Motion data from right and left hand synchronized with video.

• 

  FIGURE 2.. Eye tracking technology and data. A, Remote computer-attached eye tracker. B, Simulated hydrogel task. C, Analysis of results utilizing Scanpaths-ordered set of fixations points (depicted by circles) connected by saccades (depicted by lines), and dwell time (denoted by a larger diameter of blue circle) from a representative individual from each group. D, Analysis of results utilizing Scanpaths-ordered set of fixations points.

• 

  FIGURE 3.. Full-cap EEG technology and data. A, study participant wearing EEG cap, (B) EEG electrode locations on participant’s scalp, and (C) EEG neural activity differences among students, residents, and attendings during mental imagery of the same robotic suturing task. X-axis labels correspond to the sensor locations shown in (B).

• 

  FIGURE 4.. fNIRS FLS study and data. A, fNIRS physiology diagram and set up during FLS. B, Heatmap data showing that surgical experts generate more intense activation as shown in red.31 FLS indicates fundamentals of laparoscopic surgery; fNIRS, functional neuroimaging.