Development of performance and learning rate evaluation models in robot-assisted surgery using electroencephalography
Somayeh B Shafiei1, Saeed Shadpour2, Farzan Sasangohar3
1Intelligent Cancer Care Laboratory, Department of Urology, Roswell Park Comprehensive Cancer Center, Buffalo, NY, 14263, USA. Somayeh.besharatshafiei@roswellpark.org.
NPJ Science of Learning
|January 19, 2024
Summary
This study developed objective models for evaluating robot-assisted surgery (RAS) skills using electroencephalogram (EEG) and eye-tracking data. These models show promise for accurately assessing surgical performance and learning rates.
Area of Science:
- Robotics
- Neuroscience
- Medical Simulation
Background:
- Current robot-assisted surgery (RAS) performance evaluations are subjective, costly, and inconsistent.
- Objective metrics are needed to accurately assess surgical skills and learning curves in RAS.
Purpose of the Study:
- To develop objective models for evaluating RAS performance and learning rates using simulator tasks.
- To utilize electroencephalogram (EEG) and eye-tracking data for skill assessment.
Main Methods:
- Recorded EEG and eye-tracking data from 26 subjects during surgical simulator tasks (Tubes, Suture Sponge, Dots and Needles).
- Extracted functional brain networks from EEG data and analyzed 12 eye-tracking features.
- Developed linear models for performance and learning rate evaluation, incorporating subject-wise standardization.
Main Results:
- Models incorporating subject-wise standardization showed improved R-squared values.
- Specific eye-tracking features (pupil diameter, saccade rate, pupil diameter entropy) correlated with performance in different tasks.
- EEG-derived features (temporal flexibility, search information) were associated with performance and learning rates across various brain areas and frequencies.
Conclusions:
- Objective evaluation of RAS performance and learning is feasible using EEG and eye-tracking data.
- The developed models offer a potential pathway to standardize and improve surgical skill assessment in RAS.
- Further validation with larger sample sizes and diverse tasks is recommended.
