A transfer entropy-based framework for multimodal analysis of visuomotor and neuromuscular coordination in surgical skill assessment
Open Access
- 6 October 2026
- journal article
- research article
- Published by Wolters Kluwer Health in International Journal of Surgery
Abstract
Aim: To investigate directional information flow among eye motion, tool motion, and muscle activity across surgical skill levels during a simulated laparoscopic task to uncover how visual, motor, and neuromuscular subsystems interact. Methods: Eighteen participants (five attending surgeons, five residents, and eight novices) performed a dual-peg transfer task. Eye motion, tool motion, and surface electromyography signals (biceps, triceps, wrist flexors, and extensors) were recorded. Normalized transfer entropy ( ) quantified bidirectional information flow across eye-tool, muscle-tool, and muscle-eye pathways. Linear mixed models assessed skill-dependent trends, while t-distributed stochastic neighbor embedding (t-SNE) and K-means clustering examined participant grouping. Results: Experts demonstrated significantly greater information flow from eye-to-tool motion than tool-to-eye motion, consistent with the anticipatory gaze strategy. Conversely, novices demonstrated the opposite pattern, suggesting a greater reliance on sensory feedback during tool movement. Experts also showed that wrist-centered muscle activity conveyed more predictive information about tool motion, whereas novices demonstrated greater information flow from tool motion to elbow-centered muscle activity. No significant directional differences in information flow were observed between eye motion and muscle activity across skill levels. The linear mixed model analysis indicated that information flow in the eye-tool and wrist muscle-tool pathways was the most sensitive to skill level. The t-SNE visualization and K-means clustering provided preliminary, exploratory evidence consistent with a separation of participants by skill level. Conclusions: Multimodal analysis captures distinct, skill-dependent coordination patterns among the eye motion, tool motion, and activity of muscles of the arm, demonstrating its potential as a novel approach to surgical skill assessment in a simulated setting. These preliminary findings suggest potential for future personalized, data-driven training applications, subject to validation in larger, independent cohorts with longitudinal training outcomes before any practical application could be considered.This publication has 29 references indexed in Scilit:
- Assessment of surgical proficiency based on evaluating muscle activity, bimanual muscle coordination, and fatigue susceptibility in simulated laparoscopic tasksMedical & Biological Engineering & Computing, 2025
- Spatiotemporal characteristics of eye-hand coordination among different skill levels in laparoscopic surgerySurgical Endoscopy, 2025
- Tracking 3D motion of instruments in microsurgery: A comparative study of stereoscopic marker-based vs. deep learning method for objective analysis of surgical skillsInformatics in Medicine Unlocked, 2024
- Surgical tooltip motion metrics assessment using virtual marker: an objective approach to skill assessment for minimally invasive surgeryInternational Journal of Computer Assisted Radiology and Surgery, 2023
- A statistical framework for neuroimaging data analysis based on mutual information estimated via a gaussian copulaHuman Brain Mapping, 2016
- Effect of driving experience on anticipatory look-ahead fixations in real curve drivingAccident Analysis & Prevention, 2014
- Estimating causal interaction between prefrontal cortex and striatum by transfer entropyCognitive Neurodynamics, 2013
- Transfer entropy—a model-free measure of effective connectivity for the neurosciencesJournal of Computational Neuroscience, 2010
- A global assessment tool for evaluation of intraoperative laparoscopic skillsThe American Journal of Surgery, 2005
- Excess Length of Stay, Charges, and Mortality Attributable to Medical Injuries During HospitalizationJAMA, 2003