Evaluation of Surgical Skills during Robotic Surgery by Deep Learning-Based Multiple Surgical Instrument Tracking in Training and Actual Operations
Problem
Surgeons' skills in robotic surgery are hard to evaluate automatically and quantitatively because they depend on instrument movement.
Why it matters
Objective skill evaluation is increasingly important for training and patient safety as robotic surgery volumes grow.
What the AI did
A deep learning-based multi-instrument tracking algorithm handled occlusion and identity maintenance, and motion metrics from tracking fed surgical skill prediction models.
Results
RMSE 3.52 mm; AUC 0.7, 0.78, 0.86 at 1 mm, 2 mm, 5 mm thresholds; tested on 54 video segments.
Impact
Demonstrated feasible automatic, quantitative skill evaluation in robotic surgery on 54 video segments.
Record
| Venue | Journal of Clinical Medicine |
|---|---|
| Published | 2020 |
| Work type | article |
| Authors | Dongheon Lee, Hyeong Won Yu, Hyungju Kwon, Hyoun‐Joong Kong, Kyu Eun Lee, Hee Chan Kim |
| Citations | 107 (OpenAlex, at import time) |
| References | 64 |
| Model or method | not stated in source |
| Organization | not stated in source |
| Primary source | https://doi.org/10.3390/jcm9061964 |
| Open access | gold |
| Keywords | robotic surgery, instrument tracking, deep learning, skill assessment, motion metrics |
Importance rating
real but narrower advance
Objective robotic-surgery skill evaluation aids training and patient safety; abstract notes growing procedure volumes but gives no scale; narrow advance.