Quantitative Outcome-Oriented Assessment of Microsurgical Anastomosis
This repository contains 50 annotated microsurgical anastomosis images and the code for automated surgical skill assessment using Mask R-CNN.
- Dataset: 50 annotated images with COCO-format annotations.
- Training: Mask R-CNN training with 5-fold cross-validation.
- Inference: Automated segmentation of sutures and blood vessels.
- ALI Assessment: Detection and quantification of five surgical error categories.
- Visualization: Segmentation results and surgical error visualizations.
- Model Weights: Pretrained model checkpoints for inference.
Hu, L., Gholami, S., Dindelegan, G., Meling, T.R. & Billard, A. (2025, July). Quantitative Outcome-Oriented Assessment of Microsurgical Anastomosis. In 2025 IEEE Engineering in Medicine and Biology Society (EMBC). IEEE. [link]
Relating Surgeon Proficiency Scores to Maturation of Visuomotor Skill during Training
This dataset contains de-identified longitudinal microsurgical skill assessment data from 22 participants evaluated across two sessions, together with statistical analysis scripts and visualization results.
- Participants: 11 students, 10 residents, and 1 surgeon.
- ALI Scores: Counts of 10 surgical error types and total ALI scores for both sessions.
- Participant Information: Age, sex, surgical experience, microscope experience, and training history.
- Statistical Analysis: Within-participant comparisons, expertise-group comparisons, and significance tests.
- Visualizations: Surgical error examples, classification diagrams, and statistical figures.
- Code: Python scripts for reproducing the statistical analyses and figures.