Object Detection in Cataract Surgical Images

Iris and Pupil Detection in Cataract Surgical Data

This project focused on detecting the iris and pupil in cataract surgery images using deep learning techniques. The goal was to enable better intraoperative monitoring and support enhanced post-operative outcomes by precisely identifying these ocular features in real-time.

Description:
Implemented YOLOv5 object detection to identify iris and pupil in cataract surgery images for monitoring intraoperative deviations and optimizing post-operative outcomes.

Tools & Technologies:
YOLOv5 Object Detection (You Only Look Once)

Outcome:
Achieved a Mean Average Precision (mAP) range of 0.75–0.80 across multiple datasets.

GitHub Repository:
YOLO Object Detection

Sample Images