YOLOv8 pipeline for glomeruli detection in renal whole-slide images.

The workflow segments tissue, extracts patches, detects glomeruli, and reconstructs predictions back to slide coordinates. It is designed as an upstream step for later glomerular analysis: detected regions can become inputs for classification, feature extraction, or WSI-level aggregation.

The project shows the practical engineering around medical object detection: converting high-resolution WSI data into trainable patches, keeping annotation coordinates consistent, and producing outputs that can be traced back to the original slide.

---

← Back to projects