← Back to Projects Healthcare 
Medical Image Segmentation Tool
U-Net architecture in TensorFlow for medical image segmentation with data augmentation, MC Dropout uncertainty, DICOM handling, and Grad-CAM visualization.
Overview
Deep learning pipeline for medical image segmentation using U-Net architecture with production deployment via ONNX export.
Architecture
- U-Net encoder-decoder with skip connections
- MC Dropout for uncertainty quantification
- DICOM preprocessing and augmentation pipeline
- Grad-CAM visualization for model interpretability
- ONNX export for deployment
Key Features
- Multi-class segmentation with confidence maps
- Active learning with uncertainty sampling
- DICOM/NIfTI format support
- Interactive Streamlit visualization tool