Changelog¶
All notable changes to this project will be documented in this file.
[0.1.0] - 2025-01-29¶
Added¶
Patch extraction from whole slide images
Multiple cell segmentation models: - CellViT: Vision transformer-based nuclear instance segmentation - HoverNet: Nuclear instance segmentation - Cellpose: Deep learning cell instance segmentation
Feature extraction capabilities: - PyRadiomics_Gray: Comprehensive radiomics features - PyRadiomics_HED: Comprehensive radiomics features - Morphometrics: Morphological features
Multiple Instance Learning models: - CLAM: Clustering-constrained attention MIL - TransMIL: Transformer-based MIL with PPEG - Standard MIL: Max pooling MIL - HistoBistro: Transformer-based MIL
Command-line interface for all operations
Feature visualization tools
Graph creation tools
Dataset creation and management
Model training and evaluation pipelines
Features¶
Support for multiple WSI formats (SVS, TIFF, NDPI, etc.)
GPU acceleration for deep learning models
Configurable patch extraction parameters
Comprehensive logging and error handling
Type-safe configuration with Pydantic