API Reference
Welcome to the PrismToolBox API documentation. This section provides comprehensive reference for all modules, classes, and functions in the library.
Core Modules
🔬 WSI Core (prismtoolbox.wsicore)
The core module for whole slide image handling and preprocessing.
Main Classes:
- WSI - Primary class for WSI operations, tissue detection, patch extraction, and visualization
Key Functionality: - Slide loading and multi-resolution access - Tissue contour detection and analysis - ROI (Region of Interest) management - Patch extraction with various modes - Visualization and stitching capabilities - QuPath integration for annotations
🧠WSI Embeddings (prismtoolbox.wsiemb)
Feature extraction and embedding generation from WSI patches.
Main Classes:
- SlideEmbedder - Extract embeddings from slide patches using pretrained models
- PatchEmbedder - Extract embeddings from patch datasets
- EmbeddingProcessor - Process, analyze, and visualize embeddings
Key Functionality: - Model-based embeddings (ResNet, Vision Transformers, Foundation models) - Stain-based feature extraction (color deconvolution) - Cell-based feature extraction (morphological features) - Dimensionality reduction and clustering - Visualization and analysis tools
🔬 Nuclei Segmentation (prismtoolbox.nucleiseg)
Deep learning-based nuclei segmentation and analysis.
Main Classes:
- NucleiSegmenter - Segment nuclei in WSI patches using deep learning models
Key Functionality: - Multiple segmentation models (SOP, custom models) - Batch processing of slide patches - Post-processing and conflict resolution - QuPath export for visualization