This documentation provides comprehensive information about the Deep Phenotyping framework, which uses contrastive learning for identifying and stratifying single cells in whole slide immunofluorescence microscopy images derived from liquid biopsies. The framework achieves high accuracy in classifying diverse cell phenotypes and enables automated identification of rare cell populations.
Core Modules
Representation Learning
Core implementation of the contrastive learning framework for feature extraction from cell images.
Leukocyte Classifier
Classification models for leukocytes and other cell types in liquid biopsies.
Utilities
Common utilities for data processing, normalization, and image analysis.
Pipeline Components
Pipeline
End-to-end processing pipeline for whole slide images, from preprocessing to analysis.