Case study: Pathology Explorer
Automatically detect cell types and tissue structures from histology data
Click to enlarge
5%
Pathology Explorer outperforms SOTA with 5% better detection and 24% higher F1 classification, using 5x fewer parameters.
Converse directly with histological data
Pathology Explorer is a state-of-the-art AI agent for pan-cancer nuclei detection, segmentation and classification in H&E-stained pathology images.
- Characterize patient populations from tissue images
- Understand spatial composition of TME
- Identify biomarkers predictive of response
Development and performance
- We developed a new ML system to segment and classify cell & tissues. The pre-print available on arxiv at **https://arxiv.org/abs/2508.09926**
- Using our methodology, we built Pathology Explorer, an agent that unlocks robust detection, segmentation & classification of multiple cell types, including understudied immune populations such as neutrophils and eosinophils.
- Pathology Explorer outperforms SOTA with 5% better detection and 24% higher F1 classification, using 5x fewer parameters.
- Pathology Explorer generalizes across multiple cancer types.
Predicted cell types and structures
Pathology Explorer automatically detects multiple cell types, including:
- Lymphocytes
- Neutrophils
- Plasmocytes
- Fibroblasts
- Eosinophils
- Cancer cells
Preprint and repositories
Read the preprint at: https://arxiv.org/abs/2508.09926
The model weights and an open implementation are available.
→ GitHub repository: https://github.com/owkin/histoplus
→ Model weights: https://huggingface.co/owkin/histoplus
Share to:
No items found.