| Item Type: | Dataset | 
|---|---|
| Title: | PathoCell: a benchmark for foundational models in computational pathology | 
| Creators Name: | Reith, F.H. | 
| Abstract: | This repository contains the datasets for PathoCellBench (formerly PhenoBench), a comprehensive benchmark designed to evaluate the cell phenotyping capabilities of pathology Foundation Models. The benchmark suite includes four key datasets, all processed into the efficient LMDB (Lightning Memory-Mapped Database) format to facilitate large-scale experimentation. DATASETS INCLUDED: This collection consolidates the following datasets: PathoCell (formerly PhenoCell): Our new, challenging benchmark dataset for H&E cell phenotyping with 14 granular cell types.; PanNuke: A pan-cancer nucleus segmentation and classification dataset. Note: This dataset requires a one-time setup step after downloading (see instructions below).; Lizard: A large-scale nucleus segmentation and classification dataset in colon histology.; ARCTIQUE: A colorectal cancer histology dataset for tile-based classification. | 
| Keywords: | Pathology, Histology, Cell Phenotyping, Foundation Models, Benchmark | 
| Source: | Hugging Face | 
| Publisher: | GitHub | 
| Date: | September 2025 | 
| External Fulltext: | View full text on external repository or document server | 
| Related to: | 
Repository Staff Only: item control page


![[feed]](/style/images/feed-icon-14x14.png)
 Tools
 Tools Tools
 Tools
