PathTrove
PathTrove curates 40+ clinically important metadata fields, data modality maps, label organization, and download guidance to serve real clinical research needs
0+
Dataset records
0
Organ categories
0
Task labels
0
Task types
PathTrove maintains a broad index of computational pathology datasets across organs, stains, modalities, tasks, and access conditions.
Records capture stain, scanner, rare-disease status, clinical metadata, cohort notes, data volume, modality, label type, and QC context.
Download guidance, data-volume breakdowns, modality maps, annotation sources, and label organization are collected whenever they can be verified.
The open query agent recommends suitable datasets and dataset combinations from curated PathTrove records.
PathTrove turns a concrete clinical research need into dataset recommendation, metadata inspection, and resource delivery.
Recommended usage
The online search page is a public demo. Queries may take 1-2 minutes and can fail because of server load, Hugging Face Space availability, upstream API limits, or network issues. For stable use, install the open-source Claude Code plugin and query the curated PathTrove snapshot locally.
Claude Code plugin
/plugin marketplace add DearCaat/PathTrove
/plugin install cpath-dataset-agent@pathtrove
/reload-plugins
/cpath-dataset-agent:cpath-query public H&E lung segmentation datasetsOnline demo latency depends on question complexity and backend availability.
Local Claude Code usage is the recommended path for real research work.
The plugin includes the curated dataset index, reports, and retrieval workflow.
Dataset entries keep source papers, official pages, repositories, licenses, and provenance notes together for record-level checking.
Stain, scanner, rare disease, clinical metadata, cohort diversity, demographics, and QC fields help users avoid weak dataset matches.
Data scale, valid image count, modality map, label structure, annotation source, and download instructions are recorded when evidence is available.
Natural-language search helps assemble candidate datasets for training, validation, external testing, or survey-level comparison.