Laborator's picture
Fix bold-bullet patterns and add first person voice
eade72c verified
---
license: apache-2.0
task_categories:
- image-classification
- image-segmentation
- object-detection
tags:
- microscopy
- bioimage
- histopathology
- cell-biology
- materials-science
- dataset-index
- meta-dataset
size_categories:
- n<1K
pretty_name: Microscopy Datasets Index
extra_gated_prompt: >-
This dataset is in validation phase.
Access is granted to verified researchers and organizations.
Please describe your intended use case.
extra_gated_fields:
Full name: text
Organization: text
Intended use: text
I agree to use this data for research purposes only:
type: checkbox
---
# Microscopy Datasets Index
I kept losing track of which microscopy datasets exist and what format they're in. So I made an index. 300+ open datasets, searchable by domain, task, microscopy type, and license.
## What's in here
A single reference file (JSONL and Parquet) cataloging 309 publicly available microscopy datasets. Each entry includes:
- id — short slug
- name — human-readable dataset name
- source — who published it (Broad Institute, Kaggle, ISBI, etc.)
- url — direct link to the dataset page
- domain — cell_biology, histopathology, materials_science, geology, parasitology, environmental, quality_control, botany, entomology
- task — instance_segmentation, image_classification, cell_tracking, object_detection, semantic_segmentation, etc.
- microscopy_type — fluorescence, brightfield, electron, phase_contrast, confocal, polarized_light, etc.
- image_count — approximate number of images
- license — CC0, CC-BY-4.0, Apache-2.0, etc.
- year — publication year
- notes — one-liner description
## Coverage by domain
| Domain | Count |
|---|---|
| Cell Biology | ~110 |
| Histopathology | ~60 |
| Materials Science | ~25 |
| Quality Control | ~25 |
| Parasitology | ~20 |
| Geology | ~20 |
| Environmental | ~20 |
| Botany | ~15 |
| Entomology | ~15 |
## What's included
- The full BBBC collection (Broad Bioimage Benchmark Collection, BBBC001-BBBC048)
- Cell Tracking Challenge datasets (15 sequences)
- Kaggle microscopy competitions (DSB 2018, Sartorius, HuBMAP, etc.)
- ISBI challenge datasets
- Human Protein Atlas collections
- Allen Cell Explorer datasets
- Recursion RxRx1/RxRx2/RxRx3
- JUMP-Cell Painting consortium
- TCGA histopathology (lung, breast, kidney, colon, prostate)
- Grand Challenge competitions (Camelyon, MoNuSeg, PanNuke, TIGER, MIDOG, etc.)
- Materials science: steel defects, metallography, SEM/EBSD, composites, additive manufacturing
- Geology: thin sections, digital rocks, micro-CT, petrology
- Parasitology: malaria, helminth eggs, trypanosomes, schistosoma
- Environmental: plankton, diatoms, microplastics, pollen
- Industrial QC: MVTec AD, steel surface defects, wafer maps, PCB defects
- Electron microscopy: CREMI, SNEMI3D, FIB-SEM, EMPIAR cryo-EM
## Quick start
```python
import pandas as pd
df = pd.read_parquet("data/microscopy_index.parquet")
# All fluorescence segmentation datasets
fluoro_seg = df[(df.microscopy_type == "fluorescence") & (df.task.str.contains("segmentation"))]
# Datasets with permissive licenses
open_data = df[df.license.isin(["CC0", "CC-BY-4.0", "Apache-2.0", "public-domain"])]
# Materials science datasets with 500+ images
big_materials = df[(df.domain == "materials_science") & (df.image_count >= 500)]
```
## This is a reference index, not the actual data
The files here are a catalog. The URLs point to where the actual datasets live. I did not scrape or redistribute any images. If a URL is dead, open an issue and I'll update it.
## Format
- `data/microscopy_index.jsonl` — one JSON object per line
- `data/microscopy_index.parquet` — same data, columnar format
## Contributing
Know a microscopy dataset that's missing? Open a PR or issue with the dataset name, URL, and domain. I'll add it.