[Dataset] 150k+ Annotated Stool Images — Available For Research/commercial Licensing

I’ve built what I believe is the largest annotated stool image dataset in existence (~150k+ photos) and I’m exploring whether to license it for research or commercial use. Posting here to gauge interest and get feedback before I decide how to distribute.

What’s in it

• Size: ~150,000 images (and growing)
• Source: user submissions via {{iOS/Android consumer app, real-world in-toilet photos}}
• Resolution: {{typical resolution range, e.g. 1024×1024 up to 4032×3024}}
• Diversity: {{geographic spread, device/camera variation, lighting conditions, toilet/water conditions}}

Annotations (per image)

• Bristol Stool Scale (type 1–7)
• {{color, consistency, volume estimate, blood/mucus flags — list whatever you actually have}}
• {{any free-text notes, symptoms, or linked user-reported metadata like diet, hydration, medications}}
• Annotator: {{self-reported by user / reviewed by clinician / AI-assisted + human verified — be honest}}
• {{Inter-rater agreement or QA process, if any}}

Provenance & compliance

• Collected under {{Privacy Policy / ToS URL}} with explicit user consent for {{research use / model training}}
• {{PII stripped: no faces, no identifying EXIF, no filenames containing user IDs}}
• {{HIPAA status — usually not HIPAA since it’s a consumer app, not a covered entity, but state it clearly}}
• {{GDPR: EU users’ data handled per … / excluded / anonymized}}
• Not sourced from clinical/hospital settings — this is consumer-generated, in-the-wild data

What it’s useful for

• Training classifiers for Bristol scale, blood detection, abnormality flags
• Gut health / GI apps, telehealth triage, IBD/IBS monitoring research
• Benchmarking medical vision models on messy, non-clinical imagery

Licensing

• Open to: {{non-exclusive research license / exclusive commercial license / per-sample pricing / academic free + commercial paid}}
• Can provide a {{small sample pack, e.g. 500 images}} under NDA for evaluation

DM or comment if interested — happy to answer questions about the schema, provide sample images, or discuss licensing terms.