# WTepsan

- **Event:** [Meta Global AI Developer Hackathon](https://cerebralvalley.ai/e/meta-ai-developer-hackathon)
- **When:** Oct 26, 12:00 PM – Nov 4, 9:00 PM (EST)
- **Where:** Online
- **Team:** [Worawit Tepsan](https://cerebralvalley.ai/u/WTepsan)
- **GitHub:** https://github.com/wtepsan/WoundMeter
- **Demo video:** https://www.youtube.com/watch?v=HSMxFIOesmw
- **Gallery:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/3378

WoundMeter — AI Assists. Families Care. Professionals Decide.

As populations age, more older adults receive care at home, while healthcare professionals cannot monitor every wound daily. Family caregivers often struggle to document changes consistently, and ordinary photographs are difficult to compare because of differences in distance, angle, and lighting.

WoundMeter bridges home care and professional oversight by transforming a smartphone photo into a calibrated, clinician-verified wound record.

A caregiver photographs a wound alongside a ruler. Meta SAM 3.1 proposes the wound boundary, while deterministic algorithms calculate its physical dimensions. Muse Spark 1.3 assists with observations, explanations, and record-based questions, and Muse Voice enables speech-to-text interaction. Clinicians can correct boundaries, verify measurements, edit AI-generated notes, and approve records. Families and care teams can then follow the same wound through a shared, versioned history.

The platform also supports photographs with no visible wound, enabling zero-area records without forcing an incorrect segmentation.

Built during the hackathon (26 October–4 November 2026): I integrated Meta AI models, interactive boundary correction, colour calibration, clinician approval, audit history, connected caregiver–clinician workflows, and a mobile-installable Progressive Web App. The earlier YOLO wound detector and ruler-based measurement algorithm are explicitly identified as pre-existing components.

Inspired by my earlier elderly-care research at Chiang Mai University, WoundMeter demonstrates how AI can support accessible, collaborative home healthcare while keeping professionals in control.

WoundMeter is a research prototype, not a diagnostic or treatment tool. Its demonstration uses fictional patients and sample imagery; clinical effectiveness has not been validated.

App Demo: https://woundmeter.wtepsan.dev

## More from Meta Global AI Developer Hackathon

- [Hasrizal Husin](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/3375)
- [ClearCheckout](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/3376)
- [Emilyn](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/3377)
- [WINNERS](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/3379)
- [OvO](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/3380)
- [Alibi](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/3381)

---

Markdown version of https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/3378. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt