Skip to Main Content

Lourdus

Built at Meta Global AI Developer Hackathon · Oct 26, 2026 · Remote

Lourdus — Demo video

Oral cancer is a massive health challenge in India, driven largely by widespread tobacco use. Because specialists are usually located in distant cities, frontline village health workers are the first to see patients with early visible mouth lesions. But without a standardized system, these remote referrals often get lost, or they arrive without clear photos and clinical notes, making it hard for dentists to know who needs to travel to a clinic. OncoLens solves this by providing an offline workflow for photo documentation and referral routing. In the field, a health worker captures a photo of the mouth and completes a short questionnaire. A fixed clinical protocol then assigns a referral priority based entirely on the answers. Once the device syncs, Meta models process the case to give the remote reviewing dentist clean and structured data. Muse Spark acts as a quality gate to catch unusable uploads like severe glare or non mouth photos. SAM 3.1 filters out teeth and lets the dentist precisely outline a lesion with a single click. The dentist reviews the case, signs off, and manages follow ups through a dashboard. OncoLens exports printable slips and FHIR R4 bundles, ensuring every patient is clearly documented and the referral never gets lost.

Team