Phone-based AI screening for mouth and throat lesions in low-resource areas
Mobile phone-based deep learning algorithm for oral lesion screening in low-resource settings
['FUNDING_OTHER'] · SLOAN-KETTERING INST CAN RESEARCH · NIH-11309875
A low-cost phone camera paired with artificial intelligence will help health workers in low-resource communities spot mouth and throat lesions that might be cancer so patients can be referred sooner.
Quick facts
| Phase | ['FUNDING_OTHER'] |
|---|---|
| Study type | Nih_funding |
| Sex | All |
| Sponsor | SLOAN-KETTERING INST CAN RESEARCH (nih funded) |
| Locations | 1 site (NEW YORK, UNITED STATES) |
| Trial ID | NIH-11309875 on ClinicalTrials.gov |
What this research studies
If this comes to your clinic, a health worker will use a specially equipped mobile phone that captures polarized light and autofluorescence images of the mouth. An on-device deep learning algorithm will read the images and flag lesions that look potentially dangerous, guiding who should be sent to a specialist. The team will test the device and AI in collaboration between U.S. centers and a partner hospital in Mumbai to ensure it works without internet access. Results will be compared to expert readings and clinical follow-up to see how well the tool helps with triage in real-world low-resource settings.
Who could benefit from this research
Good fit: People with visible mouth or throat lesions or who are at high risk for oral cancer presenting at participating clinics in low-resource settings would be the ideal candidates.
Not a fit: Patients whose problems are not visible in the oral cavity or who already have access to immediate specialist care and diagnostic biopsy may not gain direct benefit from this screening tool.
Why it matters
Potential benefit: Could lead to earlier detection and faster referral of potentially cancerous oral lesions in places that lack specialists.
How similar studies have performed: Preliminary work by the team showed hardware feasibility and an earlier cloud-based algorithm reached about 79% sensitivity and 82% specificity, but on-device, offline validation is still needed.
Where this research is happening
NEW YORK, UNITED STATES
- SLOAN-KETTERING INST CAN RESEARCH — NEW YORK, UNITED STATES (ACTIVE)
Researchers
- Principal investigator: RAJADHYAKSHA, MILIND — SLOAN-KETTERING INST CAN RESEARCH
- Study coordinator: RAJADHYAKSHA, MILIND
About this research
- This is an active NIH-funded research project — typically early-stage science, not a clinical trial accepting patient enrollment.
- Some NIH-funded labs run parallel clinical studies or seek volunteers for related work. To check, contact the principal investigator or institution listed above.
- For full project details, budget, and progress reports, visit the official NIH RePORTER page below.