Accepted for/Published in: JMIR mHealth and uHealth
Date Submitted: Sep 8, 2025
Date Accepted: Feb 9, 2026
A Community-Based Usability Study of an AI-Enabled Oral Cancer Screening App Operated by Village Health Volunteers
ABSTRACT
Background:
Oral cancer is a major public health concern in low- and middle-income countries, where access to specialist care and early detection remains limited. Mobile health (mHealth) technologies supported by artificial intelligence (AI) offer a scalable approach to extend screening services into underserved communities. In Thailand, village health volunteers (VHVs) are key frontline workers who provide preventive services and bridge gaps between rural populations and specialist care.
Objective:
This study aimed to describe the technical development of RiskOCA, a smartphone-based, AI-assisted oral cancer screening platform, and to evaluate its usability when deployed by VHVs in a rural Thai province.
Methods:
RiskOCA was developed using a three-tier architecture comprising a patient-facing interface for risk factor profiling and guided imaging, an embedded deep learning engine (DeepLab v3+ with ResNet-50 backbone) for lesion detection, classification, and segmentation, and a secure specialist portal for expert review of all cases. The AI model was trained on 2,591 annotated intraoral images and validated for real-world use. Field testing was conducted in Phu Kamyao District, Phayao Province, where 1,242 adults (≥40 years) were screened with assistance from VHVs. Usability was evaluated through a structured 25-item questionnaire completed by 250 VHVs, with responses rated on a 5-point Likert scale.
Results:
The AI model achieved a mean classification accuracy of 87.6% across three diagnostic categories. Usability evaluation indicated high satisfaction across all domains, with an overall mean score of 4.17 out of 5. The highest ratings were for the app’s impact on elderly surveillance (M = 4.30), while all domains were rated “satisfied” or “very satisfied.”
Conclusions:
RiskOCA demonstrated strong technical performance and high user acceptance among VHVs, supporting its feasibility for community-based oral cancer screening. By integrating AI-assisted triage with expert review, the platform has potential to reduce diagnostic delays, expand screening coverage, and serve as a scalable model for oral cancer prevention in resource-limited settings.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.