Introduction: Oral cancer remains a major public health challenge in low- and middle-income countries (LMICs), where high incidence, late-stage diagnosis, and limited access to specialized care contribute to elevated morbidity and mortality. Although visual oral inspection is a potentially effective screening strategy, its reliance on highly trained professionals limits scalability within public health systems. Portable, low-cost mHealth-based technologies offer an opportunity to expand access to early detection among vulnerable populations. Objectives: To evaluate the feasibility, usability, diagnostic performance, and capacity-building potential of mDOC (mobile Detection of Oral Cancer), a portable imaging-based technology integrated with artificial intelligence, implemented within the oral cancer screening program of Brazil’s Unified Health System (SUS). Methods: This multicenter, prospective study is conducted within the largest oral cancer screening program in Brazil, operated through a mobile unit serving 18 cities in the state of São Paulo. The project includes: (1) a pilot study with 50 participants to assess operational feasibility; (2) a formal usability evaluation of the mDOC among different categories of healthcare professionals, following international standards; (3) a field study involving 1,000 high-risk individuals (≥35 years, tobacco and/or alcohol users), comparing the diagnostic performance of mDOC, using automated and expert image analysis, with conventional clinical examination and histopathology; and (4) the development and implementation of a low-cost, hands-on training model to support task shifting in primary care. The study involves international collaboration with U.S. based academic institutions responsible for technological development, algorithm optimization, and scientific validation, while Brazilian partners lead field implementation, workforce training, and integration into the SUS, strengthening local capacity. Results: The mDOC is expected to demonstrate high sensitivity for the detection of oral cancer and oral potentially malignant disorders, with diagnostic performance comparable or superior to conventional clinical examination. Automated image analysis is anticipated to reduce diagnostic subjectivity and improve decision support for non-specialist health professionals. High usability and acceptability across different professional categories are expected, supporting scalability in resource-limited settings. The training model is designed to enhance local competencies and promote sustainable implementation. Conclusions: The integration of mDOC into oral cancer screening within the SUS represents a scalable, low-cost innovation for LMICs, combining mobile imaging, artificial intelligence, and workforce capacity building. This approach has the potential to reduce inequities in access to early diagnosis, strengthen cancer prevention strategies, and serve as a replicable model for other countries facing similar health system challenges.

Screenshots of the customized mDOC app for Android that guides healthcare professionals through the data collection process