Abstract Oral cancer is the sixth most common malignancy worldwide. In the United States, 53,000 new cases of oral and oropharyngeal cancer are diagnosed annually. With early detection and treatment, patients with oral cancer can have excellent outcomes. However, most patients are not diagnosed until their disease is at a late stage when treatment is more invasive, more expensive, and less effective. Improving early detection of oral cancer and its precursors represents the best opportunity to reduce the incidence, morbidity, and mortality of oral cancer. In the US, rural areas are especially likely to lack effective programs for early detection of oral cancer. Early diagnosis of oral cancer in rural settings is frequently hampered by a lack of personnel with appropriate expertise, lack of health care infrastructure, limited access to health services, and long travel distances. Patients in rural counties experience longer delays and travel greater distances for diagnosis and treatment by a specialist, compared to patients in urban areas. In this project we will develop two new tools to improve early detection of oral cancer in rural areas of the US: (1) a low-cost, robust, mobile phone-based imaging system for mobile Detection of Oral Cancer (mDOC), and (2) a low-cost training model to aid in teaching oral cancer examination procedures, including the use of mDOC. These tools will provide the means to improve detection of oral cancer in rural areas through autofluorescence and white-light imaging technology, objective automated image analysis, and expert review by off-site dentists or doctors. In Aim 1 we will develop the mDOC instrument and the interactive oral exam training model. In Aim 2 we will use the mDOC device to image 120 patients referred to an oral specialist for evaluation of suspicious oral lesions. We will use this data set to develop and validate mDOC automated analysis algorithms to objectively identify high risk oral mucosal lesions. In parallel, we will conduct a usability study to evaluate and optimize the oral exam training model and the mDOC device. In Aim 3, we will use mDOC in a pilot study to image 50 patients seeking care at dental and primary care clinics in the Rio Grande Valley of south Texas, to evaluate the feasibility of oral cancer exam using mDOC in rural and underserved healthcare settings; and we will train local healthcare providers to perform oral exams using the training model and the mDOC device. Our partnership combines expertise in biomedical imaging with clinical expertise in detection, diagnosis, and treatment of oral cancer in diverse populations. The innovative mobile imaging technology to be developed in this proposal can enable scale-up of effective programs for early detection of oral cancer in rural and medically underserved regions of the United States.

Project Narrative Oral cancer is the sixth most common malignancy worldwide, and 53,000 new cases of oral cavity and oropharyngeal cancer are diagnosed in the United States each year. Primary care physicians and dentists are unprepared to recognize and distinguish premalignant and cancerous lesions from benign conditions, and patients in rural counties of the US experience a 16.8 month delay and travel twice the distance to have their oral lesions accurately diagnosed and treated by an oral health specialist. In this project we will overcome these barriers by developing a mobile phone-based imaging system for mobile Detection of Oral Cancer (mDOC) and an oral examination training model that will enable less experienced providers to carry out highly effective programs for early detection of oral cancer in rural settings.