Deep Learning for Oral Cancer Screening and Referral: DL-CANSCREEN, an artificial intelligence tool
Next Project
The overall aim of this project was to investigate the feasibility of different deep learning (DL) multiclass digital image classifiers to classify digital photographic pictures of the oral cavity into oral cancer and precancer lesions by differentiating them from healthy mucosal variants and common benign lesions. We aimed to identify the best fitting model with several systematic experiments, to classify individuals in the ‘healthy-to-diseased’ spectrum that may further help in referral pathways and may serve as an automated care linkage tool, in future, for patient care continuum. It could serve as a decision support system for primary care clinicians at clinical set-ups, an adjunct tool for community health workers for screening & referral and an automated linkage mechanism for patients leading to care continuum. This project was funded by Google – AI for Social Good, India (Google-AISG, 2021-2022) with Carnegie Mellon University, Pittsburgh, USA, as the academic partner.
