Project: Support for Early Detection of Oral Cancer Through Artificial Intelligence
Function: Late diagnosis of oral cancer is associated with worse prognoses and higher morbidity and mortality, making early identification of suspicious lesions a challenge, especially in areas with limited acc... Late diagnosis of oral cancer is associated with worse prognoses and higher morbidity and mortality, making early identification of suspicious lesions a challenge, especially in areas with limited access to specialized professionals. To contribute to addressing this problem, we developed an Artificial Intelligence-based technology capable of analyzing images of the oral cavity and automatically identifying visual patterns associated with potentially malignant lesions. The solution uses Computer Vision and Convolutional Neural Network (CNN) techniques to classify images and support initial screening, without replacing professional evaluation. Preliminary results obtained from public databases indicate an accuracy of around 80%, demonstrating the viability of the approach. The technology aims to offer an accessible decision support tool with the potential to expand screening and promote the early referral of suspicious cases.
Target Product: Health professionals and primary and dental care services that perform screening for oral lesions.
