Abstract
<h3>Objectives</h3> Oral squamous cell carcinoma is a major challenge for all health professionals, particularly for general dentists. These challenges include the management of this malignancy and the early diagnosis to minimize sequelae and mortality. There are several methods for early detection for squamous cell carcinoma. The objective of this research is the development of an app capable of assisting the dental surgeon in the early diagnosis of oral cancer. <h3>Study Design</h3> This system was based on convolutional neural networks. Initially, a survey of the database of images of potentially malignant lesions and squamous cell carcinomas was performed. In the selected images, the marking occurred to delimit the lesions; therefore, the system was trained from the markings. <h3>Results</h3> In total, 493 cases were marked: 248 squamous cell carcinomas and 245 potentially malignant lesions. This app will identify the lesions, marking squamous cell carcinomas in red, potentially malignant oral lesions in green, and normal mucosa in blue. <h3>Conclusions</h3> The system developed should be used in collaboration with health agents, mainly dental surgeons, to guide the clinical diagnosis of these injuries.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have