Real – Time Detection of Throat Cancer Using Deep Learning Algorithm

Abstract- Early detection of throat cancer is basic for effective treatment and higher survival rates. Be that as itmay, conventional symptomatic strategies like imaging and biopsies take a lot of time and got to be interpreted byexperts, which as often as possible causes delays in determination. This extend proposes a real-time throat cancerdiscovery framework that employments equipment integration and profound learning to overcome these obstacles. Theframeworkclassifies restorative images as generous orharmful with high accuracy by utilizing pre-trained deep learningmodels such as VGG16, ResNet50, and RCNN, in conjunction with advanced picture handling strategies. The proposedarrangement incorporates equipment components such as a GSM module, LCD display, and Arduino microcontroller.This configuration sends SMS alarms to medical experts in cases of malignancy and gives instant feedback on thediagnosis through the LCD screen. In order to ensure accurate feature extraction for classification, the system is madeto process endoscopic or X-ray images utilizing MATLAB for pre-processing and segmentation. This system provides aworkable, effective, and scalable solution for early throat cancer detection by combining state-of-the-art profoundlearning algorithms with real-time equipment capabilities. Its ability to provide immediate results and alert medicalprofessionals ensures provoke mediation, enormously progressing persistent care and cutting down on treatmentdelays.