THE ROLE OF VISUAL SENSOR-BASED CT IMAGING FOR RAPID DIAGNOSIS OF LUNG CANCER MARKERS IN ATHLETIC PATIENTS

Abstract

Lung cancer has the greatest fatality rate, requiring a biopsy to confirm its subtype before therapy can begin. Deep learning has recently developed significant tools for lung cancer diagnosis and treatment planning in athletic patients. In any case, diagnosing the neurotic sort of cellular breakdown in the lungs in its beginning phases utilizing CT pictures stays troublesome because of an absence of accessible preparing information and incredible counterfeit shrewd models. The wide range of diagnoses makes a precise prediction and subtype classification of lung cancer in athletic patients critical. Pathologists now use Computer Aided Diagnosis (CAD) to make accurate diagnoses. Our research shows how to use a pre-trained Deep Convolutional Neural Networks (DCNN) architecture called Visual Geometry Group – 16 (VGG16) to detect, predict, and categorize ovarian cancer in athletic patient’s subtypes from histopathology pictures. VGG-16 comprises of 16 layers (13 convolution layers, three completely associated layers, five max-pooling layers, and one softmax layer). For this examination, we obtained 1000 CT filter pictures of lungs in four gatherings (adenocarcinoma, huge cell carcinoma, squamous cell carcinoma and ordinary). By using just 613 CT scan images our model can detect and classify lung cancer with having an accuracy of 77.61% in less than 2 seconds which is higher the accuracy of manual detection as the dataset increases more accurate our model became which shows us a possible way for more accurate and fast cancer detection.