Development of Medical Image Analytics by Deep Learning Model for Prediction and Classification of CT Image Diseases

The CT images of Lung illnesses or diseases that damage the lungs and weaken the respiratory system. Lung cancer is one of the topmost causes of death in humans around the world. Humans have a better chance of surviving if they are detected early. The average survival rate of persons with lung cancer increases from 14 to 49 percent if the disease is detected early. While computed tomography (CT) is significantly more effective than X-ray, a complete diagnosis requires a combination of imaging techniques that complement each other. But, because there are multiple phases of cancer that develop into different types of tumors with varying sizes and risks, finding lung cancer does not predict the risk of cancer. A deep neural network is constructed and tested for detecting lung cancer CT images. This research work analyses different types of tumor sizes such as large cell carcinoma, normal, squamous cell carcinoma, and adenocarcinoma. Also, the lung tumors are detected and predicted with the help of computer vision methods such as Residual neural network (ResNet), Convolutional neural network (CNN). Finally, the results of all the methods are compared and various parameters were calculated. Thus, the proposed method (ResNet) gives an optimal solution on comparison with respect to all the parameters.