Detection of Malignant Lung Tumors Using Convolutional Neural Networks

Cancer is amongst the leading reasons of untimely demise in human beings. Detection of nodules present in lung in earlier stages of cancer can help in bringing down the mortality rate of this disease. Hence, the most pressing priority in the field of cancer treatment is the need of new methodologies that can aid in detecting the disease in its earlier stages. In this project, convolutional neural network is put to use for the purpose of detecting cancerous nodules in the lungs. Using this system, not only lung cancer can be detected in a person but also the chances of a person acquiring lung cancer can be estimated. In every phase of this system, the given image is enhanced and segmented separately. For the purpose of image enhancement, the image is resized, contrasts are enhanced and the colour space is converted from one form to another. This step is followed by segmentation and classification. For the purpose of classification, a binary CNN classifier is deployed. This methodology can provide greater precision for both diagnostic and predictive purposes