Detection of Lung Cancer Using CT-Scan Image - Deep Learning Approach

Abstract

Cancer is a disease that comes in many forms and is the largest cause of death worldwide for men and women alike. Early detection of cancer has the highest chance of saving a person's life. Some of the procedures used to diagnose cancer include CT scans, bone scans, MRIs, PET (Positron Emission Tomography), ultrasound, and X-rays. Cancers such as lung cancer are among the deadliest worldwide, killing approximately five million people every year. This chapter focuses on lung cancer detection. The diagnosis of Cancer is usually a very difficult task in the biomedical and the bioinformatics field. Now, computed tomography (CT) scans can provide useful information for lung cancer diagnosis. In recent advances, deep learning approaches have improved to outperform humans in some tasks like classifying objects in images and also predicting better accuracy. Therefore, these techniques have been utilized in this model for the treatment of cancerous conditions. We detect lung cancer nodules from a given input and classify cancer as Adenocarcinoma, Large Cell Carcinoma, or Squamous Cell Carcinoma in our research. To detect the location of lung nodules, researchers used revolutionary deep learning approaches. In this paper basically, we used three deep learning case studies to diagnose lung cancer such as VGG16, INCEPTIONV3 and RESNET50 and also, we are discussing various measures for evaluating the performance of our model to get better accuracy. SS