Prostate Cancer: Cancer Detection and Classification Using Deep Learning

Abstract

Cancer has already taken the second leading position as the cause of death making the global burden world wide. There is an estimate of 9.6 million deaths in 2018 as per the survey of US cancer statistics. Prostate cancer is an abnormal, malignant growth of cells in the prostate gland uncontrollably. This disease mainly affects either middle-aged or older thousands of men each year. Cancer till now is a major issue that is continuously creating a burden to the individuals as well as health systems. Therefore, it is a big concern in front of health systems demanding a strong system which can detect early and help for treatment. Here, we have applied some deep learning models to study the classification for the detection of clinically significant prostate cancer from the large-scale cancer image data for each patient. This paper investigates the effect of deep learning architecture like VGG16, EfficientNet, Dense Net121, ResNext50 in the large-scale cancer image data classification setting. Our main contribution is to focus on the high-level accuracy because these deep learning algorithms have the capability of transfer learning with image instant segmentation.