Deep Learning-Based Cancer Detection Technique

Abstract

The time is now for deep learning (DL)-dependent analysis of healthcare images to move from the realm of exploratory research projects to that of translational ones, and eventually into clinical practise. This process has been sped up by developments in data availability, DL methods, and computer power over the last decade. As a result of this experience, the authors now know more about the potential benefits and drawbacks of incorporating DL into clinical treatment, two factors that, in the authors' opinion, will propel progress in this area over the next several years. The most significant of these difficulties are the widespread need of strength of commonly utilized DL training approaches to various pervasive pathological properties of healthcare images and storages, the need of an properly digitised environment in hospitals, and the need of sufficient open datasets on which DL approaches may be trained and tested.