Early Esophageal Cancer detection using Deep learning Techniques. (Review Article)

Esophageal cancer is one of the deadliest diseases for humans, since it is discovered in very advanced stages. As result, pathologists are increasingly relying in image recognition and artificial intelligence tools to aid in the early identification and evaluation of this lesion. We examined a number of papers that dealt with this issue during the time span in order to shed light on the studies that were performed in this area (2017 and 2020). We have looked at experiments that used Convolutional Neural Network (CNN) technologies in the study of endoscopic images to help with early detection or diagnosis of esophageal cancer and its various forms. More research on esophageal malignant growth is required, as well as improving the disease’s indicative existence and employing more proven techniques for feature selection/extraction of endoscopic images. The aim of this review is to highlight the research conducted on endoscopic images of the esophagus using deep learning algorithms, including CNN, Support Vector Machine (SVM), Random Forests (RF) and other techniques that were used to design the Computer-Aided Detection (CAD) system. In this review we covered some but not all articles that was of great contact with our master’s thesis research in this regard.