Study on segmentation and prediction of lung cancer based on machine learning approaches

Abstract

Lung cancer is a dangerous disease in human health. At the early stage, lung cancer detection provides a way to save human life. As a result, improvements in Deep Learning (DL), a technique, a branch of Machine Learning (ML), have helped to identify and classify lung cancer in clinical photographs. DL technology has also outperformed traditional methods in a variety of fields. Researchers are exploring various DL techniques for disease detection to improve the accuracy of the CAD systems in CT lung cancer detection. In this experiment, cutting-edge ML  and DL methods for lung disease have been recommended as CAD systems after thoroughly analysing existing frameworks. It can be separated into FP reduction systems and system to detect nodule. The primary characteristics of various approaches are analyzed. The CT lung datasets existing for examination and evaluation with the various approaches are also presented and discussed.