Machine Learning Approaches for Early Detection of Lung Cancer

Abstract

Lung cancer is a very harmful form of cancer that is hard to find. It happens when cancerous cell grow in the airways. This type of cancer can be deadly. Doctors struggle to detect and treat it properly. They still don't know the exact cause or cure for cancer. If cancer is found early, it can be treated. It causes death in both men and women, so it's important to quickly and accurately examine any nodules. To detect lung cancer early, different techniques have been used. Finding cancer early can save lives and give patients a better chance of recovery. Technology plays a significant part in accurately identifying cancer. Researcher have proposed different approaches based on their studies. In recent years, machine learning algorithms have become powerful tools for helping detect lung diseases. Different algorithms have been used to study medical data and find patterns that show lung diseases. This paper reviews the utilization of ML methods for detecting lung cancer. Four methods—ANN (Artificial Neural Networks), SVM(Support Vector Machines), KNN(k-Nearest Neighbors), and Naive Bayes—were tested to see how well they detected lung cancer. KNN had the highest accuracy, with an average of 98.5%. The paper also looked at different datasets to find the most suitable one for detecting lung cancer. After reviewing multiple papers, the LIDC dataset was considered the best choice for this task. This review paper will assist researchers in efficiently reviewing relevant literature without having to refer to numerous papers.