Lung Cancer Detection with Prediction Employing Machine Learning Algorithms: A Recent Study

Abstract

Every year, the number of people dying from lung cancer rises around the world. It is second most cancer affecting among population worldwide. The ability to forecast the onset of cancer in patients can aid clinicians in making decisions about their drugs and therapies.This study suggests a new technique for detecting and predicting the existence of malignant nodules in the lungs of patients. To conduct the classification, the suggested system uses a machine learning technique called support vector machine (SVM) and a deep learning algorithm called convolutional neural network (CNN) and a large lung cancer repository database called the UCI repository. Images are pre-processed and then post-processed in the initial step of cleaning. The RGB to greyscale conversion is included in the pre-processing step, and the noise is removed using the Non-Local Means (NLM) filter in the post-processing step. Image segmentation was achieved using Otsu's method in the second stage of development, and feature extraction was achieved using Grey Level Co-occurrence Matrix (GLCM). Finally, the two classifiers are used to classify lung malignant images, and the accuracy of their classifications is compared and recorded.