Identifikasi Penyakit Kanker Paru-Paru Dengan Menggunakan Metode GLCM dan Convolutional Neural Network

Abstract

Cancer is the uncontrolled growth and spread of cells. One of the deadliest types of cancer is lung cancer. Lung cancer is the uncontrolled growth of cancer cells in lung tissue. Lung cancer is the leading cause of death of all cancer deaths in both men and women. Lung cancer is diagnosed through a chest X-ray (CXR) or better known as an X-ray. This CXR radiological examination is very helpful in the process of medical diagnosis and identification of lung disease. The lack of public knowledge in reading X-ray results requires experts such as doctors or other medical personnel to read them manually. Reading manually still allows for errors in diagnosing diseases. Diagnosing this disease requires good skills and facilities. Because if it is not treated immediately, it can spread and metastasize and eventually increase the degree of severity. The methods proposed in this study are GLCM and Convolutional Neural Networks to identify whether an image is normal or cancerous. The steps involved in this identification are pre-processing, segmentation with k-means clustering, feature extraction with Gray Level Co-Occurrence Matrix and identification. In this study it was shown that the proposed method was able to identify cancer with an accuracy of 75%, a sensitivity of 75% and a specificity of 75%.