KLASIFIKASI CITRA RONTGEN DADA UNTUK MENDETEKSI PENYAKIT TUBERKULOSIS

Helmy Amalia Ariesta, NIM.: 21206051005 (2023) KLASIFIKASI CITRA RONTGEN DADA UNTUK MENDETEKSI PENYAKIT TUBERKULOSIS. Masters thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

Preview	Text (KLASIFIKASI CITRA RONTGEN DADA UNTUK MENDETEKSI PENYAKIT TUBERKULOSIS) 21206051005_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf - Published Version Download (2MB) | Preview
	Text (KLASIFIKASI CITRA RONTGEN DADA UNTUK MENDETEKSI PENYAKIT TUBERKULOSIS) 21206051005_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf Restricted to Registered users only Download (3MB) | Request a copy

Abstract

An infectious disease that is a significant contributor to poor health and one of the leading causes of death in the world is tuberculosis (TB). Radiologists can do clinical work, examining chest X-ray results, but the limited number of hospital radiologists may not handle the high demand for chest X-ray readings. Therefore, automation is needed to give tuberculosis negative results and positive tuberculosis results quickly. In this case machine learning is an intelligent system that can be used. This study performs digital image processing, namely active countor segmentation, then inputs the results of segmentation to find the value of feature extraction of first-order statistics, GLCM and second-order statistics and classifies using the KNN, LDA, NB, SVM and decision tree algorithms. The results showed that the best feature extraction was obtained from GLCM feature extraction using four parameters (Contrast, Correlation, Energy and Homogeneity) with the KNN classification algorithm resulting in accuracy, sensitivity and specificity of 92.5%, 100% and 86.9%, respectively.

Share this knowledge with your friends :

Actions (login required)

View Item