Gray Level Co-Occurrence Matrix For Miliary Tuberculosis Pathologies

Abstract

Tuberculosis (Tb) Is Currently One Of The Most Dangerous Diseases Worldwide As Announced By The World Health Organization. Sputum Tests Or Thorax X-Rays Are Used To Identify Infected Patients. In Some Regions Of Africa, The Infection Rate Is Up To 90% While The Density Of Radiologists Is Too Low To Screen Patients Area-Wide While Diagnosing Tuberculosis Still Remains A Challenge. When Left Undiagnosed And Thus Untreated, Mortality Rates Of Patients With Tuberculosis Are High. In An Effort To Reduce The Burden Of The Disease, This Paper Presents Our Automated Approach For Detecting Miliary Tuberculosis In Conventional Poster Anterior Chest Radiographs. We First Preprocess The Lung Image Using Different Preprocessing Techniques And Segment The Image Using A Thresholding Algorithm In Order To Clearly Identify The Texture Feature Of Miliary Tuberculosis. For This Lung Region, We Compute A Set Of Statistical Texture Features Using Gray Label Co-Occurrence Matrix (Glcm) Algorithms, Which Enable The X-Rays To Be Diagnosis As Normal Or Miliary Tb Using Four Statistical Features. We Measure The Performance Of Our System Datasets: The Set Collected By Download From A Various Website Like Https://Ceb.Nlm.Nih.Gov/Repositories/Tuberculosischest Xray Image Data Sets/ And Japanese Society Of Radiological Technology (Jsrt) Built A Database. The Proposed Computer-Aided Diagnostic System For Miliary Tb Screening Achieves A Performance That Approaches The Performance Of Human Experts. We Achieve A Contrast Under The Confusion Matrix Of 20 Image Dataset From These 10 Miliary Tb Chest X-Ray Image And 10 Normal Chest X-Ray Image, From This Data Set We Get Accuracy 95% And Sensitivity 100% And Specificity 100% According To This Result This Work Reduce False Negative Result And Increase True Negative Result And Also Decrease False Positive Result And Increase True Positive Result.