Development of Linear Contrast Stretching (LCS) and Region of Interest (ROI) Method for Detecting Spinal Fractures Based on CT-Scan Images

Abstract

This study aims to develop linear contrast stretching (LCS) and region of interest (ROI) methods to detect spinal fractures based on CT-scan images. LCS is applied to enhance image contrast by expanding the pixel intensity range, thereby clarifying bone structure. ROI is used to focus the analysis on relevant areas, such as sections of bone suspected to have fractures. The development of these LCS and ROI methods is expected to improve fracture detection visibility, as evidenced by more precise measurement of the detected fracture pixel area and increased detection accuracy, which is anticipated to facilitate the clinical diagnosis process. The study consists of several stages, including image preprocessing, edge detection, image sharpening, and the application of LCS + ROI on the targeted area. The study findings indicate that the proposed method is capable of detecting fractures with high accuracy, thereby assisting radiologists in diagnosing spinal fractures more quickly and accurately, achieving an accuracy rate of 95%. The implementation of this method is expected to make a significant contribution to CT-scan image-based medical diagnosis.