Stress Fracture Detection using Adaptive Histogram Equalization (Ahe) and Contrast Limited Adaptive Histogram Equalization (Clahe): a comparative study / Faikah Zakaria and Hamizah Emran

• Main Authors: Zakaria, Faikah, Emran, Hamizah
• Format: Article
• Language: English
• Published: Faculty of Health Sciences, Universiti Teknologi MARA 2020
• Subjects: Examination. Diagnosis. Including radiography; Fractures (General)
• Online Access: https://ir.uitm.edu.my/id/eprint/55009/1/55009.pdf; https://ir.uitm.edu.my/id/eprint/55009/; http://healthscopefsk.com/

A lower extremity stress fracture is account about 80%-90% of all stress fractures and are standard among the person who involved in strength and substantial loading activities. Early detection and diagnosis manage to prevent further severe complication. However, initial plain radiography is unable to identify the presence of stress fracture approximately for an about up to six to ten weeks after the onset of the fracture; especially in the early stages without displacement of the bone fragment. The purpose of this paper is to discuss the enhancement of stress fracture detection of AP ankle digital X-ray by using image enhancement techniques. Two techniques are proposed in this paper, which is Adaptive Histogram Equalization (AHE) and Contrast Limited Adaptive Histogram Equalization (CLAHE). These study data images were retrieved from the online database. A total of 10 original images were retrieved, and 10 test images for each Set A and B are produced. The quantitative evaluation of this study was done, which is the structural similarity index measurement (SSIM) and peak noise to signal ratio (PSNR) in MATLAB software. AHE showed that it has better image enhancement and high image quality for SSIM and PSNR value compared to CLAHE. At p>0.05, there is a significant difference between SSIM and PSNR in AHE and CLAHE. From this study, it shows that the image enhancements techniques can be done as pre-processing steps to improve the enhancement of the stress fracture detection in digital AP ankle X-ray images.