Publication: Histogram equalization based contrast enhancement method for COVID-19 chest x-ray images
Loading...
Date
2024-07
Authors
Oh, She Quen
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The pandemia caused by the SARS-CoV-2 virus has triggered an unprecedented health and economic crisis. Chest X-Ray examinations become the most popular diagnostic method due to its affordability and ability to assess various conditions quickly and simply. However, X-Ray images are in low contrast due to the high value of Kilovoltage peak (kVp) and the scatter radiation that travels in all directions during the radiographic examination. This study aims to enhance the contrast of COVID-19 Chest X-Ray images through Histogram Equalization (HE) based contrast enhancement methods, facilitating more accurate disease identification and assessment by clinicians. Four techniques namely Histogram Equalization (HE), Adaptive Histogram Equalization (AHE), Contrast Limited Adaptive Histogram Equalization (CLAHE), and Brightness Preserving Bi-Histogram Equalization (BBHE) are employed and compared for image contrast improvement. Qualitative assessment through visual inspection and quantitative evaluation utilizing entropy, Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Method (SSIM), and Contrast Improvement Index (CII) are conducted on a dataset comprising 200 COVID 19 Chest X-Ray images. The quantitative analysis show that CLAHE method with a Clip Limit of 0.005, is the best method in enhancing the contrast of the COVID-19 Chest X-Ray images, with the highest PSNR (22.59), SSIM (0.89), and CII (1.03) values when compared with the other three methods. The qualitative analysis show that CLAHE perform the best in producing the enhanced images with more natural look while preserving most of the details in the image. CLAHE achieves a balanced enhancement of contrast and achieving well-distributed brightness levels. The results show CLAHE is an effective approach for COVID-19 Chest X-Ray images enhancement and can be used as a pre-processing tool for medical image understanding and analysis.