MRI-Brain Image Contrast Enhancement based on Comparative of Multiple Histogram Equalization Techniques

Abstract

The poor image quality in Magnetic Resonance Imaging (MRI), notably lack of contrast between the tissues, artifacts inherent to these types of images, and interindividual variability, could make the image analysis challenging and compromise the clinical diagnosis accuracy. In order to enhance the important image contents by minimizing the noise and maintaining the actual detailed features, image enhancement methods are required. This research compares the majority of histogram-based methods used to denoise and improve brain MRI images, primarily Contrast Limited Adaptive Histogram Equalization (CLAHE), Histogram Equalization (HE), Bi-histogram Equalization with a Plateau Limit (BHEPL), and Brightness Preserving Histogram Equalization with Maximum Entropy (BPHEME) methods. Based on an analysis of the quality measurement metrics Mean Square Error (MSE), peak signal to noise ratio (PSNR), Entropy, and Root Mean Square Error (RMSE), an experimental investigation with the use of real brain MRI images has been conducted. The effectiveness of utilizing CLAHE to enhance the contrast regarding extremely high-quality MRI medical images has been demonstrated in experimental results based on the quality measurement metrics that the average rate for the CLAHE ( <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\text{MSE} =542.15, \text{PSNR}=22.21, \text{RMSE} =21.63, \text{entropy}=5.21$</tex> ).