Enhancement Methods of Brain MRI Images : A Review

Abstract

Image processing shows an important part of collecting information on brain images. Magnetic resonance imaging (MRI) technique provides important information for doctors to diagnose diseases. The image processing technique begins with image pre-processing to improve the quality of the original image. The procedures of images pre-processing cover artifact elimination, skull despoil, noise elimination, and image quality enhancement. Detecting tumors easily requires processed images. This study is a review of the current methods used in the process of enhancing the quality of brain MRI images. The study aims to review current methods for enhancing the quality of MRI images to identify the strengths and weaknesses of each method to proceed to the next stage in detecting tumors. The strengths and weaknesses of each method are considered in selecting the best method for handling a variety of different cases. The summary of each method is presented in a table followed by a brief explanation. This study reveals that the Average Intensity Reinstatement placed on Adaptive Histogram Equalization is the best pre-processing method for clinical datasets with the highest PSNR values of 87.370 and the Brainweb dataset shows that the combined Contrast Guided Interpolation and Iterative back-projection methods are the best pre-processing method with the highest PSNR values of 30.196. Meanwhile, Non-Local Means Filter is the best pre-processing method for the clinical dataset because it has the lowest MSE value of 0.025 compared to others.