A novel brightness preserving joint histogram equalization technique for contrast enhancement of brain MR images

Low contrast is a challenging factor in brain magnetic resonance (MR) images due to its structural complexity. Histogram equalization (HE) approach is often used in enhancing the contrast in brain MR images. However, the spatial information is not taken into account in this approach. Further, the problem of preserving structural details while retaining the brightness is also an important concern. To solve these, we suggest a novel stationary wavelet transform based brightness preserving joint histogram equalization (SWT-BPJHE) scheme for brain MR image contrast enhancement. Our contributions are - i) use of SWT to extract the low sub-band wavelength coefficients from the low contrast input image for enhancement, ii) to isolate the high frequency wavelength coefficients from enhance-ment, retaining the structural details, iii) to preserve brightness. The suggested scheme is experimented with synthetic brain MR images from BrainWeb and clinical images from Howard Whole Atlas databases. The performance is evaluated in terms of several valida-tion indices followed by statistical analysis. The outcomes reveal the superiority of the sug-gested scheme in comparison to state-of-the-art methods. (c) 2021 Nalecz Institute of Biocybernetics and Biomedical Engineering of the Polish Academy of Sciences. Published by Elsevier B.V. All rights reserved.