MR Image Enhancement using Stationary Wavelet Transform based Approach
Pranaba Kumar Mishro, Sanjay Agrawal, Rutuparna Panda, Ku. Tapaswini Hansdah · 2020
Magnetic resonance (MR) imaging is a widely used modality in clinical image diagnosis. Usually, these images contain anatomical structures, tissue regions along with the artefacts. Low contrast is a challenging issue in the diagnosis of medical images. It may occur in MR images due to poor imaging environment and instrument errors. Further, the quality of an image may degrade during processing and transmission. The spatial domain techniques used for enhancement fails to retain the complex edge/structural features of the MR images. So to address this problem, we suggest a stationary wavelet transform based approach for enhancing the MR image quality. The enhancement is done in the transform domain. Only the detail component of the transformed image is enhanced using adaptive histogram equalization technique. The proposed technique is experimented using synthetic and real MR images of different anatomical structures of the human body. The method is validated using evaluation indices: absolute mean brightness error, edge based contrast measure and relative contrast measure. The evaluation parameters show the superiority of the suggested technique in comparison to the state-of-the-art methods.