Multimodal Medical Image Fusion Using DWT and SIDWT: A Wavelet-Based Approach for Enhanced Diagnostic Accuracy

Abstract

The modalities of various medical images give only limited details corresponding to soft details like muscles or hard details like bones or skull, etc. Hence using a single modality, the prediction of diseases limits the diagnosis capability. This further results in compromise of accuracy. In the case of images, fusion methods enhance the accuracy by merging multimodality medical images like CT, MRI, PET, etc. This paper details about DWT and SIDWT techniques for fusion of images. In these techniques, wavelet transforms are used to cobineCT and MRI images. The DWT and SIDWT techniques are used with maximum and saliency features. The improvement in PSNR is atleast by 1%, MSE by 5.9%, Entropy by 29.37%, standarad deviation by 29.86% and mutual information by 11.8% when compared with input images.

METHODS:DWT and SIDWT

FINDINGS: The improvement in PSNR is atleast by 1%, MSE by 5.9%, Entropy by 29.37%, standarad               

deviation by 29.86% and mutual information by 11.8% when compared with input images.

NOVELTY: fusion of images from different molarities.