Quantitative Evaluation of Multi-Modal Medical Image Fusion Using Diagnostic Performance Metrics

This multi-modal medical imaging is very important for the precise identification of a certain disorder because many imaging modalities give the user not only anatomical but also functional information at no cost. However, relying on a single modality of images often limits diagnostic accuracy because of the lack of features. In this work, the goal is to quantitatively evaluate multi-modal medical picture fusion techniques with traditional medical diagnostic performance metrics. These images are merged together from a variety of modalities to create a composite image that contains a lot of information from all modalities. The ability to distinguish fused images is carefully measured against single-modality images, with a number of parameters such as accuracy, sensitivity, and specificity. Experimental results show that the fused images regularly outperform the individual modality images in all the criteria tested, which means that they provide better feature discrimination and enhanced diagnosis support. The results indicate the effectiveness of multi-modal picture fusion as an effective tool to assist in clinical decision-making and highlight its potential in computer-aided diagnostic systems.