Development of Efficient Information Preservation Techniques for Medical Image Fusion

Abstract

Medical Imaging approaches such as CT and MRI scans capture images within the human body which are helpful for medical practitioners to find out the abnormalities if any. Different approaches give diagnostic information at different level. During certain cases, an MRI scan is suggested when a CT scan doesn t provide enough details. The fusion of CT and MRI scan images overcome the disadvantage of missing details obtained through a single approach. Fusion of images obtained through two same or different imaging approaches in order to provide more finer details and to avoid information loss is called medical image fusion. The objective of the research is to develop an improved framework for the fusion of medical images. newlineIn the first contribution, the different fusion mechanisms are studied, analyzed and implemented. The results are compared. This led to the selection of transformation for the decomposition of images. In the second contribution, a novel fusion rule is proposed, and thus an efficient fusion algorithm with improved values of QAB/f is developed. The performance measures are comparatively improved than the existing algorithms. In the third and final contribution, the efficiency of the fusion mechanism is improved by newline6 newlineoptimizing the fusion rule. The fusion rule plays a key role in the selection of pixels that appear in the fused image. Thus the optimization of those pixel value selections improves the fused image parameters. Preservation of borders and edges are improved and important details are preserved. newline