Improvement of the medical fusion process of images by fuzzy entropy and transformation of the contourlet

Abstract

Many types of medical pictures have to be fused, as single-modality medical images can give limited information because of the imagery and the complicated architecture of the human organ. This study proposes to offer a platform on which to make clinical diagnoses and to increase the accuracy of the target identification and the quality of the fused pictures by combining the benefits of nonsubsampled contourlet transform (NSCT) and fuzzy entropy. A picture is first broken down into low frequency or high frequency subbands through NSCT. In line with the various features of the low and high frequency components the respective fusion rules must be implemented. It calculates the level of membership of low frequency coefficients. The fusion of coefficients is also calculated and then utilized to retain picture features. By increasing regional energy, high-frequency components are merged. Inverse transformation produces the final fused picture. Experimental results have shown that, based on subjective visual effect and objective assessment standards, the suggested technique produces a satisfactory fusion effect. This process may also achieve high average gradient, standard deviation (SD), and edge preservation and maintain the fused picture features well. Effective reference can be provided by the outcome of the suggested algorithm for patients' assessment.