Fast curvelet transform through genetic algorithm for multimodal medical image fusion

Abstract

Currently, medical imaging modalities produce different types of medical images to help doctors to diagnose illnesses or injuries. Each modality of images has its specific intensity. Many researchers in medical imaging have attempted to combine redundancy and related information from multiple types of medical images to produce fused medical images that can provide additional concentration and image diagnosis inspired by the information for the medical examination. We propose a new method and method of fusion for multimodal medical images based on the curvelet transform and the genetic algorithm (GA). The application of GA in our method can solve the suspicions and diffuse existing in the input image and can further optimize the characteristics of image fusion. The proposed method has been tested in many sets of medical images and is also compared to recent medical image fusion techniques. The results of our quantitative evaluation and visual analysis indicate that our proposed method produces the best advantage of medical fusion images over other methods, by maintaining perfect data information and color compliance at the base image.