Multimodal Medical Image Fusion using Hybrid Domains

Abstract

In a variety of clinical applications, image fusion is critical for merging data from multiple sources into a single, more understandable outcome. The use of medical image fusion technologies to assist the physician in executing combination procedures can be advantageous. The diagnostic process includes preoperative planning, intra operative supervision, an interventional treatment. In this thesis, a technique for image fusion was suggested that used a combination model of PCA and CNN. A method of real-time image fusion that employs pre-trained neural networks to synthesize a single image from several sources in real-time. A innovative technique for merging the images is created based on deep neural network feature maps and a convolution network. Picture fusion has become increasingly popular as a result of the large variety of capturing techniques available. The proposed design is implemented using deep learning technique. The accuracy of the proposed design is around 15% higher than the existing design. The proposed fusion algorithm is verified through a simulation experiment on different multimodality images. Experimental results are evaluated by the number of well-known performance evaluation metrics