An approach for medical image registration using a Double Exponential-Dynamic Group Based Cooperative Optimization algorithm
Innasi Lineta P.,J. Jerusalin Carol
DOI: https://doi.org/10.1007/s11042-024-18429-z
IF: 2.577
2024-02-14
Multimedia Tools and Applications
Abstract:Medical image registration is the process of aligning multiple medical images, volumes, or surfaces to a common coordinate system In multiple medical image analysis applications, image registration plays a significant role. For medical image registration, conventional approaches are chiefly optimization-driven, which finds the optimal deformations that exploit the similarity between two images. Accordingly, this paper plans to develop a novel optimization approach for the registration, with the assistance of two images, such as a reference image as well as a moving image. Here, the Convolutional Neural Network (CNN) features are extracted from transformation-applied image and the reference image. Then, with the assistance of the proposed multi-similarity model, a similarity between the features of the transformed image and the reference image is found. For the proposed multi-similarity model, the Logarithmic Normalized Mutual Information (LNMI), Modified Normalized Mutual Information (MNMI), and Tversky's feature contrast model are used. Finally, the proposed Double Exponential-Dynamic Group Based Cooperative Optimization algorithm (DE-DGCO), which is developed by combining Dynamic Group Based Cooperative Optimization (DGCO) and Double Exponential smoothing (DeS) is used to update the weight and the transformation parameters. Finally, experimentation is done to test the efficiency of the proposed model. Here, the DE-DGCO algorithm attained enhanced performance than the traditional techniques regarding maximum PSNR value of 47.737; dice score value of 0.855, and minimal RMSE value of 0.113.
computer science, information systems, theory & methods,engineering, electrical & electronic, software engineering