A New Medical Image Fusion Approach Using Spatial Attention and Weighted Local Energy

Abstract

∗Medical image fusion technology has become a powerful tool for assisting physicians in diagnosing diseases and locating lesions since it can integrate various pathological information given by single modal medical images. This paper proposes a novel multi-modality medical image fusion approach in generalized rieszwavelet transform (GRWT) domain. Two source images are deconstructed by GRWT into low-frequency sub-band and highfrequency sub-band. Then a spatial attention based fusion rule and a weighted local energy (WLE) fusion rule are adopted for low-frequency sub-band and high-frequency sub-band respectively. After performing the inverse GRWT, final fused image is obtained. The proposed method is tested on two cases of MRI-CT group and MRI-PET group. Qualitative analysis and quantitative evaluations are performed on the proposed method with better performance in AG,Q_NCIE,PSNR,SF and RMSE. The results suggest that our ∗Corresponding author: YingHan, Biomedical Engineering Institute, Hainan University, Haikou, China 570228; Department of Neurology, Xuanwu Hospital of Capital Medical University, Beijing, China 100053; Center of Alzheimer’s Disease, Beijing Institute for Brain Disorders, Beijing, China 100053; hanying@xwh.ccmu.edu.cn. †Corresponding author: Guixia Kang, Key Laboratory of Universal Wireless Communications, Ministry of Education, Beijing University of Posts and Telecommunications, Beijing, China; gxkang@bupt.edu.cn. Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from permissions@acm.org. ICBBE ’21, November 12–15, 2021, Kyoto, Japan © 2021 Association for Computing Machinery. ACM ISBN 978-1-4503-8507-7/21/11. . . $15.00 https://doi.org/10.1145/3502871.3502872 method is an effective and valuable approach to help clinicians to diagnose diseases.