The significance of image fusion in nuclear medicine and molecular imaging

Abstract

Nuclear medicine molecular imaging (NMMI) typically employs radioactive isotopes to label cells or molecules and then utilizes imaging devices such as positron emission tomography and single photon emission computed tomography to generate images. However, the images produced by these devices often suffer from problems such as signal noise, low resolution, and poor soft-organ contrast. To address these limitations, image fusion technology merges images from different imaging modalities, combining multiple types of medical image information obtained through various imaging techniques. This process generates a more comprehensive and accurate image, significantly improving image quality, reducing noise, and ultimately enhancing diagnostic accuracy and treatment effectiveness. Image fusion technology has found widespread applications in NMMI, achieving significant results in various fields. This review provides an overview of the development of image fusion technology, introduces traditional image fusion techniques, explores deep learning-based image fusion methods, and finally discusses the challenges and future directions of image fusion technology in NMMI.