Medical image fusion aims to combine information from different sources acquired with different imaging modalities to improve the diagnosis and treatment of diseases. In recent years, medical image fusion has become a very active topic for which a variety of fusion methods have been proposed. Most of them focus on voxel-level algorithms for multimodal brain images. Many effective image representation approaches such as multiscale transforms, sparse representations, and deep learning have been introduced for the development of fusion algorithms. Despite recent advances, medical image fusion still faces challenges, such as few fusion methods for real-world scenarios like inaccurate registration context, a lack of benchmarks for performance evaluation of methods, and few studies on specific applications of fusion methods.
This Research Topic focuses on reporting advanced studies related to multimodal brain image fusion, including methods, evaluations, and applications, with the aim of promoting the development of multimodal medical image fusion technology. More specifically, this Research Topic expects high-quality work on novel brain image fusion methods that are robust to real-world scenarios such as inaccurate registration, effective approaches for performance evaluation of multimodal brain image fusion, practical applications of multimodal image fusion for classification, recognition and segmentation tasks related to diagnosis and treatment of brain diseases. Multimodal image fusion methods for the brain developed for modalities such as computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET), single photon emission computed tomography (SPECT), can be presented based on both conventional imaging and deep neural networks. Performance evaluation studies can be performed to establish benchmarks with datasets, objective metrics, and baseline methods. In addition, specific applications such as brain anomaly detection and glioma segmentation can be investigated using image fusion techniques.
This Research Topic is therefore proposed to advance the fundamentals and technologies of multimodal image fusion in the brain, including methods, assessments, and applications. Both original and review articles are welcome. Topics of interest may include, but are not limited to:
- Multimodal brain image registration and fusion;
- Machine learning/deep learning for brain image fusion;
- Datasets and benchmarks for multimodal brain image fusion;
- Objective evaluation for brain image fusion;
- Applications of multimodal brain images in diagnosis of relevant diseases
Medical image fusion aims to combine information from different sources acquired with different imaging modalities to improve the diagnosis and treatment of diseases. In recent years, medical image fusion has become a very active topic for which a variety of fusion methods have been proposed. Most of them focus on voxel-level algorithms for multimodal brain images. Many effective image representation approaches such as multiscale transforms, sparse representations, and deep learning have been introduced for the development of fusion algorithms. Despite recent advances, medical image fusion still faces challenges, such as few fusion methods for real-world scenarios like inaccurate registration context, a lack of benchmarks for performance evaluation of methods, and few studies on specific applications of fusion methods.
This Research Topic focuses on reporting advanced studies related to multimodal brain image fusion, including methods, evaluations, and applications, with the aim of promoting the development of multimodal medical image fusion technology. More specifically, this Research Topic expects high-quality work on novel brain image fusion methods that are robust to real-world scenarios such as inaccurate registration, effective approaches for performance evaluation of multimodal brain image fusion, practical applications of multimodal image fusion for classification, recognition and segmentation tasks related to diagnosis and treatment of brain diseases. Multimodal image fusion methods for the brain developed for modalities such as computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET), single photon emission computed tomography (SPECT), can be presented based on both conventional imaging and deep neural networks. Performance evaluation studies can be performed to establish benchmarks with datasets, objective metrics, and baseline methods. In addition, specific applications such as brain anomaly detection and glioma segmentation can be investigated using image fusion techniques.
This Research Topic is therefore proposed to advance the fundamentals and technologies of multimodal image fusion in the brain, including methods, assessments, and applications. Both original and review articles are welcome. Topics of interest may include, but are not limited to:
- Multimodal brain image registration and fusion;
- Machine learning/deep learning for brain image fusion;
- Datasets and benchmarks for multimodal brain image fusion;
- Objective evaluation for brain image fusion;
- Applications of multimodal brain images in diagnosis of relevant diseases