DAFNet: Dual-path Adaptive Fusion Network for High-Fidelity Cross-Modality Brain MR Image Synthesis
Abstract
Cross-modality MRI synthesis using deep learning can streamline clinical workflows, compensate for missing image contrasts, and reduce examination time; however, existing methods often require substantial computational resources, limiting their practical accessibility. To address this limitation, we propose and investigate a lightweight yet effective framework, termed the Dual-path Adaptive Fusion Network (DAFNet), for T1-to-T2 images synthesis. DAFNet is based on cGAN (conditional generative adversarial networks) structure, and the generator employs a dual-path encoder that adaptively combines depth-wise separable convolutions and dilated convolutions to capture both fine-grained details and broader contextual features. A novel Dual-path Adaptive Fusion Module (DAFM) is introduced to dynamically combine these feature streams through channel-wise complementary weighting, enabling efficient and adaptive feature integration. This fusion mechanism is further applied to the final skip connection to enhance reconstruction fidelity, while the decoder adopts a standard transposed convolution architecture. Coupled with a conventional conditional GAN discriminator, the proposed model maintains low computational complexity while improving perceptual image fidelity. Experimental results demonstrate that DAFNet achieves superior performance compared with baseline U-Net and conventional cGAN models, reaching a peak signal-to-noise ratio (PSNR) of 26.43 dB while significantly reducing model size. This result indicates that DAFNet provides an effective balance between synthesis fidelity and computational efficiency, making it a promising solution for deployment on standard computing hardware and for broader clinical and research applications in MRI contrast synthesis.
Related
- Bridging cone-beam CT and MRI to stopping power ratio maps: a modality-agnostic Brownian bridge approach to support robust adaptive proton therapy
- High-fidelity dual-domain photoacoustic data synthesis with latent diffusion models
- Adaptive Multi-Fidelity Structural Optimization under Fluid-Structure Interaction
- XPAR-Net: cross-paradigm alignment and reconstruction network for high-fidelity lens-free on-chip microscopy
- Load-Balanced Neural-Network-Aided Adaptive Tabulation of Phase Equilibria for Transcritical Multiphase Flows
BibTeX
@misc{liu2026dafnet,
title = {{DAFNet: Dual-path Adaptive Fusion Network for High-Fidelity Cross-Modality Brain MR Image Synthesis}},
author = {Jialin Liu and Xiaoliang Zhang},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.33849},
eprint = {2609.33849},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.33849},
}
RIS
TY - UNPB
TI - DAFNet: Dual-path Adaptive Fusion Network for High-Fidelity Cross-Modality Brain MR Image Synthesis
AU - Jialin Liu
AU - Xiaoliang Zhang
PY - 2026
DA - 2026/09/27
PB - arXiv
DO - 10.48550/arxiv.2609.33849
UR - https://arxiv.org/abs/2609.33849
AB - Cross-modality MRI synthesis using deep learning can streamline clinical workflows, compensate for missing image contrasts, and reduce examination time; however, existing methods often require substantial computational resources, limiting their practical accessibility. To address this limitation, we propose and investigate a lightweight yet effective framework, termed the Dual-path Adaptive Fusion Network (DAFNet), for T1-to-T2 images synthesis. DAFNet is based on cGAN (conditional generative adversarial networks) structure, and the generator employs a dual-path encoder that adaptively combines depth-wise separable convolutions and dilated convolutions to capture both fine-grained details and broader contextual features. A novel Dual-path Adaptive Fusion Module (DAFM) is introduced to dynamically combine these feature streams through channel-wise complementary weighting, enabling efficient and adaptive feature integration. This fusion mechanism is further applied to the final skip connection to enhance reconstruction fidelity, while the decoder adopts a standard transposed convolution architecture. Coupled with a conventional conditional GAN discriminator, the proposed model maintains low computational complexity while improving perceptual image fidelity. Experimental results demonstrate that DAFNet achieves superior performance compared with baseline U-Net and conventional cGAN models, reaching a peak signal-to-noise ratio (PSNR) of 26.43 dB while significantly reducing model size. This result indicates that DAFNet provides an effective balance between synthesis fidelity and computational efficiency, making it a promising solution for deployment on standard computing hardware and for broader clinical and research applications in MRI contrast synthesis.
ER -