arXiv 2412.107762024-12-14eess.IVcs.AIcs.CV

从频率调制、空间纯化和尺度多样化的角度提升基于ViT的MRI重建

Boosting ViT-based MRI Reconstruction from the Perspectives of Frequency Modulation, Spatial Purification, and Scale Diversification

Fudan University（复旦大学）；Digital Medical Research Center（数字医学研究中心）；School of Basic Medical Science（基础医学院）；Shanghai Key Laboratory of Medical Image Computing and Computer Assisted Intervention（上海医学影像计算与计算机辅助干预重点实验室）；Academy for Engineering and Technology（工程与技术学院）

Yucong Meng, Zhiwei Yang, Yonghong Shi, Zhijian Song

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出FPS-Former，从频率调制、空间纯化和尺度多样化三个角度解决ViT在MRI重建中的高频捕捉、注意力噪声和多尺度建模问题，在三个公共数据集上以更低计算成本超越现有方法。

摘要：

加速MRI重建过程由于k空间中的大量欠采样而呈现出具有挑战性的病态逆问题。近年来，视觉Transformer（ViTs）已成为该任务的主流方法，展示了显著的性能提升。然而，仍有三个重要问题尚未解决：（1）ViTs难以捕捉图像的高频分量，限制了其检测局部纹理和边缘信息的能力，从而阻碍了MRI恢复；（2）先前的方法在内容相关和不相关的token之间计算多头自注意力（MSA），引入了噪声并显著增加了计算负担；（3）ViTs中的朴素前馈网络无法建模对图像恢复至关重要的多尺度信息。在本文中，我们提出了FPS-Former，一个基于ViT的强大框架，从频率调制、空间纯化和尺度多样化的角度来解决这些问题。具体而言，针对问题（1），我们引入了频率调制注意力模块，通过在拉普拉斯金字塔中自适应地重新校准频率信息来增强自注意力图。针对问题（2），我们定制了空间纯化注意力模块，以捕捉紧密相关token之间的交互，从而减少冗余或不相关的特征表示。针对问题（3），我们提出了一种基于混合尺度融合策略的高效前馈网络。在三个公共数据集上进行的全面实验表明，我们的FPS-Former在降低计算成本的同时，优于最先进的方法。

英文摘要：

The accelerated MRI reconstruction process presents a challenging ill-posed inverse problem due to the extensive under-sampling in k-space. Recently, Vision Transformers (ViTs) have become the mainstream for this task, demonstrating substantial performance improvements. However, there are still three significant issues remain unaddressed: (1) ViTs struggle to capture high-frequency components of images, limiting their ability to detect local textures and edge information, thereby impeding MRI restoration; (2) Previous methods calculate multi-head self-attention (MSA) among both related and unrelated tokens in content, introducing noise and significantly increasing computational burden; (3) The naive feed-forward network in ViTs cannot model the multi-scale information that is important for image restoration. In this paper, we propose FPS-Former, a powerful ViT-based framework, to address these issues from the perspectives of frequency modulation, spatial purification, and scale diversification. Specifically, for issue (1), we introduce a frequency modulation attention module to enhance the self-attention map by adaptively re-calibrating the frequency information in a Laplacian pyramid. For issue (2), we customize a spatial purification attention module to capture interactions among closely related tokens, thereby reducing redundant or irrelevant feature representations. For issue (3), we propose an efficient feed-forward network based on a hybrid-scale fusion strategy. Comprehensive experiments conducted on three public datasets show that our FPS-Former outperforms state-of-the-art methods while requiring lower computational costs.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。