Efficient Complex-Valued Vision Transformers for MRI Classification Directly from k-Space

TL;DR

This paper introduces a complex-valued Vision Transformer that directly classifies MRI k-Space data, preserving phase information and significantly reducing computational resources while maintaining high accuracy.

Contribution

It presents a novel complex-valued Vision Transformer with a radial k-Space patching strategy for direct MRI classification, improving efficiency and robustness.

Findings

Achieves competitive classification accuracy with state-of-the-art image-based models.

Exhibits superior robustness to high acceleration factors.

Reduces VRAM consumption during training by up to 68 times.

Abstract

Deep learning applications in Magnetic Resonance Imaging (MRI) predominantly operate on reconstructed magnitude images, a process that discards phase information and requires computationally expensive transforms. Standard neural network architectures rely on local operations (convolutions or grid-patches) that are ill-suited for the global, non-local nature of raw frequency-domain (k-Space) data. In this work, we propose a novel complex-valued Vision Transformer (kViT) designed to perform classification directly on k-Space data. To bridge the geometric disconnect between current architectures and MRI physics, we introduce a radial k-Space patching strategy that respects the spectral energy distribution of the frequency-domain. Extensive experiments on the fastMRI and in-house datasets demonstrate that our approach achieves classification performance competitive with state-of-the-art image-domain baselines (ResNet, EfficientNet, ViT). Crucially, kViT exhibits superior robustness to high acceleration factors and offers a paradigm shift in computational efficiency, reducing VRAM consumption during training by up to 68$\times$ compared to standard methods. This establishes a pathway for resource-efficient, direct-from-scanner AI analysis.