TEAM PILOT -- Learned Feasible Extendable Set of Dynamic MRI Acquisition Trajectories

TL;DR

A novel deep-compressed sensing approach that uses 3D window attention and flexible, temporally extendable acquisition trajectories that significantly reduces both training and inference times compared to existing approaches, while also adapting to different temporal dimensions during inference without requiring additional training.

Abstract

Dynamic Magnetic Resonance Imaging (MRI) is a crucial non-invasive method used to capture the movement of internal organs and tissues, making it a key tool for medical diagnosis. However, dynamic MRI faces a major challenge: long acquisition times needed to achieve high spatial and temporal resolution. This leads to higher costs, patient discomfort, motion artifacts, and lower image quality. Compressed Sensing (CS) addresses this problem by acquiring a reduced amount of MR data in the Fourier domain, based on a chosen sampling pattern, and reconstructing the full image from this partial data. While various deep learning methods have been developed to optimize these sampling patterns and improve reconstruction, they often struggle with slow optimization and inference times or are limited to specific temporal dimensions used during training. In this work, we introduce a novel deep-compressed sensing approach that uses 3D window attention and flexible, temporally extendable acquisition trajectories. Our method significantly reduces both training and inference times compared to existing approaches, while also adapting to different temporal dimensions during inference without requiring additional training. Tests with real data show that our approach outperforms current state-of-theart techniques. The code for reproducing all experiments will be made available upon acceptance of the paper.

Figures (10)

• Figure 1: 3D attention-based reconstruction model. The model receives the downsampled frames $\mathbf{\tilde{Z}}$ in the image domain as the input, processes them using a combination of 3D convolution and windowed attention blocks, outputting the reconstructed frames $\mathbf{\hat{Z}}$.
• Figure 2: Full data processing pipeline including the emulation of data acquisition and image reconstruction. Fully sampled frames $\mathbf{Z}$ serving as the "ground-truth" are fed into the pipeline alongside with the acquisition trajectories $\mathbf{K}$. The reconstructed frames are received at the output.
• Figure 3: Acquisition Time Minimization. Our results achieves results on-par with Multi-PILOT using 50% less sampling points.
• Figure 4: Mean Temporal Derivative $\mu_{2k}$ - with (blue) and without (orange) trajectory refinement.
• Figure 5: Reconstruction Results In Sequence Transition And Intermediate Frames - For inference of a 24 length sequence with trajectory stacking applied before (B,D) and after (A,C) trajectory refinement. Frame indices are in the bottom left.