Motion-resolved real-time 4D flow MRI with low-rank and subspace modeling
Aiqi Sun,Bo Zhao,Yichen Zheng,Yuliang Long,Peng Wu,Bei Wang,Rui Li,He Wang
DOI: https://doi.org/10.1002/mrm.29557
Abstract:Purpose: To develop a new motion-resolved real-time four-dimensional (4D) flow MRI method, which enables the quantification and visualization of blood flow velocities with three-directional flow encodings and volumetric coverage without electrocardiogram (ECG) synchronization and respiration control. Methods: An integrated imaging method is presented for real-time 4D flow MRI, which encompasses data acquisition, image reconstruction, and postprocessing. The proposed method features a specialized continuous ( k , t ) $$ \left(\mathbf{k},t\right) $$ -space acquisition scheme, which collects two sets of data (i.e., training data and imaging data) in an interleaved manner. By exploiting strong spatiotemporal correlation of 4D flow data, it reconstructs time-series images from highly-undersampled ( k , t ) $$ \left(\mathbf{k},t\right) $$ -space measurements with a low-rank and subspace model. Through data-binning-based postprocessing, it constructs a five-dimensional dataset (i.e., x-y-z-cardiac-respiratory), from which respiration-dependent flow information is further analyzed. The proposed method was evaluated in aortic flow imaging experiments with ten healthy subjects and two patients with atrial fibrillation. Results: The proposed method achieves 2.4 mm isotropic spatial resolution and 34.4 ms temporal resolution for measuring the blood flow of the aorta. For the healthy subjects, it provides flow measurements in good agreement with those from the conventional 4D flow MRI technique. For the patients with atrial fibrillation, it is able to resolve beat-by-beat pathological flow variations, which cannot be obtained from the conventional technique. The postprocessing further provides respiration-dependent flow information. Conclusion: The proposed method enables high-resolution motion-resolved real-time 4D flow imaging without ECG gating and respiration control. It is able to resolve beat-by-beat blood flow variations as well as respiration-dependent flow information.