[Paper Review] Free-Breathing Myocardial T1 Mapping using Inversion-Recovery Radial FLASH and Motion-Resolved Model-Based Reconstruction

Xiaoqing Wang, Sebastian Rosenzweig+8 more|arXiv (Cornell University)|Nov 17, 2021

Advanced MRI Techniques and Applications42 references4 citations

TL;DR

This paper presents a free-breathing myocardial T₁ mapping technique using inversion-recovery radial FLASH and motion-resolved model-based reconstruction, enabling accurate, precise, and repeatable T₁ mapping in under two minutes without breath-holding. The method corrects for incomplete T₁ recovery between inversions, estimates respiratory motion from k-space data via SSA-FARY, and uses calibrationless, motion-resolved reconstruction with spatio-temporal total variation and sparsity constraints to improve image quality and parameter accuracy.

ABSTRACT

Purpose: To develop a free-breathing myocardial T1 mapping technique using inversion-recovery (IR) radial fast low-angle shot (FLASH) and calibrationless motion-resolved model-based reconstruction. Methods: Free-running (free-breathing, retrospective cardiac gating) IR radial FLASH is used for data acquisition at 3T. First, to reduce the waiting time between inversions, an analytical formula is derived that takes the incomplete T1 recovery into account for an accurate T1 calculation. Second, the respiratory motion signal is estimated from the k-space center of the contrast varying acquisition using an adapted singular spectrum analysis (SSA-FARY) technique. Third, a motion-resolved model-based reconstruction is used to estimate both parameter and coil sensitivity maps directly from the sorted k-space data. Thus, spatiotemporal total variation, in addition to the spatial sparsity constraints, can be directly applied to the parameter maps. Validations are performed on an experimental phantom, eleven human subjects, and a young landrace pig with myocardial infarction. Results: In comparison to an IR spin-echo reference, phantom results confirm good T1 accuracy when reducing the waiting time from five seconds to one second using the new correction. The motion-resolved model-based reconstruction further improves precision compared to the spatial regularization-only reconstruction. Aside from showing that a reliable respiratory motion signal can be estimated using modified SSA-FARY, in vivo studies demonstrate that dynamic myocardial T1 maps can be obtained within two minutes with good precision and repeatability. Conclusion: Motion-resolved myocardial T1 mapping during free-breathing with good accuracy, precision and repeatability can be achieved by combining inversion-recovery radial FLASH, self-gating and a calibrationless motion-resolved model-based reconstruction.

Motivation & Objective

• To develop a free-breathing T₁ mapping technique that eliminates the need for breath-holding, improving patient comfort and scan feasibility.
• To correct for incomplete T₁ recovery between inversion pulses when reducing the repetition time from 5 s to 1 s, improving scan efficiency.
• To estimate respiratory motion directly from k-space data using an adapted SSA-FARY method, enabling self-gating without external devices.
• To reconstruct motion-resolved T₁ maps with high precision by combining spatio-temporal total variation and sparsity constraints in a calibrationless model-based framework.
• To validate the method across phantom, healthy human, and animal models with myocardial infarction, demonstrating accuracy and repeatability.

Proposed method

• The method uses inversion-recovery radial FLASH for k-space data acquisition at 3T with free-breathing and retrospective cardiac gating.
• An analytical correction is derived to account for incomplete T₁ recovery during short repetition times, improving T₁ accuracy.
• Respiratory motion is estimated from the k-space center of the contrast-varying radial data using a modified singular spectrum analysis (SSA-FARY) technique.
• A motion-resolved model-based reconstruction is applied, directly estimating T₁ maps and coil sensitivity maps from sorted k-space data.
• The reconstruction employs spatio-temporal total variation and spatial sparsity constraints via an IRGNM-ADMM algorithm to enhance precision and reduce noise.
• The method is calibrationless, avoiding the need for separate coil sensitivity calibration scans.

Experimental results

Research questions

• RQ1 Can T₁ mapping accuracy be preserved when reducing the repetition time from 5 seconds to 1 second using an analytical correction for incomplete T₁ recovery?
• RQ2 Can respiratory motion be reliably estimated from the k-space center without external monitoring devices using an adapted SSA-FARY method?
• RQ3 Does motion-resolved model-based reconstruction with spatio-temporal total variation and sparsity constraints improve T₁ map precision compared to spatial regularization alone?
• RQ4 Can free-breathing T₁ mapping be achieved with high accuracy, precision, and repeatability in both healthy and pathological myocardial tissue within a clinically feasible scan time of under two minutes?
• RQ5 Is the proposed method robust across diverse imaging scenarios, including phantoms, healthy volunteers, and animal models with myocardial infarction?

Key findings

• Phantom studies confirmed that the proposed T₁ correction maintains high accuracy even when the repetition time is reduced from 5 s to 1 s.
• Motion-resolved model-based reconstruction significantly improved T₁ precision compared to spatial regularization-only reconstruction, particularly in dynamic tissue regions.
• The modified SSA-FARY method successfully estimated respiratory motion signals from k-space data, enabling effective self-gating without external hardware.
• In vivo studies demonstrated that dynamic myocardial T₁ maps could be acquired in less than two minutes with high precision and good repeatability in healthy human subjects.
• The method achieved reliable T₁ quantification in a young landrace pig with myocardial infarction, indicating potential for clinical patient applications.
• Reconstruction time was reduced to 25 minutes using GPU-accelerated implementation, though further optimization is needed for clinical workflow integration.