Accelerated CEST imaging through deep learning quantification from reduced frequency offsets.

PURPOSE: To shorten CEST acquisition time by leveraging Z-spectrum undersampling combined with deep learning for CEST map construction from undersampled Z-spectra. METHODS: Fisher information gain analysis identified optimal frequency offsets (termed "Fisher offsets") for the multi-pool fitting model, maximizing information gain for the amplitude and the FWHM parameters. These offsets guided initial subsampling levels. A U-NET, trained on undersampled brain CEST images from 18 volunteers, produced CEST maps at 3 T with varied undersampling levels. Feasibility was first tested using retrospective undersampling at three levels, followed by prospective in vivo undersampling (15 of 53 offsets), reducing scan time significantly. Additionally, glioblastoma grade IV pathology was simulated to evaluate network performance in patient-like cases. RESULTS: Traditional multi-pool models failed to quantify CEST maps from undersampled images (structural similarity index [SSIM] <0.2, peak SNR <20, Pearson r <0.1). Conversely, U-NET fitting successfully addressed undersampled data challenges. The study suggests CEST scan time reduction is feasible by undersampling 15, 25, or 35 of 53 Z-spectrum offsets. Prospective undersampling cut scan time by 3.5 times, with a maximum mean squared error of 4.4e-4, r = 0.82, and SSIM = 0.84, compared to the ground truth. The network also reliably predicted CEST values for simulated glioblastoma pathology. CONCLUSION: The U-NET architecture effectively quantifies CEST maps from undersampled Z-spectra at various undersampling levels.

Free-breathing 3D CEST MRI of human liver at 3.0 T.

PURPOSE: To develop a novel 3D abdominal CEST MRI technique at 3 T using MR multitasking, which enables entire-liver coverage with free-breathing acquisition. METHODS: k-Space data were continuously acquired with repetitive steady-state CEST (ss-CEST) modules. The stack-of-stars acquisition pattern was used for k-space sampling. MR multitasking was used to reconstruct motion-resolved 3D CEST images of 53 frequency offsets with entire-liver coverage and 2.0 × 2.0 × 6.0 mm3 spatial resolution. The total scan time was 9 min. The sensitivity of amide proton transfer (APT)-CEST (magnetization transfer asymmetry [MTRasym ] at 3.5 ppm) and glycogen CEST (glycoCEST) (mean MTRasym around 1.0 ppm) signals generated with the proposed method were tested with fasting experiments. RESULTS: Both APT-CEST and glycoCEST signals showed high sensitivity between post-fasting and post-meal acquisitions. APT-CEST and glycoCEST MTRasym signals from post-mean scans were significantly increased (APT-CEST: -0.019 ± 0.017 in post-fasting scans, 0.014 ± 0.021 in post-meal scans, p < 0.01; glycoCEST: 0.003 ± 0.009 in post-fasting scans, 0.027 ± 0.021 in post-meal scans, p < 0.01). CONCLUSION: The proposed 3D abdominal steady-state CEST method using MR multitasking can generate CEST images of the entire liver during free breathing.

Whole-brain steady-state CEST at 3 T using MR Multitasking.

PURPOSE: To perform fast 3D steady-state CEST (ss-CEST) imaging using MR Multitasking. METHODS: A continuous acquisition sequence with repetitive ss-CEST modules was developed. Each ss-CEST module contains a single-lobe Gaussian saturation pulse, followed by a spoiler gradient and eight FLASH readouts (one "training line" + seven "imaging lines"). Three-dimensional Cartesian encoding was used for k-space acquisition. Reconstructed CEST images were quantified with four-pool Lorentzian fitting. RESULTS: Steady-state CEST with whole-brain coverage was performed in 5.6 s per saturation frequency offset at the spatial resolution of 1.7 × 1.7 × 3.0 mm3 . The total scan time was 5.5 min for 55 different frequency offsets. Quantitative CEST maps from multipool fitting showed consistent image quality across the volume. CONCLUSION: Three-dimensional ss-CEST with whole-brain coverage can be done at 3 T within 5.5 min using MR Multitasking.

A Patch-Wise Deep Learning Approach for Myocardial Blood Flow Quantification with Robustness to Noise and Nonrigid Motion.

Quantitative analysis of dynamic contrast-enhanced cardiovascular MRI (cMRI) datasets enables the assessment of myocardial blood flow (MBF) for objective evaluation of ischemic heart disease in patients with suspected coronary artery disease. State-of-the-art MBF quantification techniques use constrained deconvolution and are highly sensitive to noise and motion-induced errors, which can lead to unreliable outcomes in the setting of high-resolution MBF mapping. To overcome these limitations, recent iterative approaches incorporate spatial-smoothness constraints to tackle pixel-wise MBF mapping. However, such iterative methods require a computational time of up to 30 minutes per acquired myocardial slice, which is a major practical limitation. Furthermore, they cannot enforce robustness to residual nonrigid motion which can occur in clinical stress/rest studies of patients with arrhythmia. We present a non-iterative patch-wise deep learning approach for pixel-wise MBF quantification wherein local spatio-temporal features are learned from a large dataset of myocardial patches acquired in clinical stress/rest cMRI studies. Our approach is scanner-independent, computationally efficient, robust to noise, and has the unique feature of robustness to motion-induced errors. Numerical and experimental results obtained using real patient data demonstrate the effectiveness of our approach.Clinical Relevance- The proposed patch-wise deep learning approach significantly improves the reliability of high-resolution myocardial blood flow quantification in cMRI by improving its robustness to noise and nonrigid myocardial motion and is up to 300-fold faster than state-of-the-art iterative approaches.