CineVN: Variational network reconstruction for rapid functional cardiac cine MRI.

PURPOSE: To develop a reconstruction method for highly accelerated cardiac cine MRI with high spatiotemporal resolution and low temporal blurring, and to demonstrate accurate estimation of ventricular volumes and myocardial strain in healthy subjects and in patients. METHODS: The proposed method, called CineVN, employs a spatiotemporal Variational Network combined with conjugate gradient descent for optimized data consistency and improved image quality. The method is first evaluated on retrospectively undersampled cine MRI data in terms of image quality. Then, prospectively accelerated data are acquired in 18 healthy subjects both segmented over two heartbeats per slice as well as in real time with 1.6 mm isotropic resolution. Ventricular volumes and strain parameters are computed and compared to a compressed sensing reconstruction and to a conventional reference cine MRI acquisition. Lastly, the method is demonstrated in 46 patients and ventricular volumes and strain parameters are evaluated. RESULTS: CineVN outperformed compressed sensing in image quality metrics on retrospectively undersampled data. Functional parameters and myocardial strain were the most accurate for CineVN compared to two state-of-the-art compressed sensing methods. CONCLUSION: Deep learning-based reconstruction using our proposed method enables accurate evaluation of cardiac function in real-time cine MRI with high spatiotemporal resolution. This has the potential to improve cardiac imaging particularly for patients with arrhythmia or impaired breath-hold capability.

Inline automatic quality control of 2D phase-contrast flow MRI for subject-specific scan time adaptation.

PURPOSE: To develop an inline automatic quality control to achieve consistent diagnostic image quality with subject-specific scan time, and to demonstrate this method for 2D phase-contrast flow MRI to reach a predetermined SNR. METHODS: We designed a closed-loop feedback framework between image reconstruction and data acquisition to intermittently check SNR (every 20 s) and automatically stop the acquisition when a target SNR is achieved. A free-breathing 2D pseudo-golden-angle spiral phase-contrast sequence was modified to listen for image-quality messages from the reconstructions. Ten healthy volunteers and 1 patient were imaged at 0.55 T. Target SNR was selected based on retrospective analysis of cardiac output error, and performance of the automatic SNR-driven "stop" was assessed inline. RESULTS: SNR calculation and automated segmentation was feasible within 20 s with inline deployment. The SNR-driven acquisition time was 2 min 39 s ± 67 s (aorta) and 3 min ± 80 s (main pulmonary artery) with a min/max acquisition time of 1 min 43 s/4 min 52 s (aorta) and 1 min 43 s/5 min 50 s (main pulmonary artery) across 6 healthy volunteers, while ensuring a diagnostic measurement with relative absolute error in quantitative flow measurement lower than 2.1% (aorta) and 6.3% (main pulmonary artery). CONCLUSION: The inline quality control enables subject-specific optimized scan times while ensuring consistent diagnostic image quality. The distribution of automated stopping times across the population revealed the value of a subject-specific scan time.

Increasing the scan-efficiency of pulmonary imaging at 0.55 T using iterative concomitant field and motion-corrected reconstruction.

PURPOSE: To develop an iterative concomitant field and motion corrected (iCoMoCo) reconstruction for isotropic high-resolution UTE pulmonary imaging at 0.55 T. METHODS: A free-breathing golden-angle stack-of-spirals UTE sequence was used to acquire data for 8 min with prototype and commercial 0.55 T MRI scanners. The data was binned into 12 respiratory phases based on superior-inferior navigator readouts. The previously published iterative motion corrected (iMoCo) reconstruction was extended to include concomitant field correction directly in the cost function. The reconstruction was implemented within the Gadgetron framework for inline reconstruction. Data were retrospectively reconstructed to simulate scan times of 2, 4, 6, and 8 min. Image quality was assessed using apparent SNR and image sharpness. The technique was evaluated in healthy volunteers and patients with known lung pathology including coronavirus disease 2019 infection, chronic granulomatous disease, lymphangioleiomyomatosis, and lung nodules. RESULTS: The technique provided diagnostic-quality images, and image quality was maintained with a slight loss in SNR for simulated scan times down to 4 min. Parenchymal apparent SNR was 4.33 ± 0.57, 5.96 ± 0.65, 7.36 ± 0.64, and 7.87 ± 0.65 using iCoMoCo with scan times of 2, 4, 6, and 8 min, respectively. Image sharpness at the diaphragm was comparable between iCoMoCo and reference images. Concomitant field corrections visibly improved the sharpness of anatomical structures away from the isocenter. Inline image reconstruction and artifact correction were achieved in <5 min. CONCLUSION: The proposed iCoMoCo pulmonary imaging technique can generate diagnostic quality images with 1.75 mm isotropic resolution in less than 5 min using a 6-min acquisition, on a 0.55 T scanner.

Quantifying 3D MR fingerprinting (3D-MRF) reproducibility across subjects, sessions, and scanners automatically using MNI atlases.

PURPOSE: Quantitative MRI techniques such as MR fingerprinting (MRF) promise more objective and comparable measurements of tissue properties at the point-of-care than weighted imaging. However, few direct cross-modal comparisons of MRF's repeatability and reproducibility versus weighted acquisitions have been performed. This work proposes a novel fully automated pipeline for quantitatively comparing cross-modal imaging performance in vivo via atlas-based sampling. METHODS: We acquire whole-brain 3D-MRF, turbo spin echo, and MPRAGE sequences three times each on two scanners across 10 subjects, for a total of 60 multimodal datasets. The proposed automated registration and analysis pipeline uses linear and nonlinear registration to align all qualitative and quantitative DICOM stacks to Montreal Neurological Institute (MNI) 152 space, then samples each dataset's native space through transformation inversion to compare performance within atlas regions across subjects, scanners, and repetitions. RESULTS: Voxel values within MRF-derived maps were found to be more repeatable (σT1 = 1.90, σT2 = 3.20) across sessions than vendor-reconstructed MPRAGE (σT1w = 6.04) or turbo spin echo (σT2w = 5.66) images. Additionally, MRF was found to be more reproducible across scanners (σT1 = 2.21, σT2 = 3.89) than either qualitative modality (σT1w = 7.84, σT2w = 7.76). Notably, differences between repeatability and reproducibility of in vivo MRF were insignificant, unlike the weighted images. CONCLUSION: MRF data from many sessions and scanners can potentially be treated as a single dataset for harmonized analysis or longitudinal comparisons without the additional regularization steps needed for qualitative modalities.

Accuracy of free-breathing multi-parametric SASHA in identifying T1 and T2 elevations in pediatric orthotopic heart transplant patients.

T1/T2 parametric mapping may reveal patterns of elevation ("hotspots") in myocardial diseases, such as rejection in orthotopic heart transplant (OHT) patients. This study aimed to evaluate the diagnostic accuracy of free-breathing (FB) multi-parametric SAturation recovery single-SHot Acquisition (mSASHA) T1/T2 mapping in identifying hotspots present on conventional Breath-held Modified Look-Locker Inversion recovery (BH MOLLI) T1 and T2-prepared balanced steady-state free-precession (BH T2p-bSSFP) maps in pediatric OHT patients. Pediatric OHT patients underwent noncontrast 1.5T CMR with BH MOLLI T1 and T2p-bSSFP and prototype FB mSASHA T1/T2 mapping in 8 short-axis slices. FB and BH T1/T2 hotspots were segmented using semi-automated thresholding (ITK-SNAP) and their 3D coordinate locations were collected (3-Matic, Materialise, Leuven, Belgium). Receiver operator characteristic curve analysis and measures of central tendency were utilized. 40 imaging datasets from 23 pediatric OHT patients were obtained. FB mSASHA yielded a sensitivity of 82.8% for T1 and 80% for T2 maps when compared to the standard BH MOLLI, as well as 100% specificity for both T1 and T2 maps. When identified on both FB and BH maps, hotspots overlapped in all cases, with an average long axis offset between FB and BH hotspot centers of 5.8 mm (IQR 3.5-8.2) on T1 and 5.9 mm (IQR 3.5-8.2) on T2 maps. FB mSASHA T1/T2 maps can identify hotspots present on conventional BH T1/T2 maps in pediatric patients with OHT, with high sensitivity, specificity, and overlap in 3D space. Free-breathing mapping may improve patient comfort and facilitate OHT assessment in younger patient populations.

The impact of water exchange on estimates of myocardial extracellular volume calculated using contrast enhanced T1 measurements: A preliminary analysis in patients with severe aortic stenosis.

PURPOSE: Guidelines recommend measuring myocardial extracellular volume (ECV) using T1 -mapping before and 10-30 min after contrast agent administration. Data are then analyzed using a linear model (LM), which assumes fast water exchange (WX) between the ECV and cardiomyocytes. We investigated whether limited WX influences ECV measurements in patients with severe aortic stenosis (AS). METHODS: Twenty-five patients with severe AS and 5 healthy controls were recruited. T1 measurements were made on a 3 T Siemens system using a multiparametric saturation-recovery single-shot acquisition (a) before contrast; (b) 4 min post 0.05 mmol/kg gadobutrol; and (c) 4 min, (d) 10 min, and (e) 30 min after an additional gadobutrol dose (0.1 mmol/kg). Three LM-based ECV estimates, made using paired T1 measurements (a and b), (a and d), and (a and e), were compared to ECV estimates made using all 5 T1 measurements and a two-site exchange model (2SXM) accounting for WX. RESULTS: Median (range) ECV estimated using the 2SXM model was 25% (21%-39%) for patients and 26% (22%-29%) for controls. ECV estimated in patients using the LM at 10 min following a cumulative contrast dose of 0.15 mmol/kg was 21% (17%-32%) and increased significantly to 22% (19%-35%) at 30 min (p = 0.0001). ECV estimated using the LM was highest following low dose gadobutrol, 25% (19%-38%). CONCLUSION: Current guidelines on contrast agent dose for ECV measurements may lead to underestimated ECV in patients with severe AS because of limited WX. Use of a lower contrast agent dose may mitigate this effect.

Left atrial remodeling in hypertrophic cardiomyopathy and Fabry disease: A CMR-based head-to-head comparison and outcome analysis.

BACKGROUND: Hypertrophic cardiomyopathy (HCM) and Fabry disease cardiomyopathy (FD) are phenocopies, as they show left ventricular hypertrophy (LVH). The left atrium (LA) is emerging as a potential marker of disease severity in both cardiomyopathies. The present study compares HCM and FD cardiomyopathy with similar degree of LVH, exploring LA morpho-functional parameters and the correlates of clinical outcome. METHODS: We performed a comprehensive CMR-based comparison between 30 HCM and 30 FD patients matched on age, sex, BSA, LV mass and major cardiovascular risk factors affecting LA remodeling (arterial hypertension and diabetes). 30 healthy controls were also included. CMR feature tracking (CMR-FT) analysis, T1 mapping and conventional parameters were evaluated. Patients also underwent transthoracic echocardiography for LV diastolic function assessment. Clinical events at follow-up were collected (atrial and ventricular events, bradyarrhythmia, heart failure (HF) hospitalization and death). RESULTS: HCM patients showed greater LA remodeling compared to FD patients, namely higher LA end-systolic volume index (LAVi max), lower LA-ejection fraction (LA-EF) and worse reservoir (εs) and booster function (εa) (all p < 0.05). Accordingly, these parameters have demonstrated good potential for distinguishing between FD and HCM (AUC 0.68-0.73, all p < 0.05), with LAVi max being an independent predictor for HCM diagnosis (OR 1.07, 95%CI 1.011-1.132, p 0.02). Moreover, in HCM patients a significant association between εs and HF occurrence was observed at 2-year follow-up (OR 0.85, 95%CI 0.72-0.99, p 0.04). CONCLUSIONS: In HCM, LA remodeling is greater than in FD cardiomyopathy with similar LVH, and reservoir strain is associated with HF at follow-up.

Accelerated Cardiac MRI Cine with Use of Resolution Enhancement Generative Adversarial Inline Neural Network.

Background Cardiac cine can benefit from deep learning-based image reconstruction to reduce scan time and/or increase spatial and temporal resolution. Purpose To develop and evaluate a deep learning model that can be combined with parallel imaging or compressed sensing (CS). Materials and Methods The deep learning model was built on the enhanced super-resolution generative adversarial inline neural network, trained with use of retrospectively identified cine images and evaluated in participants prospectively enrolled from September 2021 to September 2022. The model was applied to breath-hold electrocardiography (ECG)-gated segmented and free-breathing real-time cine images collected with reduced spatial resolution with use of generalized autocalibrating partially parallel acquisitions (GRAPPA) or CS. The deep learning model subsequently restored spatial resolution. For comparison, GRAPPA-accelerated cine images were collected. Diagnostic quality and artifacts were evaluated by two readers with use of Likert scales and compared with use of Wilcoxon signed-rank tests. Agreement for left ventricle (LV) function, volume, and strain was assessed with Bland-Altman analysis. Results The deep learning model was trained on 1616 patients (mean age ± SD, 56 years ± 16; 920 men) and evaluated in 181 individuals, 126 patients (mean age, 57 years ± 16; 77 men) and 55 healthy subjects (mean age, 27 years ± 10; 15 men). In breath-hold ECG-gated segmented cine and free-breathing real-time cine, the deep learning model and GRAPPA showed similar diagnostic quality scores (2.9 vs 2.9, P = .41, deep learning vs GRAPPA) and artifact score (4.4 vs 4.3, P = .55, deep learning vs GRAPPA). Deep learning acquired more sections per breath-hold than GRAPPA (3.1 vs one section, P < .001). In free-breathing real-time cine, the deep learning showed a similar diagnostic quality score (2.9 vs 2.9, P = .21, deep learning vs GRAPPA) and lower artifact score (3.9 vs 4.3, P < .001, deep learning vs GRAPPA). For both sequences, the deep learning model showed excellent agreement for LV parameters, with near-zero mean differences and narrow limits of agreement compared with GRAPPA. Conclusion Deep learning-accelerated cardiac cine showed similarly accurate quantification of cardiac function, volume, and strain to a standardized parallel imaging method. © RSNA, 2023 Supplemental material is available for this article. See also the editorial by Vannier and Wang in this issue.

Automated Inline Myocardial Segmentation of Joint T1 and T2 Mapping Using Deep Learning.

Purpose: To develop an artificial intelligence (AI) solution for automated segmentation and analysis of joint cardiac MRI short-axis T1 and T2 mapping. Materials and Methods: In this retrospective study, a joint T1 and T2 mapping sequence was used to acquire 4240 maps from 807 patients across two hospitals between March and November 2020. Five hundred nine maps from 94 consecutive patients were assigned to a holdout testing set. A convolutional neural network was trained to segment the endocardial and epicardial contours with use of an edge probability estimation approach. Training labels were segmented by an expert cardiologist. Predicted contours were processed to yield mapping values for each of the 16 American Heart Association segments. Network segmentation performance and segment-wise measurements on the testing set were compared with those of two experts on the holdout testing set. The AI model was fully integrated using open-source software to run on MRI scanners. Results: A total of 3899 maps (92%) were deemed artifact-free and suitable for human segmentation. AI segmentation closely matched that of each expert (mean Dice coefficient, 0.82 ± 0.07 [SD] vs expert 1 and 0.86 ± 0.06 vs expert 2) and compared favorably with interexpert agreement (Dice coefficient, 0.84 ± 0.06 for expert 1 vs expert 2). AI-derived segment-wise values for native T1, postcontrast T1, and T2 mapping correlated with expert-derived values (R 2 = 0.96, 0.98, and 0.87, respectively, vs expert 1, and 0.97, 0.99, and 0.92 vs expert 2) and fell within the range of interexpert reproducibility (R 2 = 0.97, 0.99, and 0.90, respectively). The AI model has since been deployed at two hospitals, enabling automated inline analysis. Conclusion: Automated inline analysis of joint T1 and T2 mapping allows accurate segment-wise tissue characterization, with performance equivalent to that of human experts.Keywords: MRI, Neural Networks, Cardiac, Heart Supplemental material is available for this article. © RSNA, 2022.

Effect of Migalastat on cArdiac InvOlvement in FabRry DiseAse: MAIORA study.

BACKGROUND: A small but significant reduction in left ventricular (LV) mass after 18 months of migalastat treatment has been reported in Fabry disease (FD). This study aimed to assess the effect of migalastat on FD cardiac involvement, combining LV morphology and tissue characterisation by cardiac magnetic resonance (CMR) with cardiopulmonary exercise testing (CPET). METHODS: Sixteen treatment-naïve patients with FD (4 women, 46.4±16.2 years) with cardiac involvement (reduced T1 values on CMR and/or LV hypertrophy) underwent ECG, echocardiogram, troponin T and NT-proBNP (N-Terminal prohormone of Brain Natriuretic Peptide) assay, CMR with T1 mapping, and CPET before and after 18 months of migalastat. RESULTS: No change in LV mass was detected at 18 months compared to baseline (95.2 g/m2 (66.0-184.0) vs 99.0 g/m2 (69.0-121.0), p=0.55). Overall, there was an increase in septal T1 of borderline significance (870.0 ms (848-882) vs 860.0 ms (833.0-875.0), p=0.056). Functional capacity showed an increase in oxygen consumption (VO2) at anaerobic threshold (15.50 mL/kg/min (13.70-21.50) vs 14.50 mL/kg/min (11.70-18.95), p=0.02), and a trend towards an increase in percent predicted peak VO2 (72.0 (63.0-80.0) vs 69.0 (53.0-77.0), p=0.056) was observed. The subset of patients who showed an increase in T1 value and a reduction in LV mass (n=7, 1 female, age 40.5 (28.6-76.0)) was younger and at an earlier disease stage compared to the others, and also exhibited greater improvement in exercise tolerance. CONCLUSION: In treatment-naïve FD patients with cardiac involvement, 18-month treatment with migalastat stabilised LV mass and was associated with a trend towards an improvement in exercise tolerance. A tendency to T1 increase was detected by CMR. The subset of patients who had significant benefits from the treatment showed an earlier cardiac disease compared to the others. TRIAL REGISTRATION NUMBER: NCT03838237.

Hemodynamic Evaluation of Type B Aortic Dissection Using Compressed Sensing Accelerated 4D Flow MRI.

BACKGROUND: 4D Flow MRI is a quantitative imaging technique to evaluate blood flow patterns; however, it is unclear how compressed sensing (CS) acceleration would impact aortic hemodynamic quantification in type B aortic dissection (TBAD). PURPOSE: To investigate CS-accelerated 4D Flow MRI performance compared to GRAPP-accelerated 4D Flow MRI (GRAPPA) to evaluate aortic hemodynamics in TBAD. STUDY TYPE: Prospective. POPULATION: Twelve TBAD patients, two volunteers. FIELD STRENGTH/SEQUENCE: 1.5T, 3D time-resolved cine phase-contrast gradient echo sequence. ASSESSMENT: GRAPPA (acceleration factor [R] = 2) and two CS-accelerated (R = 7.7 [CS7.7] and 10.2 [CS10.2]) 4D Flow MRI scans were acquired twice for interscan reproducibility assessment. Voxelwise kinetic energy (KE), peak velocity (PV), forward flow (FF), reverse flow (RF), and stasis were calculated. Plane-based mid-lumen flows were quantified. Imaging times were recorded. TESTS: Repeated measures analysis of variance, Pearson correlation coefficients (r), intraclass correlation coefficients (ICC). P < 0.05 indicated statistical significance. RESULTS: The KE and FF in true lumen (TL) and PV in false lumen (FL) did not show difference among three acquisition types (P = 0.818, 0.065, 0.284 respectively). The PV and stasis in TL were higher, KE, FF, and RF in FL were lower, and stasis was higher in GRAPPA compared to CS7.7 and CS10.2. The RF was lower in GRAPPA compared to CS10.2. The correlation coefficients were strong in TL (r = [0.781-0.986]), and low to strong in FL (r = [0.347-0.948]). The ICC levels demonstrated moderate to excellent interscan reproducibility (0.732-0.989). The FF and net flow in mid-descending aorta TL were significantly different between CS7.7 and CS10.2. CONCLUSION: CS-accelerated 4D Flow MRI has potential for clinical utilization with shorter scan times in TBAD. Our results suggest similar hemodynamic trends between acceleration types, but CS-acceleration impacts KE, FF, RF, and stasis more in FL. EVIDENCE LEVEL: 1 Technical Efficacy: Stage 2.

Open-source MR imaging and reconstruction workflow.

PURPOSE: This work presents an end-to-end open-source MR imaging workflow. It is highly flexible in rapid prototyping across the whole imaging process and integrates vendor-independent openly available tools. The whole workflow can be shared and executed on different MR platforms. It is also integrated in the JEMRIS simulation framework, which makes it possible to generate simulated data from the same sequence that runs on the MRI scanner using the same pipeline for image reconstruction. METHODS: MRI sequences can be designed in Python or JEMRIS using the Pulseq framework, allowing simplified integration of new sequence design tools. During the sequence design process, acquisition metadata required for reconstruction is stored in the MR raw data format. Data acquisition is possible on MRI scanners supported by Pulseq and in simulations through JEMRIS. An image reconstruction and postprocessing pipeline was implemented into a Python server that allows real-time processing of data as it is being acquired. The Berkeley Advanced Reconstruction Toolbox is integrated into this framework for image reconstruction. The reconstruction pipeline supports online integration through a vendor-dependent interface. RESULTS: The flexibility of the workflow is demonstrated with different examples, containing 3D parallel imaging with controlled aliasing in volumetric parallel imaging (CAIPIRINHA) acceleration, spiral imaging, and B0 mapping. All sequences, data, and the corresponding processing pipelines are publicly available. CONCLUSION: The proposed workflow is highly flexible and allows integration of advanced tools at all stages of the imaging process. All parts of this workflow are open-source, simplifying collaboration across different MR platforms or sites and improving reproducibility of results.

An inline deep learning based free-breathing ECG-free cine for exercise cardiovascular magnetic resonance.

BACKGROUND: Exercise cardiovascular magnetic resonance (Ex-CMR) is a promising stress imaging test for coronary artery disease (CAD). However, Ex-CMR requires accelerated imaging techniques that result in significant aliasing artifacts. Our goal was to develop and evaluate a free-breathing and electrocardiogram (ECG)-free real-time cine with deep learning (DL)-based radial acceleration for Ex-CMR. METHODS: A 3D (2D + time) convolutional neural network was implemented to suppress artifacts from aliased radial cine images. The network was trained using synthetic real-time radial cine images simulated using breath-hold, ECG-gated segmented Cartesian k-space data acquired at 3 T from 503 patients at rest. A prototype real-time radial sequence with acceleration rate = 12 was used to collect images with inline DL reconstruction. Performance was evaluated in 8 healthy subjects in whom only rest images were collected. Subsequently, 14 subjects (6 healthy and 8 patients with suspected CAD) were prospectively recruited for an Ex-CMR to evaluate image quality. At rest (n = 22), standard breath-hold ECG-gated Cartesian segmented cine and free-breathing ECG-free real-time radial cine images were acquired. During post-exercise stress (n = 14), only real-time radial cine images were acquired. Three readers evaluated residual artifact level in all collected images on a 4-point Likert scale (1-non-diagnostic, 2-severe, 3-moderate, 4-minimal). RESULTS: The DL model substantially suppressed artifacts in real-time radial cine images acquired at rest and during post-exercise stress. In real-time images at rest, 89.4% of scores were moderate to minimal. The mean score was 3.3 ± 0.7, representing increased (P < 0.001) artifacts compared to standard cine (3.9 ± 0.3). In real-time images during post-exercise stress, 84.6% of scores were moderate to minimal, and the mean artifact level score was 3.1 ± 0.6. Comparison of left-ventricular (LV) measures derived from standard and real-time cine at rest showed differences in LV end-diastolic volume (3.0 mL [- 11.7, 17.8], P = 0.320) that were not significantly different from zero. Differences in measures of LV end-systolic volume (7.0 mL [- 1.3, 15.3], P < 0.001) and LV ejection fraction (- 5.0% [- 11.1, 1.0], P < 0.001) were significant. Total inline reconstruction time of real-time radial images was 16.6 ms per frame. CONCLUSIONS: Our proof-of-concept study demonstrated the feasibility of inline real-time cine with DL-based radial acceleration for Ex-CMR.

Impact of deep learning architectures on accelerated cardiac T1 mapping using MyoMapNet.

The objective of the current study was to investigate the performance of various deep learning (DL) architectures for MyoMapNet, a DL model for T1 estimation using accelerated cardiac T1 mapping from four T1 -weighted images collected after a single inversion pulse (Look-Locker 4 [LL4]). We implemented and tested three DL architectures for MyoMapNet: (a) a fully connected neural network (FC), (b) convolutional neural networks (VGG19, ResNet50), and (c) encoder-decoder networks with skip connections (ResUNet, U-Net). Modified Look-Locker inversion recovery (MOLLI) images from 749 patients at 3 T were used for training, validation, and testing. The first four T1 -weighted images from MOLLI5(3)3 and/or MOLLI4(1)3(1)2 protocols were extracted to create accelerated cardiac T1 mapping data. We also prospectively collected data from 28 subjects using MOLLI and LL4 to further evaluate model performance. Despite rigorous training, conventional VGG19 and ResNet50 models failed to produce anatomically correct T1 maps, and T1 values had significant errors. While ResUNet yielded good quality maps, it significantly underestimated T1 . Both FC and U-Net, however, yielded excellent image quality with good T1 accuracy for both native (FC/U-Net/MOLLI = 1217 ± 64/1208 ± 61/1199 ± 61 ms, all p < 0.05) and postcontrast myocardial T1 (FC/U-Net/MOLLI = 578 ± 57/567 ± 54/574 ± 55 ms, all p < 0.05). In terms of precision, the U-Net model yielded better T1 precision compared with the FC architecture (standard deviation of 61 vs. 67 ms for the myocardium for native [p < 0.05], and 31 vs. 38 ms [p < 0.05], for postcontrast). Similar findings were observed in prospectively collected LL4 data. It was concluded that U-Net and FC DL models in MyoMapNet enable fast myocardial T1 mapping using only four T1 -weighted images collected from a single LL sequence with comparable accuracy. U-Net also provides a slight improvement in precision.

Improved accuracy and precision with three-parameter simultaneous myocardial T1 and T2 mapping using multiparametric SASHA.

PURPOSE: To develop and validate a three-parameter model for improved precision multiparametric SAturation-recovery single-SHot Acquisition (mSASHA) cardiac T1 and T2 mapping with high accuracy in a single breath-hold. METHODS: The mSASHA acquisition consists of nine images of variable saturation recovery and T2 preparation in 11 heartbeats with T1 and T2 values calculated using a three-parameter model. It was validated in simulations and phantoms at 3 T with comparison to a four-parameter joint T1 -T2 technique. The mSASHA acquisition was compared with MOLLI, SASHA, and T2 -prepared balanced SSFP in 10 volunteers. RESULTS: The mSASHA technique had high accuracy in phantoms compared to spin echo, with -0.2 ± 0.3% T1 error and -2.4 ± 1.3% T2 error. The mSASHA coefficient of variation in phantoms for T1 was similar to MOLLI (0.7 ± 0.2% for both) and T2 -prepared balanced SSFP for T2 (1.3 ± 0.7% vs 1.4 ± 0.3%, adjusted p > .05 for both). In simulations, three-parameter mSASHA had higher precision than four-parameter joint T1 -T2 for both T1 and T2 (46% and 11% reductions in T1 and T2 interquartile range for native myocardium). In vivo myocardial mSASHA T1 was similar to SASHA (1523 ± 18 ms vs 1520 ± 18 ms) with similar coefficient of variation to both MOLLI and SASHA (3.3 ± 0.6% vs 3.1 ± 0.6% and 3.3 ± 0.5% respectively, adjusted p > .05 for all). Myocardial mSASHA T2 was 37.1 ± 1.1 ms with similar precision to T2 -prepared balanced SSFP (6.7 ± 1.7% vs 6.0 ± 1.6%, adjusted p > .05). CONCLUSION: Three-parameter mSASHA provides high-accuracy cardiac T1 and T2 quantification in a single breath-hold with similar precision to MOLLI and T2 -prepared balanced SSFP. Further study is required to both establish normative values and demonstrate clinical utility in patient populations.

Accelerated cardiac T1 mapping in four heartbeats with inline MyoMapNet: a deep learning-based T1 estimation approach.

PURPOSE: To develop and evaluate MyoMapNet, a rapid myocardial T1 mapping approach that uses fully connected neural networks (FCNN) to estimate T1 values from four T1-weighted images collected after a single inversion pulse in four heartbeats (Look-Locker, LL4). METHOD: We implemented an FCNN for MyoMapNet to estimate T1 values from a reduced number of T1-weighted images and corresponding inversion-recovery times. We studied MyoMapNet performance when trained using native, post-contrast T1, or a combination of both. We also explored the effects of number of T1-weighted images (four and five) for native T1. After rigorous training using in-vivo modified Look-Locker inversion recovery (MOLLI) T1 mapping data of 607 patients, MyoMapNet performance was evaluated using MOLLI T1 data from 61 patients by discarding the additional T1-weighted images. Subsequently, we implemented a prototype MyoMapNet and LL4 on a 3 T scanner. LL4 was used to collect T1 mapping data in 27 subjects with inline T1 map reconstruction by MyoMapNet. The resulting T1 values were compared to MOLLI. RESULTS: MyoMapNet trained using a combination of native and post-contrast T1-weighted images had excellent native and post-contrast T1 accuracy compared to MOLLI. The FCNN model using four T1-weighted images yields similar performance compared to five T1-weighted images, suggesting that four T1 weighted images may be sufficient. The inline implementation of LL4 and MyoMapNet enables successful acquisition and reconstruction of T1 maps on the scanner. Native and post-contrast myocardium T1 by MOLLI and MyoMapNet was 1170 ± 55 ms vs. 1183 ± 57 ms (P = 0.03), and 645 ± 26 ms vs. 630 ± 30 ms (P = 0.60), and native and post-contrast blood T1 was 1820 ± 29 ms vs. 1854 ± 34 ms (P = 0.14), and 508 ± 9 ms vs. 514 ± 15 ms (P = 0.02), respectively. CONCLUSION: A FCNN, trained using MOLLI data, can estimate T1 values from only four T1-weighted images. MyoMapNet enables myocardial T1 mapping in four heartbeats with similar accuracy as MOLLI with inline map reconstruction.

Left Ventricular Fibrosis Assessment by Native T1, ECV, and LGE in Pulmonary Hypertension Patients.

Cardiac magnetic resonance imaging (MRI) is emerging as an alternative to right heart catheterization for the evaluation of pulmonary hypertension (PH) patients. The aim of this study was to compare cardiac MRI-derived left ventricle fibrosis indices between pre-capillary PH (PrePH) and isolated post-capillary PH (IpcPH) patients and assess their associations with measures of ventricle function. Global and segmental late gadolinium enhancement (LGE), longitudinal relaxation time (native T1) maps, and extracellular volume fraction (ECV) were compared among healthy controls (N = 25; 37% female; 52 ± 13 years), PH patients (N = 48; 60% female; 60 ± 14 years), and PH subgroups (PrePH: N = 29; 65% female; 55 ± 12 years, IpcPH: N = 19; 53% female; 66 ± 13 years). Cardiac cine measured ejection fraction, end diastolic, and end systolic volumes and were assessed for correlations with fibrosis. LGE mural location was qualitatively assessed on a segmental basis for all subjects. PrePH patients had elevated (apical-, mid-antero-, and mid-infero) septal left ventricle native T1 values (1080 ± 74 ms, 1077 ± 39 ms, and 1082 ± 47 ms) compared to IpcPH patients (1028 ± 53 ms, 1046 ± 36 ms, 1051 ± 44 ms) (p < 0.05). PrePH had a higher amount of insertional point LGE (69%) and LGE patterns characteristic of non-vascular fibrosis (77%) compared to IpcPH (37% and 46%, respectively) (p < 0.05; p < 0.05). Assessment of global LGE, native T1, and ECV burdens did not show a statistically significant difference between PrePH (1.9 ± 2.7%, 1056.2 ± 36.3 ms, 31.2 ± 3.7%) and IpcPH (2.7 ± 2.7%, 1042.4 ± 28.1 ms, 30.7 ± 4.7%) (p = 0.102; p = 0.229 p = 0.756). Global native T1 and ECV were higher in patients (1050.9 ± 33.8 and 31.0 ± 4.1%) than controls (28.2 ± 3.7% and 1012.9 ± 29.4 ms) (p < 0.05). Cardiac MRI-based tissue characterization may augment understanding of cardiac involvement and become a tool to facilitate PH patient classification.

Cardiac and cardiometabolic phenotyping of trastuzumab-mediated cardiotoxicity: a secondary analysis of the MANTICORE trial.

AIMS: An improved understanding of the pathophysiology of trastuzumab-mediated cardiotoxicity is required to improve outcomes of patients with human epidermal growth factor receptor 2 (HER2)-positive breast cancer. We aimed to characterize the cardiac and cardiometabolic phenotype of trastuzumab-mediated toxicity and potential interactions with cardiac pharmacotherapy. METHODS AND RESULTS: This study was an analysis of serial magnetic resonance imaging (MRI) and circulating biomarker data acquired from patients with HER2-positive early-stage breast cancer participating in a randomized-controlled clinical trial for the pharmaco-prevention of trastuzumab-associated cardiotoxicity. Circulating biomarkers (B-type natriuretic peptide, troponin I, MMP-2 and -9, GDF-15, neuregulin-1, and IGF-1) and MRI of cardiac structure and function and abdominal fat distribution were acquired prior to trastuzumab, post-cycle 4 and post-cycle 17. Ninety-four participants (51 ± 8 years) completed the study with 30 on placebo, 33 on perindopril, and 31 on bisoprolol. Post-cycle 4, global longitudinal strain deteriorated from baseline in both placebo (+2.0 ± 2.7%, P = 0.002) and perindopril (+0.9 ± 2.5%, P = 0.04), but not with bisoprolol (-0.2 ± 2.1%, P = 0.55). In all groups combined, extracellular volume fraction and GDF-15 increased post-cycle 4 (+1.3 ± 4.4%, P = 0.004; +130 ± 150%, P ≤ 0.001, respectively). However, no significant change in troponin I was detected throughout trastuzumab. In all groups combined, visceral and intermuscular fat volume increased post-cycle 4 (+7 ± 17%, P = 0.02, +8 ± 23%, P = 0.02, respectively), while muscle volume and IGF-1 decreased from post-cycle 4 to 17 (-2 ± 10%, P = 0.008, -18 ± 28%, P < 0.001, respectively). CONCLUSION: Trastuzumab results in impaired cardiac function and early myocardial inflammation. Trastuzumab is also associated with deleterious changes to the cardiometabolic phenotype which may contribute to the increased cardiovascular risk in this population.

Trabecular complexity as an early marker of cardiac involvement in Fabry disease.

AIMS: Fabry cardiomyopathy is characterized by glycosphingolipid storage and increased myocardial trabeculation has also been demonstrated. This study aimed to explore by cardiac magnetic resonance whether myocardial trabecular complexity, quantified by endocardial border fractal analysis, tracks phenotype evolution in Fabry cardiomyopathy. METHODS AND RESULTS: Study population included 20 healthy controls (12 males, age 32±9) and 45 Fabry patients divided into three groups: 15 left ventricular hypertrophy (LVH)-negative patients with normal T1 (5 males, age 28±13; Group 1); 15 LVH-negative patients with low T1 (9 males, age 33±9.6; Group 2); 15 LVH-positive patients (11 males, age 53.5±9.6; Group 3). Trabecular fractal dimensions (Dfs) (total, basal, mid-ventricular, and apical) were evaluated on cine images. Total Df was higher in all Fabry groups compared to controls, gradually increasing from controls to Group 3 (1.27±0.02 controls vs. 1.29±0.02 Group 1 vs. 1.30±0.02 Group 2 vs. 1.34±0.02 Group 3; P<0.001). Group 3 showed significantly higher values of all Dfs compared to the other Groups. Both basal and total Dfs were significantly higher in Group 1 compared with controls (basal: 1.30±0.03 vs. 1.26±0.04, P =0.010; total: 1.29±0.02 vs. 1.27±0.02, P=0.044). Total Df showed significant correlations with: (i) T1 value (r=-0.569; P<0.001); (ii) LV mass (r=0.664, P<0.001); (iii) trabecular mass (r=0.676; P <0.001); (iv) Mainz Severity Score Index (r=0.638; P<0.001). CONCLUSION: Fabry cardiomyopathy is characterized by a progressive increase in Df of endocardial trabeculae together with shortening of T1 values. Myocardial trabeculation is increased before the presence of detectable sphingolipid storage, thus representing an early sign of cardiac involvement.