Publications

Background: Myocardial feature tracking (FT) provides a comprehensive analysis of myocardial deformation from cine balanced steady-state free-precession images (bSSFP). However, FT remains time-consuming, precluding its clinical adoption.
Purpose: To compare left-ventricular global radial strain (GRS) and global circumferential strain (GCS) values measured using automated DeepStrain analysis of short-axis cine images to those calculated using manual commercially available FT analysis.
Study Type: Retrospective, single-center.
Population: A total of 30 healthy subjects and 120 patients with cardiac disease for DeepStrain development. For evaluation, 47 healthy subjects (36 male, 53 5 years) and 533 patients who had undergone a clinical cardiac MRI (373 male, 59 ±14 years).
Field Strength/Sequence: bSSFP sequence at 1.5 T (Phillips) and 3 T (Siemens).
Assessment: Automated DeepStrain measurements of GRS and GCS were compared to commercially available FT (Circle, cvi42) measures obtained by readers with 1 year and 3 years of experience. Comparisons were performed overall and stratified by scanner manufacturer.
Statistical Tests: Paired t-test, linear regression slope, Pearson correlation coefficient (r).
Results: Overall, FT and DeepStrain measurements of GCS were not significantly different (P=0.207), but measures of GRS were significantly different. Measurements of GRS from Philips (slope =1.06 [1.03 1.08], r=0.85) and Siemens (slope =1.04 [0.99 1.09], r=0.83) data showed a very strong correlation and agreement between techniques. Measurements of GCS from Philips (slope =0.98 [0.98 1.01], r=0.91) and Siemens (slope =1.0 [0.96 1.03], r=0.88) data similarly showed a very strong correlation. The average analysis time per subject was 4.1 1.2 minutes for FT and 34.7 3.3 seconds for DeepStrain, representing a 7-fold reduction in analysis time.
Data Conclusion: This study demonstrated high correlation of myocardial GCS and GRS measurements between freely available fully automated DeepStrain and commercially available manual FT software, with substantial time-saving in the analysis.

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.

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.

 

Purpose: To improve the accuracy and robustness of T1 estimation by MyoMap-Net, a deep learning–based approach using 4 inversion-recovery T1-weighted images for cardiac T1 mapping.

Methods: MyoMapNet is a fully connected neural network for T1 estimation of an accelerated cardiac T1 mapping sequence, which collects 4 T1-weighted images by a single Look-Locker inversion-recovery experiment (LL4).MyoMap-Net was originally trained using in vivo data from the modified Look-Locker inversion recovery sequence, which resulted in significant bias and sensitivity to various confounders. This study sought to train MyoMapNet using signals generated from numerical simulations and phantom MR data under multiple simulated confounders. The trained model was then evaluated by phantom data scanned using new phantom vials that differed from those used for training. The performance of the new model was compared with modified Look-Locker inversion recovery sequence and saturation-recovery single-shot acquisition for measuring native and postcontrast T1 in 25 subjects.

Results: In the phantom study, T1 values measured by LL4 with MyoMapNet were highly correlated with reference values from the spin-echo sequence. Furthermore, the estimated T1 had excellent robustness to changes in flip angle and off-resonance. Native and postcontrast myocardium T1 at 3 Tesla measured by saturation-recovery single-shot acquisition, modified Look-Locker inversion recovery sequence, and MyoMapNet were 1483±46.6 ms and 791±45.8 ms, 1169±49.0 ms and 612±36.0 ms, and 1443±57.5 ms and 700±57.5 ms, respectively. The corresponding extracellular volumes were 22.90%±3.20%, 28.88%±3.48%, and 30.65%±3.60%, respectively.

Conclusion: Training MyoMapNet with numerical simulations and phantom data will improve the estimation of myocardial T1 values and increase its robustness to confounders while also reducing the overall T1 mapping estimation time to only 4 heartbeats.

Purpose: To develop and evaluate a free breathing non-electrocardiograph (ECG) myocardial T1* mapping sequence using radial imaging to quantify the changes in myocardial T1* between rest and exercise (T1*reactivity) in exercise cardiac MRI (Ex-CMR).
Methods: A free-running T1* sequence was developed using a saturation pulse followed by three Look-Locker inversion-recovery experiments. Each Look-Locker continuously acquired data as radial trajectory using a low flip-angle spoiled gradient-echo readout. Self-navigation was performed with a temporal resolution of∼100 ms for retrospectively extracting respiratory motion. The mid-diastole phase for every cardiac cycle was retrospectively detected on the recorded electrocardiogram signal using an empirical model. Multiple measurements were performed to obtain mean value to reduce effects from the free-breathing acquisition. Finally, data acquired at both mid-diastole and end-expiration are picked and reconstructed by a low-rank plus sparsity constraint algorithm. The performance of this sequence was evaluated by simulations, phantoms, and in vivo studies at rest and after physiological exercise.
Results: Numerical simulation demonstrated that changes in T1* are related to the changes in T1; however, other factors such as breathing motion could influence T1* measurements. Phantom T1* values measured using free-running T1* mapping sequence had good correlation with spin-echo T1 values and was insensitive to heart rate. In the Ex-CMR study, the measured T1* reactivity was 10% immediately after exercise and declined over time.

Conclusion: The free-running T1* mapping sequence allows free-breathing non-ECG quantification of changes in myocardial T1* with physiological exercise. Although, absolute myocardial T1* value is sensitive to various confounders

OBJECTIVES: We implemented an explainable machine learning (ML) model to gain insight into the association between cardiac magnetic resonance markers and adverse outcomes of cardiovascular hospitalization and all-cause death (composite endpoint) in patients with nonischemic dilated cardiomyopathy (NICM).

BACKGROUND: Risk stratification of patients with NICM remains challenging. An explainable ML model has the potential to provide insight into the contributions of different risk markers in the prediction model.

METHODS: An explainable ML model based on extreme gradient boosting (XGBoost) machines was developed using cardiac magnetic resonance and clinical parameters. The study cohorts consist of patients with NICM from 2 academic medical centers: Beth Israel Deaconess Medical Center (BIDMC) and Brigham and Women’s Hospital (BWH), with 328 and 214 patients, respectively. XGBoost was trained on 70% of patients from the BIDMC cohort and evaluated based on the other 30% as internal validation. The model was externally validated using the BWH cohort. To investigate the contribution
of different features in our risk prediction model, we used Shapley additive explanations (SHAP) analysis.

RESULTS: During a mean follow-up duration of 40 months, 34 patients from BIDMC and 33 patients from BWH experienced the composite endpoint. The area under the curve for predicting the composite endpoint was 0.71 for the internal BIDMC validation and 0.69 for the BWH cohort. SHAP analysis identified parameters associated with right ventricular
(RV) dysfunction and remodeling as primary markers of adverse outcomes. High risk thresholds were identified by SHAP analysis and thus provided thresholds for top predictive continuous clinical variables.

CONCLUSIONS: An explainable ML-based risk prediction model has the potential to identify patients with NICM at risk for cardiovascular hospitalization and all-cause death. RV ejection fraction, end-systolic and end-diastolic volumes (as indicators of RV dysfunction and remodeling) were determined to be major risk markers.

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.

Background: The role of cardiac magnetic resonance (CMR) in predictingoutcome in heart failure preserved ejection fraction (HFpEF) remains to be fully investigated. We sought to investigate CMR predictors for HF hospitalization in HFpEF patients using a machine learning risk model (ML).
Methods: In a retrospective study, we identified203 HFpEF patients (64±12 years of age, 48% women) who were referred for CMR. Left atrial (LA) and right ventricular (RV) strains were measured using CVI42® software. An explainable ML using XGBoost® was developed based on CMR and clinical data to predict future HF hospitalization as primary outcome. SHAP (SHapley Additive exPlanations) values were calculated to interpret contributions of different risk markers to outcome.
Results: During follow-up(50±39 months), 85 patients(42%) met the primary outcome. Demographics and ventricular functions were similar between groups with and without outcome (p>0.05). However, hospitalized patients had impaired LA (19.1±8.3% vs. 9.6±6.4%, p<0.001) and RV (-19.6±4.4% vs. -14.6±4.6%, p<0.001) strains. Figure 1A shows the performance of model with and without addition of strain data, demonstrating the incremental value of strain in predicting outcome. SHAP values demonstrated that LA and RV strains are the most prognostic predictors (Figure 1B&1C).
Conclusion: An explainable ML can identify HFpEF patients with high likelihood of hospitalization. SHAP analysis identifies LA and RV strains as major predictors of adverse outcome.