Deep Learning Synthesis of White-Blood From Dark-Blood Late Gadolinium Enhancement Cardiac Magnetic Resonance.

OBJECTIVES: Dark-blood late gadolinium enhancement (DB-LGE) cardiac magnetic resonance has been proposed as an alternative to standard white-blood LGE (WB-LGE) imaging protocols to enhance scar-to-blood contrast without compromising scar-to-myocardium contrast. In practice, both DB and WB contrasts may have clinical utility, but acquiring both has the drawback of additional acquisition time. The aim of this study was to develop and evaluate a deep learning method to generate synthetic WB-LGE images from DB-LGE, allowing the assessment of both contrasts without additional scan time. MATERIALS AND METHODS: DB-LGE and WB-LGE data from 215 patients were used to train 2 types of unpaired image-to-image translation deep learning models, cycle-consistent generative adversarial network (CycleGAN) and contrastive unpaired translation, with 5 different loss function hyperparameter settings each. Initially, the best hyperparameter setting was determined for each model type based on the Fréchet inception distance and the visual assessment of expert readers. Then, the CycleGAN and contrastive unpaired translation models with the optimal hyperparameters were directly compared. Finally, with the best model chosen, the quantification of scar based on the synthetic WB-LGE images was compared with the truly acquired WB-LGE. RESULTS: The CycleGAN architecture for unpaired image-to-image translation was found to provide the most realistic synthetic WB-LGE images from DB-LGE images. The results showed that it was difficult for visual readers to distinguish if an image was true or synthetic (55% correctly classified). In addition, scar burden quantification with the synthetic data was highly correlated with the analysis of the truly acquired images. Bland-Altman analysis found a mean bias in percentage scar burden between the quantification of the real WB and synthetic white-blood images of 0.44% with limits of agreement from -10.85% to 11.74%. The mean image quality of the real WB images (3.53/5) was scored higher than the synthetic white-blood images (3.03), P = 0.009. CONCLUSIONS: This study proposed a CycleGAN model to generate synthetic WB-LGE from DB-LGE images to allow assessment of both image contrasts without additional scan time. This work represents a clinically focused assessment of synthetic medical images generated by artificial intelligence, a topic with significant potential for a multitude of applications. However, further evaluation is warranted before clinical adoption.

Hybrid artificial intelligence outcome prediction using features extraction from stress perfusion cardiac magnetic resonance images and electronic health records.

Background: Prediction of clinical outcomes in coronary artery disease (CAD) has been conventionally achieved using clinical risk factors. The relationship between imaging features and outcome is still not well understood. This study aims to use artificial intelligence to link image features with mortality outcome. Methods: A retrospective study was performed on patients who had stress perfusion cardiac magnetic resonance (SP-CMR) between 2011 and 2021. The endpoint was all-cause mortality. Convolutional neural network (CNN) was used to extract features from stress perfusion images, and multilayer perceptron (MLP) to extract features from electronic health records (EHRs), both networks were concatenated in a hybrid neural network (HNN) to predict study endpoint. Image CNN was trained to predict study endpoint directly from images. HNN and image CNN were compared with a linear clinical model using area under the curve (AUC), F1 scores, and McNemar's test. Results: Total of 1,286 cases were identified, with 201 death events (16%). The clinical model had good performance (AUC =80%, F1 score =37%). Best Image CNN model showed AUC =72% and F1 score =38%. HNN outperformed the other two models (AUC =82%, F1 score =43%). McNemar's test showed statistical difference between image CNN and both clinical model (P<0.01) and HNN (P<0.01). There was no significant difference between HNN and clinical model (P=0.15). Conclusions: Death in patients with suspected or known CAD can be predicted directly from stress perfusion images without clinical knowledge. Prediction can be improved by HNN that combines clinical and SP-CMR images.

Automated inversion time selection for late gadolinium-enhanced cardiac magnetic resonance imaging.

OBJECTIVES: To develop and share a deep learning method that can accurately identify optimal inversion time (TI) from multi-vendor, multi-institutional and multi-field strength inversion scout (TI scout) sequences for late gadolinium enhancement cardiac MRI. MATERIALS AND METHODS: Retrospective multicentre study conducted on 1136 1.5-T and 3-T cardiac MRI examinations from four centres and three scanner vendors. Deep learning models, comprising a convolutional neural network (CNN) that provides input to a long short-term memory (LSTM) network, were trained on TI scout pixel data from centres 1 to 3 to identify optimal TI, using ground truth annotations by two readers. Accuracy within 50 ms, mean absolute error (MAE), Lin's concordance coefficient (LCCC) and reduced major axis regression (RMAR) were used to select the best model from validation results, and applied to holdout test data. Robustness of the best-performing model was also tested on imaging data from centre 4. RESULTS: The best model (SE-ResNet18-LSTM) produced accuracy of 96.1%, MAE 22.9 ms and LCCC 0.47 compared to ground truth on the holdout test set and accuracy of 97.3%, MAE 15.2 ms and LCCC 0.64 when tested on unseen external (centre 4) data. Differences in vendor performance were observed, with greatest accuracy for the most commonly represented vendor in the training data. CONCLUSION: A deep learning model was developed that can identify optimal inversion time from TI scout images on multi-vendor data with high accuracy, including on previously unseen external data. We make this model available to the scientific community for further assessment or development. CLINICAL RELEVANCE STATEMENT: A robust automated inversion time selection tool for late gadolinium-enhanced imaging allows for reproducible and efficient cross-vendor inversion time selection. KEY POINTS: • A model comprising convolutional and recurrent neural networks was developed to extract optimal TI from TI scout images. • Model accuracy within 50 ms of ground truth on multi-vendor holdout and external data of 96.1% and 97.3% respectively was achieved. • This model could improve workflow efficiency and standardise optimal TI selection for consistent LGE imaging.

High Prevalence of New Clinically Significant Findings in Patients With Embolic Stroke of Unknown Source Evaluated by Cardiac Magnetic Resonance Imaging.

BACKGROUND: Embolic stroke of unknown source (ESUS) accounts for 1 in 6 ischemic strokes. Current guidelines do not recommend routine cardiac magnetic resonance (CMR) imaging in ESUS, and beyond the identification of cardioembolic sources, there are no data assessing new clinical findings from CMR in ESUS. This study aimed to assess the prevalence of new cardiac and noncardiac findings and to determine their impact on clinical care in patients with ESUS. METHODS AND RESULTS: In this prospective, multicenter, observational study, CMR imaging was performed within 3 months of ESUS. All scans were reported according to standard clinical practice. A new clinical finding was defined as one not previously identified through prior clinical evaluation. A clinically significant finding was defined as one resulting in further investigation, follow-up, or treatment. A change in patient care was defined as initiation of medical, interventional, surgical, or palliative care. From 102 patients recruited, 96 underwent CMR imaging. One or more new clinical findings were observed in 59 patients (61%). New findings were clinically significant in 48 (81%) of these patients. Of 40 patients with a new clinically significant cardiac finding, 21 (53%) experienced a change in care (medical therapy, n=15; interventional/surgical procedure, n=6). In 12 patients with a new clinically significant extracardiac finding, 6 (50%) experienced a change in care (medical therapy, n=4; palliative care, n=2). CONCLUSIONS: CMR imaging identifies new clinically significant cardiac and noncardiac findings in half of patients with recent ESUS. Advanced cardiovascular screening should be considered in patients with ESUS. REGISTRATION: URL: https://www.clinicaltrials.gov; Unique identifier: NCT04555538.

High-resolution quantification of stress perfusion defects by cardiac magnetic resonance.

Aims: Quantitative stress perfusion cardiac magnetic resonance (CMR) is becoming more widely available, but it is still unclear how to integrate this information into clinical decision-making. Typically, pixel-wise perfusion maps are generated, but diagnostic and prognostic studies have summarized perfusion as just one value per patient or in 16 myocardial segments. In this study, the reporting of quantitative perfusion maps is extended from the standard 16 segments to a high-resolution bullseye. Cut-off thresholds are established for the high-resolution bullseye, and the identified perfusion defects are compared with visual assessment. Methods and results: Thirty-four patients with known or suspected coronary artery disease were retrospectively analysed. Visual perfusion defects were contoured on the CMR images and pixel-wise quantitative perfusion maps were generated. Cut-off values were established on the high-resolution bullseye consisting of 1800 points and compared with the per-segment, per-coronary, and per-patient resolution thresholds. Quantitative stress perfusion was significantly lower in visually abnormal pixels, 1.11 (0.75-1.57) vs. 2.35 (1.82-2.9) mL/min/g (Mann-Whitney U test P < 0.001), with an optimal cut-off of 1.72 mL/min/g. This was lower than the segment-wise optimal threshold of 1.92 mL/min/g. The Bland-Altman analysis showed that visual assessment underestimated large perfusion defects compared with the quantification with good agreement for smaller defect burdens. A Dice overlap of 0.68 (0.57-0.78) was found. Conclusion: This study introduces a high-resolution bullseye consisting of 1800 points, rather than 16, per patient for reporting quantitative stress perfusion, which may improve sensitivity. Using this representation, the threshold required to identify areas of reduced perfusion is lower than for segmental analysis.

Follow-up of CRT-D patients downgraded to CRT-P at the time of generator exchange.

Background: Some patients with cardiac resynchronisation therapy (CRT) experience super-response (LVEF improvements to ≥50%). At generator exchange (GE), downgrading (DG) from CRT-defibrillator (CRT-D) to CRT-pacemaker (CRT-P) could be an option for these patients on primary prevention ICD indication and no required ICD therapies. Long-term data on arrhythmic events in super-responders is scarce. Methods: CRT-D patients with LVEF improvement to ≥50% at GE were identified in four large centres for retrospective analysis. Mortality, significant ventricular tachyarrhythmia and appropriate ICD-therapy were determined, and patient analysis was split into two groups (downgraded to CRT-P or not). Results: Sixty-six patients (53% male, 26% coronary artery disease) on primary prevention were followed for a median of 129 months [IQR: 101-155] after implantation. 27 (41%) patients were downgraded to CRT-P at GE after a median of 68 [IQR: 58-98] months (LVEF 54% ± 4%). The other 39 (59%) continued with CRT-D therapy (LVEF 52% ± 6%). No cardiac death or significant arrhythmia occurred in the CRT-P group (median follow-up (FU) 38 months [IQR: 29-53]). Three appropriate ICD-therapies occurred in the CRT-D group [median FU 70 months (IQR: 39-97)]. Annualized event-rates after DG/GE were 1.5%/year and 1.0%/year in the CRT-D group and the whole cohort, respectively. Conclusions: No significant tachyarrhythmia were detected in the patients downgraded to CRT-P during follow-up. However, three events were observed in the CRT-D group. Whilst downgrading CRT-D patients is an option, a small residual risk for arrhythmic events remains and decisions regarding downgrade should be made on a case-by-case basis.

Hemodynamic force assessment by cardiovascular magnetic resonance in HFpEF: A case-control substudy from the HFpEF stress trial.

BACKGROUND: The diagnosis of heart failure with preserved ejection fraction (HFpEF) remains challenging. Exercise-stress testing is recommended in case of uncertainty; however, this approach is time-consuming and costly. Since preserved EF does not represent normal systolic function, we hypothesized comprehensive cardiovascular magnetic resonance (CMR) assessment of cardiac hemodynamic forces (HDF) may identify functional abnormalities in HFpEF. METHODS: The HFpEF Stress Trial (DZHK-17; Clinicaltrials.gov: NCT03260621) prospectively recruited 75 patients with exertional dyspnea, preserved EF (≥50%) and signs of diastolic dysfunction (E/e' ≥8) on echocardiography. Patients underwent rest and exercise-stress right heart catheterisation, echocardiography and CMR. The final study cohort consisted of 68 patients (HFpEF n = 34 and non-cardiac dyspnea n = 34 according to pulmonary capillary wedge pressure (PCWP)). HDF assessment included left ventricular (LV) longitudinal, systolic peak and impulse, systolic/diastolic transition, E-wave deceleration as well as A-wave acceleration forces. Follow-up after 24 months evaluated cardiovascular mortality and hospitalisation (CVH) - only two patients were lost to follow-up. FINDINGS: HDF assessment revealed impairment of LV longitudinal function in patients with HFpEF compared to non-cardiac dyspnoea (15.8% vs. 18.3%, p = 0.035), attributable to impairment of systolic peak (38.6% vs 51.6%, p = 0.003) and impulse (20.8% vs. 24.5%, p = 0.009) forces as well as late diastolic filling (-3.8% vs -5.4%, p = 0.029). Early diastolic filling was impaired in HFpEF patients identified at rest compared with patients identified during stress only (7.7% vs. 9.9%, p = 0.004). Impaired systolic peak was associated with CVH (HR 0.95, p = 0.016), and was superior to LV global longitudinal strain assessment in prediction of CVH (AUC 0.76 vs. 0.61, p = 0.048). INTERPRETATION: Assessment of HDF indicates impairment of LV systolic ejection force in HFpEF which is associated with cardiovascular events. FUNDING: German Centre for Cardiovascular Research (DZHK).

High-Resolution Free-Breathing Quantitative First-Pass Perfusion Cardiac MR Using Dual-Echo Dixon With Spatio-Temporal Acceleration.

Introduction: To develop and test the feasibility of free-breathing (FB), high-resolution quantitative first-pass perfusion cardiac MR (FPP-CMR) using dual-echo Dixon (FOSTERS; Fat-water separation for mOtion-corrected Spatio-TEmporally accelerated myocardial peRfuSion). Materials and Methods: FOSTERS was performed in FB using a dual-saturation single-bolus acquisition with dual-echo Dixon and a dynamically variable Cartesian k-t undersampling (8-fold) approach, with low-rank and sparsity constrained reconstruction, to achieve high-resolution FPP-CMR images. FOSTERS also included automatic in-plane motion estimation and T 2 * correction to obtain quantitative myocardial blood flow (MBF) maps. High-resolution (1.6 x 1.6 mm2) FB FOSTERS was evaluated in eleven patients, during rest, against standard-resolution (2.6 x 2.6 mm2) 2-fold SENSE-accelerated breath-hold (BH) FPP-CMR. In addition, MBF was computed for FOSTERS and spatial wavelet-based compressed sensing (CS) reconstruction. Two cardiologists scored the image quality (IQ) of FOSTERS, CS, and standard BH FPP-CMR images using a 4-point scale (1-4, non-diagnostic - fully diagnostic). Results: FOSTERS produced high-quality images without dark-rim and with reduced motion-related artifacts, using an 8x accelerated FB acquisition. FOSTERS and standard BH FPP-CMR exhibited excellent IQ with an average score of 3.5 ± 0.6 and 3.4 ± 0.6 (no statistical difference, p > 0.05), respectively. CS images exhibited severe artifacts and high levels of noise, resulting in an average IQ score of 2.9 ± 0.5. MBF values obtained with FOSTERS presented a lower variance than those obtained with CS. Discussion: FOSTERS enabled high-resolution FB FPP-CMR with MBF quantification. Combining motion correction with a low-rank and sparsity-constrained reconstruction results in excellent image quality.