Late Gadolinium Enhancement Cardiac MRI Research Articles

Quality assurance of late gadolinium enhancement cardiac MRI images: a deep learning classifier for confidence in the presence or absence of abnormality with potential to prompt real-time image optimisation

BackgroundLate gadolinium enhancement (LGE) of the myocardium has significant diagnostic and prognostic implications, with even small areas of enhancement being important. Distinguishing between definitely normal and definitely abnormal LGE images is usually straightforward; but diagnostic uncertainty arises when reporters are not sure whether the observed LGE is genuine or not. This uncertainty might be resolved by repetition (to remove artefact) or further acquisition of intersecting images, but this must take place before the scan finishes. Real-time quality assurance by humans is a complex task requiring training and experience, so being able to identify which images have an intermediate likelihood of LGE while the scan is ongoing, without the presence of an expert is of high value. This decision-support could prompt immediate image optimisation or acquisition of supplementary images to confirm or refute the presence of genuine LGE. This could reduce ambiguity in reports. MethodsShort-axis, phase sensitive inversion recovery (PSIR) late gadolinium images were extracted from our clinical CMR database and shuffled. Two, independent, blinded experts scored each individual slice for ‘LGE likelihood’ on a visual analogue scale, from 0 (absolute certainty of no LGE) to 100 (absolute certainty of LGE), with 50 representing clinical equipoise. The scored images were split into 2 classes – either “high certainty” of whether LGE was present or not, or “low certainty”. The dataset was split into training, validation and test sets (70:15:15). A deep learning binary classifier based on the EfficientNetV2 convolutional neural network architecture was trained to distinguish between these categories. Classifier performance on the test set was evaluated by calculating the accuracy, precision, recall, F1-score, and area under the receiver operating characteristics curve (ROC AUC). Performance was also evaluated on an external test set of images from a different centre. Results1645 images (from 272 patients) were labelled and split at the patient level into training (1151 images), validation (247 images) and test (247 images) sets for the deep learning binary classifier. Of these, 1208 images were ‘high certainty’ (255 for LGE, 953 for no LGE), and 437 were ‘low certainty’). An external test comprising 247 images from 41 patients from another centre was also employed. After 100 epochs the performance on the internal test set was: accuracy = 94%, recall = 0.80, precision = 0.97, F1-score = 0.87 and ROC AUC = 0.94. The classifier also performed robustly on the external test set (accuracy = 91%, recall = 0.73, precision = 0.93, F1-score = 0.82 and ROC AUC = 0.91). These results were benchmarked against a reference inter-expert accuracy of 86%. ConclusionsDeep learning shows potential to automate quality control of late gadolinium imaging in CMR. The ability to identify short-axis images with intermediate LGE likelihood in real-time may serve as a useful decision support tool. This approach has the potential to guide immediate further imaging while the patient is still in the scanner, thereby reducing the frequency of recalls and inconclusive reports due to diagnostic indecision.

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

Abstract 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.

Myocardial scar characteristics by 3D-LGE cannot fully explain different arrhythmic event rates in primary and secondary prevention of sudden cardiac death

Abstract Background There is a noticeable difference in the incidence of ventricular arrhythmias among patients receiving defibrillator devices (ICD or CRT-D) for primary vs. secondary prevention of sudden cardiac death (SCD). The underlying reasons for this difference remain to be fully explained. Purpose To assess for differences in myocardial scar characteristics between patients who had defibrillator devices implanted in primary vs. secondary prevention scenarios, and their correlation with arrhythmic events. Methods In this single center retrospective study, patients who underwent late gadolinium enhancement (LGE) cardiac MRI for clinical purposes before the implantation of an ICD or CRT-D were included. Patients with channelopathies (n=2) or inappropriate imaging quality (n=7) were excluded. We used ADAS software to perform myocardial scar characterization in 3D-LGE datasets in all but 16 patients, in which 2D datasets were used. The primary endpoint was a composite of appropriate ICD therapy (appropriate shock or ATP), sustained ventricular tachycardia or SCD. Results A total of 116 patients (mean age 66±14 years, 81% male) were included, 40 (35%) with devices implanted in secondary prevention. During a median follow-up of 28 months (IQR 16-24), 23 events were identified (18 appropriate ICD therapy, 9 shocks and 9 ATP; 2 SCD; 3 sustained VT), 7 of which (30.4%) in the primary prevention group, and 16 (69.6%) in the secondary prevention group. The event rate was significantly higher in the secondary prevention setting (15.0 events per 100 persons-year [95% CI 7.7 – 22.4] vs. 4.1 events per 100 persons-year [95% CI 1.1 – 7.1]; p < 0.001). Despite a higher LVEF in the secondary prevention group (41 ± 14% vs. 30 ± 13%; p-value < 0.001), no statistically significant differences were found regarding scar tissue characteristics, namely scar and borderzone (BZ) mass, total channel mass, largest channel mass, number of channels and scar heterogeneity (BZ mass / scar mass ratio) – Table. Conclusion Despite the higher event rate in patients receiving defibrillator devices in secondary vs. prevention, no differences in myocardial scar characteristics were found between both groups. These findings suggest that arrhythmic risk is unlikely to be explained solely by the anatomical substrate, and support a greater role for the interplay between substrate and transient triggers.

Segmentation of cardiac infarction in delayed-enhancement MRI using probability map and transformers-based neural networks

Background and Objective: Automatic segmentation of myocardial infarction is of great clinical interest for the quantitative evaluation of myocardial infarction (MI). Late Gadolinium Enhancement cardiac MRI (LGE-MRI) is commonly used in clinical practice to quantify MI, which is crucial for clinical diagnosis and treatment of cardiac diseases. However, the segmentation of infarcted tissue in LGE-MRI is highly challenging due to its high anisotropy and inhomogeneities.Methods: The innovative aspect of our work lies in the utilization of a probability map of the healthy myocardium to guide the localization of infarction, as well as the combination of 2D U-Net and U-Net transformers to achieve the final segmentation. Instead of employing a binary segmentation map, we propose using a probability map of the normal myocardium, obtained through a dedicated 2D U-Net. To leverage spatial information, we employ a U-Net transformers network where we incorporate the probability map into the original image as an additional input. Then, To address the limitations of U-Net in segmenting accurately the contours, we introduce an adapted loss function.Results: Our method has been evaluated on the 2020 MICCAI EMIDEC challenge dataset, yielding competitive results. Specifically, we achieved a Dice score of 92.94% for the myocardium and 92.36% for the infarction. These outcomes highlight the competitiveness of our approach.Conclusion: In the case of the infarction class, our proposed method outperforms state-of-the-art techniques across all metrics evaluated in the challenge, establishing its superior performance in infarction segmentation. This study further reinforces the importance of integrating a contour loss into the segmentation process.

Detection of Early Diffuse Myocardial Fibrosis and Inflammation in Chagas Cardiomyopathy with T1 Mapping and Extracellular Volume

Purpose To evaluate myocardial T1 mapping and extracellular volume (ECV) parameters in different stages of Chagas cardiomyopathy and determine whether they are predictive of disease severity and prognosis. Materials and Methods Prospectively enrolled participants (July 2013 to September 2016) underwent cine and late gadolinium enhancement (LGE) cardiac MRI and T1 mapping with a precontrast (native) or postcontrast modified Look-Locker sequence. The native T1 and ECV values were measured among subgroups that were based on disease severity (indeterminate, Chagas cardiomyopathy with preserved ejection fraction [CCpEF], Chagas cardiomyopathy with midrange ejection fraction [CCmrEF], and Chagas cardiomyopathy with reduced ejection fraction [CCrEF]). Cox proportional hazards regression and the Akaike information criterion were used to determine predictors of major cardiovascular events (cardioverter defibrillator implant, heart transplant, or death). Results In 107 participants (90 participants with Chagas disease [mean age ± SD, 55 years ± 11; 49 men] and 17 age- and sex-matched control participants), the left ventricular (LV) ejection fraction and the extent of focal and diffuse or interstitial fibrosis were correlated with disease severity. Participants with CCmrEF and participants with CCrEF showed significantly higher global native T1 and ECV values than participants in the indeterminate, CCpEF, and control groups (T1: 1072 msec ± 34 and 1073 msec ± 63 vs 1010 msec ± 41, 1005 msec ± 69, and 999 msec ± 46; ECV: 35.5% ± 3.6 and 35.0% ± 5.4 vs 25.3% ± 3.5, 28.2% ± 4.9, and 25.2% ± 2.2; both P < .001). Remote (LGE-negative areas) native T1 and ECV values were also higher (T1: 1056 msec ± 32 and 1071 msec ± 55 vs 1008 msec ± 41, 989 msec ± 96, and 999 msec ± 46; ECV: 30.2% ± 4.7 and 30.8% ± 7.4 vs 25.1% ± 3.5, 25.1% ± 3.7, and 25.0% ± 2.2; both P < .001). Abnormal remote ECV values (>30%) occurred in 12% of participants in the indeterminate group, which increased with disease severity. Nineteen combined outcomes were observed (median follow-up time: 43 months), and a remote native T1 value greater than 1100 msec was independently predictive of combined outcomes (hazard ratio, 12 [95% CI: 4.1, 34.2]; P < .001). Conclusion Myocardial native T1 and ECV values were correlated with Chagas disease severity and may serve as markers of myocardial involvement in Chagas cardiomyopathy that precede LGE and LV dysfunction. Keywords: MRI, Cardiac, Heart, Imaging Sequences, Chagas Cardiomyopathy Supplemental material is available for this article. © RSNA, 2023

PO-03-107 MYOCARDIAL SCAR CHARACTERISTICS BY 3D-LGE CANNOT FULLY EXPLAIN DIFFERENT ARRHYTHMIC EVENT RATES IN PRIMARY AND SECONDARY PREVENTION OF SUDDEN CARDIAC DEATH

There is a noticeable difference in the incidence of ventricular arrythmias among patients receiving defibrillator devices (ICD or CRT-D) for primary vs. secondary prevention of sudden cardiac death (SCD). The underlying reasons for this difference remain to be fully explained. To assess for differences in myocardial scar characteristics between patients who had defibrillator devices implanted in primary vs. secondary prevention scenarios, and their correlation with arrhythmic events. In this single center retrospective study, patients who underwent late gadolinium enhancement (LGE) cardiac MRI for clinical purposes before the implantation of an ICD or CRT-D were included. Patients with channelopathies (n=2) or inappropriate imaging quality (n=7) were excluded. We used ADAS software to perform myocardial scar characterization in 3D-LGE datasets in all but 16 patients, in which 2D datasets were used. The primary endpoint was a composite of appropriate ICD therapy (appropriate shock or ATP), sustained ventricular tachycardia or SCD. A total of 116 patients (mean age 66±14 years, 81% male) were included, 40 (35%) with devices implanted in secondary prevention. During a median follow-up of 28 months (IQR 16-24), 23 events were identified (18 appropriate ICD therapy, 9 shocks and 9 ATP; 2 SCD; 3 sustained VT), 7 of which (30.4%) in the primary prevention group, and 16 (69.6%) in the secondary prevention group. The event rate was significantly higher in the secondary prevention setting (15.0 events per 100 persons-year [95% CI 7.7 – 22.4] vs. 4.1 events per 100 persons-year [95% CI 1.1 – 7.1]; p < 0.001). Despite a higher LVEF in the secondary prevention group (41 ± 14% vs. 30 ± 13%; p-value < 0.001), no statistically significant differences were found regarding scar tissue characteristics, namely scar and borderzone (BZ) mass, total channel mass, largest channel mass, number of channels and scar heterogeneity (BZ mass / scar mass ratio) – Table. Despite the higher event rate in patients receiving defibrillator devices in secondary vs. prevention, no differences in myocardial scar characteristics were found between both groups. These findings suggest that arrhythmic risk is unlikely to be explained solely by the anatomical substrate, and support a greater role for the interplay between substrate and transient triggers.Tabled 1Primary PreventionSecondary Preventionp-valueScar + BZ mass (g)26.2 (IQR 13.3 - 35.0)23.7 (IQR 11.4 - 33.4)p = 0.322Scar + BZ percentage (%)18.0 (IQR 11.0 – 27.0)17.0 (IQR 9.0 - 26.0)p = 0.545BZ mass (g)17.7 (IQR 11.1 - 24.2)14.8 (IQR 8.8 - 22.0)p = 0.350BZ percentage (%)13.0 (IQR 8.0 - 18.0)11.0 (IQR 6.0 - 17.0)p = 0.677Scar mass (g)6.6 (IQR 3.4 - 11.0)6.9 (IQR 2.0 - 10.3)p = 0.693Scar percentage (%)4.0 (IQR 2.0 - 9.0)5.0 (IQR 1.0 - 8.0)p = 0.695Scar heterogeneity2.4 (IQR 1.7 - 3.9)2.7 (IQR 1.5 - 4.4)p = 0.797Total channels mass (g)3.0 (IQR 0.7 - 4.9)1.8 (IQR 0.0 - 5.0)p = 0.215Biggest channel mass (g)1.4 (IQR 0.5 - 2.5)0.8 (IQR 0.0 - 3.0)p = 0.163Number of channels3.0 (IQR 1.0 - 5.0)3.0 (IQR 0.0 - 5.0)p = 0.520BZ, Borderzone Open table in a new tab

Image Quality of Submillimeter High-Spatial-Resolution 2D Late Gadolinium-enhanced Images in Cardiac MRI: A Feasibility Study.

To evaluate the image quality of high-spatial-resolution two-dimensional (2D) late gadolinium enhancement (LGE) cardiac MRI compared with conventional normal-resolution LGE MRI. This prospective study included participants suspected of having cardiomyopathy who underwent cardiac MRI between March 2021 and December 2021. Normal-resolution and high-resolution 2D LGE sequences (inversion recovery [IR] and phase-sensitive inversion recovery [PSIR]) were performed at 3 T. Resolution was compared between normal-resolution and high-resolution images obtained in a quality assurance phantom. In vivo image quality and resolution were evaluated qualitatively using a five-point scoring system. Receiver operating characteristic curve analysis was used for LGE detection performance. Border sharpness was assessed with profile curve measurement. The contrast-to-noise ratio (CNR) between hyperenhancement and remote myocardium and LGE detection performance were calculated using normal-resolution IR images as the reference. In total, 120 participants were evaluated (mean age, 56 years ± 17 [SD]; 72 men). Features smaller than 1 mm were detectable only on high-resolution images of the phantom. In vivo, the image resolution score with high-resolution LGE was 4.14-4.24, which was higher than the normal-resolution LGE reference score of 2.99 (P < .05). Border sharpness was higher in high-resolution images (P < .001). Receiver operating characteristic curve analysis revealed no evidence of a difference in LGE detection between normal-resolution and high-resolution images. There was also no evidence of a change in CNR of LGE in IR and PSIR magnitude compared with reference images. Comparison of image quality in 2D high-resolution and normal-resolution LGE cardiac MRI demonstrated the highest resolution for high-resolution IR and high-resolution PSIR magnitude sequences.Keywords: Cartilage Imaging, MRI, Cardiac, Heart, Imaging Sequences, Comparative Studies Supplemental material is available for this article. © RSNA, 2022.

Multidimensional immunoimaging to characterize the inflammatory reaction after permanent and temporary coronary artery occlusion in mice

Abstract Background The immune response following acute myocardial infarction encompasses a delicate balance between inflammatory and reparative programs. Our knowledge of these complex mechanisms is mainly derived from studies using a murine model of permanent coronary artery occlusion. In this study we developed, validated and implemented multiparametric imaging methods to investigate cardiac function and the systemic immune response in transient or permanent coronary artery occlusion mouse models. Methods The myocardial infarction models encompassed either transient (40 min) or permanent LAD occlusion in C57BL/6 mice, and non-infarcted mice were used as controls. Two or seven days later, the animals subjected to systemic immunoimaging of the bone marrow, spleen and myocardium with late gadolinium enhancement cardiac MRI (LGE cMRI), 18F-fluorodeoxyglucose (18F-FDG) PET, 18F-Fluorothymidine (18F-FLT) PET, 64Cu-CCR2 PET targeting inflammatory monocytes, 89Zr-CD11b nanobody PET and 19F-HDL-PERFECTA MRI, both targeting myeloid cells. In addition, the same myocardial infarction models were applied to atherosclerosis-prone Apoe−/− and systemic inflammation and plaque progression were assessed by flow cytometry and immunohistochemistry four weeks after infarction. Results Through LGE cMRI and 18F-FDG PET, we observed that temporary coronary occlusion resulted in a smaller infarct size, better cardiac function and viability compared to permanent occlusion. Multiparametric immunoimaging targeting CD11b+ cells by 89Zr-CD11b nanobody PET and Ly6Chi inflammatory monocytes by 64Cu-CCR2 PET demonstrated that mice subjected to transient coronary occlusion had less immune cell influx to the ischemic myocardium. This finding was confirmed by flow cytometry analysis of the infarct zone. In contrast, both myocardial infarction models cause a similar systemic immune response in the bone marrow and spleen as observed with multimodal imaging with subsequent similar numbers of CD11b+ cells in the blood. Both permanent and temporary coronary artery occlusion aggravate atherosclerosis in Apoe−/− mice with higher macrophage and Ly6Chi monocyte numbers in aortas and larger plaque size compared to Apoe−/− mice without myocardial infarction. Conclusions We developed and employed multimodal, multiparametric imaging protocols to characterize the immune response in the heart, bone marrow and spleen in two models of myocardial infarction. While cardiac function was superior in the ischemia reperfusion model, both types of myocardial infarction accelerated atherosclerosis. Funding Acknowledgement Type of funding sources: Other. Main funding source(s): NIH and DFG

Correlation between Cardiac MRI and Voltage Mapping in Evaluating Atrial Fibrosis: A Systematic Review.

To provide an overview of existing literature on the association between late gadolinium enhancement (LGE) cardiac MRI and low voltage areas (LVA) obtained with electroanatomic mapping (EAM) or histopathology when assessing atrial fibrosis. A systematic literature search was conducted in the PubMed, Embase, and Cochrane Library databases to identify all studies published until June 7, 2022, comparing LGE cardiac MRI to LVA EAM and/or histopathology for evaluation of atrial fibrosis. The study protocol was registered at PROSPERO (registration no. CRD42022338243). Two reviewers independently evaluated the studies for inclusion. Risk of bias and applicability for each included study were assessed using Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) criteria. Data regarding demographics, electrophysiology, LGE cardiac MRI, and study outcomes were extracted. The search yielded 1048 total results, of which 22 studies were included. Nineteen of the 22 included studies reported a significant correlation between high signal intensity at LGE cardiac MRI and LVA EAM or histopathology. However, there was great heterogeneity between included studies regarding study design, patient samples, cardiac MRI performance and postprocessing, and EAM performance. Current literature suggests a correlation between LGE cardiac MRI and LVA EAM or histopathology when evaluating atrial fibrosis but high heterogeneity between studies, demonstrating the need for uniform choices regarding cardiac MRI and EAM acquisition in future studies.Keywords: Cardiac, MR Imaging, Left Atrium Supplemental material is available for this article. © RSNA, 2022.

Myocardial Fibrosis at Cardiac MRI Helps Predict Adverse Clinical Outcome in Patients with Mitral Valve Prolapse.

Background Patients with mitral valve prolapse (MVP) may develop adverse outcomes even in the absence of mitral regurgitation or left ventricular (LV) dysfunction. Purpose To investigate the prognostic value of mitral annulus disjunction (MAD) and myocardial fibrosis at late gadolinium enhancement (LGE) cardiac MRI in patients with MVP without moderate-to-severe mitral regurgitation or LV dysfunction. Materials and Methods In this longitudinal retrospective study, 118 144 cardiac MRI studies were evaluated between October 2007 and June 2020 at 15 European tertiary medical centers. Follow-up was from the date of cardiac MRI examination to June 2020; the minimum and maximum follow-up intervals were 6 months and 156 months, respectively. Patients were excluded if at least one of the following conditions was present: cardiomyopathy, LV ejection fraction less than 40%, ischemic heart disease, congenital heart disease, inflammatory heart disease, moderate or worse mitral regurgitation, participation in competitive sport, or electrocardiogram suggestive of channelopathies. In the remainder, cardiac MRI studies were reanalyzed, and patients were included if they were aged 18 years or older, MVP was diagnosed at cardiac MRI, and clinical information and electrocardiogram monitoring were available within 3 months from cardiac MRI examination. The end point was a composite of adverse outcomes: sustained ventricular tachycardia (VT), sudden cardiac death (SCD), or unexplained syncope. Multivariable Cox regression analysis was performed. Results A total of 474 patients (mean age, 47 years ± 16 [SD]; 244 women) were included. Over a median follow-up of 3.3 years, 18 patients (4%) reached the study end point. LGE presence (hazard ratio, 4.2 [95% CI: 1.5, 11.9]; P = .006) and extent (hazard ratio, 1.2 per 1% increase [95% CI: 1.1, 1.4]; P = .006), but not MAD presence (P = .89), were associated with clinical outcome. LGE presence had incremental prognostic value over MVP severity and sustained VT and aborted SCD at baseline (area under the receiver operating characteristic curve, 0.70 vs 0.62; P = .03). Conclusion In contrast to mitral annulus disjunction, myocardial fibrosis determined according to late gadolinium enhancement at cardiac MRI was associated with adverse outcome in patients with mitral valve prolapse without moderate-to-severe mitral regurgitation or left ventricular dysfunction. © RSNA, 2022 Online supplemental material is available for this article. See also the editorial by Gerber in this issue.

Optimized automated cardiac MR scar quantification with GAN‐based data augmentation

BackgroundThe clinical utility of late gadolinium enhancement (LGE) cardiac MRI is limited by the lack of standardization, and time-consuming postprocessing. In this work, we tested the hypothesis that a cascaded deep learning pipeline trained with augmentation by synthetically generated data would improve model accuracy and robustness for automated scar quantification. MethodsA cascaded pipeline consisting of three consecutive neural networks is proposed, starting with a bounding box regression network to identify a region of interest around the left ventricular (LV) myocardium. Two further nnU-Net models are then used to segment the myocardium and, if present, scar. The models were trained on the data from the EMIDEC challenge, supplemented with an extensive synthetic dataset generated with a conditional GAN. ResultsThe cascaded pipeline significantly outperformed a single nnU-Net directly segmenting both the myocardium (mean Dice similarity coefficient (DSC) (standard deviation (SD)): 0.84 (0.09) vs 0.63 (0.20), p < 0.01) and scar (DSC: 0.72 (0.34) vs 0.46 (0.39), p < 0.01) on a per-slice level. The inclusion of the synthetic data as data augmentation during training improved the scar segmentation DSC by 0.06 (p < 0.01). The mean DSC per-subject on the challenge test set, for the cascaded pipeline augmented by synthetic generated data, was 0.86 (0.03) and 0.67 (0.29) for myocardium and scar, respectively. ConclusionA cascaded deep learning-based pipeline trained with augmentation by synthetically generated data leads to myocardium and scar segmentations that are similar to the manual operator, and outperforms direct segmentation without the synthetic images.

Extent of Left Atrial Fibrosis Correlates with Descending Aorta Proximity at 3D Late Gadolinium Enhancement Cardiac MRI in Patients with Atrial Fibrillation.

To determine whether the distance between the descending aorta and left atrial (LA) wall correlates with the amount of fibrosis quantified in the posterior left inferior pulmonary vein (LIPV) area of the LA in patients with atrial fibrillation (AF). In this retrospective study, patients with AF underwent cardiac MRI in sinus rhythm prior to a pulmonary vein isolation procedure (July 2018 to February 2020). The mean distance (distancemean) and shortest distance (distanceshort) between the descending aorta and the LA wall were measured on three-dimensional (3D) contrast-enhanced MR angiograms; distancemean was defined as the average of five measurements at different levels between the descending aorta and the LA wall. The extent of LA fibrosis, both global fibrosis and regional fibrosis within the LIPV area, was derived from postprocessed, 3D, late gadolinium-enhanced images. Associations between the extent of fibrosis and the proximity of the descending aorta were analyzed by using correlative and multivariable analyses. A total of 47 (mean age, 60 years ± 8 [standard deviation]; 31 men) patients were included for analysis. The extent of fibrosis in the posterior LIPV area was correlated with the distancemean (r s = -0.48; P < .01) and distanceshort (r s = -0.49; P < .01). Patients with a short distance between the descending aorta and LA wall (defined as a distanceshort < 2 mm) had a higher percentage of fibrosis in the posterior LIPV area than patients with a distanceshort greater than 2 mm (38.7% ± 22.7 vs 21.2% ± 17.8; P < .01). The distance between the descending aorta and LA was correlated with the extent of quantified fibrosis within the posterior LIPV area.Keywords: MRI, Cardiac, Left Atrium Supplemental material is available for this article. © RSNA, 2022.

Cardiac MRI to Visualize Myocardial Damage after ST-Segment Elevation Myocardial Infarction: A Review of Its Histologic Validation.

Cardiac MRI is a noninvasive diagnostic tool using nonionizing radiation that is widely used in patients with ST-segment elevation myocardial infarction (STEMI). Cardiac MRI depicts different prognosticating components of myocardial damage such as edema, intramyocardial hemorrhage (IMH), microvascular obstruction (MVO), and fibrosis. But how do cardiac MRI findings correlate to histologic findings? Shortly after STEMI, T2-weighted imaging and T2* mapping cardiac MRI depict, respectively, edema and IMH. The acute infarct size can be determined with late gadolinium enhancement (LGE) cardiac MRI. T2-weighted MRI should not be used for area-at-risk delineation because T2 values change dynamically over the first few days after STEMI and the severity of T2 abnormalities can be modulated with treatment. Furthermore, LGE cardiac MRI is the most accurate method to visualize MVO, which is characterized by hemorrhage, microvascular injury, and necrosis in histologic samples. In the chronic setting post-STEMI, LGE cardiac MRI is best used to detect replacement fibrosis (ie, final infarct size after injury healing). Finally, native T1 mapping has recently emerged as a contrast material-free method to measure infarct size that, however, remains inferior to LGE cardiac MRI. Especially LGE cardiac MRI-defined infarct size and the presence and extent of MVO may be used to monitor the effect of new therapeutic interventions in the treatment of reperfusion injury and infarct size reduction. © RSNA, 2021 Online supplemental material is available for this article.

Cine and late gadolinium enhancement MRI registration and automated myocardial infarct heterogeneity quantification.

To develop an approach for automated quantification of myocardial infarct heterogeneity in late gadolinium enhancement (LGE) cardiac MRI. We acquired 2D short-axis cine and 3D LGE in 10 pigs with myocardial infarct. The 2D cine myocardium was segmented and registered to the LGE images. LGE image signal intensities within the warped cine myocardium masks were analyzed to determine the thresholds of infarct core (IC) and gray zone (GZ) for the standard-deviation (SD) and full-width-at-halfmaximum (FWHM) methods. The initial IC, GZ, and IC + GZ segmentations were postprocessed using a normalized cut approach. Cine segmentation and cine-LGE registration accuracies were evaluated using dice similarity coefficient and average symmetric surface distance. Automated IC, GZ, and IC + GZ volumes were compared with manual results using Pearson correlation coefficient (r), Bland-Altman analyses, and intraclass correlation coefficient. For n = 87 slices containing scar, we achieved cine segmentation dice similarity coefficient = 0.87 ± 0.12, average symmetric surface distance = 0.94 ± 0.74 mm (epicardium), and 1.03 ± 0.82 mm (endocardium) in the scar region. For cine-LGE registration, dice similarity coefficient was 0.90 ± 0.06 and average symmetric surface distance was 0.72 ± 0.39 mm (epicardium) and 0.86 ± 0.53 mm (endocardium) in the scar region. For both SD and FWHM methods, automated IC, GZ, and IC + GZ volumes were strongly (r > 0.70) correlated with manual measurements, and the correlations were not significantly different from interobserver correlations (P > .05). The agreement between automated and manual scar volumes (intraclass correlation coefficient = 0.85-0.96) was similar to that between two observers (intraclass correlation coefficient = 0.81-0.99); automated scar segmentation errors were not significantly different from interobserver segmentation differences (P > .05). Our approach provides fully automated cine-LGE MRI registration and LGE myocardial infarct heterogeneity quantification in preclinical studies.

Prognostic utility of differential contrast-enhancement patterns on late gadolinium enhancement cardiac MRI (LGE-CMR) in patients with cancer-associated cardiac masses

Abstract Background LGE-CMR tissue characterization is widely used to identify cardiac masses (CMASS) in cancer patients – including neoplasm (NEO) and thrombus (THR). Prognostic utility of their differential LGE patterns is unknown. Purpose To determine incremental prognostic utility of LGE patterns in CMASS Methods The population comprised of cancer patients with CMASS on LGE-CMR, for which etiology was classified based on presence (NEO) or absence (THR) of enhancement, and controls matched for cancer type/stage. LGE-CMR tissue properties of NEO was classified based on extent of contrast enhancement – diffusely enhancing (DE), mixed (ME), and predominantly avascular (PA). Clinical follow up was performed for embolic events within 6 months of CMR and all-cause mortality. Results 330 cancer patients (55% M; 55±16yo) with an array of cancer diagnoses (19% sarcoma, 17% GI, 13% GU) were studied. Among CMASS+ pts (n=190), 66% had NEO and 34% had THR on LGE. All THR were non-enhancing. Among NEO, LGE pattern was variable (46% DE, 41% ME, 13% PA); ME lesions were larger than other groups (Fig. 1A). Quantitative tissue properties were consistent with qualitative groups, as evidenced by stepwise variation in signal intensity and CNR. Cumulative embolic events were 3-fold higher in CMASS+ than controls (All: 20% vs. 7%, p=0.001; PE: 13% vs. 5%, p=0.02; CVA/systemic embolism: 10% vs. 3%, p=0.01). Median time to event was 1.3 months [IQR 0.1–2.3] from CMR. Aggregate events were similar between NEO and THR, reflecting similar rates of PE and CVA (p=NS). Among CMASS pts with embolic events, 56% were on anticoagulation at time of event (59% NEO, 50% THR, p=0.61). Regarding CMASS morphology, emboli were 3-fold higher among intracavitary (IC) or highly mobile (HM) CMASS (IC: 25% vs 7%, p&amp;lt;0.001; HM: 38% vs 12%, p=0.001). Regarding location, right sided CMASS were associated with a 3–5 fold increase in PE (IC: 19% vs 6%; HM: 35% vs 7%, both p&amp;lt;0.001) and similar CVA events among left sided CMASS (IC: 17% vs. 6%, p=0.02; HM: 33% vs 6%, p=0.05). Embolic events were similar when partitioned based on quantitative LGE patterns between patients with and without embolic events. As for all-cause mortality, NEO on CMR conferred increased mortality than THR (HR 3.06 [CI=1.84–5.1], p&amp;lt;0.001) and matched controls (HR 2.08 [CI=1.42–3.04], p&amp;lt;0.001) during a median follow-up of 9.4 months [IQR 3.6–23.2]. Among NEO subgroups (Fig. 1B), survival was lower in patients with heterogeneous LGE patterns vs matched controls: the lowest survival in ME (p=0.002) suggests increased vascularity and tumor hypoxia/necrosis associated with aggressive tumors and hence larger lesions. Conclusions Among cancer patients, CMR-evidenced CMASS confers high short-term embolic risk, which are equivalently common between NEO and THR. Intra-cavitary location and increased mobility augment embolic risk irrespective of CMASS tissue properties whereas differential LGE patterns on CMR strongly impact prognosis. Funding Acknowledgement Type of funding source: None

P5657Fibrosis detected by late-gadolinium enhancement cardiac MRI is associated with atrial fibrillation and poorer ablation outcome: A meta-analysis

Abstract Background Fibrosis is a hallmark of atrial fibrillation (AF) substrate. Recent data suggests that fibrosis detected by late-gadolinium enhancement (LGE) cardiac MRI (CMR) can predict AF. However, this relationship is not well described. Objective To delineate the association of cardiac fibrosis detected by LGE CMR with AF prevalence, AF recurrence after catheter ablation. Methods PubMed, Embase, Web of Science and Ovid MEDLINE were searched through November 2018, using the keywords: LGE AND Fibrosis AND CMR AND AF. Inclusion criteria: 1. LGE CMR of left atrial (LA LGE), ventricular wall (LV LGE) or right ventricular wall (RV LGE); 2. Studies reporting AF or recurrent arrhythmia after ablation; 3. Patient ≥18 years; and 4. ≥50 participants. Included studies were pooled in a random effects meta-analysis and reported as: mean difference (MD); unadjusted risk ratios (RR); adjusted hazard ratios (HR); and 95% confidence intervals (95% CI). Results After exclusions, we identified 9 studies (2,307 patients [65.9% males, 34.1% females]) conducted between 2003 and 2015 for LGE and AF. Fibrosis was present in 666 (35.1%) and detected by LV LGE in 7 (78%) and RV LGE in 2 (22%). The presence of AF was higher in patients positive for ventricular LGE than those negative, trending towards significance (RR: 1.51, 95% CI: 0.94–2.45, p=0.09). Pooled LV fibrosis associated with AF progression (RR [NPAF vs. PAF]: 2.2, 95% CI: 1.22–3.94, p=0.009). We identified 8 studies (2,041 patients [65.8% males, 34.2% females]) conducted between 2006 and 2016 reporting LGE and AF recurrence after catheter ablation, with fibrosis detected in 644 (31.6%) by LA LGE in 8 (88.9%, biased towards one centre). After 17.8±14.2 follow-up years, atrial fibrosis was significantly greater in recurrent AF than controls (MD: 4.97%, 95% CI: 1.23–8.7, p&lt;0.01), and predicted 16% increased risk of AF recurrence (RR: 1.16, 95% CI: 1.07–1.26, p&lt;0.05). Conclusion Myocardial fibrosis detected by LGE associates with prevalence and progress of AF and is predictive of AF recurrence post ablation. This further supports the proarrhythmic role of fibrosis and selection of patients for ablation therapy based on LGE.

Metal Artifact Reduction in Cardiovascular MRI for Accurate Myocardial Scar Assessment in Patients With Cardiac Implantable Electronic Devices.

OBJECTIVE. An important application of late gadolinium enhancement (LGE) cardiac MRI is accurate assessment of myocardial scar before ablation. However, this is often limited in patients with cardiac implantable electronic devices (CIEDs) because of metal device-induced artifacts. The purpose of this study was to determine whether a modified wideband inversion recovery (IR) LGE MRI technique decreases artifact volume to allow the assessment of myocardial scar. SUBJECTS AND METHODS. Fifty patients (17 women and 33 men; mean age ± SD, 61 ± 12 years; mean ejection fraction ± SD, 35.9% ± 13.3%) with CIEDs underwent cardiac MRI using conventional and modified wideband IR LGE techniques before ablation. The volume of device-induced artifact was quantified and stratified by tertiles on mild, moderate, and severe. Ordinal logistic regression analysis assessed the association between artifact volume on conventional and wideband images adjusted for patients' demographics. RESULTS. Conventional LGE MRI resulted in device-induced hyperintense artifacts that obscured ventricular segments in 32 of 50 (64%) cases. Wideband LGE MRI significantly reduced severe artifact volume (p < 0.0001) and completely resolved all mild and most moderate artifacts. Overall, wideband techniques resulted in a 56% reduction in total artifact volume for the cohort (p < 0.0001). The wideband LGE MRI sequence minimized artifacts in the most commonly obscured segments on the conventional LGE MRI sequence, with persistent artifacts in seven, eight, and four of 32 cases at the basal anterior, midventricular anterior, and midventricular anteroseptal segments, respectively. CONCLUSION. The modified wideband IR technique completely resolves mild and moderate device-induced hyperintense artifacts and significantly reduces the volume of severe artifact to allow accurate identification of myocardial scar in patients with CIEDs before ablation.

Ventricular Tachycardia (VT) Substrate Characteristics: Insights from Multimodality Structural and Functional Imaging of the VT Substrate Using Cardiac MRI Scar, 123I-Metaiodobenzylguanidine SPECT Innervation, and Bipolar Voltage.

Postischemic adaptation results in characteristic myocardial structural and functional changes in the ventricular tachycardia (VT) substrate. The aim of this study was to compare myocardial structural and functional adaptations (late gadolinium enhancement/abnormal innervation) with detailed VT mapping data to identify regional heterogeneities in postischemic changes. Methods: Fifteen patients with ischemic cardiomyopathy and drug-refractory VT underwent late gadolinium enhancement cardiac MRI (CMR), 123I-metaiodobenzylguanidine SPECT, and high-resolution bipolar voltage mapping to assess fibrosis (>3 SDs), abnormal innervation (<50% tracer uptake), and low-voltage area (<1.5 mV), respectively. Three-dimensional reconstructed CMR/123I-metaiodobenzylguanidine models were coregistered for further comparison. Results: Postischemic structural and functional adaptations in all 3 categories were similar in size (reported as median [quartile 1-quartile 3]: CMR scar, 46.1 cm2 [33.1-86.9 cm2]; abnormal innervation, 47.8 cm2 [40.5-68.1 cm2]; and low-voltage area, 29.5 cm2 [24.5-102.6 cm2]; P > 0.05). However, any single modality underestimated the total VT substrate area defined as abnormal in at least 1 of the 3 modalities (76.0 cm2 [57.9-143.2 cm2]; P < 0.001). Within the total VT substrate area, regions abnormal in all 3 modalities were most common (25.2%). However, significant parts of the VT substrate had undergone heterogeneous adaptation (abnormal in <3 modalities); the most common categories were "abnormal innervation only" (18.2%), "CMR scar plus abnormal innervation only" (14.9%), and "CMR scar only" (14.6%). All 14 VT channel/exit sites (0.88 ± 0.74 mV) were localized to myocardium demonstrating CMR scar and abnormal innervation. This specific tissue category accounted for 68.3% of the CMR scar and 31.2% of the total abnormal postischemic VT substrate area. Conclusion: Structural and functional imaging demonstrated regional heterogeneities in the postischemic VT substrate not appreciated by any single modality alone. The coexistence of abnormal innervation and CMR scar may identify a particularly "proarrhythmic" adaptation and may represent a potential novel target for VT ablation.

High spatial resolution free-breathing 3D late gadolinium enhancement cardiac magnetic resonance imaging in ischaemic and non-ischaemic cardiomyopathy: quantitative assessment of scar mass and image quality

PurposeTo compare breath-hold (BH) with navigated free-breathing (FB) 3D late gadolinium enhancement cardiac MRI (LGE-CMR)Materials and methodsFifty-one patients were retrospectively included (34 ischaemic cardiomyopathy, 14 non-ischaemic cardiomyopathy, three discarded). BH and FB 3D phase sensitive inversion recovery sequences were performed at 3T. FB datasets were reformatted into normal resolution (FB-NR, 1.46x1.46x10mm) and high resolution (FB-HR, isotropic 0.91-mm voxels). Scar mass, scar edge sharpness (SES), SNR and CNR were compared using paired-samples t-test, Pearson correlation and Bland-Altman analysis.ResultsScar mass was similar in BH and FB-NR (mean ± SD: 15.5±18.0 g vs. 15.5±16.9 g, p=0.997), with good correlation (r=0.953), and no bias (mean difference ± SD: 0.00±5.47 g). FB-NR significantly overestimated scar mass compared with FB-HR (15.5±16.9 g vs 14.4±15.6 g; p=0.007). FB-NR and FB-HR correlated well (r=0.988), but Bland-Altman demonstrated systematic bias (1.15±2.84 g). SES was similar in BH and FB-NR (p=0.947), but significantly higher in FB-HR than FB-NR (p<0.01). SNR and CNR were lower in BH than FB-NR (p<0.01), and lower in FB-HR than FB-NR (p<0.01).ConclusionNavigated free-breathing 3D LGE-CMR allows reliable scar mass quantification comparable to breath-hold. During free-breathing, spatial resolution can be increased resulting in improved sharpness and reduced scar mass.Key Points• Navigated free-breathing 3D late gadolinium enhancement is reliable for myocardial scar quantification.• High-resolution 3D late gadolinium enhancement increases scar sharpness• Ischaemic and non-ischaemic cardiomyopathy patients can be imaged using free-breathing LGE CMR.

Abstract 121: Embolic Stroke of Unknown Source is Associated With Higher Left Atrial Fibrosis Than Other Stroke Subtypes

Introduction: Atrial cardiomyopathy without atrial fibrillation is suspected to lead to embolic stroke of unknown source (ESUS). Late gadolinium enhancement cardiac MRI (LGE-MRI) is a non-invasive technique used to assess atrial fibrosis and cardiomyopathy. Hypothesis: We hypothesized that patients with ESUS have a higher percent of atrial fibrosis, quantified using LGE-MRI, compared to other stroke subtypes. Methods: Cardiac LGE-MRIs were obtained in ischemic stroke patients treated at two comprehensive stroke centers. Acquisition of images took approximately 30 minutes and was well tolerated by all of the patients. Each patient had a complete standard of care evaluation. Stroke neurologists blinded to LGE-MRI results characterized the stroke subtypes according to TOAST methods modified by the addition of ESUS to the cryptogenic category. Atrial enhancement was defined on LGE-MRI by an intensity threshold that was determined by expert inspection with initial visualization via a volume-rendering tool and expressed as a percentage. We used a t-test to compare mean percent atrial fibrosis between ESUS and non-ESUS patients. Results: We enrolled 17 patients: 3 (17%) female; mean age 53 years (SD 16); and 10 (59%) with ESUS. Age and sex did not differ between ESUS and non-ESUS patients. Atrial fibrosis estimation by LGE-MRI was successfully quantified in all patients. Mean left atrial fibrosis for the entire sample was 9.8% (SD 7.9) and was significantly higher in ESUS (14.1%, SD 6.6%) versus non-ESUS patients (3.6%, SD 5.2%), p=0.003. See Figure. Conclusions: Compared to non-ESUS stroke patients, ESUS patients had significantly higher percentages of atrial fibrosis, supporting atrial cardiomyopathy as one possible mechanism for ESUS. Larger sample sizes are needed to confirm that LGE-MRI is a useful tool to reduce uncertainty about the cause of stroke in the ESUS population.