Myocardial Scar Research Articles

BOATMAP: Bayesian Optimization Active Targeting for Monomorphic Arrhythmia Pace-mapping

Recent advances in machine learning and deep learning have presented new opportunities for learning to localize the origin of ventricular activation from 12-lead electrocardiograms (ECGs), an important step in guiding ablation therapies for ventricular tachycardia. Passively learning from population data is faced with challenges due to significant variations among subjects, and building a patient-specific model raises the open question of where to select pace-mapping data for training. This work introduces BOATMAP, a novel active learning approach designed to provide clinicians with interpretable guidance that progressively assists in locating the origin of ventricular activation from 12-lead ECGs. BOATMAP inverts the input–output relationship in traditional machine learning solutions to this problem and learns the similarity between a target ECG and a paced ECG as a function of the pacing site coordinates. Using Gaussian processes (GP) as a surrogate model, BOATMAP iteratively refines the estimated similarity landscape while providing suggestions to clinicians regarding the next optimal pacing site. Furthermore, it can incorporate constraints to avoid suggesting pacing in non-viable regions such as the core of the myocardial scar. Tested in a realistic simulation environment in various heart geometries and tissue properties, BOATMAP demonstrated the ability to accurately localize the origin of activation, achieving an average localization accuracy of 3.9±3.6mm with only 8.0±4.0 pacing sites. BOATMAP offers real-time interpretable guidance for accurate localization and enhancing clinical decision-making.

Diagnostic performance of the fragmented QRS complex on electrocardiogram for detecting myocardial scars assessed by 3.0 Tesla cardiac magnetic resonance imaging

Abstract Background Fragmented QRS complex (f-QRS) on a 12-lead electrocardiogram (EKG) with a 100 Hz low-pass filter is known to be related to ischemic myocardial scars. Cardiac magnetic resonance (CMR) imaging enhances tissue characterization capability resulting in a better myocardial scar assessment over other noninvasive imaging modalities. However, the diagnostic values of f-QRS on non-ischemic scars and f-QRS from EKG with a 40 Hz low-pass filter (routine filter in clinical practice) are unknown. Therefore, this study aims to evaluate the diagnostic performance of f-QRS (from EKG with 40 Hz and 100 Hz low-pass filters) for detecting any myocardial scars assessed by 3.0 Tesla CMR. Methods This cross-sectional study included patients who underwent a 3.0 Tesla CMR scan from May 2020 to May 2023. A 12-lead EKG with 0.15-40 and 0.15-100 Hz low-pass filters, performed on the same day of the CMR scan, was assessed for the presence of f-QRS. The ECG leads were divided into 3 categories (e.g., anteroseptal leads V1-V4; lateral leads I, aVL, V5-V6; and inferior leads II, III, aVF). The f-QRS was defined as the presence of R’ wave or notching in the nadir of the S wave in 2 contiguous leads. The primary outcome was the diagnostic performance of f-QRS from EKG in myocardial scar detection in the corresponding left ventricle (LV) segments. The secondary outcomes were to compare diagnostic performance between f-QRS from EKG with 0.15-40 and 0.15-100 Hz low-pass filters. Results The study involved 1,692 participants with a median age of 67 (IQR: 59-85) years old and 52.5% males. Myocardial scars were found in 826 (49%) participants. Male, history of CAD, and myocardial scars were significantly more frequent in the participants with f-QRS (59.4% vs 46%, 26.4% vs 20.6%, and 48.9% vs 37.3%, respectively). The sensitivity, specificity, positive predictive value, negative predictive value, and AUC of f-QRS from EKG with 0.15-100 Hz low-pass filter for detecting myocardial scars were 25.6%, 88.7%, 45.1%, 76.8%, and 0.57 for anteroseptal segments; 22.1%, 91.5%, 36.8%, 84.1%, and 0.57 for lateral segments; and 42.9%, 63.4%, 36.9, 68.9%, and 0.53 for inferior segments. The f-QRS from 0.15-100 Hz showed a higher sensitivity but lower specificity and PPV for all LV segments. NPV and AUC were not significantly different between the two filters. Conclusion This study demonstrates a high specificity and negative predictive value of f-QRS from a 12-lead EKG with 0.15-40 and 0.15-100 Hz low-pass filters in diagnosing myocardial scars when correlated to the corresponding LV segments.central illustrationbaseline charateristic

Three-dimensional deformation imaging to detect myocardial scar

Abstract Introduction Myocardial scar or fibrosis has significant implications in patient management and prognosis. Currently, cardiovascular magnetic resonance (CMR) is the reference technique for detection and quantification of scar though other imaging modalities have been proven to be useful as well. Purpose This study aims to investigate the usefulness of 3D strain echocardiography to identify myocardial scar in patients with ischaemic and non-ischaemic myocardial fibrosis using late gadolinium enhancement (LGE) CMR as a reference. Methods 149 individuals with ischaemic (89), non-ischaemic scar (26) and no scar (34) on LGE CMR underwent 3D transthoracic echocardiography. The 3D data sets were processed with the 4D LV analysis tool by TOMTEC Imaging Systems GmbH, Germany. The longitudinal, circumferential, radial, and principal tangential strain were calculated. Using a 16 LV-segment model the areas of LGE were assessed on CMR images, and a correlation between 3D strain values and LGE was investigated. Results A multivariate binary logistic regression model was built with the conditional backward elimination method using LGE as the dependent variable. The predictors made a significant contribution to the model (Wald test: 49.985, p<0.001). The Cox & Snell R-square was 0.141, and Nagelkerke R-square was 0.232, suggesting that the final model explains 14.1% or 23.2% of LGE variability. By logistic regression analysis, less negative peak principal tangential strain (P-PTS), peak circumferential strain (P-SC), and end-systolic circumferential strain (ES-CS) increased the odds, whereas increased peak radial strain (P-RS) decreased the likelihood of LGE (Table 1). The LGE extent was calculated as the average LGE transmurality of all 16 segments and was used as the dependent variable in linear regression models. The LGE extent was significantly correlated with all strain indices (Table 2). By stepwise linear regression analysis, P-PTS (B=0.029, 95%CI:0.019-0.040, P<0.001) outperformed the remaining strain variables in terms of linear association with LGE extent. Association between strain indices and the presence or absence of myocardial fibrosis in each of the 16 segments was tested by point biserial correlation (Table 3). PTS and CS indices were associated with myocardial fibrosis in 10/16 segments; RS indices were associated with LGE in 9/16 segments, and LS was not related to the presence of LGE in 15/16 segments. Septal (basal-mid-apical) and mid/apical anterior and inferior segmental strain showed more consistent associations with corresponding myocardial fibrosis. Conclusion 3D strain echocardiography is a promising and easily accessible technique, that can be used to identify areas of myocardial scar as it correlates well with LGE in many LV segments. Echocardiography is a bedside and widely available technique, and it could be used in patients with contraindications or no access to CMR or myocardial perfusion scan.Table 3

Interplay between strain pattern-based classification of dyssynchrony and myocardial scar in CRT patients

Abstract Introduction Echocardiography-derived septal strain patterns are valuable for predicting volumetric reverse remodelling and survival in patients undergoing cardiac resynchronization therapy (CRT). Nevertheless, presence of myocardial scarring in the left ventricle (LV) is known to attenuate CRT treatment response. Purpose This study aims to investigate the interplay between septal strain patterns and myocardial scarring in CRT candidates, and how they link to outcome after CRT. Methods In this prospective multicentre study, CRT candidates from five European institutions underwent pre-implantation assessments using speckle-tracking strain analysis on echocardiography and cardiac magnetic resonance imaging (CMR) with late gadolinium enhancement (LGE). Dyssynchrony was graded through analysis of septal strain curves, categorizing patients into five distinct patterns, denoted as strain Stage-0 through strain Stage-4 (Fig. A). LGE of the LV wall was quantitatively assessed using CMR (as a percentage). The study endpoints were volumetric reverse remodelling (volume responder = more than 15% decrease LV end-systolic volume) after one year, and all-time all-cause mortality. Results A total of 267 patients, with a mean age of 66±11 years (69% males, 90% left bundle branch block [LBBB]) presented with an average LV ejection fraction (EF) of 30±7%. With each stepwise increase in strain stage, global LV scarring decreased (p<0.01; Fig. B) and the proportion of volume responders increased (all p<0.0001; Fig. C). ROC curves demonstrated that the predictive power of strain stages for CRT volumetric response was not significantly improved when scarring percentage was added (AUC 0.78 [95% CI: 0.69-0.83] vs. 0.81 [95% CI: 0.71-0.90]; p>0.05; Fig. D). In a multivariate cox regression model – including clinical characteristics that were significantly related to survival in univariate analysis – a patients’ strain stage was the only significant predictor of survival, also outperforming scar burden as determinant of outcome (HR 0.68; p=0.041; Fig. E). Conclusions Septal strain patterns show a strong association with the extent of reverse remodelling and outcome in CRT-treated patients. An inverse relationship is observed between pre-implant strain stages and the extent of myocardial scarring. However, scar burden did not provide incremental predictive value over the strain stages. Strain-pattern-based dyssynchrony assessment is a meaningful predictive marker which integrates information on myocardial dysfunction due to dyssynchrony as well as scar burden.

Prediction of scar and microvascular obstruction after myocardial infarction

Abstract Background Myocardial scar, demonstrated by late gadolinium enhancement (LGE), and microvascular obstruction (MVO) at cardiac magnetic resonance imaging (CMR) are well described risk factors for cardiac events after myocardial infarction (MI). Purpose This study aimed to evaluate the prediction of myocardial scar and MVO based on cardiac biomarker levels. Methods We evaluated 631 patients with CMR 3 days [interquartile range 2-4 days] after myocardial infarction (420 with ST-segment elevation and 211 with non-ST-segment elevation myocardial infarction). Cardiac biomarker levels were measured at admission as well as 8, 16, 24 and 48 hours after percutaneous coronary intervention (PCI). Results A myocardial scar was detectable in 586 patients (92.9%), microvascular obstruction in 170 patients (29.3%). Patients with LGE showed significantly higher peak hs-cTnT levels than those without (3908.2±4374.0 ng/l vs 416.8±426.8 ng/l, p<0.001). A cut-off value of hs-cTnT >688.5 ng/l was a predictor for scar with a sensitivity of 81.7% and a specificity of 84.4%. Patients with MVO showed significantly higher peak hs-cTnT levels than those without (6909.8±5525.1 ng/l vs 2467.1±2993.0 ng/l, p<0.001). A cut-off value of hs-cTnT >1500.0 ng/l was a predictor for MVO with a sensitivity of 92.9% and a specificity of 43.4%. Conclusions Higher peak hs-cTnT levels are strong predictors for the occurrence of myocardial scar and MVO at CMR after MI as risk factors for future myocardial events. We propose the new cut-off value of hs-cTnT of >688.5 ng/l within 48 hours after PCI for prediction of myocardial scar and hs-cTnT of >1500.0 ng/l for prediction of MVO. This could be used as a screening tool for post-myocardial infarction patients, who may benefit from further evaluation using CMR.

Cardiac structural, functional remodelling, and perfusion impairments across the spectrum of aortic stenosis

Abstract Background Aortic stenosis (AS) accounts for substantial global morbidity and premature mortality even in moderate AS (Mod-AS). Whilst myocardial remodeling response is considered critical in the adverse prognosis of Mod-AS, the precise mechanisms remain poorly understood. We aimed to prospectively assess myocardial remodeling, perfusion and energetics differences in Mod-AS and severe AS (Severe-AS). Methods Fifty-two Severe-AS and 25 Mod-AS patients and 18 demographically-matched controls underwent cardiovascular magnetic resonance and phosphorus-magnetic resonance spectroscopy to define left ventricular (LV) mass and function, global longitudinal shortening (GLS), rest and adenosine-stress myocardial blood flow (MBF), myocardial perfusion reserve (MPR), layer-specific perfusion metrics (subendocardial [Endo], subepicardial [Epi] MBF and MPR, and Endo-Epi-MBF ratio [Endo/Epi]), myocardial scar on late gadolinium enhancement (LGE) imaging, and myocardial energetics (phosphocreatine:ATP ratio [PCr/ATP]). Results Compared to controls, with increasing AS severity, there was progressive increase in LV concentricity (LV mass-index (controls: 46[40,51],Mod-AS:58[51,65],Severe-AS: 70[65,75]g/m2; P<0.0001), and stepwise decline in GLS (controls:19.9[17.6,22.2], Mod-AS:17.7[16.6,18.8], Severe-AS:13.4[12.5,14.4]%; P<0.0001) with significant differences between Mod-AS and Severe-AS in all three comparisons. Both stress MBF(controls:2.1[1.9,2.3],Mod-AS:1.9[1.6,2.2],Severe-AS:1.3[1.2,1.5]ml/min/g; P<0.0001) and MPR(controls:3.3[2.8,3.6],Mod-AS:2.8[2.4,3.2], Severe-AS:1.9[1.8,2.1]; P<0.0001) were only significantly reduced in Severe-AS compared to controls, with significant differences also detected between Mod-AS and Severe-AS. However, stress-endo-MBF (controls:2.0[1.8,2.3], Mod-AS:1.7[1.5,2.0], Severe-AS:1.2[1.1,1.3]ml/min/g; P<0.0001), stress-Endo/Epi(controls:1.00[0.93,1.07],Mod-AS:0.87[0.80,0.94],Severe-AS:0.81[0.75,0.82];P<0.0001), rest-Endo/Epi (controls:1.12[1.10,1.14], Mod-AS:1.06[1.03,1.09], Severe-AS:1.03[1.02,1.06]; P<0.0001) and endo-MPR (controls:3.2[2.7,3.6],Mod-AS:2.5[2.1,2.9],Severe-AS:1.7[1.5,1.8]; P<0.0001) were all significantly reduced in both Mod-AS and Severe-AS. Compared to controls, both AS groups showed significantly lower PCr/ATP (controls:2.2[2.0,2.3],Mod-AS:1.8[1.6,2.0],Severe-AS:1.7[1.6,1.8]; P<0.0001). Only the Severe-AS group had evidence of non-ischemic myocardial scarring on LGE (2.9[0.0,6.2]%), which was detected in 65%(n=34) of patients. AS severity (peak aortic valve velocity) correlated with stress-MBF (r=-0.45, P=0.0003), MPR (r=-0.44, P=0.0005) and GLS (r=-0.47, P=0.0001). Conclusions Moderate and severe-AS are both associated with cardiac concentric hypertrophy, reductions in myocardial energetics, and subendocardial hypoperfusion. Patients with Severe-AS exhibit a more pronounced phenotype with worse LV hypertrophy, contractile dysfunction and myocardial scarring compared to Mod-AS patients.

Accuracy of a non-invasive mapping system for the localisation of re-entrant VT site of origin and its relationship to myocardial scar on cross-sectional imaging

Abstract Introduction View in ventricular onset (VIVO) is a non-invasive mapping system used for localising the site of earliest activation in ventricular arrhythmia, utilising a mathematical algorithm, patient specific cardiac model (constructed using cross-sectional imaging data), 3D images of the patient’s torso (for localising surface electrodes), and a 12-lead electrocardiogram (ECG).(1, 2) Several publications have examined its role in the localisation of premature ventricular contractions (PVCs), and a 5 patient case series assessed its role in scar related re-entrant VT.(1-4) However, its validity in a larger cohort and of the relationship to the relevant myocardial scar has not been investigated. Objective Assess the accuracy of the VIVO mapping system in localising the VT-SoO for patients with scar related re-entrant VT and its relationship to the relevant myocardial scar on cross-sectional imaging. Methods 20 patients with structural heart disease (18 ischaemic cardiomyopathy (ICM), 1 dilated cardiomyopathy (DCM), 1 hypertrophic cardiomyopathy (HCM)) (Male n=18, 63±14 years) and recent cross-sectional cardiac imaging, were recruited over a 16-month period (table 1). All patients had a clinical indication for VT ablation and were on optimal medical therapy, 19 had an implantable cardiac defibrillator. Invasive electro-anatomical mapping (EAM) was performed with the Advisor HD Grid multipolar catheter and maps were generated using Omnipolar electrograms (EGMs). The VT-SoO was identified using an activation- or pace-map by an experienced operator and the location defined using the American heart association’s 17 segment model of the left ventricle during the procedure.(5) VIVO maps were reviewed by a second independent operator and a segment allocated, scar segments were obtained from cross-sectional imaging. A "complete match" was defined as exact segment concordance between allocated segments, "partial match" as adjacent segments, and "no match" if it does not satisfy either of those requirements. Results Mean left ventricular ejection fraction was 35.5 ± 10.6%. Mean procedure time was 242 ± 70 minutes, with a mean ablation time of 20 ± 11.1 minutes. A total of 32 re-entrant VTs were mapped. The VT exit site was identified in all cases (11 activation-map; 21 pace-map). A complete match between the EAM and VIVO map was seen in 75% of VTs and partial match in a further 15% (table 2). The VT-SoO was located withing the myocardial scar (or directly adjacent to is) in 83% of accurately mapped VTs. Procedural success (defined as freedom from device treated episodes) was seen in 90% of patients at mean follow up of 7.3 ± 4.7 months. Conclusion VIVO non-invasive mapping system was able to accurately map the VT-SoO in scar dependent VT, and identify the relevant myocardial scar as seen on cross sectional imaging. Further research assessing its ability to accurately identify relevant ablation targets is ongoing.

Prevalence of cardiomyopathies and myocardial scar in patients with non-complicated premature ventricular beats: a CMR study

Abstract Introduction Frequent or symptomatic premature ventricular beats (PVBs) in subjects with no personal or family history of cardiac disease or sudden death is a recurrent motive for cardiac magnetic resonance (CMR) prescriptions. This represents a significant burden on healthcare services, given the low availability of CMR technique. If CMR is a recommended imaging modality in patients with history of cardiac arrest due to ventricular arrhythmia, sustained ventricular arrhythmias or when first-line tests are abnormal, its diagnostic value in this specific setting has not been evaluated. Purpose To determine the prevalence of significant CMR abnormalities (myocardial fibrosis, arrhythmogenic cardiomyopathies) in patients with no personal nor family history of cardiac disease and non-complicated PVBs. Methods Between September 2010 and July 2021, 416 healthy adults with no personal nor family cardiac disease history underwent CMR at our institution to explore non complicated PVBs, and were retrospectively included (Figure 1). Right and left ventricular (RV/LV) function and dimensions were systematically evaluated on CMR, as the presence of myocardial late gadolinium enhancement (LGE). Criteria for cardiomyopathies (hypertrophic cardiomyopathy (HCM), dilated cardiomyopathy (DCM), ARVD), the presence and location of myocardial fibrosis were looked after. Results A total of 605 patients were referred for CMR to explore frequent PVBs, representing 14% of all tests performed during this period. Among them, 12 patients (3%) presented significant myocardial abnormalities possibly related to PVBs: 2 patients had sequelae of myocardial infarction and 3 of myocarditis, 1 patient had criteria of HCM and 1 of ARVD, finally 5 presented non-specific myocardial fibrosis (septal mid-wall LGE). In addition, after a mean follow-up of 5 years, there was no death in patients with abnormal CMR, and only 2 patients underwent transcatheter ablation for focal non complicated but symptomatic PVBs. There was no hospitalization for heart failure. Conclusion In patients with non-complicated PVBs and no history of cardiac disease, the rate of cardiomyopathy pattern or myocardial scar on CMR is very low (3%). In this specific population, events were very rare: frequent isolated PVBs, even when associated with myocardial abnormalities, were a benign condition in our study. Given the high volume of patients referred for this indication in daily practice (14% of total CMRs in our center), the limited availability and high costs of CMR, and the difficulties of CMR sequences acquisition in the context of frequent PVBs (possibly leading to misdiagnosis), these data suggest that we should refine the selection of patients undergoing CMR for PVBs exploration.

Ringlike late gadolinium enhancement: prevalence, association with cardiac phenotype, and patient outcome in an unselected cohort referred for cardiovascular magnetic resonance imaging

Abstract Background The identification of a non-ischemic ringlike enhancement in the left ventricular (LV) myocardium by cardiac magnetic resonance imaging (CMR) has emerged as a significant biomarker associated with arrhythmogenic cardiomyopathy (CMP) and adverse patient outcomes. Recently, the term "scarring CMP" has been also proposed to encompass various conditions characterized by a high burden of non-ischemic, often ring-like, myocardial scarring, including both genetic and acquired etiologies. Purpose To systematically evaluate CMR phenotypic traits and etiologic backgrounds in patients with ringlike enhancement and to assess the burden of adverse outcomes during follow-up (FU). Methods This is a single-center, retrospective analysis of prospectively enrolled patients undergoing CMR (from 2002 to 2024). Ringlike late gadolinium enhancement (LGE) was defined as continuous enhancement in three or more adjacent segments based on the AHA 17 segment model. Patients records were reviewed for etiologic classification and FU assessment. Results Among 23.472 patients, 19.696 (84%) underwent LGE assessment. In 152 patients (0.77%; 73% male; mean age 48.5 ± 17 years) a ringlike LGE was identified. The mean number of LV segments with LGE was 10 ± 3. Right ventricular (RV) LGE was present in 23% of subjects. LV fatty metaplasia was detected in 22%. Infero-lateral segments were most frequently affected. CMR showed the following morpho-functional features: LV dilatation (53%), LV hypokinesia (69%), LV hypertrophy (10.5%), LV non-compaction (11%), RV dilation (26%), and RV ejection fraction < 40% (21%). Phenotypic classification according to 2023 ESC guidelines identified dilated CMP phenotype (43%), non-dilated LV CMP phenotype (39%), predominant arrhythmogenic right ventricular phenotype (6%), hypertrophic phenotype (7%). 5% of cases remained unclassified, all of them showing LV dilated non-hypokinetic phenotype. Genetic analysis was performed in 101 patients (66%). Etiologic classification revealed non-desmosomal gene related CMP (24%), desmosomal gene related CMP (8%), inflammatory CMP (16%), genetic neuromuscular diseases (5%), congenital metabolic errors (3%), familiar gene elusive CMP (2%) and unknown etiology (45%). Notably, only 17% of patients classified as having inflammatory CMP and 66% of those with unknown etiology underwent genetic testing, with negative results. The median follow-up duration was 3 (IQR: 1-7) years. All-cause mortality occurred in 10.5% of patients, while 17% experienced sustained ventricular tachycardia, appropriate implantable cardioverter-defibrillator therapies, or sudden cardiac death. Heart transplantation or left ventricular assist device implantation was required in 8% of cases. Conclusion Ringlike LGE is a rare, non-disease-specific feature that can be found in different morpho-functional CMP phenotypes. It is associated with frequent genetic-determined etiology and high burden of arrhythmic and non-arrhythmic outcomes.

ECGWiz: a unified and generalizable AI model for comprehensive ECG diagnosis

Abstract Background Deep learning has demonstrated diagnostic performance in electrocardiogram (ECG) analysis comparable to that of clinicians. However, existing models primarily excel in diagnosing common conditions such as arrhythmias and heart blocks, which are easily identifiable using standard predictor variables. Conversely, models aiming to uncover novel insights, such as detecting structural changes or the presence of myocardial scar, often lack generalizability due to insufficient data, model architecture limitations, and other factors. Moreover, these models frequently encounter challenges, such as a decline in performance when applied to external datasets. To address these issues comprehensively, we propose the ECGWiz — a unified model capable of tackling diverse ECG-related problems with consistently high performance. Objective To develop a generalizable Foundation AI model for ECG diagnosis capable of detecting diverse cardiac abnormalities and diseases using few training samples. Methods We utilized publicly available datasets like PTB-XL, CPSC, etc. To build our model, totaling approximately 80,000 ECGs. These 12 lead ECGs are scaled and filtered before using them for training. The AI model is pretrained using Self Supervised Learning and Contrastive Learning strategies. Further, the model undergoes supervised training using a hierarchy of labels that increase in granularity, encompassing arrhythmias, heart blocks, other cardiac abnormalities, heart failure, sudden cardiac death and more. Lastly, the model is prepared for Few-Shot Learning — where the model learns from a minimal set of samples (5 to 10 ECGs) of the target disease. For testing the model, we are using unique datasets like PROSE which consists of non-ischemic cardiomyopathy patients with and without myocardial scaring detected using cardiac MRI imaging at the Johns Hopkins University. Results Our model, trained on a dataset of only 5 ECG recordings, demonstrated a 74% accuracy in identifying myocardial scarring as detected by LGE MRI, utilizing a standard 12-lead ECG. This performance is comparable to that of conventional deep learning methods, which necessitate datasets containing upwards of 120 ECG recordings. Conclusion Our model represents a promising stride in creating a unified and adaptable foundation AI model for comprehensive ECG-based diagnosis. Its capacity to learn from a minimal set of examples signifies a robust framework that can evolve and adapt to new challenges in ECG diagnosis. Ongoing improvements, particularly through dataset expansion, underscore its potential to enhance cardiovascular diagnostics, offering precise and efficient patient care.

Deep learning for quantification of myocardial scar and microvascular obstruction in late gadolinium enhancement cardiovascular magnetic resonance images

Abstract Objectives Detecting and measuring myocardial scar and microvascular obstruction (MVO) accurately following reperfusion therapy is clinically vital for patients with acute myocardial infarction (AMI). Current clinical practice relies on manual delineation of endocardial borders as well as hyper- and hypo-enhanced regions on late gadolinium enhancement (LGE) cardiovascular magnetic resonance (CMR) images for scar and MVO quantification, respectively. However, this approach is subjective, labor-intensive, and time-consuming. The objective of this study is to develop and assess a convolutional neural network (CNN)-based method for the automated quantification of LGE scar and MVO in AMI patients. Methods A total of 1836 LGE images from 252 patients with AMI were analyzed, with the cohort comprising 232 males and an average age of 56±9 years. Our approach, utilizing a combined architecture of cascaded YOLOv8 (You Only Look Once version 8) and nnU-Net, was implemented for the simultaneous segmentation of the left ventricular (LV) myocardium, scar, and MVO in LGE images. Independent datasets for training, validation, and testing were maintained in an 8:1:1 ratio. Performance evaluation was conducted by comparing the results to manual scar and MVO quantification, utilizing metrics such as the Dice similarity coefficient (DSC) and Bland-Altman analysis. Results The automated segmentation process took 0.05 seconds per image. The segmentation accuracy, as assessed by the DSC, yielded values of 0.96±0.04 (per-patient, n=25) and 0.96±0.05 (per-slice, n=184) for scar segmentation, and 0.86±0.17 (per-patient) and 0.89±0.15 (per-slice) for MVO segmentation. In per-patient analysis, the volumes of automatically segmented scar and MVO demonstrated strong agreement with manually derived results, with biases [limits of agreement] of -0.3 [-3.0, 2.4] cm³ and 0.1 [-0.2, 0.3] cm³, respectively. Furthermore, there was strong concordance between automatically and manually derived scar size as a percentage of LV mass (-0.2 [-3.1, 2.7] %) and MVO size as a percentage of scar (1.2 [-5.6, 8.0] %). Conclusions The findings demonstrate excellent agreement between our proposed automated deep-learning solution and clinical expert annotation, suggesting its potential as a predictive tool for prognosis and clinical outcomes in patients with AMI.

T2 mapping as a marker of active myocardial injury in Anderson-Fabry disease patients

Abstract Background In Anderson-Fabry Disease (AFD) with cardiac involvement, left ventricular hypertrophy (LVH) is often associated with myocardial scarring by cardiovascular magnetic resonance (CMR) late gadolinium enhancement (LGE) imaging. Myocardial inflammation, detected as increased T2 weighted/T2 mapping (T2m) signal, has also been reported, although its role in disease progression is still unclear. Purpose To investigate the relationship between T2m and LGE/LVH progression in AFD patients. Methods 67 paired CMR scans (1.5T) from 30 AFD patients were included. Three groups were defined according to LGE: no LGE at either scan (LGE-/-); LGE at baseline, unchanged at follow-up (LGE+/-); LGE at follow-up, appeared or increased compared to baseline (LGE+/+). T2m was measured in the infero-lateral basal segment in all scans; in LGE+ scans only, LGE was also manually contoured (ROI_LGE), and the ROI_LGE copied and pasted on the matching T2m image (T2m_ROI_LGE). See Figure 1. Troponin I (TnI) was also dosed at each scan. A univariate repeated measures analysis of variance was used to compare the groups and Tukey-Kramer correction was applied. Results As expected, the presence of LGE was strongly associated with older age, higher global and regional T2m, left ventricular (LV) mass and TnI values. As a novelty, in the LGE+/+ scans (scar appearance/progression) T2m in the ROI_LGE was higher than in the LGE+/- scans (scar unchanged) (p value 0.02). None of the other parameters, such as LV mass, TnI and global T2m values, differ between the the LGE+/+ compared to the LGE +/- group. See Table 1. Conclusions In our AFD patients’ cohort, scar progression was associated with focal increase in T2m values, i.e. T2m may identify patients with active myocardial injury. The potential implication on management/disease progression monitoring needs to be tested in future research, with longer follow-up. The role of sex, with females known to have higher average T2m values, will also need to be explored in a wider patient’s cohort.

Myocardial scar in end-stage hypertrophic cardiomyopathy: correlation with systolic function and prognostic significance

Abstract Background Hypertrophic cardiomyopathy with systolic impairment, often referred to as end-stage HCM (ES-HCM), has a poor prognosis. Scar is a significant contributor but the patterns and prognostic significance are unknown. Aim To understand role of scar in ES-HCM and investigate potential predictors of outcome. Methods A retrospective 4 centre study of ES-HCM (LVEF≤55%) who had undergone CMR between 2006 and 2022 for unexplained LVH. Exclusion criteria were: scaring therapies (septal reduction), phenocopies (amyloidosis, storage) and dual pathologies (severe aortic valve disease, previous infarction). CMR followed standardized protocols. All images were core-lab analysed de-novo with scar quantified using the full width at half maximum technique for late gadolinium enhancement (LGE) quantification in grams (LGEm) and percentage of total LV mass (LGE%). Cox models used to identify risk factors for all-cause mortality. Results From 3810 HCM CMR scans, 128 were ES-HCM pts (3%), 25(20%) were excluded due to known infarction. 103 (male n=82, 81%; age 57yrs[IQR51–68]) formed the study cohort. Of the 54% genotyped, 62% carried a (likely)pathogenic sarcomere mutation: MYBPC3 (n=20), MYH7 (n=7), TNNT2 (n=4), TMP1 (n=2),ALPK3 (n=1). Heart morphology was: 38(37%) isolated basal septal hypertrophy, 30(29%) reverse septal curvature, 7(7%) mid-cavity LVH with apical aneurysm, 3(3%) apical HCM, 24%other. The median EF was 46%(IQR 39-50). In only one third (34%) the ventricle was dilated (LVEDd mean 97ml/m2 IQR 79-112). The RV was impaired in 19(18%). Median max wall thickness was 17mm(IQR15-20) and LVmass 74mg/Kg(IQR65-90). Scar was present in 94(93%) and was typically extensive: median LGE% 23%(IQR 13-33), LGEm 31g(IQR 16-45) – although 7% had no scar. The LGE pattern was diffuse in 38%, patchy in 35%, ring-like in 9%, and infarct-like in 7% (all without coronary narrowing at invasive/non-invasive coronary artery assessment). LGE and cardiac function were effectively independent with low or no correlation between scar and LVEF% or global longitudinal strain (r2=0.015,p=0.227 or r2= 0.048,p=0.028 for LGE% and LGEm against LVEF; r2=0.059,p=0.021 or r2=0.196,p<0.001 for LGE% and LGEm against GLS) with poor correlation also with indexed stroke volume. During a 5 year (IQR 3-7) follow-up (516patient-years), 30 deaths occurred: mean age at death was 62(52-74). Using multivariate Cox regression analysis EF and demographics such as age were not predictive, but LGE% (HR 1.046, 95%IC 1.007-1.086, p=0.019 and GLS (HR 1.172, 95%IC 1.004-1.368, p=0.044) were independent all-cause mortality predictors (C-index 0.708; table). Conclusion Myocardial scar is almost always present in ES-HCM and is typically extensive – but with some unexpected features: 7% of patients have no LGE, dual pathology(infarction) is not rare, and ringlike LGE (ALVClike) may occur. LGE correlates poorly with any measure of LV impairment but is a strong predictor of mortality, as is GLS.

Non-invasive imaging to guide percutaneous coronary revascularization in chronic total occlusion: a systematic review and meta-analysis

Abstract Background and aims Whether recanalization of coronary chronic total occlusions (CTO) leads to improved patient outcomes is uncertain. We set out to investigate the use of noninvasive imaging modalities to guide patient selection for percutaneous recanalization of coronary chronic total occlusions (CTO-PCI) and its association with outcomes. Methods Systematic review and meta-analysis of studies involving patients undergoing CTO-PCI to detail use of preprocedural non-invasive imaging, change in clinical endpoints and the occurrence of clinical events in these patient cohorts. PubMed and Embase, the Cochrane Database of Systematic Reviews, the PROSPERO database and the Clinical Trials Registry were searched for relevant randomized and observational studies up to May 2023. This study is registered with PROSPERO. The study primary endpoint was a composite of all-cause death and nonfatal myocardial infarction (MI), with a secondary composite endpoint of cardiovascular death, nonfatal MI and target vessel revascularization (TVR). Additional endpoints were individual components of the composite endpoints along with changes in angina burden, LV ejection fraction, ischemic burden, and infarct extent after PCI. Results Of 1264 publications retrieved, 31 studies (26 observational and 5 randomized) were finally included for a total of 7260 patients. Guidance to CTO-PCI varied greatly among included studies, with only a minority implementing routine preprocedural inducible ischemia and myocardial viability assessment. At a median of 3.1 years (range 1-5 years), the primary endpoint occurred at a rate of 3.1 events per 100 patient-years (95%CI: 2.4-3.7). When non-invasive imaging was used to guide CTO-PCI, patients with myocardial ischemia and/or viability, a reduced risk for the primary endpoint was observed (OR 0.3 95%CI 0.2-0.5) compared to CTO-PCI without evidence of myocardial ischemia and/or viability. CTO-PCI was associated with improvements in SAQ summary score, LVEF, reduction of ischemic burden, improvement in myocardial perfusion (all P-values < 0.01), with no change in the burden of myocardial scarring. At meta-regression analysis, a greater improvement in LVEF was observed in studies enrolling patients with lower average baseline LVEF (≤ 45%) (R2 = 63%, p < 0.0001). Conclusions CTO-PCI led to a significant improvement in symptom status (reduction of chest pain) and LV ejection fraction. CTO-PCI guided by the presence of ischemia and/or viability was associated with improved patient outcomes compared with CTO-PCI in the absence of ischemia and/viability, or in patients where non-invasive imaging testing was not performed. Future randomized clinical trials testing whether ischemia and viability-guided CTO-PCI improves clinical outcomes are lacking and warranted.PRISMA Flow diagramGraphical abstract

Histopathological Evaluation of Somatostatin Receptor 2 Expression in Myocarditis—Rationale for the Diagnostic Use of Somatostatin Receptor Imaging

Background/Objectives: Myocarditis is an inflammatory disease of the myocardium and remains to this day a challenging diagnosis. A promising novel imaging method uses the expression of somatostatin receptors (SSTRs) on inflammatory cells to visualize myocardial inflammation. However, little is known about the histopathological correlate of SSTR imaging in different forms of myocarditis. Methods: In the present retrospective histopathological study, we systematically analysed the expression of SSTR subtype 2 (SSTR2) on inflammatory cells of 33 patients with biopsy- or explant-proven myocarditis (lymphocytic myocarditis (n = 5), giant-cell myocarditis (n = 11), and cardiac sarcoidosis (n = 17)), and in eight controls (multi-organ donors) without signs of myocardial inflammation and/or scars. Results: In all patients, immunohistochemical staining for SSTR2 was positive in areas with CD68-positive macrophages and multinucleated giant cells. Staining for SSTR2 was most prominent in the presence of multinucleated giant cells. The colocalization of both SSTR2 and CD68 on the same cell could be confirmed using immunofluorescence microscopy. Western blotting confirmed the upregulated expression of SSTR2 in cases of granulomatous inflammation (sarcoidosis) of the skeletal and heart muscle, in comparison with controls. Conclusions: In conclusion, our findings demonstrate the expression of SSTR2 on the protein level on CD68-positive macrophages and multinucleated giant cells in various forms of myocarditis, which provides a clear rationale for the diagnostic use of SSTR imaging in this patient group.

Dynamic Voltage Mapping of the Post-infarct Ventricular Tachycardia Substrate: A Practical Technique to Help Differentiate Scar from Borderzone Tissue

During catheter ablation of post-infarct ventricular tachycardia (VT), substrate mapping is used when VT is non-inducible or poorly tolerated. Substrate mapping aims to identify regions of slowly conducting myocardium (borderzone) within and surrounding myocardial scar for ablation. Historically, these tissue types have been identified using bipolar voltage mapping, with areas of low bipolar voltage (<0.50 mV) defined as scar, and areas with voltages between 0.50 mV and 1.50 mV as borderzone. In the era of high-density mapping, studies have demonstrated slow conduction within areas of bipolar voltage <0.50 mV, suggesting that this historical cut-off is outdated. While electrophysiologists often adapt voltage cut-offs to account for this, the optimal scar-borderzone threshold is not known. In this review, we discuss dynamic voltage mapping, a novel substrate mapping technique we have developed, which superimposes data from both activation and voltage maps, to help delineate the post-infarct VT circuit through identification of the optimal scar-borderzone voltage threshold.

Return-to-Play Post-Myocarditis for Athletes: To Play or Not to Play?

Myocarditis is a condition marked by inflammation of the heart muscle, which can lead to serious outcomes such as sudden cardiac death (SCD) and life-threatening arrhythmias. While myocarditis can affect any population, athletes, especially those engaged in high-intensity training, are at increased risk due to factors such as reduced immunity and increased exposure to pathogens. This review examines the clinical presentation, current guidelines, diagnostic challenges, and the significance of cardiac magnetic resonance imaging (CMR) in detecting myocardial inflammation and scarring. Current guidelines recommend a period of exercise restriction followed by thorough reassessment before athletes can return-to-play (RTP). However, there are several knowledge gaps, including the implications of persistent late gadolinium enhancement (LGE) on CMR and the optimal duration of exercise restriction. Additionally, the psychological impact of myocarditis on athletes highlights the importance of incorporating mental health support in the recovery process. A shared decision-making approach should be encouraged in RTP, considering the athlete's overall health, personal preferences, and the potential risks of resuming competitive sports. We have proposed an algorithm for RTP in athletes following myocarditis, incorporating CMR. Future research is warranted to refine RTP protocols and improve risk stratification, particularly through longitudinal studies that examine recovery and outcomes in athletes.

The Role of Nuclear Medicine in the Diagnostic Work-Up of Athletes: An Essential Guide for the Sports Cardiologist.

Athletes with heart disease are at increased risk of malignant ventricular arrhythmias and sudden cardiac death compared to their sedentary counterparts. When athletes have symptoms or abnormal findings at preparticipation screenings, a precise diagnosis by differentiating physiological features of the athlete's heart from pathological signs of cardiac disease is as important as it is challenging. While traditional imaging methods such as echocardiography, cardiac magnetic resonance, and computed tomography are commonly employed, nuclear medicine offers unique advantages, especially in scenarios requiring stress-based functional evaluation. This article reviews the use of nuclear medicine techniques in the diagnostic work-up of athletes with suspected cardiac diseases by highlighting their ability to investigate myocardial perfusion, metabolism, and innervation. The article discusses the application of single photon emission computed tomography (SPECT) and positron emission tomography (PET) using radiotracers such as [99mTc]MIBI, [99mTc]HDP, [18F]FDG, and [123I]MIBG. Several clinical scenarios are explored, including athletes with coronary atherosclerosis, congenital coronary anomalies, ventricular arrhythmias, and non-ischemic myocardial scars. Radiation concerns are addressed, highlighting that modern SPECT and PET equipment significantly reduces radiation doses, making these techniques safer for young athletes. We conclude that, despite being underutilized, nuclear medicine provides unique opportunities for accurate diagnosis and effective management of cardiac diseases in athletes.

Lipomatous Metaplasia Facilitates Ventricular Tachycardia in Patients WithNonischemic Cardiomyopathy.

Ventricular tachycardia (VT) substrate in patients with nonischemic cardiomyopathy (NICM) is complex in distribution and intramural location. This study sought to test the hypothesis that myocardial lipomatous metaplasia (LM) is a vital anatomic substrate for VT corridors in patients with NICM and VT, and that LM stabilizes current propagation in VT corridors. Among 49 patients with NICM in the 2-center INFINITY (Prospective Intra-Myocardial Fat Deposition and Ventricular Tachycardia in Cardiomyopathy) Study, potential VT viable corridors within the myocardial scar and/or LM were computed from late gadolinium enhancement cardiac magnetic resonance images and were registered with electroanatomical maps. Corridors passing through VT entrance, isthmus, and/or exit sites, estimated by entrainment or pace mapping, were defined as VT corridors. LM was separately distinguished from scar using computed tomography. The SD of current amplitude along each corridor was measured. Compared with 151 non-VT corridors, 35 VT corridors traversed a substantially higher volume of LM, with a median 236.6mg (IQR: 13.5-903.4mg) vs 5.8mg (IQR: 0.0-57.9mg) (P< 0.001). Among corridors with computable current amplitude, 28 VT corridors exhibited substantially lower current variation along the corridors, with SD 8.0 μA (25th-75th percentile: 6.1-10.3 μA) vs 14.9 μA (25th-75th percentile: 8.5-23.7 μA) among 71 non-VT corridors (P<0.001). Individual VT circuit sites (95 out 118) were highly colocalized with LM. VT circuitry corridors in NICM are more likely to traverse LM and exhibit reduced current amplitude variation compared with bystander corridors.

Diagnosis of arrhythmogenic cardiomyopathy: 20 years of progress and innovation

Arrhythmogenic cardiomyopathy (CMA) is a cardiac disease characterized by non-ischemic ventricular scarring and electrical instability. The diagnosis of CMA still remains challenging today and requires the use of a set of criteria, since no single diagnostic test represents the gold standard. The first diagnostic criteria were defined and disseminated in 1994 and then revised in 2010, focusing mainly on right ventricular involvement. In 2019, an international panel of experts identified the limitations of the previous diagnostic criteria. The 2020 Padua criteria included a specific pathway for the diagnosis of left ventricular variants and emphasized the need for the use of cardiac magnetic resonance imaging in the characterization of myocardial scarring. These criteria were further refined and published in 2023 as European Task Force (TF) criteria, thus gaining international recognition.Exploring the history of CMA and its diagnosis, in this review we analyze the changes and progress in the 20 years that have occurred from the first version of the criteria in 1994 to the latest in European TF of 2023, highlighting the evolution of our knowledge of the pathobiology and morpho-functional characteristics of the disease. One of the most relevant updates is undoubtedly the introduction of the concept of "scarring/arrhythmogenic cardiomyopathy", a definition that enhances the main features of the pathology and emphasizes the multiplicity of phenotypes and clinical presentations independent of etiology.