scholarly journals The Role of AI in Characterizing the DCM Phenotype

2021 ◽  
Vol 8 ◽  
Author(s):  
Clint Asher ◽  
Esther Puyol-Antón ◽  
Maleeha Rizvi ◽  
Bram Ruijsink ◽  
Amedeo Chiribiri ◽  
...  
Keyword(s):  
Artificial Intelligence ◽  
Deep Learning ◽  
Risk Stratification ◽  
Dilated Cardiomyopathy ◽  
Tissue Characterization ◽  
Left Ventricular ◽  
Muscle Disease ◽  
Left Ventricular Ejection ◽  
Ventricular Ejection Fraction ◽  
Care Assessment

Dilated Cardiomyopathy is conventionally defined by left ventricular dilatation and dysfunction in the absence of coronary disease. Emerging evidence suggests many patients remain vulnerable to major adverse outcomes despite clear therapeutic success of modern evidence-based heart failure therapy. In this era of personalized medical care, the conventional assessment of left ventricular ejection fraction falls short in fully predicting evolution and risk of outcomes in this heterogenous group of heart muscle disease, as such, a more refined means of phenotyping this disease appears essential. Cardiac MRI (CMR) is well-placed in this respect, not only for its diagnostic utility, but the wealth of information captured in global and regional function assessment with the addition of unique tissue characterization across different disease states and patient cohorts. Advanced tools are needed to leverage these sensitive metrics and integrate with clinical, genetic and biochemical information for personalized, and more clinically useful characterization of the dilated cardiomyopathy phenotype. Recent advances in artificial intelligence offers the unique opportunity to impact clinical decision making through enhanced precision image-analysis tasks, multi-source extraction of relevant features and seamless integration to enhance understanding, improve diagnosis, and subsequently clinical outcomes. Focusing particularly on deep learning, a subfield of artificial intelligence, that has garnered significant interest in the imaging community, this paper reviews the main developments that could offer more robust disease characterization and risk stratification in the Dilated Cardiomyopathy phenotype. Given its promising utility in the non-invasive assessment of cardiac diseases, we firstly highlight the key applications in CMR, set to enable comprehensive quantitative measures of function beyond the standard of care assessment. Concurrently, we revisit the added value of tissue characterization techniques for risk stratification, showcasing the deep learning platforms that overcome limitations in current clinical workflows and discuss how they could be utilized to better differentiate at-risk subgroups of this phenotype. The final section of this paper is dedicated to the allied clinical applications to imaging, that incorporate artificial intelligence and have harnessed the comprehensive abundance of data from genetics and relevant clinical variables to facilitate better classification and enable enhanced risk prediction for relevant outcomes.

Abstract 15759: Arrhythmic Risk Stratification in Patients With Idiopathic Dilated Cardiomyopathy and No Conventional Indications for Implantable Cardiac Defibrillator

Circulation ◽  
2014 ◽  
Vol 130 (suppl_2) ◽  
Author(s):  
Marco Merlo ◽  
Fabrizio Pirozzi ◽  
Davide Stolfo ◽  
D’Angelo Gianluca ◽  
Marco Alonge ◽  
...  
Keyword(s):  
Risk Stratification ◽  
Dilated Cardiomyopathy ◽  
Previous History ◽  
Left Ventricular ◽  
Muscle Disease ◽  
Idiopathic Dilated Cardiomyopathy ◽  
Left Ventricular Ejection ◽  
Ventricular Ejection Fraction ◽  
History Of ◽  
Duration Of Disease

Background: Primary prevention of major ventricular arrhythmias (MVA) in patients with idiopathic dilated cardiomyopathy (IDCM) is primarily based on left ventricular ejection fraction (LVEF) assessment. Nonetheless, unexpected MVA still affect patients considered at low-risk (LR), attesting the limitation of the current risk stratification. We sought to identify the clinical predictors of MVA in patients with IDCM and no conventional indications for implantable cardioverter defibrillator (ICD). Methods: In this retrospective, case-control study, among 922 patients enrolled in the Heart Muscle Disease Registry of Trieste from 1988 to 2013, we analyzed 414 IDCM patients considered at LR according to the following criteria: LVEF≥36% and no previous episodes of MVA (sustained ventricular tachycardia/non-fatal ventricular fibrillation (SVT/VF), appropriate ICD intervention and sudden cardiac death (SD)). Data were recorded at the last available medical examination before the index event. Patients were optimally treated at the time of index evaluation (87% and 84% of ACE-inhibitors/sartans and beta-blockers, respectively). Results: Over a median follow-up of 43 (IQR 16-116) months, 45 patients (11 % of patients at LR; 31% of the all MVA in whole IDCM population) experienced MVA. The mean age at the time of event was 51±14 years. They were characterized by a previous history of syncope in 10 patients (22%), LVEF of 42±6% with a normal value (>50%) in 10 patients (22%), significant LV dilation (VTDi>90ml/m 2 ) in 16 patients (35%) and left bundle branch block 12 patients (27%). Independent risk factors for MVA in LR subgroup were previous history of syncope (OR 3.41, 1.45-8.03, p=0.004), larger left ventricular dilation (OR 2.62, 1.72-3.99, p<0.0001) and longer duration of disease (OR 1.39, 1.42-2.63, p<0.0001). Conclusions: in a large cohort of IDCM patients nearly one-third of the MVA occurred in a population apparently at LR of events. History of syncope, larger LV dilatation and significant duration of disease emerged as strong predictors of MVA and should be considered in the arrhythmic risk stratification of patients without conventional criteria for SD primary protection.

Arrhythmic risk stratification in non-ischaemic dilated cardiomyopathy beyond ejection fraction

Heart ◽  
2020 ◽  
Vol 106 (9) ◽  
pp. 656-664 ◽  
Author(s):  
Antonio Cannatà ◽  
Giulia De Angelis ◽  
Andrea Boscutti ◽  
Camilla Normand ◽  
Jessica Artico ◽  
...  
Keyword(s):  
Ejection Fraction ◽  
Risk Stratification ◽  
Dilated Cardiomyopathy ◽  
Left Ventricular ◽  
Left Ventricular Ejection ◽  
Ventricular Ejection Fraction ◽  
Novel Approach ◽  
Magnetic Resonance Parameters ◽  
Reverse Remodelling ◽  
Life Threatening

Sudden cardiac death and arrhythmia-related events in patients with non-ischaemic dilated cardiomyopathy (NICM) have been significantly reduced over the last couple of decades as a result of evidence-based pharmacological and non-pharmacological therapeutic strategies. Nevertheless, the arrhythmic stratification in patients with NICM remains extremely challenging, and the simple indication based on left ventricular ejection fraction appears to be insufficient. Therefore, clinicians need to go beyond the current criteria for implantable cardioverter-defibrillator implantation in the direction of a multiparametric evaluation of arrhythmic risk. Several parameters for arrhythmic risk stratification, ranging from electrocardiographic, echocardiographic, imaging-derived and genetic markers, are crucial for proper arrhythmic risk stratification and a multiparametric evaluation of risk in patients with NICM. In particular, integration of cardiac magnetic resonance parameters (mostly late gadolinium enhancement) and specific genetic information (ie, presence of LMNA, PLN, FLNC mutations) appears fundamental for proper implementation of the current arrhythmic risk stratification. Finally, a novel approach focused on both arrhythmic risk and prediction of left ventricular reverse remodelling during follow-up might be useful for effective multiparametric and dynamic arrhythmic risk stratification in NICM. In the future, a complete and integrated evaluation might be mandatory to implement arrhythmic risk prediction in patients with NICM and to discriminate the competing risk between heart failure-related events and life-threatening arrhythmias.

scholarly journals 359 Transient vs. persistent improved ejection fraction in non-ischaemic dilated cardiomyopathy

2021 ◽  
Vol 23 (Supplement_G) ◽  
Author(s):  
Paolo Manca ◽  
Davide Stolfo ◽  
Marco Merlo ◽  
Caterina Gregorio ◽  
Antonio Cannatà ◽  
...  
Keyword(s):  
Ejection Fraction ◽  
Dilated Cardiomyopathy ◽  
Multivariable Analysis ◽  
Left Ventricular ◽  
Muscle Disease ◽  
Left Ventricular Ejection ◽  
Longitudinal Assessment ◽  
Ventricular Assist ◽  
Ventricular Ejection Fraction ◽  
Increased Risk

Abstract Aims The recent definition of heart failure with improved ejection fraction (HFimpEF) outlined the importance of the longitudinal assessment of left ventricular ejection fraction (LVEF). However, long-term progression and outcomes of this subgroup are poorly explored. We sought to assess the LVEF trajectories and the correlations with outcome in non-ischaemic dilated cardiomyopathy (NICM) with improved ejection fraction (impEF). Methods and results Consecutive NICM patients with baseline LVEF ≤40% enrolled in the Trieste Heart Muscle Disease Registry with ≥1 LVEF assessment after baseline were included. ImpEF was defined as a baseline LVEF ≤40%, a ≥ 10 point increase from baseline LVEF, and a second measurement of LVEF &gt;40%. Transient impEF was defined by the documentation of recurrent LVEF ≤40% during follow-up. The primary endpoint was a composite of all-cause death, heart transplantation and left ventricular assist device (D/HT/LVAD). Among 800 patients, 460 (57%) had impEF (median time to improvement 13 months). Higher heart rate, smaller left atrium, absence of severe mitral regurgitation and shorter duration of disease were associated with impEF. ImpEF was independently associated with lower risk of D/HT/LVAD (HR: 0.36; 95% CI: 0.27–0.48; P &lt; 0.001). Transient impEF was observed in 189 patients (41% of the overall impEF group) and was associated with higher risk of D/HT/LVAD compared with persistent impEF at multivariable analysis (HR: 2.54; 95% CI: 1.60–4.04). Conclusions In NICM, we observed a 57% rate of impEF. However, recurrent decline in LVEF was observed in ≈40% of impEF patients and it was associated with an increased risk of D/HT/LVAD.

scholarly journals Diagnosing and sports counselling of athletes with myocarditis

2019 ◽  
Vol 67 (2) ◽  
Keyword(s):  
Risk Stratification ◽  
Tissue Characterization ◽  
T2 Mapping ◽  
Left Ventricular ◽  
Left Ventricular Ejection ◽  
Individual Risk ◽  
Cardiac Events ◽  
Ventricular Ejection Fraction ◽  
Characterization Methods ◽  
Imaging Tool

Myocarditis is defined as an inflammation of the heart muscle and its presentation, especially in athletes, is heterogeneous. Underlying causes include in most of the cases viruses, and less often bacteria, toxins, vasculitic diseases or pharmaceutical agents. Cardiac magnetic resonance (CMR) imaging is the primary imaging tool to diagnose myocarditis following laboratory test, electrocardiogram and echocardiography. In certain cases, endomyocardial biopsy is required, especially in unclear cases with reduced systolic left ventricular ejection fraction. Although, athletes and sport physicians face the dilemma of significant performance decline in competitive athletes against the risk of adverse cardiac events, currently abstinence from competitive sports is recommended for at least 3–6 months in myocarditis. Sports recommendations are currently based mainly on autopsy studies and experts’ opinions and better risk stratification tools are imperatively needed. New tissue characterization methods, namely T1 mapping and T2 mapping in CMR continue to improve sensitivity and specificity of diagnosing myocarditis and may further enhance individual risk assessment. In the future, sports physicians may be able to rely more on these novel noninvasive tissue characterisation methods in risk stratification and sports restriction recommendations of athletes with suspected myocarditis.

scholarly journals Familial dilated cardiomyopathy with RBM20 mutation in an Indian patient: a case report

2021 ◽  
Vol 73 (1) ◽  
Author(s):  
Soumi Das ◽  
Sandeep Seth
Keyword(s):  
Ejection Fraction ◽  
Dilated Cardiomyopathy ◽  
Age Of Onset ◽  
Arrhythmogenic Right Ventricular Cardiomyopathy ◽  
Left Ventricular ◽  
Left Ventricular Ejection ◽  
Early Age ◽  
Indian Patient ◽  
Ventricular Ejection Fraction ◽  
First Case

Abstract Background Dilated cardiomyopathy (DCM) is a disease of the heart muscle characterized by ventricular dilation and a left ventricular ejection fraction of less than 40%. Unlike hypertrophic cardiomyopathy (HCM) and arrhythmogenic right ventricular cardiomyopathy (ARVC), DCM-causing mutations are present in a large number of genes. In the present study, we report a case of the early age of onset of DCM associated with a pathogenic variant in the RBM20 gene in a patient from India. Case presentation A 19-year-old Indian male diagnosed with DCM was suggested for heart transplantation. His ECG showed LBBB and echocardiography showed an ejection fraction of 14%. He had a sudden cardiac death. A detailed family history revealed it to be a case of familial DCM. Genetic screening identified the c.1900C>T variant in the RBM20 gene which led to a missense variant of amino acid 634 (p.Arg634Trp). Conclusion To the best of our knowledge, the variant p.Arg634Trp has been earlier reported in the Western population, but this is the first case of p.Arg634Trp in an Indian patient. The variant has been reported to be pathogenic at an early age of onset; therefore, close clinical follow-up should be done for the family members caring for the variant.

scholarly journals Temporal trends in outcome and patient characteristics in dilated cardiomyopathy, data from the Swedish Heart Failure Registry 2003–2015

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Helen Sjöland ◽  
Jonas Silverdal ◽  
Entela Bollano ◽  
Aldina Pivodic ◽  
Ulf Dahlström ◽  
...  
Keyword(s):  
Heart Failure ◽  
Dilated Cardiomyopathy ◽  
Temporal Trends ◽  
Physical Symptoms ◽  
Left Ventricular ◽  
Left Ventricular Ejection ◽  
Continuous Change ◽  
Ventricular Ejection Fraction ◽  
Increased Risk ◽  
Over Time

Abstract Background Temporal trends in clinical composition and outcome in dilated cardiomyopathy (DCM) are largely unknown, despite considerable advances in heart failure management. We set out to study clinical characteristics and prognosis over time in DCM in Sweden during 2003–2015. Methods DCM patients (n = 7873) from the Swedish Heart Failure Registry were divided into three calendar periods of inclusion, 2003–2007 (Period 1, n = 2029), 2008–2011 (Period 2, n = 3363), 2012–2015 (Period 3, n = 2481). The primary outcome was the composite of all-cause death, transplantation and hospitalization during 1 year after inclusion into the registry. Results Over the three calendar periods patients were older (p = 0.022), the proportion of females increased (mean 22.5%, 26.4%, 27.6%, p = 0.0001), left ventricular ejection fraction was higher (p = 0.0014), and symptoms by New York Heart Association less severe (p < 0.0001). Device (implantable cardioverter defibrillator and/or cardiac resynchronization) therapy increased by 30% over time (mean 11.6%, 12.3%, 15.1%, p < 0.0001). The event rates for mortality, and hospitalization were consistently decreasing over calendar periods (p < 0.0001 for all), whereas transplantation rate was stable. More advanced physical symptoms correlated with an increased risk of a composite outcome over time (p = 0.0043). Conclusions From 2003 until 2015, we observed declining mortality and hospitalizations in DCM, paralleled by a continuous change in both demographic profile and therapy in the DCM population in Sweden, towards a less affected phenotype.

Risk stratification in hf with mid-range LVEF: the role of cardiopulmonary exercise testing

2020 ◽  
Vol 41 (Supplement_2) ◽  
Author(s):  
B.M.L Rocha ◽  
G.J Lopes Da Cunha ◽  
P.M.D Lopes ◽  
P.N Freitas ◽  
F Gama ◽  
...  
Keyword(s):  
Risk Stratification ◽  
Exercise Testing ◽  
Primary Endpoint ◽  
Prognostic Significance ◽  
Cardiopulmonary Exercise Testing ◽  
Left Ventricular ◽  
Left Ventricular Ejection ◽  
Ventricular Ejection Fraction ◽  
Cardiopulmonary Exercise

Abstract Background Cardiopulmonary exercise testing (CPET) is recommended in the evaluation of selected patients with Heart Failure (HF). Notwithstanding, its prognostic significance has mainly been ascertained in those with left ventricular ejection fraction (LVEF) &lt;40% (i.e., HFrEF). The main goal of our study was to assess the role of CPET in risk stratification of HF with mid-range (40–49%) LVEF (i.e., HFmrEF) compared to HFrEF. Methods We conducted a single-center retrospective study of consecutive patients with HF and LVEF &lt;50% who underwent CPET from 2003–2018. The primary composite endpoint of death, heart transplant or HF hospitalization was assessed. Results Overall, 404 HF patients (mean age 57±11 years, 78.2% male, 55.4% ischemic HF) were included, of whom 321 (79.5%) had HFrEF and 83 (20.5%) HFmrEF. Compared to the former, those with HFmrEF had a significantly higher mean peak oxygen uptake (pVO2) (20.2±6.1 vs 16.1±5.0 mL/kg/min; p&lt;0.001), lower median minute ventilation/carbon dioxide production (VE/VCO2) [35.0 (IQR: 29.1–41.2) vs 39.0 (IQR: 32.0–47.0); p=0.002) and fewer patients with exercise oscillatory ventilation (EOV) (22.0 vs 46.3%; p&lt;0.001). Over a median follow-up of 28.7 (IQR: 13.0–92.3) months, 117 (28.9%) patients died, 53 (13.1%) underwent heart transplantation, and 134 (33.2%) had at least one HF hospitalization. In both HFmrEF and HFrEF, pVO2 &lt;12 mL/kg/min, VE/VCO2 &gt;35 and EOV identified patients at higher risk for events (all p&lt;0.05). In Cox regression multivariate analysis, pVO2 was predictive of the primary endpoint in both HFmrEF and HFrEF (HR per +1 mL/kg/min: 0.81; CI: 0.72–0.92; p=0.001; and HR per +1 mL/kg/min: 0.92; CI: 0.87–0.97; p=0.004), as was EOV (HR: 4.79; CI: 1.41–16.39; p=0.012; and HR: 2.15; CI: 1.51–3.07; p&lt;0.001). VE/VCO2, on the other hand, was predictive of events in HFrEF but not in HFmrEF (HR per unit: 1.03; CI: 1.02–1.05; p&lt;0.001; and HR per unit: 0.99; CI: 0.95–1.03; p=0.512, respectively). ROC curve analysis demonstrated that a pVO2 &gt;16.7 and &gt;15.8 mL/kg/min more accurately identified patients at lower risk for the primary endpoint (NPV: 91.2 and 60.5% for HFmrEF and HFrEF, respectively; both p&lt;0.001). Conclusions CPET is a useful tool in HFmrEF. Both pVO2 and EOV independently predicted the primary endpoint in HFmrEF and HFrEF, contrasting with VE/VCO2, which remained predictive only in latter group. Our findings strengthen the prognostic role of CPET in HF with either reduced or mid-range LVEF. Funding Acknowledgement Type of funding source: None

Abstract 13785: Left Ventricular Contractile Entropy in 99mTc-Sestamibi SPECT is a Novel Prognostic Predictor in Patients with Non-ischemic Dilated Cardiomyopathy

Circulation ◽  
2014 ◽  
Vol 130 (suppl_2) ◽  
Author(s):  
Naoaki Kano ◽  
Takahiro Okumura ◽  
Akinori Sawamura ◽  
Naoki Watanabe ◽  
Hiroaki Mori ◽  
...  
Keyword(s):  
Dilated Cardiomyopathy ◽  
Left Ventricular ◽  
Left Ventricular Ejection ◽  
Sampling Point ◽  
Cardiac Events ◽  
Ventricular Ejection Fraction ◽  
Qrs Duration ◽  
Non Ischemic Dilated Cardiomyopathy ◽  
The Mean ◽  
99Mtc Sestamibi

Background: It has been reported that mechanical dispersion of myocardial contraction is increased in failing myocardium. However little is known about the association between contractile entropy evaluated by myocardial scintigraphy and prognosis in patients with non-ischemic dilated cardiomyopathy (NIDCM). Purpose: We aimed to investigate the prognostic value of contractile entropy in patients with NIDCM. Methods: Forty-seven patients (38 male, 55.1 years) with NIDCM were performed gated 99mTc-sestamibi myocardial perfusion SPECT (GMPS) and endomyocardial biopsy. Entropy was automatically calculated as a result of contractile phase analysis for each myocardial sampling point from GMPS, and it reflects a dispersion of global mechanical contraction. All patients were allocated into two groups based on the median of entropy; HE-group: entropy≥0.61 and LE-group: entropy<0.61. All patients were followed up at the mean of 2.8 years. Results: The mean QRS duration, left ventricular ejection fraction (LVEF) and plasma brain natriuretic peptide (BNP) levels were 114 msec, 35% and 225 pg/mL, respectively. Although there were no significant differences in QRS duration and plasma BNP levels between the two groups, LVEF was lower in the HE-group than in the LE-group (31.1% vs 39.8%, p=0.002). In Kaplan-Meier survival analysis, cardiac event rate was significantly higher in the HE-group (Figure). Cox proportional hazard analysis revealed that the HE-group was a significant determinant of cardiac events (Hazard Ratio: 7.66; 95%CI: 0.070-2.532; p=0.033). The mRNA expression level of sarcoplasmic endoplasmic reticulum Ca2+ ATPase (SERCA2a) in biopsy specimens was significantly lower in the LE-group (p=0.015). Conclusion: Contractile entropy, reflecting an impairment of global left ventricular contraction, might be useful to predict a poor prognosis in patients with NIDCM.

scholarly journals Serum autotaxin level predicts future cardiac events in patients with dilated cardiomyopathy

2021 ◽  
Vol 42 (Supplement_1) ◽  
Author(s):  
T Araki ◽  
T Okumura ◽  
T Mizutani ◽  
Y Kimura ◽  
S Kazama ◽  
...  
Keyword(s):  
Survival Analysis ◽  
Dilated Cardiomyopathy ◽  
Volume Fraction ◽  
Left Ventricular ◽  
Cox Proportional Hazards ◽  
Left Ventricular Ejection ◽  
Myocardial Inflammation ◽  
Prognostic Impact ◽  
Cardiac Events ◽  
Ventricular Ejection Fraction

Abstract Background Autotaxin (ATX) has been reported to promote myocardial inflammation and subsequent cardiac remodeling through lysophosphatidic acid (LPA) production. However, the prognostic impact of ATX has not been clarified in dilated cardiomyopathy (DCM). Purpose We aimed to investigate the prognostic impact of ATX in patients with DCM. Methods We enrolled 104 DCM patients (49.8 years, 76 males). The subjects underwent blood sampling, echocardiography, cardiac catheterization, and endomyocardial biopsy. Gender differences in serum ATX levels have been reported, thus we divided the subjects into two groups using median serum ATX levels for men and women: High-ATX group and Low-ATX group. All patients were followed up by expert cardiologists. The cardiac event was defined as a composite of cardiac death and hospitalization for worsening heart failure. Results Eighty-nine percent of the subjects were classified as New York Heart Association functional class I or II. Female patients had higher serum ATX levels than male patients, with median values of 257.0 ng/mL and 203.5 ng/mL, respectively (Figure A). The average left ventricular ejection fraction and brain natriuretic peptide levels were 30.6% and 122.5 pg/mL. In survival analysis, cumulative event-free probability was significantly lower in High ATX group (p=0.007, Figure B). In Cox proportional hazards analysis, High-ATX was one of the independent predictors of composite cardiac events (Hazards Ratio, 2.575; p=0.043). On the other hand, high sensitive C-reactive protein and collagen volume fraction in myocardial samples were not significant predictors. Conclusion High serum ATX level was associated with poor prognosis in patients with DCM. FUNDunding Acknowledgement Type of funding sources: None. Gender difference in autotaxin levels Survival analysis of cardiac events

scholarly journals Detecting Cardiomyopathies in Pregnancy and the Postpartum Period with an Electrocardiogram-Based Deep Learning Model

2021 ◽  
Author(s):  
Demilade A Adedinsewo ◽  
Patrick W Johnson ◽  
Erika J Douglass ◽  
Itzhak Zachi Attia ◽  
Sabrina D Phillips ◽  
...  
Keyword(s):  
Deep Learning ◽  
Natriuretic Peptides ◽  
Postpartum Period ◽  
Learning Model ◽  
Left Ventricular ◽  
Left Ventricular Ejection ◽  
Clinical Parameters ◽  
Multivariable Model ◽  
Ventricular Ejection Fraction ◽  
Deep Learning Model

Abstract Aims Cardiovascular disease is a major threat to maternal health, with cardiomyopathy being among the most common acquired cardiovascular diseases during pregnancy and the postpartum period. The aim of our study was to evaluate the effectiveness of an electrocardiogram (ECG)-based deep learning model in identifying cardiomyopathy during pregnancy and the postpartum period. Methods and Results We used an ECG-based deep learning model to detect cardiomyopathy in a cohort of women who were pregnant or in the postpartum period seen at Mayo Clinic. Model performance was evaluated using area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity. We compared the diagnostic probabilities of the deep learning model with natriuretic peptides and a multivariable model consisting of demographic and clinical parameters. The study cohort included 1,807 women; 7%, 10%, and 13% had left ventricular ejection fraction (LVEF) of 35% or less, less than 45%, and less than 50%, respectively. The ECG-based deep learning model identified cardiomyopathy with AUCs of 0.92 (LVEF ≤35%), 0.89 (LVEF &lt;45%), and 0.87 (LVEF &lt;50%). For LVEF of 35% or less, AUC was higher in Black (0.95) and Hispanic (0.98) women compared to White (0.91). Natriuretic peptides and the multivariable model had AUCs of 0.85 to 0.86 and 0.72, respectively. Conclusions An ECG-based deep learning model effectively identifies cardiomyopathy during pregnancy and the postpartum period and outperforms natriuretic peptides and traditional clinical parameters with the potential to become a powerful initial screening tool for cardiomyopathy in the obstetric care setting.

Sign in / Sign up

Export Citation Format

Share Document