Search

Valvular heart disease is an important contributor to cardiovascular morbidity and mortality and remains underdiagnosed. Deep learning analysis of electrocardiography (ECG) may be useful in detecting aortic stenosis (AS), aortic regurgitation (AR), and mitral regurgitation (MR).This study aimed to develop ECG deep learning algorithms to identify moderate or severe AS, AR, and MR alone and in combination.A total of 77,163 patients undergoing ECG within 1 year before echocardiography from 2005-2021 were identified and split into train (n = 43,165), validation (n = 12,950), and test sets (n = 21,048; 7.8% with any of AS, AR, or MR). Model performance was assessed using area under the receiver-operating characteristic (AU-ROC) and precision-recall curves. Outside validation was conducted on an independent data set. Test accuracy was modeled using different disease prevalence levels to simulate screening efficacy using the deep learning model.The deep learning algorithm model accuracy was as follows: AS (AU-ROC: 0.88), AR (AU-ROC: 0.77), MR (AU-ROC: 0.83), and any of AS, AR, or MR (AU-ROC: 0.84; sensitivity 78%, specificity 73%) with similar accuracy in external validation. In screening program modeling, test characteristics were dependent on underlying prevalence and selected sensitivity levels. At a prevalence of 7.8%, the positive and negative predictive values were 20% and 97.6%, respectively.Deep learning analysis of the ECG can accurately detect AS, AR, and MR in this multicenter cohort and may serve as the basis for the development of a valvular heart disease screening program.

Aims Determining the aetiology of left ventricular hypertrophy (LVH) can be challenging due to the similarity in clinical presentation and cardiac morphological features of diverse causes of disease. In particular, distinguishing individuals with hypertrophic cardiomyopathy (HCM) from the much larger set of individuals with manifest or occult hypertension (HTN) is of major importance for family screening and the prevention of sudden death. We hypothesized that an artificial intelligence method based joint interpretation of 12-lead electrocardiograms and echocardiogram videos could augment physician interpretation. Methods and results We chose not to train on proximate data labels such as physician over-reads of ECGs or echocardiograms but instead took advantage of electronic health record derived clinical blood pressure measurements and diagnostic consensus (often including molecular testing) among physicians in an HCM centre of excellence. Using more than 18 000 combined instances of electrocardiograms and echocardiograms from 2728 patients, we developed LVH-fusion. On held-out test data, LVH-fusion achieved an F1-score of 0.71 in predicting HCM, and 0.96 in predicting HTN. In head-to-head comparison with human readers LVH-fusion had higher sensitivity and specificity rates than its human counterparts. Finally, we use explainability techniques to investigate local and global features that positively and negatively impact LVH-fusion prediction estimates providing confirmation from unsupervised analysis the diagnostic power of lateral T-wave inversion on the ECG and proximal septal hypertrophy on the echocardiogram for HCM. Conclusion These results show that deep learning can provide effective physician augmentation in the face of a common diagnostic dilemma with far reaching implications for the prevention of sudden cardiac death.

The global burden of diabetes is rapidly increasing, from 451 million people in 2019 to 693 million by 20451. The insidious onset of type 2 diabetes delays diagnosis and increases morbidity2. Given the multifactorial vascular effects of diabetes, we hypothesized that smartphone-based photoplethysmography could provide a widely accessible digital biomarker for diabetes. Here we developed a deep neural network (DNN) to detect prevalent diabetes using smartphone-based photoplethysmography from an initial cohort of 53,870 individuals (the ‘primary cohort’), which we then validated in a separate cohort of 7,806 individuals (the ‘contemporary cohort’) and a cohort of 181 prospectively enrolled individuals from three clinics (the ‘clinic cohort’). The DNN achieved an area under the curve for prevalent diabetes of 0.766 in the primary cohort (95% confidence interval: 0.750–0.782; sensitivity 75%, specificity 65%) and 0.740 in the contemporary cohort (95% confidence interval: 0.723–0.758; sensitivity 81%, specificity 54%). When the output of the DNN, called the DNN score, was included in a regression analysis alongside age, gender, race/ethnicity and body mass index, the area under the curve was 0.830 and the DNN score remained independently predictive of diabetes. The performance of the DNN in the clinic cohort was similar to that in other validation datasets. There was a significant and positive association between the continuous DNN score and hemoglobin A1c (P ≤ 0.001) among those with hemoglobin A1c data. These findings demonstrate that smartphone-based photoplethysmography provides a readily attainable, non-invasive digital biomarker of prevalent diabetes. A deep neural network applied to smartphone-based vascular imaging can detect diabetes, opening new possibilities for non-invasive diagnosis.

IMPORTANCE: Millions of clinicians rely daily on automated preliminary electrocardiogram (ECG) interpretation. Critical comparisons of machine learning–based automated analysis against clinically accepted standards of care are lacking. OBJECTIVE: To use readily available 12-lead ECG data to train and apply an explainability technique to a convolutional neural network (CNN) that achieves high performance against clinical standards of care. DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study was conducted using data from January 1, 2003, to December 31, 2018. Data were obtained in a commonly available 12-lead ECG format from a single-center tertiary care institution. All patients aged 18 years or older who received ECGs at the University of California, San Francisco, were included, yielding a total of 365 009 patients. Data were analyzed from January 1, 2019, to March 2, 2021. EXPOSURES: A CNN was trained to predict the presence of 38 diagnostic classes in 5 categories from 12-lead ECG data. A CNN explainability technique called LIME (Linear Interpretable Model-Agnostic Explanations) was used to visualize ECG segments contributing to CNN diagnoses. MAIN OUTCOMES AND MEASURES: Area under the receiver operating characteristic curve (AUC), sensitivity, and specificity were calculated for the CNN in the holdout test data set against cardiologist clinical diagnoses. For a second validation, 3 electrophysiologists provided consensus committee diagnoses against which the CNN, cardiologist clinical diagnosis, and MUSE (GE Healthcare) automated analysis performance was compared using the F1 score; AUC, sensitivity, and specificity were also calculated for the CNN against the consensus committee. RESULTS: A total of 992 748 ECGs from 365 009 adult patients (mean [SD] age, 56.2 [17.6] years; 183 600 women [50.3%]; and 175 277 White patients [48.0%]) were included in the analysis. In 91 440 test data set ECGs, the CNN demonstrated an AUC of at least 0.960 for 32 of 38 classes (84.2%). Against the consensus committee diagnoses, the CNN had higher frequency-weighted mean F1 scores than both cardiologists and MUSE in all 5 categories (CNN frequency-weighted F1 score for rhythm, 0.812; conduction, 0.729; chamber diagnosis, 0.598; infarct, 0.674; and other diagnosis, 0.875). For 32 of 38 classes (84.2%), the CNN had AUCs of at least 0.910 and demonstrated comparable F1 scores and higher sensitivity than cardiologists, except for atrial fibrillation (CNN F1 score, 0.847 vs cardiologist F1 score, 0.881), junctional rhythm (0.526 vs 0.727), premature ventricular complex (0.786 vs 0.800), and Wolff-Parkinson-White (0.800 vs 0.842). Compared with MUSE, the CNN had higher F1 scores for all classes except supraventricular tachycardia (CNN F1 score, 0.696 vs MUSE F1 score, 0.714). The LIME technique highlighted physiologically relevant ECG segments. CONCLUSIONS AND RELEVANCE: The results of this cross-sectional study suggest that readily available ECG data can be used to train a CNN algorithm to achieve comparable performance to clinical cardiologists and exceed the performance of MUSE automated analysis for most diagnoses, with some exceptions. The LIME explainability technique applied to CNNs highlights physiologically relevant ECG segments that contribute to the CNN’s diagnoses.

Laboratory blood testing is routinely used to assay biomarkers to provide information on physiologic state beyond what clinicians can evaluate from interpreting medical imaging. We hypothesized that deep learning interpretation of echocardiogram videos can provide additional value in understanding disease states and can predict common biomarkers results. Using 70,066 echocardiograms and associated biomarker results from 39,460 patients, we developed EchoNet-Labs, a video-based deep learning algorithm to predict anemia, elevated B-type natriuretic peptide (BNP), troponin I, and blood urea nitrogen (BUN), and abnormal levels in ten additional lab tests. On held-out test data across different healthcare systems, EchoNet-Labs achieved an area under the curve (AUC) of 0.80 in predicting anemia, 0.82 in predicting elevated BNP, 0.75 in predicting elevated troponin I, and 0.69 in predicting elevated BUN. We further demonstrate the utility of the model in predicting abnormalities in 10 additional lab tests. We investigate the features necessary for EchoNet-Labs to make successful predictions and identify potential prediction mechanisms for each biomarker using well-known and novel explainability techniques. These results show that deep learning applied to diagnostic imaging can provide additional clinical value and identify phenotypic information beyond current imaging interpretation methods.

Deep Reinforcement Learning has been applied in a number of fields to directly optimize non-differentiable reward functions, including in sequence to sequence settings using Self Critical Sequence Training (SCST). Previously, SCST has primarily been applied to bring conditional language models closer to the distribution of their training set, as in traditional neural machine translation and abstractive summarization. We frame the generation of search engine text ads as a sequence to sequence problem, and consider two related goals: to generate ads similar to those a human would write, and to generate ads with high click-through rates. We jointly train a model to minimize cross-entropy on an existing corpus of Landing Page/Text Ad pairs using typical sequence to sequence training techniques while also optimizing the expected click-through rate (CTR) as predicted by an existing oracle model using SCST. Through joint training we achieve a 6.7% increase in expected CTR without a meaningful drop in ROUGE score. Human experiments demonstrate that SCST training produces significantly more attractive ads without reducing grammatical quality.

The global burden of diabetes is rapidly increasing, from 451 million people in 2019 to 693 million by 2045

The combinatorial effect of genetic variants is often assumed to be additive. Although genetic variation can clearly interact non-additively, methods to uncover epistatic relationships remain in their infancy. We develop low-signal signed iterative random forests to elucidate the complex genetic architecture of cardiac hypertrophy. We derive deep learning-based estimates of left ventricular mass from the cardiac MRI scans of 29,661 individuals enrolled in the UK Biobank. We report epistatic genetic variation including variants close to CCDC141 , IGF1R , TTN , and TNKS. Several loci where variants were deemed insignificant in univariate genome-wide association analyses are identified. Functional genomic and integrative enrichment analyses reveal a complex gene regulatory network in which genes mapped from these loci share biological processes and myogenic regulatory factors. Through a network analysis of transcriptomic data from 313 explanted human hearts, we found strong gene co-expression correlations between these statistical epistasis contributors in healthy hearts and a significant connectivity decrease in failing hearts. We assess causality of epistatic effects via RNA silencing of gene-gene interactions in human induced pluripotent stem cell-derived cardiomyocytes. Finally, single-cell morphology analysis using a novel high-throughput microfluidic system shows that cardiomyocyte hypertrophy is non-additively modifiable by specific pairwise interactions between CCDC141 and both TTN and IGF1R . Our results expand the scope of genetic regulation of cardiac structure to epistasis.