High-Throughput Precision Phenotyping of Left Ventricular Hypertrophy With Cardiovascular Deep Learning

Abstract

Importance: Early detection and characterization of increased left ventricular (LV) wall thickness can markedly impact patient care but is limited by under-recognition of hypertrophy, measurement error and variability, and difficulty differentiating causes of increased wall thickness, such as hypertrophy, cardiomyopathy, and cardiac amyloidosis.

Objective: To assess the accuracy of a deep learning workflow in quantifying ventricular hypertrophy and predicting the cause of increased LV wall thickness.

Design, settings, and participants: This cohort study included physician-curated cohorts from the Stanford Amyloid Center and Cedars-Sinai Medical Center (CSMC) Advanced Heart Disease Clinic for cardiac amyloidosis and the Stanford Center for Inherited Cardiovascular Disease and the CSMC Hypertrophic Cardiomyopathy Clinic for hypertrophic cardiomyopathy from January 1, 2008, to December 31, 2020. The deep learning algorithm was trained and tested on retrospectively obtained independent echocardiogram videos from Stanford Healthcare, CSMC, and the Unity Imaging Collaborative.

Main outcomes and measures: The main outcome was the accuracy of the deep learning algorithm in measuring left ventricular dimensions and identifying patients with increased LV wall thickness diagnosed with hypertrophic cardiomyopathy and cardiac amyloidosis.

Results: The study included 23 745 patients: 12 001 from Stanford Health Care (6509 [54.2%] female; mean [SD] age, 61.6 [17.4] years) and 1309 from CSMC (808 [61.7%] female; mean [SD] age, 62.8 [17.2] years) with parasternal long-axis videos and 8084 from Stanford Health Care (4201 [54.0%] female; mean [SD] age, 69.1 [16.8] years) and 2351 from CSMS (6509 [54.2%] female; mean [SD] age, 69.6 [14.7] years) with apical 4-chamber videos. The deep learning algorithm accurately measured intraventricular wall thickness (mean absolute error [MAE], 1.2 mm; 95% CI, 1.1-1.3 mm), LV diameter (MAE, 2.4 mm; 95% CI, 2.2-2.6 mm), and posterior wall thickness (MAE, 1.4 mm; 95% CI, 1.2-1.5 mm) and classified cardiac amyloidosis (area under the curve [AUC], 0.83) and hypertrophic cardiomyopathy (AUC, 0.98) separately from other causes of LV hypertrophy. In external data sets from independent domestic and international health care systems, the deep learning algorithm accurately quantified ventricular parameters (domestic: R2, 0.96; international: R2, 0.90). For the domestic data set, the MAE was 1.7 mm (95% CI, 1.6-1.8 mm) for intraventricular septum thickness, 3.8 mm (95% CI, 3.5-4.0 mm) for LV internal dimension, and 1.8 mm (95% CI, 1.7-2.0 mm) for LV posterior wall thickness. For the international data set, the MAE was 1.7 mm (95% CI, 1.5-2.0 mm) for intraventricular septum thickness, 2.9 mm (95% CI, 2.4-3.3 mm) for LV internal dimension, and 2.3 mm (95% CI, 1.9-2.7 mm) for LV posterior wall thickness. The deep learning algorithm accurately detected cardiac amyloidosis (AUC, 0.79) and hypertrophic cardiomyopathy (AUC, 0.89) in the domestic external validation site.

Conclusions and relevance: In this cohort study, the deep learning model accurately identified subtle changes in LV wall geometric measurements and the causes of hypertrophy. Unlike with human experts, the deep learning workflow is fully automated, allowing for reproducible, precise measurements, and may provide a foundation for precision diagnosis of cardiac hypertrophy.

Figure 1.. Deep Learning Workflow Combining Evaluation of Ventricular Dimensions and Suspicion for Underdiagnosed Diseases

A, The deep learning algorithm used parasternal long-axis echocardiogram video as input to derive key points and estimate ventricular dimensions. After identifying patients with left ventricular hypertrophy (LVH), the deep learning workflow used a video-based architecture to distinguish common causes of LVH. B, Correlation of human annotations vs model predictions for ventricular dimensions in data sets from Stanford Health Care (SHC; n = 1200), Cedars-Sinai Medical Center (CSMC; n = 1309), and Unity Imaging Collaborative (n = 1791). C, Model variation on the 3 data sets vs human clinical annotation variation. Middle lines represent means; upper and lower bounds of the boxes, 25th and 75th percentiles; and points, values greater than 1.5 times the IQR. D, Receiver operating characteristic curves for diagnosis of amyloidosis in the SHC validation (n = 813) and test (n = 812) sets. AS indicates aortic stenosis; AUC, area under the curve; CA, cardiac amyloid; HCM, hypertrophic cardiomyopathy; HTN, hypertension; IVS, intraventricular septum; LVID, LV internal dimension; LVPW, LV posterior wall.

Figure 2.. Beat-to-Beat Evaluation of Ventricular Dimensions

A, Model prediction of key points on an individual frame of parasternal long-axis video. B, Frame-by-frame prediction of wall thickness and ventricular dimension and automated detection of systole and diastole allowing for beat-to-beat prediction of ventricular hypertrophy. C, Waterfall plot of individual video variation in beat-to-beat evaluation of ventricular hypertrophy (n = 2320) in the internal test data set. Each video is represented by multiple points along a line representing the measurement of each beat and a line signifying the range of predictions. IVS indicates intraventricular septum; IVSd, intraventricular septum (diastole); LVID, left ventricular internal dimension; LVIDd, left ventricular internal dimension (diastole); LVPW, left ventricular posterior wall.

Figure 3.. Performance of Disease Cause Classification in the Independent External Validation Cohort

A, Receiver operating characteristic curves for detection of cardiac amyloidosis and hypertrophic cardiomyopathy in the Cedars-Sinai Medical Center independent external test set (n = 2351). B, Precision-recall curves for detection of amyloidosis and hypertrophic cardiomyopathy. C, Representative images for selected cases and controls for each cause. AP indicates average precision; AUC, area under the curve.