Automated interpretation of systolic and diastolic function on the echocardiogram: a multicohort study

Abstract

Background Echocardiography is the diagnostic modality for assessing cardiac systolic and diastolic function to diagnose and manage heart failure. However, manual interpretation of echocardiograms can be time consuming and subject to human error. Therefore, we developed a fully automated deep learning workflow to classify, segment, and annotate two-dimensional (2D) videos and Doppler modalities in echocardiograms. MethodsWe developed the workflow using a training dataset of 1145 echocardiograms and an internal test set of 406 echocardiograms from the prospective heart failure research platform (Asian Network for Translational Research and Cardiovascular Trials; ATTRaCT) in Asia, with previous manual tracings by expert sonographers. We validated the workflow against manual measurements in a curated dataset from Canada (Alberta Heart Failure Etiology and Analysis Research Team; HEART; n=1029 echocardiograms), a real-world dataset from Taiwan (n=31 241), the USbased EchoNet-Dynamic dataset (n=10 030), and in an independent prospective assessment of the Asian (ATTRaCT) and Canadian (Alberta HEART) datasets (n=142) with repeated independent measurements by two expert sonographers. Findings In the ATTRaCT test set, the automated workflow classified 2D videos and Doppler modalities with accuracies (number of correct predictions divided by the total number of predictions) ranging from 0•91 to 0•99. Segmentations of the left ventricle and left atrium were accurate, with a mean Dice similarity coefficient greater than 93% for all. In the external datasets (n=1029 to 10 030 echocardiograms used as input), automated measurements showed good agreement with locally measured values, with a mean absolute error range of 9-25 mL for left ventricular volumes, 6-10% for left ventricular ejection fraction (LVEF), and 1•8-2•2 for the ratio of the mitral inflow E wave to the tissue Doppler e' wave (E/e' ratio); and reliably classified systolic dysfunction (LVEF <40%, area under the receiver operating characteristic curve [AUC] range 0•90-0•92) and diastolic dysfunction (E/e' ratio ≥13, AUC range 0•91-0•91), with narrow 95% CIs for AUC values. Independent prospective evaluation confirmed less variance of automated compared with human expert measurements, with all individual equivalence coefficients being less than 0 for all measurements. Interpretation Deep learning algorithms can automatically annotate 2D videos and Doppler modalities with similar accuracy to manual measurements by expert sonographers. Use of an automated workflow might accelerate access, improve quality, and reduce costs in diagnosing and managing heart failure globally.

Effectiveness of Ultrasound Cardiovascular Images in Teaching Anatomy: A Pilot Study of an Eight-Hour Training Exposure

“…In the near future, Artificial Intelligence (AI) is expected to assist novices in the interpretation of US images, with companies developing deep learning algorithms that can make measurements on US images with less variability than manual measurements performed by expert sonographers [ 30 ]. The US examination of the heart is time-consuming, operator dependent and subject to errors [ 31 ], but it is a low-cost and non-invasive examination [ 32 ].…”

Artificial Intelligence-Enhanced Echocardiography for Systolic Function Assessment

“…The model successfully classified 23 views, segmented cardiac structures, assessed LVEF, and diagnosed three diseases (hypertrophic cardiomyopathy, cardiac amyloid, and pulmonary arterial hypertension). In 2022, Tromp J. et al [ 30 ] developed a fully automated AI workflow to classify, segment, and interpret two-dimensional and Doppler modalities based on international and interracial datasets. The results showed that the algorithms successfully assessed LVEF with the area under the receiver operating characteristic curve (AUC) of 0.90–0.92.…”

Clinical implications of the universal definition for the prevention and treatment of heart failure

“…6 Deep learning algorithms have shown potential in annotating 2D videos and Doppler modalities with similar accuracy and lower variability in automated compared to manual measurements. 63 Automated measurements of LVEF could feature in future iterations of the universal definition as it appears to help improve diagnostic accuracy and overcome the variability issues ladened in manual measurements.…”

“…Other imaging modalities such as global longitudinal strain look promising in better characterizing structural and functional abnormalities that are critical in the development of HF, 62 but substantial variation (~5%–7%) between modalities still exists 6 . Deep learning algorithms have shown potential in annotating 2D videos and Doppler modalities with similar accuracy and lower variability in automated compared to manual measurements 63 . Automated measurements of LVEF could feature in future iterations of the universal definition as it appears to help improve diagnostic accuracy and overcome the variability issues ladened in manual measurements.…”