Artificial intelligence-based screening for cardiomyopathy in an obstetric population: A pilot study

Abstract

Background: Cardiomyopathy is a leading cause of pregnancy-related mortality and the number one cause of death in the late postpartum period. Delay in diagnosis is associated with severe adverse outcomes.

Objective: To evaluate the performance of an artificial intelligence-enhanced electrocardiogram (AI-ECG) and AI-enabled digital stethoscope to detect left ventricular systolic dysfunction in an obstetric population.

Methods: We conducted a single-arm prospective study of pregnant and postpartum women enrolled at 3 sites between October 28, 2021, and October 27, 2022. Study participants completed a standard 12-lead ECG, digital stethoscope ECG and phonocardiogram recordings, and a transthoracic echocardiogram within 24 hours. Diagnostic performance was evaluated using the area under the curve (AUC).

Results: One hundred women were included in the final analysis. The median age was 31 years (Q1: 27, Q3: 34). Thirty-eight percent identified as non-Hispanic White, 32% as non-Hispanic Black, and 21% as Hispanic. Five percent and 6% had left ventricular ejection fraction (LVEF) <45% and <50%, respectively. The AI-ECG model had near-perfect classification performance (AUC: 1.0, 100% sensitivity; 99%-100% specificity) for detection of cardiomyopathy at both LVEF categories. The AI-enabled digital stethoscope had an AUC of 0.98 (95% CI: 0.95, 1.00) and 0.97 (95% CI: 0.93, 1.00), for detection of LVEF <45% and <50%, respectively, with 100% sensitivity and 90% specificity.

Conclusion: We demonstrate an AI-ECG and AI-enabled digital stethoscope were effective for detecting cardiac dysfunction in an obstetric population. Larger studies, including an evaluation of the impact of screening on clinical outcomes, are essential next steps.

Keywords: Cardiomyopathies; ECG; Heart failure; Obstetrics; Postpartum; Pregnancy.

Graphical abstract

Figure 1

Flow diagram for study participants.

Figure 2

Digital stethoscope positions. A: Digital stethoscope connects to mobile app via Bluetooth. B: Fifteen-second single-lead electrocardiogram recordings were obtained at 3 locations across the chest: placed vertically at the left sternal border (V2), angled across the left upper chest (angled), and placed horizontally over the left subclavicular area (subclavicular).

Figure 3

Twelve-lead electrocardiogram (ECG) receiver operating characteristic (ROC) curves and confusion matrix. A, B: ROC curve (A) and confusion matrix (B) for detection of left ventricular ejection fraction (LVEF) <50% using a 12-lead artificial intelligence–enabled ECG (AI-ECG) algorithm among pregnant and postpartum women. C, D: ROC curve (C) and confusion matrix (D) for detection of LVEF <45% using 12-lead AI-ECG algorithm among pregnant and postpartum women. The 1 false-positive case had an ejection fraction of 49%.

Figure 4

Standard 12-lead electrocardiogram (ECG) examples. ECG tracings for 2 patients in our study sample, with clinical interpretations: 1 with left ventricular ejection fraction (LVEF) of 35% (AI-predicted probability 0.796) and the other with LVEF of 66% (AI-predicted probability 0.180). The previously determined AI-predicted probability threshold for a positive flag is 0.256.