Point-of-care screening for heart failure with reduced ejection fraction using artificial intelligence during ECG-enabled stethoscope examination in London, UK: a prospective, observational, multicentre study

Abstract

Summary Background Most patients who have heart failure with a reduced ejection fraction, when left ventricular ejection fraction (LVEF) is 40% or lower, are diagnosed in hospital. This is despite previous presentations to primary care with symptoms. We aimed to test an artificial intelligence (AI) algorithm applied to a single-lead ECG, recorded during ECG-enabled stethoscope examination, to validate a potential point-of-care screening tool for LVEF of 40% or lower. Methods We conducted an observational, prospective, multicentre study of a convolutional neural network (known as AI-ECG) that was previously validated for the detection of reduced LVEF using 12-lead ECG as input. We used AI-ECG retrained to interpret single-lead ECG input alone. Patients (aged ≥18 years) attending for transthoracic echocardiogram in London (UK) were recruited. All participants had 15 s of supine, single-lead ECG recorded at the four standard anatomical positions for cardiac auscultation, plus one handheld position, using an ECG-enabled stethoscope. Transthoracic echocardiogram-derived percentage LVEF was used as ground truth. The primary outcome was performance of AI-ECG at classifying reduced LVEF (LVEF ≤40%), measured using metrics including the area under the receiver operating characteristic curve (AUROC), sensitivity, and specificity, with two-sided 95% CIs. The primary outcome was reported for each position individually and with an optimal combination of AI-ECG outputs (interval range 0–1) from two positions using a rule-based approach and several classification models. This study is registered with ClinicalTrials.gov , NCT04601415 . Findings Between Feb 6 and May 27, 2021, we recruited 1050 patients (mean age 62 years [SD 17·4], 535 [51%] male, 432 [41%] non-White). 945 (90%) had an ejection fraction of at least 40%, and 105 (10%) had an ejection fraction of 40% or lower. Across all positions, ECGs were most frequently of adequate quality for AI-ECG interpretation at the pulmonary position (979 [93·3%] of 1050). Quality was lowest for the aortic position (846 [80·6%]). AI-ECG performed best at the pulmonary valve position (p=0·02), with an AUROC of 0·85 (95% CI 0·81–0·89), sensitivity of 84·8% (76·2–91·3), and specificity of 69·5% (66·4–72·6). Diagnostic odds ratios did not differ by age, sex, or non-White ethnicity. Taking the optimal combination of two positions (pulmonary and handheld positions), the rule-based approach resulted in an AUROC of 0·85 (0·81–0·89), sensitivity of 82·7% (72·7–90·2), and specificity of 79·9% (77·0–82·6). Using AI-ECG outputs from these two positions, a weighted logistic regression with l2 regularisation resulted in an AUROC of 0·91 (0·88–0·95), sensitivity of 91·9% (78·1–98·3), and specificity of 80·2% (75·5–84·3). Interpretation A deep learning system applied to single-lead ECGs acquired during a routine examination with an ECG...

Artificial intelligence-based opportunistic screening for the detection of arterial hypertension through ECG signals

“…The applications of artificial intelligence on ECG are evolving rapidly with tremendous future implications on cardiovascular medicine [10,11,[33][34][35] . ECG signals and patterns largely unrecognizable to human eye interpretation can be detected by machine learning algorithms making the ECG a more powerful, non-invasive clinical tool.…”

Artificial intelligence in cardiology: The past, present and future

“…This was evident in the AI-ECG study wherein an AI based algorithm applied to a single-lead ECG recorded during ECG-enabled stethoscope examination reported good diagnostic accuracy for detection of LVEF <40%. 26 Apart from identification of at-risk patients, AI based system can also play a part in hospitalization prevention. Traditional statistical models for readmission prediction have marked limitations and ML algorithms tend to better identify those at risk for HF readmission.…”

“… 23 Smartwatch identification of AF Participants receiving notification of an irregular pulse: 34% had AF on subsequent ECG patch readings and 84% of notifications were concordant with AF Heart Failure Artificial Intelligence-Clinical Decision Support System for HF diagnosis Choi et al. 25 Evaluate the diagnostic accuracy of an AI-CDSS for heart failure AI-CDSS - high diagnostic accuracy for HF: concordance rate between AI-CDSS and heart failure specialists - 98% AI algorithm applied to a single-lead ECG recorded during ECG-enabled stethoscope examination Bachtiger P et al, 26 Validate a potential point-of-care screening tool (ECG-enabled stethoscope) for LVEF of 40% or lower AI-ECG-enabled stethoscope can detect LVEF of ≤40% with good accuracy- AUROC: 0·91, sensitivity: 91·9% and specificity: 80·2% Predicting response to therapy (CRT) AI-ECG, Predicting CRT outcomes Predicted death or HF hospitalization within 12 months - AUC of 0.74 Novel characterization of HF phenogroups Kalscheur et al. 27 Predicting CRT outcomes Predicted echocardiographic CRT response better than current guidelines (AUC: 0.70 vs. 0.65) - greater discrimination of long-term survival (c-index: 0.61 vs. 0.56) Feeny et al.…”

Towards an artificial intelligence-augmented, ECG-enabled physical exam

“…1,2 In The Lancet Digital Health, Patrik Bachtiger and colleagues tie into that clinical workflow by using a novel, ECGenabled stethoscope in a realistic clinical encounter to identify patients with reduced left ventricular (LV) function. 3 Studies using AI-augmented ECGs to idefntify patients with asymptomatic LV dysfunction marked by a reduced ejection fraction (LVEF) have a clinical purpose. [4][5][6] Although these previous studies are novel in providing prospective, real-world data, Bachtiger and colleagues take this one step further by integrating the algorithm as part of the auscultatory physical exam.…”