What is the clinical evidence from this study?

Study design: Observational. Population: Reduced left ventricular ejection fraction (n=3580). Intervention: AI analysis of continuous-wave Doppler spectra. Primary outcome: Detection of LVEF ≤40% (AUC 0.906).

What does this research mean for the field?

An artificial intelligence model analyzing continuous-wave Doppler spectra from the aortic valve can accurately detect reduced left ventricular ejection fraction (≤40%) without the need for two-dimensional imaging or ECG-gated volumetric analysis. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

To develop an artificial intelligence approach analyzing aortic valve continuous-wave Doppler spectra to identify reduced left ventricular ejection fraction (≤40%) without requiring 2D imaging or ECG gating.

August 23, 2026European Heart Journal - Digital HealthOpen Access

Artificial Intelligence Analysis of Continuous Wave Doppler Spectra to Detect Reduced Left Ventricular Ejection Fraction

View Full Paper

Ask AI

Bookmark

Why the study?

Current AI solutions for LVEF assessment rely on 2D imaging and ECG signals, limiting utility when data quality is poor.

Does artificial intelligence analysis of continuous-wave Doppler spectra accurately detect reduced left ventricular ejection fraction (≤40%) in patients undergoing echocardiography?

Population

3,580 patients (3,988 examinations, 4,231 aortic CW Doppler recordings)

Comparison

AI analysis of aortic CW Doppler spectra to detect LVEF ≤40%

Design

Retrospective study

Key result

An AI model analyzing continuous-wave Doppler spectra detected left ventricular ejection fraction ≤40% with 85.2% accuracy, 79.0% sensitivity, 86.7% specificity, and an AUC of 0.906.

Authors

EKEdyta Kaczmarska-Dyrda KSKarol Sadowski DWDamian Waląg

Discussion

Member takes

Overview

May enable LVEF assessment from limited Doppler data; leaves open prospective validation before clinical use.

Key Points

To develop an artificial intelligence approach analyzing aortic valve continuous-wave Doppler spectra to identify reduced left ventricular ejection fraction (≤40%) without requiring 2D imaging or ECG gating.
Retrospectively analyzed 4,231 aortic continuous-wave Doppler recordings from 3,988 examinations across 3,580 patients, generating 13,359 single-peak spectral images.
Trained a CoAtNet-2 neural network on patient-level cohorts to classify single-peak Doppler images and averaged the output probabilities per examination.
Reduced ejection fraction (≤40%) occurred in 20.8% of examinations, associating with lower peak aortic valve velocity and higher mean pixel intensity.
In the held-out test cohort of 782 examinations, the model achieved an AUC of 0.906, 85.2% accuracy, 79.0% sensitivity, 86.7% specificity, 94.3% NPV, and 59.9% PPV.

Study Design

Type

Observational (n=3,580)

Structured PICO

Does artificial intelligence analysis of continuous-wave Doppler spectra accurately detect reduced left ventricular ejection fraction (≤40%) in patients undergoing echocardiography?

Population

3,580 patients (3,988 examinations) evaluated retrospectively to develop and test an AI model for detecting reduced left ventricular ejection fraction from continuous-wave Doppler spectra.

Exposure

Artificial intelligence (CoAtNet-2 neural network) analysis of continuous-wave (CW) Doppler spectra from the aortic valve

Outcome

Detection of reduced left ventricular ejection fraction (LVEF ≤40%)surrogate

Main Result

Effect estimate: AUC 0.906

An AI model analyzing continuous-wave Doppler spectra from the aortic valve can accurately detect reduced LVEF (≤40%) without requiring 2D imaging or ECG gating.

Cite This Study

Kaczmarska-Dyrda et al. (2026) conducted an observational in Reduced left ventricular ejection fraction (n=3,580). AI analysis of continuous-wave Doppler spectra was evaluated on Detection of LVEF ≤40% (AUC 0.906). An AI model analyzing continuous-wave Doppler spectra detected left ventricular ejection fraction ≤40% with 85.2% accuracy, 79.0% sensitivity, 86.7% specificity, and an AUC of 0.906.

synapsesocial.com/papers/6a8aae407677a3411444730b https://doi.org/10.1093/ehjdh/ztag137

View Full Paper

Ask AI

Bookmark

August 23, 2026European Heart Journal - Digital HealthOpen Access

Artificial Intelligence Analysis of Continuous Wave Doppler Spectra to Detect Reduced Left Ventricular Ejection Fraction

View Full Paper

Ask AI

Bookmark

Why the study?

Current AI solutions for LVEF assessment rely on 2D imaging and ECG signals, limiting utility when data quality is poor.

Does artificial intelligence analysis of continuous-wave Doppler spectra accurately detect reduced left ventricular ejection fraction (≤40%) in patients undergoing echocardiography?

Population

3,580 patients (3,988 examinations, 4,231 aortic CW Doppler recordings)

Comparison

AI analysis of aortic CW Doppler spectra to detect LVEF ≤40%

Design

Retrospective study

Key result

An AI model analyzing continuous-wave Doppler spectra detected left ventricular ejection fraction ≤40% with 85.2% accuracy, 79.0% sensitivity, 86.7% specificity, and an AUC of 0.906.

Authors

EKEdyta Kaczmarska-Dyrda KSKarol Sadowski DWDamian Waląg

Discussion

Member takes

Overview

May enable LVEF assessment from limited Doppler data; leaves open prospective validation before clinical use.

Key Points

To develop an artificial intelligence approach analyzing aortic valve continuous-wave Doppler spectra to identify reduced left ventricular ejection fraction (≤40%) without requiring 2D imaging or ECG gating.
Retrospectively analyzed 4,231 aortic continuous-wave Doppler recordings from 3,988 examinations across 3,580 patients, generating 13,359 single-peak spectral images.
Trained a CoAtNet-2 neural network on patient-level cohorts to classify single-peak Doppler images and averaged the output probabilities per examination.
Reduced ejection fraction (≤40%) occurred in 20.8% of examinations, associating with lower peak aortic valve velocity and higher mean pixel intensity.
In the held-out test cohort of 782 examinations, the model achieved an AUC of 0.906, 85.2% accuracy, 79.0% sensitivity, 86.7% specificity, 94.3% NPV, and 59.9% PPV.

Study Design

Type

Observational (n=3,580)

Structured PICO

Does artificial intelligence analysis of continuous-wave Doppler spectra accurately detect reduced left ventricular ejection fraction (≤40%) in patients undergoing echocardiography?

Population

3,580 patients (3,988 examinations) evaluated retrospectively to develop and test an AI model for detecting reduced left ventricular ejection fraction from continuous-wave Doppler spectra.

Exposure

Artificial intelligence (CoAtNet-2 neural network) analysis of continuous-wave (CW) Doppler spectra from the aortic valve

Outcome

Detection of reduced left ventricular ejection fraction (LVEF ≤40%)surrogate

Main Result

Effect estimate: AUC 0.906

An AI model analyzing continuous-wave Doppler spectra from the aortic valve can accurately detect reduced LVEF (≤40%) without requiring 2D imaging or ECG gating.

Cite This Study

synapsesocial.com/papers/6a8aae407677a3411444730b https://doi.org/10.1093/ehjdh/ztag137

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark