What does this research mean for the field?

An artificial intelligence algorithm predicts left ventricular hypertrophy (LVH) unrelated to abnormal loading conditions with a sensitivity of 73% and specificity of 64% in a cohort of 890 patients. Novelty: ClaimNovelty.NOVEL_FINDING. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

The aim is to develop and validate an AI algorithm to detect unexplained left ventricular hypertrophy (LVH) using electrocardiogram (ECG) data.

February 8, 2026European Heart Journal

Pathophysiological stratification of left ventricular hypertrophy using artificial intelligence-based electrocardiogram analysis

View Full Paper

Ask AI

Bookmark

Key result

ECG-based AI detects LVH unrelated to abnormal loading with ~94% NPV.

Why the study?

In 20-30% of cases, LVH remains unexplained and can result from HCM, CA, or AFD, making early detection crucial for risk stratification, genetic counseling, and timely treatment.

Does an AI-based ECG algorithm accurately detect left ventricular hypertrophy unrelated to abnormal loading conditions in patients with TTE-confirmed LVH?

Population

Patients with TTE-confirmed LVH (1,899 training set, 890 test set) at University Medical Center Utrecht

Comparison

AI ECG algorithm prediction vs clinical composite of CA, AFD, or HCM

Design

Retrospective cohort study

Authors

DADino AhmetagićUniversity Medical Center Utrecht BAB K O ArendsUniversity Medical Center Utrecht RLR R Van De LeurUniversity Medical Center Utrecht

Discussion

Member takes

Implication

May aid ruling out unexplained LVH on ECG; leaves open prospective validation before clinical adoption.

Key Points

The aim is to develop and validate an AI algorithm to detect unexplained left ventricular hypertrophy (LVH) using electrocardiogram (ECG) data.
Retrospective analysis of ECG-transthoracic echocardiogram pairs from patients with confirmed LVH.
Patients were categorized based on left ventricular wall thickness and absence of hypertension-mediated damage.
A convolutional neural network was trained on a subset of ECGs to identify primary endpoint causes.
A test set was used to validate algorithm performance.
89 out of 890 patients (10.0%) had LVH unrelated to abnormal loading conditions.
Algorithm achieved an area under the curve of 0.74, with sensitivity of 0.73 and specificity of 0.64.
Positive predictive value was 0.23, while negative predictive value was 0.94.

Structured PICO

Does an AI-based ECG algorithm accurately detect left ventricular hypertrophy unrelated to abnormal loading conditions in patients with TTE-confirmed LVH?

Population

2,789 patients with TTE-confirmed left ventricular hypertrophy (1,899 in training set, 890 in test set), test set median age 66, 59.9% male, at University Medical Center Utrecht.

Intervention

Artificial intelligence (AI) algorithm (convolutional neural network) using electrocardiogram (ECG) data

Outcome

Composite of cardiac amyloidosis (CA), Anderson-Fabry disease (AFD), or hypertrophic cardiomyopathy (HCM)composite

An AI-based ECG algorithm can identify left ventricular hypertrophy caused by hypertrophic cardiomyopathy, cardiac amyloidosis, or Anderson-Fabry disease with moderate accuracy and high negative predictive value.

Cite This Study

Ahmetagić et al. (2025) studied this question. AI algorithm using ECG detected LVH unrelated to abnormal loading with AUC 0.74, sensitivity 0.73, specificity 0.64, and NPV 0.94 in 890 patients.

synapsesocial.com/papers/698828620fc35cd7a8847d85 https://doi.org/10.1093/eurheartj/ehaf784.2555

Topics

Hypertrophic cardiomyopathy Artificial intelligence in cardiology

View Full Paper

Ask AI

Bookmark

Pathophysiological stratification of left ventricular hypertrophy using artificial intelligence-based electrocardiogram analysis

Discussion

Member takes

Implication

Key Points

Structured PICO

Cite This Study

Topics

Pathophysiological stratification of left ventricular hypertrophy using artificial intelligence-based electrocardiogram analysis

Discussion

Member takes

Implication

Key Points

Structured PICO

Cite This Study

Also Consider

Topics

Also Consider