Aortic Stenosis, Low-gradient Aortic Stenosis
Conditions
Keywords
low-gradient aortic stenosis, Echocardiography, Artificial intelligence, Cardiac imaging, Machine learning
Brief summary
The purpose of this study is to evaluate the effectiveness of an artificial intelligence (AI) model developed by the investigators for identifying severe low-gradient aortic valve stenosis. Accurate assessment of stenosis severity is crucial for proper qualification for surgical treatment. It is expected that the use of AI will improve diagnostic accuracy and thereby support better clinical outcomes. Patients with suspected significant low-gradient aortic stenosis will be enrolled. The study is observational and involves no additional risk for participants. Standard imaging studies performed for clinical indications will be additionally analyzed by the AI model, which will classify aortic stenosis as severe or moderate. The model's results will not influence the clinical management of participants but will be compared with physicians' assessments to validate its diagnostic performance. The study will be conducted in 2025-2026. The findings will provide insights into the usefulness of AI in the diagnosis of severe aortic stenosis and may contribute to the development of advanced clinical decision-support tools.
Detailed description
This study is a prospective multicenter observational validation of an artificial intelligence (AI) model for differentiating severe low-gradient from moderate aortic stenosis using transthoracic echocardiography images. The model, developed and published by the investigators, demonstrated promising diagnostic performance in retrospective data. In the present trial, approximately 300 participants with suspected significant low-gradient aortic stenosis will be enrolled during 2025-2026. Standard imaging studies performed for clinical indications will be analyzed by the AI model, which will classify aortic stenosis as severe or moderate. The AI-derived results will not influence clinical decision-making but will be compared with physicians assessments to evaluate diagnostic accuracy and reproducibility in real-world practice.
Interventions
DIAGNOSTIC_TESTAI diagnostic test for severe low-gradient aortic stenosis
All participants will undergo standard transthoracic echocardiography performed for clinical indications. Echocardiographic images will be analyzed both by experienced physicians and by the investigational AI model. Additional diagnostic tests (such as cardiac CT, low-dose dobutamine stress echocardiography or transesophageal echocardiography) may be performed if clinically indicated, according to current guideline recommendations. The AI-derived results will not influence clinical decision-making.
Sponsors
The Institute of Bioorganic Chemistry, Polish Academy of Sciences
National Institute of Cardiology, Warsaw, Poland
Study design
Observational model
COHORT
Time perspective
PROSPECTIVE
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* Age ≥ 18 years * Clinical suspicion of significant low-gradient aortic stenosis * Echocardiographic examination performed for clinical indications * Ability to provide informed consent
Exclusion criteria
* Previous aortic valve intervention (surgical or transcatheter) * Inadequate image quality precluding echocardiographic analysis * Concomitant severe valvular disease (severe mitral stenosis or mitral/aortic regurgitation) that could confound assessment * Patients unwilling or unable to provide informed consent
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Area Under the Receiver Operating Characteristic Curve (AUC) describing the sensitivity-specificity relationship of the AI model. | At the time of the nearest Heart Team meeting following the echocardiographic examination (typically within 1 week). | AUC will be calculated to assess the ability of the AI model to differentiate between severe low-gradient and moderate aortic stenosis. The analysis will use physician assessment and guideline-based diagnostic criteria as the reference standard. AUC will be reported with 95% confidence intervals. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Diagnostic performance of the AI model in clinically relevant subgroups. | At the nearest Heart Team meeting following the echocardiographic examination (typically within 1 week). | Diagnostic performance of the AI model (AUC, sensitivity, specificity) in clinically relevant subgroups, such as patients with atrial fibrillation or subtypes of low-gradient aortic stenosis (classic, paradoxical, normal flow). |
Countries
Poland
Contacts
Primary ContactMichał Wrzosek, MD
Backup ContactTomasz Hryniewiecki, Professor of Medicine
Outcome results
None listed