Valvular & structural heart disease
Automatic Diagnosis of Left Valvular Heart Disease Based on Artificial Intelligence Stethoscope
Authors' conclusion
Artificial intelligence-based stethoscope is capable of diagnosing left-sided VHD accurately and may make routine screening for VHD more practical.
Study context
- Population
- Patients assessed for left-sided valvular heart disease (aortic stenosis, aortic regurgitation, mitral stenosis, mitral regurgitation)
- Participants
- 514 patients
- Comparison
- AI-based stethoscope model vs echocardiography
- Trainingn = 304
- Testingn = 210
- Outcome
- AUC of the model for detecting left-sided valvular heart disease, against echocardiography
Methodological detailsderivation and testing groups
- Derivation and testing groups
Performance
As reported in the abstract
Left-sided VHD detection (vs echocardiography)
| Test or model | Sensitivity | Specificity | AUC |
|---|---|---|---|
| Testing | 70.00% | 73.68% | 0.7554 |
Testing
Sensitivity70.00%
Specificity73.68%
AUC0.7554
Further findings
Diagnostic performance of ML model by lesion (model development)
- Aortic stenosis AUC
- 0.7621
- Aortic regurgitation AUC
- 0.7075
- Mitral stenosis AUC
- 0.6426
- Mitral regurgitation AUC
- 0.7906
- Left-sided valvular heart disease (VHD) AUC
- 0.8541
- Left-sided VHD sensitivity
- 83.07%
- Left-sided VHD specificity
- 78.26%
Original publication
Automatic Diagnosis of Left Valvular Heart Disease Based on Artificial Intelligence Stethoscope. JACC Adv. Published 20 August 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →