Faculty mentor/PI email address
ekoman@virtua.org
Keywords
Valvular Heart Disease, Artificial Intelligence, Digital Auscultation, Point-of-Care Screening, Diagnostic Accuracy
IRB or IACUC Protocol Number
Non-IRB research
Date of Presentation
5-6-2026 12:00 AM
Poster Abstract
Background: Valvular heart disease (VHD) affects >10% of adults aged 75+ yet remains underdiagnosed when asymptomatic due to declining auscultatory proficiency. AI-augmented digital auscultation offers a point-of-care screening solution, though no meta-analysis has pooled diagnostic accuracy across VHD subtypes in adults using echocardiography as the reference.
Methods: A systematic review and meta-analysis were conducted per PRISMA guidelines. PubMed, Embase, Cochrane, IEEE Xplore, and Scopus were searched without date restriction. Studies applying AI or machine learning to digital auscultation or phonocardiography for VHD classification in adults with echocardiographic reference and patient-level diagnostic metrics were included. Studies using only public datasets, pediatric populations, or non-patient-level data were excluded. A bivariate random-effects model was used for primary analysis with secondary AUROC pooling.
Results: Of 223 records identified, 160 title/abstract were screened after deduplication. Twelve studies were included in qualitative synthesis; nine were suitable for pooling (n=3,542 patients, 6 countries). Pooled sensitivity was 82.0% and specificity 87.2%, with an SROC AUC=0.91. Secondary AUROC pooling yielded 0.94. Aortic stenosis subgroup (5 studies) demonstrated sensitivity 91.6% and specificity 89.4%, while composite VHD endpoints (4 studies) showed sensitivity 67.3% with specificity 84.1%. Heterogeneity was moderate (I²=34.3–64.3%). 4 studies included clinician comparator; AI demonstrated higher sensitivity in 3 of 4.
Conclusions: AI-augmented digital auscultation demonstrates high diagnostic accuracy for VHD detection, with optimal performance for aortic stenosis and attenuated sensitivity in heterogeneous populations. Results support its role as a point-of-care screening tool in primary care and resource-limited settings. Standardization of device type and reference standards is warranted.
Disciplines
Artificial Intelligence and Robotics | Cardiovascular Diseases | Medicine and Health Sciences
AI-Augmented Digital Auscultation for Point-of-Care Screening of Valvular Heart Disease: A Systematic Review and Meta-Analysis
Background: Valvular heart disease (VHD) affects >10% of adults aged 75+ yet remains underdiagnosed when asymptomatic due to declining auscultatory proficiency. AI-augmented digital auscultation offers a point-of-care screening solution, though no meta-analysis has pooled diagnostic accuracy across VHD subtypes in adults using echocardiography as the reference.
Methods: A systematic review and meta-analysis were conducted per PRISMA guidelines. PubMed, Embase, Cochrane, IEEE Xplore, and Scopus were searched without date restriction. Studies applying AI or machine learning to digital auscultation or phonocardiography for VHD classification in adults with echocardiographic reference and patient-level diagnostic metrics were included. Studies using only public datasets, pediatric populations, or non-patient-level data were excluded. A bivariate random-effects model was used for primary analysis with secondary AUROC pooling.
Results: Of 223 records identified, 160 title/abstract were screened after deduplication. Twelve studies were included in qualitative synthesis; nine were suitable for pooling (n=3,542 patients, 6 countries). Pooled sensitivity was 82.0% and specificity 87.2%, with an SROC AUC=0.91. Secondary AUROC pooling yielded 0.94. Aortic stenosis subgroup (5 studies) demonstrated sensitivity 91.6% and specificity 89.4%, while composite VHD endpoints (4 studies) showed sensitivity 67.3% with specificity 84.1%. Heterogeneity was moderate (I²=34.3–64.3%). 4 studies included clinician comparator; AI demonstrated higher sensitivity in 3 of 4.
Conclusions: AI-augmented digital auscultation demonstrates high diagnostic accuracy for VHD detection, with optimal performance for aortic stenosis and attenuated sensitivity in heterogeneous populations. Results support its role as a point-of-care screening tool in primary care and resource-limited settings. Standardization of device type and reference standards is warranted.