Wearable-derived acoustic biomarkers for ascending aortic dilation evaluation in bicuspid aortic valve disease

Abstract Background Bicuspid aortic valve (BAV) is the most common congenital cardiac anomaly, affecting 1-2% of the population. A significant proportion of BAV patients (30-50%) develop ascending aortic dilatation, with a notable trend toward younger onset. Current clinical challenges include the lack of reliable non-invasive biomarkers for early detection and monitoring of aortic dilation, leading to delayed interventions and increased risks of aortic complications. Purpose This study aimed to explore the characteristics of acoustic cardiography (ACG) in BAV patients using wearable devices and investigate their potential as novel indicators for early warning of ascending aortic dilation. Methods We retrospectively analyzed 200 valve disease patients from Ruijin Hospital, including 74 confirmed BAV cases. All participants underwent synchronized 1-minute phonocardiogram and electrocardiogram recordings at the mitral (M), Erb’s point (E), and aortic (A) regions, and echocardiogram was performed on the same day. Morphological features of heart sounds were evaluated, and parameters including electromechanical activation time/RR (EMAT%), left ventricular ejection time(LVET), pre-ejection period /RR(PEP%), PEP/LVET, maximum energy points of S1/S2 (S1Emax, S2Emax), maximum amplitudes of S1/S2 (S1Amax, S2Amax), and frequencies of S1/S2 energy peaks (S1EmaxFreq, S2EmaxFreq) were extracted. Subgroup analyses compared these parameters between patients with ascending aortic diameters ≥45 mm and <45 mm. Results The BAV cohort exhibited a significantly younger mean age (41.5 years) compared to non-BAV valve disease patients (74 years, p<0.001). Echocardiography revealed larger ascending aortic diameters in BAV patients (39.00 mm vs. 36.00 mm, p=0.01) and a higher prevalence of aortic dilation ≥45 mm (21.62% vs. 7.14%, p=0.003). Distinct acoustic patterns were observed in BAV patients, including a novel "Extra Energy Peak at S1" morphology (45.95% vs. 1.59% in non-BAV, p<0.001). In patients with aortic diameters ≥45 mm, S2EmaxFreq was markedly elevated (46.00 Hz vs. 38.00 Hz, p<0.001). ROC analysis demonstrated moderate predictive performance of S2EmaxFreq for aortic dilation ≥45 mm, with an AUC of 0.783 (sensitivity 76.0%, specificity 74.3%, p<0.001) in the overall cohort. Notably, in the BAV subgroup, S2EmaxFreq achieved superior discriminative power (AUC=0.862, sensitivity 93.7%, specificity 79.3%, p<0.001). Conclusion The elevated S2EmaxFreq in aortic dilation and its robust predictive accuracy, particularly within the BAV cohort, suggest its potential as a novel acoustic biomarker for ascending aortic enlargement. This wearable-enabled strategy could optimize imaging surveillance intervals (echocardiography/CTA/MRI) through dynamic risk stratification, while providing continuous, non-invasive monitoring of aortic remodeling progression in young BAV patients at elevated cardiovascular risk.Baseline for valvular disease Acoustic characteristics of BAV