Wearable Neonatal Cardiac Monitoring Through Chest Vibration Analysis

This non-invasive screening system uses chest motion sensors and deep learning to measure a newborn's cardiac mechanical vibrations. By analyzing these kinetic signals immediately after birth, it enables rapid, accurate detection of critical congenital heart defects.

Background

Neonatal cardiology relies heavily on the early identification of critical congenital heart diseases (CCHDs) to ensure infant survival and prevent long-term morbidity. Structural heart defects—including tetralogy of Fallot, transposition of the great arteries, and hypoplastic left heart syndrome—often require urgent medical or surgical intervention shortly after delivery. Because affected newborns may initially appear asymptomatic in the nursery, universal and reliable neonatal heart screening is essential to detect these life-threatening conditions before clinical deterioration or hospital discharge occurs. Consequently, healthcare providers require rapid, accessible, and highly sensitive diagnostic tools to evaluate cardiovascular health immediately after birth.

Despite the necessity for prompt detection, the current clinical standard for neonatal screening—pulse oximetry—suffers from significant diagnostic limitations. Because pulse oximetry relies exclusively on measuring peripheral blood oxygen saturation, it frequently fails to detect structural anomalies that do not immediately alter blood oxygen levels, such as coarctation of the aorta. Furthermore, optical oxygen sensors exhibit reduced accuracy in infants with darker skin tones, creating healthcare disparities and increasing the risk of false-negative or false-positive results. Finally, because transitional neonatal circulation takes time to stabilize, pulse oximetry cannot be reliably administered within the first 24 to 48 hours of life, delaying critical diagnoses.

Technology

This non-invasive neonatal heart screening technology detects critical congenital heart diseases (CCHDs) by analyzing the mechanical activity of a newborn’s heart. Utilizing contact-based sensors—such as wearable accelerometers and gyroscopes—or contactless optical imaging, the system captures synchronized electrocardiogram (ECG), tri-axial seismocardiography (SCG), and gyrocardiography (GCG) signals from the chest. A dedicated processing pipeline filters out respiratory and motion artifacts, segments cardiac cycles using R-peaks, and extracts time-domain, time-frequency, and three-dimensional spatial features. These isolated mechanical and spectral biomarkers, alongside clinical covariates like gestational age and birth weight, are analyzed by a multi-task deep neural network. The system then outputs a binary clinical risk status and specific CCHD classifications—such as coarctation of the aorta or tetralogy of Fallot—directly to electronic health records or clinician dashboards.

This solution is highly differentiated from the current standard of care, pulse oximetry, which relies on blood oxygen saturation levels. Because pulse oximetry cannot detect anomalies that do not alter blood oxygen, suffers from reduced accuracy in newborns with darker skin tones, and is unreliable during the first 24 to 48 hours of life, many critical conditions go undiagnosed. By directly measuring cardiac mechanics and precordial vibrations rather than optical blood oxygenation, this technology provides an equitable, skin-tone-independent diagnostic tool. Furthermore, it can be deployed immediately after birth—within the first 12 hours—to accurately identify subtle hemodynamic abnormalities and structural defects, offering a rapid, scalable, and low-cost screening method for both clinical and home settings.

Benefits

  • Enables early detection of critical congenital heart diseases within 12 hours of birth, significantly earlier than traditional pulse oximetry
  • Detects heart defects that do not alter blood oxygen levels, such as coarctation of the aorta, which are typically missed by standard screening
  • Eliminates racial and ethnic disparities in screening accuracy by directly measuring mechanical heart activity rather than relying on optical sensors affected by skin tone
  • Provides a non-invasive, low-cost, and scalable diagnostic solution utilizing accessible wearable sensors, smartphones, or contactless optical methods
  • Automates complex signal processing and deep learning analysis to classify specific structural heart anomalies with high sensitivity
  • Integrates seamlessly with existing hospital infrastructure, including electronic health records, nursing dashboards, and pulse oximetry systems
  • Supports versatile deployment across clinical settings and home-use environments for continuous or spot-check neonatal monitoring

Applications

  • Hospital newborn congenital heart screening
  • Wearable neonatal cardiac monitoring patches
  • Smartphone-based infant heart screening
  • Contactless NICU infant cardiac monitoring
  • Home-use infant cardiovascular health monitoring

Additional info

Dr. Amirtahà Taebi is an Assistant Professor with the Department of Bioengineering at Lehigh University. His work focuses on developing AI-powered sensing technologies, wearable and contactless monitoring systems, and digital twin modeling to advance noninvasive cardiovascular diagnostics. His research has been supported by local, state, and federal agencies, including the NSF, NIH, and Mississippi Institutions of Higher Learning.