Vibration-Based Tools Aim to Improve Newborn Screening for Heart Defects

Early recognition is essential because affected infants require medication or surgery within the first year of life. Hospitals rely on pulse oximetry to screen newborns, yet the test can miss defects and may not perform equally across skin tones. To help address this challenge, researchers at Lehigh University are developing new vibration-based and contactless tools to strengthen newborn screening for these conditions.

Lehigh University’s (Bethlehem, PA, USA) P.C. Rossin College of Engineering and Applied Science is advancing seismocardiography, a noninvasive method that uses a small chest sensor to record vibrations generated by cardiac activity. Structural abnormalities create distinctive vibration patterns that can be analyzed for disease signatures. The approach has yielded clinically meaningful information in conditions such as heart failure. The team is applying this experience to identify critical congenital heart diseases in newborns.

According to the Centers for Disease Control and Prevention, about one in four babies born with a heart defect has a critical congenital heart disease, and treatment must occur within the first year of life. All babies born in U.S. hospitals undergo pulse oximetry screening to estimate blood oxygen saturation. The method detects light absorption by hemoglobin, but its algorithms were trained primarily on lighter skin tones, leading to overestimated oxygen levels in some patients with darker skin. About half of U.S. newborns have darker skin tones, which heightens the risk of missed disease. In addition, some critical congenital heart diseases do not reduce oxygen levels, limiting what pulse oximetry can reliably detect.

Earlier work at Mississippi State University enabled the group to collaborate with Nemours Children’s Hospital in Orlando by sending sensors to capture chest vibrations in newborns with and without critical congenital heart diseases. Now based at Lehigh, the investigators are processing those recordings and plan to expand data collection with Nemours. The immediate objective is to demonstrate that vibration signals can detect these defects in the clinical newborn population.

The team is developing a contactless option in which parents could use a cell phone to record short videos of a newborn’s chest for motion analysis. Working with clinical collaborators, they are studying whether video-based vibration extraction can indicate the presence of a critical congenital heart disease and prompt timely evaluation. 

A complementary digital twin of the thoracic region, created from patient computed tomography or magnetic resonance imaging, will model how cardiac vibrations propagate to improve algorithm training. The long-term goal is to pair these tools with pulse oximetry and ultimately deliver a primary screening approach that performs across all skin tones and disease types.

“Right now, there are newborns who are suffering because the method for detecting their disease is ineffective. Their parents take them home, full of joy over a baby that means the world to them. But these infants can become very sick. Their skin turns blue, and they can't breathe properly. The parents are devastated. If we could help families avoid these situations, nothing would make me happier,” said Amirtahà Taebi, assistant professor of bioengineering in Lehigh University's P.C. Rossin College of Engineering and Applied Science.

Related Links
Lehigh University P.C. Rossin College of Engineering and Applied Science