Soft Fabric Necklace Tracks Care of Preterm, Low-Birth-Weight Infants
A Columbia Engineering team, led by Xia Zhou, associate professor of computer science, decided to see if they could make KMC monitoring easier. An expert in mobile computing and networks, Zhou’s research is focused on wireless systems and mobile health. Her interest was sparked after a conversation with her colleague Fred Jiang, associate professor of electrical engineering at Columbia Engineering, whose team was exploring solutions to KMC monitoring by using microphones. Zhou’s group was already studying fabric sensing, so they started to investigate a fabric approach to the problem.
They knew that the electrocardiogram (ECG) waveform yields critical clinical metrics that are essential for the diagnosis and management of various cardiac issues and diseases. So they proposed a fabric-based wearable technology for KMC monitoring, and designed Joey as a lightweight, soft fabric necklace worn by the caregiver on her chest skin.
Qijia Shao, Zhou’s PhD student and the study’s lead author, came up with Joey’s elegant design, which continuously monitors the chest-to-chest skin contact duration between the caregiver and an infant, as well as two vital signs essential to an infant’s well-being: heart rate and respiration rate. And all without needing to place any sensors on the baby’s body.
“Our work is the first study that systematically explores the transmission of electrocardiogram (ECG) signals across human bodies as well as the conditions for sensing mixed ECG signals,” Zhou said. “This work extends the application of fabric-based sensing to multi-user scenarios and opens exciting possibilities for new developments in multi-user fabric sensing and interaction.”
How Joey works
Joey achieves the team’s monitoring goals by exploiting the transmission of ECG signals across individuals during skin-to-skin contact when the sensing path crosses their hearts, using fabric sensors to measure the electrical activity in response to cardiac activity. It can detect the presence of mixed ECG signals of the two bodies in contact to measure the chest-to-chest skin contact duration. Zhou’s team then developed computational algorithms to extract infants’ ECG signals for vital sign monitoring and to mitigate the impact of motion to ensure reliable, accurate sensing.
Supporting Kangaroo Mother Care with Computational Fabrics
The sensing monitor used by Joey to oversee the Kangaroo Mother Care of preterm, low-birth-weight infants. Credit: Courtesy of Zhou lab.
Collaboration with Columbia’s neonatal pediatricians
The researchers worked with 35 participants to demonstrate Joey’s accuracy in estimating KMC duration with an average accuracy of 96% and vital sign estimations with clinically acceptable accuracy. Interviews with eight pediatricians in the Division of Neonatology at the Columbia University Irving Medical Center (CUIMC) further confirmed the clinical usability of Joey’s sensing fabric for infant skin. The system is also robust against various practical factors such as motion noise, repeated washing, and variations in skin conditions.
“I am very excited about our findings because they demonstrate the promising potential of physiological sensing using everyday conductive fabrics, a ubiquitous and natural sensing medium,” said Shao. “The comfort and ease of wear of these soft, sensing materials offer a significant advantage over rigid, adhesive sensors, which have been the mainstream methods for physiological sensing.”
The study was presented on June 5 by Shao, who is graduating this summer and joining Hong Kong University of Science and Technology as an assistant professor, at the ACM MobiSys 2024 conference in Minato-ku, Tokyo, Japan, where it won Best Paper Award and the Joey demonstration video won People’s Choice Demo Award.
What’s next
Zhou’s team is continuing to work with pediatricians at CUIMC, discussing potential collaboration ideas, including conducting clinical studies of the technology. The group is also exploring other applications of fabric-based physiological sensing, such as sleep monitoring and wearable robotics.