# AI-powered obstetric technology research project co-led by Dartmouth professor awarded $4.8 million federal contract - The Dartmouth

*Источник: The Dartmouth*
*Дата: 2026-07-31*
*Язык: en*

**Кратко:** In June, the Advanced Research Projects Agency for Health awarded a $4.8 million federal contract to a team co-led by Dartmouth computer science professor Tam Vu and Columbia University computer science professor Xia Zhou to support their first year of research for their obstetric technology research project, Hypoxia Assessment via Real-time Maternal-fetal Oxygenation and Neurophysiological Integration. The project seeks to create an AI-powered wearable belt and ear device that non-invasively tracks fetal and maternal heart rates in real time during pregnancy.

In June, the Advanced Research Projects Agency for Health awarded a $4.8 million federal contract to a team co-led by Dartmouth computer science professor Tam Vu and Columbia University computer science professor Xia Zhou to support their first year of research for their obstetric technology research project, Hypoxia Assessment via Real-time Maternal-fetal Oxygenation and Neurophysiological Integration. The project seeks to create an AI-powered wearable belt and ear device that non-invasively tracks fetal and maternal heart rates in real time during pregnancy.
The Dartmouth sat down with Vu and Zhou separately to discuss AI in healthcare, their research on medical technology, and HARMONI.
Can you tell me more about your research background?
XZ: I started my faculty career at Dartmouth thirteen years ago. Our lab has been working on wireless sensing and communication. These days, we’re looking more into the use of light and textiles as a sensing modality for human sensing and object tracking. The most relevant experience to this ARPA-H project is a prior project called Joey, a wearable textile technology that non-invasively monitors newborn kangaroos’ vital signs continuously. That line of work directed my attention to neonatal and maternal care. We’re now broadening the technology to maternal and fetal monitoring and hoping to get a more holistic view of the mother and the fetus.
TV: I research mobile and wearable technologies. I’ve spent my time on a new concept called “earable computing,” which turns the space inside and around the ear into a platform that can sense health signals. From there, we were able to capture many other vital signals. From that work, my work broadened out to building a wide range of wearable and wireless sensors for healthcare and medicine. We build small and low-power devices that can measure things, and then we use that information and AI to make a decision.
How does HARMONI work? What is your role in the project?
TV: HARMONI has an array of sensors that help monitor the baby and mother at the same time, and then we use machine learning to identify the connection between the mother’s and baby’s signals. The Dartmouth team is the integrator of all the pieces. We have folks making an ultrasound device or sensors, others making the smart belt at Columbia and others making the earpiece; all of that would then come together into a single integrated system.
From the institutional perspective, Columbia is the lead, but from the management perspective, I’m a principal investigator and system architect for the whole project. We started with designing the approach and identifying and communicating with potential collaborators and healthcare partners.
XZ: I’m a co-principal investigator on the project. At Columbia, we have multiple subteams. I’m leading a computer science team. There are also two teams at the medical center at Columbia Irving Medical Center, one on the OB/GYN side and another in psychiatry. We will be doing the visibility study at the medical center, and we will also work with them on the data analytics to know how to process these physiological signals for hypoxia index determination.
What do you hope to achieve with HARMONI?
TV: We proposed HARMONI as a system for monitoring the health of a baby during labor and delivery. The goal is for the system to tell clinicians in real time whether a baby is getting enough oxygen for decision-making purposes. [Supplemental oxygen may be administered to newborns struggling to begin breathing after birth to prevent hypoxia.]
XZ: We are hoping to develop wearable, non-invasive fetal-maternal monitoring technologies for hypoxia detection that can ultimately be used at every bedside in the delivery room. The goal is to collect both fetal and maternal physiological signals to detect or identify the degree of hypoxia and provide information on the etiology so that clinicians and doctors can take more informed actions in the labor room.
What is the anticipated timeline of your project?
XZ: We are currently funded for Phase I, or the initial prototyping. We’ve started the design of the belt and earpiece, are testing different components, and have joined the circuit design of various components. The goal is to have an initial integrated prototype around month 5 or 6. In the meantime, we are also talking to other ARPA-H-funded teams and seeing how their sensors can be integrated into our platform.
What role do you see AI taking in this project?
XZ: One thing we hope AI can help with is to reduce motion-induced noise so that we can glean a signal from our sensors. We’re working with multimodal data from different individual sensors, and we can use AI to process that data, do a more comprehensive analysis of these signals, and keep useful information on the hypoxia status, including whether something is fetus-induced or mother-induced. We hope to use AI as a last resort, using as much modeling and signal processing as possible to have a deep understanding of the whole supply chain of hypoxia.
These interviews have been edited for clarity and length.
Sahil Gandhi ’29 is a reporter from Staten Island, N.Y., and is majoring in environmental studies and government modified with philosophy and economics. He loves word searches and falling down internet and Wikipedia rabbit holes.

[Оригинал](https://www.thedartmouth.com/article/2026/07/ai-powered-obstetric-technology-research-project-co-led-by-dartmouth-professor-awarded-4-8-million-federal-contract)