Researchers at the University of Washington have developed a smartphone app that tracks fetal heart rate as accurately as clinical tools. No additional hardware, no gel, just the phone’s speaker and microphone.
It’s called DopFone. The system plays an 18 kHz tone and waits for the echo. A machine learning model estimates heart rate based on subtle changes in reflected sound.
In a study of 23 pregnant patients, the app landed a medical-grade Doppler in about 2 beats per minute. This is well within the 8 BPM margin accepted by doctors.
The idea is to give more people access to regular fetal monitoring, without expensive equipment or constant clinic visits. But the technology has its limitations and is not yet ready for prime time.
How it works and where it stumbles
You hold the phone’s microphone against your stomach for a minute. The speaker emits a deep sound that penetrates the tissue. When this sound bounces back, the baby’s heartbeats create tiny frequency shifts that the phone picks up.
The team tested the app on patients between 19 and 39 weeks of pregnancy. It held up in different phone positions and angles. This is important at home, where no one guides you. However, accuracy decreased slightly in people with higher BMI. The readings remained within normal limits, just not as close.
The study also excluded people with an irregular fetal heartbeat. These are emergencies, not a place for experimental technology. Therefore, we don’t yet know how the app deals with abnormal rhythms.
What the app can’t yet do
DopFone currently only works for singleton pregnancies. The team didn’t test it on twins or triplets. Two heartbeats would complicate the signal.
All testing was conducted on iPhones in a controlled medical environment. How well does it work on Android? In a noisy living room? We don’t know yet. The researchers plan to collect more real-world data.
The study focused on the second and third trimesters, there were no data for first trimester monitoring.
And since all participants in the study had healthy heart rates, the app’s ability to flag distress signals like tachycardia is unproven.
When you might actually use it
Don’t search for DopFone in app stores yet. The team first needs further data on everyday conditions. They also train the model on more diverse hardware and patient populations.
The long-term vision is a free or low-cost app that works on any phone. That could change for people in rural areas or low-resource regions where traditional Doppler devices are rare.
It could also help high-risk patients get more frequent checks between doctor visits. The researchers emphasize that this is not a replacement for medical care. When it arrives it will be a gauge, not a diagnostic device. You still need a doctor to interpret the data.
