In a labor and delivery unit somewhere right now, a strip of paper is scrolling out of a machine. It traces a baby's heartbeat against the rhythm of contractions. A nurse studies the squiggles. A doctor frowns. And a decision gets made - sometimes the right one, sometimes a precautionary cesarean that, with better information, might never have happened. This is modern childbirth: high stakes, low resolution. Raydiant Oximetry was built to change the resolution.
The company is small - around 25 people working out of San Ramon, in the quieter eastern stretch of the San Francisco Bay Area. It is clinical-stage, which is the polite industry term for "not yet selling anything." And yet it has a partnership with GE HealthCare, an FDA Breakthrough Device designation, money from the Gates Foundation, and a first-place finish in a national NIH innovation challenge. For a company that hasn't shipped a commercial product, that is a remarkable amount of institutional confidence.
The monitors used to watch babies during labor have, by the company's own framing, roughly the accuracy of a coin toss when it comes to predicting fetal distress.
A monitor that mostly tells you to worry
Here is the uncomfortable thing about electronic fetal monitoring, the standard of care for half a century: it is sensitive but not specific. It is very good at flagging that something might be wrong and quite bad at telling you whether something is. A heart-rate dip can mean a baby in genuine oxygen distress. It can also mean a baby that is simply being squeezed by a contraction and is perfectly fine. The tracing cannot tell the difference. The clinician has to guess.
Guessing has a cost. Roughly one in three American births is now a cesarean - a major abdominal surgery, with its own risks for mother and child, that is sometimes ordered on the strength of an ambiguous squiggle. Childbirth is the most common reason people are admitted to U.S. hospitals. The tool at the center of it has barely changed since the Beatles broke up.
The question every fetal heart-rate monitor quietly dodges is the only one that matters: is the baby actually getting enough oxygen?
A pediatric anesthesiologist with a hunch
Neil Ray is a pediatric anesthesiologist by training, which means he has spent a great deal of his career in operating rooms, including the ones where C-sections happen. He noticed something that bothered him: he was helping deliver babies via surgery without ever being certain the surgery was necessary. The data didn't exist to know. So in 2016 he started a company to create it.
The bet was elegant, if not exactly easy. Pulse oximetry - the little clip on your finger that reads your blood oxygen - is one of the most trusted measurements in medicine. The trouble is that you can't clip anything onto a baby who is still inside the womb. Ray's wager was that you could measure fetal oxygen through the mother's abdomen using light, and then use machine learning to pull the baby's faint signal out of the noise. It is, on paper, an absurdly hard signal-processing problem. That is rather the point - if it were easy, someone would have done it.
Pulse oximetry is everywhere in medicine. The one place it had never reliably worked was the one place a wrong call leads to surgery.
Lumerah, in plain terms
Lumerah is a noninvasive, transabdominal fetal pulse oximeter. Unpacked: it sits on the outside of the mother's belly, shines light through tissue, and reads the oxygen saturation of the baby's blood directly. No scalp electrode screwed into the baby's head. No waiting for the water to break. Just a sensor and a model that turns a messy optical signal into a number a clinician can act on.
The "directly" is what separates it from everything in the room today. Conventional monitoring infers a baby's wellbeing from heart-rate patterns - a proxy for a proxy. Lumerah skips the inference and measures the thing itself. Combine that reading with risk-stratification software, and the promise is a delivery room where the question "is this baby okay?" has an actual answer instead of an anxious shrug.
Technology
Biophotonics sensing + machine-learning signal processing
How it sits
Noninvasive, on the maternal abdomen - nothing inside
What it reads
Fetal blood-oxygen saturation, in real time during labor
The goal
Fewer unnecessary C-sections, fewer missed emergencies
The road so far
Who's betting on it, and how much
Validation in medical devices doesn't come from press releases - it comes from regulators, clinicians, and the kind of money that does its homework. Raydiant has collected all three. The FDA granted Lumerah Breakthrough Device designation, a status reserved for technologies that could meaningfully improve care for serious conditions. A 35-patient Early Feasibility Study has been completed, and an IDE pilot study is now running at LSU Health Shreveport.
Then there's the cap table, which reads like a who's-who of people who care about mothers and babies for a living: Cross-Border Impact Ventures, RH Capital, the Global Health Impact Fund, VCapital, and the March of Dimes Innovation Fund. Add grants from the Bill & Melinda Gates Foundation, the Government of Ireland, and the NIH. When the Gates Foundation and a 130-year-old maternal-health charity are both writing checks, the mission is doing some convincing of its own.
Where the $31.7M came from
You can fake a lot of things in a startup deck. An FDA Breakthrough designation and a 35-patient clinical study are not among them.
A public-health goal wearing a startup's clothes
Most startups want to win a market. Raydiant wants to move a statistic. The mission is to reduce maternal and neonatal mortality and morbidity - the kind of language you'd expect from a nonprofit, backed here by the structure of a venture-funded device company. That hybrid shows up everywhere: impact funds beside traditional VCs, foundation grants beside Series A money, a partnership with one of the largest imaging companies on earth.
It also explains the patience. Device companies measure progress in regulatory milestones, not viral growth, and the people funding Raydiant clearly knew what they were signing up for. The reward, if it lands, isn't just market share. It's a delivery room that surprises fewer families.
Back to that scrolling strip of paper
Return to where we started: the nurse, the doctor, the squiggles, the decision. Now imagine a second number on the screen - not the baby's heart rate, but the baby's actual oxygen level. The contraction that would have triggered an anxious call to the OR instead shows a baby that is well-oxygenated and fine. The labor continues. The cesarean doesn't happen. Or the opposite: a tracing that looked reassuring is overruled by an oxygen reading that isn't, and a real emergency gets caught earlier.
That is the room Raydiant Oximetry is building toward. Not a flashy one. Just one where the most important question in childbirth finally has a measurement instead of a guess. The paper may keep scrolling. But the people reading it will, for the first time in decades, have something better to go on.