Preflight AI: Early chorioamnionitis infection warning

Earlier detection of dangerous infections during pregnancy to enable action before harm occurs.

About the Technology

One out of every 10 babies are born early each year in the U.S., and nearly half of these preterm births are caused by infections, called chorioamnionitis, that force early delivery. Doctors today must wait until they see an infection to act, by which time damage to the fetus has often already occurred. Preflight AI changes that, allowing care providers to move from reactive to proactive care.

Using a dataset of more than 44,000 pregnancies, our system detects the subtle warning signs of an infection in patient vitals a median of 16.5 hours before a clinical diagnosis. Those extra hours give doctors time to begin antibiotics before a fever starts, reducing the number of early births and their attendant risks. Preflight AI analyzes 30+ data points simultaneously, offering a continuous health monitor that achieves 95% accuracy and meets a critical, currently unmet need among doctors.

Working with the HIT Fund, we’ll implement Preflight AI at Stanford Labor & Delivery, allowing us to gather real-world performance data and compare outcomes. We’ll also conduct customer discovery to test our value hypothesis for additional use cases and evaluate go-to-market strategies.