Deepecho

Deepecho

Deepecho

Preeclampsia early screening from blind B-mode ultrasound sweeps acquired by minimally trained operators

Rabat, Maroc

MAR

For-profit, including B-Corp or similar models

—

Preeclampsia and gestational diabetes are among the leading causes of maternal and perinatal morbidity and mortality worldwide, particularly in low-resource settings where access to specialist care and advanced diagnostics is limited. Early detection is critical—prophylaxis initiated in the first trimester can reduce the incidence of preeclampsia by up to 60%. Yet, existing clinical screening algorithms (e.g., FMF, PREDICT) require laboratory biomarkers, accurate blood pressure measurements, and trained providers—resources often unavailable in under-resourced areas. In Morocco and Nigeria, where we are leading this work, maternal mortality remains unacceptably high. In sub-Saharan Africa, the lifetime risk of maternal death is 1 in 38, compared to 1 in 5,400 in high-income countries. Preeclampsia alone accounts for 10–15% of maternal deaths globally. We aim to solve the early detection gap by enabling AI-powered screening for preeclampsia and gestational diabetes using blind-sweep B-mode ultrasound scans, collected by minimally trained health workers. This approach bypasses the need for labs or specialized skills and can be deployed widely through handheld devices. By shifting screening earlier in pregnancy and closer to the community, we hope to significantly improve outcomes for millions of women annually in under-served regions.

Our solution is an AI-powered screening tool that analyzes blind-sweep ultrasound videos collected by minimally trained health workers to detect early signs of preeclampsia and gestational diabetes in pregnancy. Using a low-cost, portable ultrasound device, health workers perform a 2–3 minute standardized scan across the abdomen without needing to identify specific fetal structures. These blind-sweep videos are then analyzed by our deep learning models, trained to recognize subtle patterns linked to adverse pregnancy outcomes—patterns that are invisible to the human eye. Unlike traditional risk assessment tools, our solution does not rely on lab tests, blood pressure measurements, or experienced sonographers. It works entirely off the ultrasound video, making it uniquely suited for low-resource settings. Results are available within minutes via a mobile app or cloud platform, helping frontline workers identify high-risk patients early—while referral and preventive treatment are still possible. We’ve tested the system on over 822 patients and demonstrated promising predictive performance. Our goal is to integrate this tool into routine antenatal care, bringing accurate, early screening to communities where traditional methods are not feasible.

Our solution directly serves pregnant women in low-resource settings, particularly in sub-Saharan Africa and South Asia, where maternal mortality remains unacceptably high and access to quality prenatal screening is limited or nonexistent. These women often live in rural or peri-urban areas where healthcare systems are overstretched, there is a shortage of trained obstetricians and sonographers, and access to laboratory testing is inconsistent. As a result, conditions like preeclampsia and gestational diabetes frequently go undetected until they cause serious harm—contributing to maternal deaths, stillbirths, and long-term complications. Our AI-powered ultrasound screening tool enables early, accurate risk prediction using a simple scan performed by minimally trained workers, bringing diagnostic capability to the community level. By identifying high-risk pregnancies in the first trimester, our solution allows for timely referrals, preventive treatments (e.g., low-dose aspirin), and closer monitoring—saving lives and reducing complications. In the long term, this tool empowers local health systems to offer equitable, accessible antenatal care, regardless of geography or income. It also reduces dependence on specialist infrastructure, making quality care more resilient and scalable. Our ultimate goal is to improve outcomes for millions of underserved mothers and babies globally.