How AI is revolutionising maternal health in remote villages

Maternal mortality is a global problem, with nearly 95 percent of deaths occurring in low- and middle-income countries. [iStockphoto]

Artificial Intelligence (AI) is transforming maternal and child healthcare in remote regions of Africa, providing innovative solutions to challenges faced by under-served communities.

Across Africa, where maternal mortality rates remain alarmingly high, due to limited access to healthcare, AI-powered technologies are stepping in to bridge the gap.

One such breakthrough, is BabyChecker, an AI-powered portable ultrasound device developed by Delft Imaging, a Netherlands-based social enterprise.

Highlighted during the Africa Health Agenda International Conference (AHAIC) 2025 in Kigali, Rwanda, BabyChecker represents a significant milestone in maternal healthcare.

Designed specifically for low- and middle-income countries, BabyChecker provides a cost-effective and efficient solution for early pregnancy risk detection.

This portable device comprises a smartphone, an ultrasound probe, a mobile application, and a user instructions card.

Unlike conventional ultrasound machines, which require trained professionals, BabyChecker leverages AI to analyse ultrasound images.

Keep Reading

The solution helps community health workers (CHWs) with minimal training to conduct ultrasound screenings easily.

Akshay Rajagopal, BabyChecker Project Manager at Delft Imaging, explains that the technology analyses obstetric ultrasound scans to identify gestational age and high-risk pregnancies, such as foetal malposition, placenta previa, and multiple pregnancies.

Delft Imaging has a long history in screening solutions, dating back to the early 20th century when it pioneered X-ray screening for Tuberculosis (TB) in the Netherlands.

Over time, Delft Imaging transitioned to digital health solutions, incorporating AI to provide more portable and effective screening tools for diseases such as TB and maternal health complications.

BabyChecker focuses on communities with limited access to healthcare due to high maternal and child deaths.

The device consists of a smartphone, an ultrasound probe, a mobile application, and a user instructions card.

“With a long experience of providing solutions that detect diseases in remote villages, we developed the idea of BabyChecker to address the high maternal mortality that can be easily prevented by timely intervention,” explained Akshay.

“BabyChecker is not replacing the standard ultrasound, it is meant to provide ultrasound services in regions where expecting mothers cannot access these services,” he noted.

While Delft Imaging has been active in more than 85 countries, BabyChecker has been deployed in over 10 nations in collaboration with technical advisors, funders, and local implementation partners to integrate AI into maternal healthcare.

BabyChecker has been successfully piloted and commercialised in multiple regions, including Sub-Saharan Africa, Central and Latin America, and Pakistan.

The technology is designed for use by community health workers and consists of an ultrasound probe connected to a mobile phone running the BabyChecker app.

The app guides healthcare workers to perform a six-sweep protocol involving three vertical and three horizontal sweeps across the pregnant belly.

Once the six sweeps are completed, AI processes the ultrasound images in just five seconds, providing critical insights.

The results are displayed on the smartphone as traffic lights, namely green, yellow, and red.

The green signal indicates that pregnancy is progressing well and that the woman can continue antenatal care.

Yellow indicates potential complications, but they are not high-risk, while red indicates a risky pregnancy that requires immediate referral to save the lives of a mother and child.

“The results of the scan determine the care of pregnancy; for example, when the scan indicates yellow strap, the pregnant woman must visit the hospital within two weeks for check-up and management,” explained the expert.

The traffic light system is adaptable to different national health protocols, ensuring relevance across various countries.

Akshay said that the application will soon include other risk parameters.

Currently, in most remote areas, community health workers scan women using only their bare hands, often leading to inaccurate results, delayed referrals, and an increased risk of death.

The device has contributed to a rise in ANC visits, particularly during the first trimester, when many women may not yet be aware of their pregnancy.

“In Kenya, the AI technology was adopted in October 2024, and at least 500 screenings have already been conducted,” Akshay stated.

Since its initial rollout, over 9,000 scans have been performed globally by more than 200 community health workers.

Akshay emphasised that BabyChecker is tailored for rural and remote settings, requiring no internet connection or electricity.

This makes it an ideal solution for low-resource environments where maternal mortality remains alarmingly high.

“BabyChecker is designed for community health workers. No internet, no electricity; this is for rural settings,” he explained. “By equipping CHWs with AI-driven screening tools, we can reduce maternal mortality significantly.”

Health data reveals that every day, about 800 women die from preventable causes related to pregnancy and childbirth globally.

Maternal mortality is a global problem, with nearly 95 percent of deaths occurring in low- and middle-income countries.

In Kenya, about 6,000 to 8,000 women die at birth every year, with common causes of maternal deaths being obstetric haemorrhage, non-obstetric complications/indirect maternal deaths, and hypertensive disorders associated with pregnancy.

According to Akshay, AI-powered solutions like BabyChecker also reduce the burden on hospitals, ensuring that only high-risk cases are referred, saving time and healthcare costs.

Unlike traditional ultrasound machines, which require specialised training, BabyChecker can be used after watching a three-minute training tutorial, making it a game changer for untrained health workers.

When fully charged, the app can conduct 30 to 50 scans.

For BabyChecker to scale further, Delft Imaging is working closely with ministries of health, NGOs, and international organisations focused on maternal and newborn health.

The Sustainable Development Goals (SDGs) aim to reduce maternal deaths to less than 70 deaths per 100,000 live births by 2030, but low-income countries still face a significant gap in achieving this target.

Akshay urged governments and funders to invest in AI-driven solutions to support maternal healthcare in the hardest-to-reach areas.

“Anything portable and handy that can go door-to-door, screening pregnant women at their bed and home, is highly valued,” he said.

“We are bringing AI to communities, empowering health workers with high-tech screening tools to prevent maternal deaths.”

With the increasing adoption of AI in healthcare, Baby-Checker is set to play a critical role in making pregnancy screening more accessible and saving countless lives across Africa.