AI-Enabled POCUS: Transforming Obstetric Ultrasound in LMICs

In many low- and middle-income countries (LMICs), access to obstetric ultrasound is limited due to various challenges such as lack of trained providers, workload, and insufficient resources for sustainability. To address these barriers, artificial intelligence (AI) has emerged as a powerful tool for automating image acquisition and interpretation in obstetric point-of-care ultrasound (POCUS). This innovative technology has the potential to revolutionize antenatal care services in LMICs by improving attendance, enhancing provider-client trust, and boosting providers’ confidence in clinical decision-making.

A recent study explored stakeholders’ perspectives on how AI-enabled POCUS could transform the current obstetric ultrasound landscape in LMICs and identified key considerations for its introduction. The study involved midwives, doctors, researchers, and implementors from African LMICs. Participants completed a survey and took part in interviews or focus groups to provide insights on standard POCUS, perceptions of AI-enabled POCUS, and the potential impact on ANC quality, services, and clinical outcomes.

The findings revealed several key themes, including the priority AI capabilities, potential impact on ANC utilization and experience of care, health system integration considerations, and research priorities. Stakeholders highlighted the importance of algorithm accuracy, the need for diverse datasets, and concerns about over-reliance on AI at the expense of clinical acumen. While there was optimism about the ability of both standard and AI-enabled POCUS to improve ANC attendance and provider confidence, there were also uncertainties and concerns surrounding AI technology.

Health system integration considerations focused on task sharing with midwives, training requirements, policy issues, referral systems, and facility readiness. Stakeholders emphasized the importance of protecting providers, ensuring proper training, and addressing liability risks associated with AI-enabled POCUS. Additionally, the study highlighted the need for robust research to assess the impact of AI technology on ANC quality, early attendance, and referral systems.

In conclusion, while AI-enabled POCUS shows great promise in expanding access to ultrasound in LMICs, there are important considerations and challenges that must be addressed before its widespread implementation. By carefully navigating these issues and conducting further research, AI technology has the potential to enhance maternal and neonatal care in resource-constrained settings.