skip to content

Department of Computer Science and Technology

A project focus area in Ethiopia is applications of AI to improve maternal health and safety in childbirth.

A local expert on the technical opportunities is Mahlet Woreta, who has created some impressive demonstrations of potential IT services for the Amhara region around Bahir Dar. Mahlet grew up in a rural community, and brings that knowledge to her current work as a lecturer in Computer Science at Bahir Dar University.

After finishing her undergraduate degree, Mahlet created a prototype system using SMS to improve communication links between maternity hospitals and women expecting babies in remote locations where there is limited healthcare provision. That work won a national competition, but sadly has not found further support that might fund trial deployment.

Her subsequent Master’s thesis described implementation of a question-answering system that can respond to spoken queries in Amharic with relevant text from the local reference handbook for maternal health.

We have been discussing the opportunities to build on these experiments, with systems that are structurally compatible with local infrastructure and healthcare systems. Rural communities are supported by a network of extension workers, operating from local health posts. The Amhara region also has hundreds of local health centres that provide general healthcare, nursing and midwifery, but Mahlet identifies the need for coordination with ambulance services that can bring women to a maternity hospital for assisted births.

Mahlet has explored the use of SMS services (implemented in a PHP backend), Interactive Voice Response, and natural language speech interaction - all of which help cater to the diverse range of users involved in maternal health from medical professionals and expert midwives, to ambulance drivers, community workers, and rural women with low literacy levels.

These seem to be fascinating opportunities to explore system architectures that include a mix of human and automated systems, traditional and professional knowledge, and different capabilities in technology and language use. We are working on suitable proposals, and would welcome expert advice, or suggestions of collaborators and/or funders.