Don touts AI use to tackle health challenges

He gave the advice yesterday when officiating at the inception meeting for the Dataset for Maternal Health risks stratification in Tanzania for predictive and responsible machine learning.

Prof Kusiluka said the initiative marks an important moment in the journey towards leveraging advanced technologies to address critical health care challenges in the country.

He said AI and Machine Learning have emerged as powerful tools in addressing health challenges by revolutionising diagnostics, treatment and decision prediction.

Tanzania, he said, is yet to fully harness the potential of thetechnologies in improving healthcare outcomes due to insufficient data availability, limited AI infrastructure and shortage of trained professionals.

“Data is like food to AI. Without Data, machine learning cannot do much and innovation won’t thrive” he said.

“Embracing AI and machine learning holds immense promise forenhancing health care delivery particularly in remote and underserved areas, by enabling proactive interventions, personalised treatment plans and efficient resource allocation”.

He said the country must prioritise investment in data collection, machine learning and AI infrastructure to unlock
their transformative potential and ensure equitable access to quality healthcare.

To address the challenge of insufficient data availability for maternal health, the project aims to create an open and inclusive dataset for maternal health risk stratification for equitable and responsible machine learning applications in the country.

The project will also develop an AI model that will be able to predict maternal health risks (as high, medium, or low) to enable referral hospitals to focus their constrained resources on high-risks pregnancies while low risks pregnancies being managed at the community health centres.

The project will bring together academic experts in medicine, data science and machine learning from UDOM and Muhimbili National Hospital (MNH), while the regional administration and local government ministry
will be responsible for local government which plays a crucial role inensuring the project reaches the communities that needed it most.

The Ministry of Health in the project will provide invaluable guidance and ensures the project aligns with the national healthcare priorities.

The collaboration is meant for the development of impactful solutions that address the specific needs of underserved communities in the country.

The government understands the importance of investing in emerging technologies, including machine learning and AI as a strategy towards the realisation of the UN Sustainable Development Goals.

The need for machine learning, AI and other emerging technologies has been emphasised in the third National Five-Year Development Plan (FYDPIII) for 2021/2026 where the government will prioritise harnessing of development possibilities presented by big data, automation, machine learning and robotisation.

UDOM’s Principal (College of Informatics and Virtual Education), Dr Florence Rashid said the project has been funded by Lakuna in collaboration with the International Development Research Centre (IDRC) and IBM.

She said Artificial Intelligence is one of the strong areas, amongother disciplines at the college and recently they have attracted many other projects in blockchain, cyber security, school connectivity, telecommunication engineering, and Information systems and Engineering as their commitment of using ICT to address
challenges facing the community.

The project will collect maternal health risk data for underserved communities, build an annotated dataset from the collected data, to develop a predictive machine learning decision model that predicts maternal health risk and publish the dataset in open platform.

• 5