AI in Heath: Tackling Healthcare Challenges in Kenya Through Innovation

By Melisa Mong'ina 

For years, Artificial Intelligence (AI) has been viewed as a threat to many professions. However, as technology continues to evolve, this narrative is shifting—particularly in the healthcare sector. AI is increasingly proving to be a lifesaver for both doctors and patients, aiding in faster diagnoses and enabling individuals to access better health advice.

Dr. Terry Mochire, an AI researcher at IntelliSOFT Consulting Limited, emphasizes the importance of aligning AI technologies with real, existing healthcare challenges. She underscores that thoughtful planning is essential to the success of AI-driven health initiatives.

“When we work with AI as technology builders, it’s easy to get excited about trying out new tools,” said Dr. Mochire. “But if you can’t connect that to a validated, existing problem, then what you have is just software that solves nothing.”

Dr. Mochire highlights three key AI-driven health projects at IntelliSOFT. The first, Afia AI, is a mobile health app designed to address non-communicable disease (NCD) risk factors among Kenyan youth. The second focuses on using AI to improve cervical cancer management and monitoring in Kenya. The third involves using large-language models to generate patient-facing materials after clinical visits.

“Our first project is Afia AI, which tackles NCD risk factors among Kenyan youth,” she explained. “Secondly, we have cervical cancer registry data modeling, where we leverage AI to improve end-to-end management of cervical cancer care. Lastly, we’re working on a project that uses large-language models to create educational materials for patients at the end of their clinic visits.”

Non-communicable diseases are a leading cause of death globally, with 60–70% of these deaths occurring in low- to middle-income countries. Dr. Mochire notes that many of these diseases are linked to behaviors adopted during adolescence, such as unhealthy diets, tobacco use, alcohol consumption, and inactivity.

“Seventy percent of NCD-related deaths result from habits that begin in adolescence. By raising awareness of these risk factors among Kenyan youth, we can ease the burden on our health systems and reduce financial strain on individuals,” she said.

Afia AI was developed as a mobile app to help youth understand and manage NCD risks through personalized health information, delivered by large-language models like ChatGPT and Bard (now Gemini).

In the Afia AI project, users interact with these models to ask questions about NCD risk factors, such as how smoking relates to diseases like diabetes, hypertension, or cancer. The information provided is verified using WHO and Kenyan health guidelines to ensure accuracy. After each session, users are encouraged to reconsider their health-related choices.

“Afia AI was tested among youth in Nairobi and Busia counties,” Dr. Mochire said. “We used two versions of large-language models, ChatGPT and Bard (now Gemini), to deliver behavior-changing, personalized health information. We also ensured that all information provided was fact-checked using WHO and Kenyan guidelines.”

Turning to cancer care, Dr. Mochire highlights that cervical cancer management in Kenya, and similar low- to middle-income settings, faces multiple challenges, including inconsistent care, inadequate infrastructure, and lack of centralized data systems to track patient outcomes.

“Cancer management in Kenya faces major gaps: inadequate infrastructure, continuity of care, and scaling challenges,” she said.

In response, IntelliSOFT partnered with the National Cancer Institute of Kenya to develop an advanced national cancer registry. This registry enables better tracking of cancer incidence and care outcomes, and provides insights into cancer prevalence across regions, hospitals, and populations.

“Cancer remains a leading cause of mortality worldwide and disproportionately affects low- and middle-income countries,” Dr. Mochire explained. “In 2023, IntelliSOFT collaborated with the National Cancer Institute of Kenya to develop a national cancer registry to better manage cancer.”

The registry is designed to answer critical questions about cancer burden and treatment outcomes, thereby supporting evidence-based decision-making. By leveraging AI and registry data, healthcare providers can improve treatment plans and deliver more personalized care, helping to address systemic gaps.

“With this registry, we can answer crucial questions, like how significant is the cancer burden in the country? Or more specifically, in a particular county, hospital, region, or even tribe,” she concluded.