Homegrown AI Solutions for COVID-19 Response and Recovery in Uganda

Authors: COAST Project Uganda

Thursday, 16 December 2021

Data is the lifeline of today’s interconnected society and can be harnessed responsibly to address the growing impact of the COVID-19 pandemic. In Uganda, where there has been a significant contraction in economic activity, added strain on the public health system and a range of social impacts, there is a burgeoning opportunity to explore data innovations and technological approaches to design homegrown, contextualised solutions. Seizing this opportunity, Makerere University Artificial Intelligence lab, in partnership with the Infectious Disease Institute, launched the COAST Project with a view to developing end-to-end Artificial Intelligence (AI) and data systems for targeted surveillance and management of COVID-19 and future pandemics that could affect the country.  

The project aims to use AI and integrated data systems to optimise and improve the efficiency, response and recovery from COVID-19 and future pandemics, with emphasis on the needs of underserved cohorts within Uganda’s population. As a multidisciplinary approach, COAST leverages AI, epidemiology and computer science to build a  set of synergistic, contextualised and equitable end-to-end AI data systems that can generate insights to inform decision-makers and the public as part of the ongoing COVID-19 response. 

To achieve this mission, COAST has outlined three specific objectives that are being operationalised across four workstreams:

Objective 1 - Inclusive and Equitable Datasets

The first objective is to strengthen existing data systems in the country, which can help produce more usable and equitable datasets for AI-driven inventions. COAST is combining multiple datasets such as existing and ongoing near-real-time radio broadcast data, allowing for the analysis of male and female community voices on COVID-19 topics and issues. In Uganda, there has been an uptick in mis- and disinformation about COVID-19, not least about how the virus spreads, preventative measures, and vaccine efficacy. To address arising public concerns, COAST is mining audio radio data to better understand community perceptions using a speech-to-text model. The team is also collecting call records from health facilities and community healthcare worker questions about the prevention and management of COVID-19 and other infectious diseases. 

Objective 2 - AI-Powered Detection and Diagnostic Tools

The second objective is to develop and deploy AI-driven detection and diagnostic tools for improved patient care and management. COAST is engineering a decision support system called Call For Life (CFL), a Ugandan COVID-19 AI chatbot and machine learning guided screening tool. The chatbot will be an automated, AI-driven virtual assistant that provides individuals with health care assistance through both audio and text, and through enhanced modelling will be able to produce user-specific responses to a wide variety of questions. The chatbot will serve both English and Luganda speakers, further expanding the potential for enhanced automated health telecommunications within communities. 

The financial and logistical limitations related to PCR testing often hamper efforts intended to reduce community infections. Therefore, there is a need to fasten and automate the detection of COVID-19, for instance through lung ultrasound imaging which is cheaper, more portable and repeatable. COAST aims to develop a screening model that can help with improving COVID-19 patient care and management, through the use of AI and point of care lung ultrasound imaging. With this model, a COVID-19 risk assessment can be evaluated in less than a minute and without added costs or risk of exposure to a suspected patient.

Objective 3 - Targeted Government Responses and Inventions

The third objective is to model and evaluate COVID-19 interventions for targeted government responses, based on the fused datasets described above in the first and second objectives. AI and mathematical forecasting models will be developed for real-time tracking of COVID-19 spread and associated risks to inform the design and development of regional and population-specific interventions. Using AI techniques, COAST will analyse radio conversations to identify issues affecting males and females within communities who may not be connected online, as well as to better public perceptions towards COVID-19 measures and public policies, particularly in vulnerable communities. The team is also modelling the association between urban air quality, COVID-19 cases, and human mobility to assess the impact and effectiveness of government interventions. 

Creating Future-Proof Interventions  

Since the pandemic was declared in March 2020, Uganda like the rest of the world has been greatly impacted by the COVID-19 crisis. The number of cumulative cases has since grown to more than 100,000 with more than 3,000 deaths as of early November 2021, according to the Ministry of Health. By leveraging existing and emerging datasets as well as adopting a multidisciplinary research approach, COAST hopes to advocate for the responsible and ethical use of digital, AI-driven health solutions in Uganda. Overall, the pandemic has highlighted the need to embrace homegrown, contextualised AI solutions in the country to effectively address challenges being faced in the public health system, and create future-proof interventions to ensure better preparedness. 

The COAST project is being implemented with grant support from Canada’s International Development Research Centre (IDRC) and the Swedish International Development Cooperation Agency (Sida) as part of the Global South AI4COVID Program. To learn more about our work follow us on Twitter and visit our website.

Tags: #COVID19, #InformationSystems, #ArtificialIntelligence, #PublicHealth