The report “AI in Global Health: Defining a Collective Path Forward” was funded by USAID’s Center for Innovation and Impact (CII) https://www.usaid.gov/cii and the Rockefeller Foundation https://www.rockefellerfoundation.org in close coordination with the Bill & Melinda Gates Foundation https://www.gatesfoundation.org.The report analyzes the thinking across the global health field on deploying and scaling Artificial Intelligence (AI) in global health.
The report identifies opportunities for donors, governments, investors, the private sector, and other stakeholders to explore and accelerate the appropriate development and most cost effective use of AI in global health.
The report explores use cases with the highest potential in the global health context and the challenges to scaling AI in “Low and Middle Income Countries” (LMIC). The goal is to understand which barriers may require more strategic and deliberate intervention, and potential investments as part of a coordinated approach to funding AI in global health.
Management tools using AI will help to better understand future outbreaks of diseases. The use of “AI in Medical Epidemiology” (AIME) would help a country’s Ministry of Health predict future outbreaks of diseases like Zika and dengue in a specific geography.
For example, the Ministry of Health in Malaysia, has fully digitized and integrated EMRs from health facilities across the country and has applied AI to this data. This has enabled the Ministry of Health to map various health burdens and disease outbreaks occurring across the country by using machine learning to identify correlations among multiple variables across complex data sets to identify risk factors.
Tools supporting AI are available to frontline health workers to help workers triage and diagnose patients often outside of health facilities, assist with clinical decision support, and monitor compliance of their patients.
Also, the availability of tools to patients can help patients direct their own care by identifying the type and severity of a patient’s condition and provide health recommendations directly to the patient. These recommendations may include how and where to seek care if needed, guidelines for self-care, and behavioral health changes outside of the health system.
The availability of quality data and the quantity of the data is essential for AI tools to be developed. However, due to the many current efforts by global health funders to capture digital health data, the report recommends that stakeholders strengthen their coordination efforts to maximize the overall potential impact.
Some steps need to be taken to deal with the massive need for data by health funders such as:
- Identify or create an open data platform for sharing datasets relevant to AI in global health
- Develop a platform for publication of datasets relevant to AI and publicize the availability of the data
- Add datasets to the platform to ensure that the data is representative of a variety of contexts
- Encourage funded innovators to collect and share data where possible
- Encourage developing and deploying at-scale data collection systems which may include EMR systems and the use of smart phones by frontline health workers to capture data
- Develop digitization hubs to convert historical analog data.
One of the main challenges to address interoperability. Several actions should entail setting up a common language, developing standards so that developers can build applications. For example, app developers could create integration points for the AI data system so one system can transmit data to another system, develop international standards bodies, and governments could create standards for the data sharing to facilitate alignment with global and local health protocols and systems.
To view and/or download the report, go to https://www.usaid.gov/cii/ai-in-global-health.