Is AI The Way Forward For Global Health?

Despite huge advancements and progress in the world of global health over the past decades, many middle and low-income countries are still falling behind, unable to reach their sustainable development goals. This, in turn, is creating an urgency to prioritize wellbeing, and AI holds enormous promise in transforming the provision of healthcare in resource strained environments.

The Artificial Intelligence in Global Health report, funded by the USAID’s Center for Innovation and Impact, Rockefeller Foundation, and the Bill & Melinda Gates Foundation, outlined 27 cases of AI in global healthcare, and the massive potential it holds for drastically improving health in LEDC’s. The use of AI was split into four key areas – population health, patient and front line health worker virtual assistants, and physician clinical decision support. Not only does the report provide solutions that could improve the access, quality, and effectiveness of global healthcare, but it also takes into account the current maturity of AI systems and the feasibility of these solutions. Overall the report outlines an optimistic yet practical foundation for better investment and development in AI while selling a futuristic vision of healthcare and reassuring critics of artificial intelligence by addressing the issues of data ownership, privacy, and ethics.

India is one such place that could hugely benefit from the implementation of AI into its healthcare program. With half a billion people in India needing much better access to essential health services, the country’s growing digital expertise and infrastructure have great potential to improve lives. Approximately 400 million people have a smartphone, more than 50% of the population use the internet regularly, and the mobile data costs are a tenth of the global average. Aadhaar, a system of biometrically verifiable unique identification numbers, is now available to most of the country’s residents. Additionally, India has strong medical, technological and applied research institutions including the India Institutes of Medical Sciences, Indian Institutes of Technology and the Council of Scientific and Industrial Research. The country’s democracy — the largest in the world — provides accountability.

Currently, there is a considerable shortage of trained physicians and nurses, but many subsidiary health workers available shows great scope to use artificial intelligence to support medical staff in their work. India has begun to digitize public health data banks, incorporating the use of data visualization and monitoring. National health guidelines are being integrated into smartphone-based applications for community health workers and public servants. Artificial intelligence can help tackle problems created by the lack of doctors for essential tasks such as X-ray screening for tuberculosis, trauma imaging for fractures, or screening for diabetic retinopathy. The quality of AI interpretations of diabetic screenings is now even better than human analysis, not only being to diagnose diabetic retinopathy (the leading cause of preventable blindness) but can also identify age, gender, and cardiovascular risk.

However, there are emerging concerns about the applications of technology, ranging from the common skepticism of data privacy and ethics to the questionability of the effectiveness of AI compared to physical nurses and doctors. Despite the versatility of the diabetic retinopathy AI screener, the technology raised issues of uncertainty as a black-box application – where the inner workings were not fully understood – which may lead to misclassification errors. For example, algorithms to read chest X-rays that worked well in Africa have been shown to yield false positives in India, proving the AI technology to be misdirected by a patient’s skin color.

Another challenge is appealing to the diverse range of stakeholders in public health interventions, who all have contrasting principles and opinions. Firstly the collection of private data has to be collected from healthcare centers or individuals to train algorithms, which is then developed and rigorously tested by medical researchers and professionals, to be effective and accurate. Finally, there must be a large enough platform with strong infrastructure to implement the digitization of healthcare using specific technology, to accelerate distribution and uptake, not to mention the various concerns by individuals about their data privacy.

AI has the potential to revolutionize healthcare in India and globally, creating systems that could potentially save lives and improve healthcare in low and middle-income countries drastically. Yet the systems still seem to be intrusive, especially in access to rural areas and there are still many concerns over security. While Indian courts have allowed the usage of AI, it is largely restricted. Despite being a huge task, proper governance and maintenance on AI systems in healthcare could improve the health and well being of the most excluded in low-income countries: women and children. With the rapid growth of technology, it is likely that the AI systems will soon be capable enough of supporting national healthcare, but will governments take on the challenge? –AI Daily