While artificial intelligence stands to bring rapid improvements to the healthcare sector, director-general of the World Health Organisation Margaret Chan has warned that it must be for the good of everybody – not just the wealthiest countries.
“What good does it do to get early diagnoses of skin or breast cancer if a country does not provide the opportunity for treatment or if the price of medicines are not affordable?” she asked the audience at the UN’s AI Summit for Good. “Many developing countries don’t have health data to mine. And they don’t have functioning systems for registering vital causes of death stats.
Read more: DeepMind accused of accessing NHS data on an 'inappropriate legal basis'
“Enthusiasms for smart machines reflect the perspectives of well-resourced companies and wealthy countries. We need a wider perspective.”
Chan shared her first-hand experiences of this disparity of wealth in her role at WHO, encountering medical facilities that are forced to function without electricity or running water. “I would be hard pressed to promote to these countries the advantages of AI when even standard machines for analysing samples or sterilising equipment cannot run for want of electricity. Any discussion of smart machines revolutionising healthcare must be alert to these huge gaps in capacities.”
One thing that all countries share however, she says, is a move away from communicable diseases being the greatest threat, to non-communicable diseases such as heart disease, diabetes and hypertension. It’s the result of the “same dominant forces” shaping our health everywhere: “ageing, rapid unplanned urbanisation and the globalised marketing of unhealthy products.” AI can help address these though, she added, and wearables have already been deployed to track and monitor health.
"AI experts, what gadgets can you come up with to empower individuals to make healthy choices?" she implored the room.
Many of the concerns Chan raised about our AI-driven future are already relevant. Machine learning can help speed up diagnoses, for instance, however Chan emphasises the need for a human overseer: “medical decisions are complex and I doubt a machine will ever be able to imitate genuine human compassion.” Now that administrative practices are digitised and updating systems has become another task for medical professionals, there are already concerns that physicians get less face time with patients.
“Machines can aid the work of doctors and streamline processes leading to decisions. But AI cannot replace doctors and nurses in their interactions with patients,” Chan added.
She also warned about over-reliance on technological tools. Medical history, she says, is full of examples of new tools not being granted licenses or being withdrawn because the risk of problems outweighs the potential benefits. “Wearables for monitoring cardiovascular performance are already being questioned...tools can sometimes give a false sense of safety and security.”
Patient confidentiality will also remain a priority, as institutions strive to mine increasing amounts of data to improve systems. In the UK we have already seen the potential problems posed by technology companies collaborating with the health sector: DeepMind’s collaboration with the NHS resulted in complaints over a lack of transparency, a "special relationship" between a public body and private company, and the National Data Guardian accused the NHS of handing 1.6 million patient records given to DeepMind on an "inappropriate legal basis".
Regulatory issues around the use of AI in healthcare are abundant and go beyond privacy, however. “What happens if a smartphone app misses a symptom that signifies a severe underlying disease?" asks Chan. "Can you sue a machine for medical malpractice? Medical devices are heavily regulated and with good reason. Doctors and nurses are licensed to practice medicine and undergo continuing study. The question is how do we programme a machine to think like a human?”
“AI is a new frontier for the health sector...But we do not have the answers to many questions around AI - we’re not even sure we know all the questions that need to be asked.”
This article was originally published by WIRED UK
