AI & Healthcare: The Prescription Algorithm
AI in the medical sphere has been progressing nicely in Scandinavia and Asia, but digital health is curiously stunted in the UK.
Unless you believe the click-bait —i.e. “Google is using Machine Learning to predict a patient’s death with 95% accuracy!” — AI’s progress into the medical system has been a slow, meandering slog.
The Wire summed up the reason only last week:
“AI has no place in the NHS if patient privacy isn’t assured,” reported .
In the next five years, however, AI’s assimilation into the health system is set to advance. An expert panel at London’s recent Connected World Summit discussed the inherent pitfalls and rampant potential of this impending clash of tech and ethics.
Digital Medical Records
Despite the ubiquitous digitisation of data (notably our friendships and finances), the most important information out there — our health — hasn’t made the leap to digitisation.
The obvious push-back is personal data being abused and sold for commercial gain.
Nonetheless, The Guardian, reported in 2017 that our medical data is, effectively, already out there:
“Although information is anonymised, data miners and brokers can build up detailed dossiers on individual patients by cross-referencing with other sources.”
And yet we are not experiencing the potential benefit this can offer.
An individual’s medical history, linked to their blood-kin for predictive analysis of hereditary and later-in-life illness, is a profound value-proposition.
So what’s taking so long?
Indeed, accessible digital medical records would require citizens to take more responsibility for their own health.
Daily input on one’s actual health—what you eat, when you exercise, how you actually feel—would enable AI to revel in a constant stream of new data and provide precision prognoses on you and wider society.
So the development of AI in healthcare isn’t a tech issue at all. Rather, it’s a cultural one?
In fact, she continues, emerging countries could lead the way as they may not have such outdated preferential systems that make technological innovation difficult to implement.
Optimists belief that once medical records become digitised, accessible and readily updated, then AI will be able to perform to a much higher level.
Panel guest Dodsworth’s D-Health is a pan-European consultancy. They are expert advisers in the commercial aspects of digital health, health tech and the life sciences.
Dodsworth points to Sweden where AI in healthcare is progressing fast. Min Doktor is a Malmö-based app that provides doctor-patient communications through voice, video, and text messaging. They already have a 100,000 users and a €22 million investment to grow the technology across borders. They even have a rival medical-consultation app, KRY.
The discerning benefit of these apps — and they are already prevalent in Asia (see ODOC in India and Sri Lanka) — is that once the AI processes the user’s medical data, the information is sent to an uber-like pool of doctors; some of whom will be specialists in fields that are unreachable at the consultation level of a GP meeting.
“40% of cases [on Min Doktor] are dealt with without 1–2–1 contact. It’s a superb example of AI working in healthcare: users benefit from the around the clock convenience, while the medical system has it’s workload eased”, affirms Dodsworth.
What About The NHS?
The third and final panelist is Declan Hadley, Digital Health Lead at Lancashire and Cumbria Change Program (NHS). The health of 1.7 million people in the north of England is under his remit, but he is late to the discussion as his train had been delayed after someone had a heart attack onboard.
“Truly, honestly”, says Hadley, “we’re not really using AI in healthcare. We have pioneers across the system, and there are interesting projects with plenty of promise, but there has not been any real transformation”.
Is that set to change in the next 5 years?
“For our region, our challenge is not around funding, it’s about resources i.e. people.”
“Brexit will affect this further. We [therefore] need tech [AI] to take mundanities out of the process”.
Towards an AI Of Global Medical Knowledge
This short panel discussion reflects the fragmented political picture and the need for a more unified approach to tech.
For certain, there are huge developments that need to happen around data and regulation before AI can securely support healthcare. But, to my mind, the limitlessness of AI’s potential in healthcare hasn’t been truly realised either.
What if a bonafide AI of healthcare could think outside the box of Western Biomedicine?
What if 10,000 years of medicine — Ayurvedic, Chinese, Folk and Shamanic — could instantly be accessed by AI to create a digital diagnosis that incorporates myriad medical systems?
Surely only then would we be embarking on new era of healthcare: a holistic health experience that is globally aware, culturally sensitive, but intelligently artificial.
Craig writes for Calcey Technologies, a boutique software product engineering agency with roots in the Silicon Valley, that lends its software development muscle to start-ups and scale-ups around the world. Cacley works with some of the largest bio-science and medical equipment companies in the world, assisting their efforts to build software tools for sales, marketing and analysis. Calcey’s client portfolio also includes global names such as PayPal and Stanford University, alongside numerous exciting startups, including Nutrifix (UK), Nelly.com (Sweden) and MyBudget (Australia). The team of 100+ engineers, based at its development centre in Sri Lanka, are looking to engage with more exciting product development opportunities.