The stage of development of the use of AI in medicine has been compared to that of computer-driven financial trading in the 2000s. In 2005, reflects a recent article in the Financial Times, computer-driven hedge funds using algorithms to follow the markets and take buy and sell decisions, termed quantamental investors, managed assets totalling less than $50 billion.

Hedge fund data and market research company Hedge Fund Research, says that had leapt to $270 billion by 2014. In October 2018, MarketWatch reported the total was approaching $1 trillion – a level surely now comfortably exceeded. High frequency trading executed by algorithms, where hundreds and thousands of short term trades, designed to capture incremental gains that accumulate, account for more than half of daily US equity trading turnover.

AI is now increasingly making its way into practical use in modern healthcare systems. Sensors for remote monitoring of vital signs with medical-grade accuracy have started to be approved by the FDA. They are already being used to monitor patients suffering from conditions and symptoms that range from type 2 diabetes to sepsis, respiratory rate, resting heart rate, skin temperature and even body position.

That data is run through machine learning algorithms that can spot the tell-tale signs a particular individual needs more attention. That is helping with earlier intervention and a reduction in emergency hospital visits.

The data can then be fed into computers that use machine learning to spot people who might need more attention, allowing for early intervention and the avoidance of emergency hospital visits.

Richard Zane, chief innovation officer at UCHealth, which runs 12 Colorado hospitals, describes his company’s use of new AI-enabled monitors patients are being fitted with and the efficiencies resulting with:

“Now, instead of one nurse monitoring eight people on a ward, she can monitor 8,000 people at home.”

With healthcare costs one of the most significant burdens that individuals and governments struggle with across the globe, a problem that is only getting worse, the potential cost savings promised by increased use of AI is attractive. That’s reflected in the amount of private equity and venture capital going into AI start-ups focused on healthcare – $4 billion invested across 367 young companies last year, says CB Insights.

Healthtech attracting investment ranges from algorithms to improve hospital administration and logistics to systems able to read scans such as mammograms and surgical assistants that offer real-time advice during operations. Accenture forecasts $150 billion in annual healthcare savings by 2026, that will result from the use of AI in the USA alone. In the UK, the NHS has set aside a significant budget to trial new technologies, many of which are powered by AI.

However, despite the increasing use of machine learning algorithms in healthcare and medicine, it can be expected that progress towards the kind of influence technology has in financial markets trading will prove to be slower. With the consequences of mistakes far higher, and irreversible, algorithms used in healthcare need to prove an almost faultless degree of accuracy before regulatory approval.

Perhaps a better comparison is with the development of driverless vehicles technology by tech companies such as Google’s Waymo and Uber. While they have now developed the technology to the point it can handle 99% of situations encountered on the roads, it is not yet completely flawless. And to win both regulatory approval and public trust, it has to be.

But AI used in healthcare is perhaps more like kind of autopilot functionalities that allow modern cars such as Tesla models to drive themselves on the motorway or park in narrow spaces. But the driver still has an active role and constantly approves the AI’s decisions, or takes control to override them when deemed necessary.

Bill Evans, managing director of Rock Health Venture fund, which has invested in a number of AI technologies for healthcare explains:

“Our strategy is to enable the frontal cortex of doctors rather than lobotomising them.”

So AI won’t be replacing doctors and nurses anytime soon. But it is starting to make a real impact when it comes to helping them do their jobs better and more efficiently.