AI in healthcare: can robots save the day?


Date: 24 August 2018

AI in healthcare: can robots save the day?Unless you've been under a rock for the past few months, you'll know that the predicament the NHS is in is worsening. The medical sector was promised £20bn by the Prime Minister a little over a month ago - but news soon followed that this would not be given until 2019, and is subject to a number of conditions.

Some MPs are arguing that artificial intelligence (AI) is the key to improving the situation long-term. Already robots are being tried and tested in surgery and diagnosis. The PRECEYES model, for example, has recently been shown to improve the efficacy of routing eye surgeries.

Meanwhile, researchers at the John Radcliffe Hospital in Oxford have been busy on an intelligent diagnostic system that can identify and diagnose heart disease with greater accuracy than a human doctor.

Embracing robots in healthcare?

Robots, then, can bring clear benefits to the NHS. Research and development is being carried out by businesses large and small, with new projects being published on a regular basis.

From diagnostic tools and surgery enhancers to innovative apps that enable patients to track their health and feed back to their GP, new tech is paving the way for improved efficiency and the freeing up of healthcare professionals' time.

But there are concerns about the use of AI and the limitations of machines. One example is the use of neural network technology. Able to handle vast amounts of data, this system - which is designed to mimic the working of the human brain - can identify patterns and adapt in response to what it learns.

Due to the vast amounts of data being processed, however, the machines soon become so complex that their creators can't identify how an answer has been reached. This could lead to, say, health insurance being refused on the basis of an algorithm no-one understands, or an inexplicable dose of a drug being recommended.

Machines are also subject to inadvertent human biases that may be programmed into them by their developers. Medical research, for example is largely dominated by white men, whose subjects are also predominantly white men.

A drug, therefore, that has only proven to be notably effective on men could be recommended to both genders due to this bias - and the machine would not be able to compensate for it.

Harnessing the power of AI

The solution, then, is to stick - at least in the present - to AI that simply extends the abilities of humans, without cutting them out of the equation altogether. Developers are also being encouraged to put adequate safeguarding measures into place during the refinement of their technologies.

This, of course, means extensive research and development, for which R&D tax relief is available from the government. All businesses engaged in developing AI for the healthcare industry - from the largest corporate to the smallest agile start-up - can access this, provided their project meets eligibility criteria.

If you think your activity may qualify for tax credits, contact a tax adviser today for advice on how to claim.

Sponsored post. Copyright © 2018 R&D Tax Solutions