Artificial intelligence will bring NHS patients a greater quality of care by better diagnosing medical conditions and personalising medicine, according to the chairman of NHS England.
Prof Sir Malcolm Grant said the health service would benefit hugely from the use of machine learning and robots, suggesting that if such technology could outperform humans, it would be daft not to use it.
He acknowledged that the subject was “fraught with ethical issues” but suggested that the medical profession needed to be “more focused” in the way it used treatments.
He said diagnoses could only be strengthened with the use of computers to scan the wealth of information available through patients’ records in order to identify similarities.
“I do believe that artificial intelligence and machine learning has the capacity to hugely enhance our ability to diagnose illnesses and to understand how to treat them better,” he said.
Artificial intelligence will bring NHS patients a greater quality of care by better diagnosing medical conditions
“This has an enormous potency to personalise medicine and to get us away from current practice where we tend to use one set of pharmaceutical products to benefit 20 or 30 per cent of the population.”
Increasing numbers of robots are being used in the NHS, taking jobs in some hospitals such as dispensing drugs and ferrying linen and food. But technological advances mean that they are also starting to play a much greater role in medicine, this year assisting surgery for the first time.
In February, Health Secretary Jeremy Hunt said people would be diagnosed at home by a computer within two years.
Sir Malcolm told BBC Radio 4’s Inside Science programme that health care was one of the greatest challenges facing developed societies.
He said that gaps in medical profiles that were due to missing information or a lack of understanding could be filled due to the “intuitive nature” of technology.
“By comparing this patient’s profile with of hundreds of thousands of other patients we can start to pick out what it might be that is causing the disorder is that we can see,” he said.
Artificial intelligence could be used to spot patterns and use each individual’s genomic structure
“For me, the overriding thing is that if a machine can outperform a human doing a task such as diagnosis, given the complexity of modern medicine, the complexities of modern science, why wouldn’t we use it? Why wouldn’t we use it to inform the human and to bring to our patients the greater quality and greater safety of care that I think AI really can achieve?”
Sir Malcolm, who last week took part in a panel discussion called Socioeconomic Impacts of Machine Learning at the Royal Society’s Machine Learning conference, said artificial intelligence could spot patterns and use each individual’s genomic structure to understand what would work best for them.
He told the conference that around 50 per cent of morbidity was due to genetics, around 10 per cent was healthcare and 40 per cent largely down to lifestyle factors.
"We know that much of the cost of healthcare is down to admissions and treatment which were avoidable," he said.
"That is one major area where I am sure that machine learning and other technologies can help us enormously in not just reducing the cost of healthcare but also the burden of disease in modern society."
He acknowledged that a critical component of data usage within the NHS was privacy and data protection.
