Issue of the day: Artificial intelligence in healthcare

NEW research has suggested artificial intelligence can diagnose breast cancer more accurately than trained doctors. AI seems set to revolutionise healthcare.

AI in healthcare?

Complex algorithms and software are used to emulate human behaviour in the analysis of complex medical data to diagnose and treat patients.

It’s already in use?

Microsoft UK has reported a rise in the use of AI technologies in healthcare. Its survey in October, found 46 per cent of healthcare leaders reported their organisation used the technology in some form, up 8 per cent on 2018.

So what are examples of current use?

Wide-ranging, from identifying patients most likely to miss appointments and giving them reminder phone calls, to the use of robots to analyse data from pre-op medical records to guide a surgeon’s instrument during surgery.

It’s a developing field?

The UK government pledged £250 million into AI health tech in August.

The latest findings?

An algorithm developed by researchers, at Northwestern University in Chicago and Imperial College London, working alongside Google Health, slashed the number of missed breast cancer cases from one in 10 to one in 37.

How?

By noticing tumours undetected by radiologists in a study of nearly 26,000 women who had mammograms at NHS hospitals, as well as just over 3,000 US cases. When the algorithm assessed the scans, 2.7 per cent were missed, in comparison to the 9.4 per cent missed by a panel of six radiologists.

It could become routine?

Researchers hope the breast cancer detection system will become as common as simply using “spell check”, reducing life-threatening delays from “false negatives”.

AI has surpassed medics before?

In 2018, an international study by researchers from France, Germany and the US used machines trained to detect signs of skin cancer to compare the results against 58 dermatologists.

The machines correctly diagnosed malignant cases in 95 per cent of cases – the dermatologists diagnosed 87 per cent.

Negatives?

The reality of integrating sophisticated AI systems into day-to-day healthcare would require intensive training and there are concerns about the abilities of already ageing hospital

IT systems. Safeguarding data would also be key, as well as an awareness of a need not to become totally reliant on AI to the abandonment of human intuition.

There will always be a place for people?

Co-author of the latest study, Dr Mozziyar Etemadi, an assistant professor of anaesthesiology at Northwestern, said more research is required, adding that the “ultimate goal will be to find the best way to combine the two – the magic of the human brain isn’t going anywhere any time soon”.