Study: AI Not Much Better Than Humans in Medical Diagnosis

Despite all the hype surrounding AI outperforming humans in medical diagnosis, it turns out they’re about par, at best.

According to The Guardian, a recent study found that artificial intelligence and human experts were equally skilled at interpreting medical images.

Prof Alastair Denniston, at the University Hospitals Birmingham NHS foundation trust and a co-author of the study, said the results were encouraging but the study was a reality check for some of the hype about AI.

Dr Xiaoxuan Liu, the lead author of the study and from the same NHS trust, agreed. “There are a lot of headlines about AI outperforming humans, but our message is that it can at best be equivalent,” she said.

The study was reportedly the first comprehensive review of the issue. Previous efforts were said to be based on a small number of studies and, some note, “poor-quality research.”

“This excellent review demonstrates that the massive hype over AI in medicine obscures the lamentable quality of almost all evaluation studies,” said Prof David Spiegelhalter, chair of the Winton centre for risk and evidence communication at the University of Cambridge.

That said, the researchers themselves say that artificial intelligence has enormous potential in medicine. Denniston, for instance, said that deep learning systems could be a diagnostic tool that could even help the backlog of scans and images. Liu meanwhile believes that they could be quite valuable for places without experts to interpret images.

Photo: iStock