Disease diagnosis by AI similar to that by doctors

Artificial intelligence (AI) is just as good at diagnosing a disease based on a medical image as healthcare professionals, says the first systematic review and meta-analysis on the topic.

Researchers at University Hospitals Birmingham NHS Foundation Trust in the United Kingdom, who set about examining whether AI can diagnose illnesses as effectively as healthcare professionals from medical images, found that AI could correctly diagnose disease in 87 percent of the cases, whereas detection by healthcare professionals yielded an 86 percent accuracy rate. Furthermore, when it came to assessing the AI’s diagnostic performance compared with that of healthcare professionals, the researchers looked at two outcomes: specificity and sensitivity.

‘Sensitivity’ defines the probability that a diagnostic tool gets a positive result in people who have the disease. ‘Specificity’ refers to the accuracy of the diagnostic test, which complements the sensitivity measure. The researchers found that the specificity for deep learning algorithms was 93 percent, compared with humans at 91 percent.

Though the sample cases we studied were few, they were of high quality and within that high quality of studies we found that deep learning could indeed detect diseases ranging from cancers to eye diseases as accurately as health professionals. But it’s important to note that AI did not substantially outperform human diagnosis, said the team behind the research.