Data Science Investment Counter —Funding raised by UK data science companies in 2018.
£ 5.640 Million

Artificial intelligence beats doctors at diagnosing skin cancer

Artificial intelligence would be superior to humans in diagnosing pigmented skin lesions, according to a new study led by experts at the Medical University of Vienna.

Published in The Lancet Oncology journal on Tuesday, the study examined the results of tests where 511 physicians competed against 139 computer algorithms.

The AI was trained on a database of benign and malignant pigmented lesions counting 10.000 images created by a researchers’ team at the Department of Dermatology of the Medical University of Vienna in cooperation with the University of Queensland in Australia.

As part of the test, every participant had to diagnose 30 images randomly selected out of a 1511 test-set. While the best-scoring humans diagnosed correctly 18.8 out of 30 cases, the best machines scored much higher, with 25.4 correct diagnoses.

“Two thirds of all participating machines were better than humans; this result had been evident in similar trials during the past years”, said first-author of the study Philipp Tschandl.

However, according to the researcher, for how more precise than humans, the algorithms’ current abilities are not yet sufficient to replace human doctors altogether.

“The computer only analyses an optical snapshot and is really good at it”, Tschandl said. “In real life, however, the diagnosis is a complex task. Physicians usually examine the entire patient and not just single lesions.

“When humans make a diagnosis they also take additional information into account, such as the duration of the disease, whether the patient is at high or low risk, and the age of the patient, which was not provided in this study”, he explained.

While artificial intelligence in healthcare is not able to replace trained clinicians yet, it has already proven to be extremely useful for advancements in the medical field.

In January this year machine learning was used to translates brain signals directly into speech to improve the lives of people who cannot speak, and more recently an AI program by the University of Surrey was developed to help in predicting cancer symptoms.


Co-working space and blog dedicated to all things data science.

Subscribe to our newsletter