Artificial intelligence already trumps humans in screening malignant tumours from tissue scans

Professor of Systems Biology Sampsa Hautaniemi believes that artificial intelligence will be a great help in clinical practice, as long as the laws that regulate its use are drawn up wisely.

Originally published in Finnish in issue 5/2023 of the Yliopisto magazine.

According to Professor of Systems Biology Sampsa Hautaniemi, artificial intelligence already interprets mammograms better than humans.

Hautaniemi points to a British–American study published in 2020 where artificial intelligence looked for evidence of malignant tumours in nearly 30,000 mammograms. It both was more accurate at identifying tumours and made fewer false-positive diagnoses compared to humans.
Using artificial intelligence to screen the scans first reduced oncologists’ workload by almost 90%.

“And it's worth noting that this result is already a few years old. Artificial intelligence is developing at a dizzying rate, and has only progressed in accuracy,” Hautaniemi emphasises.

In another case from last year, an AI solution identified with 88% certainty cancer cells from images taken of pathology specimens.

Use of artificial intelligence becoming increasingly common

Currently, at least 71 AI-augmented tools for clinical use have already been approved in the United States. 

Hautaniemi estimates that the utilisation of artificial intelligence will become increasingly common in Finnish healthcare. 

“Artificial intelligence is heavily involved in research. We have excellent opportunities to utilise it also in clinical practice.”

Of course, there is a risk that overly tight regulation hinders the use AI for patients' benefit.

“We have to take care to not make the adoption of artificial intelligence as difficult as register-based research is due to the Act on the Secondary Use of Health and Social Data. Excessive caution in permit-related matters can make effective methods unreasonably cumbersome for patient care.”

Professor Sampsa Hautaniemi spoke on the topic AI in Cancer Research and Oncology at an AI symposium organised by Biocenter Finland.
Yliopisto is a scholarly magazine published by the University of Helsinki committed to observing the journalistic guidelines of the Council for Mass Media in Finland.