An artificial intelligence (AI) system has demonstrated a high level of accuracy in diagnosing prostate cancer in tissue samples.
The study, which was led by Karolinska Institutet in Sweden, found the AI system, called OncoWatch, was comparable with the abilities of 23 leading uropathologists in determining a sample’s Gleason score, a key prognostic marker for prostate cancer. The results have been published in The Lancet Oncology.
Karolinska Institutet associate professor Martin Eklund said: “Our AI tool has the potential to reduce the workload of uropathologists, allowing them to focus on the most difficult cases and at the same time act as a safety net to improve quality.
“It also has the potential to speed up diagnostics and reduce costs for healthcare services.”
To train and test the system, the researchers took high-resolution images of more than 8,000 biopsies taken from 1,200 Swedish men aged 50-69. Around 6,600 of the samples were used to train the AI system to spot the difference between biopsies with and without cancer, while the remaining samples were used to test the system.
As well as a near-perfect ability to assign Gleason scores, the AI system was on par with the experts in determining whether a sample contained cancer or not as well as estimating the length of the biopsied tumour.
Eklund said: “The idea is not that AI should replace the human involvement, but rather act as a safety net to ensure that pathologists don’t miss some cancers, and assist in standardization of grading.”
A multicentre study spanning nine European countries is now building upon the initial research, aiming to train the AI system to recognise cancer in biopsies taken from different laboratories, with different types of digital scanners and with very rare growth patterns. Researchers will also look into how the AI model could be implemented in Sweden’s healthcare system.
The project received network and financial support from the European Institute of Technology’s Health division, EIT Health.