Recent findings by researchers at Radboud University medical centre could indicate a potential use for artificial intelligence systems in predicting, detecting and diagnosing the spread of breast cancer.

Wednesday 13 December 2017      Latest research
""

In a new study published in the Journal of the American Medical Association, researchers at Radboud University medical centre in the Netherlands evaluated the use of computer algorithms – which had been developed as part of a competition – to interpret lymph node tissue slides, an assessment normally carried out by pathologists.

The team found that the very best algorithms were able to spot lymph node metastases (secondary tumour cells) just as well as a pathologist when there were no time constraints. In addition, some of the algorithms were better at diagnosing metastases than a panel of 11 pathologists in a simulation exercise designed to mimic pathologists’ normal working routine.

The findings could indicate a potential use for artificial intelligence systems in predicting, detecting and diagnosing the spread of breast cancer. 

Katherine Woods, Senior Research Communications Manager at Breast Cancer Now, said: 

“The prospect of using computer intelligence to more accurately predict and detect the spread of breast cancer is exciting, but clinical testing is needed to assess whether this might be feasible and effective in patients. 

“When breast cancer spreads to another part of the body, it sadly becomes incurable. We urgently need to find new ways to prevent the disease spreading, to detect it earlier and to develop new treatments to improve outcomes for these women. 

“Accurately determining the staging of a patient’s breast cancer is absolutely key to informing and planning their treatment in the clinic. Currently, the assessment of the spread of an individual’s breast cancer is done by analysing their lymph nodes using microscope slides, an essential task performed by expert pathologists. While it’s fascinating that computer algorithms could enhance our ability to predict and spot metastasis, further studies are now needed to test whether this would work in the clinic.”

For more information visit The Times website.