 |
| Diagram showing a transperineal prostate biopsy – Credit Cancer Research UK via Wikipedia |
Led by researchers at the Karolinska Institutet in Sweden, a group of more than 1000 AI experts engaged in a competition to test AI algorithms for accurately grading prostate cancer.
The study, dubbed PANDA, was designed to accelerate innovation for solving the problem of inconsistency in diagnosing prostate cancer, according to Martin Eklund, an associate professor at the Department of Medical Epidemiology and Biostatistics at the Karolinska Institute and senior author of the study.
“The results from PANDA show for the first time that AI systems can produce an equally accurate diagnosis and grading of prostate cancer in an international setting as human pathologists," Eklund said in a press release. "The next step is controlled studies for evaluating how to best introduce AI systems in patient care."
One of the problems with diagnosing prostate cancer is that different pathologists can reach different conclusions even when based the same tissue samples, leaving room for uncertainty in making treatment decisions, according to information in the press release. Using AI technology the researchers hope to reduce inconsistency and increase reproducibility in the assessment of tissue samples.
The study involved colleagues at Radboud University Medical Center in the Netherlands, Google Health in the US and researchers at the University of Turku, Finland. Using a variety of AI algorithms 1,290 developers evaluated 10,616 digitized prostate biopsies. In the end the algorithms achieved agreements with expert uropathologists 86 percent of the time. That is deemed good enough to move the algorithms into clinical testing in prospective clinical trials.
The goal is not to replace human experts, but to function as a safety net to avoid missing cancer cases and improve consistency of diagnosis.
Sources: Karolinska Institute press release and Nature Medicine
“The results from PANDA show for the first time that AI systems can produce an equally accurate diagnosis and grading of prostate cancer in an international setting as human pathologists," Eklund said in a press release. "The next step is controlled studies for evaluating how to best introduce AI systems in patient care."
One of the problems with diagnosing prostate cancer is that different pathologists can reach different conclusions even when based the same tissue samples, leaving room for uncertainty in making treatment decisions, according to information in the press release. Using AI technology the researchers hope to reduce inconsistency and increase reproducibility in the assessment of tissue samples.
The study involved colleagues at Radboud University Medical Center in the Netherlands, Google Health in the US and researchers at the University of Turku, Finland. Using a variety of AI algorithms 1,290 developers evaluated 10,616 digitized prostate biopsies. In the end the algorithms achieved agreements with expert uropathologists 86 percent of the time. That is deemed good enough to move the algorithms into clinical testing in prospective clinical trials.
The goal is not to replace human experts, but to function as a safety net to avoid missing cancer cases and improve consistency of diagnosis.
Sources: Karolinska Institute press release and Nature Medicine
No comments:
Post a Comment