PRACTICE CHANGING research recently published has demonstrated that AI systems can identify and grade prostate cancer tissue samples as accurately as pathologists. These findings are now being used as evidence that AI systems are ready for introduction into prostate cancer care as a complementary tool.
Study assessed results from a competition called PANDA. The three-month competition challenged AI experts to develop systems for accurately grading prostate cancer samples and garnered rapid responses. “Only 10 days into the competition, algorithms matching average pathologists were developed.” Stated Kimmo Kartasalo, Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden.
Currently, the diagnosis and treatment of prostate cancer can be challenging. Often different pathologists can interpret tissue samples differently, coming to variable conclusions. This can impact on treatment decisions as clinicians are working from uncertain information. The study authors believe AI technology holds the key to improving reproducibility, consistency and accuracy in diagnosis and treatment.
Previous study with AI has demonstrated the technology’s ability to indicate whether a tissue sample contains cancerous cells or not and estimate the amount of tumour tissue in a biopsy. However, AI accuracy often suffers when adapting to hospital settings in different countries after highly sensitive training.
Martin Eklund, Department of Medical Epidemiology, Karolinska Institutet, Stockholm, Sweden explained the value of the study’s findings, “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. The next step is controlled studies for evaluating how to best introduce AI systems in patient care.”