Story at a glance
- Researchers at UCLA Health found AI had a 17% greater accuracy rate of discovering the extent, or contour, of prostate cancer.
- Seven urologists and three radiologists analyzed 50 removed tumors, specifically for residual cancer cells.
- AI-assisted contours had an 84.7% discovery success rate, compared to that of 67.2% by doctors’ standardized tumor delineation methods.
(NewsNation) — Artificial intelligence could provide a clearer path to identifying the scope of prostate cancer.
Researchers at UCLA Health found AI had a 17% greater accuracy rate of discovering the extent, or contour, of prostate cancer. The study was published in The Journal of Urology.
Seven urologists and three radiologists analyzed 50 removed tumors, specifically for residual cancer cells.
AI-assisted contours had an 84.7% discovery success rate, compared to that of 67.2% by doctors’ standardized tumor delineation methods.
In doing so, AI helped more accurately define the extent of prostate cancer in males and simultaneously decreases the margin for error.
Doctor-based assessments of cancer sizes were greatly underestimated compared to AI-assisted analysis. The former had a negative margin rate — the absence of cancer cells around removed tissue — with an accuracy of 1.6% to 72.8% of AI.
“We saw the use of AI assistance made doctors both more accurate and more consistent, meaning doctors tended to agree more when using AI assistance,” Shyam Natarajan, senior author of the study said in the release.