PITTSBURGH – Human error can be charming in an endearing way, but no one appreciates mistakes when it comes to such a serious topic as cancer. In that regard, researchers at the University of Pittsburgh developed a new artificial intelligence program with the most accurate record to date when it comes to recognizing prostate cancer.
“Humans are good at recognizing abnormalities, but they have their own biases or past experience,” says lead author Dr. Rajiv Dhir, chief of pathology and vice president of pathology at UPMC Shadyside and professor of biomedical informatics at UPitt, in a release. “The machines are separate from the whole story. There is definitely an element of standardization of care. “
What separates this AI from the rest of the robotic package? Dr. Dhir and his team “fed” images from their program of more than a million parts of slides extracted from biopsies of prostate cancer patients. The AI program was then tested on 1,600 different slide images collected from 100 suspected prostate cancer patients.
The AI performed incredibly well on that test. Results show 98% sensitivity and 97% specificity for finding and identifying prostate cancer. Those statistics are much higher than the scores recorded by previous cancer detection algorithms.
Plus, this is the first algorithm that does much more than just detect cancer. This program also scores well in categories including tumor size and classification, and evaluation of surrounding nerve invasion by cancer cells.
CLICK HERE TO SUBSCRIBE TO OUR NEWSLETTER AND GET THE LATEST STUDY FROM STUDYFINDS.ORG BY EMAIL!
The AI even detects cancer on six slides that were slid by a human pathologist.
While all of this is very promising, the study authors caution that the AI program is not yet ready to completely replace human doctors. For example, regarding those six slides that went unnoticed by a human physician, that pathologist might have seen enough evidence to warrant a cancer diagnosis. before reaching those particular slides.
Still, Dr. Dhir says the algorithm, at least, can serve as great insurance.
“Algorithms like this are especially helpful in injuries that are atypical,” says Dr. Dhir. “A non-specialized person may not be able to make the correct evaluation. That’s a great advantage of this type of system. “
Of course, this project only focuses on prostate cancer. While a completely new algorithm would have to be trained for each type of cancer, the research team is optimistic, and its results can be recreated with other variations of the cancer.
The study is published in The Lancet Digital Health.
Do you like studies? Follow us on Facebook!
