Chinese researchers have developed AI-driven system that can diagnose cancer. A special type of prostate cancer can be diagnosed using the artificial intelligence system.
The cancer diagnosis is the single most difficult factor of the treatment. The self-learning artificial intelligence (AI) system can identify cancerous prostate samples. The research report suggests that the accuracy of this system is as high as any pathological test. The system can eliminate the shortage of trained pathologists in the world.
The scientists from China are positioning the system as something that can be used to streamline and eliminate the variation in cancer diagnosis process. The further advancements in the AI algorithm can lead to automated or partially-automated prostate cancer diagnosis.
The team of researchers from China’s Nanjing University presented the system at the European Association of Urology Conference in Copenhagen. The system presented by researchers shows similar levels of accuracy as a human pathologist.
The lead scientist, Hongqian Guo said, “This is not going to replace a human pathologist. We still need an experienced pathologist to take responsibility for the final diagnosis. What it will do is help pathologists make better, faster diagnosis, as well as eliminating the day-to-day variation in judgment which can creep into human evaluations.”
The researchers have fed the data of 918 prostate whole mount pathology samples from 283 patients in the system. The system was run through the analysis, which the algorithm is gradually learning and improving the diagnosis.
Prostate cancer is one of the most common male cancer. Over 1.1 million cases are diagnosed each year. To confirm the diagnosis, a biopsy sample is required. The sample is later examined by a pathologist. The AI-system capable of taking care of diagnosis process.
The research results show an accuracy of 99.38%. Which is as high as a human pathologist. Guo further added, “The system was programmed to learn and gradually improve how it interpreted the samples. Our result shows that the diagnosis the AI reported was at a level comparable to that of a pathologist. Furthermore, it could accurately classify the malignant levels of prostate cancer.”