Researchers at Seoul National University Boramae Medical Center have published a study demonstrating the effectiveness of their newly developed XAI (explanatory artificial intelligence) in diagnosing prostate cancer.

XAI refers to artificial intelligence that provides a basis for judgment on its results so that doctors can understand. Unlike conventional AI, XAI has improved the reliability of results by resolving uncertainty in judgment. As a result, research is actively underway to apply the technology.

The team, led by Professor Jeong Hyeon, developed an AI algorithm to predict prostate cancer based on data from 2,843 people who visited the hospital from March 2009 to October 2019 and had a prostate biopsy. Professors Suh Jun-hyo at Seoul National University Hospital also participated in the study.

Afterward, the team randomly assigned 948 people to analyze the effectiveness of the AI system.

As a result, the overall prostate cancer prediction rate was about 87 percent. Notably, the clinically significant prostate cancer predictive rate rose as high as 95 percent. Therefore, the researchers determined that its AI algorithm could effectively diagnose prostate cancer, the hospital said.

“This study confirmed the possibility of effectively predicting prostate cancer using AI technology,” Professor Jeong said. “The XAI technology used in this study has the advantage that medical staff can easily understand the reason for the AI’s judgment.”

The team expects that the platform will have a high impact on the clarity of diagnosis when applied in clinical practice, Jeong added.