"A research group led by Professor Hidenori Kage of the University of Tokyo Graduate School of Medicine and Assistant Professor Hiroaki Ikushima of the University of Tokyo Hospital used clinical and cancer gene information from pancreatic cancer patients across Japan, and based on the clinical information of 2,220 pancreatic cancer patients who underwent cancer gene panel testing, developed a machine learning model that individually estimates the probability of detecting cancer genes leaking into the blood from pancreatic cancer cells through a blood test."