The invention discloses an acute kidney injury incidence probability prediction system based on intensive care detection items. The acute kidney injury incidence probability prediction system obtainsdata of multiple detection items of a patient within 24 hours and demographic information related to gender, age and weight from a hospital database, marks the data of each item in chronological orderfrom far to near, performs front-to-back arrangement, completes extraction of related features according to the classified demographic data, serum creatinine, systolic blood pressure, urine volume,blood gas analysis, body temperature heart rate, and medication information, wherein the related features relate to age, gender, the change value, the mean value, the standard deviation, the minimum value, the maximum value, the most recent value and other digital features, of the related detection items within 24 hours, and the like: the data of total 38 detection items, and combined with the artificial intelligence machine learning algorithm, predicts the incidence of acute kidney injury of the patient 24 hours in advance, for creating conditions for early clinical intervention.