Assessment of AI Prediction Models in Prediction of Acute Kidney Injury in Critical Patients

Baseline characteristics, including demographic information, comorbidities, vital signs, laboratory results, medical interventions, disease severity scores, etc. were carefully reviewed and collected. The definitions of comorbidities including congestive heart failure, peptic ulcer disease, myocardial infarction, peripheral vascular disease, diabetes, dementia, chronic pulmonary disease, rheumatic disease, cerebrovascular disease, cancer, paraplegia, liver disease, renal disease, and acquired immunedeficiency syndrome. Severe organ failure due to ineffective immune response to infection was identified as sepsis. During the first 24 h when the patient was admitted to the ICU, the average values of the patient's vital signs (heart rate, mean arterial pressure, respiration rate, body temperature, and SpO2) were measured,, and the highest value of the biochemical laboratory tests (hematocrit, hemoglobin, platelets, white blood cell, blood urea nitrogen, international normalized ratio, Scr,