Prediction of Improving Outcomes in Upper Urinary Tract Tumor Patients Treated with Enhanced Recovery After Surgery in The Artificial Neural Network System

To explore the clinic value of ERAS (enhanced recovery after surgery) in postoperative recovery (laparoscopic surgery) for curative treatment on UUTC (upper urinary tract tumor), and predict the outcomes with an artificial neural network (ANN) system. After laparoscopic surgery for curative treatment on UUTC, patients were randomly assigned to the two groups ERAS and control group, the perioperative and postoperative short-term outcomes were compared. Clinical characteristics, handing methods, and clinical outcomes data were analyzed using an adapted ANN model. The strength of the ANN prediction was measured between -1 and 1 with -1 representing bad outcome and 1 representing good outcome. In ERAS group, postoperative exhaust and defecation time, as well as the time of beginning taking food or off bed activities for patients were shorter obviously than control. Moreover, the length of hospital stay and post-operative pain in ERAS patients were obviously less than in control group. The ANN applied here was able to accurately predict all patients outcomes.