Artificial intelligence system and traditional infusion of propofol in thoracic and abdominal surgery: A prospective single-blind randomized controlled study

Objective To explore the effectiveness and superiority of artificial intelligence Propofol TCI (Target controlled Infusion) system based on Bispectral Index (BIS) feedback in the anesthesia process of thoracic and abdominal surgery patients under general anesthesia.

Methods A total of 140 patients scheduled by thoracic or abdominal surgery were randomly divided into the artificial intelligence group and control group. All patients were given intravenous anesthesia with propofol and remifentanil. The target BIS value of sedation was adjusted to 50 during the operation. The two control systems were evaluated by comparing the percentage of time that the patient's BIS was maintained between 40 and 60, overall GS score, drug consumption, and hemodynamic parameters.

Results During anesthesia, the percentage of time BIS value between 40 and 60 in the artificial intelligence group was higher than that in control group (P ＜ 0.001). The GS value (overall score) in the AI group was lower than that in control group (P ＜ 0.001). The number of automatic adjustment of propofol in the artificial intelligence group and control group were (29.37 ± 11.20) times/h and (3.42 ± 1.52) times/h, respectively.

Conclusions The artificial intelligence propofol infusion system is better than the control group in maintaining the ideal sedation depth, and reduces the workload of the anesthesiologist.