Patients may develop postoperative complications after surgeries. Postoperative complications not only affect patients’ recovery, delay the timing of discharge, serious complications may also endanger the health and cause death. Lots of factors are associated with postoperative complications. Preoperative health status and intraoperative physiological changes both may affect the occurrence of postoperative complications. The preoperative health status consists of the patient’s history of systemic disease (diabetes, hypertension, etc.) and the test results before surgery (blood electrolytes, liver function tests, etc.). The intraoperative physiological changes include anesthetic drugs or surgical stimulation induced hemodynamic fluctuations (intraoperative blood pressure abnormalities, arrhythmias, bleeding and blood transfusion, etc.). Some researches have revealed the incidences and risk factors of some postoperative complications. However, each patient’s condition is different so he may have different probability to develop different postoperative complications. Clinically there are no definite rules to access the risks and probabilities of postoperative complications for an individual patient. In recent years, artificial neural networks (ANN) have been used in many fields. In 2003 Das and his colleagues developed the ANN model to predict the prognosis of patients with lower gastrointestinal bleeding. In 2008 we developed the ANN model to identify the risk of hypotension for patients undergoing general anesthesia and the original paper has been published in Medical Decision Making journal. In this study, we will use anesthesia information system to collect patients’ health status parameters preoperatively, and then collect hemodynamic and surgical related parameters during operation. After surgeries we follow up if the patients develop postoperative complications. Using the collected data we will develop ANN postoperative complication prewarning system. The system can help to predict the risk of postoperative complications for each patient. The medical personnel can be alert and may take some strategies to prevent the postoperative complications.
|Effective start/end date||8/1/12 → 7/31/13|
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.