Department of Surgery, Rutgers New Jersey Medical School, Newark, New Jersey.
Clinical Outcomes Research Group, CORG LLC, Grantham, New Hampshire.
J Surg Res. 2021 Mar;259:372-378. doi: 10.1016/j.jss.2020.09.021. Epub 2020 Oct 21.
Inguinal hernia repair is one of the most commonly performed surgical procedures. We developed and validated an artificial neural network (ANN) model for the prediction of surgical outcomes and the analysis of risk factors for inguinal hernia repair.
The American College of Surgeons National Surgical Quality Improvement Program was used to find patients who underwent inguinal hernia repair. Using logistic regression and ANN models, we evaluated morbidity, readmission, and mortality using the area under the receiver operating characteristic curves, true-positive rate, true-negative rate, false-positive rate, and false-negative rates.
There was no significant difference in the power of the ANN and logistic regression for predicting mortality, readmission, and all morbidities after inguinal hernia repair. Risk factors for morbidity, readmission, and mortality outcomes identified using ANN were consistent with logistic regression analysis.
ANNs perform comparably to logistic regression models in the prediction of outcomes after inguinal hernia repair. ANNs may be a useful tool in risk factor analysis of hernia surgery and clinical applications.
腹股沟疝修补术是最常施行的外科手术之一。我们开发并验证了一个人工神经网络（ANN）模型，用于预测手术结果和分析腹股沟疝修补术的风险因素。
使用美国外科医师学院国家外科质量改进计划（American College of Surgeons National Surgical Quality Improvement Program）来寻找接受腹股沟疝修补术的患者。使用逻辑回归和 ANN 模型，我们使用接收者操作特征曲线下面积、真阳性率、真阴性率、假阳性率和假阴性率来评估发病率、再入院率和死亡率。
在预测腹股沟疝修补术后死亡率、再入院率和所有发病率方面，ANN 和逻辑回归的效能没有显著差异。使用 ANN 确定的发病率、再入院率和死亡率结果的风险因素与逻辑回归分析一致。
ANN 在预测腹股沟疝修补术后的结果方面与逻辑回归模型相当。ANN 可能是疝手术风险因素分析和临床应用的有用工具。
