肝移植受者他克莫司血药浓度早期预测方案及评估

Establishment and evaluation of early prediction of tacrolimus concentration in liver transplantation recipients

目的:建立人工神经网络用于估算他克莫司血药浓度。方法:收集36例肝移植受者口服他克莫司的150份稳态全血浓度数据,采用遗传算法配合动量法优化网络参数,建立人工神经网络。结果:人工神经网络平均预测误差(MPE)与平均绝对误差(MAE)分别为0.05±3.01ng·mL-1和2.09±2.12ng·mL-1,78.3%血药浓度数据绝对预测误差≤3.0ng·mL-1。人工神经网络准确性及精密度优于多元线性回归。结论:人工神经网络预测可用于预测他克莫司血药浓度。

OBJECTIVE To establish a prediction method for tacrolimus concentration in liver transplantation recipients by artificial intelligence. METttODS 150 tacrolimus concentration samples from 36 Chinese liver transplantation recipients were collected. Artificial neural network （ANN） was established after that network parameters were optimized by using momentum method combined with genetic algorithm. Furthermore,the performance of ANN was compared with that of multiple linear re gression （MLR）. RESULTS The mean prediction error and mean absolute prediction error of ANN were 0. 05 ＋ 3.01 ng,mL-1 and 2. 09 ＋ 2. 12 ng. mL-~ , respectively. The absolute prediction error of 78. 3~ of testing data sets was less than 3. 0 ng. mL- t. Accuracy and precision of ANN was superior to that of MLR. CONCLUSION ANN was suitable to predict tacrolimus concentration.