神经网络方法在胃收缩识别中的应用

胃收缩运动在消化过程中起着重要作用．传统的测量胃运动的方法是侵 袭性的．本文提出一种运用人工神经网络由体表胃电图（ｅｌｅｃｔｒｏｇａｓｔｒｏｇｒａｍ） 无损识别胃收缩运动的方法．以５个受试者的胃电图作为训练集，另５个受试 者的胃电图作为测试集．以同时检测的与每段胃电图对应的胃腔内压力记录 作为评价标准，运用经过优化的具有单个隐层的反向传播神经网络，以分段 后的每段胃电图的时－频表征作为网络的输入，实验结果表明：识别胃运动静 止期的准确度达９０呢，识别收缩运动期的准确度达９４％．

Application of neural network approach in recognition of gastric contractions

Gastric contractions play an important role in the digestive process of the stomach. The established method for the measurement of gastric contractions is invasive by intubating a manometric probe in the stomach. A noninvasive method is introduced in this paper for the noninvasive identification of gastric contractions from the surface electrogastrogram using an artificial neural network. The electrogastrogram (EGG) was measured by placing surface electrodes on the abdominal skin over the stomach in an overnight study in ten subjects. Gastric contractions were simultaneous monitored using an intraluminal manometric probe. The exponential distribution method was used for the time-frequency representation of the EGG. The spectral domain energy distribution data of the EGG were used as the input of the neural network. Back-propagation neural networks were designed. Experiments were conducted for the optimization of the network structure and parameters. Using the EGG data in five subjects as the training set and the EGG data in another five subjects as the testing set ,an accuracy of 90 % for the identification of gastric contractions from the EGG was achieved using a three-layer back-propagation neural network.