肝炎患者血浆蛋白分布信息熵及其人工神经网络识别模型

Information-Entropy of Blood Plasma of Hepatitis and Recognition Model Based on Artificial Neural Network

研究了肝炎患者的血浆蛋白分布信息熵值，发现慢性肝炎患者的信息熵较之正常人大。其中慢性活动性肝炎（CAH）又大于慢性持续性肝炎（CPH）。这说明肝炎给血浆蛋白的成分造成了紊乱，而漫性活动性肝炎造成的紊乱更大一些。利用这些指标建立了以人工神经网络为基础的识别模型，它在经过训练后能对新的作本进行联想式的识别，在研究中共识别率达到了100％。

In this paper the in formation-entropy (I) of protein distribution in blood plasma of hepatitis was observed- The human electrophoresis spectrum of blood plasrna showed that the value of I in control group was higher than the normal, whereas the group CAH was the highest with respect to CPH. This may be expressed as hepatitis, producing disturbance in blood plasma and CAH could affect more than CPH. Us-ing I as an index, we propose a recognition model based on artificial neural network, after some learning cy-cles, it can identify new samples.