血清蛋白质质谱模型在大肠癌诊断中的应用

目的 建立蛋白质芯片技术检测血清蛋白质质谱的方法,探讨基于人工神经网络的血清蛋白质质谱模型在大肠癌诊断中的应用价值。方法 应用表面增强激光解吸电离飞行时间质谱仪(SELDI-TOF-MS),测定了147例血清标本(其中大肠癌55例,健康人92例)的蛋白质质谱,用随机抽取的87例标本(大肠癌32例,健康人55例)作为训练组,进行训练与交叉验证,将筛选出来的5910,8930,4476和8817的4个质荷比峰作为输入,建立人工神经网络预测模型。并用另外测试组(大肠癌23例,健康人37例)的血清标本盲法验证该模型。结果 利用从训练组得出的基于人工神经网络的血清蛋白质质谱模型,对测试组的60例(包括Dukes A)未知血清进行预测,得到该方法对大肠癌的检出率为82．6％(19／23),排除率为91．9％(34／37)。结论 蛋白质芯片技术检测血清蛋白质质谱法在大肠癌的诊断中较以往的传统方法具有更高的检出率和排除率,值得进一步研究与应用。

Application of serum protein pattern model in diagnosis of colorectal cancer

Objective To explore the application of serum protein pattern mod els in diagnosis of colorectal cancer (CRC) by proteinchip technology. Methods One hundred and forty-seven serum samples ( 55 CRC patients and 92 healthy indi viduals) randomly divided into training set ( n =87,32 CRC patients and 55 he alth y individuals) and test set( n =60),were subjected for analysis by surface en hanc ed laser desorption/ionization-time of flight-mass spectrometry (SELDI-TOF-M S). Four top-scored peaks in 5910,8930,4476 and 8817 were detected by pr o teinchip software version 3.0. and were trained by a multi-layer artificial neu r al network (ANN) with a back propagation algorithm. An artificial neural network classifier had developed for separating CRC from the healthy group. The classif ier was then challenged with the test set (60 samples including 23 CRC patients and 37 healthy individuals) to determine the validity and accuracy of the classi fication system. Results The artificial neural network classifier separated th e CRC from the healthy samples,with sensitivity of 82.6% and specificity of 91. 9%. Conclusion Combination of SELDI-TOF-MS with the artificial neural network yields significant higher sensitivity and specificity than CEA in the diagnosis of CRC,which should be further studied.