[Value of auto-fluorescence spectrum combined with tumor markers in diagnosis of lung cancer].

To improve the diagnostic efficiency of cancer, serum fluorescence spectrum combined with tumor marker groups was proved more powerful, especially when used with mathematical evaluation model, that is, artificial neural network (ANN) modeling. ANN modeling is very suitable for the discrimination of lung cancer. ANN has evident superiority in solving nonlinear, multi-parameter and uncertain complicated problems. In the present paper, serum fluorescence spectrum was applied to study the difference among normal, benign and malignant groups and develop the relevant method of determination. On the other hand, combined with tumor markers, CEA, NSE, SCC-Ag, CYFRA21-1 and p16 methylation, artificial neural network and Fisher linear discriminatory analysis were used to develop the prediction models of diagnosis of lung cancer, and compared by ROC. It was shown that the result of the fluorescence spectrum combined with tumor markers based on ANN model is superior to that of the fluorescence spectrum ANN model. The performance of ANN model is superior to that of Fisher linear discriminatory analysis.