Classification of nitric oxide assessed by hybrid kernel function in lacunar stroke

Nitric oxide (NO) is a key oxidative stress marker. Real time NO measurement in pA/nM by electrochemistry is the powerful tool explores pathophysiological process of many disease including stroke subtypes, especially, lacunar stroke. In this study, we investigate the performance of classification models by hybrid nonlinear kernel function of NO obtained from healthy control and lacunar stroke. The results show that hybrid kernel function classifier has higher performance than those of linear, polynomial, radial basis function (RBF) and sigmoid kernel functions and also gives the best classification of NO in normal and lacunar stroke. In conclusion, hybrid kernel function will be applied and further studied in acute lacunar stroke, chronic hypertension and hyperlipidemia.