ABSTRACT
ABSTRACT: A simple and rapid discrimination method of four lung cancer-related Volatile Organic Compounds (VOCs) with a colorimetric sensor array was presented. The sensor system showed high sensitivity to selected lung cancer biomarkers with concentrations of 50-500 ppb. By means of Fisher Linear Discriminant (FLD), it exhibited good selectivity and fine discrimination of selected gases via extracting array pattern information. And dierent concentrations of selected VOCs were discriminated by the employment of Back Propagation Neural Network (BPNN). It is also suggested that the proposed method could not only distinguish dierent kinds of lung cancerrelated VOCs but also perform quantitative determination of some VOC concentrations with the correct rate almost up to 100%. It is expected that this method will have potential application for clinical and commercial value in the near future.
