Discrimination of Lung Cancer Related Volatile Organic Compounds with a Colorimetric Sensor Array

Abstract

A simple and rapid discrimination of four lung cancer related volatile organic compounds (VOCs) was achieved with a novel colorimetric sensor array. Based on the cross-responsive mechanism, the sensor system exhibited high sensitivity to selected lung cancer biomarkers, including p-xylene, styrene, isoprene, and hexanal with concentrations varying from 50 ppb to 500 ppb. By extracting color information, it provided good selectivity and fine discrimination of selected gases via pattern recognition with Fisher linear discriminant (FLD). Additionally, with the employment of the Takagi-Sogeno Fuzzy Neural Network (TSFNN), different concentrations of selected VOCs were discriminated. It also suggested that the colorimetric sensor array proposed could not only distinguish different lung cancer related VOCs but also discriminate specific VOCs of different concentrations with an average rate of classification up to 95%. Our preliminary study demonstrated that the cross-responsive sensor array had infinite potential for further clinical and commercial use for early diagnosis of lung cancer.