Recognizing biomarker gases by monitoring the flow rates in a microfluidic channel

Microfluidics has become popular as a diagnostic methodology owing to its generality in the examination of variety of biofluids. More recently, discrimination among gaseous analytes in microfluidic channels has also been demonstrated. Breath analysis for early detection of different diseases is important owing to its numerous advantages over other diagnostic techniques. In this paper, we introduce a simple, inexpensive and disposable device capable of discriminating between 8 important biomarkers by monitoring their flow rates along a microchannel. The data related to the flow rates of different analytes are recorded as response patterns, and their discriminative information content is extracted with signal processing techniques and presented as feature vectors in a 3-D feature space. Discrimination among acetone, benzene, carbon monoxide, ethanol, hydrogen, isopropanol, methanol and toluene at wide concentration ranges in air is experimentally demonstrated.