| AVS 54th International Symposium | |
| Biomaterial Interfaces | Friday Sessions |
| Session BI-FrM |
| Session: | Microbioanalytical Systems |
| Presenter: | K. Artyushkova, University of New Mexico |
| Authors: | K. Artyushkova, University of New Mexico M. Bore, University of New Mexico A. Evangelista-Lara, University of New Mexico G.P. Lopez, University of New Mexico |
| Correspondent: | Click to Email |
This study investigates the potential of multivariate methods (MVA) for identifying electrokinetic separation and estimating velocities of moving species based on analysis of imaging datasets from microfluidic and nanofluidic devices. We have developed an image analysis methodology based on MVA of temporal datasets that is capable of identifying velocities of at least two molecular species from the images where no visible separation of the species has occurred. Among multivariate analysis methods examined are Principal Component Analysis (PCA), Multivariate Curve Resolution (MCR), PARAFAC (parallel factor analysis) and Independent Component Analysis (ICA). These methods allow one to fully exploit the data by analyzing all pixels within images and using the temporal dimension, in contrast with manual methods of visual inspection of images or traditional image processing methods. The methodology has been developed and tested temporal images acquired by fluorescence microscopy capturing separation within nanochannels, microchannels and gel electrophoresis of charged dyes and model protein receptor/ligand systems.