The specific aim of the project is to evaluate whether novel information concerning the biological characteristics of living cells can be extracted from digitally-captured flow cytometric pulse waveforms, using a specific new class of mathematical functions termed wavelet transformations. This project, therefore, is directed towards increasing the ability to empirically detect differences between cell types using cytometry. The characterization of differences in the biological properties of cells is central to the understanding of the process of growth and development of multicellular organisms, and equally is central to the detection and treatment of disease states. If this project is successful, it is believed that it will have profound biomedical and clinical relevance, particularly in the areas of hematology, oncology, and virology. In previous work, the investigators have built and tested hardware that permits digital capture of the pulse waveforms produced by flow cytometric analysis of biological cells. They have created a library of pulse waveforms from a variety of different cell types. These captured waveforms can be then processed in the digital domain, and the investigators have been able to demonstrate how novel information can be extracted from the pulse shapes using the discrete fourier transform. One of the characteristics of inflow cytometric pulse waveforms is that they comprise individual pulses having a generally Gaussian shape, which increases from zero to a maximum as the cell passes into the laser beam, and then decreases again to zero as the cell exits. Technically, this means the pulse waveforms are non-stationary, being signals that evolve in time. A recent development in signal processing has been the development of transformations particularly suited for analysis of non-stationary waveforms; one of these is called the wavelet transformation. The investigators now wish to subject these pulse waveforms to wavelet analysis in order to determine whether this analysis can extract novel information concerning the optical properties of these cells. If this work is successful, it should lead to the design of dedicated hardware, compatible with commercial flow cytometric instrumentation, that can efficiently implement wavelet transformation of flow cytometric data. The ultimate goal is the realtime processing of pulse waveforms, to permit sorting of cells based on wavelet characteristics. Through assignment of novel wavelet characteristics to particular cell types or pathologies, specific analytical, diagnostic, and prognostic procedures should then emerge.
