Study Of Polarization Diffraction Imaging Flow Cytometry By One TDI Camera

Morphology and function of the cell are highly correlated.Current approaches of cell morphology study and research are primarily based on microscopy methods.The rapid advances in the research and development of monoclonal antibodies,fluorescent reagents and flow cytometry makes it possible for rapid assay of cell functions by fluorescent signals,and conventional flow cytometry(FCM)is widely used in clinical and cell biology research.The existing approach of FCM,however,has three significant drawbacks: the cell must be stained for measurement of fluorescent signals as the major source of information,FCM provides very limited information on cell morphology,and small amount of information per cell.A new method of polarization diffraction imaging flow fytometry(p-DIFC)has been developed to solve these problems.Despite these progresses,the motion blur still occurs in the acquired diffraction images of flowing cells using conventional CCD camera by the current experimental systems,which will reduce the detection speed and the classification accuracy of the p-DIFC method.In this dissertation study,we have designed and constructed a new imaging unit based on an in-house developed time-delay-integration(TDI)CCD camera to significantly improve the experimental system by eliminating the motion blur.Acquisitions of diffraction images from microspheres and human cancer cells were performed to validate the new imaging unit and investigate quantitatively the effect of motion blur on cell classification.The TDI based imaging unit has been designed and completed for acquiring cross-polarized diffraction images(p-DI)with one camera.The new design allows measurement of angle-resolved distributions of certain combinations of the Mueller matrix elements.Finally,theoretically analysis shows that the system detection speed can be improved by a factor of 50 in comparison to the previous p-DIFC systems with two conventional CCD cameras.Through this dissertation research,we have investigated the motion blur in the images collected by the different sets of cameras and developed a quantitative analysis model for analyzing the motion blur.The motion blur simulation method for cells diffraction image gotten by TDI-CCD camera,the analysis and evaluation of motion blur effect on the diffraction image texture parameter model are also proposed and verified.Then the data analysis proves that the TDI-CCD camera and imaging system in this paper can be used to get diffraction image in high signal-to-noise ratio.Through an experiment of the particles in different velocity,this paper quantitatively studies the influence of motion blur on the diffraction image of microspheres and human cells of MCF-7,K562.The experimental data shows that motion blur effects on the diffraction images of the cells from lateral scattering light are far greater than microspheres.The quantitative analysis of the motion blur to the influence of diffraction image texture parameters and cell classification results are completed.Finally,a cell classification study on HL-60 and MCF-7 cells has been carried out to investigate the accuracy and robustness of the classification model.Three groups of measurements have been performed on the HL-60 and MCF-7 cells in different days.The GLCM parameters are calculated with different GLCM steps,and then SVM and three data sets are used to classify the two cells,then we calculate and compare the accuracy and consistency of the classification on different GLCM steps.The results demonstrate that the significant reduction of motion blur and choice of different pixel distances can markedly improve the consistency of best performing SVM models.