Research Of Phase Imaging Technology Applied In Micro-channel Cell's Detection

The morphological characteristics of cells are one of the focuses of cytology and medical research.Most of the living cells are pure phase objects,which only affect the phase of the light wave.Traditional optical methods that record object information using the amplitude and intensity of light cannot achieve observation of living cells.Therefore,phase imaging techniques such as digital holographic microscopy based on coherent light and differential interference based on partially coherent light have emerged.The phase imaging technology effectively uses the phase information of the cell to recover its morphological characteristics by inverting the phase change,and can perform nondestructive detection on the cell.There are many spatial microstructures according to different uses of microfluidic chips,which provides a convenient and safe operating platform for cell culture,capture,separation and other operations,and is an important tool for studying microfluidic technology.The microfluidic channel is an important part of the microfluidic chip.When detecting the state of the cells in the microfluidic chip,it is an important task to observe the microfluidic channel cells.In response to the characteristics and needs of the above problems,this article builds a microscopic imaging system composed of a programmable LED array as a light source,and applies phase imaging technology to microfluidic channel cell detection.The imaging system is partially coherent light illumination,combined with the phase gradient transfer function to complete the construction of the imaging model from the frequency domain.The imaging result of the system is determined by the LED lighting mode,and different lighting schemes can be achieved through coding control without mechanical operation.Combined with asymmetric lighting,differential phase contrast imaging results of cells in the microfluidic channel can be obtained.In addition,according to the characteristics of cell movement in the microchannel,this paper further studies the effect of camera acquisition mode on the imaging of moving cells,compares the efficiency and accuracy of different deblurring algorithms in the restoration of moving cell images,and uses the optimal restoration method The image quality is improved.At the same time,the circle detection method is used to obtain the target on the phase imaging result.After the target feature is extracted,the application of the phase imaging technology in the microfluidic channel cell detection is realized by combining with the machine learning technology.After verification,the phase microscopic imaging system of LED array illumination constructed in this paper can establish a relationship between the phase gradient and the amplitude image.Moreover,imaging does not require complicated optical elements,and the implementation cost is low.The combination of image matching and restoration proposed in this paper solves the problem of how to obtain differential phase contrast images when the cells in the microchannel are moving.At the same time,BP neural network and support vector machine were used to realize the detection and recognition of two samples,which proved the availability of phase imaging technology in microfluidic channel cell detection.