Multiple Marker Detection Of Biological Samples Based On Hyperspectral Imaging Technique

In the field of tumor microenvironment,the ability to obtain different cell phenotypes on the same tissue sample and obtain their spatial location information is a problem that needs to be solved urgently.Spectral aliasing among different fluorescent probes makes it difficult for different phenotypes of cells to be labeled and detected simultaneously,and repeated elution of the sample will damage the sample and can’t be accurately detected.Hyperspectral microscopic imaging technology effectively combines microscopic imaging technology and spectral detection technology,microscopic imaging technology is used to obtain sample morphological information.Spectral detection technology is mainly used to obtain a certain range of spectral information of a single point of a tissue sample.The combination of the two can obtain the morphological information and spectral information of biological samples simultaneously.Morphological information can effectively obtain the spatial distribution of cells,while spectral information can effectively solve spectral aliasing and autofluorescence interference,and achieve the acquisition of different cell phenotypes on tissue samples.In this paper,an automatic push-sweep hyperspectral microscopy imaging system was established by combining prism-grating spectroscopic technology,microscopic imaging technology,autofocus technology and multiple fluorescence labeling technology to realize the study of tumor microenvironment.The system was designed based on Nikon research grade inverted fluorescence microscope,a prism-grating component was used for spectrum splitting,a high precision two-dimensional motorized stage was applied for pushbroom,an motor focus module was controlled for focusing,and the hyperspectral microscopic image was collected through a highly sensitive SCMOS camera.After the hardware system was built,the hyperspectral microscopy imaging control software was written based on the Labview development software,including the high-precision motion control of the twodimensional motion platform,the electric focusing system for automatic focusing control,and the camera imaging control.The position of the camera was adjusted and the motion speed of the platform was calibrated to make the camera,the light splitting system and the two-dimensional motion platform work together,obtained the push-broom image of the target sample,collected and store the original image data,evaluated the definition of the image,and selected the appropriate one the clarity evaluation function of the system tests the spatial resolution,calibrated the spectrum,and completed the performance test of the system.The spectral range of the system was from 420 nm to 800 nm,the spectral sampling rate is2.06 nm,the spectral resolution was better than 3.5nm,and the spatial resolution was better than 0.87 μm.In the meantime,aiming at the difficulty of auto-focusing in the push-broom hyperspectral microscopy imaging system,experiments were carried out using passive and active focusing methods.The advantages and disadvantages of the two focusing methods were compared and analyzed from three aspects of imaging clarity,imaging speed and system complexity.The problem that focusing could not be realized in the hyperspectral microscopic imaging system was solved,and the large field of view of 3.25mm×3.25 mm under 40 times objective lens was realized.At last,the pathological section labeled with 4-color quantum dots excited by a monochromatic laser was the research object,perform multiple fluorescent label imaging and unmix the mixed signal spectrum.The results show that the hyperspectral microscopic imaging system can eliminate the interference of fluorescence spectrum overlap,and can effectively detect multiple fluorescent labels,and is expected to achieve simultaneous acquisition of different cell phenotypes and spatial location information on the same tissue sample,provide help for research in the field of tumor microenvironment.