Research On Heterogeneous Detection In Biological Tissues Based On Multi-wavelength Transmission Image Information

Multi-wavelength non-destructive transmission imaging has become a research hotspot and has been widely used in many fields.However,there are few research applications of multi-wavelength transmission imaging in the detection of biological tissue heterogeneity.This is mainly due to the strong scattering of light source in the transmission process of biological tissue,which makes the signal in the multiwavelength transmission image weak and can’t obtain rich heterogeneous characteristic information.However,due to the existence of a large number of new blood vessels and hemoglobin around the heterogeneity of biological tissues,a large shadow is generated in the process of transmission imaging,which provides a possibility for optical transmission imaging to detect heterogeneity in biological tissues.And optical properties of different tissues are different at different wavelengths.Therefore,in order to promote the clinical application of multiwavelength transmission imaging in the early screening of biological tissue heterogeneity,this paper improves the accuracy of heterogeneous detection in multiwavelength transmission images based on multi-wavelength information,multiwavelength information combination,multi-wavelength information fusion and optimal wavelength combination.The main research works of this paper include:1.Aiming at the fuzzy problem of heterogeneous edge in multi-wavelength transmission image.1)Based on multi-wavelength information,a multi-wavelength image fusion method in frequency domain is proposed.The low-frequency and highfrequency regions of image are respectively reconstructed by regional variance significance and regional space energy weighting fusion strategies to break the connectivity of image between the same frequency domain.2)Based on multiwavelength information,an image enhancement method based on Weight Constraint Decision is proposed.The method eliminates the erosion effect and checkerboard effect in the process of histogram equalization.Compared with the current popular methods,both methods can not only improve the contrast between heterogeneous region and normal region in image,but also can detect the location of heterogeneous edge more accurately.2.Aiming at the problem that only weak signals pass through biological tissues during transmission imaging,it is impossible to obtain rich heterogeneous characteristic information.Based on the modulation and demodulation-frame accumulation technique and the optical properties of different tissues at different wavelengths are different multi-spectral characteristic attributes,the combination of multi-wavelength information realizes the improvement of heterogeneous classification accuracy in transmission image.And as the number of wavelength combinations increases,the average classification accuracy of model increases gradually.3.Aiming at the obvious scattering problem of light source in the transmission process of biological tissues.1)The spatial pyramid matching model is used to modify the scattering coefficient of images to obtain high quality transmission images.2)In terms of multi-type multi-wavelength image fusion,the addition of gradient fusion color map further improves the recognition accuracy of heterogeneity in transmission image.3)In the aspect of multi-wavelength image super-resolution reconstruction,the multi-wavelength fusion color map is processed by various superresolution image reconstruction models in turn to obtain high quality multiwavelength transmission image.The above three aspects of multi-wavelength image fusion can improve the image quality,and gradually improve the recognition accuracy of heterogeneity in multi-wavelength image.4.Aiming at the problem of high information redundancy between bands in multi-wavelength transmission imaging.A variety of nonlinear transformations are used to transform multi-wavelength images into spectral data,and then the set optimal wavelength combination is selected based on the successive projection algorithm,which provides a new idea of wavelength combination optimization for multi-wavelength transmission imaging screening heterogeneity.