The Research Of Microcirculation Parameters Measurement Based On Videos

Microcirculation system is an important regulatory system to maintain the normal function of tissues and organs in human body.It is the basis of local and systemic circulation stability.Microcirculation system research has been widely used in clinical practice,such as sepsis,diabetes and liver disease.These diseases are accompanied by changes in microcirculation parameters.Therefore,quantitative analysis of microcirculation parameters can greatly assist doctors in the diagnosis and treatment of diseases.The traditional microcirculation parameter measurement methods require a lot of manual participation,not only time consuming,but also cannot guarantee the accuracy of the measurement results.In recent years,the research of automated measurement of microcirculation parameters has been widely concerned,and the microcirculation parameters based on video and image processing have higher accuracy and can reduce manual participation to speed up the calculation.We study the microcirculation parameter measurement methods based on microcirculation video and image processing.The research mainly includes three parts:microcirculation video and image preprocessing,the measurement of microcirculation shape parameters,and the measurement of microcirculation dynamic parameters.Preprocessing of microcirculation video and image includes video jitter correction,image denoising,and contrast enhancement.Video jitter correction using the SIFT corner detection combined with electronic image stabilization algorithm,which effectively removes the initial microcirculation video instability.Microcirculation image denoising and contrast enhancement are both based on contourlet transform,which can remove noise and improve the contrast of blood vessels and background parts.It effectively solves the problem of uneven illumination,large noise and low contrast of microcirculation image.The microcirculation image preprocessing improves the visualization of the microcirculation system and facilitates the measurement of microcirculation parameters.Measurements of microcirculation shape parameters include measurements of functional capillary density,curvature,and diameter.Functional capillary density measurement based on the vessel segmentation algorithm.Three kinds of vessel segmentation algorithms are compared:CLAHE,Gabor transform and wavelet transform.These methods can effectively separate the vessels from the background.The result shows that the noise sensitivity of wavelet transform is lower and the result is more accurate.Vessel curvature calculation is based on vessel centerline detection and fitting.Vessel diameter measurement is based on the detection of vessel boundaries.In this paper,the measurement of microcirculation parameters achieves a high degree of automation with high accuracy.Microcirculation dynamic parameters refer to the measurement of microcirculatory flow index and microcirculatory blood flow velocity.The microcirculatory flow index is measured based on the observer's assessment of the video.In this paper,the blood cell tracking and blood flow velocity calculation are based on ST image.The ST image is processed by Gaussian filter,noise filtering and direction filtering.These filters can effectively remove the noise,especially the horizontal stripe noise,and improve the contrast of the track path and the background part.In addition,for the vessels with large diameter or curvature,we put forward considering three tracking line of blood flow,which effectively improve the microcirculation blood flow velocity measurement accuracy.