基于广度优先搜索算法计算视网膜血管弯曲度和长度

摘要

目的：通过多年临床经验可知，眼底血管是人体唯一可以通过非创伤性直接观察的较深层的微血管，通过这些微血管的弯曲度和长度可以表征视网膜状态，进一步诊断疾病。方法本文介绍一种基于广度优先搜索算法，将视网膜血管分布看成血管树，利用遍历法计算视网膜中血管长度（即每个节点间的距离）及弯曲度。结果通过imagej软件利用此方法测量十五种不同形态位置的血管段长度和弯曲度，并与人工测量的血管数据进行对比，对比所得数据，验证此计算方法是否能准确有效的表征血管长度及弯曲度，将其作为诊断疾病的有力依据。结论通过对比2组数据，此方法能快速准确的计算出血管长度，减少人为测量的误差。%Objective Through many years clinical experience, the retinal blood vessels is the body only can be observed directly by non traumatic deep capillaries, curvature and the length can be characterized by these capillaries retinal status, further diagnosis.Methods This paper introduces a based on the breadth-first search algorithm, the distribution of retinal blood vessels as a vascular tree,through the calendar is used to calculate the length of the blood vessels in the retina (i.e., the distance between each node) and bending.Results By imagej software use this way to measure the 35 different forms of bending and vessel length, and comparison with data of artificial measurement of blood vessels, comparing the data obtained, to validate the calculation method can accurately effective characterization of blood vessels and bending length, make it as a powerful reason to diagnose diseases.Conclusion By comparing the two groups of data, this method can fast calculate the length of blood vessels, reduce man-made error of the measurement.