一种眼底黄斑水肿OCT图像分割方法

黄斑水肿(ME)被认为是增殖性糖尿病视网膜病变的主要适应症之一,由于视网膜下层中蛋白质沉积物的积累,ME引起视网膜肿胀。光学相干断层成像(OCT)通过显示黄斑病变的横截面视图提供了ME的早期检测。文章提出了一种新的算法,即改进遗传算法和最佳熵法相结合的方法,可以在前期工作的基础上提高时间效率,从5s提高至0.7s。算法首先对图像预处理,排除血管和散斑噪声对分割效果的影响,再将信息熵作为遗传算法(GA)的适应度函数,将最大熵作为遗传算法的收敛准则。经过遗传操作,得到最佳阈值,对黄斑水肿OCT图像进行分割,然后对得到的二值图像进行边界平滑处理,得到最终的分割图像。所述方法对300张黄斑水肿OCT图像进行了分割,与Canny算子、OTSU分割算法和临床医生手动分割结果相比较。结果表明,算法能够快速准确提取黄斑水肿区域,为临床诊断和治疗提供了定量分析的工具。     

A miniaturized system for measurement of the refractive index of sub-microliter liquid

This study introduced the research and development of a portable and miniaturized system for the measurement of the refractive index of sub-microliter liquid based on a microfluidic chip. A technical method of double-beam interference, was proposed for use in the measurement. Based on this, by using a laser diode as a light source,changes in the refractive index were calculated by utilizing a complementary metal–oxide–semiconductor to detect the movement of interference fringes of the liquid. Firstly, this study simulated the effects of influencing factors on the interference infringes of two Gaussian beams, such as their spot sizes, distance between two beam spots, and detection range. Secondly, this research introduced the system design and construction of the doublebeam interference method and analyzed the results of refractive index tests on sub-microliter aqueous glucose solutions with different concentrations. The measurement accuracy reached 10~(-4) refractive index units. This system has a compact structure and is rendered portable by using batteries for its power supply. The entire system is designed to be a double Z-shaped structure with a length of about 15 cm, a width of 5 cm, and a height of about 10 cm. It can be used to measure changes in the refractive index of sub-microliter to nanoliter liquids based on the use of a microfluidic chip.     

基于分数阶积分算法的OCT图像去噪研究

散斑存在于光学相干层析成像(OCT)信号中,不可避免地影响OCT图像质量。通过对OCT系统中的噪声源进行分析,提出了一种傅里叶域OCT图像散斑噪声降噪的分数阶积分算法。为了克服单纯主观视觉判别图像质量的局限性,均方误差、峰值信号噪声比和边缘保护系数被选为图像去噪评估标准。通过实验与中值滤波和维纳滤波方法进行比较,结果表明,该算法可以有效地保留OCT图像中的重要边缘细节信息,同时有效消除噪声,使图像细节清晰,提高图像质量。