【摘要】 酿酒酵母作为经典的模式细胞，常被用于衰老研究。高通量微流控芯片作为酵母细胞增殖衰老过程长期监测分析平台，通常需面对大量图像数据处理问题。为了解决酵母单细胞形态参数提取图像处理算法中计算体量大、数据容错低的问题，提出了一种基于计算机视觉的优化算法，通过Canny边缘检测、Hough变换、形态学运算和线性插值等步骤，针对微捕获单元中酵母细胞轮廓提取误差和处理效率进行了优化。利用所提算法实现了高精度自动化酵母细胞体积参数提取和生长速率分析，探究了细胞复制衰老过程中体积的动态变化及其与复制寿命的相关性。更多还原

【Abstract】 Budding yeast, the scientific name of which is Saccharomyces cerevisiae, is a typical model organism for cell aging studies. High-throughput microfluidic devices, as long-term monitoring and analysis platforms to investigate cell dynamics along yeast replicative aging, usually encounter challenges in processing high-throughput micrographs. To address the challenges, such as the substantial computational load and limited data tolerance in image processing for extracting single-cell morphological parameters of budding yeast, an algorithm based on computer vision is proposed to extract mother-cell contours in single-cell traps. The algorithm integrates Canny edge detection, Hough transform, morphological operations, and linear interpolation. Several tailored optimizations have been implemented to deal with errors in contour extraction and enhance overall efficiency in image processing. Furthermore, the proposed algorithm is utilized to achieve high-precision automated extraction of cell volume and growth rate of budding yeast, and analyze the dynamic changes in cell volume during the process of yeast replicative aging and its correlation with replication lifespan.更多还原