Lu Jianfeng, Wu Tao, Cheng Lei, Zhu Zhenwen, Yang Jingyu
School of Computer Science and Technology, Nanjing University of Science and Technology, Nanjing, China.
Int J Comput Biol Drug Des. 2012;5(1):75-88. doi: 10.1504/IJCBDD.2012.045953. Epub 2012 Mar 21.
Quantitative analysis of image-based phenotype is of significant importance in zebrafish image analysis and high-content screening. In this paper, an automatic analysis pipeline is proposed for quantification of the curvedness of zebrafish embryo tails. First, zebrafish embryos are extracted based on the active contour model, and then its skeleton line is generated by the principal graph algorithm. Several curvedness measures of tail skeleton are designed and compared, and experimental results show that joint consideration of the tail shape and its relative position relationship can result in better performance. The proposed pipeline has comparable performance to the manual analysis.
基于图像的表型定量分析在斑马鱼图像分析和高内涵筛选中具有重要意义。本文提出了一种自动分析流程,用于量化斑马鱼胚胎尾巴的弯曲度。首先,基于主动轮廓模型提取斑马鱼胚胎,然后通过主曲线算法生成其骨架线。设计并比较了几种尾巴骨架的弯曲度测量方法,实验结果表明,综合考虑尾巴形状及其相对位置关系能取得更好的性能。所提出的流程与人工分析具有相当的性能。