Dow J A, Lackie J M, Crocket K V
Department of Cell Biology, University of Glasgow, Scotland.
J Cell Sci. 1987 Feb;87 ( Pt 1):171-82. doi: 10.1242/jcs.87.1.171.
An image analysis package based on a BBC microcomputer has been developed, which can simultaneously track many moving cells in vitro. Cells (rabbit neutrophil leucocytes, BHK C13 fibroblasts, or PC12 phaeochromocytoma cells) are viewed under phase optics with a monochrome TV camera, and the signal digitized. Successive frames are acquired by the computer as a 640 X 256 pixel array. Under controlled lighting conditions, cells can readily be isolated from the background by binary filtering. In real-time tracking, the positions of a given cell in successive frames are obtained by searching the area around the cell's centroid in the previous frame. A simple box-search algorithm is described, which proves highly successful at low cell densities. The resilience of different search algorithms to various exceptional conditions (such as collisions) is discussed. The success of this system in real-time tracking is largely dependent upon the leisurely speed of movement of cells, and on obtaining a clean, high quality optical image to analyse. The limitations of this technique for different cell types, and the possible configurations of more sophisticated hardware, are outlined. This system provides a versatile and automated solution to the problem of studying the movement of tissue cells.
已开发出一种基于BBC微型计算机的图像分析软件包,它能够在体外同时追踪多个移动细胞。在相差光学条件下,用单色电视摄像机观察细胞(兔嗜中性白细胞、BHK C13成纤维细胞或PC12嗜铬细胞瘤细胞),并将信号数字化。计算机以640×256像素阵列采集连续的帧。在受控照明条件下,通过二值滤波可轻松将细胞与背景分离。在实时追踪中,通过在前一帧中搜索细胞质心周围的区域来获取给定细胞在连续帧中的位置。描述了一种简单的盒式搜索算法,该算法在低细胞密度下非常成功。讨论了不同搜索算法对各种异常情况(如碰撞)的适应性。该系统在实时追踪方面的成功很大程度上取决于细胞悠闲的移动速度,以及获取清晰、高质量的光学图像进行分析。概述了该技术对不同细胞类型的局限性以及更复杂硬件的可能配置。该系统为研究组织细胞运动问题提供了一种通用且自动化的解决方案。