Dorn Jonas F, Danuser Gaudenz, Yang Ge
Laboratory for Computational Cell Biology, Department of Cell Biology, CB167 The Scripps Research Institute La Jolla, California 92037, USA.
Methods Cell Biol. 2008;85:497-538. doi: 10.1016/S0091-679X(08)85022-4.
With the many modes of live cell fluorescence imaging made possible by the rapid advances of fluorescent protein technology, researchers begin to face a new challenge: How to transform the vast amounts of unstructured image data into quantitative information for the discovery of new cell behaviors and the rigorous testing of mechanistic hypotheses? Although manual and semiautomatic computer-assisted image analysis are still used extensively, the demand for more reproducible and complete image measurements of complex cellular dynamics increases the need for fully automatic computational image processing approaches for both mechanistic studies and screening applications in cell biology. This chapter provides an overview of the issues that arise with the use of computational algorithms in live cell imaging studies, with particular emphasis on the close coordination of sample preparation, image acquisition, and computational image analysis. It also aims to introduce the terminology and central concepts of computer vision to facilitate the communication between cell biologists and computer scientists in collaborative imaging projects.
随着荧光蛋白技术的迅速发展，活细胞荧光成像出现了多种模式，研究人员开始面临一项新挑战：如何将大量非结构化图像数据转化为定量信息，以发现新的细胞行为并严格检验机制假说？尽管手动和半自动计算机辅助图像分析仍被广泛使用，但对于复杂细胞动力学进行更可重复和完整的图像测量的需求，增加了在细胞生物学的机制研究和筛选应用中采用全自动计算图像处理方法的必要性。本章概述了在活细胞成像研究中使用计算算法时出现的问题，特别强调样本制备、图像采集和计算图像分析的紧密协调。它还旨在介绍计算机视觉的术语和核心概念，以促进细胞生物学家和计算机科学家在协作成像项目中的交流。
