SGER: Detection of Subtle Spatio-Temporal Patterns in Biological Imaging Using Permutation Methods

在现代生物学中，计算机在获取、处理、可视化和存储复杂的四维(空间和时间)数据方面已经成为不可或缺的角色。然而，生物学研究的最后阶段是从获取、处理和存储数据到新的生物学知识。计算机化的可视化在这一阶段得到了应用，但能够补充生物学家工具包的计算机化推理工具的开发还很少。该项目正在开发发现工具--以新算法和软件的形式--来处理复杂的生物成像4D数据。这些工具将有助于检测用标准分析手段很难注意到的微妙的时空模式(即非随机性)。主要的技术是基于排列测试的，这种测试在其他科学中得到了广泛的应用，但在生物成像中还没有得到应用。挑战是设计算法来处理典型成像数据带来的特定问题。

Computers have become indispensable in much of current biology, in their roles of acquiring, processing, visualizing, and storing complex four- dimensional (space & time) data. However, the final stage in biological research is going from acquired, processed, and stored data to new biological knowledge. Computerized visualization is used in this stage, but there has been less development of computerized reasoning tools that can complement the biologist's toolkit. This project is developing discovery tools - in the form of novel algorithms and software - to deal with the complex 4D data of biological imaging. These tools will help detect subtle spatio-temporal patterns (i.e., non-randomness) that would be difficult to notice with standard means of analysis. The main technique is based on permutation tests, which have been applied widely in other sciences but not in biological imaging. The challenge is to design algorithms to handle the specific problems posed by typical imaging data.