Title
Measuring spatiotemporal dependencies in bivariate temporal random sets with applications to cell biology.
Abstract
Analyzing spatio-temporal dependencies between different types of events is highly relevant to numerous biological phenomena (e.g. signalling and trafficking) especially as advances in probes and microscopy have facilitated imaging of dynamic processes in living cells. For many types of events, the segmented areas can overlap spatially and temporally forming random clumps. In this paper, we model binary image sequences of two different event types as a realization of a bivariate temporal random set and propose a non-parametric approach to quantify spatial and spatio-temporal interrelations using the pair-correlation, cross-covariance and the Ripley IK functions. Based on these summary statistics we propose a randomization procedure to test independence between event types by applying random toroidal shifts and Monte Carlo tests. A simulation study assessed the performance of the proposed estimators and showed that these statistics capture the spatio-temporal dependencies accurately. The estimation of the spatio-temporal interval of interactions was also obtained. The method was successfully applied to analyze the interdependencies of several endocytic proteins using image sequences of living cells and validated the procedure as a new way to automatically quantify dependencies between proteins in a formal and robust manner.
Year
DOI
Venue
2008
10.1109/TPAMI.2007.70821
IEEE Trans. Pattern Anal. Mach. Intell.
Keywords
Field
DocType
random toroidal shift,bivariate temporal random sets,cell biology,measuring spatiotemporal dependencies,random clump,bivariate temporal random set,different type,event type,binary image sequence,different event type,spatio-temporal interval,spatio-temporal dependency,spatio-temporal interrelation,set theory,clathrin mediated endocytosis,covariance function,stochastic process,image segmentation,genetics,pair correlation function,pattern analysis,molecular biophysics,cross covariance,binary image,monte carlo methods,testing,proteins,stochastic processes,random processes,signal analysis,statistical testing,applications,image analysis,signal processing
Covariance function,Pattern recognition,Cross-covariance,Computer science,Stochastic process,Nonparametric statistics,Temporal database,Artificial intelligence,Statistical hypothesis testing,Covariance,Estimator
Journal
Volume
Issue
ISSN
30
9
0162-8828
Citations 
PageRank 
References 
1
0.40
4
Authors
7
Name
Order
Citations
PageRank
Ester Diaz1183.03
Rafael Sebastian23812.23
Guillermo Ayala39516.13
María Elena Díaz4215.69
Roberto Zoncu5112.03
Derek Toomre6467.75
Stéphane Gasman710.40