Back to Search Start Over

Tab S2: Statistical relationships between tERGM parameters and patients' behavior from Temporal exponential random graph models of longitudinal brain networks after stroke

Authors :
Obando, Catalina
Rosso, Charlotte
Siegel, Joshua
Corbetta, Maurizio
De Vico Fallani, Fabrizio
Publication Year :
2022
Publisher :
The Royal Society, 2022.

Abstract

tERGMs are fitted over the first two time visits covering the acute phase of stroke, i.e. 2 weeks and 3 months. The corresponding parameter values ��E (interhemispheric temporal edges) and ��T (intrahemispheric temporal triangles) �������� corrected for experimental nuisance variables - are correlated with the multidomain behavioral scores in the chronic phase, i.e. 1 year. In panel A), Spearman correlation coefficients are reported along with their p-values between parentheses. 95% confidence intervals via bootstrap with replacement sampling are reported next to significant correlations. Panel B) shows the R2 values of the linear models used to explain patient's behavior along with the p-values of the independent variables, ie tERGM parameters, lesion size and age. Panel C) shows the R2 values of the linear models used to predict the behavior in hold-out patients, along with the p-values associated with Pearson's correlation between predicted and actual behavioral scores. In all panels, one asterisk denotes a statistical significance of p

Details

Database :
OpenAIRE
Accession number :
edsair.doi.dedup.....a549dd5c86a1226eb2ef061bce40c882
Full Text :
https://doi.org/10.6084/m9.figshare.18855089.v1