Description Length Guided Unified Granger Causality Analysis

In this article, we propose a description length guided unified Granger causality analysis (uGCA) framework for sequential medical imaging. While existing efforts of GCA focused on causal relation design and statistical methods their improvement, our strategy adopts the minimum (MDL) principle in procedure where MDL offers model selection criteria deciding optimal sense length. Under framework, present different forms linear representations under several coding schemes that all achieve lower bounds redundancy, thus producing valid criteria. The are validated using 5-node network synthetic experiment, illustrating its potential advantage over conventional two-stage approach. subtle distinction between performance uGCA is investigated as well. More importantly, proposed approach gives more similar topology than challenging fMRI dataset, which neural correlates mental calculation elicited by visual auditory stimulation (respectively) same task paradigm, allowing one to evaluate methods.