AUTHOR=Jiang Xi , Li Xue , Xing Haoyang , Huang Xiaoqi , Xu Xin , Li Jing
TITLE=Brain Entropy Study on Obsessive-Compulsive Disorder Using Resting-State fMRI
JOURNAL=Frontiers in Psychiatry
VOLUME=12
YEAR=2021
URL=https://www.frontiersin.org/journals/psychiatry/articles/10.3389/fpsyt.2021.764328
DOI=10.3389/fpsyt.2021.764328
ISSN=1664-0640
ABSTRACT=Object: Brain entropy is a potential index in the diagnosis of mental diseases, but there are some differences in different brain entropy calculation, which may bring confusion and difficulties to the application of brain entropy. Based on the resting-state function magnetic resonance imaging (fMRI) we analyzed the differences of the three main brain entropy in the statistical significance, including approximate entropy (ApEn), sample entropy (SampEn) and fuzzy entropy (FuzzyEn), and studied the physiological reasons behind the difference through comparing their performance on obsessive-compulsive disorder (OCD) and the healthy control (HC).
Method: We set patients with OCD as the experimental group and healthy subjects as the control group. The brain entropy of the OCD group and the HC are calculated, respectively, by voxel and AAL region. And then we analyzed the statistical differences of the three brain entropies between the patients and the control group. To compare the sensitivity and robustness of these three kinds of entropy, we also studied their performance by using certain signal mixed with noise.
Result: Compare with the control group, almost the whole brain's ApEn and FuzzyEn of OCD are larger significantly. Besides, there are more brain regions with obvious differences when using ApEn comparing to using FuzzyEn. There was no statistical difference between the SampEn of OCD and HC.
Conclusion: Brain entropy is a numerical index related to brain function and can be used as a supplementary biological index to evaluate brain state, which may be used as a reference for the diagnosis of mental illness. According to an analysis of certain signal mixed with noise, we conclude that FuzzyEn is more accurate considering sensitivity, stability and robustness of entropy.