Localizing Spectral Interactions in the Resting State Network Using the Hilbert–Huang Transform

Brain Sciences

Ai-Ling Hsu¹  Men-Tzung Lo²  Chia-Wei Li³  Changwei W. Wu⁴  Pengmin Qin⁵ 

[1] Bachelor Program in Artificial Intelligence, Chang Gung University, Taoyuan 33305, Taiwan;Department of Biomedical Sciences and Engineering, National Central University, Taoyuan 32049, Taiwan;Department of Radiology, Wan Fang Hospital, Taipei Medical University, Taipei 11696, Taiwan;Graduate Institute of Mind Brain and Consciousness, Taipei Medical University, Taipei 11031, Taiwan;Key Laboratory of Brain, Cognition and Education Sciences (South China Normal University, Ministry of Education), Center for Studies of Psychological Application and Guangdong Key Laboratory of Mental Health and Cognitive Science, South China Normal University, Guangzhou 510631, China;

关键词: resting-state fMRI;    ensemble spectral interaction;    Hilbert–Huang transform;    amplitude-to-amplitude coupling;    time-frequency map;    wavelet analysis;   

DOI  :  10.3390/brainsci12020140

来源: DOAJ

【 摘 要 】

Brain synchronizations are orchestrated from neuronal oscillations through frequency interactions, such as the alpha rhythm during relaxation. Nevertheless, how the intrinsic interaction forges functional integrity across brain segregations remains elusive, thereby motivating recent studies to localize frequency interactions of resting-state fMRI (rs-fMRI). To this point, we aim to unveil the fMRI-based spectral interactions using the time-frequency (TF) analysis; however, Fourier-based TF analyses impose restrictions on revealing frequency interactions given the limited time points in fMRI signals. Instead of using the Fourier-based wavelet analysis to identify the fMRI frequency of interests, we employed the Hilbert–Huang transform (HHT) for probing the specific frequency contribution to the functional integration, called ensemble spectral interaction (ESI). By simulating data with time-variant frequency changes, we demonstrated the Hilbert TF maps with high spectro-temporal resolution and full accessibility in comparison with the wavelet TF maps. By detecting amplitude-to-amplitude frequency couplings (AAC) across brain regions, we elucidated the ESI disparity between the eye-closed (EC) and eye-open (EO) conditions in rs-fMRI. In the visual network, the strength of the spectral interaction within 0.03–0.04 Hz was amplified in EC compared with that in EO condition, whereas a canonical connectivity analysis did not present differences between conditions. Collectively, leveraging from the instantaneous frequency of HHT, we firstly addressed the ESI technique to map the fMRI-based functional connectivity in a brand-new AAC perspective. The ESI possesses potential in elucidating the functional connectivity at specific frequency bins, thereby providing additional diagnostic merits for future clinical neuroscience.

【 授权许可】

Unknown