Synopsis

Thalamus has the capacity to generate oscillations characterized by different frequency and levels of synchrony. However, little is known of what function the generated EEG dynamic pattern may serve in thalamocortical interactions. In this study, we used psychophysiological interaction framework to detect fMRI thalamocortical coupling modulated by dynamic EEG connectivity patterns, and tested the alteration of the modulatory couplings in juvenile myoclonic epilepsy. The results found frequency-dependent characteristics for the thalamocortical coupling. The group comparison revealed the couplings were altered with certain thalamus sets involved in JME, which may have implications for better understanding of the thalamocortical circuit influenced by epileptic activity.

Abstract

Introduction: Thalamus subserves cognitive functions by interacting with cortical and other subcortical areas, generating different types of oscillation characterized by their frequency and levels of synchrony1. However, little is known of what function the generated synchrony pattern may serve in thalamocortical and cortiocortical interactions. Epilepsy, as is a system disease, has been frequently reported involving aberrant thalamocortical coupling2-4. Moreover, computational analysis suggests the spike-wave generation may be associated with increased cortical firing, then resulting in the low-frequency thalamocortical oscillation5, 6. The association between EEG oscillation variation and thalamocortical may contribute to our understanding of the origins of the epileptic activity. Technically, simultaneous EEG-fMRI has been a powerful approach to link the EEG activity and the whole brain, even the deep sub-cortical structures7, 8. Therefore, we hypothesized that the thalamocortical connectivity can be modulated by the dynamic connectivity patterns of the rhythmic EEG, and this modulation may be altered in epilepsy.

Methods: In the current study, simultaneous EEG-fMRI data was acquired from 20 juvenile myoclonic epilepsy (JME), and 20 healthy controls (HC). Adapted directed transfer function (ADTF) was applied to construct scalp EEG time-varying connectivity patterns, and the dynamic ADTF variation of different frequency bands was then used as a modulator in psychophysiological interaction (PPI) framework to detect fMRI thalamocortical functional coupling. Here, the Power-264 meta-analysis cortical regions of interest 9 and 8 thalamus sets of Human Brainnetome Atlas 10 were used. The alteration of thalamocortical coupling with EEG modulation in JME was analyzed. Furthermore, the association between corticocortical connectivity and modulatory thalamocortical coupling was examined.

Results: The results revealed that these thalamocortical couplings have frequency band specificity, and posterior and lateral thalamus sets were more sensitive to the modulations. For theta frequency band, positive modulatory effects were found mainly in thalamus sets to the sensorimotor network (SMN) and cingulo-opercular network (CON), while the modulations were mainly decreased in JME (Figure 1). For alpha, most thalamus sets demonstrated negative modulatory couplings to visual network (VN) and default mode network (DMN) in HC group, while JME showed inversed positive modulatory couplings in VN and DMN, as well as distinct increased coupling to SMN. For beta band, JME showed significant decreased connectivity to CON (Figure 3). In addition, modulatory thalamocortical coupling was highly associated with corticocortical functional connectivity, while mismatch in visual and sensorimotor pathway was prevailingly presented in JME patients (Figure 4).

Discussion: In HC group, negative relationship has been found between the alpha power and the connectivity between thalamus sets and visual network/DMN, which may be associated with specific functional inhibition during alpha activity11. The positive modulation from theta and beta bands mainly act on the coupling from thalamus to SMN, CON areas, which was consistent with the main projections originating from thalamus and limbic thalamus12, 13. Thalamocortical network is the vulnerable system in epilepsy. This study confirmed the abnormal thalamocortical couplings in JME, and established the association between the phenomenon and modulatory EEG variation inputs. Inversed thalamus-VN, thalamus-DMN coupling may be related with the abnormal visual pathway and self-reference system in epilepsy group. Also, altered thalamus-SMN and thalamus-CON coupling with theta and beta modulation in epilepsy implied that theta and beta are the main precursor frequency bands positively modulating the connectivity between thalamus and cortex14. In addition, we found that the parietal, occipital and temporal thalamus were all sensitive areas to low frequency dynamic oscillation, which may represent the spatial pattern of thalamocortical as a mechanism due to the spontaneous oscillation feedback. Furthermore, the specific increased association of motor and visual pathway in terms of alpha modulation may imply the potential mechanism of altered intrinsic network connectivity in JME.

Conclusion: In summary, this study investigated the modulation from rhythmic variation in thalamocortical system, and found frequency-dependent characteristics for these modulations. Moreover, the modulatory thalamocortical coupling was altered with certain thalamus sets involved in JME patients. In addition, the association between thalamocortical and corticocortical connecitivty was also detected. This modulation analysis may have implications for a better understanding of the descending control of thalamocortical circuit and the influences by epileptic activity.

Acknowledgements

References