EEG Microstate-Specific Functional Connectivity and Stroke-Related Alterations in Brain Dynamics

 No clue, but lots of big words. I would require readability so survivors can take it to their doctor to implement.  This does nothing of the sort.

EEG Microstate-Specific Functional Connectivity and Stroke-Related Alterations in Brain Dynamics

Zexuan Hao¹, Xiaoxue Zhai², Dandan Cheng², Yu Pan^2*† and Weibei Dou^1*†

• ¹Department of Electronic Engineering, Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua University, Beijing, China
• ²Department of Rehabilitation Medicine, School of Clinical Medicine, Beijing Tsinghua Changgung Hospital, Tsinghua University, Beijing, China

The brain, as a complex dynamically distributed information processing system, involves the coordination of large-scale brain networks such as neural synchronization and fast brain state transitions, even at rest. However, the neural mechanisms underlying brain states and the impact of dysfunction following brain injury on brain dynamics remain poorly understood. To this end, we proposed a microstate-based method to explore the functional connectivity pattern associated with each microstate class. We capitalized on microstate features from eyes-closed resting-state EEG data to investigate whether microstate dynamics differ between subacute stroke patients (N = 31) and healthy populations (N = 23) and further examined the correlations between microstate features and behaviors. An important finding in this study was that each microstate class was associated with a distinct functional connectivity pattern, and it was highly consistent across different groups (including an independent dataset). Although the connectivity patterns were diminished in stroke patients, the skeleton of the patterns was retained to some extent. Nevertheless, stroke patients showed significant differences in most parameters of microstates A, B, and C compared to healthy controls. Notably, microstate C exhibited an opposite pattern of differences to microstates A and B. On the other hand, there were no significant differences in all microstate parameters for patients with left-sided vs. right-sided stroke, as well as patients before vs. after lower limb training. Moreover, support vector machine (SVM) models were developed using only microstate features and achieved moderate discrimination between patients and controls. Furthermore, significant negative correlations were observed between the microstate-wise functional connectivity and lower limb motor scores. Overall, these results suggest that the changes in microstate dynamics for stroke patients appear to be state-selective, compensatory, and related to brain dysfunction after stroke and subsequent functional reconfiguration. These findings offer new insights into understanding the neural mechanisms of microstates, uncovering stroke-related alterations in brain dynamics, and exploring new treatments for stroke patients.

Introduction

It is now a consensus that the brain at rest is not truly “at rest”, and the spontaneous brain activity exhibits complex dynamic spatiotemporal configurations (Raichle et al., 2001; Fox and Greicius, 2010; Pirondini et al., 2017). Of note, spontaneous brain activity can predict behavioral performance, and intrinsic activity plays a basic and functional role in brain function (Spisak et al., 2020). On the other hand, brain injury (e.g., stroke) causes behavioral deficits as well as widespread structural and functional network dysfunction (Salvalaggio et al., 2020). Motor and sensory impairments are the two most common deficits after stroke, affecting approximately 85 and 50% of stroke patients, respectively (Wu et al., 2015; Ramsey et al., 2017). To date, using neurophysiological techniques, a tremendous number of studies based on spectral analysis, functional connectivity, and graph theory analysis have been reported in the field of brain injury mechanisms and stroke rehabilitation (Wu et al., 2015; Stinear, 2017; Trujillo et al., 2017; Mane et al., 2019; Chiarelli et al., 2020; Ros et al., 2022). These studies laid an important foundation for our understanding of stroke-related neurological alterations and neuroplasticity and for finding neural markers that can indicate prognosis and outcomes. However, given the intrinsic non-stationary nature of brain neural signals, static features are not time-resolved and lose the information about the time dimension. Dynamic analysis methods may better reflect the full profile of brain activity and capture critical features of the spatiotemporal dimensions in health and disease (Kabbara et al., 2017; Chang et al., 2018; O’Neill et al., 2018; Bonkhoff et al., 2020; Wang et al., 2020). Dynamic functional connectivity is the most commonly used method in functional magnetic resonance imaging (fMRI) studies to investigate the connectivity dynamics in stroke patients. It has shown that stroke alters the brain’s preference for distinct connectivity states (Bonkhoff et al., 2020; Wang et al., 2020). However, the sliding-window approach, which is typically utilized in dynamic functional connectivity, has the limitation of requiring a prior unknown window width (O’Neill et al., 2018). Windows that are too short or too long will not capture the true temporal dynamics of the brain. Several alternative approaches have been applied to electrophysiological data to describe the brain dynamics, including hidden Markov models (HMMs) (Vidaurre et al., 2018; Bai et al., 2021a) and microstate analysis (Pascual-Marqui et al., 1995; van de Ville et al., 2010; Zanesco et al., 2020). Spatially distinct patterns of oscillatory power and coherence at the source level have been observed in health and disorders of consciousness by HMMs (Vidaurre et al., 2018; Bai et al., 2021a). Nonetheless, it remains unclear whether the assumptions underlying HMMs are met in resting-state EEG (Gschwind et al., 2015; von Wegner et al., 2016). In addition, little is known about the electromagnetic properties of the brain injury regions, and techniques for constructing accurate head models of patients with brain injuries (e.g., stroke) have not yet been well developed. Microstates reflect short periods (∼100 ms) of quasi-stable brain states that evolve in time, resulting from the synchronous and coordinated activity of brain networks (Pascual-Marqui et al., 1995; Koenig et al., 2002; Zanesco et al., 2020). Importantly, microstate analysis is traditionally at the sensor level and needs minimal assumptions regarding the properties of the neural signals.

Numerous studies based on EEG have demonstrated that resting-state functional connectivity and microstate analyses are successful and valuable methods for studying neurological and psychiatric diseases (Khanna et al., 2015; Zappasodi et al., 2017; Zuchowicz et al., 2018; Eldeeb et al., 2019; Musaeus et al., 2019a; Riahi et al., 2020; Tait et al., 2020). Additionally, there are associations between EEG microstates and fMRI resting-state networks (RSNs) (Britz et al., 2010; van de Ville et al., 2010; Custo et al., 2017). Furthermore, motor scores are related to EEG features (e.g., functional connectivity) of spontaneous brain activity (Riahi et al., 2020; Hoshino et al., 2021). Given the tight link between microstates and brain networks of spontaneous brain activity, we speculated that microstate dynamics could reflect the motor capacity (e.g., lower/upper limb function) to some extent (Pirondini et al., 2017; Spisak et al., 2020; Zhang et al., 2021). However, the current microstate studies have mainly focused on the cognitive functions of the brain in health (Brechet et al., 2019; Pirondini et al., 2020; Zanesco et al., 2020) and diseases like schizophrenia (da Cruz et al., 2020) and Alzheimer (Musaeus et al., 2019b; Tait et al., 2020). Only a limited number of studies focused on patients with brain injuries, such as stroke (Zappasodi et al., 2017) and consciousness of disorders (Gui et al., 2020; Bai et al., 2021b). To date, only one study (to our knowledge) with the complete microstate analysis (Zappasodi et al., 2017) focused on stroke (at the acute stage). This research suggested that microstate B duration explained the 11% of the effective recovery of the National Institute of Health Stroke Scale (NIHSS), and there was a significant difference between patients with left-sided and right-sided stroke in parameters of microstate C and D. Further in-depth research is needed to verify and explore the stroke-related alterations in microstate dynamics.

There are still three critical questions that need to be deeply investigated. First, is each microstate class associated with a specific functional connectivity pattern? Several studies have explored state-wise functional connectivity in different scenarios (Comsa et al., 2019; Duc and Lee, 2019), but the connectivity pattern corresponding to each microstate class in the resting state is still unclear. In addition, according to previous studies, each microstate template has distinct topographic distribution and bright characteristics (Michel and Koenig, 2018; Li et al., 2021; Zanesco et al., 2021). Therefore, we assumed that each microstate class corresponds to a unique pattern of functional connectivity. It is instructive for addressing this question to extend our understanding of the mechanisms underlying microstates and help us explain microstate dynamics. Second, how do the microstate parameters evolve in time after stroke, and how are they perturbed or altered by stroke? Third, do microstate parameters encode information that reflects the motor capacity of the lower limbs? Lower limb motor function is an important basis for restoring walking function and activities of daily living after stroke. However, currently few studies have assessed lower limb function by neural features derived from EEG (Sebastian-Romagosa et al., 2020; Hoshino et al., 2021). The answers to the latter two questions will help us to explore the crucial characteristics of stroke patients, the possible patterns of brain functional reorganization, and the possibility of using microstate dynamics for functional assessment.

In the current study, we proposed a microstate-based approach and leveraged the EEG datasets of patients at two-time points (i.e., before and after the rehabilitation therapy) and healthy controls to explore the three aforementioned questions. The Lower-extremity part of Fugl-Meyer Assessment (FMAL) was used to evaluate the lower limb functional status of stroke patients at the two-time points (Fugl-Meyer et al., 1975). Previous research has suggested that intra-cortical alpha oscillations primarily account for the emergence of microstate classes (Milz et al., 2017), and most microstate studies are based on larger bandwidths such as 2–20 or 1–40 Hz (Khanna et al., 2015). Here, we used the EEG signal in the 1–20 Hz frequency band for microstate analysis. Phase-locked synchronization of neural signals in the brain may be the key mechanism for brain information integration (Michel and Koenig, 2018). Therefore, in this study, functional connectivity in the alpha band was constructed by the phase-locking value (PLV) method (Lachaux et al., 1999; Duc and Lee, 2019; Yao et al., 2021). Microstate-specific connectivity was computed by connected instantaneous phase signals belonging to a particular microstate class. We compared one microstate-specific functional connectivity with the others to examine whether there were differences among microstate-wise connectivity. The results were also validated in an independent EEG dataset (Babayan et al., 2019). In addition, due to the functional and structural deficits after stroke, we predicted that some microstate parameters would be significantly different in stroke patients compared to those in healthy controls. We also tested how microstate features changed after rehabilitation training, and whether there were significant differences in microstate features between patients with left-sided and right-sided stroke. Moreover, we utilized support vector machine (SVM) models to investigate their ability to distinguish stroke patients from healthy controls by only microstate features. Finally, we explored the cross-sectional correlations between the microstate features and FMAL scores. This may allow us to identify potential neural markers that provide insight into lower limb function recovery after stroke.

In sum, we explored the functional connectivity patterns underlying microstates and investigated how microstate dynamics are altered in stroke patients at two-time points compared to healthy individuals and the associations between microstate features and FMAL scores. No such study exists for microstates.

More at link.