Skip to main content
Download PDF
Download ePub

Sub-acute stroke demonstrates altered beta oscillation and connectivity pattern in working memory

Journal of NeuroEngineering and Rehabilitation volume 21, Article number: 212 (2024) Cite this article

Abstract

Introduction

Working memory (WM) is suggested to play a pivotal role in relearning and neural restoration during stroke rehabilitation. Using EEG, this study investigated the oscillatory mechanisms of WM in subacute stroke.

Methods

This study included 48 first subacute stroke patients (26 good-recovery, 22 poor-recovery, based on prognosis after a 4-week period) and 24 matched health controls. We examined the oscillatory characteristics and functional connectivity of the 0-back WM paradigm and assessed their associations with prognosis.

Results

Patients of poor recovery are characterised by a loss of significant beta rebound, beta-band connectivity, as well as impaired working memory speed and performances. Meanwhile, patients with good recovery have preserved these capacities to some extent. Our data further identified beta rebound to be closely associated with working memory speed and performances.

Conclusions

We provided novel findings that beta rebound and network connectivity as mechanistic evidence of impaired working memory in subacute stroke. These oscillatory features could potentially serve as a biomarker for brain stimulation technologies in stroke recovery.

Introduction

Stroke is the third most common cause of morbidity and the second most common cause of dementia [1]. The most common symptoms of stroke include physical deficits (such as paralysis, and sensory loss) and dysfunction in learning, memory, and executive functions. When it comes to cognitive function, working memory (WM) has a fundamental role in performing complex behaviours and is associated with poor functional outcomes after a stroke [2, 3].

An abrupt, cliff-like decline in brain function represents a key difference between acute stroke and neurodegenerative cerebral disorders. More importantly, this pattern of cognitive decline exhibits reversibility within a defined temporal framework [4], which is critical for both neural restoration and clinical rehabilitation. In clinical practices, there is also evidence supporting the ability to regain partial functions or develop compensatory skills through the process of learning [5]. Notably, working memory underpins the learning processes [2], making it critical for neural restoration and clinical rehabilitation. It is therefore important to investigate the behavioural and neural dynamics of WM, offering insights for the development of more effective rehabilitation strategies.

Working memory is defined as a multi-component system involved in goal-directed behaviours that involve retaining and manipulating information [6]. N-back task is a well-validated and widely used means of manipulating working memory capacity and its response requirements [7]. Among them, 0-back requires less workload, which is highly suitable for post-stroke individuals with significant cognitive impairment. It includes components such as sustained attention to a stimulus and continuous memory of the instructions. It is noted that clinical assessment of vascular cognitive impairment (VCI) predominantly relies on cognitive scales, with a paucity of concurrent evidence from cognitive tasks such as WM.

Using an electroencephalogram (EEG), previous studies have established the oscillatory mechanisms of WM. Most of the studies have identified increased power in the fast oscillatory ranges such as beta (13–30 Hz) and gamma (30–100 Hz) [8, 9]. In terms of the functioning, gamma oscillation is suggested to store memories, while beta oscillation is closely associated with attention and response selection [9,10,11]. However, the effects of stroke on neural oscillations underlying WM are largely unclear. A recent scoping review indicated a correlation between decreased fast waves (such as beta) and poor cognition following a stroke [12].

This study was designed to investigate the oscillatory characteristics of working memory in subacute stroke patients. Patients in the subacute stroke underwent a visual-spatial WM 0-back with EEG recordings. They were further classified as good-recovery (n = 26) and poor-recovery (n = 22) according to the modified Rankin Score (mRS). We hypothesized that stroke patients would demonstrate lower power in beta and gamma range compared to healthy controls. Moreover, patients with a poor recovery would have a more prominent decrease in beta and gamma oscillation.

Materials and methods

Participants

We initially screened 105 patients, of whom forty-eight met our study criteria and were enrolled. All participants had evidence of stroke on diffusion-weighted MRI scans. Inclusion criteria were as follows: (1) first-time diagnosis of cerebral infarction within the middle cerebral artery region confirmed by computed tomography or magnetic resonance imaging; (2) ages between 18 and 75 years; (3) admission to the hospital within 10 days to 1.5 months of stroke onset; (4) provision of informed consent. Exclusion criteria included: (1) a prior history of stroke; (2) a history of traumatic brain injury or brain tumour; (3) the presence of neuromuscular diseases such as amyotrophic lateral sclerosis or myasthenia gravis; (4) unstable medical condition preventing participation in the clinical trial; (5) metal implants in the head, face, or heart. In addition to the patients, we included a group of twenty-four healthy controls, matched in terms of gender and age with the patients in both the good-recovery and poor-recovery groups (details in the Results section). All participants provided written informed consent, and this study was approved (Approval No. 2023 − 0694) and adhered to the Declaration of Helsinki.

Study design

This was an observational study. Clinical assessments and EEG recordings were performed when patients were initially referred to the rehabilitation department. The Modified Rankin Score (mRS) was also evaluated at the same time. Patients received regular rehabilitation treatments during hospitalisation.

Clinical assessments

Clinical assessments included the NIH Stroke Scale/Score (NIHSS) [13], Mini-Mental State Examination (MMSE) [14], Montreal Cognitive Assessment (MoCA) [15], as well as the Modified Barthel Index [16]. Medical history (mainly hypertension and diabetes) and lifestyle (mainly smoking and drinking habits) of patients were also assessed.

Working memory task

A visual 0-back working memory task was used here given the cognitive capacities of our patients. Each trial started with a white fixation cross presented for 500 ms. A white square (2.6 * 2.6 degree) was then presented against a black background for 2000 ms in one of the eight pre-determined positions (Fig. 2a). A jitter was not used as the duration was quite long (i.e. 2000ms). A white square presented in the upper left was defined as a target stimulus to which participants responded by pressing the left key in the mouse. Stimuli in other positions were non-targets and participants responded by pressing the right key. Button press was performed with the dominant hand of the participants (all right-handed, see Table 1). The task included 24 target trials and 56 non-target trials which were randomly presented. Participants took a practice test for 10 trials before the experiment. The task was presented using E-Prime 2.0 (Psychology Software Tools Inc). Working memory performance was evaluated using the d prime sensitivity (d’; z-transformed values of hit-rate minus false-alarm rate) and accurate reaction time [17].

Table 1 Patients’ demographical and clinical data characteristics
Full size table

EEG recordings

EEG recordings during working memory took place in a temperature-controlled, sound-attenuated, and electrically shielded room. Participants sat in a chair with their eyes opened and looking forward. A 32-channel EEG system (Biosemi ActiveTwo, Netherlands) was used to record continuous EEG according to the international 12–20 system. The sampling rate was set to 2,048 Hz and the impedances were kept below 5 kΩ throughout the recordings.

Data analyses

EEG data were pre-processed offline using custom-written scripts that implement functions from EEGLAB (version 13.6.5b) [18] running under Matlab R2018b (The MathWorks, Inc.). Data from malfunctioning channels were visually inspected and removed. Data were bandpass filtered between 0.5 and 100 Hz and bandstop filtered between 48 and 52 Hz using zero-phase Butterworth filters which could compensate for the filter delay [19]. Continuous data were segmented based on the onset of the image (-500 to 2,000 ms). Segmented data were re-referenced to the average reference, and the fast independent component analysis algorithm (FastICA) was used to remove stereotyped artefacts, e.g. eye blinks, lateral eye movements, muscle, and line noise [20]. Stereotyped artefacts were identified by visual inspection of the temporal and spatial representation of the independent components. Missing channels were then interpolated, and epochs were inspected again to remove any anomalous activity in the signal. Rejected channels were no more than three in each dataset.

Time-frequency analyses were performed using the hanning tapered “mtmconvol” method in the FieldTrip toolbox [21]. Power was calculated in the range of 1–100 Hz in the time window of -500 to 2,000 ms and baseline corrected (-500 ms) for each trial before averaging trials in each condition for each subject. Baseline correction was performed with the equation (frames of interest - baseline) / baseline.

EEG connectivity was calculated by computing the debiased weighted phase lag index (WPLI) based on the frequency representations obtained above. The WPLI is a measure of the phase coherence of two signals, based on the imaginary part of the cross-spectrum [22, 23]. WPLI is suggested to be able to reduce sensitivity to additional, uncorrelated noise sources, such as volume conduction, as well as to increase statistical power to detect changes in the phase-synchronization [22]. For each frequency step, the WPLI was computed for each electrode pair. Connectivity values were then averaged to the frequency and time frames of interest based on results from the power analysis (see Fig. 3, healthy control: 0.88–1.80 s, 13–30 Hz; good-recovery: 1.33–1.78 s, 13–30 Hz; pooled healthy controls and good-recovery: 1.33–1.78 s, 13–30 Hz). Other frequency bands were further examined for exploratory purposes. In order to calculate WPLI, the hanning tapered “mtmconvol” method was initially used for time-frequency analyses, in which the output was specified to ‘fourier’ to generate complex Fourier-spectra and then baseline corrected (-500 ms).

Statistical analyses

Patients were initially grouped into good- (mRS ≤ 2, n = 26) and poor recovery (mRS > 2, n = 22) based on mRS scores. For working memory performances, one-way ANOVAs with three levels (healthy control, good-recovery, poor-recovery) were conducted on reaction time and d prime sensitivity respectively (SPSS, IBM Corp, Armonk, NY, version 22). Multiple comparisons were performed with Bonferroni correction at p < 0.05.

For time-frequency data, non-parametric cluster-based permutation statistics were performed at a global level. This method is suggested to be able to control multiple comparisons across EEG channels and time [24]. The cluster-based permutation tests were applied to the time window of interest (0–2,000 ms) in all channels. Dependent t-tests were initially performed in each group to reveal event-related oscillation changes from baseline. Multiple comparison problems were solved with the alpha level (0.05) being divided by the number of comparisons (3 times here). One-way ANOVAs with three levels (healthy control, good-recovery, poor-recovery) were then conducted to reveal group differences in oscillation. Multiple comparisons were performed with Bonferroni correction at p < 0.05. Statistical analysis was initially performed in beta (13–30 Hz) and gamma (31–100 Hz) bands based on a prior hypothesis [25, 26]. Exploratory analyses were also conducted in other frequency bands (i.e. delta: 1–3 Hz, theta: 4–7 Hz, alpha: 8–12 Hz).

Statistical analyses of EEG connectivity were performed using the network-based statistic (NBS) toolbox [27]. The NBS is a nonparametric technique that uses cluster analysis to perform null hypothesis testing across networks of values from pairs of potentially connected nodes [27]. Dependent t-tests were performed. We generated 5000 permutations for statistical comparisons with a primary threshold set at p < 0.005, in order to guarantee only robust differences in connectivity between electrode pairs to be compared at the cluster level [28,29,3030]. The secondary threshold for electrode pairs was set at p < 0.05 for network-based statistics.

We also performed Pearson’s correlations to build up the relationship between EEG oscillations and working memory performances. Multiple comparison problems were solved with the alpha level (0.05) being divided by the number of correlations (4 times here due to 2 WM performances and 2 EEG features).

Results

Clinical information

Of the 105 stroke patients, we excluded 42 patients and enrolled 63 patients. Among the 63 patients, 10 lost contacts after discharge, and another 5 had other lesions in MRI data (e.g. small vessel disease signs, meningioma, or ventricular enlargement). We therefore included 48 patients (77%) in the outcome analysis, comprising 37 men and 11 women with a mean age of 59.3 ± 12.4 years (Fig. 1). Among them, 26 patients showed improved motor function (MRS ≤ 2, good-recovery), while 22 exhibited no significant change (MRS > 2, poor-recovery) (Fig. 1).

Fig. 1
figure 1

Flowchart for patient inclusion and exclusion

Full size image

Table 1 presents an overview of the clinical characteristics of enrolled patients. Among them, 40 had ischemic stroke, while 8 had hemorrhagic stroke. The total follow-up period ranged from 30 to 45 days, with an average of 36.8 days (± 7.1). All EEG recordings were obtained within 6 weeks, with a mean time of 24.3 days (± 14.5) post-stroke onset.

Upon admission, the NIHSS score for neurological impairment was 5.1 (± 2.4) across all patients. Remarkably, significant variations in NIHSS scores were observed among different prognostic groups (p = 0.000). Similar intergroup disparities were noted in cognitive function (MMSE and MoCA) and self-care ability (p = 0.012, p = 0.002, p = 0.000, respectively). Nevertheless, no significant differences were detected between the groups in age, gender, lesion location, lesion volume, cerebrovascular disease risk factors, or lifestyle.

Working memory performance

One-way ANOVA revealed a significant group effect on reaction time (F2, 69 = 6.04, P = 0.004). Post-hoc comparisons indicated longer reaction time in patients of poor-recovery (Mean = 899.29) compared to those of good-recovery (Mean = 749.84, Pcorrected = 0.034), and to healthy controls (Mean = 703.68, Pcorrected = 0.004) (Fig. 2b). There was no difference in reaction time between patients of good-recovery and healthy controls (Pcorrected > 0.5).

Fig. 2
figure 2

Working memory task and performances. a) An example of non-target and target stimulus. b) Results of reaction time. Patients with poor recovery had longer reaction time compared to those with good recovery and healthy controls. There was no difference between patients with good-recovery and healthy controls. c). Results of d prime sensitivity. Healthy controls had higher d prime sensitivity compared to patients with good recovery and to those with poor recovery. There was no difference between patients with good recovery and with poor recovery (Pcorrected > 0.5). * denotes p < 0.05, ** denotes p < 0.01

Full size image

In terms of d prime sensitivity, One-way ANOVA also revealed a significant group effect (F2, 69 = 4.68, P = 0.012). Post-hoc comparisons indicated higher d prime sensitivity in healthy controls (Mean = 189.56) compared to patients of good-recovery (Mean = 120.22, Pcorrected = 0.048) and to those of poor-recovery (Mean = 107.89, Pcorrected = 0.020) (Fig. 2c). There was no difference in d prime sensitivity between patients of good-recovery and of poor-recovery (Pcorrected > 0.5).

EEG oscillation

In the three datasets, the number of channels was rejected were 2.43 ± 0.53 (poor-recovery), 2.25 ± 0.46 (good-recovery), and 2.17 ± 0.45 (healthy control), respectively. The number of epochs were 76.95 ± 2.68 (poor-recovery), 76.50 ± 3.13 (good-recovery), and 76.79 ± 3.59 (healthy control), respectively.

Cluster-based permutation tests revealed a beta inhibition in the central regions in each group compared to baseline (healthy controls: Pcorrected = 0.004, time range = 0.17–0.73 s; good-recovery: Pcorrected = 0.003, time range = 0.22–0.81 s; poor-recovery: Pcorrected = 0.000, time range = 0.20–0.97 s) (Fig. 3a-c). A list of significant electrodes was listed here. healthy controls: C4, CP1, CP2, CP6, P3; good-recovery: FC1, FC2, C3, C4, CP5, CP1, CP2; poor-recovery: FC5, FC1, C3, C4, CP5, CP1, CP6, P3, P4.

Fig. 3
figure 3

Event-related changes from baseline. a-c) All three groups demonstrated a decrease in beta power in the central regions (healthy controls: Pcorrected = 0.004, time range = 0.17–0.73 s; good-recovery: Pcorrected = 0.003, time range = 0.22–0.81 s; poor-recovery: Pcorrected = 0.000, time range = 0.20–0.97 s). There was also an increased late beta power in the frontal and parieto-occipital regions in the healthy controls and patients of good-recovery (healthy controls: Pcorrected = 0.000, time range = 0.88–1.80 s; good-recovery: Pcorrected = 0.014, time range = 1.33–1.78 s). This pattern of beta oscillation was not significant in patients with poor recovery (Pcorrected > 0.05). Tagged electrodes in each topolot indicated the significant channels revealed by cluster statistics. The time-frequency representations were plotted with the significant electrodes in the topoplot of (c), or the shared significant electrodes in the two topoplots in (a, b). X denotes p < 0.05. * denotes p < 0.01, dB = decibel

Full size image

In addition to beta inhibition, there was also a beta rebound in the frontal and parieto-occipital regions only in the healthy controls and patients of good-recovery relative to baseline (healthy controls: Pcorrected = 0.000, time range = 0.88–1.80 s; good-recovery: Pcorrected = 0.014, time range = 1.33–1.78 s). This beta increase was not significant in patients of poor recovery (Pcorrected > 0.5) (Fig. 3a-c). A list of significant electrodes was listed here. healthy controls: AF3, F7, F8, F4, FC6, T7, T8, P7, P3, P8, PO3, PO4, O1, O2; good-recovery: AF3, AF4, F7, F3, F4, FC2, T7, CP5, P7, P3, P4, P8, PO3, PO4, O1, O2.

One-way ANOVAs also revealed group differences in beta oscillations. Specifically, patients of poor recovery demonstrated a larger beta inhibition in the frontal and parieto-occipital regions compared to healthy controls and to those of good recovery (healthy controls: Pcorrected = 0.010, time range = 0.72–1.10 s, Electrodes = F3, F8, FC1, FC6, T8, PO4, P8, O2; good-recovery: Pcorrected = 0.021, time range = 0.70–0.87 s, Electrodes = AF3, AF4, F8, F3, CP6, P7, P3, P4, P8, PO3, PO4, O1, O2) (Fig. 4a-b).

Fig. 4
figure 4

Group comparisons EEG oscillations. a-b) Patients of poor-recovery demonstrated a larger decrease in early beta power in the frontal and parieto-occipital regions compared to healthy controls and to those of good-recovery (healthy controls: Pcorrected = 0.010, time range = 0.72–1.10 s; good-recovery: Pcorrected = 0.021, time range = 0.70–0.87 s). c-d) Healthy controls demonstrated a larger increase in late beta power in the frontal and parieto-occipital regions compared to patients of good-recovery and to those of poor-recovery (good-recovery: Pcorrected = 0.010, time range = 0.90–1.70 s; poor-recovery: Pcorrected = 0.003, time range = 0.81–1.78 s). Tagged electrodes in each topolot indicated the significant channels revealed by cluster statistics. The time-frequency representations were plotted with the significant electrodes in the topoplot. X denotes p < 0.05. * denotes p < 0.01, dB = decibel

Full size image

In addition, healthy controls demonstrated a larger beta rebound in the frontal and parieto-occipital regions compared to patients of good recovery and to those of poor recovery (good-recovery: Pcorrected = 0.010, time range = 0.90–1.70 s, Electrodes = AF3, F8, FC6, T8, P8, PO3, PO4, O1, O2; poor-recovery: Pcorrected = 0.003, time range = 0.81–1.78 s, Electrodes = AF3, F7, F8, FC5, FC1, FC6, T7, T8, C3, CP6, P8, PO4, O2) (Fig. 4c-d). Patients of good recovery and those of poor recovery were not different in the late beta power (no cluster survived p < 0.05).

Correlation results

Beta rebound in the healthy controls and patients of good-recovery was associated with faster reaction (Fig. 5a), as well as with higher d prime sensitivity (Fig. 5b). Meanwhile, patients of poor-recovery were not included in this correlation as there was not significant beta rebound in the event-related oscillations. In contrast, beta inhibition was not associated with either reaction time or d prime sensitivity (all P > 0.05) (Fig. 5c-d).

Fig. 5
figure 5

Correlation results. a-b) Increased beta power in the healthy controls and patients of good-recovery was associated with the faster reaction, as well as with higher d prime sensitivity. In contrast, decreased beta power in the early phase was not associated with either reaction time or d prime sensitivity. All EEG signals were extracted from the significant clusters in the event-related analysis

Full size image

EEG connectivity

As beta rebound in the frontal and parieto-occipital regions was associated with both reaction time and d prime sensitivity, EEG connectivity was mainly analysed in this increased beta cluster. In line with event-related oscillations, in healthy controls EEG connectivity data demonstrated increased beta band WPLI compared to baseline (Pcorrected = 0.048) between the frontal regions (F4, F8) and parieto-occipital regions (P7, O1) (Fig. 6a). In the pooled dataset of healthy controls and good-recovery, EEG connectivity data indicated increased beta band WPLI (Pcorrected = 0.013) between the frontal regions (AF3, FP1, FP2, F8) and parieto-occipital regions (O1, PO3), as well as interactions between parietal and occipital regions (P3, P7, PO3, O1) (Fig. 6b). We also extended the analysis to other frequency band and the results were not significant (Pcorrected > 0.05).

Fig. 6
figure 6

EEG connectivity analysis. a) In healthy controls, EEG connectivity data demonstrated increased beta band WPLI (Pcorrected = 0.048) between the frontal regions (F4, F8) and parieto-occipital regions (P7, O1). b) In the pooled dataset of healthy controls and good-recovery, EEG connectivity data indicated increased beta band WPLI (Pcorrected = 0.013) between the frontal regions (AF3, FP1, FP2, F8) and parieto-occipital regions (O1, PO3), as well as increased interactions between parietal and occipital regions (P3, P7, PO3, O1). L and R denote left and right hemispheres respectively

Full size image

Discussion

This study investigated the neural oscillatory characterises of working memory in subacute stroke patients with good or poor recovery. We presented novel findings that patients of poor recovery are characterised by a loss of significant beta rebound, beta-band connectivity, as well as impaired working memory speed and performance. Meanwhile, patients of good recovery have preserved these capacities to some extent. Our data further identified beta rebound to be closely associated with working memory speed and performance.

Our data presented novel findings that subacute stroke is characterised by altered beta rebound in working memory. Moreover, head-to-head comparisons revealed a decreased beta rebound in patients with a good recovery compared to healthy controls, and even a complete disappearance of significant beta rebound in those with a poor recovery. Beta rebound in the healthy controls and good-recovery patients were further found to be associated with faster and better working memory performances.

A series of studies have identified the involvement of beta oscillation in working memory [10, 31,32,33]. For instance, beta power was found to increase in a memory condition compared with a control condition [11]. In terms of functioning, these findings suggest beta oscillation to support sustained attention and/or response selection. However, this line of evidence has mainly come from healthy controls. A few studies have been carried out on mild cognitive impairment (MCI), which identified an absence of significant beta oscillation in MCI, or even a desynchronisation in beta oscillation [34, 35]. Built on these studies, our data presented an impaired beta rebound in subacute stroke individuals. Moreover, beta rebound further disappeared in poor-recovery patients, presenting a covarying pattern of stroke disability and beta oscillation. It is noted that beta rebound appeared after the release of a button press, coinciding with the time frames for attending to and performing the next trial. This finding further corroborates the role of beta oscillation in sustained attention. Our findings therefore highlight beta rebound as a mechanistic evidence of impaired working memory in subacute stroke.

Beyond beta oscillation, our data further indicated an absence of beta band connectivity indexed by WPLI in poor-recovery patients, which was identified in healthy controls and patients of good recovery. A line of evidence indicated that beta band synchronisation could be influenced by working memory demand or load [36,37,38,39]. Another study observed increased beta band synchronisation in the retention of working memory tasks [40], with the same feature being induced by increased attention [41]. Our findings further indicate an impaired network connectivity in working memory particularly in stroke patients with a poor-recovery. Moreover, this network was characterised by connectivity losses between the frontal and parieto-occipital regions. In one way, these regions are highly consistent with the spatial distribution in beta rebound. In another way, the frontal and parieto-occipital regions are extensively involved in high-level cognitions such as working memory.

We also presented interesting findings that poor-recovery patients demonstrated a larger beta inhibition compared to those of good recovery and healthy controls. Task-induced beta oscillations have been well-documented in recent literature [42, 43]. Notably, Motion-Related Beta Inhibition (MRBI) is one of the most prominent manifestations [44]. This is also supported by the associations between beta power and local concentrations of the inhibitory neurotransmitter gamma-aminobutyric acid (GABA) [45,46,47,48]. Our data demonstrated a consistent beta inhibition in all three groups. Moreover, patients of poor recovery had the most prominent and long-lasting beta inhibition, coinciding with the longest reaction time in the task performance. Our findings thus highlight beta inhibition to underlie impaired response execution in stroke patients of poor recovery. It is also interesting to find that patients of good recovery had equal reaction time and beta inhibition to healthy controls, but demonstrated a smaller d’ sensitivity and lower beta rebound. This finding further corroborates the role of beta rebound in working memory performances. It is acknowledged that no significant difference in beta rebound was found between patients of poor and good recovery. It is clear that beta rebound was preserved in good-recovery patients to some extent but disappeared in those of poor recovery. A larger sample size would reveal a significant cluster between these two groups.

Notably, there is also evidence suggesting low-frequency oscillations as a biological marker for stroke recovery [49]. However, this data was recorded as resting-state EEG, making it difficult to directly compare them with the task-induced oscillations in our study. In addition, we did not observe significant gamma oscillation in our working memory task. One possibility is that our 0-back task requires relatively less workload or cognitive resources. This task was tailored to the cognitive capacities of our patients. However, this null finding does not overshadow our novel findings on altered beta inhibition and rebound in stroke patients.

Our findings have implications for stroke rehabilitation. We have identified late beta rebound to be closely correlated with working memory performances. Moreover, patients of poor recovery failed to show a significant beta rebound. These findings indicate the potential to modulate beta power for stroke rehabilitation with brain stimulation technologies, such as transcranial magnetic stimulation (TMS) and transcranial direct current stimulation (tDCS). Patients of good recovery demonstrated preserved beta rebound and working memory performances in our data. These findings suggest the need to design early and tailored rehabilitation strategies for those patients. In addition, our data indicate the significance of working memory in the assessment of stroke patients. It can be used to supplement clinical assessments, and thus provide insights into the design and assessment of rehabilitation strategies.

This study presents several limitations that warrant acknowledgment. The cohort in this study comprised stroke patients with relatively mild degrees of neurological impairment, and the sample size was relatively small. The 2-second observation window employed in this study may not fully capture specific post-task neural oscillatory phenomena. For instance, it does not account for the post-task response (PTR), a transitional reaction lasting several seconds between the conclusion of cognition/movement and the onset of rest. Investigating the intrinsic relationship between PTR and beta inhibition and rebound warrants further research. The relatively short follow-up duration represents a limitation. It is therefore important to validate our findings in future longitudinal studies.

Conclusion

We provided novel findings that beta rebound and network connectivity as mechanistic evidence of impaired working memory in subacute stroke. These oscillatory features also demonstrate different patterns in subacute stroke patients with good and poor recovery.

Data availability

No datasets were generated or analysed during the current study.

References

  1. Unibaso-Markaida I, Iraurgi I, Ortiz-Marqués N, Amayra I, Martínez-Rodríguez S. Effect of the Wii sports Resort on the improvement in attention, processing speed and working memory in moderate stroke. J Neuroeng Rehabil. 2019;16:1–8.

    Article  Google Scholar 

  2. Lindeløv JK, Overgaard R, Overgaard M. Improving working memory performance in brain-injured patients using hypnotic suggestion. Brain. 2017;140(4):1100–6.

    Article  PubMed  Google Scholar 

  3. Synhaeve NE, Schaapsmeerders P, Arntz RM, Maaijwee NA, Rutten-Jacobs LC, Schoonderwaldt HC, et al. Cognitive performance and poor long-term functional outcome after young stroke. Neurology. 2015;85(9):776–82.

    Article  PubMed  Google Scholar 

  4. Levine DA, Galecki AT, Langa KM, Unverzagt FW, Kabeto MU, Giordani B, et al. Trajectory of cognitive decline after Incident Stroke. Jama-Journal Am Med Association. 2015;314(1):41–51.

    Article  CAS  Google Scholar 

  5. Espenhahn S, Rossiter HE, van Wijk BCM, Redman N, Rondina JM, Diedrichsen J et al. Sensorimotor cortex beta oscillations reflect motor skill learning ability after stroke. Brain Commun. 2020;2(2).

  6. Lugtmeijer S, Lammers NA, de Haan EH, de Leeuw F-E, Kessels RP. Post-stroke working memory dysfunction: a meta-analysis and systematic review. Neuropsychol Rev. 2021;31:202–19.

    Article  PubMed  Google Scholar 

  7. Kas MJ, Serretti A, Marston H. Quantitative neurosymptomatics: linking quantitative biology to neuropsychiatry. Neurosci Biobehav Rev. 2019;97:1–2.

    Article  PubMed  Google Scholar 

  8. Spitzer B, Haegens S. Beyond the status quo: a role for beta oscillations in endogenous content (re) activation. Eneuro. 2017;4(4).

  9. Lundqvist M, Rose J, Herman P, Brincat SL, Buschman TJ, Miller EK. Gamma and beta bursts underlie working memory. Neuron. 2016;90(1):152–64.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Lundqvist M, Herman P, Warden MR, Brincat SL, Miller EK. Gamma and beta bursts during working memory readout suggest roles in its volitional control. Nat Commun. 2018;9(1):394.

    Article  PubMed  PubMed Central  Google Scholar 

  11. Smyrnis N, Protopapa F, Tsoukas E, Balogh A, Siettos CI, Evdokimidis I. Amplitude spectrum EEG signal evidence for the dissociation of motor and perceptual spatial working memory in the human brain. Exp Brain Res. 2014;232:659–73.

    Article  PubMed  Google Scholar 

  12. Sutcliffe L, Lumley H, Shaw L, Francis R, Price CI. Surface electroencephalography (EEG) during the acute phase of stroke to assist with diagnosis and prediction of prognosis: a scoping review. BMC Emerg Med. 2022;22(1):29.

    Article  PubMed  PubMed Central  Google Scholar 

  13. Brott T, Adams HP Jr, Olinger CP, Marler JR, Barsan WG, Biller J, et al. Measurements of acute cerebral infarction: a clinical examination scale. Stroke. 1989;20(7):864–70.

    Article  CAS  PubMed  Google Scholar 

  14. Folstein MF, Folstein SE, Mchugh PR. Mini-mental state: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12(3):0–198.

    CAS  Google Scholar 

  15. Nasreddine ZS, Phillips NA, Bédirian V, Charbonneau S, Whitehead V, Collin I, et al. The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc. 2005;53(4):695–9.

    Article  PubMed  Google Scholar 

  16. Mahoney FI. Functional evaluation: the Barthel index. Maryland State Med J. 1965;14(2):61–5.

    CAS  Google Scholar 

  17. Haatveit BC, Sundet K, Hugdahl K, Ueland T, Melle I, Andreassen OA. The validity of d prime as a working memory index: results from the Bergen n-back task. J Clin Exp Neuropsychol. 2010;32(8):871–80.

    Article  PubMed  Google Scholar 

  18. Delorme A, Makeig S. EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. J Neurosci Methods. 2004;134(1):9–21.

    Article  PubMed  Google Scholar 

  19. Widmann A, Schröger E, Maess B. Digital filter design for electrophysiological data – a practical approach. J Neurosci Methods. 2015;250:34–46.

    Article  PubMed  Google Scholar 

  20. Korhonen RJ, Hernandez-Pavon JC, Metsomaa J, Mki H, Ilmoniemi RJ, Sarvas J. Removal of large muscle artifacts from transcranial magnetic stimulation-evoked EEG by independent component analysis. Med Biol Eng Comput. 2011;49(4):397–407.

    Article  PubMed  Google Scholar 

  21. Oostenveld R, Fries P, Maris E, Schoffelen J-M, FieldTrip. Open Source Software for Advanced Analysis of MEG, EEG, and Invasive Electrophysiological Data. Computational Intelligence and Neuroscience. 2011;2011.

  22. Vinck M, Oostenveld R, Van Wingerden M, Battaglia F, Pennartz CM. An improved index of phase-synchronization for electrophysiological data in the presence of volume-conduction, noise and sample-size bias. NeuroImage. 2011;55(4):1548–65.

    Article  PubMed  Google Scholar 

  23. Che X, Cash R, Chung SW, Bailey N, Fitzgerald PB, Fitzgibbon BM. The dorsomedial prefrontal cortex as a flexible hub mediating behavioral as well as local and distributed neural effects of social support context on pain: a Theta Burst Stimulation and TMS-EEG study. NeuroImage. 2019;201:116053.

    Article  PubMed  Google Scholar 

  24. Maris E, Oostenveld R. Nonparametric statistical testing of EEG-and MEG-data. J Neurosci Methods. 2007;164(1):177–90.

    Article  PubMed  Google Scholar 

  25. Coleman SC, Seedat ZA, Whittaker AC, Lenartowicz A, Mullinger KJ. Beyond the Beta Rebound: post-task responses in Oscillatory Activity follow Cessation of Working memory processes. NeuroImage. 2023;265.

  26. Li Y, Noguchi Y. Neural correlates of a load-dependent decline in visual working memory. Cereb Cortex Commun. 2022;3(2):tgac015–tgac.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Zalesky A, Fornito A, Bullmore ET. Network-based statistic: identifying differences in brain networks. NeuroImage. 2010;53(4):1197–207.

    Article  PubMed  Google Scholar 

  28. Bailey NW, Hoy KE, Rogasch NC, Thomson RH, McQueen S, Elliot D, et al. Responders to rTMS for depression show increased fronto-midline theta and theta connectivity compared to non-responders. Brain Stimul. 2018;11(1):190–203.

    Article  CAS  PubMed  Google Scholar 

  29. Luo X, Che X, Li H. Concurrent TMS-EEG and EEG reveal neuroplastic and oscillatory changes associated with self-compassion and negative emotions. Int J Clin Health Psychol. 2023;23(1):100343.

    Article  PubMed  Google Scholar 

  30. Ye Y, Wang J, Che X. Concurrent TMS-EEG to reveal the neuroplastic changes in the Prefrontal and Insular cortices in the Analgesic effects of DLPFC-rTMS. Cereb Cortex. 2022.

  31. Coleman SC, Seedat ZA, Whittaker AC, Lenartowicz A, Mullinger KJ. Beyond the beta rebound: post-task responses in oscillatory activity follow cessation of working memory processes. NeuroImage. 2023;265:119801.

    Article  PubMed  Google Scholar 

  32. Li Y, Noguchi Y. Neural correlates of a load-dependent decline in visual working memory. Cereb Cortex Commun. 2022;3(2):tgac015.

    Article  PubMed  PubMed Central  Google Scholar 

  33. Kornblith S, Buschman TJ, Miller EK. Stimulus load and oscillatory activity in higher cortex. Cereb Cortex. 2016;26(9):3772–84.

    Article  PubMed  PubMed Central  Google Scholar 

  34. Güntekin B, Emek-Savaş DD, Kurt P, Yener GG, Başar E. Beta oscillatory responses in healthy subjects and subjects with mild cognitive impairment. NeuroImage: Clin. 2013;3:39–46.

    Article  PubMed  Google Scholar 

  35. Fraga FJ, Mamani GQ, Johns E, Tavares G, Falk TH, Phillips NA. Early diagnosis of mild cognitive impairment and Alzheimer’s with event-related potentials and event-related desynchronization in N-back working memory tasks. Comput Methods Programs Biomed. 2018;164:1–13.

    Article  PubMed  Google Scholar 

  36. Tallon-Baudry C, Bertrand O, Fischer C. Oscillatory synchrony between human extrastriate areas during visual short-term memory maintenance. J Neurosci. 2001;21(20):RC177.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Tallon-Baudry C, Mandon S, Freiwald WA, Kreiter AK. Oscillatory synchrony in the monkey temporal lobe correlates with performance in a visual short-term memory task. Cereb Cortex. 2004;14(7):713–20.

    Article  PubMed  Google Scholar 

  38. Axmacher N, Schmitz DP, Wagner T, Elger CE, Fell J. Interactions between medial temporal lobe, prefrontal cortex, and inferior temporal regions during visual working memory: a combined intracranial EEG and functional magnetic resonance imaging study. J Neurosci. 2008;28(29):7304–12.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  39. Dotson NM, Salazar RF, Gray CM. Frontoparietal correlation dynamics reveal interplay between integration and segregation during visual working memory. J Neurosci. 2014;34(41):13600–13.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Sarnthein J, Petsche H, Rappelsberger P, Shaw GL, von Stein A. Synchronization between prefrontal and posterior association cortex during human working memory. Proc Natl Acad Sci. 1998;95(12):7092–6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Makeig S, Jung T-P. Tonic, phasic, and transient EEG correlates of auditory awareness in drowsiness. Cogn Brain Res. 1996;4(1):15–25.

    Article  CAS  Google Scholar 

  42. ElShafei HA, Zhou YJ, Haegens S. Shaping Information Processing: the role of Oscillatory Dynamics in aWorking Memory Task. Eneuro. 2022;9(5).

  43. Gudi-Mindermann H, Rimmele JM, Bruns P, Kloosterman NA, Donner TH, Engel AK et al. Post-training load-related changes of auditory Working Memory - An EEG Study. Front Hum Neurosci. 2020;14.

  44. Fry A, Mullinger KJ, O’Neill GC, Barratt EL, Morris PG, Bauer M, et al. Modulation of Post-movement Beta Rebound by Contraction Force and Rate of Force Development. Hum Brain Mapp. 2016;37(7):2493–511.

    Article  PubMed  PubMed Central  Google Scholar 

  45. Gaetz W, Edgar JC, Wang DJ, Roberts TPL. Relating MEG measured motor cortical oscillations to resting gamma-aminobutyric acid (GABA) concentration. NeuroImage. 2011;55(2):616–21.

    Article  CAS  PubMed  Google Scholar 

  46. Hall SD, Stanford IM, Yamawaki N, McAllister CJ, Ronnqvist KC, Woodhall GL, et al. The role of GABAergic modulation in motor function related neuronal network activity. NeuroImage. 2011;56(3):1506–10.

    Article  CAS  PubMed  Google Scholar 

  47. Jensen O, Goel P, Kopell N, Pohja M, Hari R, Ennentrout B. On the human sensorimotor-cortex beta rhythm: sources and modeling. NeuroImage. 2005;26(2):347–55.

    Article  CAS  PubMed  Google Scholar 

  48. Muthukumaraswamy SD, Myers JFM, Wilson SJ, Nutt DJ, Lingford-Hughes A, Singh KD, et al. The effects of elevated endogenous GABA levels on movement-related network oscillations. NeuroImage. 2013;66:36–41.

    Article  CAS  PubMed  Google Scholar 

  49. Cassidy JM, Wodeyar A, Wu JN, Kaur K, Masuda AK, Srinivasan R, et al. Low-frequency oscillations are a biomarker of Injury and Recovery after Stroke. Stroke. 2020;51(5):1442–50.

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

Not applicable.

Funding

This work was supported by the Key R&D Program of Zhejiang Province under grant number 2022C03038:BLB19J014.

Author information

Author notes

  1. Lin Mao, Xianwei Che and Juehan Wang contributed equally to this work.

Authors and Affiliations

  1. Departments of Physical Medicine and Rehabilitation, The First Affiliated Hospital, Zhejiang University School of Medicine, Building 6, 58 Chengzhan Road, Hangzhou, 310003, China

    Lin Mao, Juehan Wang, Yifan Zhao, Liliang Zou, Shuang Wei, Xueqiong Zhu, Zuobing Chen & Daming Wang

  2. Centre for Cognition and Brain Disorders, The Affiliated Hospital of Hangzhou Normal University, Hangzhou, 310003, China

    Xianwei Che

  3. Department of Rehabilitation Medicine, The First People’s Hospital of Yuhang District, Hangzhou, 311100, China

    Xiaorui Jiang

  4. Department of Hyperbaric Oxygen, The Sixth Medical Center of PLA General Hospital, Beijing, 100142, China

    Shuyi Pan & Dazhi Guo

  5. Departments of Physical Medicine and Rehabilitation, The Second Affiliated Hospital, Nanchang University School of Medicine, Nanchang, 330038, China

    Dongxia Hu

  6. Emergency and Trauma Center, The First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, 310003, China

    Xiaofeng Yang

Authors
  1. Lin Mao
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Xianwei Che
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Juehan Wang
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Xiaorui Jiang
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Yifan Zhao
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Liliang Zou
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Shuang Wei
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Shuyi Pan
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Dazhi Guo
    View author publications

    You can also search for this author inPubMed Google Scholar

  10. Xueqiong Zhu
    View author publications

    You can also search for this author inPubMed Google Scholar

  11. Dongxia Hu
    View author publications

    You can also search for this author inPubMed Google Scholar

  12. Xiaofeng Yang
    View author publications

    You can also search for this author inPubMed Google Scholar

  13. Zuobing Chen
    View author publications

    You can also search for this author inPubMed Google Scholar

  14. Daming Wang
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

Lin Mao conceived of the study and made substantial contributions to conception and design of the study, performed the statistical analysis, made the analysis and interpretation ofdata and drafted the manuscript. Xianwei Che participated in the design of the study, performed statistical analysis. Juehan Wang participated in the design of the study and helped todraft the manuscript and collected the data. Xiaorui Jiang participated in the design of the study, performed statistical analysis and collected the data. Yifan Zhao and Liliang Zou and Shuang Wei participated in the selection and medical care of the patients, coordination, data collection and interpretation. Shuyi Pan and Dazhi Guo participated in the design of thestudy and interpretation of data and revised the manuscript critically for important intellectual content. Xueqiong Zhu and Dongxia Hu and Xiaofeng Yang participated in the design of the study. Zuobing Chen and Daming Wang participated in the interpretation of data, drafting the manuscript and revising it critically forimportant intellectual content. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Zuobing Chen or Daming Wang.

Ethics declarations

Ethics approval and consent to participate

This study was approved by the First Affiliated Hospital of Zhejiang University School of Medicine (Approval No. 2023 − 0694) and adhered to the Declaration of Helsinki.

Consent for publication

All participants provided written informed consent.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Mao, L., Che, X., Wang, J. et al. Sub-acute stroke demonstrates altered beta oscillation and connectivity pattern in working memory. J NeuroEngineering Rehabil 21, 212 (2024). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12984-024-01516-5

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12984-024-01516-5

Keywords

Journal of NeuroEngineering and Rehabilitation

ISSN: 1743-0003

Contact us