Analysis of complexity and dynamic functional connectivity based on resting-state EEG in early Parkinson’s disease patients with mild cognitive impairment

The temporal-insula type of temporal plus epilepsy patients with different postoperative seizure outcomes have different cerebral blood flow patterns

PURPOSE

This study retrospectively analyzed preoperative arterial spin labeling (ASL) perfusion MRI data of patients with the temporal-insula type of temporal plus epilepsy (TI-TPE). We aimed to investigate the differences in presurgical cerebral blood flow (CBF) changes in TI-TPE patients with different surgical outcomes.

METHOD

A total of 48 TI-TPE patients confirmed by SEEG were meticulously reviewed for this study. Patients were divided into the seizure-free (SF) group (Engel IA) and the non-seizure-free (NSF) group (Engel IB to IV) according to the Engel seizure classification. The 3D-ASL data of all patients before surgery were analyzed using statistical parametric mapping (SPM) and graph theory analysis. These findings were then compared to healthy controls (HC) based on whole-brain voxel-level analysis and covariance network analysis.

RESULT

At the voxel-level, both SF and NSF groups showed significantly decreased CBF in the ipsilateral transverse temporal gyrus and insula (TTG/insula), contralateral middle cingulate gyrus, precuneus (MCG/precuneus), and increased CBF in the ipsilateral superior temporal gyrus and the superior temporal pole (STG/STP). Wherein the SF group showed more lower CBF in the contralateral MCG/precuneus, with unique increased CBF in the contralateral STG/insula and decreased CBF in the contralateral calcarine as well. In terms of network attributes, the NSF group showed a significantly higher clustering coefficient (Cp), global efficiency (Eglob), local efficiency (Eloc), shorter shortest path length (Lp), and more extensive abnormal nodes compared to the SF and HC groups. While the SF group has higher synchronicity than the HC group.

CONCLUSION

Both SF and NSF groups had abnormal CBF changes at the voxel and network levels with different patterns. The SF group showed more obvious regional CBF changes, while the NSF group showed more extended network disruption, which might underlie different seizure outcomes after local surgical resection.

Sequential patterning of dynamic brain states distinguish Parkinson's disease patients with mild cognitive impairments

Parkinson's disease (PD) is a neurodegenerative disease which presents clinically with progressive impairments in motoric and cognitive functioning. Pathophysiologic mechanisms underlying these impairments are believed to be attributable to a breakdown in the spatiotemporal coordination of functional neural networks across multiple cortical and subcortical regions. The current investigation used resting state, functional magnetic resonance imaging (rs-fMRI) to determine whether the temporal characteristics or sequential patterning of dynamic functional network connectivity (dFNC) states could accurately distinguish among people with PD who had normal cognition (PD-NC, n = 18), those with PD who had mild cognitive impairment (PD-MCI, n = 15), and older-aged healthy control (HC, n = 22) individuals. Results indicated that the proportion of time during the rs-fMRI scan that was spent in each of three identified dFNC states (dwell time) differed among these three groups. Individuals in the PD-MCI group spent significantly more time in a dFNC state characterized by low functional network connectivity, relative to participants in both the PD-NC (p = 0.0226) and HC (p = 0.0027) cohorts and tend to spend less time in a state characterized by anti-correlated thalamo-cortical connectivity, relative to both the PD-NC (p = 0.016) and HC (p = 0.0562) groups. A machine-learning method using sequential pattern mining was also found to distinguish among the groups with moderate accuracies ranging from 0.53 to 0.80, revealing distinct sequential patterns in the temporal ordering of dFNC states. These findings underscore the potential of dFNC and sequential pattern mining as relevant methods for further exploration of the pathophysiologic underpinnings of cognitive impairment among people living with PD.

Significance of gender, brain region and EEG band complexity analysis for Parkinson's disease classification using recurrence plots and machine learning algorithms

Parkinson Disease (PD) is a complex neurological disorder attributed by loss of neurons generating dopamine in the SN per compacta. Electroencephalogram (EEG) plays an important role in diagnosing PD as it offers a non-invasive continuous assessment of the disease progression and reflects these complex patterns. This study focuses on the non-linear analysis of resting state EEG signals in PD, with a gender-specific, brain region-specific, and EEG band-specific approach, utilizing recurrence plots (RPs) and machine learning (ML) algorithms for classification. For this an open EEG dataset consisting of 14 PD and 14 healthy (HC) subjects is utilized. Recurrence plots and cross-recurrence plots (CRPs) were constructed for each frequency band and brain region, extracting complexity measures such as determinism (DET) and entropy (ENT). The interpretability of the ML model decisions is investigated using explainability technique. The scattered distribution of points in RPs of male PD individuals reflects the complex and dynamic nature of abnormal brain function. Also, CRPs confirms the enhanced effect of Beta Gamma synchronization during PD in the Parietal region. Low DET and high ENT corresponds to the complex non-linear characteristics of EEG signals and brain neuronal circuits during PD condition in male subjects. The extracted recurrence features served as inputs to the ML models, which achieved high classification performance, across all the scenarios. This study demonstrates the potential of recurrence plot-based complexity analysis combined with machine learning for the gender-specific, region-specific, and band-specific assessment of EEG signals during resting state in Parkinson's disease.

Development of Neurodegenerative Disease Diagnosis and Monitoring from Traditional to Digital Biomarkers

Monitoring and assessing the progression of symptoms in neurodegenerative diseases, including Alzheimer's and Parkinson's disease, are critical for improving patient outcomes. Traditional biomarkers, such as cerebrospinal fluid analysis and brain imaging, are widely used to investigate the underlying mechanisms of disease and enable early diagnosis. In contrast, digital biomarkers derived from phenotypic changes-such as EEG, eye movement, gait, and speech analysis-offer a noninvasive and accessible alternative. Leveraging portable and widely available devices, such as smartphones and wearable sensors, digital biomarkers are emerging as a promising tool for ND diagnosis and monitoring. This review highlights the comprehensive developments in digital biomarkers, emphasizing their unique advantages and integration potential alongside traditional biomarkers.

Multiple time delay induced Hopf bifurcation of a cortex - basal ganglia model for Parkinson's Disease

Exploring the origin of beta - band oscillation in the cortex - basal ganglia model plays an important role in understanding the mechanism of Parkinson's disease. In this paper, we investigate the effect of three synaptic transmission time delays in the subthalamic nucleus(STN) - the globus pallidus external segment(GPe) loop, the excitatory neurons in the cortex(EXN) - the inhibitory neurons in the cortex(INN) loop and EXN - STN loop on critical conditions of occurrence of beta - band oscillation through Hopf bifurcation theory including linear stability analysis, center manifold theorem and normal form analysis. Our results reveal that suitable transmission time delay can induce beta - band oscillation through Hopf bifurcation, and the critical condition under which Hopf bifurcation occurs is more sensitive to the transmission time delay in EXN - STN loop T 3 , where ifT 3 > 0.00185 , beta - band oscillation always occurs for any transmission time delay in STN - GPe, EXN - INN loops. Our theoretical analyses provide some ideas for the future research of Parkinson's disease.

Altered EEG Theta and Alpha Band Functional Connectivity in Mild Cognitive Impairment During Working Memory Coding

Individuals with mild cognitive impairment (MCI), the preclinical stage of Alzheimer disease (AD), suffer decline in their visual working memory (WM) functions. Using large-scale network analysis of electroencephalography (EEG), the current study intended to investigate if there are differences in functional connectivity properties extracted during visual WM coding stages between MCI patients and normal controls (NC). A total of 21 MCI patients and 20 NC performed visual memory tasks of load four, while 32-channel EEG recordings were acquired. The functional connectivity properties were extracted from the acquired EEGs by the directed transform function (DTF) via spectral Granger causal analysis. Brain network analyses revealed distinctive brain network patterns between the two groups during the WM coding stage. Compared with the NC, MCI patients exhibited a reduced visual network connectivity of the frontal-temporal in θ (4-7Hz) band. A likely compensation mechanism was observed in MCI patients, with a strong brain functional connectivity of the frontal-occipital and parietal-occipital in both θ and α (8-13Hz) band. Further analyses of the network core node properties based on the differential brain network showed that, in θ band, there was a significant difference in the out-degree of the frontal lobe and parietal lobe between the two groups, while in α band, such difference was located only in the parietal lobe. The current study found that, in MCI patients, dysconnectivity is found from the prefrontal lobe to bilateral temporal lobes, leading to increased recruitment of functional connectivity in the frontal-occipital and parietal-occipital direction. The dysconnectivity pattern of MCI is more complex and primarily driven by core nodes Pz and Fz. These results significantly expanded previous knowledge of MCI patients' EEG dynamics during WM tasks and provide new insights into the underpinning neural mechanism MCI. It further provided a potential therapeutic target for clinical interventions of the condition.

Neural quantification of emotion influencing learning based on dynamic brain network analyses

Emotions have a significant impact on learning processes. Different aspects of emotion and learning, including attention and motivation, have been widely explored. However, the underlying emotional factors that explicitly affect the learning process remain unclear. To this end, we designed an emotion-learning paradigm that incorporates videos to elicit emotions and reading materials as the learning content, complemented by a series of behavioral tests. Initially, we employed multiscale entropy (MSE) to quantify the brain complexity of EEG signals across various emotional states during the learning process. Subsequently, we extracted specific neural rhythms from EEG signals using masking empirical mode decomposition (MEMD) and constructed the dynamic brain network of the learning process by sliding window and phase locking value (PLV). Brain networks, including the prefrontal-temporal reading and the frontal-parietal cognitive control networks, in the learning stage were examined with their global efficiency, clustering coefficient, and local efficiency calculated. The results indicated that the group experiencing negative emotions exhibited a lower rate of information integration, despite higher brain complexity. This may be attributed to the additional cognitive resource consumption triggered by negative emotions. Conversely, subjects experiencing positive emotions demonstrated not only lower brain complexity during the learning process but also higher rates of information integration and improved learning performance.

A novel nonlinear bispectrum analysis for dynamical complex oscillations

In this study, we proposed a novel set of bispectrum in constructing both frequency power and complexity spectrum. The uniform phase empirical mode decomposition (UPEMD) was implemented to obtain nonlinear extraction while guaranteeing explicit frequencies. Lepel-Ziv complexity (LZC) and frequency power per mode were used for comprehensive frequency spectra. To examine the performances of the proposed method and meanwhile optimize routine methodological parameters, either chaotic logistic maps or a default non-stationary simulation in 40 ~ 60 Hz along with several challenges were designed. The simulation results showed the UPEMD-based LZC spectrum distinguishes the degree of complexity, reflecting the bandwidth and noise level of the inputs. The UPEMD-based power spectrum on the other side presents power distribution of nonlinear and nonstationary oscillation across multiple frequencies. In addition, given gait disturbance is an unsolved symptom in adaptive deep brain stimulation (DBS) for Parkinson's disease (PD), meanwhile considering the representative of deep brain activities to the complex oscillations, such data were analyzed further. Our results showed the high-frequency band (45 ~ 80 Hz) of the UPEMD-based LZC spectrum reflects the impact of auditory cues in modulating the complexity of DBS recording. Such an increase in complexity (45 ~ 60 Hz) reduces shortly after the cue was removed. As for the UPEMD-based power spectrum, decreasing power over the higher frequency region (> 30 Hz) was shown with auditory cues. These results manifest the potential of the proposed methods in reflecting gait improvement for PD. The proposed bispectrum reflected both the nonlinear complexity and power spectrum analyses, enabling examining targeted frequencies with refined resolution.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11571-023-09953-z.

Multi-Threshold Recurrence Rate Plot: A Novel Methodology for EEG Analysis in Alzheimer's Disease and Frontotemporal Dementia

This study introduces Multi-Threshold Recurrence Rate Plots (MTRRP), a novel methodology for analyzing dynamic patterns in complex systems, such as those influenced by neurodegenerative diseases in brain activity. MTRRP characterizes how recurrence rates evolve with increasing recurrence thresholds. A key innovation of our approach, Recurrence Complexity, captures structural complexity by integrating local randomness and global structural features through the product of Recurrence Rate Gradient and Recurrence Hurst, both derived from MTRRP. We applied this technique to resting-state EEG data from patients diagnosed with Alzheimer's Disease (AD), Frontotemporal Dementia (FTD), and age-matched healthy controls. The results revealed significantly higher recurrence complexity in the occipital areas of AD and FTD patients, particularly pronounced in the Alpha and Beta frequency bands. Furthermore, EEG features derived from MTRRP were evaluated using a Support Vector Machine with leave-one-out cross-validation, achieving a classification accuracy of 87.7%. These findings not only underscore the utility of MTRRP in detecting distinct neurophysiological patterns associated with neurodegenerative diseases but also highlight its broader applicability in time series analysis, providing a substantial tool for advancing medical diagnostics and research.

Differences in functional network between focal onset nonconvulsive status epilepticus and toxic metabolic encephalopathy: application to machine learning models for differential diagnosis

We aimed to compare network properties between focal-onset nonconvulsive status epilepticus (NCSE) and toxic/metabolic encephalopathy (TME) during periods of periodic discharge using graph theoretical analysis, and to evaluate the applicability of graph measures as markers for the differential diagnosis between focal-onset NCSE and TME, using machine learning algorithms. Electroencephalography (EEG) data from 50 focal-onset NCSE and 44 TMEs were analyzed. Epochs with nonictal periodic discharges were selected, and the coherence in each frequency band was analyzed. Graph theoretical analysis was performed to compare brain network properties between the groups. Eight different traditional machine learning methods were implemented to evaluate the utility of graph theoretical measures as input features to discriminate between the two conditions. The average degree (in delta, alpha, beta, and gamma bands), strength (in delta band), global efficiency (in delta and alpha bands), local efficiency (in delta band), clustering coefficient (in delta band), and transitivity (in delta band) were higher in TME than in NCSE. TME showed lower modularity (in delta band) and assortativity (in alpha, beta, and gamma bands) than NCSE. Machine learning algorithms based on EEG global graph measures classified NCSE and TME with high accuracy, and gradient boosting was the most accurate classification model with an area under the receiver operating characteristics curve of 0.904. Our findings on differences in network properties may provide novel insights that graph measures reflecting the network properties could be quantitative markers for the differential diagnosis between focal-onset NCSE and TME.

Altered functional connectivity during visual working memory state in patients with mild cognitive impairment

Patients with mild cognitive impairment (MCI) suffer from severe memory function impairment, especially working memory [1]. Based on Electroencephalogram (EEG), this study used power spectral density and large-scale network analysis to reveal the frequency changes of brain areas and the diverse network patterns during the visual WM coding stages between MCI and normal controls (NC). The results showed, compared to NC, the left and right prefrontal lobes and central regions has significant synchronization in the θ frequency band, while the left temporal lobe has significant desynchronization during the working memory coding state among MCI. Brain network analysis revealed the diverse network patterns during the WM coding stage between two group. Compared with the NC, the brain of MCI patients reduced the top-down visual network connection flow derived from frontal lobe to parietal lobe, and recruited a more up-down mechanism with a much stronger information flow from frontal lobe to occipital lobe during the WM coding state. This result provides a new perspective for the neural mechanism of WM deficits in MCI patients.Clinical Relevance-Abnormal EEG rhythm and connectivity of brain regions may be important indicators of WM disorders in patients with MCI.

Evolution of brain network dynamics in early Parkinson's disease with mild cognitive impairment

How mild cognitive impairment (MCI) is instantiated in dynamically interacting and spatially distributed functional brain networks remains an unexplored mystery in early Parkinson's disease (PD). We applied a machine-learning technology based on personalized sliding-window algorithm to track continuously time-varying and overlapping subnetworks under the functional brain networks calculated form resting state electroencephalogram data within a sample of 33 early PD patients (13 early PD patients with MCI and 20 early PD patients without MCI). We decoded a set of subnetworks that captured surprisingly dynamically varying and integrated interactions among certain brain lobes. We observed that the master expressed subnetworks were particularly transient, and flexibly switching between high and low expression during integration into a dynamic brain network. This transience was particularly salient in a subnetwork predominantly linking temporal-parietal-occipital lobes, which decreases in both expression and flexibility in early PD patients with MCI and expresses their degree of cognitive impairment. Moreover, MCI induced a regularly interrupted, slow evolution of subnetworks in functional brain network dynamics in early PD at the individual level, and the dynamic expression characteristics of subnetworks also reflected the degree of cognitive impairment in patients with early PD. Collectively, these results provide novel and deeper insights regarding MCI-induced abnormal dynamical interaction and large-scale changes in functional brain network of early PD.

Temporal and spatial variability of dynamic microstate brain network in early Parkinson's disease

Changes of brain network dynamics reveal variations in macroscopic neural activity patterns in behavioral and cognitive aspects. Quantification and application of changed dynamics in brain functional connectivity networks may contribute to a better understanding of brain diseases, and ultimately provide better prognostic indicators or auxiliary diagnostic tools. At present, most studies are focused on the properties of brain functional connectivity network constructed by sliding window method. However, few studies have explored evidence-based brain network construction algorithms that reflect disease specificity. In this work, we first proposed a novel approach to characterize the spatiotemporal variability of dynamic functional connectivity networks based on electroencephalography (EEG) microstate, and then developed a classification framework for integrating spatiotemporal variability of brain networks to improve early Parkinson's disease (PD) diagnostic performance. The experimental results indicated that compared with the brain network construction method based on conventional sliding window, the proposed method significantly improved the performance of early PD recognition, demonstrating that the dynamic spatiotemporal variability of microstate-based brain networks can reflect the pathological changes in the early PD brain. Furthermore, we observed that the spatiotemporal variability of early PD brain network has a specific distribution pattern in brain regions, which can be quantified as the degree of motor and cognitive impairment, respectively. Our work offers innovative methodological support for future research on brain network, and provides deeper insights into the spatiotemporal interaction patterns of brain activity and their variabilities in early PD.

Temporal Stability of Dynamic Default Mode Network Connectivity Negatively Correlates with Suicidality in Major Depressive Disorder

Previous studies have demonstrated that the suicidality in patients with major depressive disorder (MDD) is related to abnormal brain functional connectivity (FC) patterns. However, little is known about its relationship with dynamic functional connectivity (dFC) based on the assumption that brain FCs fluctuate over time. Temporal stabilities of dFCs within the whole brain and nine key networks were compared between 52 MDD patients and 21 age, sex-matched healthy controls (HCs) using resting-state functional magnetic resonance imaging and temporal correlation coefficients. The alterations in MDD were further correlated with the scores of suicidality item in the Hamilton Rating Scale for Depression (HAMD). Compared with HCs, the MDD patients showed a decreased temporal stability of dFC as indicated by a significantly decreased temporal correlation coefficient at the global level, as well as within the default mode network (DMN) and subcortical network. In addition, temporal correlation coefficients of the DMN were found to be significantly negatively correlated with the HAMD suicidality item scores in MDD patients. These results suggest that MDD may be characterized by excessive temporal fluctuations of dFCs within the DMN and subcortical network, and that decreased stability of DMN connectivity may be particularly associated with the suicidality in MDD.

Association of Plasma and Electroencephalography Markers With Motor Subtypes of Parkinson’s Disease

Objective

The aim of this study was to investigate the correlations of plasma neurodegenerative proteins and electroencephalography (EEG) dynamic functional network (DFN) parameters with disease progression in early Parkinson’s disease (PD) with different motor subtypes, including tremor-dominant (TD) and postural instability and gait disorder (PIGD).

Methods

In our study, 33 patients with PD (21 TD and 12 PIGD) and 33 healthy controls (HCs) were enrolled. Plasma neurofilament light chain (NfL), α-synuclein (α-syn), total-tau (t-tau), β-amyloid 42 (Aβ42), and β-amyloid 40 (Aβ40) levels were measured using an ultrasensitive single-molecule array (Simoa) immunoassay. All the patients with PD underwent EEG quantified by DFN analysis. The motor and non-motor performances were evaluated by a series of clinical assessments. Subsequently, a correlation analysis of plasma biomarkers and EEG measures with clinical scales was conducted.

Results

In the TD group, plasma NfL exhibited a significant association with MDS-UPDRS III and Montreal Cognitive Assessment (MoCA). A higher Aβ42/40 level was significantly related to a decrease in Hamilton Depression Rating Scale (HAMD) and Hamilton Anxiety Rating Scale (HAMA) in the PIGD group. In terms of the correlation between EEG characteristic parameters and clinical outcomes, trapping time (TT) delta was positively correlated with MDS-UPDRS III and MoCA scores in the TD group, especially in the prefrontal and frontal regions. For other non-motor symptoms, there were significant direct associations of k_PLI theta with HAMD and HAMA, especially in the prefrontal region, and k_PLI gamma was particularly correlated with Rapid Eye Movement Sleep Behavior Disorder Screening Questionnaire (RBDSQ) scores in the prefrontal, frontal, and parietal regions in the TD group. Furthermore, there was a significant positive correlation between plasma t-tau and k_PLI, and pairwise correlations were found among plasma NfL, theta TT, and MoCA scores in the TD group.

Conclusion

These results provide evidence that plasma neurodegenerative proteins and EEG measures have great potential in predicting the disease progression of PD subtypes, especially for the TD subtype. A combination of these two kinds of markers may have a superposition effect on monitoring and estimating the prognosis of PD subtypes and deserves further research in larger, follow-up PD cohorts.

Dynamic Connectivity Analysis Using Adaptive Window Size

In this paper, we propose a new method to study and evaluate the time-varying brain network dynamics. The proposed RICI-imCPCC method (relative intersection of confidence intervals for the imaginary component of the complex Pearson correlation coefficient) is based on an adaptive window size and the imaginary part of the complex Pearson correlation coefficient. It reduces the weaknesses of the existing method of constant sliding window analysis with narrow and wide windows. These are the low temporal precision and low reliability for short connectivity periods for wide windows, and high susceptibility to noise for narrow windows, all resulting in low estimation accuracy. The proposed method overcomes these shortcomings by dynamically adjusting the window width using the RICI rule, which is based on the statistical properties of the area around the observed sample. In this paper, we compare the RICI-imCPCC with the existing constant sliding window analysis method and describe its advantages. First, the mathematical principles are established. Then, the comparison between the existing and the proposed method using synthetic and real electroencephalography (EEG) data is presented. The results show that the proposed RICI-imCPCC method has improved temporal resolution and estimation accuracy compared to the existing method and is less affected by the noise. The estimation error energy calculated for the RICI-imCPCC method on synthetic signals was lower by a factor of 1.22 compared to the error of the constant sliding window analysis using narrow window size imCPCC, by a factor of 2.87 compared to using wide window size imCPCC, by a factor of 6.69 compared to using narrow window size wPLI, and by a factor of 4.72 compared to using wide window size wPLI. Analysis of the real signals shows the ability of the proposed method to detect a P300 response and to detect a decrease in dynamic connectivity due to desynchronization and blockage of mu-rhythms.