Skip to main content
NCBI home page
Cognitive Neurodynamics logoLink to Cognitive Neurodynamics
. 2023 Jul 31;18(4):1549–1561. doi: 10.1007/s11571-023-09994-4

Altered effective connectivity of the default mode network in juvenile myoclonic epilepsy

Ming Ke 1,✉, Feng Wang 1, Guangyao Liu 2,✉
PMCID: PMC11297871  PMID: 39104702

Abstract

Juvenile myoclonic epilepsy (JME) is associated with brain dysconnectivity in the default mode network (DMN). Most previous studies of patients with JME have assessed static functional connectivity in terms of the temporal correlation of signal intensity among different brain regions. However, more recent studies have shown that the directionality of brain information flow has a more significant regional impact on patients’ brains than previously assumed in the present study. Here, we introduced an empirical approach incorporating independent component analysis (ICA) and spectral dynamic causal modeling (spDCM) analysis to study the variation in effective connectivity in DMN in JME patients. We began by collecting resting-state functional magnetic resonance imaging (rs-fMRI) data from 37 patients and 37 matched controls. Then, we selected 8 key nodes within the DMN using ICA; finally, the key nodes were analyzed for effective connectivity using spDCM to explore the information flow and detect patient abnormalities. This study found that compared with normal subjects, patients with JME showed significant changes in the effective connectivity among the precuneus, hippocampus, and lingual gyrus (p < 0.05 with false discovery rate (FDR) correction) with most of the effective connections being strengthened. In addition, previous studies have found that the self-connection of normal subjects’ nodes showed strong inhibition, but the self-connection inhibition of the anterior cingulate cortex and lingual gyrus of the patient was decreased in this experiment (p < 0.05 with FDR correction); as the activity in these areas decreased, the nodes connected to them all appeared abnormal. We believe that the changes in the effective connectivity of nodes within the DMN are accompanied by changes in information transmission that lead to changes in brain function and impaired cognitive and executive function in patients with JME. Overall, our findings extended the dysconnectivity hypothesis in JME from static to dynamic causal and demonstrated that aberrant effective connectivity may underlie abnormal brain function in JME patients at early phase of illness, contributing to the understanding of the pathogenesis of JME.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11571-023-09994-4.

Keywords: Effective connectivity, Spectral dynamic causal modeling, fMRI, Default mode network, Independent component analysis, Juvenile myoclonic epilepsy

Introduction

Juvenile myoclonic epilepsy (JME) is the most common generalized genetic epilepsy syndrome, accounting for up to 10% of all epilepsy patients (Lin et al. 2022; Ma et al. 2022). The clinical characteristics of this syndrome include myoclonic jerks, tonic–clonic seizures, and occasional absence seizures; patients with JME also typically show various cognitive impairments, such as deficits in working memory, attention, and executive function (Carvalho et al. 2016; Pascalicchio et al. 2007; Wandschneider et al. 2012) cognitive impairments (Caciagli et al. 2020; Cendes and Cascino 2019) as well as frontal lobe dysfunction (Giorgi et al. 2016; Hattingen et al. 2014) severely affect patients’ quality of life and, to some extent, increase the social and economic burden of the disease. To solve this serious problem, people have been trying to understand the fundamental mechanisms underlying the neuropsychology of JME. In recent years, with advances in functional magnetic resonance imaging (fMRI) techniques and X-ray image acquisition (Altan and Karasu 2020), researchers have gained a better understanding of the pathogenesis of JME (Baykan and Wolf 2017; Caciagli et al. 2019). Resting-state functional connectivity has been proven to be an effective means to study the characteristics of brain activity and has been widely used to study epilepsy. An increasing number of researchers are focusing on the study of brain networks in patients with JME, and it has been reported in the literature that an impaired default mode network (DMN) in patients with JME is associated with the pathology of JME (Caciagli et al. 2019; Liu et al. 2022; Zhang et al. 2020).

The DMN has strong spontaneous activities in the resting state, but when performing cognitive tasks with a certain degree of difficulty, the activities of the DMN are inhibited to a certain extent. Inhibition increases as the difficulty of cognitive tasks increases (Mohan et al. 2016; Smallwood et al. 2021) The DMN plays a pivotal role in many cognitive processes, such as remembering, looking forward to the future, and social reasoning. It is not only engaged in the integration of cognition, but also in memory recollection and reconstruction of long-term memories during retrieval(Bacon et al. 2023; Jiang et al. 2023).The main cortical areas in the DMN include the medial prefrontal cortex (MPFC), posterior cingulate (PCC), precuneus (PCU), parietal lobule (IPL), and hippocampus (HIP) (Mason et al. 2007). Recently, DMN dysfunction was found in patients with JME. In previous studies, the functional interaction between the PCC and MPFC in patients with JME was reduced (McGill et al. 2012); the spontaneous fluctuations in the posterior region of the DMN and the coupling of functional connections changed in these patients and were related to clinical symptoms during epileptic seizures (Jia et al. 2018). In addition, cognitive dysregulation of JME is associated with hippocampal dysfunction (Zhang et al. 2020). As a mathematical extension of the traditional network model, the multilayer network can fully capture the fluctuations of brain imaging data with time, and capture subtle abnormal dynamic changes. A multilayer network study revealed that the flexibility of the opercular part of the inferior frontal gyrus was significantly correlated with the severity of JME symptoms (Ke et al. 2023). Taken together, these findings suggested that the DMN is prominently affected in patients with JME (Parsons et al., 2020) and specific studies on the DMN may lead to new insights that reveal the pathogenesis of JME in the early stages. However, most studies have focused on the changes in DMN functional connectivity strength, often ignoring the directionality of connections, in patients with JME.

Functional connectivity data have not provided an answer to the question of whether one neural area has a directional effect on another neural area. Data of this type have merely revealed the nondirectional statistical correlation among brain areas (Agosta et al. 2023; Cunningham et al. 2017). Effective connectivity (EC) refers to the pattern of directed causal influence and information flow (Friston et al. 2014a); these patterns are expressed as consistent, synchronized neural activity between brain regions. EC can establish a causal model of the interaction between neurons and analyze the direction and nature of interaction between the activity of one neural unit and others. EC analysis may be able to take on a more important role in exploring changes between patients and healthy controls, as it can provide a better understanding in terms of interpretability (Geng et al. 2018). Dynamic causal modeling (DCM) is the most common method used to study EC. DCM has been used mainly to study the activation of brain regions. Owing to the connectivity of brain regions with themselves and other regions, changes in the activation of one region affect both the subsequent behavior of that region and the activation of other regions (Geng et al. 2018), resulting in changes in the EC of brain networks. As an extension of DCM in the resting state, spectral dynamic causal modeling (spDCM) (Friston et al. 2014a, b) can be used to examine the model-based effective connectivity structure between regions, rather than their statistical dependencies, according to generative (state-space) models with a biologically realistic hemodynamic function (Zeng et al. 2022). Therefore, it has high computational efficiency. Furthermore, spDCM is more efficient than stochastic DCM (sDCM) (Li et al. 2011), is more sensitive to intergroup differences, and can be used to compare directionality and couplings within an endogenous network between different groups (Amiri et al. 2021), which is also an important reason for the use of spDCM in this study (Razi et al. 2015). In recent studies, spDCM was used in patients with internet gaming disorder has to identify dysregulation in the IPL-MPFC-PCC circuits (Wang et al. 2019); Wei et al. used spDCM to reveal that cocaine-dependent subjects had altered resting-state EC between parts of the memory and reward systems (Wei et al. 2021). The attenuated effective connectivity of default networks may be a clinical neurobiological feature of autism spectrum disorder (Wei et al. 2022). spDCM is also widely used in schizophrenia as well as depression, and the connection between the DMN and the central executive network (CEN) was found to be a possible clinically relevant neurobiological signature of schizophrenia symptoms (Gattuso et al. 2023; Xi et al. 2021) and altered sensorimotor function in depressed patients (Ray et al. 2021). However, spDCM has been used less frequently in patients with epilepsy, despite holding great promise for these patients.

In the current study, our main goal was to explore the effective connectivity within the DMN network, which is the most active network in the resting state (Wang et al. 2019). To achieve this goal, we used frequency domain-based spDCM to study the changes in the EC between key nodes in the DMN and to analyze the abnormal information flow and causality of patients with JME. These aspects are important for understanding the pathophysiological mechanism of JME. Our approach allows the identification of multiple important connections in JME patients that are closely related to their physical symptoms. We hypothesize that abnormal circuit connections may be more important than a single connection.

Materials and methods

Participants

Thirty-seven treatment-naïve newly diagnosed patients with JME were recruited from outpatients at the Epilepsy Center of Lanzhou University Second Hospital. All of them fulfilled the epilepsy classification criteria of the International League Against Epilepsy (ILAE) guidelines for JME (Engel 2001). Routine MRI scans were normal, and routine scalp epilepsy electroencephalogram (EEG) showed 4–6 Hz generalized spike-wave discharges (GSWDs) or polyspike-wave discharges. The exclusion criteria for the patient group were as follows: (a) a history of using any form of antiseizure medication; (b) other neurological or major psychiatric illness; (c) other developmental disabilities, such as autism or intellectual impairment; and (d) acute physical illness that would affect scanning. All resting-state data were acquired in the interictal phase, and there were no seizure manifestations at the time of magnetic resonance. The seizure severity of each patient was assessed using the National Hospital Seizure Severity Scale (NHS3), which is a valid, easily applicable measure of seizure severity (O'Donoghue et al. 1996). This scale covers six seizure-related factors (Chinese version), namely, generalized convulsions, falls, incontinence, loss of consciousness, duration of recovery time, and automatisms and generates a total score from 1 to 23. Thirty-seven healthy controls were recruited from the local community via advertisement and those with a history of febrile convulsions, seizures, or family history of epilepsy were excluded. This study was approved by the Ethics Committee of Lanzhou University Second Hospital. Written informed consent was obtained from all participants. After head motion exclusion, the remaining 35 patients and 37 healthy controls were included in the subsequent analyses. Detailed demographic and clinical characteristics of all participants included in this study are shown in Table 1.

Table 1.

Demographic and characteristics of all participants

Variables JME patients (n = 37) HCs (n = 37) p value
Sex (males/females) 20/17 13/24 0.16a
Age (years) 17.65 ± 5.37 20.08 ± 4.11 0.12b
Handedness (right/left) 37/0 37/0 –
Duration of epilepsy (months) 23.84 ± 18.68 – –
NHS3 total score 8.89 ± 4.07 – –
Age at seizure onset (year) 15.42 ± 3.15 – –
Open in a new tab

aChi-square t test, btwo sample t test, values are the mean ± SD

NHS3 National Hospital Seizure Severity Scale, JME juvenile myoclonic epilepsy, HCs healthy controls

Data acquisition

fMRI data were acquired at the Department of Magnetic Resonance of Lanzhou University Second Hospital on a Siemens Verio 3.0 T scanner (Siemens, Erlangen, Germany) with 16 head coils. Participants were instructed to remain still and as motionless as possible before the scan. Participants were additionally required to stay awake with their eyes closed and not to think systematically during the scan. To further minimize head motion, form pads provided by the scanner manufacturer were used to fix each participant’s head in a stable position. Resting-state functional images were acquired for each participant using an echo-planar imaging sequence with the following settings: repetition time [RT] = 2000 ms; echo time [TE] = 30 ms; flip angle = 90°; slice thickness = 4 mm; in-plane matrix resolution = 64 × 64; field of view [FOV] = 240 × 240 mm2; slices = 33; 200 volumes; and a total of 400 s. For anatomical localization and normalization, high-resolution structural 3D T1-weighted images were obtained using a magnetization-prepared rapid gradient-echo sequence (TR = 1900 ms; TE = 2.99 ms; flip angle = 90°; slice thickness = 0.9 mm; acquisition matrix = 256 × 256; FOV = 230 × 230 mm2; in-plane resolution = 0.9 × 0.9 mm2; slices = 192).

Methods

Overview of the EC analysis

A schematic diagram of the analysis framework to investigate the atypical EC of the DMN in JME patients is presented in Fig. 1. Specifically, there were four major analysis steps in this framework. First, the data were preprocessed to facilitate one-step analysis. Second, group independent component analysis (ICA) was performed to decompose the preprocessed functional data into multiple independent components (ICs), and the DM components were identified according to their spatial activation maps. Third, spDCM was used to model the selected critical nodes and determine the strength of the connection between each node. Finally, the network discovery procedure of Bayesian model reduction (BMR) was used to find the best model, and the model with the highest posterior probability was selected as the best model for the subject. The connection strength of the two groups of subjects was analyzed using statistics to identify differential connections. The detailed methods used for EC analysis are provided in the following subsections.

Fig. 1.

Fig. 1

Open in a new tab

Analysis flowchart to study the EC of the DMN in patients with JME

Data preprocessing

Before analysis, we prepared the data using MATLAB version 2017a running on the Windows 10 operating system DPARSF software (http://www.restfmri.net) and SPM12 (https://www.fil.ion.ucl.ac.uk/spm/software/spm12/), which were run on the MATLAB platform, were also used to preprocess the data. The main steps of preprocessing included (1) Conversion of DICOM files to NIFTI files and removal of the first 10 time points to eliminate data instability caused by various factors, such as the initial startup process of the machine; (2) Realignment of the functional volumes, slice-timing correction, and head-motion correction (in which translation greater than 3 mm and rotation greater than 3 degrees were corrected); (3) Co-registration of the functional data to the structural T1 volume; DARTEL registration to register the structure image to the template image of standard anatomical space without anatomical information; (4) Nonlinear warping into MNI space, whereby scan data were stretched, compressed, and wound to match the scanned brain to a standard brain template; (5) Linear detrending and regression of covariates of no interest from the blood oxygen level-dependent (BOLD) voxel wise time courses (removal of a linear tendency to heat up due to the operation of the machine or because the subjects had not adapted); (6) Filtering of the BOLD time courses using a low-pass filter with a window of 0.01 ~ 0.08 Hz; and (7) Spatial smoothing with a Gaussian kernel of 6 mm FWHM.

ICA

Upon completion of the DPARSF preprocessing, we performed spatial ICA using GIFT software (http://icatb.sourceforge.net/) on resting-state data of all participants to collect ICA components (consisting of spatial maps and time courses in pairs). ICA is a multivariate statistical method that splits a collection of different data to form distinct spatiotemporal networks. Since McKeown and others first applied this method to fMRI data analysis in 1998 (O'Donoghue et al. 1996), it has become one of the most popular methods in this field and has been widely used in fMRI brain functional connectivity detection (Liu et al. 2017; Moher Alsady et al. 2016) to determine the intrinsic neural networks and their changes during the course of diseases that affect the brain.

The use of ICA allows effective identification and characterization of functional networks in data collected by resting-state functional magnetic resonance imaging (rs-fMRI).Principal component analysis was first applied to reduce subject-specific data into 120 principal components, and then all participant-reduced data across time were concatenated and decomposed into 100 ICs using the infomax algorithm (Bell and Sejnowski 1995). The infomax ICA algorithm was repeated 20 times in ICASSO (Allen et al. 2014) to ensure reliability and stability. After estimating the group spatial maps, a back reconstruction approach was used to obtain subject-specific spatial maps and corresponding time courses. Then, we used the RSN resting-state network templates within GIFT software to calculate the spatial correlation with these 100 ICs (Allen et al. 2014; Liu et al. 2017) and to obtain the spatial calculation value of each IC corresponding to the template. After obtaining the spatial calculation value of each IC, we sorted the calculated correlation values according to the resting-state network template to select the best match with the IC of the network. The components were further evaluated according to the following criteria:

  • A)

    Peak activation coordinates were located primarily in gray matter;

  • B)

    The resting network template regression coefficient was greater than 0.2;

  • C)

    The value of PowerLF/PowerHF was greater than 1.5;

  • D)

    The time course is dominated by low-frequency fluctuations (Allen et al. 2011);

  • E)

    The most suitable images were selected visually.

Moreover, to remove remaining noise sources, postprocessing was performed on the time courses of selected default mode components, including detrended linear, quadratic, and cubic trends; six realignment parameters and their temporal derivatives were regressed out; detected outliers were distinguished using the 3DDESPIKE algorithm; and low-pass filtering was performed with a cut off frequency of 0.15 Hz. Finally, the selected nodes were analyzed by establishing a full connection model through spDCM.

SpDCM

spDCM analysis was performed with DCM 12 implemented in SPM 12 (https://www.fil.ion.ucl.ac.uk/spm/software/spm12/). spDCM is distinct from sDCM because it eschews the estimation of random fluctuations in (hidden) neural states, rendering spDCM essentially deterministic. spDCM simply estimates the time-invariant parameters of their cross spectra (Friston et al. 2014a). In other words, while sDCM models the observed BOLD time series of each node, spDCM models the observed functional connectivity between pairs of nodes. This model is the same as the traditional dynamic causal model. The only difference is that it performs model parameter estimation and model inversion in the frequency domain. Compared to the previous DCM approaches, spDCM has a faster calculation speed, smaller memory footprint, and more stable results. Therefore, in this study, we used spDCM to analyze the difference in EC in the DMN of patients with JME.

We assumed that all participants used the same model and specified a “full” model for each subject. Here, “full” means that each node was assumed to be connected to all other nodes (28 = 256 connectivity parameters, including 8 intrinsic self-connections) (Ma et al. 2015). Unlike DCM under tasks, there were no experimental conditions in the model, so we specified a fully connected 8-node model for each subject in the absence of exogenous inputs. Thus, spDCM contains only endogenous connectivity and is quantified by the A-matrix parameters (Friston et al. 2014a; Ray et al. 2016). The next step was model estimation based on the standard variational Bayesian procedures under the frequency domain. The convolution kernel of the model was converted into a spectrum and expressed in the frequency domain. After parameter estimation was completed, BMR was used to select the best model for the JME patient group and the HC group by the posterior probability of the model (Di and Biswal 2014; Ray et al. 2016).

BMR

In the absence of a particular hypothesis or model space, we used the fully connected model for exploratory analysis of all possible reduced models. After the complete DCM established by 8 nodes related to each participant was reversed, we employed a network discovery procedure using BMR to find the best model. This BMR program is an effective method that can be used to score a large model space without reversing each simplified model and reduce the robustness of violating the (commonly used) Laplace hypothesis in DCM (Friston et al. 2016). In the process of searching, the BMR method can be used to test all possible models nested in the fully connected model and select the model with the highest a posteriori probability as the best model for the subject.

Statistical analysis

Group differences in age and sex were explored using a two-sample t test and a chi-square test, respectively. Differences in effective connectivity between patients and controls were examined using a Wilcoxon rank test. All statistical analyses were performed in SPSS 25.0 (IBM Corporation, Armonk, NY). Statistical significance was established at p < 0.05 with false discovery rate (FDR) correction conducted for each brain measure separately.

Results

Default mode components

We used the ICA method to select the same nodes in the patient group and the HC group. To extract the BOLD fMRI time series of the nodes, the preprocessed data were used to build a general linear model (GLM). Six head motion parameters and WM/CSF signals were added to the model as interference regression variables. Therefore, we identified eight key nodes in the DMN area, including the anterior cingulate cortex (ACC), posterior cingulate cortex (PCC), precuneus (PCU), medial prefrontal cortex (MPFC), hippocampus (HIP), middle occipital gyrus (MOG), lingual gyrus (LG) and calcarine cortex (CL), for EC analysis. Supplementary Table S1 shows the names and cluster sizes of the eight nodes. Supplementary Fig. S1 shows the positions of the eight nodes in the brain and the corresponding time series from the axial and sagittal positions.

DCM network discovery (DND)

The DND approach compared 256 simplified model spaces and identified the best model for the two groups (Wang et al. 2019). Figure 2 shows the results of patients with JME and of participants in the HC group. The left column A represents the JME patient group, and the right column B represents the HC group. The upper left column of A (corresponding to B) shows the distribution of evidence from the log Bayesian model of the simplified model. The posterior probability distribution of 256 simplified models is shown in the upper right column of A (corresponding to B) (Tang et al. 2016). In both groups, comparing the log Bayesian model evidence of 256 simplified models, the “full” model worked best in the final comparison and was considered to be the best model with a posterior probability of almost 1. The lower graph of column (A) and column (B) represents the Bayesian parameter average (BPA) of the patients with JME and HC participants.

Fig. 2.

Fig. 2

Open in a new tab

Results of DND for the two groups. Here (A) represents patients with JME while (B) represents HC participants. The upper two panels of (A) (resp. B) show the log-posterior and posterior probability for every model nested within the full model for patients with JME (resp. HC). The lower panel of (A) (resp. B) shows the BPA for the JME (resp. HC). The horizontal axis represents the source region for the 64 connectivity parameters, and the colors refer to the target regions. JME juvenile myoclonic epilepsy, HC healthy control, BPA Bayesian parameter average, ACC anterior cingulate cortex, PCC posterior cingulate cortex, PCU precuneus, MPFC medial prefrontal cortex, HIP hippocampus, MOG middle occipital gyrus, LG lingual gyrus, CL calcarine cortex

Effective connectivity

After obtaining the optimal models of the two groups, we used BPA to summarize the group differences between patients with JME and HCs to quantify the group differences in EC. The two-sample t test was performed on the connectivity strength values of the JME patient group and the HC group under the same connectivity; Fig. 3a shows the EC enhancement from the PCU to the ACC (p < 0.05, t = − 3.539 with FDR correction) and Fig. 3b shows the EC enhancement from the MPFC to the HIP (p < 0.05, t = − 2.838 with FDR correction). Figure 3c shows the EC enhancement from the LG to the HIP (p < 0.05, t = 2.639 with FDR correction). Figure 3d shows the EC enhancement from the HIP to the PCU (p < 0.05, t = − 2.618 with FDR correction). Figure 3e shows the EC enhancement from the HIP to the LG (p < 0.05, t = − 3.517 with FDR correction). Figure 3f shows the EC enhancement from the CL to the PCU (p < 0.05, t = − 2.272 with FDR correction). As the connection strength increased, the information flow was transmitted more frequently, often leading to anomalies in the nodes receiving the information.

Fig. 3.

Fig. 3

Open in a new tab

EC enhancement demonstrated by the two-sample t test. Blue indicates patients with JME, red indicates HCs, *p < 0.05, **p < 0.01. JME juvenile myoclonic epilepsy, HC healthy control, ACC anterior cingulate cortex, PCU precuneus, MPFC medial prefrontal cortex, HIP hippocampus, LG lingual gyrus, CL calcarine cortex. (Color figure online)

We observed that the intrinsic self-connections of the LG (p < 0.05, t = 2.127 with FDR correction) in Fig. 4a and the ACC (p < 0.05, t = 2.449 with FDR corrected) in Fig. 4c were inhibitory and reached significance after the t test. Since the self-connections of nodes always showed inhibitory effects, the responses of the PCC and the ACC in JME were disinhibited compared to the HC group. As shown in Fig. 5, we performed a two-sample t-test (subtracting the BPA of the HC group from that of the JME) to identify which averaged connectivity parameters in JME patients were significantly different from those of HC subjects. The red line indicates an increase in effective connection strength in JME (larger values indicate a more pronounced increase in strength), and the black line indicates a weakening of effective connection strength in JME (larger values indicate a more pronounced weakening).

Fig. 4.

Fig. 4

Open in a new tab

EC decreased as demonstrated by the two-sample t test. Blue indicates patients with JME, red indicates HCs, *p < 0.05. JME juvenile myoclonic epilepsy, HC healthy control, ACC anterior cingulate cortex, MOG middle occipital gyrus, LG lingual gyrus. (Color figure online)

Fig. 5.

Fig. 5

Open in a new tab

EC difference between the JME subjects and HCs. Red lines indicate an increase in EC in patients with JME and black lines indicate a decrease in EC in patients with JME. a Mainly indicates the connection change in other nodes with the PCU, b mainly indicates the connection change in other nodes with the LG. ACC anterior cingulate cortex, PCC posterior cingulate cortex, PCU precuneus, MPFC medial prefrontal cortex, HIP hippocampus, MOG middle occipital gyrus, LG lingual gyrus, CL calcarine cortex. (Color figure online)

Discussion

In this study, we used two sets of samples (37 patients with JME and 37 healthy subjects) to study the changes in the interactions in the DMN of patients with JME and compared the resting state EC among the ACC, PCC, PCU, MPFC, HIP, MOG, LG, and CL in the DMN between patients with JME and healthy subjects. Of note, spDCM is based on frequency domain analysis used to detect EC. This analysis method revealed the flow and change in information flow in 8 key nodes in the DMN of patients with JME. We found that EC was abnormal among the DMN regions of patients with JME.

The DMN has strong spontaneous activity in the resting state, so ICA was used in this study to obtain our region of interest, which ensured that the region of interest we chose in the DMN was closely related to JME pathology. ICA does not rely on any prior knowledge in the selection of the region of interest. It completely depends on the data and thus prevented errors caused by insufficient knowledge of anatomical structures and inappropriate structural templates (Wang and Guo 2019) Previous studies have shown that the flow of causal influence among DMN regions is related to their neuronal activity levels (Jiao et al. 2011), and abnormal activity of the MPFC and HIP has been widely found in patients with JME (Bartolini et al. 2014; Ke et al. 2023; Zhang et al. 2020). The majority of JME studies that combine rs-fMRI with ROI analysis report the results of functional connectivity within the DMN (Chen and Calhoun 2018), which reflects the statistical properties (i.e., temporal correlation) among functional brain regions. However, correlation results cannot answer the question of whether a neural region directionally influences another. Therefore, we used spDCM analysis to investigate the directional relationships among different brain regions.

The results of this experiment using spDCM showed that the EC of the six connections in patients with JME from the PCU to the ACC, the MPFC to the HIP, the LG to the HIP, the HIP to the PCU, the HIP to the LG, and the CL to the PCU are enhanced, indicating that these regions, as central receptors and generators, are destroyed in the interaction, thus increasing the effect of the action (Zhang et al. 2020). The increase in the flow of causal influences in these regions reflect obstacles in receiving influences from other DMN regions and exerting influences on these regions (Ke et al. 2022). In contrast, the EC of the self-connection of the LG and the ACC is decreased, indicating that the flow of causal influence in these areas is inhibited, and the information transmission between various brain regions is blocked. The transmission of information in the two brain regions is inhibited, causing the transmitted information to be shared abnormally with other brain regions, leading to abnormal behaviors in the patients (Gamberini et al. 2020).

The patients in this study showed changes in EC among multiple nodes, especially among the PCU, ACC, MPFC, HIP, and LG. The PCU is a core area of the DMN related to emotion regulation (Fransson and Marrelec 2008). Studies have reported that high-frequency oscillatory waves during seizures are specific markers of epileptic foci and play an important role in the occurrence and propagation of epilepsy (Dixsaut and Gräff 2021; Jiruska et al. 2017). Long-term and chronic high-frequency oscillatory waves destroy normal neural tissue and slowly trigger cortical-subcortical pathways, which can cause the onset of epilepsy (Motoi et al. 2018; Zhang et al. 2020). Previous studies have suggested that the PCU plays an important role in the onset of spike-and-wave discharge in patients with JME. The EEG characteristic of JME is spike-and-wave discharge (SWD), which is dominant in the frontal region. Pertinently, Lee's findings suggest that the precuneus is likely a key region for SWD despite the small amount of neural activity observed (Lee et al. 2014). In addition, when using the FreeSurfer automatic cortical surface reconstruction method, changes in cortical morphology, such as the cingulate gyrus, precuneus, occipital lobe, and fusiform gyrus, were detected (Ma et al. 2022). In this study, we found an increase in the EC of the HIP to the PCU, the CL to the PCU, and the PCU to the ACC in patients with JME. We think that abnormalities in the PCU would lead to abnormalities in the connections of the PCU to other nodes, and as the strength of the connections increases, it would cause damage to the brain regions that receive information, resulting in dysfunction. Our altered ECs between the PCU and other regions are in line with previous studies, which have revealed PCU-related dysconnectivity in patients with JME (Jiang et al. 2018; Ke et al. 2023; Routley et al. 2020). Notably, the MPFC is related to internal mental states and self-referential processing. Our altered ECs of MPFC suggested that abnormal internal cognitive process may contribute to myoclonic jerks in JME. The ACC is usually related to emotional processing (Bravo et al. 2020), and alterations in its structure and function have been revealed in patients with JME (Faulkner et al. 2021; Liu et al. 2021). Additionally, the MPFC is involved in the thalamocortical circuitry, and a reduction in the connectivity between this region and the basal ganglia corresponds to a disruption of external interaction in JME, in which the execution of voluntary movements is impaired (Yang et al. 2022). Our results are consistent with previous studies, which have revealed PCU-, ACC-, and MPFC-related connection disorders in patients with JME, which indicates that mood disorders play a key role in the development of JME. The impact of various brain regions on JME is not isolated, and some studies have suggested that JME subjects showed abnormal functional connections among the HIP, ACC, and MPFC (Zhang et al. 2020). Overall, the altered ECs among these DMN components may implicate neuropsychological impairments in patients with JME. Taken together, the established associations in this study indicate that the early identified abnormalities in brain networks may be implicated in the progression of the disease and can be an indication of disrupted developmental processes in JME.

We also found that the neural pathways composed of the HIP-LG-HIP and the CL-PCU-ACC in the DMN were abnormally connected in patients with JME, showing negative values (usually interpreted as inhibitory) (Friston et al. 2014a). Most previous studies investigated the abnormal connectivity between two nodes but rarely paid attention to the abnormal connectivity between multiple nodes. Abnormalities in this kind of local pathway may cause abnormal activities of nodes near the pathway, which may further expand to a larger area, causing greater impact, and may even affect the flow of information throughout the brain. There are many local pathways composed of abnormal connection patterns in brain diseases. Leveque and his fellow researchers found that congenital amusia is associated with abnormal anatomical and functional connectivity in a right frontotemporal pathway (Leveque et al. 2016). Abnormal frontotemporal pathways in schizophrenia can cause positive symptoms: auditory hallucinations, delusions, and confusion (Eack et al. 2017; McCutcheon et al. 2020). The experimental results in this article provide a basis for the possible existence of similar pathways in patients with JME and provide a new direction for the study of brain networks in these patients.

There were various considerations and limitations in this study. First, epileptic transients have been suggested to influence resting-state effective connectivity (Zhang et al. 2020). Because it is difficult for patients to continually not move their heads, we did not record EEG data during MRI scanning in this study. It would be necessary to examine the effects of interictal epileptic discharges on effective connectivity in future simultaneous EEG-fMRI studies. Second, our sample size was insufficient and some meaningful connections may not have been found, which should be considered more in future neuroimaging studies with larger sample sizes. In addition, we should remain cautious as other networks could also be involved in the connections. Finally, further investigation regarding antipsychotic treatment should be considered to provide a more comprehensive view of the mechanism of JME.

Conclusion

In this study, ICA and spDCM analyses were combined to study the changes in causal influence within the DMN in patients. Our results showed that causal influences within the DMN are altered in patients and were associated with clinical symptoms in patients. The EC among the precuneus, hippocampus, and lingual gyrus all showed inhibition, indicating that the inhibition of the flow of information among brain nodes is the cause of the disease. In addition, we found that the abnormal flow of information was not only the abnormal connection between two nodes but also showed that multiple nodes were involved. We speculate that abnormalities in circuits or pathways composed of multiple nodes may have a greater impact on patients than in those composed of pairs of nodes.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary file 1 (DOCX 3697 kb) (3.6MB, docx)

Acknowledgements

Not applicable.

Author contributions

MK, FW, and GY designed the experiment and revised the manuscript. MK and FW wrote the manuscript. GY recorded and collected the data. FW performed the data analysis. All authors contributed to this article and approved the version submitted.

Funding

This work was supported by a grant from the National Natural Science Foundation of China [Grant numbers 61966023 and 82160326].

Data availability

The datasets analyzed in the current study are available from the corresponding author upon reasonable request.

Declarations

Conflict of interest

The authors report no conflict of interest.

Ethics approval and consent to participate

This study was approved by the Medical Research Ethics Committee of the Lanzhou University Second Hospital (No. 2019A-102). All individuals understood the purpose and latent risks and signed informed consent.

Footnotes

Publisher's Note

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

Contributor Information

Ming Ke, Email: keming@lut.edu.cn.

Guangyao Liu, Email: lgy362263779@163.com.

References

  1. Agosta F et al (2023) Functional connectivity from disease epicenters in frontotemporal dementia. Neurology 100:e2290–e2303 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Allen EA et al (2011) A baseline for the multivariate comparison of resting-state networks. Front Syst Neurosci 5:2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Allen EA, Damaraju E, Plis SM, Erhardt EB, Eichele T, Calhoun VD (2014) Tracking whole-brain connectivity dynamics in the resting state. Cereb Cortex (new York: 1991) 24:663–676 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Altan A, Karasu S (2020) Recognition of COVID-19 disease from X-ray images by hybrid model consisting of 2D curvelet transform, chaotic salp swarm algorithm and deep learning technique. Chaos Solitons Fractals 140:110071 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Amiri S et al (2021) Effective connectivity between emotional and motor brain regions in people with psychogenic nonepileptic seizures (PNES). Epilepsy Behav E & B 122:108085 [DOI] [PubMed] [Google Scholar]
  6. Bacon EJ, Jin C, He D, et al (2023) Functional and effective connectivity analysis of drug-resistant epilepsy: a resting-state fMRI analysis. Front Neurosci 17:1163111. Published 2023 Apr 20. 10.3389/fnins.2023.1163111 [DOI] [PMC free article] [PubMed]
  7. Bartolini E et al (2014) Abnormal response to photic stimulation in juvenile myoclonic epilepsy: an EEG-fMRI study. Epilepsia 55:1038–1047 [DOI] [PubMed] [Google Scholar]
  8. Baykan B, Wolf P (2017) Juvenile myoclonic epilepsy as a spectrum disorder: a focused review. Seizure 49:36–41 [DOI] [PubMed] [Google Scholar]
  9. Bell AJ, Sejnowski TJ (1995) An information-maximization approach to blind separation and blind deconvolution. Neural Comput 7:1129–1159 [DOI] [PubMed] [Google Scholar]
  10. Bravo F et al (2020) Anterior cingulate and medial prefrontal cortex response to systematically controlled tonal dissonance during passive music listening. Hum Brain Mapp 41:46–66 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Caciagli L et al (2019) Abnormal hippocampal structure and function in juvenile myoclonic epilepsy and unaffected siblings. Brain J Neurol 142:2670–2687 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Caciagli L et al (2020) Motor hyperactivation during cognitive tasks: an endophenotype of juvenile myoclonic epilepsy. Epilepsia 61:1438–1452 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Carvalho KC, UchidaGuaranha CGPMSB, Laura MFF, Wolf P, Yacubian EMT (2016) Cognitive performance in juvenile myoclonic epilepsy patients with specific endophenotypes. Seizure 40:33–41 [DOI] [PubMed] [Google Scholar]
  14. Cendes F, Cascino GD (2019) MRI endophenotypes of heritability and cognitive dysfunction in juvenile myoclonic epilepsy. Neurology 93:571–572 [DOI] [PubMed] [Google Scholar]
  15. Chen Z, Calhoun V (2018) Effect of spatial smoothing on task fMRI ICA and functional connectivity. Front Neurosci 12:15 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Cunningham SI, Tomasi D, Volkow ND (2017) Structural and functional connectivity of the precuneus and thalamus to the default mode network. Hum Brain Mapp 38:938–956 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Di X, Biswal BB (2014) Identifying the default mode network structure using dynamic causal modeling on resting-state functional magnetic resonance imaging. Neuroimage 86:53–59 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Dixsaut L, Gräff J (2021) The medial prefrontal cortex and fear memory: dynamics, connectivity, and engrams. Int J Mol Sci 22:12113 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Eack SM, Wojtalik JA, Keshavan MS, Minshew NJ (2017) Social-cognitive brain function and connectivity during visual perspective-taking in autism and schizophrenia. Schizophr Res 183:102–109 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Engel J (2001) A proposed diagnostic scheme for people with epileptic seizures and with epilepsy: report of the ILAE task force on classification and terminology. Epilepsia 42:796–803 [DOI] [PubMed] [Google Scholar]
  21. Faulkner H et al (2021) Safety and efficacy of prothrombin complex concentrate (PCC) for anticoagulation reversal in patients undergoing urgent neurosurgical procedures: a systematic review and metaanalysis. Neurosurg Rev 44:1921–1931 [DOI] [PubMed] [Google Scholar]
  22. Fransson P, Marrelec G (2008) The precuneus/posterior cingulate cortex plays a pivotal role in the default mode network: Evidence from a partial correlation network analysis. Neuroimage 42:1178–1184 [DOI] [PubMed] [Google Scholar]
  23. Friston KJ, Kahan J, Biswal B, Razi A (2014a) A DCM for resting state fMRI. Neuroimage 94:396–407 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Friston KJ, Kahan J, Razi A, Stephan KE, Sporns O (2014b) On nodes and modes in resting state fMRI. Neuroimage 99:533–547 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Friston KJ et al (2016) Bayesian model reduction and empirical Bayes for group (DCM) studies. Neuroimage 128:413–431 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Gamberini M et al (2020) Thalamic afferents emphasize the different functions of macaque precuneate areas. Brain Struct Funct 225:853–870 [DOI] [PubMed] [Google Scholar]
  27. Gattuso JJ et al (2023) Default mode network modulation by psychedelics: a systematic review. Int J Neuropsychopharmacol 26:155–188 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Geng X, Xu J, Liu B, Shi Y (2018) Multivariate classification of major depressive disorder using the effective connectivity and functional connectivity. Front Neurosci 12:38 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Giorgi FS et al (2016) Social cognition in juvenile myoclonic epilepsy. Epilepsy Res 128:61–67 [DOI] [PubMed] [Google Scholar]
  30. Hattingen E et al (2014) Frontal and thalamic changes of GABA concentration indicate dysfunction of thalamofrontal networks in juvenile myoclonic epilepsy. Epilepsia 55:1030–1037 [DOI] [PubMed] [Google Scholar]
  31. Jia X et al (2018) Disrupted coupling between the spontaneous fluctuation and functional connectivity in idiopathic generalized epilepsy. Front Neurol 9:838 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Jiang S et al (2018) Aberrant thalamocortical connectivity in juvenile myoclonic epilepsy. Int J Neural Syst 28:1750034 [DOI] [PubMed] [Google Scholar]
  33. Jiang Z, Cai Y, Liu S, et al (2023) Decreased default mode network functional connectivity with visual processing regions as potential biomarkers for delayed neurocognitive recovery: A resting-state fMRI study and machine-learning analysis. Front Aging Neurosci 14:1109485. 2023 Jan 6. 10.3389/fnagi.2022.1109485 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Jiao Q et al (2011) Granger causal influence predicts BOLD activity levels in the default mode network. Hum Brain Mapp 32:154–161 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Jiruska P et al (2017) Update on the mechanisms and roles of high-frequency oscillations in seizures and epileptic disorders. Epilepsia 58:1330–1339 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Ke M, Li H, Liu G (2022) The local topological reconfiguration in the brain network after targeted hub dysfunction attacks in patients with juvenile myoclonic epilepsy. Front Neurosci 16:864040 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Ke M, Wang C, Liu G (2023) Multilayer brain network modeling and dynamic analysis of juvenile myoclonic epilepsy. Front Behav Neurosci 17:1123534 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Lee C, Kim S-M, Jung Y-J, Im C-H, Kim DW, Jung K-Y (2014) Causal influence of epileptic network during spike-and-wave discharge in juvenile myoclonic epilepsy. Epilepsy Res 108:257–266 [DOI] [PubMed] [Google Scholar]
  39. Leveque Y et al (2016) Altered intrinsic connectivity of the auditory cortex in congenital amusia. J Neurophysiol 116:88–97 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Li B, Daunizeau J, Stephan KE, Penny W, Hu D, Friston K (2011) Generalised filtering and stochastic DCM for fMRI. Neuroimage 58:442–457 [DOI] [PubMed] [Google Scholar]
  41. Lin C-Y et al (2022) Reduced sleep quality is related to poor quality of life in patients with juvenile myoclonic epilepsy, a case-control study. Life (basel, Switzerland) 12:434 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Liu F et al (2017) Dynamic functional network connectivity in idiopathic generalized epilepsy with generalized tonic–clonic seizure. Hum Brain Mapp 38:957–973 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Liu J, Tai Y-J, Wang L-N (2021) Topiramate for juvenile myoclonic epilepsy. Cochrane Database Syst Rev 11:CD010008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Liu G et al (2022) Aberrant dynamic structure-function relationship of rich-club organization in treatment-naïve newly diagnosed juvenile myoclonic epilepsy. Hum Brain Mapp 43:3633–3645 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Ma L et al (2015) Inhibitory behavioral control: a stochastic dynamic causal modeling study using network discovery analysis. Brain Connect 5:177–186 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Ma L et al (2022) Altered cerebro-cerebellar effective connectivity in new-onset juvenile myoclonic epilepsy. Brain Sci 12:1658 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Mason MF, Norton MI, Van Horn JD, Wegner DM, Grafton ST, Macrae CN (2007) Wandering minds: the default network and stimulus-independent thought. Science (new York) 315:393–395 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. McCutcheon RA, Reis Marques T, Howes OD (2020) Schizophrenia—an overview. JAMA Psychiat 77:201–210 [DOI] [PubMed] [Google Scholar]
  49. McGill ML et al (2012) Default mode network abnormalities in idiopathic generalized epilepsy. Epilepsy Behavior E & B 23:353–359 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Mohan A et al (2016) The significance of the default mode network (DMN) in neurological and neuropsychiatric disorders: a review. Yale J Biol Med 89:49–57 [PMC free article] [PubMed] [Google Scholar]
  51. Moher Alsady T, Blessing EM, Beissner F (2016) MICA-A toolbox for masked independent component analysis of fMRI data. Hum Brain Mapp 37:3544–3556 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Motoi H et al (2018) Phase-amplitude coupling between interictal high-frequency activity and slow waves in epilepsy surgery. Epilepsia 59:1954–1965 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. O’Donoghue MF, Duncan JS, Sander JW (1996) The national hospital seizure severity scale: a further development of the Chalfont seizure severity scale. Epilepsia 37:563–571 [DOI] [PubMed] [Google Scholar]
  54. Parsons N, Bowden SC, Vogrin S, D’Souza WJ (2020) Default mode network dysfunction in idiopathic generalised epilepsy. Epilepsy Res 159:106254. 10.1016/j.eplepsyres.2019.106254 [DOI] [PubMed] [Google Scholar]
  55. Pascalicchio TF et al (2007) Neuropsychological profile of patients with juvenile myoclonic epilepsy: a controlled study of 50 patients. Epilepsy Behavior E & B 10:263–267 [DOI] [PubMed] [Google Scholar]
  56. Ray S, Di X, Biswal BB (2016) Effective connectivity within the mesocorticolimbic system during resting-state in cocaine users. Front Hum Neurosci 10:563 [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Ray D, Bezmaternykh D, Mel’nikov M, Friston KJ, Das M (2021) Altered effective connectivity in sensorimotor cortices is a signature of severity and clinical course in depression. Proc Natl Acad Sci U S A 118(40):e2105730118. 10.1073/pnas.2105730118 10.1073/pnas.2105730118 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Razi A, Kahan J, Rees G, Friston KJ (2015) Construct validation of a DCM for resting state fMRI. Neuroimage 106:1–14 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Routley B, Shaw A, Muthukumaraswamy SD, Singh KD, Hamandi K (2020) Juvenile myoclonic epilepsy shows increased posterior theta, and reduced sensorimotor beta resting connectivity. Epilepsy Res 163:106324 [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Smallwood J, Bernhardt BC, Leech R, Bzdok D, Jefferies E, Margulies DS (2021) The default mode network in cognition: a topographical perspective. Nat Rev Neurosci 22:503–513 [DOI] [PubMed] [Google Scholar]
  61. Tang R, Razi A, Friston KJ, Tang Y-Y (2016) Mapping smoking addiction using effective connectivity analysis. Front Hum Neurosci 10:195 [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Wandschneider B, Thompson PJ, Vollmar C, Koepp MJ (2012) Frontal lobe function and structure in juvenile myoclonic epilepsy: a comprehensive review of neuropsychological and imaging data. Epilepsia 53:2091–2098 [DOI] [PubMed] [Google Scholar]
  63. Wang Y, Guo Y (2019) A hierarchical independent component analysis model for longitudinal neuroimaging studies. Neuroimage 189:380–400 [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Wang M, Zheng H, Du X, Dong G (2019) Mapping Internet gaming disorder using effective connectivity: a spectral dynamic causal modeling study. Addict Behav 90:62–70 [DOI] [PubMed] [Google Scholar]
  65. Wei L, Wu G-R, Bi M, Baeken C (2021) Effective connectivity predicts cognitive empathy in cocaine addiction: a spectral dynamic causal modeling study. Brain Imaging Behav 15:1553–1561 [DOI] [PubMed] [Google Scholar]
  66. Wei L et al (2022) Attenuated effective connectivity of large-scale brain networks in children with autism spectrum disorders, vol. 16 [DOI] [PMC free article] [PubMed]
  67. Xi Y-B et al (2021) Triple network hypothesis-related disrupted connections in schizophrenia: a spectral dynamic causal modeling analysis with functional magnetic resonance imaging. Schizophr Res 233:89–96 [DOI] [PubMed] [Google Scholar]
  68. Yang T et al (2022) Altered dynamic functional connectivity of striatal-cortical circuits in Juvenile Myoclonic Epilepsy. Seizure 101:103–108 [DOI] [PubMed] [Google Scholar]
  69. Zeng N, Wang M, Dong H, Du X, Dong G-H (2022) Altered dynamic interactions within frontostriatal circuits reflect disturbed craving processing in internet gaming disorder. CNS Spectr 27:109–117 [DOI] [PubMed] [Google Scholar]
  70. Zhang Z, Liu G, Zheng W, Shi J, Liu H, Sun Y (2020) Altered dynamic effective connectivity of the default mode network in newly diagnosed drug-naïve juvenile myoclonic epilepsy. NeuroImage Clin 28:102431 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary file 1 (DOCX 3697 kb) (3.6MB, docx)

Data Availability Statement

The datasets analyzed in the current study are available from the corresponding author upon reasonable request.

Articles from Cognitive Neurodynamics are provided here courtesy of Springer Science+Business Media B.V.

RESOURCES