Dissociation of Structural and Functional Connectivity and Metabolism in the Neocortex of Idiopathic Generalized Epilepsy: A Simultaneous PET/MRI Multimodal Study.

Dissociation of Structural and Functional Connectivity and Metabolism in the Neocortex of Idiopathic Generalized Epilepsy: A Simultaneous PET/MRI Multimodal Study - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Search Log in Primary site navigation Search Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice AJNR Am J Neuroradiol . 2025 Jun;46(6):1208–1215. doi: 10.3174/ajnr.A8612 Dissociation of Structural and Functional Connectivity and Metabolism in the Neocortex of Idiopathic Generalized Epilepsy: A Simultaneous PET/MRI Multimodal Study Chenyang Yao a, b, c , Jie Hu a, b, c , Bixiao Cui a, b, c , Jingjuan Wang a, b, c , Zhen-Ming Wang a, b, c , Yaqin Hou a, b, c , Kharel Sudeep a, b, c , Hongwei Yang a, b, c , Yihe Wang d , Yongzhi Shan d, e, f , Jie Lu a, b, c, ✉ Author information Article notes Copyright and License information a From the Department of Radiology and Nuclear Medicine (C.Y., J.H., J.W., B.C., Z.-M.W., Y.H., K.S., H.Y., J.L.), Xuanwu Hospital Capital Medical University, Beijing, China b Beijing Key Laboratory of Magnetic Resonance Imaging and Brain Informatics (C.Y., J.H., J.W., B.C., Z.-M.W., Y.H., K.S., H.Y., J.L.), Beijing, China c Key Laboratory of Neurodegenerative Diseases (C.Y., J.H., J.W., B.C., Z.-M.W., Y.H., K.S., H.Y., J.L.), Ministry of Education, Beijing, China d Department of Neurosurgery (Y.W., Y.S.), Xuanwu Hospital Capital Medical University, Beijing, China e International Neuroscience Institute (China-INI) (Y.S.), Beijing, China f Clinical Research Center for Epilepsy (Y.S.), Capital Medical University, Beijing, China. ✉ Please address correspondence to Jie Lu, MD, PhD, Department of Radiology and Nuclear Medicine, Xuanwu Hospital Capital Medical University, Beijing, China; e-mail: imaginglu@hotmail.com ✉ Corresponding author. Received 2024 Jul 31; Accepted 2024 Nov 21. © 2025 by American Journal of Neuroradiology PMC Copyright notice PMCID: PMC12152784  PMID: 40081849 Graphical Abstract Open in a new tab This study uses simultaneous PET/MRI, which integrates structural connectivity, functional connectivity, and glucose metabolism in IGE to elucidate abnormalities of the neocortex and to analyze the correlations between these abnormalities and clinical presentations. The results show dissociated patterns of abnormalities: structural connectivity abnormalities in the limbic and visual networks, functional connectivity deficits in the default mode network, and metabolic reductions in the frontoparietal and somatomotor networks. Abstract BACKGROUND AND PURPOSE: Idiopathic generalized epilepsy (IGE) accounts for approximately 20% of epilepsy cases. Characterized by generalized spike-wave discharge, IGE is increasingly recognized as a network disorder with potential metabolic underpinnings. This study leverages the advantages of simultaneous PET/MRI, which enables the concurrent acquisition of MRI and PET data, to integrate structural connectivity (SC), functional connectivity (FC), and glucose metabolism into a unified framework. This study aims to elucidate the multimodal abnormalities of the neocortex in IGE, to analyze the correlations between these abnormalities and clinical presentations, and to investigate the interactions among different imaging modalities. MATERIALS AND METHODS: Twenty-one patients with IGE and 34 healthy controls (HCs) were recruited. Simultaneous PET/MRI scans were performed, incorporating DTI, resting-state fMRI, and [ 18 F]FDG-PET. DTI generated a neocortical connectivity blueprint, while resting-state fMRI provided a whole-brain connectivity matrix. [ 18 F]FDG-PET data were processed to obtain standardized uptake value ratios (SUVRs). Multivariate distance matrix regression was used to identify abnormal neocortical regions in SC and FC. Differences in SUVRs were identified by using least absolute shrinkage and selection operator regression. Statistical analyses, including t tests, linear models, mediation analysis, and Pearson correlations, were conducted to compare values of each technique between groups and explore relationships with clinical features. RESULTS: SC abnormalities were primarily found in the limbic (40% of all abnormal neocortical regions) and visual networks (31%), while FC abnormalities were mostly in the default mode network (DMN, 45%). Metabolic abnormalities were predominantly in the frontoparietal (26%) and somatomotor (22%) networks. SC in the limbic was positively correlated with onset age, while seizure frequency was negative correlated with DMN FC and positively correlated with frontoparietal metabolism. Mediation analysis showed that DMN FC mediated the relationship between limbic SC and frontoparietal and somatomotor metabolism. CONCLUSIONS: A multimodal approach reveals distinct and interrelated abnormalities in IGE, with different modalities reflecting various aspects of the disease, thus enhancing our understanding of its complex mechanisms. This integrative analysis could inform more effective treatments. SUMMARY PREVIOUS LITERATURE: Prior studies have recognized IGE as a network and metabolic disorder, with disruptions observed across structural, functional, and metabolic aspects of the brain. Neuroimaging findings indicate that structural alterations are more stable, while functional changes often reflect short-term dynamics. Hypometabolism in specific brain regions, though not visually apparent in patients with IGE, can be detected by using quantitative analyses. These imaging modalities reveal distinct yet interconnected dimensions of IGE, aiding in a comprehensive understanding of the disorder. KEY FINDINGS: This study found dissociated patterns of abnormalities in IGE across multimodalities: structural connectivity abnormalities were concentrated in the limbic and visual networks, functional connectivity deficits were most prominent in the default mode network, and metabolic reductions were observed in the frontoparietal and somatomotor networks. KNOWLEDGE ADVANCEMENT: The study advances understanding of IGE by revealing how distinct imaging modalities capture complementary aspects of disease mechanisms. The integration of structural, functional, and metabolic data underscores the importance of a multimodal approach in characterizing IGE and offers insights for developing more targeted and comprehensive treatment strategies. Idiopathic generalized epilepsy (IGE) is a common neurologic disorder, comprising approximately 20% of all epilepsy cases. 1 The prevalence of psychiatric and other comorbidities, including sudden unexpected death in epilepsy, is similar to that in focal epilepsy. 2 Generalized epilepsy is a clinical syndrome typically characterized by paroxysmal generalized spike-wave discharges (GSWD), and recent evidence suggests it to be a network disorder. 3 - 5 Additionally, epilepsy can be considered a metabolic disease, with primordial mechanisms triggered by compounds such as adenosine being highly relevant to seizures and epileptogenesis. 6 Despite advances in neuroimaging and clinical research, the precise pathophysiologic mechanisms of IGE remain elusive and warrant further investigation. Neuroimaging offers new insights into the structural, functional, and metabolic aspects of the brain that may contribute to the pathophysiology of IGE. 7 Although the term generalized implies that all brain neurons are synchronously and homogeneously affected in IGE, electrical neuroimaging, and molecular studies indicate the involvement of specific cortical networks while sparing others. 8 In patients with IGE, structural imaging studies reveal atrophy, and functional imaging studies show abnormal nodes in frontocentral cortical areas. Reduced connectivity in the default mode network (DMN) may be related to drug resistance in IGE. 9 DMN connectivity negatively correlates with seizure duration. 10 , 11 On PET, patients with epilepsy usually show hypometabolism in specific brain regions. 12 However, in patients with IGE, regions of hypometabolism are usually typically not visually apparent, but quantitative analyses at the vertex level can identify metabolic abnormalities. These different modalities reflect various facets of IGE, with distinct anatomic, functional, and metabolic dissociations observed. Previous studies suggest that fMRI captures physiologic alterations related to acute episodes, while structural changes reflect more stable and long-standing modifications in the brain. 13 Integrating these diverse imaging modalities into a unified analytical framework can enhance our understanding of the complex mechanisms of IGE. This comprehensive approach allows for a detailed examination of disease-related changes across different imaging modalities and facilitates the exploration of interrelationships among these modalities. This study focuses on the neocortex, utilizing structural blueprint techniques to pinpoint regions with disrupted connectivity to major white matter fiber tracts in patients with IGE. Utilizing multivariate distance matrix regression (MDMR) and least absolute shrinkage and selection operator (LASSO) regression, we can identify IGE-imaging correlations at the individual vertex level without relying on prior hypotheses. Additionally, we aim to identify areas with abnormal whole-brain functional connectivity (FC) and glucose metabolism and to examine the relationships between these abnormalities and clinical features from the perspective of large-scale brain networks. To achieve these goals, we utilize simultaneous PET/MRI to acquire MRI and PET data concurrently, integrating DTI, blood oxygen level–dependent (BOLD) fMRI, and [ 18 F]FDG-PET data. This study aims to elucidate the distinctive structural, functional, and metabolic abnormalities in the neocortex of patients with IGE, analyze the correlations between these abnormalities and clinical presentations, and investigate the interactions among different imaging modalities. We seek to provide a more holistic view of the pathophysiologic underpinnings of IGE and advance our understanding of how structural, functional, and metabolic alterations interrelate in this disorder. MATERIALS AND METHODS Participants Twenty-one patients with IGE (mean age, 27.67 ± 11.15 years, 8 men) were recruited between December 2021 and October 2023 according to International League Against Epilepsy classification. 14 All patients were further classified into subtypes: 5 patients had childhood absence epilepsy, 10 patients had juvenile myoclonic epilepsy, and 6 patients had epilepsy with generalized tonic-clonic seizures alone. 15 A total of 61.9% (13/21) of patients had received monotherapy, and the others had received multiple drugs. All patients in this study still experienced epileptic seizures, but 16 of the 21 drug-treated patients responded well to antiepileptic drugs, showing significantly decreased seizure frequency. None had brain lesions, developmental disabilities, or other neurologic disorders. Thirty-four sex- and age-matched healthy controls (HCs) (mean age, 31.12 ± 6.35 years, 16 men) were recruited, without neurologic or psychiatric disorders based on a health screening measure. Ethics approval was obtained from the ethics committee of Xuanwu Hospital, and written informed consent was provided by each participant. Detailed demographic and clinical information are provided in the Table . The demographics and clinical characteristics of patients with IGE and healthy controls Patients ( n =21) Controls ( n =34) P Value Age 27.67 ± 11.15 31.12 ± 6.35 .21 a Sex (male/female) 8/13 16/18 .52 b Education (years) 12.07 ± 3.30 13.35 ± 3.18 .16 a Age of onset (years) 15.90 ± 10.02 Duration of illness (years) 12.57 ± 9.27 Seizure frequency (times/week) 1.75 ± 1.10 Subtypes (CAE/JME/GTCA) 5/10/6 AED (response/nonresponse) 13/8 Therapy (single/multiple) 16/5 Open in a new tab a P value by 2-sample t test. b P value by 2-tailed χ 2 test. Note: —Data are presented as means (standard deviation). AED indicates antiepileptic drug; CAE = childhood absence epilepsy; GTCA = epilepsy with generalized tonic-clonic seizures alone; JME = juvenile myoclonic epilepsy. MRI Acquisition and Preprocessing Interictal [ 18 F]FDG-PET and MRI data were simultaneously obtained by using a hybrid TOF-PET/MRI scanner (Signa, GE Healthcare). MRI sequences included 3D T1 BRAVO (GE Healthcare), BOLD, and DTI. DTI data preprocessing generated a neocortical connectivity blueprint between the neocortex and major white matter fiber tracts. A whole-brain connectivity matrix was calculated based on the BOLD images. PET images were preprocessed to obtain a standardized uptake value ratio (SUVR) map. The preprocessing results for these modalities were projected into the fsaverage5 space for subsequent statistical analyses and comparisons. Details about PET/MRI acquisition and data postprocessing are provided in the Supplemental Data. IGE-Related Abnormal Neocortical Regions Identification For DTI and BOLD, a connectome-wide association study used MDMR. 16 The correlation coefficients for each vertex in the structural connectivity (SC) map were compared with the values of corresponding vertices in maps generated from every other participant. This resulted in a Pearson correlation coefficient r that measures the spatial correlation of maps between patients. The similarity of SC maps between participants was computed by using the distance metric 2 ( 1 − r ) . MDMR then tested the relationship between intersubject differences on a grouping variable and intersubject distances, resulting in a pseudo- F statistic for each vertex that demonstrated how intergroup differences were reflected in SC at that vertex. This process was repeated for every single vertex, resulting in a whole neocortical map of IGE reflected in SC at each vertex. A 5000-permutation test of the pseudo- F values yields the corresponding P values. Multiple comparisons were corrected by using the false discovery rate (FDR) at a significance level of P < .05. Covariates in this analysis included sex and age. The FC maps underwent the same MDMR analysis as described above. To identify neocortical regions with differences in PET SUVR maps between patients with IGE and HCs, we used a LASSO regression model. LASSO regression is an L1-norm regularization technique that introduces a shrinkage penalty term (λ) to prevent model overfitting, compelling some predictor coefficients to shrink to zero. 17 The tuning parameter (λ) was optimized through a repeated 5-fold cross-validation method and subsequently applied to predict the remaining subjects within the cycles. We selected the smallest mean-squared error to determine the coefficients of the SUVRs. SUVRs with nonzero coefficients were considered important features and were retained ( Fig 1 ). Additionally, we also used LASSO regression modeling to identify abnormalities of neocortical thickness in patients with IGE. FIG 1. Open in a new tab Flowchart of this study. The neocortex white-gray matter boundaries from 3D-T1 images were extracted as masks after the Recon-all process. These masks were combined with probabilistic tractography to calculate SC blueprints. BOLD MRI was used to calculate whole-neocortex FC for each vertex. [ 18 F]FDG-PET was used to calculate the neocortical SUVRs. MDMR was performed on each vertex’s SC and FC to obtain pseudo- F values and P values. Repeated execution of MDMR produced a neocortical structural and functional connectome-wide association map. LASSO regression was used to for SUVRs. Statistical Analysis The proportion of structurally connected, functionally connected, and metabolically abnormal brain regions in each network, and their mean values, were calculated according to the 7 functional networks. 18 Mean values were compared between the patients with IGE and HCs by using linear models with sex and age as covariates and adjusting for FDR. For features with abnormalities in each technique, an analysis of covariance was performed to determine whether they differed in IGE subtypes, and the linear model was performed to determine whether they differed in antiepileptic drug response and medication type. The features were used as the dependent variable, while the independent variable was the patient’s subtypes or clinical presentations, with age and sex as covariates. In cases of significant differences, post hoc analyses using the least significant difference method were conducted to pinpoint the specific groups that exhibited significant distinctions. Pearson correlations were performed between each functional network mean for each technique and age of onset, and seizure frequency. We focus on networks in which the distribution of significant vertices in each technique accounted for more than 50% of the total distribution across other networks (ie, the distribution accounts for more than 21%) for the mediation analysis. This can help focus the analysis on the most prominent and relevant networks, reduce noise, and increase the likelihood of identifying meaningful relationships among the variables of interest. Mediation analysis was conducted by using the “mediation” package in R . Notably, SC was used as an independent variable in this study because previous research has indicated that fMRI typically reflects short-term changes in brain activity, while structural imaging focuses on long-term structural alterations in the brain. 13 Because the relationship between FC and FDG metabolism is not well defined, we therefore analyzed FC and FDG as mediator and outcome variables, respectively. Demographic and clinical data were compared between the 2 groups by using the χ 2 test for sex distribution and the 2-sample t test for age, education, age of onset, duration of illness, and seizure frequency. All statistical analyses were conducted by using R (Version 4.2.1), with significance set at P < .05. This study methodology follows the guidelines outlined in the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis Checklist. 19 RESULTS Distribution of Each Technique and Correlation with Clinical Features SC. For SC, the neocortical regions with abnormalities in patients with IGE were primarily in the limbic (40% of the total) and the visual network (VN, 31%), with the frontoparietal network (FPN, 2%), and somatomotor network (SMN, 3%) being the least affected ( Fig 2 , Supplemental Data). The SC of limbic (Δmean, −0.001; 95% CI, −0.002 to −0.001) and VN (Δmean, −0.001; 95% CI, −0.002 to −0.0005) were significantly lower in patients with IGE than in HCs ( Fig 3 A ). Additionally, the neocortical thickness with abnormalities in patients with IGE were primarily in the limbic network (46%, Supplemental Data), which were significantly thinner in patients (Δmean, −0.170 mm; 95% CI, −0.103 to −0.237). FIG 2. Open in a new tab Distribution of abnormal brain regions for each technique in patients with IGE ( A ) and functional network affiliation of the regions ( B ). Larger spherical volumes represent larger pseudo- F values in ( A ). FIG 3. Open in a new tab Mean values of SC and FC, as well as SUVR, in patients with IGE and HCs, and their correlation with clinical characteristics. We found that SC in the limbic network of patients with IGE was positively correlated with onset age ( r = 0.45, P = .04), with SC becoming weaker at younger onset ages ( Fig 3 D ). However, cortical thickness of the limbic network was not significantly correlated with either onset age or seizure frequency. FC. For FC, the abnormal neocortical regions were mostly located in the DMN (45%, Fig 2 , Supplemental Data). For the FC of the DMN, patients with IGE showed significantly reduced connectivity in the default A (Δmean, −0.021; 95% CI, −0.034 to −0.008) and default C (Δmean, −0.034; 95% CI, −0.051 to −0.016) networks ( Fig 3 B ). FC of the DMN was negatively correlated with the seizure frequency ( r = −0.52, P = .02, Fig 3 E ). FDG-PET. The brain regions with metabolic abnormalities in patients with IGE were mainly in the FPN (26%) and SMN (22%), with the limbic (4%) and visual (5%) networks being the least affected ( Fig 2 , Supplemental Data). We found that FDG metabolism in several subregions of the FPN and SMN in patients with IGE was significantly lower than in HCs, including control A (Δmean, −0.097; 95% CI, −0.192 to −0.002), control C (Δmean, −0.217; 95% CI, −0.396 to −0.038), and somatomotor B (Δmean, −0.199; 95% CI, −0.297 to −0.101) networks ( Fig 3 C ). FDG metabolism in the FPN was positively correlated with seizure frequency ( r = 0.57, P = .01, Fig 3 F ). Please see Supplemental Data for details on the differences in vertices within the large-scale networks of the 3 models in IGE and HCs, as well as the correlations between clinical features and the vertices of SC, FC, and FDG-PET in the large-scale networks of IGE. Supplemental Data show the results of the subgroup comparisons. Mediation Analysis We observed that FC in the DMN mediates the relationship between SC in the limbic network and glucose metabolism in the FPN (average causal mediation effect [ACME] = 0.003, P < .001) and SMN (ACME = 0.004, P = .02, Fig 4 ). In contrast, glucose metabolism in the FPN and SMN did not have significant mediating effects between SC in the limbic network and FC in the DMN. Additionally, mediation analyses based on the SC of the VN were not associated with any significant mediation effects. The average causal mediation effect showed the detailed results of the above mediation analyses. FIG 4. Open in a new tab Mediation analysis between the SC of the limbic network, FC of the DMN, and FDG metabolism in the FPN and SMN networks. For the estimate and P values, the value before the “/” represents the FPN FDG as outcome variable, while the number after the “/” indicates the SMN FDG as outcome variable. ADE indicates average direct effect. DISCUSSION Our study observed a dissociated pattern of abnormal neocortical regions for SC, FC, and glucose metabolism in patients with IGE. The primary SC abnormalities in patients with IGE were found in the limbic network and VN, while whole-brain FC abnormalities were concentrated in the DMN. Abnormalities in glucose metabolism were distributed in the FPN and SMN. Additionally, SC of the limbic network was positively correlated with age at onset, whereas the FC of the DMN and glucose metabolism of the FPN were associated with seizure frequency. Moreover, the FC of the DMN mediated the relationship between SC of the limbic network and glucose metabolism of the FPN and SMN in patients with IGE. Seizure initiation and spread seizures are closely related to intracerebral circuits, such as the Papez circuit, 3 which has a high degree of anatomic overlap with the limbic network, suggesting a significant role in IGE pathogenesis. 20 Neuromodulation of these circuits has been shown to alleviate epileptic symptoms, indicating its importance in the IGE. 21 This study also found that SC in the llimbic network was disorder and positively correlated with the age at onset of IGE, suggesting that earlier onset is associated with weaker SC. During early brain development, critical periods exist when neural circuits are particularly plastic and vulnerable to disruptions. IGE may interfere with the normal development and maturation of the limbic network. 22 , 23 Early onset of epilepsy could disrupt the formation of white matter tracts and synaptic connections, leading to lower SC. 22 Additionally, cumulative damage from prolonged seizures, chronic neuroinflammation due to recurrent seizures, excitotoxicity, and metabolic stress, all of which can damage neural connections, 24 reduce structural integrity, and result in lower connectivity. 25 Consequently, SC reflects cumulative brain damage and long-term alterations in IGE. SC of the VN was also decreased, with VN and limbic changes often converging in patients with IGE, 26 which may be due to their spatial proximity and reflect the similar disease dynamics of IGE. Despite some research linking VN dysfunction with photosensitivity or disrupted visual perception in epilepsy, clinical evidence was lacking in this study. 27 , 28 The deactivation of the DMN and activation of the thalamus are consistently reported as responsible for generation and propagation epileptic activities. 29 The weak whole-brain connectivity of the DMN found in this study supports this view. Epileptic activity, particularly generalized seizures, can disrupt normal brain network functioning. The DMN, crucial for maintaining a resting state and involved in high-level cognitive processes like self-referential thinking and memory, may be particularly susceptible to these disruptions, which leads to reduced FC efficiency and coherence within the DMN. The current conceptualization of DMN anatomy resembles the unitary model of the limbic system, which, through the coordination of its subregions, subserves the elaboration of emotion, memories, and behavior. 30 , 31 This finding of mediation analysis underscores the pivotal role of the DMN in integrating structural and metabolic abnormalities in IGE. Epileptic activity can induce hypometabolism in brain regions involved in seizure propagation and generalization. 32 The FPN and SMN might be particularly affected by this hypometabolic state due to their involvement in cognitive and motor functions often disrupted in IGE. Simultaneous electroencephalogram-fMRI studies have shown hypersynchronization of SMN often occurs in later ictal and postictal periods, suggesting its passive involvement in GSWD. 33 Patients with IGE exhibit mild or severe cognitive or motor disorders, whether they have absence seizures, myoclonic, or generalized seizures. Thus, these 2 network abnormalities may cause or be affected by seizures. One interpretation of hypometabolism is that chronic epilepsy is associated with neuroinflammation and neuronal damage, affecting metabolic activity. 34 , 35 Microglial activation and cytokine release associated with chronic seizures can contribute to this metabolic impairment. Additionally, in response to recurrent seizures, the brain might undergo compensatory mechanisms and network reorganization, involving a shift in metabolic activity to other regions or networks to maintain overall brain function. 36 This compensatory reallocation of resources might result in reduced metabolic activity in the FPN and SMN as the brain attempts to mitigate epilepsy’s impact. FC in the DMN and glucose metabolism in the FPN were correlated with seizure frequency, indicating that these modalities might reflect the symptom and more dynamic and reversible aspects of the disease. These seizures impair synaptic plasticity, neurotransmitter balance, and SC, all of which are essential for maintaining functional and metabolic activity. 6 , 37 Additionally, compensatory mechanisms attempting to mitigate seizure effects may further alter FC and metabolic processes. 38 , 39 Each technique captures different aspects of brain, and the lack of correlation could reflect the complexity of brain network interactions, where changes in 1 domain (eg, SC) do not always correspond directly to changes in another (eg, metabolism) at the same regions. However, different modalities and brain regions can relate to each other. This complex interplay highlights the importance of considering multiple modalities to fully understand the pathophysiology of IGE. While no randomized controlled trials have established the efficacy of neuromodulation in IGE, open-label studies suggest potential benefits. 1 , 40 Common neuromodulation approaches target several key brain regions, including the thalamus, limbic system, and specific cortical areas. 41 - 43 However, the effectiveness of these targets is inconsistent across patients. Our study highlights abnormal brain regions, including the DMN and FPN, which could serve as novel neuromodulation targets. Modulating the excitability, activity, and energy demand of them may reduce seizure frequency in IGE. Moreover, the SC in the limbic system could serve as a valuable marker for evaluating the effectiveness of the treatments. Our study also has some limitations. First, the long-term use of antiepileptic drugs can influence brain metabolism, and the drug selection and medication dosage were not controlled in this study. The side effects of these medications could compound the structural, functional, or metabolic disturbances caused by the disease itself, leading to further neuroimaging changes. Although challenging, future studies could mitigate this concern by selecting drug-naïve patients. Second, the sample size of our study is small, so the results need to be interpreted with caution. Though there was no statistically significant difference between the ages of patients and HCs, the mean age of HCs was slightly older. Therefore, the results of this study merit replication in future studies with larger sample sizes. Additionally, a limitation of our mediation analysis is that it is based on assumptions derived from previous research, which primarily considers the progression the disease over time as the causal framework, rather than focusing on the causal relationships between the physiologic changes reflected by different imaging modalities. As a result, our analysis is exploratory in nature, and while it provides valuable insights, the interpretation and application of the findings should be approached with caution, pending further studies to validate these relationships. CONCLUSIONS Our study underscores the complexity of disease-related changes in IGE. While SC, FC, and glucose metabolism each reveal distinct abnormalities, the dissociation and limited correlation between these modalities suggests that they may reflect different aspects of the disorder, possibly operating independently or at different stages of disease progression. By integrating multiple neuroimaging modalities, we can advance our understanding of IGE and pave the way for more comprehensive and effective treatments. ABBREVIATIONS: ACME average causal mediation effect BOLD blood oxygen level–dependent DMN default mode network FC functional connectivity FDR false discovery rate FPN frontoparietal network GSWD generalized spike-wave discharges HC healthy control IGE idiopathic generalized epilepsy LASSO least absolute shrinkage and selection operator MDMR multivariate distance matrix regression SC structural connectivity SMN somatomotor network SUVR standardized uptake value ratio VN visual network Footnotes Chenyang Yao and Jie Hu contributed equally to this article. This study was supported by the National Key Research and Development Program of China (No.2022YFC2406900,2022YFC2406904) and UCB Pharma Ltd. Joint Science Research Foundation of China Association Against Epilepsy(CU-2023-023). Disclosure forms provided by the authors are available with the full text and PDF of this article at www.ajnr.org . References 1. Devinsky O, Elder C, Sivathamboo S, et al. Idiopathic generalized epilepsy: misunderstandings, challenges, and opportunities. Neurology 2024;102:e208076 10.1212/WNL.0000000000208076 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 2. Verducci C, Hussain F, Donner E, et al. SUDEP in the North American SUDEP Registry: the full spectrum of epilepsies. Neurology 2019;93:e227–36 10.1212/WNL.0000000000007778 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 3. Piper RJ, Richardson RM, Worrell G, et al. Towards network-guided neuromodulation for epilepsy. Brain 2022;145:3347–62 10.1093/brain/awac234 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 4. Moeller F, Maneshi M, Pittau F, et al. Functional connectivity in patients with idiopathic generalized epilepsy. Epilepsia 2011;52:515–22 10.1111/j.1528-1167.2010.02938.x [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 5. Acharya UR, Oh SL, Hagiwara Y, et al. Deep convolutional neural network for the automated detection and diagnosis of seizure using EEG signals. Comput Biol Med 2018;100:270–78 10.1016/j.compbiomed.2017.09.017 [ DOI ] [ PubMed ] [ Google Scholar ] 6. Rho JM, Boison D. The metabolic basis of epilepsy. Nat Rev Neurol 2022;18:333–47 10.1038/s41582-022-00651-8 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 7. Liao C, Wang K, Cao X, et al. Detection of lesions in mesial temporal lobe epilepsy by using MR fingerprinting. Radiology 2018;288:804–12 10.1148/radiol.2018172131 [ DOI ] [ PubMed ] [ Google Scholar ] 8. Larivière S, Rodríguez-Cruces R, Royer J, et al. Network-based atrophy modeling in the common epilepsies: a worldwide ENIGMA study. Sci Adv 2020;6:eabc6457 10.1126/sciadv.abc6457 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 9. Wandschneider B, Hong SJ, Bernhardt BC, et al. Developmental MRI markers cosegregate juvenile patients with myoclonic epilepsy and their healthy siblings. Neurology 2019;93:e1272–e1280 10.1212/WNL.0000000000008173 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 10. McGill ML, Devinsky O, Kelly C, et al. Default mode network abnormalities in idiopathic generalized epilepsy. Epilepsy Behav 2012;23:353–59 10.1016/j.yebeh.2012.01.013 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Hu J, Ran H, Chen G, et al. Altered neurovascular coupling in children with idiopathic generalized epilepsy. CNS Neurosci Ther 2023;29:609–18 10.1111/cns.14039 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 12. Wang J, Guo K, Cui B, et al. Individual [18F]FDG PET and functional MRI based on simultaneous PET/MRI may predict seizure recurrence after temporal lobe epilepsy surgery. Eur Radiol 2022;32:3880–88 10.1007/s00330-021-08490-9 [ DOI ] [ PubMed ] [ Google Scholar ] 13. Gong Q, Lui S, Sweeney JA. A selective review of cerebral abnormalities in patients with first-episode schizophrenia before and after treatment. Am J Psychiatry 2016;173:232–43 10.1176/appi.ajp.2015.15050641 [ DOI ] [ PubMed ] [ Google Scholar ] 14. Engel J Jr; International League Against Epilepsy (ILAE). A proposed diagnostic scheme for people with epileptic seizures and with epilepsy: report of the ILAE Task Force on Classification and Terminology. Epilepsia 2001;42:796–803 10.1046/j.1528-1157.2001.10401.x [ DOI ] [ PubMed ] [ Google Scholar ] 15. Hirsch E, French J, Scheffer IE, et al. ILAE definition of the idiopathic generalized epilepsy syndromes: position statement by the ILAE Task Force on Nosology and Definitions. Epilepsia 2022;63:1475–99 10.1111/epi.17236 [ DOI ] [ PubMed ] [ Google Scholar ] 16. Shehzad Z, Kelly C, Reiss PT, et al. A multivariate distance-based analytic framework for connectome-wide association studies. NeuroImage 2014;93(Pt 1):74–94 10.1016/j.neuroimage.2014.02.024 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 17. Dwyer DB, Falkai P, Koutsouleris N. Machine learning approaches for clinical psychology and psychiatry. Annu Rev Clin Psychol 2018;14:91–118 10.1146/annurev-clinpsy-032816-045037 [ DOI ] [ PubMed ] [ Google Scholar ] 18. Yeo BTT, Krienen FM, Sepulcre J, et al. The organization of the human cerebral cortex estimated by intrinsic functional connectivity. J Neurophysiol 2011;106:1125–65 10.1152/jn.00338.2011 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 19. Collins GS, Reitsma JB, Altman DG, et al. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): the TRIPOD statement. BMJ 2015;350:g7594 10.1136/bmj.g7594 [ DOI ] [ PubMed ] [ Google Scholar ] 20. Perucca E, Perucca P, White HS, et al. Drug resistance in epilepsy. Lancet Neurol 2023;22:723–34 10.1016/S1474-4422(23)00151-5 [ DOI ] [ PubMed ] [ Google Scholar ] 21. Wu D, Schaper F, Jin G, et al. Human anterior thalamic stimulation evoked cortical potentials align with intrinsic functional connectivity. Neuroimage 2023;277:120243 10.1016/j.neuroimage.2023.120243 [ DOI ] [ PubMed ] [ Google Scholar ] 22. Hatton SN, Huynh KH, Bonilha L, et al. White matter abnormalities across different epilepsy syndromes in adults: an ENIGMA-Epilepsy study. Brain 2020;143:2454–73 10.1093/brain/awaa200 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 23. Hogan RE. Epilepsy as a disease of white matter. Epilepsy Curr 2021;21:27–29 10.1177/1535759720975744 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 24. Scott RC. What are the effects of prolonged seizures in the brain? Epileptic Disord 2014;16(Spec No 1):S6–11 10.1684/epd.2014.0689 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 25. Thompson K. Status epilepticus and early development: neuronal injury, neurodegeneration, and their consequences. Epilepsia Open 2023;8(Suppl 1):S110–16 10.1002/epi4.12601 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 26. Jiang S, Pei H, Huang Y, et al. Dynamic temporospatial patterns of functional connectivity and alterations in idiopathic generalized epilepsy. Int J Neural Syst 2020;30:2050065 10.1142/S0129065720500653 [ DOI ] [ PubMed ] [ Google Scholar ] 27. Anderson J, Hamandi K. Understanding juvenile myoclonic epilepsy: contributions from neuroimaging. Epilepsy Res 2011;94:127–37 10.1016/j.eplepsyres.2011.03.008 [ DOI ] [ PubMed ] [ Google Scholar ] 28. Bartolini E, Pesaresi I, Fabbri S, et al. Abnormal response to photic stimulation in juvenile myoclonic epilepsy: an EEG-fMRI study. Epilepsia 2014;55:1038–47 10.1111/epi.12634 [ DOI ] [ PubMed ] [ Google Scholar ] 29. Gotman J, Grova C, Bagshaw A, et al. Generalized epileptic discharges show thalamocortical activation and suspension of the default state of the brain. Proc Natl Acad Sci U S A 2005;102:15236–40 10.1073/pnas.0504935102 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 30. Catani M, Dell’acqua F, Thiebaut de Schotten M. A revised limbic system model for memory, emotion and behaviour. Neurosci Biobehav Rev 2013;37:1724–37 10.1016/j.neubiorev.2013.07.001 [ DOI ] [ PubMed ] [ Google Scholar ] 31. Alves PN, Foulon C, Karolis V, et al. An improved neuroanatomical model of the default-mode network reconciles previous neuroimaging and neuropathological findings. Commun Biol 2019;2:370 10.1038/s42003-019-0611-3 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 32. Guo K, Wang J, Cui B, et al. [ 18 F]FDG PET/MRI and magnetoencephalography may improve presurgical localization of temporal lobe epilepsy. Eur Radiology 2022;32:3024–34 10.1007/s00330-021-08336-4 [ DOI ] [ PubMed ] [ Google Scholar ] 33. Tangwiriyasakul C, Perani S, Centeno M, et al. Dynamic brain network states in human generalized spike-wave discharges. Brain 2018;141:2981–94 10.1093/brain/awy223 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 34. Ravizza T, Scheper M, Di Sapia R, et al. mTOR and neuroinflammation in epilepsy: implications for disease progression and treatment. Nat Rev Neurosci 2024;25:334–50 10.1038/s41583-024-00805-1 [ DOI ] [ PubMed ] [ Google Scholar ] 35. Vezzani A, Balosso S, Ravizza T. Neuroinflammatory pathways as treatment targets and biomarkers in epilepsy. Nat Rev Neurol 2019;15:459–72 10.1038/s41582-019-0217-x [ DOI ] [ PubMed ] [ Google Scholar ] 36. Khambhati AN, Shafi A, Rao VR, et al. Long-term brain network reorganization predicts responsive neurostimulation outcomes for focal epilepsy. Sci Transl Med 2021;13:eabf6588 10.1126/scitranslmed.abf6588 [ DOI ] [ PubMed ] [ Google Scholar ] 37. Akyuz E, Polat AK, Eroglu E, et al. Revisiting the role of neurotransmitters in epilepsy: an updated review. Life Sci 2021;265:118826 10.1016/j.lfs.2020.118826 [ DOI ] [ PubMed ] [ Google Scholar ] 38. Pan G, Chen Z, Zheng H, et al. Compensatory mechanisms modulate the neuronal excitability in a kainic acid-induced epilepsy mouse model. Front Neural Circuits 2018;12:48 10.3389/fncir.2018.00048 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 39. Qi Y, Zhang YM, Gao YN, et al. AMPK role in epilepsy: a promising therapeutic target? J Neurol 2024;271:748–71 10.1007/s00415-023-12062-w [ DOI ] [ PubMed ] [ Google Scholar ] 40. O’Muircheartaigh J, Vollmar C, Barker GJ, et al. Abnormal thalamocortical structural and functional connectivity in juvenile myoclonic epilepsy. Brain 2012;135:3635–44 10.1093/brain/aws296 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 41. Tyvaert L, Chassagnon S, Sadikot A, et al. Thalamic nuclei activity in idiopathic generalized epilepsy: an EEG-fMRI study. Neurology 2009;73:2018–22 10.1212/WNL.0b013e3181c55d02 [ DOI ] [ PubMed ] [ Google Scholar ] 42. Sisterson ND, Kokkinos V, Urban A, et al. Responsive neurostimulation of the thalamus improves seizure control in idiopathic generalised epilepsy: initial case series. J Neurol Neurosurg Psychiatry 2022;93:491–98 10.1136/jnnp-2021-327512 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 43. Welch WP, Hect JL, Abel TJ. Case report: responsive neurostimulation of the centromedian thalamic nucleus for the detection and treatment of seizures in pediatric primary generalized epilepsy. Front Neurol 2021;12:656585 10.3389/fneur.2021.656585 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Articles from AJNR: American Journal of Neuroradiology are provided here courtesy of American Society of Neuroradiology ACTIONS View on publisher site PDF (2.3 MB) Cite Collections Permalink RESOURCES Similar articles Cited by other articles Links to NCBI Databases Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top

Related documents

• Altered Regional Brain Activity and Functional Connectivity Between Non-Diabetic and Diabetic Kidney Disease: A Resting-State fMRI Study.#28245
• Score-based generative diffusion models to synthesize full-dose FDG brain PET from MRI in epilepsy patients.#330841
• Brain protein burden is related to intravoxel incoherent motion: PET-MR imaging study.#330709
• The dynamic functional connectivity peak index: Detection of interictal epileptic activity with fMRI.#1039964
• Structural and Functional Connectivity in Partial Epilepsies Studied with MRI and MEG#861193