Dynamic network reconfiguration in hepatitis B cirrhosis secondary to mild hepatic encephalopathy: a multilayer network analysis
Original Article

Dynamic network reconfiguration in hepatitis B cirrhosis secondary to mild hepatic encephalopathy: a multilayer network analysis

Chao Ju1#, Longtao Yang1#, Zhongshang Dai2,3,4#, Yisong Wang1, Chang Li1, Wei Zhao1, Yongfang Jiang2,3,4, Jun Liu1,5,6

1Department of Radiology, The Second Xiangya Hospital of Central South University, Changsha, China; 2Department of Infectious Diseases, The Second Xiangya Hospital of Central South University, Changsha, China; 3Furong Laboratory, Changsha, China; 4Clinical Research Center for Viral Hepatitis in Hunan Province, Changsha, China; 5Department of Radiology, Quality Control Center in Hunan Province, Changsha, China; 6Clinical Research Center for Medical Imaging in Hunan Province, Changsha, China

Contributions: (I) Conception and design: Y Jiang, J Liu; (II) Administrative support: Y Jiang, J Liu; (III) Provision of study materials or patients: Z Dai; (IV) Collection and assembly of data: C Ju, L Yang; (V) Data analysis and interpretation: C Ju, L Yang; (VI) Manuscript writing: All authors; (VII) Final approval of manuscript: All authors.

#These authors contributed equally to this work as co-first authors.

Correspondence to: Jun Liu, MD, PhD. Department of Radiology, The Second Xiangya Hospital of Central South University, No. 139 Middle Renmin Road, Changsha 410011, China; Department of Radiology, Quality Control Center in Hunan Province, Changsha, China; Clinical Research Center for Medical Imaging in Hunan Province, Changsha, China. Email: junliu123@csu.edu.cn; Yongfang Jiang, MD, PhD. Department of Infectious Diseases, The Second Xiangya Hospital of Central South University, No. 139 Middle Renmin Road, Changsha 410011, China; Furong Laboratory, Changsha 410078, China; Clinical Research Center for Viral Hepatitis in Hunan Province, Changsha, China. Email: jiangyongfang@csu.edu.cn.

Background: Static functional networks of the brain are disrupted in minimal hepatic encephalopathy (MHE), but their dynamic alterations are unknown. This observational study utilized multilayer network analysis to investigate dynamic network characteristics in hepatitis B cirrhosis (HBC) with or without MHE and assess their association with neurocognitive function (registry: https://ctms.xyeyy.com/iit/project/index; trial registration number: LYF20240134; date: 2024-07-31).

Methods: A total of 33 HBC patients [15 non-MHE (NMHE; HBC patients without MHE) and 18 MHE individuals], as well as 36 matched healthy controls (HCs), underwent neurocognitive assessments, resting-state functional magnetic resonance imaging (rs-fMRI), and clinical examinations. Dynamic network variations were quantified using network switching rates, and their relationships with clinical and neurocognitive parameters were evaluated.

Results: Both HBC patients with and without MHE status exhibited a range of altered network switching rates compared to HCs. Specifically, differences were observed in subnetworks including somatomotor network (SMN), dorsal attention network (DAN), ventral attention network (VAN), frontoparietal network (FPN), and subcortical network (SUB), as well as in nodal regions such as the right precentral gyrus (rPrG), left fusiform gyrus (lFuG), left inferior parietal lobule (lIPL), and right hippocampus (rHipp). Furthermore, altered global level and lFuG switching rates positively correlated with Psychometric Hepatic Encephalopathy Score (PHES) (r=0.341, 0.339; P=0.004, 0.004, respectively, Bonferroni corrected).

Conclusions: This study firstly revealed that HBC patients exhibited imbalanced functional dynamics in subnetworks and nodes, suggesting a potential mechanism underlying cerebral dysfunction in MHE.

Keywords: Hepatitis B cirrhosis (HBC); minimal hepatic encephalopathy (MHE); multilayer network analysis; switching rate; functional dynamics

Submitted Nov 04, 2024. Accepted for publication Nov 11, 2025. Published online Jan 14, 2026.

doi: 10.21037/qims-24-2442

Introduction

According to reports from the World Health Organization, approximately 2 billion individuals worldwide have been infected with hepatitis B virus. In China, around 2.1% of hepatitis B patients develop cirrhosis annually, whereas 30–45% of cirrhosis patients subsequently experience secondary hepatic encephalopathy (HE) (1). HE, a reversible neurocognitive complication characterized by metabolic dysfunction, can arise from hepatic insufficiency, severe disorders, or portal venous shunts (2). It can be further classified into minimal HE (MHE) and overt HE (OHE) based on variations in clinical presentations (2). Although MHE individuals lack obvious clinical manifestations, they show a decrease in working, calculation, and driving ability, posing a potential risk to society.

In clinical practice, the most common diagnostic method for MHE is the Psychometric Hepatic Encephalopathy Score (PHES), which is subjective and susceptible to uncontrolled factors (e.g., age, education) (3-5). However, magnetic resonance imaging (MRI) offers a noninvasive alternative that can yield objective neuro-biomarkers through various imaging sequences [e.g., arterial spin labeling (ASL), diffusion-weighted imaging (DWI), magnetic resonance spectroscopy (MRS), and resting-state functional MRI (rs-fMRI)] (6). MRI has the potential to specifically identify MHE, offering novel avenues for quantitative assessment and early diagnosis (7). Although previous studies have suggested disrupted functional connectivity (FC) within and between brain networks as a key characteristic of MHE (8), it is important to note that the brain network comprises tightly interconnected modular structures that undergo dynamic changes, even during resting-state MRI scanning (9-12).

Multilayer network analysis is a unique graph-theoretic approach for studying the spatiotemporal organization of networks (13). In the context of functional brain networks, nodes represent specific brain regions, and between-node FC patterns reflect their functional interactions (14). In multilayer network models, nodes are connected in both temporal and spatial dimensions, where between-node dynamic associations portend network flexibility. This flexibility is manifested through the frequency of nodes switching between communities over time, known as network switching rate (15). The network switching rate is intimately linked to higher-order cognition such as working memory, reasoning, and planning (15). Nevertheless, alterations in global and local dynamic connections among hepatitis B cirrhosis (HBC) individuals with or without MHE and healthy controls (HCs) remain unclear. Therefore, investigating cerebral changes related to MHE in temporal connectivity reconfiguration will further enhance our understanding of neurophysiological mechanisms underlying cognitive decline.

In this study, we explored the topological dynamics, also called switching rate, in the functional connectome of treatment-naïve primary HBC patients with or without MHE using a multilayer network model applied to rs-fMRI data and aimed to reveal relationships between these dynamics and PHES. Our hypotheses were as follows: (I) At the global, subnetwork, and nodal levels, MHE would display significant brain changes in connectome dynamics [e.g., somatomotor network (SMN)] relative to non-MHE (NMHE; HBC patients without MHE) and HCs; and (II) these changes in network switching rate would be associated with PHES. The topological properties of brain dynamics might serve as objective phenotypes for identifying MHE-related alterations. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-24-2442/rc).

Methods

Participants

Between March 2023 and February 2024, 33 primary HBC patients were recruited from the Department of Infectious Diseases in The Second Xiangya Hospital of Central South University and were not accompanied by cirrhosis of other causes. Before they received diuretic, surgical, antiviral, hepatoprotective, and albumin supplementation treatment, rs-fMRI and cognitive examinations were conducted. Through verbal consultation about the results of physical examinations in the past year, 36 age-, gender-, and education-matched HCs from social recruitment were also included. They all signed written consent forms and took part in the MRI scan. The diagnostic criteria for primary HBC were mainly as follows: positivity of hepatitis B surface antigen; typical clinical signs (e.g., jaundice, ascites); imaging findings (e.g., nodular liver surface, splenomegaly, portal vein dilation); and exclusion of other major causes (e.g., significant alcohol use, autoimmune hepatitis). The infectious disease specialists determined the final diagnosis through investigation of various clinical manifestations. No previous diagnoses of structural brain disease, head trauma, seizures, behavioral or mental disorders, or MRI contraindications were reported in any of the participants. All of the participants were right-handed and had Han Chinese ancestry. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee at The Second Xiangya Hospital of Central South University, China (registry: https://ctms.xyeyy.com/iit/project/index; trial registration number: LYF20240134) and informed consent was provided by all the patients.

PHES and Mini-Mental State Examination (MMSE) scales were tested in both HBC patients and HCs for evaluating cognitive function, where PHES represented a gold standard for diagnosing MHE (16). Through the construction of a multivariable linear regression model for setting a reference value, we subdivided HBC patients into MHE (PHES ≤−4) and NMHE (PHES >−4) by taking PHES of −4 as the critical value. Moreover, the indexes reflecting liver function, such as total bilirubin, international normalized ratio, creatinine, Model for End-Stage Liver Disease score, serum sodium, serum ammonia, alanine transaminase, aspartate transaminase, and classification of Child-Pugh stage, were tested in HBC patients.

Image acquisition

All imaging data were acquired on a 3-T Siemens Skyra MRI scanner (Magnetom Skyra, Siemens, Erlangen, Germany) with a 32-channel head coil. During the scans, participants remained still, kept their eyes closed, and were instructed not to think about anything in particular. In addition, participants lay supine with foam padding between their head and the coil to minimize head movements. The scanning parameters of the rs-fMRI session are as follows: 36 axial slices, repetition time (TR) =2,000 ms, echo time (TE) =30 ms, flip angle (FA) =90°, voxel size =3.8×3.8×4 mm3, slice thickness =4 mm, field of view (FOV) =240×240 mm2.

Image processing

Pre-processing of fMRI image data was conducted by Gretna software (https://www.nitrc.org/projects/gretna/). The main steps were as follows: (I) deletion of the first 10 time points for steady-state magnetization; (II) slice timing correction; (III) imaging realignment for head motion correction and participants with head movement >2 mm or rotation >2° were excluded; (IV) image normalization using echo-planar imaging templates [voxel size (3×3×3 mm3)]; (V) linearly temporally detrended and temporal bandpass filtering (0.01–0.1 Hz); and (VI) regressing out nuisance signals (including white matter signal, cerebrospinal fluid signal, and Friston 24 motion parameters).

Multilayer brain network switching rates

The overall flow of analysis is shown in Figure 1.

Figure 1 Analysis of strategy. Individual images underwent preprocessing, and mean values from each region in the Brainnetome 246 Atlas were extracted to construct a dynamic functional matrix for each subject. An iterative ordinal Louvain sliding windows algorithm tracked dynamic network modulation over time. Subsequently, the network switching rates were computed and compared across participant groups at global, subnetwork, and nodal levels.

Multilayer brain network construction

As a powerful tool, a multilayer network model can quantify multi-dimensional imaging data from different perspectives (17), and can be taken as “network of networks”, including multitasking networks, different modality networks, and frequency-/time-varying networks (18-21). Here, a time-varying multilayer network was used, where each layer in the multi-layer network corresponded to a time-window. A total of 246 nodes were defined by the Brainnetome 246 Atlas, consisting of 36 subcortical and 210 cortical nodes (22). The averaged signal of each node was extracted. Thus, a sliding windows method was used to compute dynamic FC by analyzing each case’s preprocessed rs-fMRI data. Next, a series of W signal windows were obtained by exploiting Hamming windows (TR =2 s, step length =1, window length =50). Ultimately, to create an intralayer network, we calculated Pearson correlations between each pair of region signals in each window. A dynamic network matrix (N×N×W=246×246×121) for each individual was yielded, where N=246 was the number of atlas regions, and W=121 was the number of time windows.

Modularity and network switching rate

Through a multilayer modularity algorithm implemented in an open-source Matlab-based code package (https://github.com/GenLouvain/GenLouvain) (9), optimization of the multilayer modular quality factor Q, ranging from 0 to 1, was used to compartmentalize blocks in the multilayer network with the default settings topological resolution parameter (γ) = temporal coupling parameter (ω) =1 (19,23). In each rs-fMRI scan, a module assignment matrix (N×W=246×121) indicated temporal variation in module assignments for all 246 nodes.

Based on the above multilayer module assignment matrices, the calculation formula of switching rate of a node f(i), ranging from 0 to 1, was as follows: f(i)=ni/(W−1). ni refers to the number of altered times in the module assignment between consecutive layers; (W−1) is the number of maximum altered times. A greater f(i) value demonstrates a higher node transition frequency between various functional modules, but lower temporal stability. Network Community Toolbox (http://commdetect.weebly.com) was applied. In the Yeo atlas, the switching rate for brain global level is calculated by averaging the 246 nodes. Furthermore, cortical network is defined as 210 cortical nodes related to 7 neural networks [including frontoparietal network (FPN), default mode network (DMN), limbic network (LIB), dorsal attention network (DAN), ventral attention network (VAN), SMN, and visual network (VIS)], and 36 subcortical nodes related to subcortical network (SUB) (24-27). Through averaging their constituent nodes, the switching rate of each subnetwork was acquired. The detailed formula of the switching rate was as follows (28):

Qmultilayer(γ,ω)=12μ∑ijsγ[(Aijs−γskiskjs2ms)δ(S,γ)+δ(i,j)ωjsγ]

Statistical analysis

The differences in demographics among HCs, NMHE, and MHE were analyzed with the software SPSS 26.0 (IBM Corp., Armonk, NY, USA). One-way analysis of variance (ANOVA) and two-sample t-test were used for assessing differences in normally distributed continuous data. Mann-Whitney U test and Kruskal-Wallis test were used for assessing differences in non-normally distributed continuous data. Chi-squared test was employed to test whether there were gender and Child-Pugh stage differences among groups. All tests were two-tailed, and significant levels were set to P<0.05.

As for neural function, we used one-way ANOVA to compare switching rate differences among HCs, NMHE, and MHE at brain global level, eight subnetworks, and 246 nodes. Bonferroni correction was conducted for post-hoc tests. Results were visualized by BrainNet Viewer (https://www.nitrc.org/projects/bnv/). Then, partial correlation analysis was employed to reveal potential inter-relationships between significant switching rate indexes and PHES by taking age, gender, and education level as covariates.

Validation analysis

To test the veracity of the data, we investigated whether our findings were impacted by imaging processing parameters, in particular head motion, multilayer network model parameters (ω and γ), and the sliding window parameters (window size) (27,29). To eliminate effects of head motion, we repeated the switching rate analysis using mean framewise-displacement as a covariate. To investigate the effects of the network analysis strategy, we duplicated the analysis using ω=0.5, γ=0.9, and ω=1.0, γ=0.9 (28). Moreover, we further validated our results using window length =30 TR and step length =1 TR.

Results

Demographic, clinical, and neuropsychological characteristics

As exhibited in Table 1, the comparative analysis of demographic characteristics in age, gender, and education level revealed no statistically significant difference (all P>0.05) among HCs, NMHE, and MHE. It indicated that the MHE group was accompanied by greater serum ammonia concentration than the NMHE group. Both the MHE and NMHE groups had significantly lower PHES than HCs. Compared to the NMHE group, the PHES of the MHE group was also lower.

Table 1

Demographic and clinical characteristics of HCs, NMHE, and MHE

Characteristics HCs NMHE MHE P value
Age (years)a 52.42±9.76 57.47±7.89 51.06±9.71 0.125
Gender (male/female)b, n 30/6 10/5 17/1 0.140
Education level (years)c 9 [3] 9 [3] 9 [6] 0.262
Disease duration (years)d NA 3 [9.83] 3 [4] 0.927
Total bilirubin (μmol/L)d NA 40.4 [31.7] 24.9 [90.4] 0.664
INRd NA 1.32 [0.59] 1.31 [0.19] 0.625
Creatinine (μmol/L)d NA 68.2 [28.2] 87.0 [63.6] 0.060
MELD (score)e NA 19.38±9.96 13.79±6.55 0.062
Serum sodium (mmol/L)e NA 139.56±3.09 140.11±4.44 0.688
Serum ammonia (μmol/L)d NA 51.20 [18.4] 67.55 [26.1] 0.020§
ALT (U/L)d NA 35.30 [59.1] 43.95 [169.5] 0.539
AST (U/L)d NA 53.10 [57.3] 39.15 [50.6] 0.320
Child-Pugh stage (A/B/C)b, n NA 1/11/3 4/9/5 0.431
PHES (normal IQR, −15 to 5)c 0 [−3 to 3] −2 [−3 to 1] −7 [−10 to −4] <0.001†,‡,§

Normal distributed data are presented as mean ± SD; skewed data are presented as median [IQR]; unless otherwise stated. a, one-way ANOVA; b, Chi-squared test; c, Kruskal-Wallis test; d, Mann-Whitney U tests; e, two-sample t-test. The markers †, ‡, and § are used to indicate significant differences in demographic and clinical characteristics between HCs and NMHE, HCs and MHE, as well as NMHE and MHE, respectively. Educational level is defined as the education time starting from the first grade of primary school. Disease duration is defined as the time interval (in years) between the first definite diagnostic time informed by the patient and the date of study enrollment. ALT, alanine transaminase; ANOVA, analysis of variance; AST, aspartate transaminase; HBC, hepatitis B cirrhosis; HCs, healthy controls; INR, international normalized ratio; IQR, interquartile range; MELD, Model for End-Stage Liver Disease; MHE, minimal hepatic encephalopathy; NMHE, non-MHE (HBC patients without MHE); PHES, Psychometric Hepatic Encephalopathy Score; SD, standard deviation.

The metrics of network switching rate

Tables 2,3 display brain network and node findings related to switching rate, respectively. The following results corrected by the Bonferroni method were observed: (I) post-hoc tests exhibited that the NMHE group had larger switching rates than HCs, involving SMN, DAN, VAN, right precentral gyrus (rPrG), and right hippocampus (rHipp). However, compared to HCs, there was reduced switching rate of SUB in NMHE. (II) Compared to HCs, the MHE group had higher switching rates in DAN and FPN as well as lower switching rates in left fusiform gyrus (lFuG) and rHipp. (III) Compared to MHE, there were higher switching rates of VAN, global level, and left inferior parietal lobule (lIPL) in NMHE. Figure 2 provides a visual representation of these changes in switching rate as identified in the between-group analyses.

Table 2

Network switching rate at global and subnetwork level among HCs, NMHE, and MHE

Network switching rate HCs NMHE MHE F η2 partial P value
Global 0.315±0.099 0.262±0.079 0.234±0.075 5.479 0.142 0.006§
Subnetwork
   VIS 0.038±0.020 0.039±0.019 0.037±0.018 1.358 0.001 0.257
   SMN 0.035±0.020 0.039±0.020 0.037±0.016 8.541 0.007 <0.001†
   DAN 0.037±0.020 0.040±0.019 0.040±0.020 6.205 0.006 0.002†,‡
   VAN 0.034±0.019 0.040±0.019 0.037±0.018 4.919 0.006 0.007†,§
   LIB 0.043±0.020 0.040±0.021 0.042±0.017 2.253 0.003 0.105
   FPN 0.037±0.019 0.039±0.021 0.040±0.019 6.057 0.007 0.002‡
   DMN 0.036±0.020 0.037±0.021 0.036±0.019 1.155 0.001 0.315
   SUB 0.044±0.020 0.042±0.021 0.042±0.019 4.260 0.003 0.014†

Data are presented as mean ± SD, unless otherwise stated. Bonferroni was conducted for post-hoc tests of one-way ANOVA, and P<0.05 (corrected) was set as significance level. The markers †, ‡, and § are used to indicate significant differences in network switching rate between HCs and NMHE, HCs and MHE, as well as NMHE and MHE, respectively. ANOVA, analysis of variance; DAN, dorsal attention network; DMN, default mode network; FPN, frontoparietal network; HBC, hepatitis B cirrhosis; HCs, healthy controls; LIB, limbic network; MHE, minimal hepatic encephalopathy; NMHE, non-MHE (HBC patients without MHE); SD, standard deviation; SMN, somatomotor network; SUB, subcortical network; VAN, ventral attention network; VIS, visual network.

Table 3

Regions showing altered network switching rate at nodal level among HCs, NMHE, and MHE

Label Brain regions MNI (X, Y, Z) HCs NMHE MHE F η2 partial P value
62 rPrG 54, 4, 9 0.016±0.003 0.025±0.007 0.015±0.003 3.708 0.101 0.03†
103 lFuG −33, −16, −32 0.017±0.003 0.022±0.006 0.014±0.003 3.168 0.088 0.049‡
141 lIPL −56, −49, 38 0.021±0.003 0.019±0.005 0.013±0.003 3.527 0.097 0.035§
216 rHipp 22, −12, −20 0.021±0.004 0.022±0.006 0.016±0.004 3.416 0.094 0.039†,‡

Data are presented as mean ± SD, unless otherwise stated. Bonferroni was conducted for post-hoc tests of one-way ANOVA, and P<0.05 (corrected) was set as significance level. The markers †, ‡, and § are used to indicate significant differences in node switching rate between HCs and NMHE, HCs and MHE, as well as NMHE and MHE, respectively. ANOVA, analysis of variance; HBC, hepatitis B cirrhosis; HCs, healthy controls; lFuG, left fusiform gyrus; lIPL, left inferior parietal lobule; MHE, minimal hepatic encephalopathy; MNI, Montreal Neurological Institute; NMHE, non-MHE (HBC patients without MHE); rHipp, right hippocampus; rPrG, right precentral gyrus.

Figure 2 Between-group differences in network switching rates of subnetwork (A) and nodal (B) level. Results (P<0.05 corrected by Bonferroni) were visualized using the BrainNet Viewer (http://www.nitrc.org/projects/bnv). *, HCs vs. NMHE<0.05 (Bonferroni); †, HCs vs. MHE<0.05 (Bonferroni); #, NMHE vs. MHE <0.05 (Bonferroni). DAN, dorsal attention network; DMN, default mode network; FPN, frontoparietal network; HBC, hepatitis B cirrhosis; HCs, healthy controls; lFuG, left fusiform gyrus; LIB, limbic network; lIPL, left inferior parietal lobule; MHE, minimal hepatic encephalopathy; NMHE, non-MHE (HBC patients without MHE); SMN, somatomotor network; SUB, subcortical network; VAN, ventral attention network; VIS, visual network; rHipp, right hippocampus; rPrG, right precentral gyrus.

The main results were largely the same as our main findings in validation analysis (Table S1). Except for a significant difference in intramodular connectivity for module VAN between HC and MHE (P=0.025, Bonferroni corrected), there was no significant difference (P>0.05, Bonferroni corrected) in intra-/inter-modular connectivity among HCs, NMHE, and MHE (Table S2). Community assignments of significant nodes at different network layers for each participant are displayed in Figure S1. Node labels are shown in Table S3.

Correlation analysis

To eliminate covariates’ potential influence on switching rate metrics, we conducted partial correlation analyses to explore the relationship between significant switching rate metrics and clinical or neuropsychological characteristics in all participants. As illustrated in Figure 3, there were positive correlations between PHES and global level (r=0.341, P=0.004, Bonferroni corrected) as well as lFuG (r=0.339, P=0.004, Bonferroni corrected).

Figure 3 Relations of network switching rates to PHES. Positive relationships are illustrated between switching rates at global level (left panel) and lFuG (FuG_L_3_1, right panel) and individual PHES. FuG, fusiform gyrus; L, left; lFuG, left fusiform gyrus; PHES, Psychometric Hepatic Encephalopathy Score.

Discussion

Through construction of a multilayer network model based on the sliding window method of dynamic FC, the research findings support the alteration of brain modular dynamics in individuals with MHE. We observed notable differences in network switching rates among HCs and HBC individuals with or without MHE at the global, subnetwork, and nodal levels. Alterations occurred primarily in the global level, SMN, DAN, VAN, FPN, and SUB, as well as precentral, fusiform, parietal, and hippocampal regions, where switching rates in global level and lFuG showed positive correlations with PHES. These findings point to a novel perspective, namely that patients with MHE have abnormalities in their brain dynamic network configuration, which may be a crucial mechanism of neurocognitive impairment.

In a static state, conventional analysis of brain functional network pays more attention to patterns of within/between-network connectivity (30). However, a static functional network cannot fully characterize the whole cerebral spontaneous activity that dynamically operates over short time scales, and it is confined to reflecting temporally averaged features between regions. With distinct and rapid transitional shifts from one steady state to another over time, relevant cerebral areas show transient stability within a significantly modular organization during dynamics (31). Consequently, in a more nuanced approach, this study quantified dynamic configuration of the brain in HBC participants with or without MHE by introducing switching rate index, which can reflect dynamic FC changes in brain global, subnetwork, and nodal levels.

Lower switching rates in subnetworks or nodes might suggest discrete functional dysconnectivity (28). For instance, reduction in switching rate of VAN, auditory network, and DMN was found in obstructive sleep apnea patients and had relationships with neurocognitive dysfunction (32). In this context, respectively compared with NMHE and HCs, this study also uncovered lower switching rates in global level, VAN, and lIPL (IPL_L_6_4 belongs to DMN), as well as lFuG (FuG_L_3_1 belongs to LIB) and rHipp (Hipp_R_2_1 belongs to SUB) of HBC patients with MHE. Besides, NMHE had a lower switching rate in SUB than HCs. Lower switching rate indicated disrupted or decompensated functional dynamics in certain brain networks or nodes in HBC patients. However, compared to HCs, a higher switching rate was shown in SMN, DAN, VAN, rHipp (Hipp_R_2_1 belongs to SUB), and rPrG (PrG_R_6_5 belongs to VAN) of HBC individuals without MHE, as well as in DAN and FPN of HBC individuals with MHE, which might be attributed to existing compensatory mechanisms of functional dynamic configuration. Additionally, switching rate indexes at the global level and lFuG, which were positively correlated with PHES, might become potential aberrant phenotypes for exploring MHE.

At the subnetwork level: SMN has been correlated with preparation, planning, and execution of voluntary movements, where supplementary motor area (SMA) was identified as a core region of SMN (33). Compared with non-HE cirrhotic patients, a decrease in the values of regional homogeneity (ReHo) within SMA, cuneus, and posterior cingulate cortex was related to cognitive deficits in MHE condition (34). The VAN, responsible for between-task attention switching and sudden signal response is a typically unilateral and right-hemisphere brain network, making it critical for daily functions (35). Meanwhile, DAN can be thought of as an “aperture” of the camera for focusing and directing attention to a salient stimulus (35). Flexible interaction between VAN and DAN collaboratively facilitates the dynamic control of attention (35). Across cirrhotic patients with or without MHE and HCs, significant resting-state FC differences were found in VAN and DAN, the imaging features of which were helpful in MHE detection (36). Moreover, FPN participates in flexibly regulating cognitive control and instantiating (37). It has been indicated that rifaximin-treated cirrhotic patients with MHE exhibited restoration from aberrant hyperconnectivity of FPN and improved communication between FPN and DAN (38). SUB connecting the thalamus, basal ganglia, and cortex takes part in large-scale functional integration (39). Compared to NMHE, FC in basal ganglia network was decreased in cirrhotic patients with MHE and was positively correlated with PHES (40).

At the nodal level: PrG contains the primary motor cortex and allows voluntary control of skeletal muscles (41). Compared to HCs, reduction in negative FC among DMN, bilateral PrG, and language network was discovered in cirrhotic patients with MHE (42). FuG serves as a critical brain structure for high-level vision functions such as reading, object recognition, and face perception (43). Compared to HCs, increased ReHo in the right FuG was positively correlated with disease duration of cirrhotic patients with MHE (44). The lIPL (IPL_L_6_4) belongs to the DMN, which plays an important role in social cognitive processing, self-referential information, autobiographical processing, and situational memory (45). It has been indicated that there are significant differences in the resting-state FC of lIPL among MHE, NMHE, and HCs (36). As a temporal brain structure, the hippocampus is fundamentally involved in learning, emotions, and memory processing (46). Compared to HCs, a significantly lower bilateral hippocampal network FC strength was discovered in cirrhotic patients with MHE (47).

In this study, several limitations need to be acknowledged. Firstly, our sample size was relatively small, necessitating cautious interpretation of our conclusions. To bolster the robustness of our findings, we intend to enlarge our sample in future studies. Secondly, although sex × group interaction is not significant, the predominance of male cirrhotic patients in our cohort raises considerations about gender balance, which warrants further investigation. Thirdly, as a cross-sectional study, longitudinal follow-up is essential to gain deeper insights into the progression of the observed phenomena. Fourthly, we only recruited HBC patients in this study. Other factors (e.g., immunological disorder, alcohol) that induce cirrhosis need to be further considered. Fifthly, ancestry restriction was implemented to minimize population stratification bias, yet the external validity and generalizability of these results to other ancestral populations or to females are limited. Lastly, the presence of potential confounding factors such as anxiety and depression underscores the importance of incorporating mood assessment scales in future research endeavors to enhance the comprehensiveness of our evaluations.

Conclusions

Compared to HCs, HBC patients with or without MHE showed altered network switching rates, particularly evident in the SMN, DAN, VAN, FPN, and SUB as well as rPrG, lFuG, lIPL, and rHipp. These alterations in dynamic functional network reconfiguration may represent a critical mechanism underlying brain dysfunction in HBC patients. Although immediate clinical practice requires further validation, these metrics might hold potential to impact future patient management to some extent by enabling earlier recognition, objective monitoring of treatment response, and improved risk stratification for preventing cognitive decline and HE progression.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-24-2442/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-24-2442/dss

Funding: This work was supported by the Fundamental Research Funds for the Central Universities of Central South University (Nos. 1053320230244 and 2025ZZTS0873), the National Natural Science Foundation of China (Nos. U22A20303 and 82370611), the Clinical Research Center for Viral Hepatitis (No. 2023SK4009), the National Key Clinical Specialty Major Research Program (No. Z2023139), and the Key Project of Natural Science Foundation of Hunan Province (No. 2024JJ3041).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-24-2442/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee at The Second Xiangya Hospital of Central South University, China (registry: https://ctms.xyeyy.com/iit/project/index; trial registration number: LYF20240134) and informed consent was taken from all the patients.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

  1. Zacharias HD, Kamel F, Tan J, Kimer N, Gluud LL, Morgan MY. Rifaximin for prevention and treatment of hepatic encephalopathy in people with cirrhosis. Cochrane Database Syst Rev 2023;7:CD011585. [Crossref] [PubMed]
  2. Tapper EB, Parikh ND. Diagnosis and Management of Cirrhosis and Its Complications: A Review. JAMA 2023;329:1589-1602. [Crossref] [PubMed]
  3. Lv XH, Lu Q, Deng K, Yang JL, Yang L. Prevalence and Characteristics of Covert/Minimal Hepatic Encephalopathy in Patients With Liver Cirrhosis: A Systematic Review and Meta-Analysis. Am J Gastroenterol 2024;119:690-9. [Crossref] [PubMed]
  4. Duarte-Rojo A, Estradas J, Hernández-Ramos R, Ponce-de-León S, Córdoba J, Torre A. Validation of the psychometric hepatic encephalopathy score (PHES) for identifying patients with minimal hepatic encephalopathy. Dig Dis Sci 2011;56:3014-23. [Crossref] [PubMed]
  5. Li SW, Wang K, Yu YQ, Wang HB, Li YH, Xu JM. Psychometric hepatic encephalopathy score for diagnosis of minimal hepatic encephalopathy in China. World J Gastroenterol 2013;19:8745-51. [Crossref] [PubMed]
  6. Tamanini JVG, Ribeiro GAS, Kimura AT, Borella LF, Freddi TA, Reis F. Hyperintense lesions of the middle cerebellar peduncle and beyond: a pictorial essay. Radiol Bras 2024;57:e20240001. [Crossref] [PubMed]
  7. Wang Y, Yang L, Shang Y, Huang Y, Ju C, Zheng H, Zhao W, Liu J. Identifying Minimal Hepatic Encephalopathy: A New Perspective from Magnetic Resonance Imaging. J Magn Reson Imaging. 2025;61:11-24. [Crossref] [PubMed]
  8. Qi R, Zhang LJ, Xu Q, Liang X, Luo S, Zhang Z, Huang W, Zheng L, Lu GM. Abnormal functional connectivity within the default mode network in patients with HBV-related cirrhosis without hepatic encephalopathy revealed by resting-state functional MRI. Brain Res 2014;1576:73-80. [Crossref] [PubMed]
  9. Mucha PJ, Richardson T, Macon K, Porter MA, Onnela JP. Community structure in time-dependent, multiscale, and multiplex networks. Science 2010;328:876-8. [Crossref] [PubMed]
  10. Cheng Y, Zhang G, Zhang X, Li Y, Li J, Zhou J, Huang L, Xie S, Shen W. Identification of minimal hepatic encephalopathy based on dynamic functional connectivity. Brain Imaging Behav 2021;15:2637-45. [Crossref] [PubMed]
  11. Chen HJ, Lin HL, Chen QF, Liu PF. Altered dynamic functional connectivity in the default mode network in patients with cirrhosis and minimal hepatic encephalopathy. Neuroradiology 2017;59:905-14. [Crossref] [PubMed]
  12. Zhang G, Li Y, Zhang X, Huang L, Cheng Y, Shen W. Identifying Mild Hepatic Encephalopathy Based on Multi-Layer Modular Algorithm and Machine Learning. Front Neurosci 2020;14:627062. [Crossref] [PubMed]
  13. De Domenico M. Multilayer modeling and analysis of human brain networks. Gigascience 2017;6:1-8. [Crossref] [PubMed]
  14. Ji JL, Spronk M, Kulkarni K, Repovš G, Anticevic A, Cole MW. Mapping the human brain's cortical-subcortical functional network organization. Neuroimage 2019;185:35-57. [Crossref] [PubMed]
  15. Braun U, Schäfer A, Walter H, Erk S, Romanczuk-Seiferth N, Haddad L, Schweiger JI, Grimm O, Heinz A, Tost H, Meyer-Lindenberg A, Bassett DS. Dynamic reconfiguration of frontal brain networks during executive cognition in humans. Proc Natl Acad Sci U S A 2015;112:11678-83. [Crossref] [PubMed]
  16. Ferenci P, Lockwood A, Mullen K, Tarter R, Weissenborn K, Blei AT. Hepatic encephalopathy--definition, nomenclature, diagnosis, and quantification: final report of the working party at the 11th World Congresses of Gastroenterology, Vienna, 1998. Hepatology 2002;35:716-21. [Crossref] [PubMed]
  17. Boccaletti S, Bianconi G, Criado R, Del Genio CI, Gómez-Gardeñes J, Romance M, Sendiña-Nadal I, Wang Z, Zanin M. The structure and dynamics of multilayer networks. Phys Rep 2014;544:1-122. [Crossref] [PubMed]
  18. Tewarie P, Hillebrand A, van Dijk BW, Stam CJ, O'Neill GC, Van Mieghem P, Meier JM, Woolrich MW, Morris PG, Brookes MJ. Integrating cross-frequency and within band functional networks in resting-state MEG: A multi-layer network approach. Neuroimage 2016;142:324-36. [Crossref] [PubMed]
  19. Pedersen M, Zalesky A, Omidvarnia A, Jackson GD. Multilayer network switching rate predicts brain performance. Proc Natl Acad Sci U S A 2018;115:13376-81. [Crossref] [PubMed]
  20. Wu K, Wang C, Liu J. Evolutionary Multitasking Multilayer Network Reconstruction. 2022 IEEE Trans Cybern 2022;52:12854-12868. [Crossref] [PubMed]
  21. Battiston F, Nicosia V, Chavez M, Latora V. Multilayer motif analysis of brain networks. Chaos 2017;27:047404. [Crossref] [PubMed]
  22. Fan L, Li H, Zhuo J, Zhang Y, Wang J, Chen L, Yang Z, Chu C, Xie S, Laird AR, Fox PT, Eickhoff SB, Yu C, Jiang T. The Human Brainnetome Atlas: A New Brain Atlas Based on Connectional Architecture. Cereb Cortex 2016;26:3508-26. [Crossref] [PubMed]
  23. Bassett DS, Porter MA, Wymbs NF, Grafton ST, Carlson JM, Mucha PJ. Robust detection of dynamic community structure in networks. Chaos 2013;23:013142. [Crossref] [PubMed]
  24. Baltruschat S, Cándido A, Megías A, Maldonado A, Catena A. Risk proneness modulates the impact of impulsivity on brain functional connectivity. Hum Brain Mapp 2020;41:943-51. [Crossref] [PubMed]
  25. Jiang R, Zuo N, Ford JM, Qi S, Zhi D, Zhuo C, Xu Y, Fu Z, Bustillo J, Turner JA, Calhoun VD, Sui J. Task-induced brain connectivity promotes the detection of individual differences in brain-behavior relationships. Neuroimage 2020;207:116370. [Crossref] [PubMed]
  26. Wang Q, Hu K, Wang M, Zhao Y, Liu Y, Fan L, Liu B. Predicting brain age during typical and atypical development based on structural and functional neuroimaging. Hum Brain Mapp 2021;42:5943-55. [Crossref] [PubMed]
  27. Zhu Y, Li X, Qiao Y, Shang R, Shi G, Shang Y, Guo H. Widespread plasticity of cognition-related brain networks in single-sided deafness revealed by randomized window-based dynamic functional connectivity. Med Image Anal 2021;73:102163. [Crossref] [PubMed]
  28. Suo X, Zuo C, Lan H, Li W, Li L, Kemp GJ, Wang S, Gong Q. Multilayer Network Analysis of Dynamic Network Reconfiguration in Adults With Posttraumatic Stress Disorder. Biol Psychiatry Cogn Neurosci Neuroimaging 2023;8:452-61. [Crossref] [PubMed]
  29. He L, Zhuang K, Li Y, Sun J, Meng J, Zhu W, Mao Y, Chen Q, Chen X, Qiu J. Brain flexibility associated with need for cognition contributes to creative achievement. Psychophysiology 2019;56:e13464. [Crossref] [PubMed]
  30. Aiello G, Ledergerber D, Dubcek T, Stieglitz L, Baumann C, Polanìa R, Imbach L. Functional network dynamics between the anterior thalamus and the cortex in deep brain stimulation for epilepsy. Brain 2023;146:4717-35. [Crossref] [PubMed]
  31. Leng X, Qin C, Lin H, Li M, Zhao K, Wang H, Duan F, An J, Wu D, Liu Q, Qiu S. Altered Topological Properties of Static/Dynamic Functional Networks and Cognitive Function After Radiotherapy for Nasopharyngeal Carcinoma Using Resting-State fMRI. Front Neurosci 2021;15:690743. [Crossref] [PubMed]
  32. Huang Y, Shen C, Zhao W, Zhang HT, Li C, Ju C, Ouyang R, Liu J. Multilayer network analysis of dynamic network reconfiguration in patients with moderate-to-severe obstructive sleep apnea and its association with neurocognitive function. Sleep Med 2023;112:333-41. [Crossref] [PubMed]
  33. de Pasquale F, Sabatini U, Della Penna S, Sestieri C, Caravasso CF, Formisano R, Péran P. The connectivity of functional cores reveals different degrees of segregation and integration in the brain at rest. Neuroimage 2013;69:51-61. [Crossref] [PubMed]
  34. Chen HJ, Zhu XQ, Yang M, Liu B, Zhang Y, Wang Y, Teng GJ. Changes in the regional homogeneity of resting-state brain activity in minimal hepatic encephalopathy. Neurosci Lett 2012;507:5-9. [Crossref] [PubMed]
  35. Vossel S, Geng JJ, Fink GR. Dorsal and ventral attention systems: distinct neural circuits but collaborative roles. Neuroscientist 2014;20:150-9. [Crossref] [PubMed]
  36. Chen HJ, Wang Y, Zhu XQ, Li PC, Teng GJ. Classification of cirrhotic patients with or without minimal hepatic encephalopathy and healthy subjects using resting-state attention-related network analysis. PLoS One 2014;9:e89684. [Crossref] [PubMed]
  37. Marek S, Dosenbach NUF. The frontoparietal network: function, electrophysiology, and importance of individual precision mapping. Dialogues Clin Neurosci 2018;20:133-40. [Crossref] [PubMed]
  38. Casanova-Ferrer F, Gallego JJ, Fiorillo A, Urios A, Ríos MP, León JL, Ballester MP, Escudero-García D, Kosenko E, Belloch V, Montoliu C. Improved cognition after rifaximin treatment is associated with changes in intra- and inter-brain network functional connectivity. J Transl Med 2024;22:49. [Crossref] [PubMed]
  39. Bell PT, Shine JM. Subcortical contributions to large-scale network communication. Neurosci Biobehav Rev 2016;71:313-22. [Crossref] [PubMed]
  40. García-García R, Cruz-Gómez ÁJ, Mangas-Losada A, Urios A, Forn C, Escudero-García D, Kosenko E, Ordoño JF, Tosca J, Giner-Durán R, Serra MA, Avila C, Belloch V, Felipo V, Montoliu C. Reduced resting state connectivity and gray matter volume correlate with cognitive impairment in minimal hepatic encephalopathy. PLoS One 2017;12:e0186463. [Crossref] [PubMed]
  41. Giampiccolo D, Duncan JS. High plasticity within the precentral gyrus and its relevance for resective neurosurgery. Brain 2023;146:e98-9. [Crossref] [PubMed]
  42. Zhan C, Chen HJ, Gao YQ, Zou TX. Functional Network-Based Statistics Reveal Abnormal Resting-State Functional Connectivity in Minimal Hepatic Encephalopathy. Front Neurol 2019;10:33. [Crossref] [PubMed]
  43. Weiner KS, Zilles K. The anatomical and functional specialization of the fusiform gyrus. Neuropsychologia 2016;83:48-62. [Crossref] [PubMed]
  44. Sun Q, Fan W, Ye J, Han P. Abnormal Regional Homogeneity and Functional Connectivity of Baseline Brain Activity in Hepatitis B Virus-Related Cirrhosis With and Without Minimal Hepatic Encephalopathy. Front Hum Neurosci 2018;12:245. [Crossref] [PubMed]
  45. McCormick C, Protzner AB, Barnett AJ, Cohn M, Valiante TA, McAndrews MP. Linking DMN connectivity to episodic memory capacity: what can we learn from patients with medial temporal lobe damage? Neuroimage Clin 2014;5:188-96. [Crossref] [PubMed]
  46. Knierim JJ. The hippocampus. Curr Biol 2015;25:R1116-21. [Crossref] [PubMed]
  47. Lin W, Chen X, Gao YQ, Yang ZT, Yang W, Chen HJ. Hippocampal atrophy and functional connectivity disruption in cirrhotic patients with minimal hepatic encephalopathy. Metab Brain Dis 2019;34:1519-29. [Crossref] [PubMed]
Cite this article as: Ju C, Yang L, Dai Z, Wang Y, Li C, Zhao W, Jiang Y, Liu J. Dynamic network reconfiguration in hepatitis B cirrhosis secondary to mild hepatic encephalopathy: a multilayer network analysis. Quant Imaging Med Surg 2026;16(2):163. doi: 10.21037/qims-24-2442

Article Options

Download Citation

Share