![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
Exp Neurobiol 2024; 33(6): 282-294
Published online December 31, 2024
https://doi.org/10.5607/en24022
© The Korean Society for Brain and Neural Sciences
Dilara Derya1 and Christian Wallraven1,2*
1Department of Brain and Cognitive Engineering, Korea University, Seoul 02841,
2Department of Artificial Intelligence, Korea University, Seoul 02841, Korea
Correspondence to: *To whom correspondence should be addressed.
TEL: 82-2-5034-5925, FAX: 82-2-3290-5925
e-mail: wallraven@korea.ac.kr
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Research on brain aging using resting-state functional magnetic resonance imaging (rs-fMRI) has typically focused on comparing “older” adults to younger adults. Importantly, these studies have often neglected the middle age group, which is also significantly impacted by brain aging, including by early changes in motor, memory, and cognitive functions. This study aims to address this limitation by examining the resting state networks in middle-aged adults via an exploratory whole-brain ROI-to-ROI analysis. Using rs-fMRI, we compared middle-aged adults (n=30) with younger adults (n=70) via an ROI-to-ROI correlation analysis, showing lower connectivity between the cerebellar (posterior) network and the salience network (left rostral prefrontal cortex), as well as between the salience network and the visual network (occipital regions) in the middle-aged group. This reduced connectivity suggests that aging affects how these brain regions synchronize and process information, potentially impairing the integration of cognitive, sensory, and emotional inputs. Additional within-group analyses showed that middle-aged adults exhibited weakened connections between networks but increased connections within the dorsal attention, fronto-parietal, visual, and default mode networks. In contrast, younger adults demonstrated enhanced connections between networks. These results underscore the role of the cerebellar, salience, and visual networks in brain aging and reveal distinct connectivity patterns associated with signs of early aging.
Keywords: Cognitive aging, Cerebellum, Magnetic resonance imaging
Brain aging is an inevitable process that occurs across the lifespan of all humans. This process, although individual dependent, serves as a crucial biomarker for identifying neuro-cognitive decline due to neurological conditions, as well as environmental, genetic, or hormonal factors that contribute to cognitive impairment. Investigating the changes in the brain due to aging can be a valuable tool for distinguishing between the mild effects of healthy aging and the early warning signs of neurodegenerative diseases.
Resting-state functional MRI (rs-fMRI) has emerged as a pivotal tool for investigating brain plasticity, including the effects of aging, by capturing spontaneous blood-oxygen-level-dependent (BOLD) signals and providing spatial and temporal information in a task-free, neutral state. These BOLD signals encode spatial and temporal information that can be utilized to examine functional connectivity (FC) patterns within so-called resting-state networks (RSNs), which include the default mode network (DMN), somato-motor networks (SMN), dorsal-attention network (DAN), salience network (SN), visual network (VN), fronto-parietal network (FPN), language network (LAN), and cerebellar network (CER).These networks encompass key regions in the brain, such as for the DMN, which comprises the posterior cingulate cortex, medial prefrontal cortex, hippocampus, and inferior parietal cortex (FPN), which includes structures such as the inferior parietal cortex, ventral visual cortex, supramarginal gyrus, superior lateral occipital cortex, insula, and supplementary motor area.
Resting-state functional connectivity (RSFC) within these networks has been extensively studied in aging research. In particular, the DMN has been a focal point of several studies describing decreased connectivity [1-7] focusing on the changes occurring with cognitive deficits due to the areas included in this network, such as the ventral medial prefrontal (vmPFC), anterior and posterior cingulate (PCC) cortices, precuneus, inferior parietal cortices, angular gyrus, and the hippocampus.
On a larger scale, it is well known that brain aging is accompanied by a gradual decline in cognitive and motor functioning. While the decline in motor tasks is at first moderate, it still may impact daily living activities, such as coordination and motion speed [8-12]. In the context of aging, rs-fMRI has revealed alterations in RSNs that correlate with cognitive and motor function decline. As individuals age, one typically observes decreased intra-network connectivity alongside increased inter-network connectivity, a phenomenon also known as decreased brain segregation [5, 13-20]. This shift is attributed to brain plasticity and the adaptive changes to reflect a compensatory mechanism [21, 22] via the scaffolding theory of aging and cognition [23, 24] wherein the brain maintains functional performance despite age-related changes [25]. The decreased intra-network connections are typically the associated effects of neurodegeneration, whereas the increased inter-network connectives are considered a compensatory mechanism of aging [26, 27].
Most importantly, research on aging in the context of rs-fMRI has often contrasted an “older adults” group with a younger one to investigate the effects of brain aging and cognitive health [28]—what counts as “old,” however, is often vague and dependent on the constraints of the research environment. The present study therefore specifically looked at a participant age range that is less well represented in the literature: the (late) middle-age range, which according to the Medical Subject Heading (MeSH; https://www.ncbi.nlm.nih.gov/mesh) criteria includes people in the age range of 45 to 64 years. This group represents a critical transitional period in the aging process that is often characterized by adaptive changes in brain plasticity and functional connectivity that can serve as early indicators of cognitive aging. This period is particularly important to investigate, as it may reflect the early onset of compensation mechanisms or age-related declines in brain networks. An important factor to account for within this group is the fact that these individuals are still often working, and their active engagement in cognitive tasks may help preserve cognitive skills in this age group [29-31]. Focusing on this demographic allows us to capture unique brain connectivity patterns that might otherwise be masked by grouping them with older adults. These insights can provide a more nuanced understanding of brain-aging processes and their effects on both motor and cognitive functions.
The present study therefore aimed to investigate healthy aging in the brain by examining RSFC by contrasting a middle-age group with a younger-adult control group across the different RS networks. Understanding the dynamics of RSFC in aging can significantly enhance our knowledge of brain plasticity and compensatory mechanisms, offering potential biomarkers for early detection of neurodegenerative conditions. Through this study, we have aimed to contribute to the broader field of aging research by elucidating the complex interplay between different brain networks and their functional connectivity patterns in the context of healthy aging.
Thirty-six healthy middle-aged adults (MA) and seventy healthy younger adults (YA) were recruited in the rs-fMRI experiment. The classification of middle-aged adults was based on the MeSH criteria, which define middle-aged adults as individuals between 45 and 64 years old. Six participants in the MA group were removed from the study due to not meeting the cut-off score with the cognitive tests (see below in Table 1, where we note that these as well as all other individuals were not clinically diagnosed). All subjects had normal or corrected-to-normal vision. None of the participants in either age group had been diagnosed with either mild cognitive impairment or other psychiatric or neurological disorders. The study protocol was approved by the Institutional Review Board of Korea University (KUIRB-2019-0005), and all participants provided informed, written consent prior to participation. All procedures and methods were performed in accordance with the Declaration of Helsinki 1963 institutional and national guidelines and regulations.
All participants were subjected to the Montreal Cognitive Assessment Scale Korean version (MoCA-K) to test their cognitive function. The Montreal Cognitive Assessment (MoCA) is a standardized screening tool designed to measure cognitive function in order to detect mild cognitive impairment (MCI) and early signs of Alzheimer's disease. It evaluates various cognitive domains, including memory, attention, language, visuospatial skills, executive functions, and orientation. The test consists of a series of tasks, such as recalling a list of words, identifying pictures of animals, drawing a clock, and following a three-step command, and scores range from 0 to 30. The test was completed in 10~12 minutes on average by all participants. We added one point to individuals whose education level was less than 12 years [32]. Six participants in the MA group were excluded from the rest of the analysis due to not exceeding a cutoff score of 23.
A t-test analysis was performed on the MoCA scores, and mean scores were above the 25-point threshold of healthy cognition (MA-mean=25.88; YA-mean=28.59, p<0.0001). A chi-square test of independence was performed to examine the relationship between sex and group relationship (YA vs. MA). The relationship between these variables was significant, χ² (1, N=100)=10.04, p=0.0015, indicating that the distribution of males and females was significantly different between the groups.
High-resolution structural and functional imaging was performed using a 3.0 Tesla Magnetom Trio, A Tim System eco Siemens MRI scanner (Siemens Medical Solutions, Erlangen, Germany) equipped with a 32-channel head coil.
Structural images were acquired for co-registration to the functional MRI data using a three-dimensional T1-weighted magnetization-prepared rapid gradient-echo planar imaging (MPRAGE) sequence with the following acquisition parameters: repetition time=1.9 ms; echo time=2.52 ms; flip angle (FA)=9°; field of view (FOV)=256×256 mm2 matrix size; slice thickness=1×1×1 mm in-plane resolution; 180 contiguous sagittal slices.
This step was followed by obtaining rs-fMRI T2*–weighted functional images via a 7 minutes and 6 seconds gradient-echo EPI sequence with BOLD contrast acquired by TR=2000 ms; TE=30 ms; FA=90°; slice thickness=3 mm, FOV=220 mm; 64×64 matrix size; 36 slices; voxel size=2.8×2.8×3.0 mm3 with no gap (210 volumes). Since keeping the eyes open has been suggested to help network delineation when compared to eyes-closed conditions [33], during the scan, subjects were instructed to keep their eyes open and look at a fixation cross without thinking anything.
Functional and structural image analysis were performed using the default fMRI preprocessing in CONN software (version 22.a) [34] and SPM12 (Wellcome Department of Imaging Neuroscience, UCL, London, United Kingdom) in MATLAB R2020b (The MathWorks Inc., Natick, MA, United Kingdom). The preprocessing methods applied included realignment and unwarping, where framewise displacement was calculated to identify motion-related outliers based on a 0.5 mm threshold for excessive movement. Any volumes that exceeded this threshold were marked as outliers and removed from the analysis using the Artifact Detection Tools (ART) included in the CONN toolbox. Slice timing correction was applied to adjust for differences in acquisition time between slices, followed by segmentation of the anatomical images into gray matter, white matter, and cerebrospinal fluid. Functional images were then co-registered with structural images and normalized to the Montreal Neurological Institute (MNI) 152 template with 2 mm isotropic voxels. Smoothing was performed using an 8 mm full-width at half maximum (FWHM) Gaussian kernel to improve the signal-to-noise ratio. Following these steps, the CONN default denoising pipeline, anatomical component-based noise correction procedure (aCompCor) [35] was applied to estimate the physiological BOLD signal noise from the white matter and cerebrospinal fluid, including band-pass filtering of 0.008 to 0.08 Hz [36, 37] to obtain temporally specific BOLD amplitude of low-frequency fluctuations.
An ROI-to-ROI analysis was conducted using the CONN toolbox to assess functional connectivity between predefined regions of interest (ROIs) across multiple RSNs. The networks used in the rs-fMRI analysis were the pre-defined networks by Yeo and colleagues [38] commonly used in the literature. The predefined networks ensure comprehensive coverage of the key functional networks involved in cognitive, motor, sensory, and emotional processing. These eight RSNs examined were comprised of the default mode network (DMN: medial prefrontal cortex [MPFC], bilateral parietal cortex [lPC], and posterior cingulate cortex [PCC]); the salience network (SN: anterior cingulate cortex [ACC], bilateral anterior insula cortex, bilateral rostral prefrontal cortex [RPFC], and supramarginal gyrus [SMG]); the dorsal-attention network (DAN: bilateral frontal eye field [FEF] and bilateral intraparietal sulcus [IPS]); the SMN (bilateral, lateral, and superior sensorimotor regions); the visual network (VN: primary, ventral, and bilateral dorsal visual cortex); the language network (LAN: inferior frontal gyrus [IFG] and posterior superior temporal gyrus [STG]); the fronto-parietal network (FPN: dorsolateral prefrontal cortex [dLPFC] and posterior parietal cortex [PPC]); and the cerebellar network (CER: anterior and posterior cerebellum).
In the first-level analysis, for each participant, BOLD time series were extracted from predefined ROIs, representing regions within RSNs. Functional connectivity between pairs of ROIs was assessed for each participant by calculating Pearson’s correlation coefficients based on the time series data from each ROI. To ensure the data met normality assumptions required for second-level analyses, these correlation values were transformed into z-scores using Fisher's transformation. This was followed by an ROI-to-ROI analysis using a general linear model (GLM) to estimate the connectivity between brain regions for each participant. The correlation coefficient analysis distinguished pairs of ROIs at the cluster level using T-statistics or F-statistics, with multiple comparison corrections applied by false discovery rate (FDR, p<0.05).
At the group level, second-level contrasts were included to compare connectivity patterns between the YA and MA groups. Age was the only factor used in this analysis to control for its effect on functional connectivity. The main effect of age was calculated with a [1(ma) -1(ya)] contrast. In the second-level analysis, both sex and MOCA scores were included as covariates to control for potential differences in functional connectivity related to sex. We note here that we did not include years of education as a covariate since the MA group exhibited a relatively uniform range of years of education (16±3.24) and since MOCA scores themselves already were adjusted for different education levels.
Multiple comparison corrections were applied using the false discovery rate (FDR) method, with statistical significance set at p<0.05 (FDR corrected) at the cluster level. Voxel-level p-values (p-unc) are presented uncorrected for transparency and exploratory purposes to illustrate the raw connectivity strengths between specific ROIs. The main findings are based on FDR-corrected cluster-level results. Statistical significance was determined using spatial pairwise clustering (SPC) to control for Type I errors, with statistical significance determined at the cluster level using FDR correction (p<0.05). The results are reported as cluster-level FDR-corrected SPC mass/intensity (cluster threshold: p<0.05 cluster-level p-FDR corrected; connection threshold: p<0.05 p-FDR corrected), with the sum of T-statistics across voxels within each cluster used to measure the mass of the clusters. The size reported by the SPC approach is defined as the number of pairwise clusters or the sum of the test statistic over all connections comprising the pairwise cluster. For exploratory purposes, results using uncorrected thresholds (p<0.05) are also reported. The main findings were based on FDR-corrected results, and the uncorrected results should not be viewed as conclusive due to the increased risk of Type I errors.
We first focused on the between-group ROI-to-ROI analysis contrasting the MA group with the YA group. Significant reductions in functional connectivity were observed in the MA group compared to the YA group. Specifically, connectivity between the left rostral prefrontal cortex of the salience network (SN) and the posterior cerebellar region (CER) was significantly reduced (t(95)=-4.45, p-FDR=0.012). Similarly, connectivity between the occipital region of the visual network (VN) and the left rostral prefrontal cortex of the SN was reduced (t(95)=-3.89, p-FDR=0.049) (Fig. 1, Table 2).
Exploratory analyses using an uncorrected threshold (p<0.05) revealed additional group differences in ROI-to-ROI connectivity. Reduced connectivity was observed between the left rostral prefrontal cortex (RPFC) of the salience network (SN) and the lateral region of the visual network (VN) as well as between the right anterior insula (AInsula) of the salience network and the posterior cerebellar network, between the medial prefrontal cortex (MPFC) of the default mode network (DMN) and the left lateral prefrontal cortex of the fronto-parietal network (FPN), and between the right anterior insula of the salience network and the right lateral prefrontal cortex of the fronto-parietal network. Further studies will be necessary to follow up on these effects given their exploratory nature.
Separate within-group analyses were conducted to further explore connectivity patterns in the MA and YA groups. In the MA group, increased ROI-to-ROI connectivity was observed within several regions. Connections within the salience network included the left supramarginal gyrus and the left posterior superior temporal gyrus, as well as the right supramarginal gyrus and the right posterior superior temporal gyrus. Additional connections included the dorsolateral prefrontal cortex of the fronto-parietal network and the left rostral prefrontal cortex of the salience network, the left posterior superior temporal gyrus of the language network (LN) and the bilateral occipital regions of the visual network (VN), and the medial visual network and the posterior cingulate cortex of the default mode network (DMN). Reduced connectivity within the cerebellar network, particularly between the anterior and posterior regions, was also observed. Interactions between the superior sensorimotor network (SMN) and the dorsal attention network (DAN), as well as connections within the visual network and between the anterior and posterior cerebellar regions, were also identified (Fig. 2, 3, Supplementary Table 1).
In the YA group, stronger ROI-to-ROI connectivity was observed across multiple regions compared to the MA group. The most robust connectivity was observed between the posterior cingulate cortex of the DMN and the medial region of the visual network, as well as between the dorsal attention network and the sensorimotor network. The connectivity within the visual network was particularly strong (Fig. 4, 5, Supplementary Table 2).
The functional connectivity analysis revealed significant age-related differences in connectivity networks, highlighting key connections. The ROI-to-ROI rs-fMRI results demonstrated potential changes in the (early) aging brain across selected resting-state networks (RSNs), including cerebellar networks.
It has been well documented that cognitive functions often decline with retirement and the cessation of routine cognitive activities [30, 31]. Our study’s focus on the late middle-age range with adults still in the workforce offers a unique perspective on how continued engagement in cognitive tasks might preserve cognitive health (we note that although our participants were employed at the time of the study, employment status was not used as a recruitment criterion). This demographic has often been overlooked in aging research, making our findings particularly significant for understanding cognitive and brain function in early aging.
Our analysis found that the most significant age-related effect was observed in the connectivity between the posterior cerebellar (CER) and salience (SN) networks, where a decrease in functional connectivity was noted. Additional significant effects included a similar decrease in functional connectivity between the visual network (VN) occipital region and the SN left rostral prefrontal cortex. These results highlight specific reductions in functional connectivity in the middle-aged group compared to the younger group, reflecting changes in cerebellar-salience and visual-salience network connectivity with age.
The reduced connectivity observed in the cerebellar-salience network and its implications for cognitive functions align with findings from other studies, which have suggested that disruptions in cortico-cerebellar networks may serve as biomarkers for preclinical stages of Alzheimer’s disease and mild cognitive impairment [39]. Previous research has also noted similar disruptions in cortico-cerebral networks in older adults [40-42]. The decline in connectivity in these networks could be linked to reductions in cerebellar volume observed with aging [43, 44].
This may be driven by several underlying mechanisms associated with aging. One hypothesis is that this decrease in connectivity reflects a decline in neural plasticity, a process that enables the brain to reorganize and adapt to environmental changes. As individuals age, plasticity in neural circuits may be diminished, leading to less flexible and efficient communication between regions [45]. Another possible mechanism is a reduction in neural efficiency—the brain’s ability to perform cognitive tasks using minimal energy—where aging could result in a breakdown of efficient communication pathways, particularly in high-demand networks like the salience network [46]. Additionally, the observed decline may indicate a shift from localized, segregated connectivity to integration across regions as part of the brain's effort to compensate for localized neural degeneration [21, 22]. This aligns with the scaffolding theory of aging, which posits that older adults recruit additional brain regions to compensate for neural declines [23, 24]. However, in early aging, as observed in our middle-aged cohort, these compensatory mechanisms may be less developed, leading to weakened connectivity. Finally, age-related reorganization of neural networks, particularly between the cerebellum and higher cognitive regions, may explain the observed reduction in functional connectivity, as brain regions that were previously more connected may reorganize in response to cognitive aging [47, 48]. To further support these hypotheses, future studies should incorporate longitudinal and task-based fMRI analyses to better understand the progression and specific demands influencing these network changes.
Notably, two studies by Onoda and colleagues [7, 49] reported decreased inter-network connectivity between the SN and VN with aging, correlating with cognitive decline related to memory and sensory processing [50].
In terms of ROI-to-ROI connectivity, the MA group showed weaker connections within the cerebellar network, with notable reductions in connections between the anterior and posterior cerebellum. Conversely, younger adults exhibited increased inter-network connectivity, particularly between the cerebellum and the fronto-parietal network (FPN) and the dorsal attention network (DAN). This shift suggests that the cerebellum’s role in maintaining cognitive and motor tasks may differ between age groups, reflecting potential age-related differences in the dynamic nature of brain regions [12, 51].
Previous studies have reported that aging is associated with broader changes in connectivity, including decreased segregation within networks and increased integration between networks [17, 52, 53], reflecting compensatory mechanisms to support cognitive function. This pattern contrasts with developmental brain maturation, which typically shows decreased inter-network connectivity and increased intra-network connectivity [52]. The compensatory overactivation of prefrontal and parietal regions aligns with the scaffolding theory of aging, which posits that older adults recruit additional neural resources to maintain cognitive function in the face of neural decline [47, 48, 54, 55]. This theory supports the notion that despite underlying neural deterioration, the brain compensates by enhancing communication between specific regions to support cognitive performance. The concept of brain maintenance also suggests that preserving efficient connectivity between critical regions can help slow cognitive decline [56]. In our study, the MA group exhibited weaker connections between regions across different networks compared to the younger group, with the exception of connectivity between the DMN posterior cingulate cortex and the VN medial region. This result, along with stronger connections within the DAN, FPN, VN, and DMN, suggests that early aging may rely on compensatory mechanisms that are less efficient. This pattern in the MA group may reflect compensatory overactivation, emphasizing the complexity of group differences in functional connectivity across regions.
Our findings contribute to the understanding of age-related brain reorganization. While previous research has predominantly focused on decreased DMN connectivity, our results show stronger within-network connections in the DMN for older adults and robust inter-regional connections involving DMN for younger adults. The similarity of specific connections involving the DAN and SN across age groups highlights the role of these regions in maintaining cognitive functions. The older age group’s maintained cognitive skills, likely due to continued employment, might explain the observed differences.
Overall, our study offers a novel perspective on healthy aging, emphasizing the reorganization of specific connectivity patterns and highlighting the importance of the pre-retirement age group. The differences in cerebellar connectivity between younger and middle-aged adults provide new insights into how the brain adapts to maintain cognitive and motor functions during early aging.
We acknowledge several limitations in our study.
First, the older age group was not balanced in terms of sex, which should be rectified in further studies via a more balanced representation of males and females in order to better understand how sex might influence age-related changes in brain connectivity. Additionally, our study focused on a relatively narrow age range within the MA population, primarily those approaching retirement. To gain a more comprehensive understanding of age-related changes in brain connectivity, future research should include a broader range of ages, particularly those above the age of 72. This would help in identifying whether the patterns observed in our study are consistent across a wider age spectrum or if they vary significantly in older age.
The mean MoCA scores for all participant groups, including both the MA and YA groups, were above the threshold of 25, which is typically considered indicative of cognitive health. Nonetheless, the younger adults generally achieved higher MoCA scores. To address this observed difference and further investigate the progression of cognitive health over time, a longitudinal study could be conducted, which would allow us to track changes in cognitive function over an extended period and provide deeper insights into how aging impacts cognitive performance and brain connectivity.
While MoCA scores and sex were included as covariates to account for their potential influence on functional connectivity, covariate adjustment may not fully eliminate residual confounding effects. Further research is needed to explore the complex interactions between these variables and functional connectivity patterns in aging.
Finally, our study employed an ROI-to-ROI analysis, which focuses on specific predefined regions and may not fully represent broader network-level connectivity patterns. An independent component analysis (ICA) or other data-driven approaches could provide complementary insights into large-scale network dynamics and functional segregation.
In conclusion, our study illustrated the impact of aging on brain networks through resting-state functional connectivity (RSFC) analysis. We observed that an early-aging participant group exhibited notable plasticity in the brain, evidenced by changes in connections. Additionally, we identified which networks remained relatively similar from the younger age group. Ultimately, our findings highlight key mechanisms underlying cognitive decline in healthy aging, providing insights into how the brain adapts to age-related changes while maintaining cognitive function.
This study was supported by the National Research Foundation of Korea under project BK21 FOUR and grants NRF-2022R1A2C2092118, NRF-2022R1H1A2092007, as well as by Institute of Information & Communications Technology Planning & Evaluation (IITP) grants funded by the Korea government (No. RS-2019-II190079, Department of Artificial Intelligence, Korea University; No. RS-2021-II212068, Artificial Intelligence Innovation Hub).
Participant demographics
YA | MA | p-value | |
---|---|---|---|
Size of age group | 70 | 30 | - |
Mean age (years, ±std) | 24.69±2.99 | 54.59±5.44 | <0.001 |
Age range (years) | 18~31 | 48~65 | - |
Sex | 32 F, 38 M | 30 F, 6 M | 0.0015 |
Education (years, ±std) | 18±1.02 | 16±3.24 | 0.01 |
MoCA-K (score, ±std) | 28.78±1.9 | 25.88±1.62 | <0.0001 |
Main age effect of MA and YA, SPC; connection threshold, p<0.05 FDR corrected at cluster level; cluster threshold, p<0.05 FDR corrected. Results include models adjusted for MoCA and sex as covariates
Correlations | Statistic | p-unc | p-FDR |
---|---|---|---|
Salience.RPFC (L) (-32,45,27) - Cerebellar.Posterior (0,-79,-32) | t(89)=-4.45 | 0.000 | 0.012 |
Salience.RPFC (L) (-32,45,27) - Visual.Occipital (0,-93,-4) | t(89)=-3.89 | 0.000 | 0.048 |
Salience.RPFC (L) (-32,45,27) - Visual.Lateral (R) (38,-72,13) | t(89)=-3.00 | 0.003 | 0.144 |
Salience.AInsula (R) (47,14,0) - Cerebellar.Posterior (0,-79,-32) | t(89)=-2.77 | 0.006 | 0.167 |
Salience.AInsula (R) (47,14,0) - Visual.Occipital (0,-93,-4) | t(89)=3.24 | 0.001 | 0.053 |
DefaultMode.MPFC (1,55,-3) - FrontoParietal.LPFC (L) (-43,33,28) | t(89)=2.16 | 0.033 | 0.399 |
Salience.AInsula (R) (47,14,0) - FrontoParietal.LPFC (R) (41,38,30) | t(89)=2.61 | 0.010 | 0.209 |
L, left; R, right; RPFC, rostral prefrontal cortex; AInsula, anterior insula; MPFC, medial prefrontal cortex; LPFC, left lateral prefrontal cortex; p-unc, uncorrected p-value; p-FDR, false discovery-rate adjusted p-value; t, t-value; bold values indicate significance.