Fronto-parietal network dynamics to understand deficits in attention performance in multiple sclerosis

Thomas Welton, Dewen Meng, Roshan das Nair, Cris S Constantinescu, Dorothee P Auer, Rob A Dineen
doi: https://doi.org/10.1101/2023.07.08.23292404

Abstract

Introduction Impaired attention performance is a significant burden to people with multiple sclerosis (MS). Brain connectivity fluctuates with transitions between cognitive states, so measurement of network dynamics during these conditions may help to understand MS-related attention impairment.

Methods In people with MS and healthy controls, attention was measured using the Attention Network Test. 3T MRI was used to measure structural connectivity and both static and dynamic functional connectivity in the attention-related fronto-parietal network (FPN) at rest and during an attentionally-demanding task. Groups were compared on connectivity of the FPN during rest and task performance. Relationships between network connectivity and attention performance were tested using linear regression.

Results The sample comprised 37 people with MS and 23 matched controls. At rest, people with MS had significantly lower structural connectivity (R2=0.13, p=0.004), lower static functional connectivity (R2=0.07, p=0.032) and higher dynamic functional connectivity (R2=0.08, p=0.026) of the FPN. Higher dynamic connectivity was significantly associated with poorer attention performance in people with MS (R2=0.20, p=0.008). During attention-task performance, static functional connectivity was greater in people with MS than controls (R2=0.10, p=0.008). The task-induced reduction in static connectivity (relative to rest) was directly related to attention performance (R2=0.23, p<0.001).

Conclusion Increased dynamic functional connectivity of the FPN at rest may be a useful indicator of deficits in sustained attention in people with MS. The transition from rest to active-attentive state is accompanied by an increase in dynamic connectivity, and decrease in static connectivity which may be helpful in understanding aetiology and treatment of attention impairment.

Introduction

Cognitive impairment occurs in 40-65% of people with multiple sclerosis (MS) and frequently involves a reduced capacity for sustained attention or concentration [1–3]. Attention impairments are reported at rates between 20% and 50%, depending on the task used [4] and greatly impact patients’ quality of life [5]. Specifically, sustained attention abnormalities are often associated with difficulties in behavioural, learning, emotional, and cognitive aspects especially during adolescence [6] and sustained attention tasks are useful for quantifying cognitive fatigue [7]. Evidence for the effectiveness of treatments for cognitive impairment is equivocal [8] which, may, in-part, be due to difficulty in measuring cognitive symptoms. Neuroimaging may offer a non-invasive, direct and quantitative way to measure cognition-related neural changes in MS.

Neuroimaging studies have shown relationships between imaging measures and attention impairment in people with MS; for example, grey matter atrophy [9], abnormal white matter diffusivity [10] and greater lesion volumes [11]. However, normal brain structure and function is now understood to depend on the interconnectivity of anatomical regions or networks [12]. Accordingly, attention performance may be based on the connectivity of a set of attention-related areas [13–16], which includes the superior and middle frontal gyri, posterior cingulate, superior parietal lobule, intraparietal sulci, superior temporal gyri and lateral occipital gyri, known as the fronto-parietal network (FPN). Studies show that functional connectivity of these networks is disrupted in MS [17]; and in other cognitive networks such as a memory-related network to the n-back task [18], in motor networks [19] and visual networks [20]. However, the specific nature of the FPN disruption and how it relates to attention performance is not fully understood.

Dynamic functional connectivity, i.e. the variation in strength of connectivity between brain regions over time, is a recently-described technique for analysis of fMRI data, and may be important for understanding cognition in MS [21]. Dynamic functional connectivity analyses include “sliding window” (the simplest approach, measuring the variance in connectivity at the scale of seconds to minutes), “time-frequency” (incorporating information from the frequency domain at shorter timescales, for example using the wavelet transform), “point process” (measuring single time-point activations) and “graph theoretic” (variation in holistic topological measures of network integration and segregation) analyses [22]. There are now many examples of the application of these techniques in neurological disease [23]; for example abnormal default-mod sub-network “dwell times” in Alzheimer’s [24] and Parkinson’s disease [25]. Studies of resting dynamic connectivity in MS have also shown differences in state-dwelling times in cognitively impaired patients [26], reduced hippocampal dynamic connectivity related to maintained memory performance [27], greater dynamic connectivity of specific connection pairs related to lower lesion loads [28] and reduced network dynamics in cognitively impaired patients [29], mostly using the simple sliding window dynamic connectivity approach. There are also EEG studies showing patient-control differences in network dynamics during attentional tasks [30–32]. The dynamic functional connectivity approach may be useful in investigating attention deficits because attentional focus inherently varies in its target and strength based on demands [33, 34] and therefore may require a tool that can detect these variations [35].

Across this body of research, major knowledge gaps exist. Namely, in understanding how dynamic functional connectivity differs between intrinsic and stimulated states. In MS, no study has contrasted brain dynamics during rest with that during task performance which, according to research in other patient populations showing functionally relevant variations in static and dynamic connectivity between resting and task states, may be important in understanding the link to cognition [36, 37]. It is implied that network specific activation during task performance leads to an increase in functional connectivity and decrease in dynamic connectivity as the network becomes driven by the task rather than interference across other activities. However, this link is complex and likely non-linear because a deficit phase may show different associations than a compensatory phase. One possibility is that network-specific activation during task performance leads to an increase in functional connectivity and decrease in dynamic connectivity as the network signal becomes driven more by the task than interference across other activities.

We propose that dynamic functional connectivity of the FPN may enable a better understanding of mechanisms behind attention performance and its impairment in people with MS by showing the direction and magnitude of changes in an attention-related network (FPN) between groups and between states. We will apply the sliding-window variance technique to quantify dynamic functional connectivity in an attention-related network, as the most standardised, widely-used and reproducible method from the literature. We hypothesised that functional connectivity of an attention-related network during task performance would (a) be abnormal in people with MS compared to controls, with people with MS having lower static but higher dynamic connectivity, (b) correlate with measures of attention performance and alter under sustained attentionally-demanding task performance compared to rest, with better task performance relating to greater static connectivity and reduced dynamic connectivity, and (c) be different between anatomical regions with a known role in attention and whole-brain connectivity measures.

Methods

Subjects

Participants were recruited from MS clinics at a large UK National Health Service hospital according to the following criteria: aged 18-65 years, clinically-definite relapsing-remitting MS (RRMS) or secondary-progressive MS (SPMS), no other neurological or psychiatric conditions and no relapses or changes in medication 30 days prior (sample size calculation in Supplemental Information). RRMS and SPMS were diagnosed according to the McDonald criteria [38]. We also recruited age-, gender- and education-matched healthy participants via recruitment posters, who also were aged 18-65 and had no neurological or psychiatric disease. All participants underwent a 2-hour visit for cognitive testing and MRI between September 2014 and January 2017 and gave written informed consent. The study was approved by the Nottingham Research Ethics Committee 2 (14/EM/0064).

Neurocognitive Testing

Participants underwent testing for attention performance using the Attention Network Test (ANT) [39]. The ANT is a computer based test which shows a sequence of visual stimuli which can be used to predict the proceeding action. ANT measures three subscales based on reaction times resulting from a warning signal (“alerting”), in response to a cue about where the target will appear (“orienting”) and with correct directional response of a central marker (“executive”) [40]. The ANT was used as the primary measure for attention in all statistical tests, while two other secondary measures were also tested.

Secondary cognitive measures that include a significant attention component were the Paced Auditory Serial Addition Test (PASAT-3) [41] and Symbol Digit Modalities Test (SDMT) [42]. We also measured known confounding factors: fatigue using the Modified Fatigue Impact Scale (MFIS) [43], depression using the Beck Depression Inventory (BDI-II) [44] and sleep quality using the Pittsburgh Sleep Quality Index (PSQI) [45]. For further characterisation of the sample and ease of comparison to other studies, we also recorded the MS Functional Composite (MSFC) [46] which includes Nine-Hole Peg Test (9HPT) and Timed 25-Foot Walk (T25FW). The cohort was naïve to all tests except those in the MSFC.

Magnetic Resonance Imaging

MRI was performed on a 3T GE Discovery 750 scanner (General Electric Healthcare; Milwaukee, WI) with 32-channel head coil. The protocol included T1-weighted fast-spoiled gradient echo (FSPGR; FA=8°, matrix 256×256×156, voxel size 1×1×1mm, TE=3.17ms, TI=900ms, TR=8200ms, FOV=256mm, NEX=1), T2-weighted fluid attenuated inversion recovery (FLAIR; FA=111°, TE=120ms, TR=8000ms, TI=2250ms, matrix 512×512×46, voxel size 0.46×0.46×3mm, FOV=235.5mm), diffusion MRI (dMRI; four b=0 volumes and 32 diffusion weighted volumes at b=1000, matrix 128×128×66, voxel size 2mm3) and functional MRI (fMRI; FA=80°, TE=36ms, TR=2200ms, matrix 64×64×37, voxel size 3.75mm3) at rest (9 minutes, 245 volumes) and during task performance (6 minutes, 165 volumes).

fMRI Task Paradigm

Subjects performed a 6-minute sustained attention task while in the scanner. The task, the modified SDMT (mSDMT), was designed to induce cognitive load by continuous sustained attention and information processing [47], and was modelled on the SDMT but adapted for use in the MRI scanner to allow responses using buttons (Figure 1a). We implemented the task in E-Prime (version 2.0; Psychology Software Tools; Pittsburgh, PA). Error rates were recorded and used to quantify performance.

Figure 1.
Figure 1. Design of fMRI task and reaction times

The task comprised 45 trials, each 6 seconds with 2-second intervals. During each trial, the subject decides whether the bottom symbol pair matches any of the presented randomised pairs and responds by pressing a button (“match” or “no-match”).

Parcellation and Fronto-Parietal Network Definition

T1-weighted images were segmented using FreeSurfer (version 5.3; http://surfer.nmr.harvard.edu/) to generate sets of 164 grey matter ROIs based on the Destrieux atlas [48]. We defined the FPN based on literature [13, 49] and the NeuroSynth meta-analytic topic map for attention control (neurosynth.org; topic 264, version 5, July 2018, [50]). We selected the relevant labels from the FreeSurfer cortical parcellation, including the superior and middle frontal gyri, posterior cingulate, superior parietal lobule, intraparietal sulcus, superior temporal gyrus and lateral occipital gyrus bilaterally (Figure 2). We also gathered whole-brain volumes from FreeSurfer.

Figure 2.
Figure 2. Regions of the fronto-parietal network in an example subject

Lesion Volumes

The volume of T2-hyperintense lesions was measured on the T2-FLAIR images using JIM (version 5.0; Xinapse Systems; Essex, UK) for fuzzy semi-automatic lesion segmentation. A single point was manually placed in each lesion and the JIM software then automatically expands a region until the lesion is filled in a 3D mask for further use in segmentation/registration. Lesion volumes were normalised to whole-brain volume.

Static Functional Connectivity

Quality of the fMRI data was ascertained using data quality best practices from the fBIRN initiative [51] including inspection of power spectra, implemented in MATLAB (version 2013a; TheMathWorks; Natick, MA). fMRI data were pre-processed in FSL MELODIC [52] using the following steps: (1) removal of the first 5 volumes, (2) motion correction and (3) spatial smoothing with a 5mm Gaussian kernel. Regions with a timeseries average greater than 3 standard deviations below the mean were ignored in the analysis. Outlying timepoints were detected using the FSL motion outliers tool [52] and missing data points replaced using cubic interpolation. High-pass filtering with a 1-second cut-off was applied. Spatial ICA was performed using MELODIC [53] to manually identify (following the steps provided by Griffanti, Douaud [54]) and filter-out (using the “fsl_regfilt” tool) components consisting predominantly of noise, with no limit on the number of components. To apply the anatomical regions to the fMRI data, we first registered the task fMRI images to the resting-state fMRI images. We then created a transform from the resting-state fMRI to the T1-weighted image using the boundary-based registration tool in FLIRT [55]. This transformation was applied in reverse (and, for the task fMRI, the task-to-rest transformation) to the anatomical regions to produce personalised grey-matter atlases in fMRI space. The transformed masks were eroded by one voxel to reduce partial volume effects and contributions from white matter or CSF. The mean timeseries from each region was imported to MATLAB (R2017b; Natick, Massachusetts, USA) and static functional connectivity quantified as the Pearson correlation coefficient between each pair of regions in the FPN (7-by-7 matrix). As a control, the same was performed for all regions in the FreeSurfer parcellation (164-by-164). This measure was chosen to reflect the level of activity within the network.

Dynamic Functional Connectivity

Using the same timeseries data as above, dynamic functional connectivity of the FPN was measured during rest and task performance using the sliding window approach of previous studies [27, 56-58]. In short, the fMRI region timeseries were split into overlapping windows of 59.4-second length (27 TRs) and an 8.8-second shift (4 TRs), resulting in 55 windows for the resting-state fMRI and 35 for the task fMRI. Within each window, static functional connectivity was calculated. The absolute difference between consecutive windows was calculated and summed across all windows. With this approach, the resulting 7-by-7 matrix reflected the amount of variance in functional connectivity over time.

Structural Connectivity

Diffusion data were processed using FSL Diffusion Toolbox and probtrackx2 [59]: first, eddy-current induced distortions and subject motion were corrected using Eddy [60]. All image volumes were registered to the average b=0 image and non-brain structures were removed. Diffusion parameters were estimated and ROIs from the two atlases were registered to diffusion space by registering the T1 brain image to the average b=0 image by rigid transformation and applying this transformation to the atlas. Fibre tracking was applied to every pair of ROIs to create matrices of the number of reconstructed streamlines between them, from the seeded 5,000. This was measured twice (A-to-B and B-to-A). The following standard formula was then applied in order to sum the streamline counts for each direction and scale the result by their volumes [61]: aij = (sij + sji)⁄(mi + mj). Where aij is the corrected summary streamline count, sij and sji are streamline counts in each direction between ROIs i and j, and mi and mj are voxel counts in each ROI. Diffusion data were incomplete for 4 participants from the MS group.

Statistical Analysis

All variables were first tested for normality using Kolmogorov-Smirnov tests and by visual inspection of histograms. The level of statistical significance for all tests was set at 5%. Group differences in age, sex, education and head motion were assessed by chi-square test or t-test.

To test for differences in FPN volumes between groups and between hemispheres, and for differences in lesion and whole-brain volumes, we applied two-sample t-tests (Supplemental Information).

To test for between-group differences in FPN connectivity, we used multiple regression, inputting the dummy group variable as the dependent variable, age, gender, education and average head motion as a block of independents via the enter method, and the connectivity measure in a second block. For tests in the FPN, using either dynamic or static connectivity, and structural connectivity, this comprised 5 tests. We evaluated the impact of multiple comparisons against a Bonferroni-corrected alpha (α′=0.05/5=0.01).

Relationships between brain connectivity measures in the FPN and attention performance (ANT) were assessed by linear regression in the MS group. For these tests, FPN average measures of static functional connectivity, dynamic functional connectivity and structural connectivity were normalised to their whole-brain equivalents, so that they did not simply reflect global inter-subject differences. The primary analysis was of the relationship between the ANT Alerting score [62, 63] and dynamic connectivity in the FPN at rest. We also tested the relationship to static functional and structural connectivity. Again, we used a Bonferroni-corrected alpha for 3 tests to account for multiple comparisons (α′=0.05/3=0.017). Last, we calculated the task-induced change in static and dynamic functional connectivity as the difference between resting and task-active connectivities and tested its relationship to attention performance (ANT) using linear regression. For all tests, age, gender, education, average head motion and normalised lesion volume were entered as covariates (via the backward-stepwise method).

Results

Demographics and Neurocognitive Performance

Thirty-seven people with MS and 23 healthy controls were recruited (Table 1). The groups were not significantly different in age, sex, education or head motion. The MS group performed significantly worse in all tests of attention (ANT, SDMT, PASAT), depression, fatigue and sleep quality. Compared to normative MS data, our sample performed poorly in attention but was similar in terms of physical disability, sleep quality or depression (Table S1). MS patients had significantly greater lesion volumes and smaller brain volumes than controls (Supplemental Information).

View this table:
Table 1. Demographic and cognitive data

Data reported as mean ± standard deviation. Group difference was tested using two-tailed independent-samples t-test except where indicated.

Group Differences within Fronto-Parietal Network Connectivity at Rest

While at rest, there were moderate strength significant differences between the MS and control groups in terms of static and dynamic functional connectivity within the FPN (Table 2; static: F=2.91, R2=-0.07, p=0.032; dynamic: F=3.28, R2=0.08, p=0.026), with MS patients having lower static connectivity (Figure 3a) and higher dynamic connectivity. Structural connectivity (Figure 3b) was also significantly lower in the FPN in people with MS (Figure 3b; F=8.79, R2=-0.13, p=0.004) and this result persisted after Bonferroni correction, as was grey matter volume (Figure 3c). Across the whole brain, there was no significant effect for people with MS having lower structural connectivity (F=2.24, R2=0.07, p=0.063) or for functional connectivity (either static or dynamic; F=0.34, R2=0.01, p=0.551; F=0.47, R2=0.02, p=0.495).

View this table:
Table 2. Comparisons between MS and control groups on brain connectivity measures at rest.

Relationships between Connectivity of the Fronto-Parietal Network and Attention Performance in MS

Within the MS group, higher dynamic functional connectivity measured in the FPN during rest was significantly related to poorer performance in the ANT Alerting score (Figure 3d; F=3.87, p=0.008, R2=-0.20). In this model, age, gender and education were not individually predictive but greater lesion volumes were. Neither corresponding test for static functional connectivity nor structural connectivity was significant (Table 3). The corresponding tests in the whole brain were also non-significant. Of the secondary attention measures, both the PASAT and SDMT scores were related to dynamic FPN functional connectivity in the same direction as for the ANT, but these effects were non-significant when controlling for multiple comparisons and neither were significantly related to static functional connectivity (Table S4).

Figure 3.
Figure 3. Results of statistical tests on attention related network

(A) comparison of static functional connectivity between resting and task-active states in the fronto-parietal network, (B) comparison between groups of structural connectivity strength in the fronto-parietal network (C) hemispheric volumes of the fronto-parietal network, and (D) dynamic connectivity of the fronto-parietal network compared to attention performance in MS (black dots, R2=0.2) and control (white circles).

View this table:
Table 3. Relationships between attention performance and connectivity of the fronto-parietal network at rest in MS

Fronto-Parietal Network Connectivity During a Performance of a Sustained Attention Task

Reaction times decreased over the course of the task for both groups (Figure 1b) and reaction times of the MS group were consistently slower than those of controls (two-sample t-test of mean reaction times; control mean: 2240ms, SD 837; MS mean: 2572ms, SD 999; t=-6.522, p=<0.001). There was no difference in the rate of change in RT between groups (two-sample t-test of the slope of the linear regression line; control mean: −13.77, SD 7.27; MS mean: - 18.91, SD 13.45; t=1.326, p=0.196). The proportion of incorrect answers was low for both groups, with the MS group giving 59 incorrect responses out of 1485 trials (4%) and the control group giving 21 incorrect responses out of 1035 trials (2%).

During performance of the continuous sustained attention task, static functional connectivity of the FPN was significantly greater in the MS group than in controls (F=7.65, R2=0.10, p=0.008; Table S5; Figure 3a). When comparing connectivity measures during rest against those during task performance in the MS and control groups separately, we found that dynamic functional connectivity of the FPN was significantly lower during task performance than at rest in both groups (Table S2). This “task-induced decrease” in static connectivity (i.e., the difference between rest and task-active states) was significantly related to attention performance in the ANT Alerting score (both groups together; Table S3; F=9.12, p<0.001, R2=0.23), with better attention performance relating to a greater rest-to-task increase in static connectivity.

Discussion

Using functional MRI data at rest and during task performance, and a detailed attention assessment, we found that the FPN was more weakly (lower static connectivity) and more variably interconnected (higher dynamic connectivity) in people with MS, and that this corresponded to poorer attention performance at the group level. During performance of an attentionally-demanding task, the MS group had, conversely, greater synchronicity (static connectivity) of brain regions belonging to the FPN (frontal gyri, posterior cingulate, superior parietal lobule, intraparietal sulci, superior temporal gyri and lateral occipital gyri) compared to controls yet performed worse in the task overall. We then showed that task-induced changes in static functional connectivity explained deficits in attention performance better than either static or dynamic functional connectivity during rest or task alone. These findings appeared to differ between the FPN and whole brain. Together, these findings support the hypothesis that MS-related functional deficits in attention are associated with greater resting dynamic functional connectivity of the attention related FPN.

Static functional connectivity represents the amount of synchronicity between regions of the brain and is thought to reflect connection strength [64]. In our study, static functional connectivity of the FPN was significantly lower at rest in MS compared to controls but higher during task performance, and its change between resting and stimulated states was related to attention performance. This change between resting and task states may reflect functional topological network reorganisation [65]; and may be abnormal in MS due to greater metabolic requirements during the task needed to achieve the same level of maintained performance as controls [66] (as shown by the same rate of change in reaction times).

In contrast, dynamic functional connectivity represents the variability of synchronised cortical metabolic demand and, therefore, has been described as underlying a capability to “switch” between intrinsic and stimulated states [67]. In our sample, higher dynamic connectivity of the FPN at rest was related to poorer attention performance in people with MS, suggesting that a state-changing capability of these regions is functionally important. However, it did not appear to be abnormal compared to our control sample during task performance, nor was its change when switching from rest to task states, whereas static connectivity was. This may suggest that variability in dynamic functional connectivity may be less important for attention performance than static functional connectivity.

A potential limitation of this work is in the heterogeneity of our sample, which included both relapsing-remitting and secondary-progressive patients, and a range of medications. We did not distinguish people with MS based on current medications due to limited statistical power. Age could feasibly be an important confound in this type of study due to differing baseline tissue characteristics [68]. We also were not able to include as many controls in the study as we had estimated would be required, having 4 fewer participants than shown in our power calculation. While the mSDMT task performed inside the scanner was designed to replicate the cognitive process used during the normal SDMT, these two tasks differ in their formats and thus cannot be directly compared, nor can the mSDMT with the ANT. The handedness of the subjects was not homogenous, and so the fMRI findings could have been impacted by the inclusion of both left- and right-handed subjects together. The tractography method used did not take into account the effect of lesions and may be unreliable, and so could be improved by the use of other tools, e.g. MRTrix3 ACT [69, 70]. Finally, we acknowledge that the statistical power of the study was limited and should constrain the interpretation of the results.

Dynamic functional connectivity of the FPN may be a useful indicator of deficits in sustained attention in people with MS. Comparison between functional connectivity measurements at rest and during task-stimulated states may allow additional inferences which add value to standard task or resting-state fMRI individually. Further investigation of functional brain network dynamics in MS is warranted and may contribute to new mechanistic insights into cognitive dysfunction.

Data Availability

Summary statistical data generated during and/or analysed during the current study are available from the corresponding author on reasonable request. Original datasets analysed are not publicly available to protect patient information.

Supplemental

Supplemental information, Tables S1-5.

Statements and Declarations

Competing Interests

No competing financial interests exist.

Declarations

Funding

Funded by the UK MS Society (grant 988).

Author Contributions

Conceived the study: RAD, DPA, CSC, TW. Performed image analysis: TW, DM. All authors discussed the results and contributed to the final manuscript.

Data Availability

Summary statistical data generated during and/or analysed during the current study are available from the corresponding author on reasonable request. Original datasets analysed are not publicly available to protect patient information.

Acknowledgements

We thank Prof Nadina Lincoln, Ms Cara Knight and Ms Holly Griffiths for help with recruitment; and Andrew Cooper for MRI scanning.

Footnotes

  • thomas_welton{at}nni.com.sg, dewen.meng{at}nottingham.ac.uk, roshan.dasnair{at}nottingham.ac.uk, cris.constantinescu{at}nottingham.ac.uk, dorothee.auer{at}nottingham.ac.uk, rob.dineen{at}nottingham.ac.uk

References

  1. 1.↵
    Achiron A, Doniger GM, Harel Y, Appleboim-Gavish N, Lavie M, Simon ES. Prolonged response times characterize cognitive performance in multiple sclerosis. European Journal of Neurology 2007;14(10):1102–1108.
    OpenUrlCrossRefPubMed
  2. 2.
    Harel Y, Kalron A, Menascu S, et al. Cognitive function in multiple sclerosis: A long-term look on the bright side. PLOS ONE 2019;14(8):e0221784.
    OpenUrl
  3. 3.↵
    McCarthy M, Beaumont JG, Thompson R, Peacock S. Modality-specific aspects of sustained and divided attentional performance in multiple sclerosis. Archives of Clinical Neuropsychology 2005;20(6):705–718.
    OpenUrlCrossRefPubMedWeb of Science
  4. 4.↵
    Arnett PA, Strober LB. Cognitive and neurobehavioral features in multiple sclerosis. Expert Review of Neurotherapeutics 2011;11(3):411-424.
    OpenUrl
  5. 5.↵
    Baumstarck-Barrau K, Simeoni M-C, Reuter F, et al. Cognitive function and quality of life in multiple sclerosis patients: a cross-sectional study. BMC neurology 2011;11:17–17.
    OpenUrl
  6. 6.↵
    Shalev L, Kolodny T, Shalev N, Mevorach C. Attention Functioning Among Adolescents With Multiple Learning, Attentional, Behavioral, and Emotional Difficulties. J Learn Disabil 2016;49(6):582–596.
    OpenUrlCrossRefPubMed
  7. 7.↵
    Berard JA, Smith AM, Walker LAS. A Longitudinal Evaluation of Cognitive Fatigue on a Task of Sustained Attention in Early Relapsing-Remitting Multiple Sclerosis. International journal of MS care 2018;20(2):55–61.
    OpenUrl
  8. 8.↵
    das Nair R, Martin K-J, Lincoln NB. Memory rehabilitation for people with multiple sclerosis. Cochrane Database of Systematic Reviews 2016;(3).
  9. 9.↵
    Bisecco A, Stamenova S, Caiazzo G, et al. Attention and processing speed performance in multiple sclerosis is mostly related to thalamic volume. Brain Imaging Behav 2018;12(1):20–28.
    OpenUrl
  10. 10.↵
    Welton T, Kent D, Constantinescu CS, Auer DP, Dineen RA. Functionally Relevant White Matter Degradation in Multiple Sclerosis: A Tract-based Spatial Meta-Analysis. Radiology 2015;275(1):89–96.
    OpenUrlCrossRef
  11. 11.↵
    Matias-Guiu JA, Cortés-Martínez A, Montero P, et al. Structural MRI correlates of PASAT performance in multiple sclerosis. BMC Neurology 2018;18(1):214.
    OpenUrl
  12. 12.↵
    Sporns O. Structure and function of complex brain networks. Dialogues in clinical neuroscience 2013;15(3):247–262.
    OpenUrlPubMed
  13. 13.↵
    Hopfinger JB, Buonocore MH, Mangun GR. The neural mechanisms of top-down attentional control. Nature Neuroscience 2000;3:284.
    OpenUrlCrossRefPubMedWeb of Science
  14. 14.
    Vossel S, Geng JJ, Fink GR. Dorsal and ventral attention systems: distinct neural circuits but collaborative roles. The Neuroscientist : a review journal bringing neurobiology, neurology and psychiatry 2014;20(2):150–159.
    OpenUrl
  15. 15.
    Shen K-k, Welton T, Lyon M, et al. Structural core of the executive control network: A high angular resolution diffusion MRI study. Human Brain Mapping 2020;41(5):1226–1236.
    OpenUrl
  16. 16.↵
    Meng D, Welton T, Elsarraj A, et al. Dorsolateral prefrontal circuit effective connectivity mediates the relationship between white matter structure and PASAT-3 performance in multiple sclerosis. Human Brain Mapping 2021;42(2):495–509.
    OpenUrl
  17. 17.↵
    Loitfelder M, Filippi M, Rocca M, et al. Abnormalities of resting state functional connectivity are related to sustained attention deficits in MS. PLoS One 2012;7(8):e42862.
    OpenUrlCrossRefPubMed
  18. 18.↵
    Cader S, Cifelli A, Abu-Omar Y, Palace J, Matthews PM. Reduced brain functional reserve and altered functional connectivity in patients with multiple sclerosis. Brain 2006;129(Pt 2):527–37.
    OpenUrlCrossRefPubMedWeb of Science
  19. 19.↵
    Schoonheim MM, Hulst HE, Brandt RB, et al. Thalamus structure and function determine severity of cognitive impairment in multiple sclerosis. Neurology 2015;84(8):776–83.
    OpenUrlCrossRefPubMed
  20. 20.↵
    Sbardella E, Tona F, Petsas N, et al. Functional connectivity changes and their relationship with clinical disability and white matter integrity in patients with relapsing-remitting multiple sclerosis. Mult Scler 2015;21(13):1681–92.
    OpenUrlCrossRefPubMed
  21. 21.↵
    Hutchison RM, Womelsdorf T, Allen EA, et al. Dynamic functional connectivity: Promise, issues, and interpretations. NeuroImage 2013;80:doi:10.016/j.neuroimage.2013.05.079.
    OpenUrlCrossRef
  22. 22.↵
    Chen JE, Rubinov M, Chang C. Methods and Considerations for Dynamic Analysis of Functional MR Imaging Data. Neuroimaging Clin N Am 2017;27(4):547–560.
    OpenUrl
  23. 23.↵
    Filippi M, Spinelli EG, Cividini C, Agosta F. Resting state dynamic functional connectivity in neurodegenerative conditions: a review of magnetic resonance imaging findings. Frontiers in neuroscience 2019;13.
  24. 24.↵
    Jones DT, Vemuri P, Murphy MC, et al. Non-stationarity in the "resting brain’s" modular architecture. PLoS One 2012;7(6):e39731.
    OpenUrlCrossRefPubMed
  25. 25.↵
    Kim J, Criaud M, Cho SS, et al. Abnormal intrinsic brain functional network dynamics in Parkinson’s disease. Brain 2017;140(11):2955–2967.
    OpenUrlCrossRef
  26. 26.↵
    d’Ambrosio A, Valsasina P, Gallo A, et al. Reduced dynamics of functional connectivity and cognitive impairment in multiple sclerosis. Multiple Sclerosis Journal 2019;0(0):1352458519837707.
    OpenUrl
  27. 27.↵
    van Geest Q, Hulst HE, Meijer KA, Hoyng L, Geurts JJG, Douw L. The importance of hippocampal dynamic connectivity in explaining memory function in multiple sclerosis. Brain Behav 2018;8(5):e00954.
    OpenUrlCrossRef
  28. 28.↵
    Huang M, Zhou F, Wu L, et al. White matter lesion loads associated with dynamic functional connectivity within attention network in patients with relapsing-remitting multiple sclerosis. Journal of Clinical Neuroscience 2019;65:59–65.
    OpenUrl
  29. 29.↵
    Eijlers AJC, Wink AM, Meijer KA, Douw L, Geurts JJG, Schoonheim MM. Reduced Network Dynamics on Functional MRI Signals Cognitive Impairment in Multiple Sclerosis. Radiology 2019;292(2):449–457.
    OpenUrlCrossRef
  30. 30.↵
    Vázquez-Marrufo M, Galvao-Carmona A, González-Rosa JJ, et al. Neural correlates of alerting and orienting impairment in multiple sclerosis patients. PLoS One 2014;9(5):e97226.
    OpenUrl
  31. 31.
    Vazquez-Marrufo M, Sarrias-Arrabal E, Martin-Clemente R, Galvao-Carmona A, Navarro G, Izquierdo G. Altered phase and nonphase EEG activity expose impaired maintenance of a spatial-object attentional focus in multiple sclerosis patients. Sci Rep 2020;10(1):20721.
    OpenUrl
  32. 32.↵
    Veréb D, Tóth E, Bozsik B, et al. Altered brain network function during attention-modulated visual processing in multiple sclerosis. Mult Scler 2021;27(9):1341–1349.
    OpenUrl
  33. 33.↵
    Corbetta M, Patel G, Shulman GL. The reorienting system of the human brain: from environment to theory of mind. neuron 2008;58(3):306–324.
    OpenUrlCrossRefPubMedWeb of Science
  34. 34.↵
    Corbetta M, Shulman GL. Control of goal-directed and stimulus-driven attention in the brain. nature reviews neuroscience 2002;3(3):201–215.
    OpenUrlCrossRefPubMedWeb of Science
  35. 35.↵
    Itti L, Koch C. A saliency-based search mechanism for overt and covert shifts of visual attention. vision research 2000;40(10):1489–1506.
    OpenUrlCrossRefPubMedWeb of Science
  36. 36.↵
    van Geest Q, Douw L, van ’t Klooster S, et al. Information processing speed in multiple sclerosis: Relevance of default mode network dynamics. Neuroimage Clin 2018;19:507–515.
    OpenUrl
  37. 37.↵
    Hellyer PJ, Jachs B, Clopath C, Leech R. Local inhibitory plasticity tunes macroscopic brain dynamics and allows the emergence of functional brain networks. NeuroImage 2016;124:85–95.
    OpenUrl
  38. 38.↵
    Polman CH, Reingold SC, Banwell B, Clanet M, Cohen JA, Filippi M. Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol 2011;69.
  39. 39.↵
    Fan J, McCandliss BD, Sommer T, Raz A, Posner MI. Testing the efficiency and independence of attentional networks. J Cogn Neurosci 2002;14(3):340–7.
    OpenUrlCrossRefPubMedWeb of Science
  40. 40.↵
    Arora S, Lawrence MA, Klein RM. The Attention Network Test Database: ADHD and Cross-Cultural Applications. Frontiers in psychology 2020;11:388–388.
    OpenUrl
  41. 41.↵
    Tombaugh TN. A comprehensive review of the Paced Auditory Serial Addition Test (PASAT). Archives of Clinical Neuropsychology 2006;21(1):53–76.
    OpenUrlCrossRefPubMedWeb of Science
  42. 42.↵
    Smith A. Symbol digit modalities test (SDMT) manual (revised) Western Psychological Services. Los Angeles 1982.
  43. 43.↵
    Fisk JD, Pontefract A, Ritvo PG, Archibald CJ, Murray TJ. The impact of fatigue on patients with multiple sclerosis. Can J Neurol Sci 1994;21(1):9–14.
    OpenUrlCrossRefPubMedWeb of Science
  44. 44.↵
    Beck AT, Steer RA, Ball R, Ranieri W. Comparison of Beck Depression Inventories - IA and -II in psychiatric outpatients. J Pers Assess 1996;67(3):588–97.
    OpenUrlCrossRefPubMedWeb of Science
  45. 45.↵
    Buysse DJ, Reynolds CF, 3rd, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res 1989;28(2):193–213.
    OpenUrl
  46. 46.↵
    Cutter GR, Baier ML, Rudick RA, et al. Development of a multiple sclerosis functional composite as a clinical trial outcome measure. Brain 1999;122(5):871–882.
    OpenUrlCrossRefPubMedWeb of Science
  47. 47.↵
    Leavitt VM, Wylie G, Genova HM, Chiaravalloti ND, DeLuca J. Altered effective connectivity during performance of an information processing speed task in multiple sclerosis. Mult Scler 2012;18(4):409–17.
    OpenUrlCrossRefPubMed
  48. 48.↵
    Destrieux C, Fischl B, Dale A, Halgren E. Automatic parcellation of human cortical gyri and sulci using standard anatomical nomenclature. NeuroImage 2010;53(1):1–15.
    OpenUrlCrossRefPubMedWeb of Science
  49. 49.↵
    Llufriu S, Martinez-Heras E, Solana E, et al. Structural networks involved in attention and executive functions in multiple sclerosis. NeuroImage: Clinical 2017;13:288–296.
    OpenUrl
  50. 50.↵
    Yarkoni T, Poldrack RA, Nichols TE, Van Essen DC, Wager TD. Large-scale automated synthesis of human functional neuroimaging data. Nature methods 2011;8(8):665–670.
    OpenUrl
  51. 51.↵
    Friedman L, Glover GH. Report on a multicenter fMRI quality assurance protocol. Journal of Magnetic Resonance Imaging 2006;23(6):827–839.
    OpenUrlCrossRefPubMedWeb of Science
  52. 52.↵
    Beckmann CF, Smith SM. Probabilistic independent component analysis for functional magnetic resonance imaging. IEEE Trans Med Imaging 2004;23(2):137–52.
    OpenUrlCrossRefPubMedWeb of Science
  53. 53.↵
    Jenkinson M, Beckmann CF, Behrens TE, Woolrich MW, Smith SM. FSL. Neuroimage 2012;62(2):782–90.
    OpenUrlCrossRefPubMedWeb of Science
  54. 54.↵
    Griffanti L, Douaud G, Bijsterbosch J, et al. Hand classification of fMRI ICA noise components. NeuroImage 2017;154:188–205.
    OpenUrlCrossRef
  55. 55.↵
    Jenkinson M, Smith S. A global optimisation method for robust affine registration of brain images. Med Image Anal 2001;5(2):143–56.
    OpenUrlCrossRefPubMedWeb of Science
  56. 56.↵
    Leonardi N, Van De Ville D. On spurious and real fluctuations of dynamic functional connectivity during rest. Neuroimage 2015;104:430–436.
    OpenUrlCrossRefPubMed
  57. 57.
    Leonardi N, Shirer WR, Greicius MD, Van De Ville D. Disentangling dynamic networks: separated and joint expressions of functional connectivity patterns in time. Human brain mapping 2014;35(12):5984–5995.
    OpenUrlCrossRefPubMedWeb of Science
  58. 58.↵
    Douw L, Wakeman DG, Tanaka N, Liu H, Stufflebeam SM. State-dependent variability of dynamic functional connectivity between frontoparietal and default networks relates to cognitive flexibility. Neuroscience 2016;339:12–21.
    OpenUrlCrossRefPubMed
  59. 59.↵
    Behrens TE, Woolrich MW, Jenkinson M, et al. Characterization and propagation of uncertainty in diffusion-weighted MR imaging. Magn Reson Med 2003;50(5):1077–88.
    OpenUrlCrossRefPubMedWeb of Science
  60. 60.↵
    Andersson JLR, Sotiropoulos SN. An integrated approach to correction for off-resonance effects and subject movement in diffusion MR imaging. Neuroimage 2016;125:1063–1078.
    OpenUrlCrossRefPubMed
  61. 61.↵
    Cheng H, Wang Y, Sheng J, et al. Characteristics and variability of structural networks derived from diffusion tensor imaging. Neuroimage 2012;61(4):1153–1164.
    OpenUrlCrossRefPubMedWeb of Science
  62. 62.↵
    Roth AK, Denney DR, Lynch SG. Information processing speed and attention in multiple sclerosis: Reconsidering the Attention Network Test (ANT). J Clin Exp Neuropsychol 2015;37(5):518–29.
    OpenUrl
  63. 63.↵
    Xiao M, Ge H, Khundrakpam BS, et al. Attention performance measured by attention network test is correlated with global and regional efficiency of structural brain networks. Frontiers in behavioral neuroscience 2016;10:194.
  64. 64.↵
    Van Dijk KR, Hedden T, Venkataraman A, Evans KC, Lazar SW, Buckner RL. Intrinsic functional connectivity as a tool for human connectomics: theory, properties, and optimization. Journal of neurophysiology 2009;103(1):297–321.
    OpenUrlPubMed
  65. 65.↵
    Welton T, Constantinescu CS, Auer DP, Dineen RA. Graph Theoretic Analysis of Brain Connectomics in Multiple Sclerosis: Reliability and Relationship with Cognition. Brain Connectivity 2020;10(2):95–104.
    OpenUrl
  66. 66.↵
    Patel KR, Tobyne S, Porter D, Bireley JD, Smith V, Klawiter E. Structural disconnection is responsible for increased functional connectivity in multiple sclerosis. Brain Struct Funct 2018;223(5):2519–2526.
    OpenUrl
  67. 67.↵
    Hansen ECA, Battaglia D, Spiegler A, Deco G, Jirsa VK. Functional connectivity dynamics: Modeling the switching behavior of the resting state. NeuroImage 2015;105:525–535.
    OpenUrlCrossRefPubMedWeb of Science
  68. 68.↵
    Phillips OR, Clark KA, Luders E, et al. Superficial White Matter: Effects of Age, Sex, and Hemisphere. Brain Connectivity 2013;3(2):146–159.
    OpenUrl
  69. 69.↵
    Smith RE, Tournier JD, Calamante F, Connelly A. Anatomically-constrained tractography: improved diffusion MRI streamlines tractography through effective use of anatomical information. Neuroimage 2012;62(3):1924–38.
    OpenUrlCrossRefPubMed
  70. 70.↵
    Tournier JD, Smith R, Raffelt D, et al. MRtrix3: A fast, flexible and open software framework for medical image processing and visualisation. NeuroImage 2019;202:116137.
Back to top
Posted July 09, 2023.
Download PDF

Supplementary Material

Data/Code
Email
Fronto-parietal network dynamics to understand deficits in attention performance in multiple sclerosis
Thomas Welton, Dewen Meng, Roshan das Nair, Cris S Constantinescu, Dorothee P Auer, Rob A Dineen
medRxiv 2023.07.08.23292404; doi: https://doi.org/10.1101/2023.07.08.23292404
Twitter logo Facebook logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area