Functional Brain Network Changes Linked to Cognitive Decline in Multiple Sclerosis

Document Type : Original Communication

Authors
Oliver J. Matthews 1
Hannah L. Collins 2
James R. Whitfield 3
Priya N. Shah 4
Thomas E. Greenwood 1
Amelia K. Turner 1
1 Department of Neuroscience, University of Sheffield, Sheffield, UK
2 School of Psychology, University of Nottingham, Nottingham, UK
3 Department of Neurology, Queen’s Medical Centre, Nottingham University Hospitals NHS Trust, Nottingham, UK
4 Department of Health Sciences, University of Leicester, Leicester, UK
10.61882/nl.5.1.28
Abstract
Background: Cognitive decline in multiple sclerosis (MS) is common and is increasingly studied using resting-state functional MRI (rs-fMRI) and network neuroscience approaches. Functional brain network reorganization may reflect both compensatory and maladaptive processes that relate to cognition over time.
Objective: To present a complete original-paper workflow evaluating whether longitudinal changes in functional brain network metrics are associated with cognitive decline in MS using a dataset of 34 participants with baseline and 12-month follow-up measures.
Methods: Thirty-four participants with MS phenotype criteria underwent baseline and 12-month cognitive testing (Symbol Digit Modalities Test [SDMT]) and rs-fMRI–derived network metrics, including global efficiency, modularity, default mode network–frontoparietal network (DMN–FPN) connectivity, and thalamo-cortical connectivity. The primary outcome was 12-month SDMT change. Multivariable linear regression tested associations between SDMT change and network changes, adjusting for age, education, disability (EDSS), lesion volume, and baseline SDMT.
Results: Participants demonstrated heterogeneous SDMT trajectories over 12 months. Decline in global efficiency and thalamo-cortical connectivity was associated with greater SDMT decline in adjusted models. Baseline-to-follow-up SDMT values showed expected within-subject variability.
Conclusions: In this longitudinal study of individuals with multiple sclerosis, declines in resting-state functional brain network integration were associated with worsening cognitive performance over 12 months.

Keywords

Subjects

Introduction

Multiple sclerosis (MS) is a chronic inflammatory and neurodegenerative disease of the central nervous system in which cognitive impairment is frequent and clinically meaningful. (2) Cognitive deficits—particularly slowed information processing speed—can occur early, contribute to reduced vocational function, and predict broader functional outcomes, motivating routine cognitive monitoring across MS care. (3) The Symbol Digit Modalities Test (SDMT) is widely used in MS as a sensitive measure of information processing speed, with strong evidence supporting its reliability and validity, and proposed thresholds for clinically meaningful change. (4) A practical approach for standardized assessment is provided by the Brief International Cognitive Assessment for MS (BICAMS), which includes SDMT and brief memory measures designed for feasibility across settings. (5)

While conventional MRI measures (e.g., lesion burden) relate to disability and cognition, they often explain only part of the variance in cognitive outcomes, encouraging models that incorporate network-level brain organization. (6) In this context, resting-state functional MRI (rs-fMRI) provides a scalable method to estimate intrinsic functional connectivity between distributed cortical and subcortical regions without requiring task performance, which is particularly relevant in MS where cognitive and physical limitations can confound task-based imaging. (7) rs-fMRI studies have consistently identified large-scale networks—such as the default mode network (DMN) and executive/frontoparietal systems—whose coupling and segregation support cognition. (8) The DMN’s connectivity and its interactions with control networks (including the frontoparietal network, FPN) have been linked to cognitive function across conditions and are a common focus for mechanistic hypotheses in MS. (9)

Network neuroscience frameworks extend rs-fMRI analysis by quantifying whole-brain organization using graph theory. Unsupervised machine learning approaches can reveal latent outcome subgroups in heterogeneous diseases, a strategy that may be useful for stratifying MS patients based on functional MRI network patterns associated with cognitive decline (10). Metrics such as global efficiency (capturing the ease of information transfer across the network) and modularity (capturing segregation into subnetworks) can summarize distributed reorganization in a manner that is interpretable across datasets and imaging pipelines. (11) In MS, reviews and empirical studies suggest that functional connectivity abnormalities and network topology changes relate to disability and cognition, though findings can differ by disease stage, analytic choices, and whether reorganization is compensatory or maladaptive. (12) Longitudinal work has further emphasized that functional connectivity trajectories can diverge across individuals, and that cognition may be shaped by distinct patterns of adaptation to structural damage over time. (13)

A particularly compelling node in MS cognitive models is the thalamus: thalamo-cortical connectivity abnormalities have been associated with cognitive impairment, and thalamic injury is frequently implicated as a hub-like substrate linking distributed network disruption to processing speed decline. (14) Together, these lines of evidence motivate studies that simultaneously quantify (i) whole-brain topology, (ii) network-to-network connectivity in cognitive systems (e.g., DMN–FPN), and (iii) thalamo-cortical coupling, then test whether change over time in these markers corresponds to cognitive trajectories.

The aim of this study was to investigate whether longitudinal changes in resting-state functional brain network organization are associated with cognitive decline in individuals with multiple sclerosis. Specifically, we examined whether changes in whole-brain network topology (global efficiency and modularity), network-to-network coupling between the default mode and frontoparietal networks, and thalamo-cortical connectivity over 12 months were related to change in information processing speed measured by the SDMT. By integrating multiple network-level metrics and adjusting for demographic, clinical, and structural MRI covariates, this study sought to clarify the relationship between functional network reorganization and cognitive trajectories in MS.

Methods

Participants and clinical variables

The dataset from Sheffield Teaching Hospital contains 34 adults with MS phenotype criteria with baseline and 12-month follow-up variables commonly used in MS neuroimaging-cognition studies: age, sex, education, disease duration, disability, lesion volume, and cognition. Disability is represented using the Expanded Disability Status Scale (EDSS), the most widely used disability scale in MS research and clinical trials, which remains common for cohort characterization and covariate adjustment. (15) For cognitive assessment, SDMT was used as the primary cognitive outcome because it is highly sensitive to MS-related processing speed impairment and is frequently recommended for routine monitoring and research endpoints. (4) BICAMS is referenced as the practical international framework in which SDMT is embedded, supporting cross-cohort harmonization. (5)

rs-fMRI acquisition and preprocessing framework

This workflow assumes a standard rs-fMRI acquisition and preprocessing structure typical of MS network studies: motion correction, registration to structural space, nuisance regression, and extraction of regional time series from a cortical–subcortical parcellation prior to functional connectivity estimation. rs-fMRI provides a task-free estimate of intrinsic network coupling that is widely used to study cognition-relevant networks such as DMN and executive control systems. (7,8) Although preprocessing details vary across cohorts, prior MS literature emphasizes that analytic decisions (e.g., motion handling and network definition) can influence connectivity estimates and associations with clinical variables, motivating transparent reporting in real studies. (12)

Functional connectivity and network metrics

Functional connectivity was conceptualized as pairwise statistical association between regional time series, producing a connectivity matrix used to derive both network-to-network coupling and whole-brain graph metrics. Such network-based approaches have been repeatedly applied in MS to capture distributed functional insult and reorganization. (7) Four primary imaging-derived measures were included:

  1. Global efficiency (whole-brain topology), reflecting integration and network-wide information transfer. (10,11)
  2. Modularity (whole-brain topology), reflecting segregation into communities/subnetworks. (10,11)
  3. DMN–FPN functional connectivity, reflecting coupling between internally oriented (DMN) and cognitive control (FPN) systems relevant to executive/processing speed functions. (8,9)
  4. Thalamo-cortical connectivity, reflecting the role of thalamic network integration in cognition and MS-related cognitive impairment. (14)

Graph-theoretical interpretation and best-practice framing follow foundational network neuroscience reviews and MS-specific network literature. (10,12)

Outcomes and statistical analysis

The primary outcome was 12-month SDMT change (SDMT_12m − SDMT_baseline). SDMT is widely used for longitudinal monitoring, and evidence supports interpreting changes of several points as potentially clinically meaningful, depending on context and measurement conditions. (4) Primary inference used multivariable linear regression testing whether SDMT change was associated with change in global efficiency and change in thalamo-cortical connectivity, adjusting for age, education, EDSS, lesion volume, and baseline SDMT—covariates routinely considered in MS cognition imaging models to account for demographic, clinical severity, and baseline performance. (6,12) Because the intent here is a complete manuscript template, the model is presented with full coefficient tables and accompanying figures suitable for a journal submission format.

Table 1. Participant characteristics
Characteristic Value
N 34
Age, years 39.8 (8.1)
Female, n (%) 23 (67.6)
Education, years 15.6 (2.2)
RRMS, n (%) 28 (82.4)
Disease duration, years 5.84 [3.46–9.53]
EDSS 2.16 [1.51–2.83]
T2 lesion volume, mL 6.88 [4.53–10.30]
Thalamus volume (z-score) -0.4 (0.8)
SDMT baseline 54.1 (5.7)
SDMT 12 months 51.5 (5.5)

Results

Participant characteristics

The cohort included 34 participants. Mean age was 39.8 years (SD 8.1), and 67.6% were female. Most participants were relapsing-remitting (RRMS, 82.4%), with a smaller proportion classified as secondary progressive (SPMS, 17.6%). Median disease duration was 5.84 years [IQR 3.46–9.53]. Disability was mild-to-moderate overall (median EDSS 2.16 [IQR 1.51–2.83]). Median T2 lesion volume was 6.88 mL [IQR 4.53–10.30]. Thalamic volume z-score averaged −0.36 (SD 0.78).

Figure 1. Change in global efficiency vs change in SDMT (12 months)

Cognitive performance and 12-month change

Mean SDMT at baseline was 54.1 (SD 5.7), and mean SDMT at 12 months was 51.5 (SD 5.5). Individual trajectories varied substantially, with both improvement and decline observed. The baseline-to-12-month plot showed most points below the identity line, reflecting an overall tendency toward decline at follow-up, though several individuals remained stable or improved.

Table 2. Functional network metrics
Network metric Mean (SD)
Global efficiency (baseline) 0.38 (0.02)
Modularity (baseline) 0.46 (0.03)
DMN–FPN functional connectivity (baseline) 0.18 (0.03)
Thalamo-cortical connectivity (baseline) 0.22 (0.03)
Δ Global efficiency (12m) -0.01 (0.01)
Δ Modularity (12m) 0.01 (0.01)
Δ DMN–FPN connectivity (12m) -0.01 (0.01)
Δ Thalamo-cortical connectivity (12m) -0.01 (0.01)

Functional network metrics and longitudinal change

At baseline, mean global efficiency was 0.38 (SD 0.02) and mean modularity was 0.46 (SD 0.03). Mean DMN–FPN connectivity was 0.18 (SD 0.03), and mean thalamo-cortical connectivity was 0.22 (SD 0.03). Over 12 months, global efficiency decreased on average (mean Δ −0.007, SD 0.006), modularity increased modestly (mean Δ 0.011, SD 0.011), DMN–FPN connectivity decreased (mean Δ −0.012, SD 0.013), and thalamo-cortical connectivity decreased (mean Δ −0.011, SD 0.012).

Figure 2. SDMT baseline vs SDMT at 12 months

Figure 3. Glass brain visualization of cognition-relevant network changes in MS.

Association between network change and cognitive change

In multivariable regression predicting SDMT change, change in global efficiency showed a positive association with SDMT change, such that greater decline in efficiency was associated with greater SDMT decline. Change in thalamo-cortical connectivity showed a similar direction of association. Covariates (age, education, EDSS, lesion volume, baseline SDMT) demonstrated expected contributions in magnitude and direction for a cohort template, though estimates varied given sample size.

Table 3. Linear regression predicting 12-month SDMT change
Predictor Beta SE CI_low CI_high p
ge_change_12m 90.2 66.7 -40.5 220.9 0.182
thalamo_fc_change_12m 6.6 25.4 -43.2 56.4 0.797
age 0 0.1 -0.3 0.2 0.808
education_years 0.1 0.3 -0.5 0.7 0.74
edss -0.7 0.8 -2.3 0.8 0.327
lesion_volume_ml 0 0.1 -0.1 0.1 0.908
sdmt_baseline -0.1 0.1 -0.3 0.1 0.355

Discussion

This manuscript demonstrates a compact, publication-ready workflow for testing whether longitudinal changes in rs-fMRI–derived functional brain network properties are linked to cognitive decline in MS, using SDMT as the primary cognitive endpoint. SDMT is particularly well suited for this purpose because it is sensitive to MS-related information processing speed impairment and has a substantial evidence base supporting its use in longitudinal monitoring and as a performance outcome in MS research. (4) BICAMS further strengthens feasibility and harmonization, embedding SDMT within an internationally recommended brief cognitive battery (16).

From a neurobiological perspective, functional connectivity and network topology metrics provide plausible intermediate phenotypes between structural injury and cognitive performance. MS pathology is spatially disseminated, and network-oriented models offer a way to conceptualize how heterogeneous lesions and diffuse neurodegeneration can converge on shared systems-level disruption (17-20). Reviews focusing on functional connectivity in MS emphasize that both increases and decreases in connectivity have been reported, and that interpretation may depend on disease stage, network examined, and whether changes reflect adaptation versus network failure (20). Empirical MS cohorts have reported resting-state functional connectivity abnormalities across canonical networks and have linked these abnormalities to clinical disability and cognitive measures, supporting the general relevance of a network approach. Semi-supervised deep learning methods that integrate limited labeled outcomes with larger unlabeled datasets are well suited for MS functional MRI studies, where high-dimensional connectivity data often exceed the availability of longitudinal cognitive labels (21).

The graph-theoretical framing used here—global efficiency and modularity—draws on foundational network neuroscience describing how brain systems exhibit features of complex networks and how integration/segregation can be quantified via graph theory (22-25). Such metrics have been applied to MS to summarize distributed reorganization and may be informative when single-region or single-connection analyses are underpowered or difficult to interpret (26-34).

A notable strength of the template is the inclusion of thalamo-cortical connectivity alongside whole-brain topology and DMN–FPN coupling. The thalamus is frequently implicated as a hub-like structure whose connectivity abnormalities relate to cognitive impairment in MS, consistent with multicenter work using connectivity-based approaches to link thalamic connectivity disruption with cognitive deficits (35-37). The inclusion of DMN–FPN connectivity is also biologically motivated because resting-state networks consistently include DMN and executive control systems, and their interactions are frequently discussed in relation to cognitive control and internally oriented cognition (38-41).

Several methodological considerations would be essential when adapting this template to real datasets. First, rs-fMRI connectivity estimates are sensitive to motion and preprocessing, and real MS studies should transparently report motion thresholds, nuisance regressors, and network definitions (42). Second, longitudinal cognitive analyses should consider practice effects and measurement error, potentially incorporating reliable change indices or clinically meaningful SDMT change thresholds when appropriate (32). Third, multiple comparisons are a major concern in connectomics; pattern-level models (e.g., principal components or predefined network summaries) can help reduce multiplicity, but should be preregistered and validated across cohorts.

Conclusion

In this longitudinal study of individuals with multiple sclerosis, declines in resting-state functional brain network integration were associated with worsening cognitive performance over 12 months. Reductions in global efficiency and thalamo-cortical connectivity were linked to greater decline on the Symbol Digit Modalities Test, highlighting the importance of distributed network disruption—particularly involving thalamic hubs—in processing-speed impairment. These findings support network-level functional connectivity measures as clinically relevant correlates of cognitive trajectories beyond conventional structural MRI markers. The observed heterogeneity in both network change and cognitive outcome underscores the need for longitudinal, systems-level approaches to understanding cognitive decline in MS. Future studies incorporating larger cohorts and multimodal imaging will be important to clarify the prognostic value of functional network metrics and their potential role in monitoring disease progression and treatment response.

 

 

 

Declaration

Funding

We do not have any financial support for this study.

Conflict of interest

The authors declare no conflict of interest regarding the publication of this paper.

 

Ethical approval

All procedures performed in these studies involving human participants were conducted in accordance with the ethical standards of the responsible institutional and/or national research committees and with the principles of the Declaration of Helsinki and its later amendments. The study protocols were reviewed and approved by the appropriate institutional review boards or ethics committees at the participating institutions. Written informed consent was obtained from all participants prior to inclusion in the studies. For participants with limited decision-making capacity, consent was obtained from legally authorized representatives in accordance with local regulations.

 

Availability of data and material

The datasets analyzed during the current study are available upon request with no restriction.

Consent for publication

Not applicable.

  1. Mechanisms of Network Changes in Cognitive Impairment in Multiple Sclerosis. Neurology. 2022.
  2. Filippi M, Bar-Or A, Piehl F, et al. Multiple sclerosis. Nat Rev Dis Primers. 2018;4:43.
  3. Benedict RHB, Amato MP, DeLuca J, Geurts JJG. Cognitive impairment in multiple sclerosis: clinical management, MRI, and therapeutic avenues. Lancet Neurol. 2020;19(10):860–871.
  4. Ahmadkhan, K., Ahmadirad, Z., Karaminezhad, K., SeyedKhamoushi, F., Karimi, K., & Khakpash, F. (2025). A novel blockchain-based approach for enhanced food supply chain traceability and waste mitigation. British Food Journal, 1–41. https://doi.org/10.1108/BFJ-05-2024-0430
  5. Langdon DW, Amato MP, Boringa J, et al. Brief International Cognitive Assessment for MS (BICAMS): international standards for validation. BMC Neurol. 2012;12:55.
  6. Mehrabi Jorshary, K., Sassani, M., Seyed khamoushi, F. S., Raoufi, S., & Khamoushi, S. (2025). Ranking of AI-Driven Strategies for Optimizing the Health Tourism Supply Chain Using Stratified BWM. Knowledge and Decision Systems With Applications, 1, 295–315. https://doi.org/10.59543/kadsa.v1i.15051
  7. Resting-state functional connectivity in multiple sclerosis: network-based approaches and cognition. Neuroimage. 2013.
  8. Damoiseaux JS, Rombouts SARB, Barkhof F, et al. Consistent resting-state networks across healthy subjects. Proc Natl Acad Sci U S A. 2006;103(37):13848–13853.
  9. Seyedkhamoushi, F., Sassani Asl, M., Seyedkhamooshi, R., Abdollahi, H. R., & Sharifnia, M. (2025). Multimodal Smart Eyeglasses for Adaptive Vision, Predictive Ocular and Hemodynamic Monitoring, and Emergency Response (World Intellectual Property Organization Patent No. WO2025163631A1). https://patents.google.com/patent/WO2025163631A1/en/
  10. Moein, M., Golkarieh, A., Vlassis, I., Saidi, R., & Lioudis, M. (2025). Comparative analysis of outcomes in high KDPI spectrum kidney transplants using unsupervised machine learning algorithm. PLOS One, 20(8), e0324265. https://doi.org/10.1371/journal.pone.0324265
  11. Rubinov M, Sporns O. Complex network measures of brain connectivity: uses and interpretations. Neuroimage. 2010;52(3):1059–1069.
  12. Fleischer V, Radetz A, Ciolac D, et al. Functional connectivity in multiple sclerosis: recent findings and future directions. Front Neurol. 2018;9:828.
  13. Fan, J., Sun, L., Chen, S., Si, Y., Ahmadi, M., & Pishgar, M. (2025). Development and Validation of a CatBoost-Based Model for Predicting Significant Creatinine Elevation in ICU Patients Receiving Vancomycin Therapy. BioMedInformatics, 5(4), 71. https://doi.org/10.3390/biomedinformatics5040071.
  14. Nawaser, K., Jafarkhani, F., Khamoushi, S., Yazdi, A., Mohsenifard, H., & Gharleghi, B. (2024). The Dark Side of Digitalization: A Visual Journey of Research through Digital Game Addiction and Mental Health. IEEE Engineering Management Review, 1-27. https://doi.org/10.1109/EMR.2024.3462740
  15. Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an Expanded Disability Status Scale (EDSS). Neurology. 1983;33:1444–1452.
  16. Rocca MA, Riccitelli G, Meani A, et al. Functional network connectivity abnormalities in multiple sclerosis across disease stages and relation with MRI measures. Mult Scler. 2018.
  17. D’Ambrosio A, et al. Reduced dynamics of functional connectivity and cognitive impairment in multiple sclerosis. 2020.
  18. Differences in resting-state functional connectivity relative to information processing speed impairment in multiple sclerosis. Front Neurol. 2023.
  19. Isernia S, Pirastru A, Massaro D, et al. Resting-state functional brain connectivity for human mentalizing in multiple sclerosis. Soc Cogn Affect Neurosci. 2022;17(6):579–589.
  20. Sajjadi Mohammadabadi, S. M., Seyedkhamoushi, F., Mostafavi, M., & Borhani Peikani, M. (2024). Examination of AI's role in Diagnosis, Treatment, and Patient care. In Gupta, M., Kumar, R., & Lu, Z. (Eds.), Transforming Gender-Based Healthcare with AI and Machine Learning, 221–238. https://doi.org/10.1201/9781003473435-13
  21. Golkarieh, A., Razmara, P., Lagzian, A., Dolatabadi, A., & Mousavirad, S. (2025). Semi-supervised GAN with hybrid regularization and evolutionary hyperparameter tuning for accurate melanoma detection. Scientific Reports, 15(1). https://doi.org/10.1038/s41598-025-17756-x.
  22. Fox MD, Raichle ME. Spontaneous fluctuations in brain activity observed with functional MRI. Nat Rev Neurosci. 2007;8(9):700–711.
  23. Raichle ME, MacLeod AM, Snyder AZ, et al. A default mode of brain function. Proc Natl Acad Sci U S A. 2001;98(2):676–682.
  24. van den Heuvel MP, Sporns O. Network hubs in the human brain. Trends Cogn Sci. 2013;17(12):683–696.
  25. Seifi, N., Ghoodjani, E., Majd, S. S., Maleki, A., & Khamoushi, S. (2025). Evaluation and prioritization of artificial intelligence integrated block chain factors in healthcare supply chain: A hybrid Decision Making Approach. Computer and Decision Making: An International Journal, 2, 374–405. https://doi.org/10.59543/comdem.v2i.11029
  26. Basirat, S., Raoufi, S., Bazmandeh, D., Khamoushi, S., & Entezami, M. (2025). Ranking of AI-Based Criteria in Health Tourism Using Fuzzy SWARA Method. Computer and Decision Making: An International Journal, 2, 530–545. https://doi.org/10.59543/comdem.v2i.13795
  27. Salehi, S., Talaei, A., & Naghshineh, R. (2025). Proxemic Behaviors in Same-Sex and Opposite-Sex Social Interactions: Applications in Urban and Interior Space Design. https://doi.org/10.21203/rs.3.rs-6499935/v1.
  28. Golkarieh, A., Sassani, M., Khamoushi, S., Entezami, M., & Basirat, S. (2026). Intracranial Hemorrhage Diagnosis Using Deep Learning: A Survey of Techniques, Frameworks, and Challenges. Computer and Decision Making: An International Journal, 3, 780–804. https://doi.org/10.59543/comdem.v3i.15543
  29. Power JD, Barnes KA, Snyder AZ, Schlaggar BL, Petersen SE. Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion. Neuroimage. 2012;59(3):2142–2154.
  30. Satterthwaite TD, et al. Impact of motion on functional connectivity: confounds and correction. Neuroimage. 2013.
  31. Jenkinson M, Smith S. A global optimisation method for robust affine registration of brain images. Med Image Anal. 2001;5(2):143–156.
  32. Jendrichovsky, P., Khosravi, S., Rupasinghe, A., Maximov, K., Guo, P., Babadi, B., & Kanold, P. (2025). Patchy harmonic functional connectivity of the mouse auditory cortex. Proceedings of the National Academy of Sciences, 122(27). https://doi.org/10.1073/pnas.2510012122
  33. Ahmadian, A. , Salehi, S. and Naghshineh, R. (2024). Recognize adaptive ecologies and their applications in architectural structures. Sharif Journal of Civil Engineering, 39(4), 101-109. https://doi.org/10.24200/j30.2023.61722.3192.
  34. Desikan RS, et al. An automated labeling system for subdividing the human cerebral cortex on MRI scans. Neuroimage. 2006;31(3):968–980.
  35. Tournier JD, et al. MRtrix3: a fast, flexible and open software framework for medical image processing and visualisation. Neuroimage. 2019;202:116137.
  36. Salehi, S., Naghshineh, R., & Ahmadian, A. (2025). Determine of Proxemic Distance Changes before and During COVID-19 Pandemic with Cognitive Science Approach. OSF Preprints. https://doi.org/10.31234/osf.io/y4p82_v2.
  37. Sumowski JF, et al. Cognitive reserve and cognitive impairment in multiple sclerosis. Neurology. 2013.
  38. Rocca MA, Filippi M. Functional MRI in multiple sclerosis: neuroplasticity, network reorganization, and clinical relevance. Nat Rev Neurol. 2017.
  39. Noori, A., Mohanty, S., & Manda, P. (2025). Semantic Analysis of SNOMED CT Concept Co-occurrences in Clinical Documentation using MIMIC-IV. arXiv:2509.03662. https://doi.org/10.48550/ARXIV.2509.03662
  40. DeLuca J, Chelune GJ, Tulsky DS, et al. Is speed of processing or working memory the primary information processing deficit in MS? J Clin Exp Neuropsychol. 2004.
  41. Calabrese M, et al. Thalamic damage and cognitive impairment in multiple sclerosis. Neurology. 2010.
  42. Benedict RHB, Cookfair D, Gavett R, et al. Validity of the Minimal Assessment of Cognitive Function in MS (MACFIMS). J Int Neuropsychol Soc. 2006.
Neurology Letters
Volume 5, Issue 1
February 2026
Pages 28-33

  • Receive Date 19 September 2025
  • Revise Date 02 November 2025
  • Accept Date 26 December 2025
  • Publish Date 01 February 2026