Взаимосвязь между моторными и когнитивными характеристиками и кортикальной активностью у пациентов с множественным склерозом: систематический обзор функциональных почти инфракрасных спектроскопических исследований.

Introduction

Multiple sclerosis (MS) is a neurodegenerative disease characterized by inflammation, demyelination, and axonal degeneration, which leads toa range of neurological deficits and significantly impairs patients' quality of life ( 1 ). Studies have shown that 80% of MS patients experience motor impairment, while 40%–70% present with concomitant cognitive impairment (e.g., declines in executive function and working memory), and the concurrent impairment of motor and cognitive coordination is is therefore considered one of the most representative comorbid features of MS patients ( 2 – 4 ). Further evidence suggests that motor and cognitive functions are interrelated or coupled in both aging and multiple sclerosis ( 5 ). Currently, approximately 2.8 million people worldwide are living with MS, with a prevalence of 35.9 cases per 100,000 population ( 6 ). As the high incidence of MS increasingly shifts toward the elderly population, the synergistic decline in these two domains is closely associated with disease severity, disability progression, and prognosis in MS patients, thereby posing significant challenges for the evaluation and treatment of this population ( 7 ).

Motor and cognitive performance is particularly critical for patients with MS, especially in daily life scenarios (e.g., walking while conversing). Daily activities typically involve the integration of cognitive and motor tasks, a process commonly referred to as dual-task processing ( 8 ). During this process, the functional specialization and network integration of the cerebral cortex collectively determine the level of motor and cognitive performance ( 9 ). Emerging evidence indicates that MS patients experience difficulties in cognitive-motor integration, which may be attributed to abnormal neurovascular coupling (NVC) and altered metabolic cortical activation, resulting in an inability to effectively allocate attentional resources. Consequently, ccompared with single-task conditions (motor or cognitive alone), MS patients exhibit declines in both motor performance (e.g., reduced gait speed, impaired balance) and cognitive performance under dual-task conditions ( 10 – 13 ). Therefore, elucidating the changes in motor/cognitive performance and their relationship with cortical activity in MS patients will help clarify the mechanisms underlying motor-cognitive impairment and inform the development of effective rehabilitation strategies for this population.

Traditional neuroimaging techniques (e.g., fMRI) are susceptible to motion artifacts, making it difficult to capture cortical activity in patients during real-world motor tasks ( 14 ). While portable electroencephalography (EEG) can monitor neural electrical activity in real time, it has limited spatial resolution and cannot accurately localize the key brain regions involved in motor planning, execution, and regulation ( 15 ). Based on the principles of NVC and optical spectroscopy, increased cerebral neural activity leads to elevated cerebral metabolic demand, resulting in decreased deoxygenated hemoglobin (HbR) levels and increased oxygenated hemoglobin (HbO 2 ) levels ( 16 , 17 ). Functional near-infrared spectroscopy (fNIRS) is a non-invasive, neuroimaging tool with moderate spatial resolution and favorable cost-effectiveness. It quantifies cerebral hemodynamic variations of oxy/deoxyhemoglobin to reflect brain activity and has been widely applied in functional neuroscience research ( 18 – 20 ). Previous studies have shown that fNIRS can be used to evaluate motor or cognitive performance in MS patients under single or dual-task conditions, with a particular focusing on monitoring activity in brain regions closely involved in motor planning, execution, and cognitive control. These regions include the prefrontal cortex (PFC), primary motor cortex (M1), premotor cortex (PMC), supplementary motor area (SMA), premotor area (PM), sensorimotor cortex (SMC), and frontopolar area (FPA) ( 21 , 22 ). Therefore, fNIRS provides an ideal tool for analyzing the neural mechanisms underlying motor-cognitive impairment in MS patients.

In conclusion, this systematic review aims to integrate existing evidence from fNIRS studies and address the following four core questions:(1) whether cortical activation differs between MS patients and healthy controls; (2) how activation patterns vary under single- vs. dual-task conditions; (3) whether activation is associated with motor/cognitive performance; and (4) the potential mechanisms underlying altered cortical activation.

Materials and methods

Study selection criteria

This systematic review was preregistered with PROSPERO (CRD420261402862) and followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses protocol. We established the inclusion criteria based on the PICOS framework. Studies meeting the following criteria were included: (1) Population: Adults aged ≥18 years with a confirmed diagnosis of MS; (2) Intervention: Any single/dual-task paradigm combining cognitive tasks or motor tasks (including single cognitive tasks, single motor tasks, or dual tasks) administered by researchers; (3) Comparison: Control groups performing single cognitive tasks, single motor tasks, or dual tasks combining cognitive and motor tasks; (4) Outcome: Studies that evaluated cerebral cortical hemodynamic responses during single or dual tasks using fNIRS; (5) Study design: Cross-sectional studies, randomized controlled trials (RCT), cohort studies, and pre-post studies, etc. Exclusion criteria: (1) Studies not employing fNIRS; (2) Non-original articles, such as reviews, editorials, commentaries, study protocols, conference abstracts, and letters; (3) Studies restricted to healthy individuals, patients with other diseases, or animals; (4) Studies using other cortical activation detection modalities, such as EEG, functional magnetic resonance imaging (fMRI), or transcranial magnetic stimulation (TMS); (5)Non-English article.

Search strategy

We conducted systematic searches in three electronic databases: PubMed, Web of Science, and Cochrane Library, with the search cutoff date set to December 31, 2025. The search strategy comprised all possible combinations of the following three groups of keywords and their corresponding Medical Subject Headings (MeSH) terms: (1) Multiple sclerosis-related terms: multiple sclerosis, MS, adults, older adults/elderly; (2) Task paradigm-related terms: dual-task, cognitive-motor performance, gait, walking, cognition, executive function, memory, motor function, mobility; (3) fNIRS-related terms: fNIRS, functional near-infrared spectroscopy, near-infrared spectroscopy. Manual supplementary searches were also conducted in Embase and Scopus.

Study eligibility was determined by screening the titles, abstracts, and full texts of the retrieved articles. 2 independent reviewers (WJW and SHJ) screened the titles, abstracts, and full texts to assess eligibility. Any discrepancies were resolved through discussion.

Data extraction

The following details were extracted from each included study: title, publication year, authors; characteristics of MS patients (sample size, age, disease subtype); fNIRS techniques used to measure brain activity (device type, wavelength, regions of interest, motion artifacts, extracerebral contamination, short-separation regression, preprocessing variability, physiological noise correction); task type and paradigm (i.e., single-task, dual-task); intervention duration, task performance, and cortical activation outcomes, including statistical correlations, gait characteristics, cognitive performance, etc.

Risk of bias in individual studies

2 independent reviewers (WJW and SHJ) assessed the quality of each included study using the research quality assessment tools from the US National Institutes of Health (NIH) ( 23 ). Cross-sectional studies were evaluated using the Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies, RCT were assessed with the Quality Assessment of Controlled Intervention Studies, and case-control studies were evaluated using the Quality Assessment Tool for Case-Control Studies. The quality assessment was performed to measure the strength of scientific evidence but was not used to determine study eligibility.

Results

Overview of included studies

As shown in Figure 1 , a total of 4,031 studies were identified through keyword searches. Among them, 531 studies were removed due to duplication, 3,240 studies were excluded during title and abstract screening, and 260 studies underwent full-text review. Of these 260 studies, 242 were excluded for failing to meet the above screening criteria. The remaining 18 studies were included in this review ( Figure 1 ).

Basic characteristics of included studies

A total of 572 patients with MS and 291 healthy controls (HC) were enrolled across the included studies. Among the MS patients, 140 had relapsing-remitting MS, 60 had progressive MS, and the remaining patients were not subtyped. The mean age of MS patients ranged from 29 to 69 years, while that of HC ranged from 30 to 75 years. Among all included studies, 1 was a RCT, 2 were case-control studies, and the rest were cross-sectional studies. Cognitive tasks included reciting alternate letters, arithmetic task, serial 7 ， s subtraction, memory tasks, etc. Motor tasks included normal walking, stepping over known/unknown obstacles, 6-minute normal walking on a treadmill, dishwashing tasks in virtual reality (VR) and real-world settings, assisted walking, and other related tasks ( Table 1 ). When investigating the activation patterns of MS patients across different tasks, different cortical regions were taken into consideration. The results indicated that the PFC was the most frequently studied area across the included studies ( n = 12). In contrast, only a small number of studies ( n = 6) examined hemodynamic responses in multiple cortical regions, including motor-related areas such as the SMA, PMC, FPA, SMC, primary motor cortex (MC), and the entire cerebral cortex ( Table 2 ).

Methodological quality of included studies

The methodological quality of included studies was assessed using the National Institutes of Health (NIH) Study Quality Assessment Tools. 12 studies were rated as “good” and 6 as “fair”. Overall, several limitations were identified across studies: some failed to report an adequate sample size, did not provide sufficient duration for study procedures, and did not control for confounding variables in their analyses. However, task types and paradigms were described in detail across all studies, and the tasks performed by participants were comparable ( Table 3 ).

Association between task performance and cortical activation under single-task conditions

Single cognitive tasks

3 studies reported cortical activation in patients with MS during single cognitive tasks. Compared with HC, MS patients exhibited higher PFC activation levels under single cognitive task conditions ( 24 ). A linear regression study found that reduced activation in the dorsolateral prefrontal cortex (DLPFC) was negatively correlated with increased levels of cognitive fatigue (CF) in MS patients, relative to healthy individuals ( 25 ). Within the MS patient population, higher disability severity was associated with stronger PFC activation ( 26 ).

Single motor tasks

5 studies demonstrated that during single walking tasks, oxygenation levels in the PFC, DLPFC, PMC, frontal pole cortex (FPC), and FPA were higher in patients with MS compared with healthy controls ( 24 , 27 – 30 ). 1 study also confirmed that PFC cortical activation levels were higher in MS patients than in healthy individuals, regardless of whether body weight support was provided ( 31 ). Another RCT using a treadmill walking task assessed the M1 and PM by calculating the hemispheric oxygenated hemoglobin area. At baseline, cortical activation was higher in MS patients than in HC. After 4 weeks of rehabilitation training, cortical activation decreased in the robot-assisted gait training (RAGT) group but increased in the observation group (OW) ( 32 ). Compared with forward walking conditions, activation in the DLPFC/FEF cortex of MS patients and the FPC/FEF cortex of HC was stronger during backward walking ( 33 ). Stratified analysis of MS patients based on disability status using the Expanded Disability Status Scale (EDSS, ≥4) and Patient Determined Disease Steps (PDSS, ≥3) revealed that PFC activation levels were higher in MS patients with severe disability ( 26 , 29 ).

Cortical activation under dual-task conditions

Dual-task vs. single motor tasks

Compared with single motor tasks, dual-task conditions further alter cortical activation levels, with distinct patterns observed between patients with MS and HC. 3 studies showed that during dual-task performance, HC exhibited greater activation in the PFC and its subregions (DLPFC/rPMC/FPC/FEF), as well as faster gait speed, compared with MS patients ( 10 , 24 , 33 ). 4 other studies indicated that MS patients generally walked slower and showed stronger PFC activation than HC under dual-task conditions ( 29 – 31 , 34 ). Some studies also reported no significant cortical activation changes in MS patients, despite poorer gait performance relative to healthy individuals ( 35 ).

Regarding MS disease subtypes, compared with relapsing-remitting MS patients, progressive MS patients exhibited reduced gait speed and increased PFC activation under dual-task conditions ( 36 , 37 ). For relapsing MS patients, however, greater activation in the PFC and SMA in response to reduced gait speed was associated with a lower risk of falls ( 10 , 36 ). In addition, MS patients with higher disability status showed stronger PFC cortical activation under dual-task conditions than those with lower disability status ( 26 ).

Dual-task vs. single cognitive tasks ( 24 )

Compared with single cognitive tasks, one study found that activation in the FPA increased as cognitive performance declined in healthy individuals ( 28 ). Most studies have shown that compared with HC, patients with MS generally exhibit higher PFC activation but poorer cognitive performance under dual-task conditions ( 10 , 24 , 29 , 34 ). 2 studies indicated that cortical activation remained stable in MS patients even when cognitive performance declined ( 28 , 35 ). However, 1 study also reported that as the complexity of dual cognitive tasks increased, PFC cortical activation decreased in MS patients compared with healthy controls ( 38 ).

2 studies showed that, compared with relapsing MS patients, progressive MS patients exhibited increased PFC activation but poorer cognitive performance (lower number of correct letters) under dual-task conditions ( 36 , 37 ). In addition, compared with single cognitive tasks, MS patients with high disability status showed higher PFC activation levels and poorer cognitive performance under dual-task conditions than those with low disability status ( 26 ).

Task load

Compared with single cognitive tasks, as the complexity of dual cognitive tasks increased, PFC activation decreased under dual-task conditions in patients with MS relative to HC ( 38 ). When walking while crossing predictable/unpredictable obstacles, MS patients showed higher activation in the PMC, MC, and somatosensory association cortex (SAC) compared with HC. During walking with unpredictable obstacles, MS patients exhibited increased attentional demands and cognitive load, accompanied by more widespread reductions in cortical activation ( 39 ). An assessment of attentional demands during dual tasks using partial body weight support to reduce individual burden in MS patients revealed that poorer cognitive performance was associated with enhanced PFC cortical activation, compared with conditions without body weight support ( 31 ).

Virtual reality vs. real-world tasks

1 upper limb motor study used a combined virtual and real-world approach to assess cortical activation. In both VR and real-world single motor tasks, patients with MS showed lower total hemoglobin concentration and weaker activation in the PMC and SMA compared with HC, indicating reduced neurovascular activation in MS patients. Under VR dual-task conditions, PMC and SMA activation in MS patients was higher than in healthy individuals. No significant differences in cortical activation were observed in MS patients across the four task conditions (VR and real-world single/dual tasks), suggesting impaired neurovascular adaptability in this population ( 40 ).

Effects of MS subtype and disability Status on cortical activation

Disease subtype in patients with MS profoundly affects the efficiency of neural compensation in the brain. Patients with relapsing-remitting MS compensate by increasing PFC activation to maintain gait stability and reduce fall risk. In contrast, patients with progressive MS exhibit increased cortical activation but poorer motor and cognitive performance, indicating excessive recruitment of additional brain resources, impaired neural compensation, and reduced plasticity. Concurrently, MS patients show a consistent pattern across single- and dual-task conditions: higher disability severity is associated with greater cortical activation. This suggests that individuals with low disability can flexibly modulate brain activation to adapt to task demands, reflecting preserved neural regulation capacity. In contrast, patients with high disability must engage more neural resources to meet complex task conditions, resulting in stronger cortical activation but poorer performance, indicative of reduced neural efficiency. Notably, MS disease subtype and disability status often coexist. For patients with progressive MS and severe disability, cortical compensation imbalance may be particularly pronounced, potentially manifesting as overload compensation or decompensation.

Effects of cognitive fatigue and task load on cortical activation

Cognitive fatigue disrupts PFC hemodynamic responses and functional connectivity, leading to impaired neurovascular coupling. As task load increases (e.g., dual tasks, unpredictable obstacle avoidance, backward walking) and individual fatigue levels rise, cortical activation patterns in patients with MS become further disorganized. Some MS patients even exhibit reduced cortical activation levels, indicating a decline in neural compensatory mechanisms.

Effects of interventions on cortical activation

Compared with conventional overground training, robot-assisted gait training can significantly reduce excessive activation of the motor cortex and improve neural resource utilization efficiency in MS patients. Partial body weight support effectively reduces cognitive load during dual tasks, alleviating neural compensatory pressure in patients with high disability. VR and augmented reality (AR) induce similar hemodynamic responses, yet they prominently reveal impaired neurovascular adaptability in MS patients.

Discussion

This is the systematic review to analyze cortical hemodynamic characteristics in patients with MS under single/dual-task conditions using fNIRS. Overall, MS patients showed a tendency toward increased cortical activation across multiple measured brain regions (e.g., FPC, PMC, SMA, SMC) under both task conditions, with greater activation typically observed during dual tasks. However, these patterns were not consistent across all studies: some reported stronger activation in HC, or no significant changes in activation despite MS poorer performance, or reduced activation in specific regions as task complexity increased. Notably, this apparent hyperactivation occurred alongside poorer motor and cognitive performance, suggesting potential compensatory mechanisms. The heterogeneity in findings, related to cortical regions, task characteristics, population diversity, disability status, and disease stage, aligns with previous studies ( 41 ).

Possible explanations for cortical activation

The cortical activation patterns observed in this review may be closely related to the connectivity, efficiency, and interactions between cortical networks. In patients with MS, inflammatory demyelination, axonal damage, and cortical lesions in the central nervous system disrupt structural connectivity between the motor network, cognitive control network, and salience network, reducing the efficiency of information transfer across brain networks ( 42 ). Based on the compensatory redistribution model, the brain recruits additional cortical resources for functional reorganization to maintain normal behavioral performance ( 43 ). However, this reorganization is not uniform but shows clear task-specific characteristics. For example, in studies involving walking combined with reciting or mental arithmetic, activation in the prefrontal cortex (especially the DLPFC and FPC) is more prominent, as this region is involved in multiple functions such as working memory, motor planning, decision-making, and the integration of cognition and movement ( 44 ). In studies involving spatial judgment tasks, the PMC, SMA, and SMC show activation advantages, as these regions are involved in sensorimotor integration, primarily responsible for motor planning, selection, and integration of sensory input and motor output ( 45 , 46 ). This difference in activation patterns reflects the dynamic allocation of brain resources based on task demands, representing interactions between cortical regions. These findings are consistent with previous studies of neurological disorders using fNIRS and fMRI, which reported significant PFC activation in older adults and healthy young individuals during obstacle walking, serial subtraction, and letter generation tasks ( 22 , 47 ). Additionally, MS patients exhibit extensive white matter tract damage and thinning of the prefrontal cortex and superior frontal gyrus (SFG) ( 48 ). During walking tasks, the brain requires additional cognitive resources, such as integrating multisensory information, implementing purposeful movements, and adjusting posture. Despite the diffuse damage in the prefrontal regions of MS patients, compensatory recruitment is maintained to support adequate task performance. However, as task complexity increases, the demand for brain resources may exceed available capacity, ultimately leading to functional decompensation, manifested as increased cortical activation but deteriorated behavioral performance ( 49 ).

Under normal conditions, the human body performs automated motor control via the M1, cerebellar motor areas, and spinal circuits, requiring only minimal activation of the PFC to complete tasks without sustained attentional monitoring or executive control ( 50 ). In patients with MS, however, these pathways for automated control are damaged. Consequently, indirect pathways projecting from the PFC and SMA to the basal ganglia, thalamus, and midbrain motor regions become activated, forcing reliance on top-down cognitive control by higher-order cortical areas. This results in increased cortical activation levels ( 51 ). Under single-task conditions, this compensation can partially support motor and cognitive performance, yet the increased activation indicates greater neural resource consumption and reduced information processing efficiency, suggesting that the patients' cerebral functional reserve has been excessively recruited ( 52 ). Under dual-task conditions, cortical activation is further augmented. For MS patients, impaired automatic control and compensatory reliance on executive control may lead to excessive competition for limited executive resources, resulting in declines in both motor and cognitive performance—a phenomenon widely recognized as a marker of reduced neural efficiency ( 53 , 54 ).

Elevated attentional load is a key factor exacerbating abnormal cortical activation. The inability to effectively allocate attentional resources to motor and cognitive tasks, regardless of whether patients have MS or other neurological conditions, is a critical cause of reduced motor and cognitive performance during dual tasks ( 55 – 57 ). Posner et al. conceptualized attention as an anatomical network whose primary function is to modulate the operation of other brain networks ( 58 ). In MS patients, damage to the attentional network leads to overload in attentional demands and impaired allocation processing during dual tasks, which often results in increased competition and information processing interference in the brain. This forces the cortex to increase both the intensity and scope of activation, yet task performance remains uncompensated ( 59 ). This aligns with previous studies showing that both patients with neurological disorders and healthy individuals exhibit poorer cognitive and walking performance when attentional resources are divided between multiple tasks under dual-task conditions ( 60 , 61 ). Thus, impaired attentional capacity or insufficient resource allocation may act synergistically with reduced neural efficiency, manifesting as the dual-task “cost” characterized by high cortical activation accompanied by low behavioral performance ( 62 , 63 ).

Notably, changes in cortical activation in MS patients must be interpreted in the context of NVC theory. NVC refers to the dynamic relationship between neuronal activity and cerebral blood flow, where increased neuronal activity leads to a corresponding rise in local cerebral blood flow to meet metabolic demand, a process reflected as an increase in oxygenated hemoglobin signals in fNIRS ( 64 ). Under dual-task conditions, MS patients must simultaneously process motor and cognitive information, significantly increasing cerebral metabolic demand. To cope with this load, activation in the PFC (particularly the DLPFC) is enhanced to support postural control and working memory load, consistent with previous findings ( 65 , 66 ). Furthermore, dynamic changes in neuroplasticity and functional reorganization influence the compensatory efficiency of NVC. In the early stages of MS, neuroplasticity can maintain basic function by recruiting additional cortical resources. However, as the disease progresses, this compensatory capacity gradually diminishes, and patients can only compensate for functional loss by further increasing brain activation, ultimately forming a maladaptive activation pattern ( 67 – 69 ). Previous studies have shown that older adults with mild disability and those with progressive multiple sclerosis often recruit more neural resources even during relatively simple tasks ( 70 , 71 ). In addition, chronic inflammatory responses and demyelinating lesions in MS can directly damage vascular endothelial function, leading to abnormal platelet adhesion, reduced cerebral vascular autoregulation, and further disrupting the matching between neural activity and blood flow ( 72 , 73 ). Under the high metabolic demand of dual tasks, the brain cannot effectively regulate blood flow supply to meet neuronal activity requirements, exacerbating cognitive and motor dysfunction ( 74 ). In summary, the combined effects of dynamic neuroplastic changes, inflammatory responses, and vascular dysfunction disrupt neurovascular coupling balance in MS patients, resulting in the paradoxical phenomenon of abnormal cortical activation alongside functional decline under dual-task conditions.

Clinical implications

The findings of this study provide guidance for disease assessment and rehabilitation interventions in patients with MS. First, cortical activation levels and oxygenated hemoglobin metabolic characteristics measured by fNIRS can be used as objective evaluation indicators. Combined with activation differences between single- and dual-task conditions, these data can help assess patients' neural functional status, identify the risk of cerebral compensation overload, and assist in evaluating disease severity and rehabilitation potential. Second, considering the characteristics of reduced neural efficiency and poor dual-task load tolerance in MS patients, a stepped dual-task rehabilitation training program can be implemented to gradually optimize cortical activation patterns, improve neuroplasticity, and enhance task performance.

Limitations

This review has several limitations. First, most included studies had small sample sizes and were cross-sectional in design. Large-scale cohort or longitudinal studies are needed to improve the generalizability of the conclusions. Second, participants' ages ranged from approximately 30 to 75 years, with diverse task types and varying disease progression, which may have contributed to result heterogeneity. Third, confounding factors were not well controlled in the included studies, potentially affecting both cortical activation levels and behavioral performance. Additionally, the fNIRS methodology used in this research has inherent limitations, particularly prominent in dynamic scenarios such as walking and dual tasks. Walking activities are prone to introducing motion artifacts, and physiological signals can be easily confused with cortical activation signals. The limited detection depth of fNIRS devices only allows coverage of superficial cortical layers. Furthermore, calibration standards for short-distance reference channels are inconsistent across studies, affecting the comparability of results. Finally, this study employed only a single neuroimaging modality, making it difficult to simultaneously integrate neural electrical activity information, which imposes limitations on the interpretation of brain function.

Future research perspectives

Advances in multimodal neuroimaging techniques have made the combined use of EEG and fNIRS an important direction in mobile neuroimaging research. fNIRS and EEG exhibit significant complementarity in such studies: fNIRS provides good spatial resolution, reflects cortical hemodynamic changes, and is more tolerant to motion artifacts, making it particularly suitable for dynamic tasks such as walking ( 75 ). In contrast, EEG captures neural electrical activity with millisecond-level temporal resolution, revealing the dynamic temporal characteristics of neural activity ( 76 ). Combined multimodal neuroimaging can simultaneously achieve high temporal and spatial resolution monitoring of neural activity, offering a more comprehensive perspective for understanding the relationship between neural activity and hemodynamic coupling in MS patients during mobile scenarios. Meanwhile, building on the limitations of the current study, future research can advance in several directions: First, Conduct large-sample, multi-center prospective data studies using unified task paradigms and data processing standards. Second, Implement study grouping with strict control of confounding factors such as disease subtype, disability severity, and cognitive status. Third, Actively carry out interventional trials to explore the effects of rehabilitation training and clinical medications on improving cortical activation patterns and behavioral abilities, thereby tapping into the application value of fNIRS in therapeutic efficacy evaluation.

Conclusions

Certain MS patients demonstrate increased task-related cortical activation accompanied by reduced performance. While this may reflect alterations in neural efficiency or neurovascular coupling, it could also indicate compensatory recruitment or increased attentional effort, depending on the task, disability severity, and disease-related factors. fNIRS may serve as a useful monitoring tool, and future multimodal EEG-fNIRS research is warranted to clarify the underlying mechanisms.

Acknowledgments

We genuinely thank the authors of the primary study who provided the data and each of the researchers who participated in this study.

Edited by: Markey Cierra Olson , Barrow Neurological Institute (BNI), United States

Reviewed by: Marc Sebastián-Romagosa , BarcelonaBeta Brain Research Center, Spain

Béatrice Degraeve , Lille Catholic University, France

Data availability statement

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.

Author contributions

HS: Writing – original draft, Data curation. HX: Supervision, Writing – original draft. KH: Writing – review & editing. ZW: Writing – review & editing, Supervision. JW: Writing – review & editing.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Correction Note

A correction has been made to this article. Details can be found at: 10.3389/fresc.2026.1941006 .

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

1. Ghasemi N Razavi S Nikzad E . Multiple sclerosis: pathogenesis, symptoms, diagnoses and cell-based therapy . Cell J . ( 2017 ) 19 ( 1 ): 1 – 10 . 10.22074/cellj.2016.4867 28367411 PMC5241505

2. Comber L Galvin R Coote S . Gait deficits in people with multiple sclerosis: a systematic review and meta-analysis . Gait Posture . ( 2017 ) 51 : 25 – 35 . 10.1016/j.gaitpost.2016.09.026 27693958

3. Hansen S Muenssinger J Kronhofmann S Lautenbacher S Oschmann P Keune PM . Cognitive screening in multiple sclerosis: the five-point test as a substitute for the PASAT in measuring executive function . Clin Neuropsychol . ( 2017 ) 31 ( 1 ): 179 – 92 . 10.1080/13854046.2016.1241894 27707354

4. Li S Zhang Q Zheng S Li G Li S He L . Efficacy of non-invasive brain stimulation on cognitive and motor functions in multiple sclerosis: a systematic review and meta-analysis . Front Neurol . ( 2023 ) 14 : 1091252 . 10.3389/fneur.2023.1091252 36779055 PMC9911042

5. Demnitz N Esser P Dawes H Valkanova V Johansen-Berg H Ebmeier KP . A systematic review and meta-analysis of cross-sectional studies examining the relationship between mobility and cognition in healthy older adults . Gait Posture . ( 2016 ) 50 : 164 – 74 . 10.1016/j.gaitpost.2016.08.028 27621086 PMC5081060

6. Walton C King R Rechtman L Kaye W Leray E Marrie RA . Rising prevalence of multiple sclerosis worldwide: insights from the atlas of MS, third edition . Mult Scler . ( 2020 ) 26 ( 14 ): 1816 – 21 . 10.1177/1352458520970841 33174475 PMC7720355

7. Marrie RA Bernstein CN Peschken CA Hitchon CA Chen H Fransoo R . Intensive care unit admission in multiple sclerosis: increased incidence and increased mortality . Neurology . ( 2014 ) 82 ( 23 ): 2112 – 9 . 10.1212/WNL.0000000000000495 24808019 PMC4118502

8. Siegert V Köhler LJ Schreff L Hamacher D Oschmann P Rothhammer V . Evaluating the impact of task-sequencing on cognitive and motor performance in MS: PASAT and 3-minute walk test order effects . Front Neurol . ( 2025 ) 16 : 1654656 . 10.3389/fneur.2025.1654656 41098379 PMC12520191

9. Gordon EM Chauvin RJ Van AN Rajesh A Nielsen A Newbold DJ . A somato-cognitive action network alternates with effector regions in motor cortex . Nature . ( 2023 ) 617 ( 7960 ): 351 – 9 . 10.1038/s41586-023-05964-2 37076628 PMC10172144

10. Saleh S Sandroff BM Vitiello T Owoeye O Hoxha A Hake P . The role of premotor areas in dual tasking in healthy controls and persons with multiple sclerosis: an fNIRS imaging study . Front Behav Neurosci . ( 2018 ) 12 : 296 . 10.3389/fnbeh.2018.00296 30618658 PMC6297844

11. Wajda DA Motl RW Sosnoff JJ . Dual task cost of walking is related to fall risk in persons with multiple sclerosis . J Neurol Sci . ( 2013 ) 335 ( 1-2 ): 160 – 3 . 10.1016/j.jns.2013.09.021 24090757

12. Leone C Patti F Feys P . Measuring the cost of cognitive-motor dual tasking during walking in multiple sclerosis . Mult Scler . ( 2015 ) 21 ( 2 ): 123 – 31 . 10.1177/1352458514547408 25178543

13. Allali G Laidet M Assal F Armand S Lalive PH . Walking while talking in patients with multiple sclerosis: the impact of specific cognitive loads . Neurophysiol Clin . ( 2014 ) 44 ( 1 ): 87 – 93 . 10.1016/j.neucli.2013.10.136 24502909

14. Mehagnoul-Schipper DJ Van Der Kallen BFW Colier WNJM Van Der Sluijs MC Van Erning LJTO Thijssen HOM . Simultaneous measurements of cerebral oxygenation changes during brain activation by near-infrared spectroscopy and functional magnetic resonance imaging in healthy young and elderly subjects . Hum Brain Mapp . ( 2002 ) 16 ( 1 ): 14 – 23 . 10.1002/hbm.10026 11870923 PMC6871837

15. Uludağ K Roebroeck A . General overview on the merits of multimodal neuroimaging data fusion . Neuroimage . ( 2014 ) 102 ( Pt 1 ): 3 – 10 . 10.1016/j.neuroimage.2014.05.018 24845622

16. Liao L-D Tsytsarev V Delgado-Martínez I Li M-L Erzurumlu R Vipin A . Neurovascular coupling: in vivo optical techniques for functional brain imaging . Biomed Eng Online . ( 2013 ) 12 : 38 . 10.1186/1475-925X-12-38 23631798 PMC3655834

17. Villringer A Chance B . Non-invasive optical spectroscopy and imaging of human brain function . Trends Neurosci . ( 1997 ) 20 ( 10 ): 435 – 42 . 10.1016/S0166-2236(97)01132-6 9347608

18. Leff DR Orihuela-Espina F Elwell CE Athanasiou T Delpy DT Darzi AW . Assessment of the cerebral cortex during motor task behaviours in adults: a systematic review of functional near infrared spectroscopy (fNIRS)studies . Neuroimage . ( 2011 ) 54 ( 4 ): 2922 – 36 . 10.1016/j.neuroimage.2010.10.058 21029781

19. Scholkmann F Kleiser S Metz AJ Zimmermann R Mata Pavia J Wolf U . A review on continuous wave functional near-infrared spectroscopy and imaging instrumentation and methodology . Neuroimage . ( 2014 ) 85 ( Pt 1 ): 6 – 27 . 10.1016/j.neuroimage.2013.05.004 23684868

20. Scarapicchia V Brown C Mayo C Gawryluk JR . Functional magnetic resonance imaging and functional near-infrared spectroscopy: insights from combined recording studies . Front Hum Neurosci . ( 2017 ) 11 : 419 . 10.3389/fnhum.2017.00419 28867998 PMC5563305

21. Wu Y Dong Y Tang Y Wang W Bo Y Zhang C . Relationship between motor performance and cortical activity of older neurological disorder patients with dyskinesia using fNIRS: a systematic review . Front Physiol . ( 2023 ) 14 : 1153469 . 10.3389/fphys.2023.1153469 37051020 PMC10083370

22. Bishnoi A Holtzer R Hernandez ME . Brain activation changes while walking in adults with and without neurological disease: systematic review and meta-analysis of functional near-infrared spectroscopy studies . Brain Sci . ( 2021 ) 11 ( 3 ): 291 . 10.3390/brainsci11030291 33652706 PMC7996848

23. National Institutes of Health . Study Quality Assessment Tools ( 2014 ). Available online at: https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools (Accessed July 6, 2026) .

24. Hernandez ME O’Donnell E Chaparro G Holtzer R Izzetoglu M Sandroff BM . Brain activation changes during balance-and attention-demanding tasks in middle-and older-aged adults with multiple sclerosis . Motor Control . ( 2019 ) 23 ( 4 ): 498 – 517 . 10.1123/mc.2018-0044 30987505

25. Borragán G Gilson M Atas A Slama H Lysandropoulos A De Schepper M . Cognitive fatigue, sleep and cortical activity in multiple sclerosis disease. A behavioral, polysomnographic and functional near-infrared spectroscopy investigation . Front Hum Neurosci . ( 2018 ) 12 : 378 . 10.3389/fnhum.2018.00378 30294266 PMC6158319

26. Hernandez ME Motl RW Foley FW Picone MA Izzetoglu M Lipton ML . Disability moderates dual task walking performance and neural efficiency in older adults with multiple sclerosis . Neurorehabil Neural Repair . ( 2024 ) 38 ( 11–12 ): 795 – 807 . 10.1177/15459683241273411 39177188 PMC11970354

27. Broscheid K-C Hamacher D Lamprecht J Sailer M Schega L . Inter-session reliability of functional near-infrared spectroscopy at the prefrontal Cortex while walking in multiple sclerosis . Brain Sci . ( 2020 ) 10 ( 9 ). 10.3390/brainsci10090643 32957682 PMC7565127

28. Santinelli FB Veldkamp R Vitório R Kos D Vos M Nijssen R . Hemodynamics of the frontopolar and dorsolateral pre-frontal Cortex in people with multiple sclerosis during walking, cognitive subtraction, and cognitive-motor dual-task . Neurorehabil Neural Repair . ( 2024 ) 38 ( 11-12 ): 820 – 31 . 10.1177/15459683241279066 39256995

29. Hernandez ME Holtzer R Chaparro G Jean K Balto JM Sandroff BM . Brain activation changes during locomotion in middle-aged to older adults with multiple sclerosis . J Neurol Sci . ( 2016 ) 370 : 277 – 83 . 10.1016/j.jns.2016.10.002 27772776

30. de Aratanha MA Balardin JB Cardoso do Amaral C Lacerda SS Sowmy TAS Huppert TJ . The use of functional near infrared spectroscopy and gait analysis to characterize cognitive and motor processing in early-stage patients with multiple sclerosis . Front Neurol . ( 2022 ) 13 : 937231 . 10.3389/fneur.2022.937231 36105774 PMC9464830

31. Chaparro G Balto JM Sandroff BM Holtzer R Izzetoglu M Motl RW . Frontal brain activation changes due to dual-tasking under partial body weight support conditions in older adults with multiple sclerosis . J Neuroeng Rehabil . ( 2017 ) 14 ( 1 ): 65 . 10.1186/s12984-017-0280-8 28662727 PMC5493004

32. Lamberti N Manfredini F Baroni A Crepaldi A Lavezzi S Basaglia N . Motor cortical activation assessment in progressive multiple sclerosis patients enrolled in gait rehabilitation: a secondary analysis of the RAGTIME trial assisted by functional near-infrared spectroscopy . Diagnostics (Basel) . ( 2021 ) 11 ( 6 ): 1068 . 10.3390/diagnostics11061068 34207923 PMC8227480

33. Kupchenko Y Dreyer-Alster S broscheid K-C Kalron A . Prefrontal hemodynamics during forward and backward walking, with and without a cognitive task, in people with multiple sclerosis . Eur J Phys Rehabil Med . ( 2023 ) 59 ( 2 ): 164 – 73 . 10.23736/S1973-9087.23.07812-7 36940180 PMC10171363

34. Holtzer R Choi J Motl RW Foley FW Wagshul ME Hernandez ME . Brain control of dual-task walking can be improved in aging and neurological disease . Geroscience . ( 2024 ) 46 ( 3 ): 3169 – 84 . 10.1007/s11357-023-01054-3 38221528 PMC11009168

35. Broscheid K-C Behrens M Dettmers C Jöbges M Schega L . Effects of a 6-min treadmill walking test on dual-task gait performance and prefrontal hemodynamics in people with multiple sclerosis . Front Neurol . ( 2022 ) 13 : 822952 . 10.3389/fneur.2022.822952 35463151 PMC9022001

36. Holtzer R Foley FW Motl RW Wagshul ME Hernandez ME Lipton ML . Brain hemodynamic responses and fall prediction in older adults with multiple sclerosis . Mult Scler . ( 2024 ) 30 ( 13 ): 1664 – 73 . 10.1177/13524585241277400 39258434 PMC12145860

37. Hernandez ME Motl RW Foley FW Izzetoglu M Wagshul M Holtzer R . Comparison of practice-related changes in dual task walking performance and neural efficiency between older adults with progressive and relapsing-remitting multiple sclerosis . Mult Scler Relat Disord . ( 2025 ) 93 : 106224 . 10.1016/j.msard.2024.106224 39693703 PMC12035826

38. Baldasso BD Raza SZ Islam SS Burry IB Newell CJ Hillier SR . Disrupted hemodynamic response within dorsolateral prefrontal cortex during cognitive tasks among people with multiple sclerosis-related fatigue . PLoS One . ( 2024 ) 19 ( 6 ): e0303211 . 10.1371/journal.pone.0303211 38837991 PMC11152297

39. Al-Shargie F Glassen M DeLuca J Saleh S . Obstacle avoidance in healthy adults and people with multiple sclerosis: preliminary fNIRS study . IEEE Trans Neural Syst Rehabil Eng . ( 2024 ) 32 : 3966 – 76 . 10.1109/TNSRE.2024.3487526 39466870

40. Lavi R Pau M Dreyer-Alster S Drori T Cocco E Kalron A . Real-time cortical activity during virtual reality practice in people with multiple sclerosis: a pilot fNIRS study . J Neuroeng Rehabil . ( 2025 ) 22 ( 1 ): 148 . 10.1186/s12984-025-01681-1 40615921 PMC12226833

41. Hamacher D Herold F Wiegel P Hamacher D Schega L . Brain activity during walking: a systematic review . Neurosci Biobehav Rev . ( 2015 ) 57 : 310 – 27 . 10.1016/j.neubiorev.2015.08.002 26306029

42. Barbi C Pizzini FB Tamburin S Martini A Pedrinolla A Laginestra FG . Brain structural and functional alterations in multiple sclerosis-related fatigue: a systematic review . Neurol Int . ( 2022 ) 14 ( 2 ): 506 – 35 . 10.3390/neurolint14020042 35736623 PMC9228847

43. Verghese J Wang C Ayers E Izzetoglu M Holtzer R . Brain activation in high-functioning older adults and falls: prospective cohort study . Neurology . ( 2017 ) 88 ( 2 ): 191 – 7 . 10.1212/WNL.0000000000003421 27927937 PMC5224713

44. Szczepanski SM Knight RT . Insights into human behavior from lesions to the prefrontal cortex . Neuron . ( 2014 ) 83 ( 5 ): 1002 – 18 . 10.1016/j.neuron.2014.08.011 25175878 PMC4156912

45. Milner D . Cognitive neuroscience: the biology of the mind and findings and current opinion in cognitive neuroscience . Trends Cogn Sci . ( 1998 ) 2 ( 11 ): 463 . 10.1016/S1364-6613(98)01226-1 21227278

46. Zwergal A Linn J Xiong G Brandt T Strupp M Jahn K . Aging of human supraspinal locomotor and postural control in fMRI . Neurobiol Aging . ( 2012 ) 33 ( 6 ): 1073 – 84 . 10.1016/j.neurobiolaging.2010.09.022 21051105

47. Kazui H Kitagaki H Mori E . Cortical activation during retrieval of arithmetical facts and actual calculation: a functional magnetic resonance imaging study . Psychiatry Clin Neurosci . ( 2000 ) 54 ( 4 ): 479 – 85 . 10.1046/j.1440-1819.2000.00739.x 10997866

48. Lorefice L Coghe G Fenu G Porta M Pilloni G Frau J . Timed up and go'and brain atrophy: a preliminary MRI study to assess functional mobility performance in multiple sclerosis . J Neurol . ( 2017 ) 264 ( 11 ): 2201 – 4 . 10.1007/s00415-017-8612-y 28894919

49. Takakusaki K . Functional neuroanatomy for posture and gait control . J Mov Disord . ( 2017 ) 10 ( 1 ): 1 – 17 . 10.14802/jmd.16062 28122432 PMC5288669

50. Nielsen JB . How we walk: central control of muscle activity during human walking . Neuroscientist . ( 2003 ) 9 ( 3 ): 195 – 204 . 10.1177/1073858403009003012 15065815

51. Herold F Wiegel P Scholkmann F Thiers A Hamacher D Schega L . Functional near-infrared spectroscopy in movement science: a systematic review on cortical activity in postural and walking tasks . Neurophotonics . ( 2017 ) 4 ( 4 ): 041403 . 10.1117/1.NPh.4.4.041403 28924563 PMC5538329

52. Stuart S Mancini M . Prefrontal cortical activation with open and closed-loop tactile cueing when walking and turning in Parkinson disease: a pilot study . J Neurol Phys Ther . ( 2020 ) 44 ( 2 ): 121 – 31 . 10.1097/NPT.0000000000000286 31425309

53. Seidler RD Bernard JA Burutolu TB Fling BW Gordon MT Gwin JT . Motor control and aging: links to age-related brain structural, functional, and biochemical effects . Neurosci Biobehav Rev . ( 2010 ) 34 ( 5 ): 721 – 33 . 10.1016/j.neubiorev.2009.10.005 19850077 PMC2838968

54. Clark DJ . Automaticity of walking: functional significance, mechanisms, measurement and rehabilitation strategies . Front Hum Neurosci . ( 2015 ) 9 : 246 . 10.3389/fnhum.2015.00246 25999838 PMC4419715

55. Learmonth YC Sandroff BM Pilutti LA Klaren RE Ensari I Riskin BJ . Cognitive motor interference during walking in multiple sclerosis using an alternate-letter alphabet task . Arch Phys Med Rehabil . ( 2014 ) 95 ( 8 ): 1498 – 503 . 10.1016/j.apmr.2014.03.007 24681387

56. Sosnoff JJ Boes MK Sandroff BM Socie MJ Pula JH Motl RW . Walking and thinking in persons with multiple sclerosis who vary in disability . Arch Phys Med Rehabil . ( 2011 ) 92 ( 12 ): 2028 – 33 . 10.1016/j.apmr.2011.07.004 22133253

57. Hall CD Echt KV Wolf SL Rogers WA . Cognitive and motor mechanisms underlying older adults’ability to divide attention while walking . Phys Ther . ( 2011 ) 91 ( 7 ): 1039 – 50 . 10.2522/ptj.20100114 21527384

58. Posner MI Sheese BE Odludaş Y Tang YY . Analyzing and shaping human attentional networks . Neural Netw . ( 2006 ) 19 ( 9 ): 1422 – 9 . 10.1016/j.neunet.2006.08.004 17059879

59. Chen Q Hattori T Tomisato H Ohara M Hirata K Yokota T . Turning and multitask gait unmask gait disturbance in mild-to-moderate multiple sclerosis: underlying specific cortical thinning and connecting fibers damage . Hum Brain Mapp . ( 2023 ) 44 ( 3 ): 1193 – 208 . 10.1002/hbm.26151 36409700 PMC9875928

60. Yogev-Seligmann G Hausdorff JM Giladi N . The role of executive function and attention in gait . Mov Disord . ( 2008 ) 23 ( 3 ): 329 – 42 ; quiz 472 . 10.1002/mds.21720 18058946 PMC2535903

61. Lu C-F Liu Y-C Yang Y-R Wu Y-T Wang R-Y . Maintaining gait performance by cortical activation during dual-task interference: a functional near-infrared spectroscopy study . PLoS One . ( 2015 ) 10 ( 6 ): e0129390 . 10.1371/journal.pone.0129390 26079605 PMC4469417

62. Tripathi S Verghese J Blumen HM . Gray matter volume covariance networks associated with dual-task cost during walking-while-talking . Hum Brain Mapp . ( 2019 ) 40 ( 7 ): 2229 – 40 . 10.1002/hbm.24520 30664283 PMC6445705

63. Hamilton F Rochester L Paul L Rafferty D O'Leary CP Evans JJ . Walking and talking: an investigation of cognitive-motor dual tasking in multiple sclerosis . Mult Scler . ( 2009 ) 15 ( 10 ): 1215 – 27 . 10.1177/1352458509106712 19667011

64. Gagliano G Monteverdi A Casali S Laforenza U Gandini Wheeler-Kingshott CAM D’Angelo E . Non-linear frequency dependence of neurovascular coupling in the cerebellar cortex implies vasodilation-vasoconstriction competition . Cells . ( 2022 ) 11 ( 6 ). 10.3390/cells11061047

65. Hou L Yang J Xu L Peng J Joyce Law CY Chen T . Activation of brain regions associated with working memory and inhibitory control in patients with attention-deficit/hyperactivity disorder in functional near-infrared spectroscopy: a systematic review . Curr Med Imaging . ( 2023 ) 19 ( 8 ): 865 – 73 . 10.2174/1573405618666220822101019 35996256 PMC10291870

66. Specogna I Casagrande F Lorusso A Catalan M Gorian A Zugna L . Functional MRI during the execution of a motor task in patients with multiple sclerosis and fatigue . Radiol Med . ( 2012 ) 117 ( 8 ): 1398 – 407 . 10.1007/s11547-012-0845-3 22729506

67. Rocca MA Valsasina P Leavitt VM Rodegher M Radaelli M Riccitelli GC . Functional network connectivity abnormalities in multiple sclerosis: correlations with disability and cognitive impairment . Mult Scler . ( 2018 ) 24 ( 4 ): 459 – 71 . 10.1177/1352458517699875 28294693

68. Rocca MA Valsasina P Absinta M Riccitelli G Rodegher ME Misci P . Default-mode network dysfunction and cognitive impairment in progressive MS . Neurology . ( 2010 ) 74 ( 16 ): 1252 – 9 . 10.1212/WNL.0b013e3181d9ed91 20404306

69. Schoonheim MM Meijer KA Geurts JJ . Network collapse and cognitive impairment in multiple sclerosis . Front Neurol . ( 2015 ) 6 : 82 . 10.3389/fneur.2015.00082 25926813 PMC4396388

70. Lassmann H Van Horssen J Mahad D . Progressive multiple sclerosis: pathology and pathogenesis . Nat Rev Neurol . ( 2012 ) 8 ( 11 ): 647 – 56 . 10.1038/nrneurol.2012.168 23007702

71. Baek CY Kim HD Yoo DY Kang KY Lee JW . Change in activity patterns in the prefrontal cortex in different phases during the dual-task walking in older adults . J Neuroeng Rehabil . ( 2023 ) 20 ( 1 ): 86 . 10.1186/s12984-023-01211-x 37420235 PMC10327141

72. Mittal M Siddiqui MR Tran K Reddy SP Malik AB . Reactive oxygen species in inflammation and tissue injury . Antioxid Redox Signal . ( 2014 ) 20 ( 7 ): 1126 – 67 . 10.1089/ars.2012.5149 23991888 PMC3929010

73. Alexander JS Zivadinov R Maghzi A-H Ganta VC Harris MK Minagar A . Multiple sclerosis and cerebral endothelial dysfunction: mechanisms . Pathophysiology . ( 2011 ) 18 ( 1 ): 3 – 12 . 10.1016/j.pathophys.2010.04.002 20663648

74. Vestergaard MB Frederiksen JL Larsson HBW Cramer SP . Cerebrovascular reactivity and neurovascular coupling in multiple sclerosis-A systematic review . Front Neurol . ( 2022 ) 13 : 912828 . 10.3389/fneur.2022.912828 35720104 PMC9198441

75. Jiaqi F . Research on physiological signal analysis based on clinical statistics and machine learning; proceedings of the ProcSPIE, F, 2023[C] .

76. Shin J Kwon J Im CH . A ternary hybrid EEG-NIRS brain-computer interface for the classification of brain activation patterns during mental arithmetic, motor imagery, and idle state . Front Neuroinform . ( 2018 ) 12 : 5 . 10.3389/fninf.2018.00005 29527160 PMC5829061