Anti-N-methyl-D-aspartate receptor (anti-NMDAR) encephalitis is an autoimmune disorder in which conventional MRI often appears normal, leading to clinical-radiologic dissociation and hindering early diagnosis and monitoring. We retrospectively studied five pediatric patients with anti-NMDAR encephalitis and compared their pretreatment diffusion MRI to age- and sex-matched controls. Using fixel-based analysis (FBA), we quantified tract-specific white matter abnormalities at the individual level. All patients showed significantly reduced fiber density and cross-section, with patterns ranging from focal to widespread involvement. In two patients, FBA abnormalities corresponded to seizure lateralization on EEG despite normal MRI, emphasizing FBA's added value in detecting seizure-concordant injury. One patient, exhibiting the highest disease burden and marked CSF pleocytosis, demonstrated extensive white matter tract disruption. These findings indicate that FBA derived from pretreatment MRI can identify clinically significant white matter injury at disease onset in pediatric anti-NMDAR encephalitis. Future large-scale studies should explore its potential as an individualized biomarker, which may offer diagnostic and prognostic value beyond conventional imaging.

Multiple Sclerosis negatively affects hand function in 60% of cases. Upper extremity dysfunction in persons with Multiple Sclerosis (PwMS) has previously been linked to slower, more variable movement, increased visual response delays (Tv), and neuroimaging evidence of altered brain activity; yet no one knows how these aspects relate to each other. This work combines clinical, kinematic, sensorimotor control, and neural imaging techniques to gain a more complete understanding of how upper extremity dysfunction arises in PwMS. Twenty PwMS and 20 Controls completed a reach and hold task with simultaneous electroencephalography recorded to determine if increased Tv in PwMS was associated with greater levels of kinematic and neurologic impairment. Compared to Controls, PwMS moved slower and with greater variability, had increased sensory response delays, and decreased neural responses in occipital and parietal regions. Increased Tv was correlated with the Nine Hole Peg Test (NHPT) performance for PwMS who had more impairment (assessed via NHPT) and showed varying effects on the neurologic response to the task. Specifically, Tv was correlated with temporal delays in parietal channels for PwMS with less impairment while Tv of more impaired PwMS was inversely correlated with peak amplitude across parietal, somatosensory, and frontal channels. It could be that PwMS with less impairment are able to compensate for increased processing delays by moving slower, while those with greater impairment are more limited in how compensation may occur.

Functional Near Infrared Spectroscopy (fNIRS) is becoming a popular metric to noninvasively identify brain activity through assessing changes in oxygenated (HbO) and deoxygenated (HbR) hemoglobin concentration. Several aspects of fNIRS imaging are appealing for clinical applications but the reproducibility of fNIRS signals has yet to be determined over several sessions. To address this, four participants completed at least ten sessions on separate days where they performed Motor and Visual tasks while fNIRS signals were measured from 102 channels spanning the entire head. Reproducibility was quantified as the percentage of significant task related activity occurring across sessions at the channel and source levels (anatomically specific and default head models) via Region of Interest (ROI) and vertex-wise analyses. To improve source localization, digitized optode positions from each session were used with the anatomy specific source localization. Individual differences in reproducibility were present yet task-related changes in HbO were significantly more reproducible over sessions than changes in HbR (F(1, 66) = 5.03, p<0.05). Increased shifts in optode position correlated with less spatial overlap across sessions for each participant. Further steps can be taken to increase the reliability of capturing brain activity with fNIRS by improving upon data acquisition and analysis techniques.

Functional connectivity (FC) analyses of intracranial EEG (iEEG) signals can potentially improve the mapping of epileptic networks in drug-resistant focal epilepsy. However, it remains unclear whether FC-based metrics provide additional value beyond established epilepsy biomarkers such as epileptic spikes and high-frequency oscillations (HFOs). Using interictal iEEG data from 26 patients, we estimated FC across eight frequency bands (4-290 Hz) using amplitude envelope correlation (AEC) and phase locking value (PLV). From the resulting FC-matrices, we estimated two graph metrics each to derive 32 FC-based features. We also extracted features related to spikes, HFOs, and power spectral densities (PSD). A trained support vector machine (SVM) classifier predicted seizure onset zones (SOZs) with an area under the ROC curve (AUC) of 0.91 for node-level 4-fold cross-validation (CV), 0.69 for patient-level 4-fold CV, and 0.73 for patient-level leave-one-out CV. Notably, gamma-band graph features from AECs outperformed spikes and HFOs in SOZ prediction when using an equivalent number of features. Our results strongly suggest that AEC-based features may provide more information about epileptogenicity compared to PLV-based features. Furthermore, machine learning provides a robust approach for identifying useful FC-based features and integrating information from putative biomarkers of epilepsy to better localize epileptogenic networks.

BACKGROUND:Infantile epileptic spasms syndrome (IESS) is a devastating developmental epileptic encephalopathy (DEE) and patients exhibit diffuse white matter alterations and structural remodeling. However, the correlation between these structural changes and brain network properties, or their effect on the efficacy of treatment outcomes in MRI non-lesional IESS patients is not clear. METHOD:This retrospective study was conducted on IESS patients using fixel-based analysis (FBA) of diffusion MRI and graph theory analysis of structural connectivity, involving 26 non-lesional IESS patients aged 2 to 12 months and 120 age-matched controls. We further examined the differences between antiseizure medication (ASM) responders and non-responders within the IESS cohort. FBA was performed across three age groups (2-5, 6-7, and 8-12 months) to evaluate white matter integrity at the micro- and macroscale using fiber density (FD), fiber cross-section (FC), and combined fiber density and cross-section (FDC). Graph theory analysis was used to assess global and local network properties. RESULTS:When compared to the control group, IESS patients exhibited significantly lower FD, FC, and FDC across major white matter tracts, including the corticospinal tract, corpus callosum, superior longitudinal fasciculus, optic radiations, and thalamic radiations (family-wise error-corrected, p < 0.05). Graph theory analysis revealed significant alterations in brain network properties, particularly in the age group of 2-5 months, where IESS patients exhibited a significantly lower mean clustering coefficient (p < 0.001, d = -0.74) and global efficiency (p = 0.001, d = -0.69. Small-world network analysis demonstrated a shift toward a more randomized network structure in IESS patients, particularly in the age group of 6-7 months (p = 0.001, d = -0.6182). In the secondary analysis, ASM treatment responders showed higher FD values in regions critical for seizure control, such as the hippocampus. Meanwhile, the ASM treatment non-responders exhibited increased FC in areas such as the pons and brainstem. Although subgroup differences did not achieve statistical significance, trends suggest that white matter integrity and network organization may influence treatment outcomes. CONCLUSION:The results highlight widespread changes in white matter integrity and network connectivity in non-lesional IESS patients, with preliminary evidence suggesting a relationship between structural brain differences and treatment responsiveness. These findings underscore the potential of advanced neuroimaging analyses to guide personalized interventions in IESS.

Methods to quantify cortical hyperexcitability are of enormous interest for mapping epileptic networks in patients with focal epilepsy. We hypothesize that, in the resting state, cortical hyperexcitability increases firing-rate correlations between neuronal populations within seizure onset zones (SOZs). This hypothesis predicts that in the gamma frequency band (40-200 Hz), amplitude envelope correlations (AECs), a relatively straightforward measure of functional connectivity, should be elevated within SOZs compared to other areas. To test this prediction, we analyzed archived samples of interictal electrocorticographic (ECoG) signals recorded from patients who became seizure-free after surgery targeting SOZs identified by multiday intracranial recordings. We show that in the gamma band, AECs between nodes within SOZs are markedly elevated relative to those elsewhere. AEC-based node strength, eigencentrality, and clustering coefficient are also robustly increased within the SOZ with maxima in the low-gamma band (permutation test Z-scores > 8) and yield moderate discriminability of the SOZ using ROC analysis (maximal mean AUC similar to 0.73). By contrast to AECs, phase locking values (PLVs), a measure of narrow-band phase coupling across sites, and PLV-based graph metrics discriminate the seizure onset nodes weakly. Our results suggest that gamma band AECs may provide a clinically useful marker of cortical hyperexcitability in focal epilepsy.

We examined the extent to which concussion impacts how implicit sensorimotor memories are used to compensate for changes in hand-held loads during goal-directed reaching. Recently concussed individuals performed computerized cognition tests and a robotic test of sensorimotor adaptation on three occasions: as soon as possible after injury; after clearance to return to activity; three months after injury. Non-concussed individuals (controls) were tested at inter-session intervals mimicking concussed group intervals. During robotic testing, subjects grasped the handle of a horizontal planar robot while reaching repeatedly to a target. The robot exerted spring-like forces that changed unpredictably between trials; this allowed us to estimate contributions of implicit sensorimotor memories to trial-by-trial performance by fitting a computational model to the time series of reach errors and robot forces. Symptom severity varied considerably within the concussed group at the first session. Computerized cognition tests revealed longer reaction times in the concussed group relative to control group in Session 1 only. Concussed subjects likewise had slower reaction time in the reaching task during the first but not later sessions. Computational modeling found abnormally high values of effective limb compliance in concussed individuals relative to the control group in the first session only, but did not find group differences in how sensorimotor memories contribute to reach adaptation in any session. Analysis of control group models identified a practice effect affecting the memory coefficients that may have masked initial effects of concussion on how implicit memories contribute to sensorimotor adaptation. Although a practice effect and a heterogeneous concussed cohort preclude strong conclusions, our findings suggest procedural improvements that may decrease the robotic test’s sensitivity to practice effects and increase its sensitivity to concussion-related changes in how implicit memories contribute to sensorimotor adaptation to unpredictable hand-held loads during reaching.### Competing Interest StatementThe authors have declared no competing interest.

Functional near infrared spectroscopy (fNIRS) and functional magnetic resonance imaging (fMRI) both measure the hemodynamic response, and so both imaging modalities are expected to have a strong correspondence in regions of cortex adjacent to the scalp. To assess whether fNIRS can be used clinically in a manner similar to fMRI, 22 healthy adult participants underwent same-day fMRI and whole-head fNIRS testing while they performed separate motor (finger tapping) and visual (flashing checkerboard) tasks. Analyses were conducted within and across subjects for each imaging approach, and regions of significant task-related activity were compared on the cortical surface. The spatial correspondence between fNIRS and fMRI detection of task-related activity was good in terms of true positive rate, with fNIRS overlap of up to 68% of the fMRI for analyses across subjects (group analysis) and an average overlap of up to 47.25% for individual analyses within subject. At the group level, the positive predictive value of fNIRS was 51% relative to fMRI. The positive predictive value for within subject analyses was lower (41.5%), reflecting the presence of significant fNIRS activity in regions without significant fMRI activity. This could reflect task-correlated sources of physiologic noise and/or differences in the sensitivity of fNIRS and fMRI measures to changes in separate (vs. combined) measures of oxy and de-oxyhemoglobin. The results suggest whole-head fNIRS as a noninvasive imaging modality with promising clinical utility for the functional assessment of brain activity in superficial regions of cortex physically adjacent to the skull.

This study describes temporal patterns of cortical activity during a simple finger movement in people with stroke to understand how temporal patterns of cortical activation and network connectivity align with prolonged muscle contraction at the end of a task. We investigated changes in the EEG temporal patterns in the beta band (13-26Hz) of people with chronic stroke (N = 10, 7F/3M) and controls (N = 10, 7F/3M), during and after a cued movement of the index finger. We quantified the change in beta band EEG power relative to baseline as activation at each electrode and the change in beta band task-based coherence (tbCoh) relative to baseline coherence as connectivity between EEG electrodes. Contrary to controls, finger tap cortical activity in the stroke group was spatially distributed bilaterally, and measurements from the post task period lacked a positive change in beta power relative to baseline, which has been described as event-related synchronization in controls. In addition, the stroke group exhibited no discernible reduction in tbCoh between the ipsilesional sensorimotor and frontal regions of the cortex during the post task period, which was a notable feature of tbCoh in controls. Our results suggest that divergent cortical activation patterns coupled with changes in connectivity between the sensorimotor and frontal cortices in the stroke group might explain clinical observations of prolonged muscle activation in people with stroke. This prolonged activation might be attributed to the combination of cortical reorganization and changes to sensory feedback post-stroke.

The purpose of this study was to characterize changes in cortical activity and connectivity in stroke survivors when vibration is applied to the wrist flexor tendons during a visuomotor tracking task. Data were collected from 10 chronic stroke participants and 10 neurologically-intact controls while tracking a target through a figure-8 pattern in the horizontal plane. Electroencephalography (EEG) was used to measure cortical activity (beta band desynchronization) and connectivity (beta band task-based coherence) with movement kinematics and performance error also being recorded during the task. All participants came into our lab on two separate days and performed three blocks (16 trials each, 48 total trials) of tracking, with the middle block including vibration or sham applied at the wrist flexor tendons. The order of the sessions (Vibe vs. Sham) was counterbalanced across participants to prevent ordering effects. During the Sham session, cortical activity increased as the tracking task progressed (over blocks). This effect was reduced when vibration was applied to controls. In contrast, vibration increased cortical activity during the vibration period in participants with stroke. Cortical connectivity increased during vibration, with larger effect sizes in participants with stroke. Changes in tracking performance, standard deviation of hand speed, were observed in both control and stroke groups. Overall, EEG measures of brain activity and connectivity provided insight into effects of vibration on brain control of a visuomotor task. The increases in cortical activity and connectivity with vibration improved patterns of activity in people with stroke. These findings suggest that reactivation of normal cortical networks via tendon vibration may be useful during physical rehabilitation of stroke patients.

Abstract Background Despite performance improvements in active lower limb prostheses, there remains a need for control techniques that incorporate direct user intent (e.g., myoelectric control) to limit the physical and cognitive demands and provide continuous, natural gait across terrains. Methods The ability of a nonlinear autoregressive neural network with exogenous inputs (NARX) to continuously predict future (up to 142 ms ahead of time) ankle angle and moment of three transtibial amputees was examined across ambulation conditions (level overground walking, stair ascent, and stair descent) and terrain transitions. Within-socket residual EMG of the prosthetic side, in conjunction with sound-limb shank velocity, were used as inputs to the single-network NARX model to predict sound-limb ankle dynamics. By overlaying the ankle dynamics of the sound limb onto the prosthesis, the approach is a step forward to establish a more normal gait by creating symmetric gait patterns. The NARX model was trained and tested as a closed-loop network (model predictions fed back as recurrent inputs, rather than error-free targets) to ensure accuracy and stability when implemented in a feedback control system. Results Ankle angle and moment predictions of amputee models were accurate across ambulation conditions and terrain transitions with root-mean-square errors (RMSE) less than 3.7 degrees and 0.22 Nm/kg, respectively, and cross-correlations (R2) greater than 0.89 and 0.93, respectively, for predictions 58 ms ahead of time. The closed-loop NARX model had similar performance when characterizing normal ranges of ankle dynamics across able-bodied participants (n = 6; RMSEθ < 2.7°, R2θ > 0.95, RMSEM < 0.11 Nm/kg, R2M > 0.98 for predictions 58 ms ahead of time). Model performance was stable across a range of different EMG profiles, leveraging both EMG and shank velocity inputs for the prediction of ankle dynamics across ambulation conditions. Conclusions The use of natural, yet altered in amputees, muscle activity with information about limb state, coupled with the closed-loop predictive design, could provide intuitive user-driven and robust control by counteracting delays and proactively modifying gait in response to observed changes in terrain. The model takes an important step toward continuous real-time feedback control of active ankle-foot prostheses and robotic devices.

Background: Meaningful integration of functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) requires knowing whether these measurements reflect the activity of the same neural sources, i.e., estimating the degree of coupling and decoupling between the neuroimaging modalities. New method: This paper proposes a method to quantify the coupling and decoupling of fMRI and EEG signals based on the mixing matrix produced by joint independent component analysis (jICA). The method is termed fMRI/EEG-jICA. Results: fMRI and EEG acquired during a syllable detection task with variable syllable presentation rates (0.25-3 Hz) were separated with jICA into two spatiotemporally distinct components, a primary component that increased nonlinearly in amplitude with syllable presentation rate, putatively reflecting an obligatory auditory response, and a secondary component that declined nonlinearly with syllable presentation rate, putatively reflecting an auditory attention orienting response. The two EEG subcomponents were of similar amplitude, but the secondary fMRI subcomponent was ten folds smaller than the primary one. Comparison to existing method: FMRI multiple regression analysis yielded a map more consistent with the primary than secondary fMRI subcomponent of jICA, as determined by a greater area under the curve (0.5 versus 0.38) in a sensitivity and specificity analysis of spatial overlap. Conclusion: fMRI/EEG-jICA revealed spatiotemporally distinct brain networks with greater sensitivity than fMRI multiple regression analysis, demonstrating how this method can be used for leveraging EEG signals to inform the detection and functional characterization of fMRI signals. fMRI/EEG-jICA may be useful for studying neurovascular coupling at a macro-level, e.g., in neurovascular disorders.

Background People with multiple sclerosis (PwMS) have balance deficits while ambulating through environments that contain moving objects or visual manipulations to perceived self-motion. However, their ability to parse object from self-movement has not been explored. The purpose of this research was to examine the effect of medial–lateral oscillations of the visual field and of objects within the scene on gait in PwMS and healthy age-matched controls using virtual reality (VR). Methods Fourteen PwMS (mean age 49 ± 11 years, functional gait assessment score of 27.8 ± 1.8, and Berg Balance scale score 54.7 ± 1.5) and eleven healthy controls (mean age: 53 ± 12 years) participated in this study. Dynamic balance control was assessed while participants walked on a treadmill at a self-selected speed while wearing a VR headset that projected an immersive forest scene. Visual conditions consisted of (1) no visual manipulations (speed-matched anterior/posterior optical flow), (2) 0.175 m mediolateral translational oscillations of the scene that consisted of low pairing (0.1 and 0.31 Hz) or (3) high pairing (0.15 and 0.465 Hz) frequencies, (4) 5 degree medial–lateral rotational oscillations of virtual trees at a low frequency pairing (0.1 and 0.31 Hz), and (5) a combination of the tree and scene movements in (3) and (4). Results We found that both PwMS and controls exhibited greater instability and visuomotor entrainment to simulated mediolateral translation of the visual field (scene) during treadmill walking. This was demonstrated by significant (p < 0.05) increases in mean step width and variability and center of mass sway. Visuomotor entrainment was demonstrated by high coherence between center of mass sway and visual motion (magnitude square coherence = ~ 0.5 to 0.8). Only PwMS exhibited significantly greater instability (higher step width variability and center of mass sway) when objects moved within the scene (i.e., swaying trees). Conclusion Results suggest the presence of visual motion processing errors in PwMS that reduced dynamic stability. Specifically, object motion (via tree sway) was not effectively parsed from the observer’s self-motion. Identifying this distinction between visual object motion and self-motion detection in MS provides insight regarding stability control in environments with excessive external movement, such as those encountered in daily life.

We examined a key aspect of sensorimotor skill: the capability to correct performance errors that arise mid-movement. Participants grasped the handle of a robot that imposed a nominal viscous resistance to hand movement. They watched a target move pseudo-randomly just above the horizontal plane of hand motion and initiated quick interception movements when cued. On some trials, the robot's viscosity or the target's speed changed without warning coincident with the GO cue. We fit a sum-of-Gaussians model to mechanical power measured at the handle to determine the number, magnitude, and relative timing of submovements occurring in each interception attempt. When a single submovement successfully intercepted the target, capture times averaged 410 ms. Sometimes, two or more submovements were required. Initial error corrections typically occurred before feedback could indicate the target had been captured or missed. Error corrections occurred sooner after movement onset in response to mechanical viscosity increases (at 154 ms) than to unprovoked errors on control trials (215 ms). Corrections occurred later (272 ms) in response to viscosity decreases. The latency of corrections for target speed changes did not differ from those in control trials. Remarkably, these early error corrections accommodated the altered testing conditions; speed/viscosity increases elicited more vigorous corrections than in control trials with unprovoked errors; speed/viscosity decreases elicited less vigorous corrections. These results suggest that the brain monitors and predicts the outcome of evolving movements, rapidly infers causes of mid-movement errors, and plans and executes corrections—all within 300 ms of movement onset.

A hallmark of human locomotion is that it continuously adapts to changes in the environment and predictively adjusts to changes in the terrain, both of which are major challenges to lower limb amputees due to the limitations in prostheses and control algorithms. Here, the ability of a single-network nonlinear autoregressive model to continuously predict future ankle kinematics and kinetics simultaneously across ambulation conditions using lower limb surface electromyography (EMG) signals was examined. Ankle plantarflexor and dorsiflexor EMG from ten healthy young adults were mapped to normal ranges of ankle angle and ankle moment during level overground walking, stair ascent, and stair descent, including transitions between terrains (i.e., transitions to/from staircase). Prediction performance was characterized as a function of the time between current EMG/angle/moment inputs and future angle/moment model predictions (prediction interval), the number of past EMG/angle/moment input values over time (sampling window), and the number of units in the network hidden layer that minimized error between experimentally measured values (targets) and model predictions of ankle angle and moment. Ankle angle and moment predictions were robust across ambulation conditions with root mean squared errors less than 1° and 0.04 Nm/kg, respectively, and cross-correlations (R2) greater than 0.99 for prediction intervals of 58 ms. Model predictions at critical points of trip-related fall risk fell within the variability of the ankle angle and moment targets (Benjamini-Hochberg adjusted p > 0.065). EMG contribution to ankle angle and moment predictions occurred consistently across ambulation conditions and model outputs. EMG signals had the greatest impact on noncyclic regions of gait such as double limb support, transitions between terrains, and around plantarflexion and moment peaks. The use of natural muscle activation patterns to continuously predict variations in normal gait and the model's predictive capabilities to counteract electromechanical inherent delays suggest that this approach could provide robust and intuitive user-driven real-time control of a wide variety of lower limb robotic devices, including active powered ankle-foot prostheses.

Background: Multiple sclerosis (MS) is associated with an increased risk of falls, degeneration of sensory organization, and possible increased reliance on vision for balance control. Research question: The aim of this study was to assess differences in standing postural control between people with MS and age and sex matched controls during medial-lateral (ML) oscillations of the visual field, with and without blinders to the lower periphery. Methods: Ten persons with MS (mean age 54.0 +/- 5.3 years) and ten age and sex matched controls (mean age: 56.3 +/- 6.0 years) participated in this study. Balance control was assessed while participants stood in a Christie Cave system while wearing stereoscopic glasses that projected an immersive forest scene. Visual conditions consisted of 2 m ML visual oscillations of the scene at five frequencies (0.0, 0.3, 0.6, 0.7 and 0.8 Hz) with and without blinders to block the lower periphery. Results and significance: The results demonstrated that, in comparison to controls, participants with MS had a significantly larger center of pressure sway in both the ML and AP direction to ML visual oscillations. Additionally, participants with MS and controls both increased center of pressure frequency content to the visual oscillation frequency, while participants with MS also increased relative power at the visual oscillation frequency in the AP direction. Blinders of lower periphery reduced the percent power at the visual oscillation frequency in both groups and reduced overall sway in participants with MS during visual oscillations. Overall, results indicate that postural balance is sensitive to visual feedback in people with MS. The elicited AP sway to ML visual oscillation could reflect errors in visual processing for the control of balance, and decreased sway in response to blocking vision of the lower peripheral field could indicate an increased reliance on visual cues to maintain balance.

Continuous myoelectric prediction of intended limb dynamics has the ability to provide transparent control of a prosthesis by the user. However, the impact on these models of adding a human user into the control loop is less clear. Here, the ability of a User Response Model (URM) to continuously predict EMG activity from gait kinematics and kinetics collected during three mobility tasks (level-ground walking, stair ascent, and stair descent) was examined. Multiple-input, multiple-output NARX-based URMs were developed with two outputs (ankle plantarflexor and dorsiflexor) and variable inputs (ankle kinetics, and shank and/or ankle kinematics). Accuracy in predicting the tibialis anterior and medial gastrocnemius EMG was comparable across URMs regardless of the number of inputs. Stair descent had the lowest accuracy among the mobility tasks. No significant differences in normalized root-mean-square error and cross-correlation were found between URMs with five and nine inputs. A URM that continuously predicts EMG activity from gait kinetics and kinematics could be used to simulate human-in-the-loop myoelectric control of a transtibial prosthesis and examine the stability of the system to changes in the environment or due to control errors.