Brain Communications

Lesion anatomy has been widely used to study post stroke cognitive outcomes, but it is unclear whether lesion-based measures provide clinically meaningful prognostic information beyond established predictors. Stroke survivors (n = 408) completed the Oxford Cognitive Screen (OCS) during acute hospitalisation and at chronic (6-month) follow-up. Lesion characteristics and structural disconnection profiles associated with chronic OCS scores were identified using ROI-level, voxel-level and structural network disconnection lesion mapping approaches. The incremental predictive value of these measures, relative to acute behaviour and pre-morbid brain health, was evaluated using regression analyses, receiver operating curve (ROC) and support vector regression (SVR) models predicting continuous chronic scores. Significant lesion and disconnection correlates of chronic cognitive impairment were identified for 9/10 OCS subtests. The extent of damage to these correlates was significantly associated with chronic cognitive scores, but their diagnostic utility for identifying persistent impairment was low under conventional thresholds (AUC mean = 0.59, range= 0.46-0.66). Acute cognitive task performance was the single best predictor of chronic cognition (AUC mean = 0.66, range = 0.4-0.95). In multivariate analyses, SVR models trained on acute cognitive performance and regional atrophy severity scores both outperformed models trained on lesion anatomy or structural disconnection across most cognitive domains. SVR models combining anatomical, disconnection and behavioural predictors did not improve predictions accuracy relative to behaviour or atrophy-only models. Together, these findings demonstrate that statistically significant lesion-outcome relationships do not necessarily translate into clinically useful prognostic indicators. In a large, clinically representative stroke cohort, detailed lesion-based measures provided limited incremental prognostic value beyond acute cognitive assessment and coarse brain health markers. These results highlight the importance of explicitly evaluating predictive utility when developing prognostic models for post-stroke cognitive outcomes.

BackgroundDespite the central role of the temporal lobes in conceptual processing, the impact of temporal lobe epilepsy (TLE) on semantic knowledge remains unclear. ObjectivesThis systematic review and meta-analysis aimed to investigate semantic functioning in TLE and the impact that seizure lateralisation and surgical intervention have on semantic outcomes. MethodologyA comprehensive literature search was conducted using Medline, Embase, and PsychINFO. Studies were eligible if they included participants aged [≥]18 years with TLE and evaluated semantic functioning. Risk of bias was assessed using the Newcastle-Ottawa scale, and a narrative synthesis summarised the review findings. Meta-analyses compared semantic performance across individuals with TLE and healthy controls and left and right TLE. Results141 studies, encompassing 8,241 participants (TLE: n = 5,623, controls: n = 2,618), were included, reporting over 30 different semantic measures. Both narrative review and meta-analysis showed significantly poorer semantic performance in people with TLE compared to controls, with impairments in semantic fluency (g = -1.35), WAIS-IV Vocabulary (g = -1.08), and Camels and Cactus Test (g = -1.37), but not on Pyramids and Palm Trees (g = -0.44). Some lateralisation effects were evident, with verbal semantic impairments more prominent in left TLE. ConclusionsTLE is associated with a mild semantic impairment. While left-sided lesions are associated with worse verbal semantic impairment, lateralisation effects more broadly were mild and inconsistent. Our findings emphasise the need to conduct routine semantic assessments in TLE to support more precise cognitive deficit monitoring, better-informed surgical risk discussions, and the development of personalised rehabilitation plans. HighlightsTLE is commonly associated with a mild impairment in semantic functioning. LTLE is generally associated with relatively poorer performance on verbal-based semantic measures. However, both LTLE and RTLE tend to demonstrate impairment across both verbal- and visual-based semantic measures. In sum, semantic measures should be included in routine cognitive evaluations to support the clinical care of individuals with TLE.

Quantitative susceptibility mapping (QSM) is an advanced post-processing technique of magnetic resonance imaging data that can be leveraged as a surrogate marker of iron accumulation in the brain following mild traumatic brain injury (mTBI). However, subtle tissue content changes characteristic of this complex injury may be lost to group-wise averaging when standard statistical models are employed. To provide more clinically- and individually-relevant information, z-tests can be used to build personalised profiles of positive susceptibility as a marker of abnormal iron homeostasis. Here, we mapped subject-specific deviations in iron-related positive susceptibility across 10 bilateral segmentations of the hippocampal subfields and 15 basal nuclei. The healthy normal susceptibility distribution for each region-of-interest (ROI) was derived from the aggregate data of 25 age-matched male controls (M = 21.10 years [range: 16-32], SD = 4.35) using z-tests. Region-wise z-scores for each of the 35 males aged between 16 and 33 years (M = 21.60, SD = 4.98) with acute (< 14 days) sports-related mTBI (sr-mTBI) were compared against the healthy reference range. Of the sr-mTBI participants, 43% exhibited abnormal iron markers in at least one ROI, which involved the hippocampal subfields in a majority (87%) of cases. Across all ROIs, particularly dense concentrations were observed in the parasubiculum and mammillary nucleus. Injury severity scores were not significantly different between sr-mTBI participants with abnormal iron markers (M = 41.7, SD = 34.5) and those without (M = 35.6, SD = 30.8), p = 0.5, however, abnormal iron markers in certain hippocampal subfields and the mammillary nucleus were observationally linked to clinical symptom phenotype. Taken together, these data allude to a region-of-risk model in which areas of the anteromedial hippocampal head, which is proximal to the sphenoid ridge, and midline structures are vulnerable to iron-mediated pathology. These findings underscore the importance of subject-specific analyses and how these sensitive methods can be used to map regional iron dyshomeostasis against cranial-dural morphology and established injury biomechanics.

In recent years, the functional importance of the anterior temporal lobes (ATLs) has come to prominence in two active, albeit unconnected branches of the literature. In one branch, neuropsychology and functional neuroimaging evidence emphasises the role of the ATLs in face recognition and linking faces to biographical knowledge. In the other, cognitive and clinical neuroscience investigations have shown that the ATLs are critical to all forms of semantic memory. To draw these literatures together and generate a unified account of ATL function, we test the predictions arising from each literature and examine the effects of bilateral versus unilateral ATL damage on face recognition, person knowledge and semantic memory. Sixteen people with bilateral ATL atrophy from semantic dementia (SD), 17 people with unilateral ATL resection for temporal lobe epilepsy (TLE; left=10, right=7), and 14 controls completed a test battery encompassing general semantic processing, person knowledge and perceptual face matching. SD patients were severely impaired across all semantic tasks, including person knowledge. Despite commensurate total ATL damage, unilateral resection generated mild impairments, with minimal differences between left- and right-ATL resection. Face matching performance was largely preserved but slightly reduced in SD and right TLE. All groups displayed the classic familiarity effect in face matching; however, this benefit was reduced in SD and right TLE groups and was aligned with the level of item-specific semantic knowledge in all participants. We propose a unified neurocognitive framework whereby the ATLs underpin a resilient bilateral representation system that supports semantic memory, person knowledge and face recognition.

Despite vulnerability to microstructural tissue damage following mild traumatic brain injury (mTBI), key subcortical brain regions have been overlooked in quantitative susceptibility mapping (QSM) studies. Alterations to tissue composition in the functionally and structurally distinct hippocampal subfields and basal ganglia regions may reflect distinct symptomatology, and better characterisation of these regions is needed to increase our understanding of mTBI pathophysiology. To address this issue, we analysed differences in positive and negative QSM values between 25 males with acute (< 14 days) sports-related mTBI (sr-mTBI) and 25 age-matched male controls across 10 hippocampal subfields and 16 basal nuclei. Additional variables of interest including age, injury severity, and days since injury at time of the magnetic resonance imaging (MRI) scan were also correlated with both positive and negative susceptibility values. Primary analyses indicated no significant difference in positive susceptibility values between sr-mTBI participants and controls for hippocampal and basal ganglia ROIs. For negative sign values, susceptibility was significantly less negative for sr-mTBI participants in the cornu ammonis 4 (CA4) region only (pFDR = 0.04). In line with the known linear relationship between iron deposition and age in deep grey matter, particularly within the first three decades of life, significant positive relationships were observed between net positive susceptibility and age in the putamen, caudate, red nucleus, parabrachial pigmented nucleus, and ventral pallidum (pFDR < 0.05). Positive relationships were also observed between absolute negative susceptibility values and age in the hippocampal fimbria, caudate, and extended amygdala (pFDR < 0.05), suggesting age-related calcifications in these regions. A negative relationship was observed between absolute negative values and age in the ventral pallidum (pFDR = 0.04), indicating potential changes to myelin content in this region. No significant associations were observed between any other variable and signed susceptibility values. The results of this study contribute to, and extend, prior literature regarding the temporal kinetics of biomagnetic substrates as a function of ageing. Decreased negative susceptibility after mTBI in the CA4 region also suggests potential injury-related effects on myelin content or neuron loss; a particularly interesting finding in light of the well-established vulnerability of cell populations in this region and susceptibility to pathology in chronic traumatic encephalopathy (CTE). The lack of other significant between-group differences suggest that alterations to tissue content may not be quantifiable at the acute stage of injury in subcortical ROIs or may be masked by age-related tissue susceptibility changes as a common feature across all participants in this young cohort. Future research should consider the use of longitudinal study designs to mitigate the influence of these factors.

BackgroundDistinguishing epilepsy from functional/dissociative seizures (FDS) is an ongoing diagnostic challenge. Misdiagnosis delays appropriate treatment and puts patients at significant risk. Quantitative analyses of clinical EEG offer a potential avenue for developing decision-support tools in the diagnosis of seizure disorders. Recent work using univariate features demonstrated that reliably identifying diagnostic traits in the presence of confounding factors remains challenging. However, diagnostic information might be available in multivariate features such as network-based measures. Using a well-controlled dataset, we run the first diagnostic accuracy study assessing the potential of multivariate resting-state EEG markers to directly discriminate between a diagnosis of epilepsy and one of FDS at the time when a diagnosis is suspected and prior to treatment initiation. MethodsThe dataset, previously examined in a published study, includes 148 age- and sex-matched individuals with suspected seizure disorders who were later diagnosed with non-lesional epilepsy (n=75) or FDS (n=73). Eyes-closed, resting-state EEG data used for the analyses were normal on visual inspection, and acquired while participants were medication-free. Functional network measures in the 6-9 Hz range were extracted and machine learning implemented to assess their predictive potential; different model configurations (including varying model types, dimensionality reduction methods, and approaches to enhance feature stability) were tested to identify the most promising approach for future translational implementations. ResultsNetwork measures derived from resting-state EEG discriminate between conditions at levels significantly above chance (maximum balanced accuracy: 67.5%). Their sensitivity to epilepsy (81.8%) is consistently higher than their sensitivity to FDS (53.3%). A systematic assessment of model choices indicates that improving the temporal stability of network features through epoch-wise averaging improves classification accuracy (62.6% to 67.5%). Multiple nonlinear model types succeed on the classification problem, with the three-best performing assigning a consistent diagnostic label to 77.5% of the individuals; however, model choice remains a strong determinant of overall classification accuracy. Dimensionality reduction did not provide a significant advantage in our models. ConclusionWe establish evidence for the clinical validity of selected network-based markers to discriminate between a diagnosis of non-lesional epilepsy and FDS prior to treatment initiation, highlighting the measures potential to support post-test probability estimation in the clinic. Our models, configured to optimise balanced accuracy, classified people with epilepsy more accurately than people with FDS, indicating that these measures are specific to epilepsy and should not be interpreted as markers of a positive diagnosis of FDS.

Parkinsons disease (PD) is associated with changes in neural activity in the sensorimotor alpha and beta bands. Using magnetoencephalography (MEG), we investigated the role of spontaneous neuronal activity within the somatosensory cortex in a large cohort of early-to mid-stage PD patients (N = 78) and age- and sex matched healthy controls (N = 60) using source reconstructed resting-state MEG. We quantified features of the time series data in terms of oscillatory alpha power, beta power, and 1/f broadband characteristics using power spectral density, and also characterised transient beta burst events in the time-domain signals. We examined the relationship between these signal features and the patients disease state, symptom severity, age, sex, and cortical thickness. PD patients and healthy controls differed on PSD broadband characteristics, with PD patients showing a steeper 1/f exponential slope and higher 1/f offset. PD patients further showed a steeper age-related decrease in the burst rate. Out of all the signal features of the sensorimotor activity, only burst rate was associated with increased severity of bradykinesia. Our study shows that general non-oscillatory features (broadband PSD slope and offset) of the sensorimotor signals are related to disease state and oscillatory burst rate scales with symptom severity in PD.

Recovery from aphasia after stroke has been hypothesized to depend on neuroplasticity in surviving brain regions. Many studies have investigated this process, but progress has been impeded by methodological limitations relating to task performance confounds, contrast validity, and sample sizes. Furthermore, few studies have accounted for the complex relationships that exist between patterns of structural damage, distributed networks of functional activity, and behavioral outcomes. The present cross-sectional study aimed to overcome these critical methodological limitations and to disentangle the relationships between structure, function, and behavior. We recruited 70 individuals with post-stroke aphasia and 45 neurologically normal comparison participants. We used a valid and reliable language mapping fMRI paradigm that adapted dynamically to each participants task performance, and carried out whole-brain permutation analyses along with hypothesis-driven analyses of individually defined functional regions of interest (ROIs). Multivariable models were constructed that incorporated lesion load estimates derived from machine learning and language activations across multiple brain regions. We found strong evidence that left posterior temporal cortex is the most critical region for language processing in post-stroke aphasia: functional activity in this region was reduced in aphasia, predictive of aphasia outcomes in a whole-brain analysis above and beyond the contribution of lesion load, and remained predictive even above and beyond other functional predictors, with a medium effect size (f2 = 0.15). We also found that right posterior temporal cortex made an independent contribution to aphasia outcomes: functional activity was attenuated in aphasia, suggesting diaschisis, yet was predictive of aphasia outcomes above and beyond left hemisphere lesion load and functional predictors, with a small effect size (f2 = 0.08). We corroborated the importance of left frontal cortex: functional activity was attenuated in aphasia and predictive of aphasia outcomes over and beyond the contribution of lesion load; however, unlike in the bilateral temporal regions, functional activity in the left frontal lobe did not remain predictive once other functional predictors were included in the model. There was no support for other potential compensatory mechanisms such as recruitment of the right frontal lobe, the bilateral multiple demand network, or perilesional regions. Taken together, our findings demonstrate that functional imaging can provide critical insights into language processing in aphasia that cannot be obtained from structural imaging alone, with the left and right posterior temporal cortices making independent contributions to aphasia outcomes after stroke.

The Disconnection Symptom Discoverer (DSD) model proposes to predict long-term performance on neuropsychological tests from stroke lesion disconnection profiles. The model requires external validation to determine reproducibility and generalizability to new and different patients. Here, we investigated whether the DSD supports accurate multi-domain cognitive outcome predictions at three different timepoints post stroke, in a clinically representative independent cohort. In this study, the DSD was used to predict visuospatial attention, verbal memory, and language scores in an independent cohort of 74 stroke survivors (mean age = 69.2, 39% female) with 3 repeated cognitive assessments. DSD-predicted scores were compared to observed neuropsychological scores collected at <2 weeks, six months, and > 2 years post-stroke. DSD-predicted language outcomes were significantly correlated with observed behaviour at the <2 weeks timepoint, but no other significant correlations between DSD-predicted scores were identified. Importantly, DSD-predicted verbal memory and visuospatial domain scores were not significantly correlated with observed behaviour at any of the considered timepoints (minimum p-value = 0.33). Across all tests and timepoints, DSD-predicted scores had an average Mean Absolute Error (MAE) of 0.21 (SD = 0.13, range = 0.04-0.43), with the highest errors occurring between predicted and observed memory scores. Larger stroke lesions were associated with higher MAE, indicating that the DSD performance was modulated by stroke severity. Overall, these results indicate that the DSD did not yield informative predictions of long-term cognitive outcomes in this external dataset. This finding provides an important illustration of potential overfitting issues within cognitive outcome prediction models, highlighting the need for caution when aiming to predict long-term post-stroke cognitive outcomes and further external validation of proposed models.

Post-stroke cognitive impairment (PSCI) is highly heterogeneous, reflecting both domain-specific deficits associated with focal lesions and broader impairments linked to premorbid health and demographic factors. In the current study we investigated whether PSCI can be distinguished into distinct across-domain cognitive profiles and how such PSCI profiles are associated with lesion neuroanatomy, demographic and premorbid health factors. This cross-sectional study employed data-driven analyses (i.e., Latent Class Analysis) on 2172 stroke survivors who completed a domain-specific cognitive screen (i.e., Oxford Cognitive Screen) within 6 months after stroke (M = 19; Mdn = 7, SD = 35.7 days). In addition, the association of the PSCI profiles with lesion and demographic characteristics was investigated. We identified two viable cognitive class solutions: a 5-class model capturing classical and novel PSCI profiles and a more detailed 13-class model reflecting a broader set of distinct cognitive profiles that commonly occur after stroke. The 5-class solution distinguished classical lateralized deficits (e.g., aphasia, neglect) alongside a minimal impairment and non-lateralized global impairment profile. In contrast, the 13-class solution provided finer-grained differentiation, particularly for non-lateralized cognitive profiles which were more strongly associated with premorbid health and education level. Importantly, lesion anatomy alone could not fully account for class distinctions. While lesion location was predictive, particularly, in hyper-acute stages, profiles for patients tested 2 weeks post-stroke revealed less influence of lesion location and more of lesion volume. These findings underscore the importance of considering both anatomical and premorbid factors in understanding PSCI. By providing a nuanced classification of PSCI profiles, this study establishes a foundation for future translational research aimed at improving clinical care and predicting cognitive trajectories.

Background The field of implantable Brain-Computer Interfaces (iBCIs) is rapidly advancing, with individuals with amyotrophic lateral sclerosis (ALS) as key beneficiaries. However, ALS-related cortical degeneration may impair iBCI effectiveness. This study investigated whether structural magnetic resonance imaging (MRI) and functional MRI (fMRI) metrics are associated with the quality of electrocorticography (ECoG) signals critical for iBCI use. Methods Six late-stage ALS participants and 76 controls underwent T1-weighted structural MRI and task-based fMRI during right-hand movement or attempts thereof. ECoG data of ALS participants was benchmarked using ECoG data acquired in epilepsy patients. Grey matter thickness in the sensorimotor cortex and fMRI activation in the motor-hand area were measured. Results Four ALS participants showed >0.4 mm thinning in the precentral gyrus, while the postcentral gyrus was spared. ECoG signal quality was significantly associated with precentral grey matter thickness, but not with fMRI activity. Conclusions These findings suggest that presurgical assessment of precentral grey matter thickness could potentially prove useful for iBCI candidate selection in advanced ALS.

Lesion network mapping (LNM) is increasingly used to link focal brain lesions to distributed functional networks. Recent work has raised concerns that LNM results may be spatially biased by dominant features of the normative connectome. If this were the case, three testable predictions would follow: (i) a consistent spatial pattern of false positives across LNM studies, (ii) that this pattern can be consistently explained by intrinsic connectome organization, and (iii) that symptom-associated LNM findings preferentially occur in regions with high spatial bias. We tested these predictions across three independent LNM datasets (n = 49/101/200), evaluating each prediction in all cohorts. Spatial bias maps derived from 4,000,000 random permutations under the null hypothesis showed minimal correspondence across cohorts (R2 = 0.4-0.8%), indicating strong cohort specificity. Moreover, dominant connectome features--captured by the first 10 principal components of connectivity profiles from 1,000 atlas regions--did not systematically explain these bias maps. Finally, symptom-associated results showed no enrichment in high-bias regions. Together, these findings provide strong evidence that spatial bias in LNM is not driven by dominant connectome features. With appropriate inferential statistics and rigorous study design, LNM remains a valid approach for mapping symptom-related brain networks.

Slowed responding to sensory inputs presented in contralesional space is pervasive following unilateral cerebral stroke, but the causal neurophysiological pathway by which this occurs remains unclear. To this end, here we leverage a perceptual decision-making framework to disambiguate information processing stages between sensation and action in 30 unilateral stroke patients (18 right hemisphere, 12 left hemisphere) and 27 neurologically healthy adults. By recording neural activity using electroencephalography (EEG) during task performance, we show that the relationship between strokes in either hemisphere and slowed contralesional response times is sequentially mediated by weaker target selection signals in the contralateral hemisphere (the N2c ERP), and subsequently delayed evidence accumulation signals (the centroparietal positivity). Notably, asymmetries in CPP and response times across hemispheres are associated with everyday functioning. Together, these data suggest a plausible neurophysiological pathway by which post-stroke contralesional slowing arises and highlight the utility of neurophysiological assessments for tracking clinically relevant behaviour.

ObjectiveDirect cortical stimulation (DCS) is the gold standard for language mapping during SEEG but is prone to false negatives and false positives that may contribute to post-operative dysphasia or else overly conservative resections. Task-induced high-frequency activity (HFA, 30-200Hz) is an emerging functional biomarker that may augment DCS, but its clinical utility remains uncertain. We aimed to quantify HFAs diagnostic concordance with DCS, assessing its potential as both a surrogate marker and a screening tool. MethodsIn this single-centre prospective study, 23 adults undergoing SEEG completed language mapping with DCS and HFA. HFA was mapped using auditory and visual naming tasks (ANT/VNT), quantified via Morlet wavelet transforms with baseline-normalised z-scores. DCS-positive channels were those where 50Hz stimulation elicited language disruption. HFA distribution was examined independently of DCS. HFA-DCS concordance was assessed for individual and combined (ANT+VNT; maximal HFA across tasks) conditions at channel and sublobar levels across two thresholds: a specificity-optimized stringent threshold (Z>0.8) to examine HFA as a surrogate for DCS, and a sensitivity-optimized permissive threshold (Z>0.3) to evaluate its potential as a screening tool. ResultsTwelve (52%) participants were female, and 17 (74%) were MRI-negative. HFA patterns differed by task: VNT produced greater HFA magnitude in the dominant frontal lobe (p=0.0498), while ANT produced greater magnitude and activation rate in the non-dominant temporal lobe (p=0.015; p=0.0189), highest in the non-dominant superior temporal gyrus. In the combined condition, concordance with DCS was low at the stringent threshold (channel-wise sensitivity/specificity=0.24/0.88; region-wise=0.43/0.77). Sensitivity improved at the permissive threshold (channel-wise 0.56, NPV=0.96), with region-wise sensitivity of 0.75, specificity=0.45, and NPV=0.94. SignificanceRegion-level HFA at a permissive threshold is useful for identifying language-negative regions and prioritising DCS testing. Poor concordance at a stringent threshold suggests HFA and DCS index distinct functional properties and are not interchangeable. Anatomically plausible HFA localisation supports the need for further multimodal validation to clarify its role in presurgical mapping. Key PointsO_LIHFA and DCS show threshold- and scale-dependent diagnostic concordance for language mapping during SEEG C_LIO_LISensitivity-optimized sublobar HFA shows high negative predictive value and moderate sensitivity for DCS-positive language sites C_LIO_LIThese metrics support sublobar HFA as a screening tool to exclude non-eloquent regions and streamline DCS language mapping C_LIO_LISpecificity-optimized HFA concords poorly with DCS, indicating these markers index distinct properties and are not interchangeable C_LIO_LICombined HFA/DCS profiles may help stratify surgical risk: HFA-/DCS-regions as low risk, while HFA+/DCS+ sites denote high risk C_LI

RationaleApathy is a common symptom in many neurological and psychiatric conditions, associated with poor prognosis and increased caregiver burden. There are currently no proven treatments, in part due to a lack of mechanistic understanding. We have proposed a novel framework for apathy, based on the failure of active inference due to imprecise priors on the outcome of actions. The loss of precision on action outcomes causes apathy by reducing the expected difference between the state of the world following action versus non-action. Here we test the hypothesis that the loss of prior precision on action outcomes is reversible and mediated neuronally by GABAergic gain on the superficial pyramidal neurons in a prefrontal-motor decision-making hierarchy. We test this in a healthy cohort and people with two syndromes associated with frontotemporal lobar degeneration. MethodsTwenty healthy controls, twenty people with behavioural variant frontotemporal dementia (bvFTD) and twenty people with progressive supranuclear palsy (PSP) took part in a randomised placebo-controlled double-blind trial using zolpidem, an established GABAA agonist. This study was registered with ISRCTN registry (ISRCTN10616794). We use the Goal Prior Assay task and dynamic causal modelling of MEG resting-state data to explore the cognitive and neural concomitants of prior precision, and the effect of GABAergic regulation on both prior precision and superficial pyramidal gain. Apathy was primarily assessed with the Apathy Evaluation Scale (Self and Carer). Principal analyses were conducted using Bayesian statistics, supplemented by classical frequentist tests. ResultsForty-three participants (20 controls and 23 patients) were included in the final analysis. We found strong evidence for a difference in measures of apathy (B>1000, p<0.001) and prior precision (B=20.4, p<0.01) between healthy controls and people with bvFTD and PSP. This difference in prior precision was not found in the drug condition (B=0.86, p=0.11). There was strong evidence of a correlation between apathy and prior precision across groups (B>100, p<0.001). Dynamic causal modelling of MEG resting-state data confirmed reductions in gain on the prefrontal superficial pyramidal neurons in patients. The prefrontal superficial pyramidal gain was partially restored on zolpidem and linked to participants prior precision on their action outcomes. ConclusionWe confirm that apathy in conditions associated with frontotemporal lobar degeneration is underwritten by a reduction in prior precision on action outcomes, mediated by reduced synaptic gain of prefrontal superficial pyramidal neurons. GABAergic regulation using zolpidem partially restores prior precision by acting on this gain and reinstating neuronal message-passing within the prefrontal-motor decision-making hierarchy.

Recent advances in the treatment of acute ischemic stroke contribute to improved patient outcomes, yet the mechanisms driving long-term disease trajectory are not well-understood. Current trends in the literature emphasize the distributed disruptive impact of stroke lesions on brain network organization. While most studies use population-derived data to investigate lesion interference on healthy tissue, the potential for individualized treatment strategies remains underexplored due to a lack of availability and effective utilization of the necessary clinical imaging data. To validate the potential for individualized patient evaluation, we explored and compared the differential information in network models based on normative and individual data. We further present our novel deep learning approach providing usable and accurate estimates of individual stroke impact utilizing minimal imaging data, thus bridging the data gap hindering individualized treatment planning. We created normative and individual disconnectomes for each of 78 patients (mean age 65.1 years, 32 females) from two independent cohort studies. MRI data and Barthel Index, as a measure of activities of daily living, were collected in the acute and early sub-acute phase after stroke (baseline) and at three months post stroke incident. Disconnectomes were subsequently described using 12 network metrics, including clustering coefficient and transitivity. Metrics were first compared between disconnectomes and further utilized as features in a classifier to predict a patients disease trajectory, as defined by three months Barthel Index. We then developed a deep learning architecture based on graph convolution and trained it to predict properties of the individual disconnectomes from the normative disconnectomes. Both disconnectomes showed statistically significant differences in topology and predictive power. Normative disconnectomes included a statistically significant larger number of connections (N=604 for normative versus N=210 for individual) and agreement between network properties ranged from r2=0.01 for clustering coefficient to r2=0.8 for assortativity, highlighting the impact of disconnectome choice on subsequent analysis. To predict patient deficit severity, individual data achieved an AUC score of 0.94 compared to an AUC score of 0.85 for normative based features. Our deep learning estimates showed high correlation with individual features (mean r2=0.94) and a comparable performance with an AUC score of 0.93. We were able to show how normative data-based analysis of stroke disconnections provides limited information regarding patient recovery. In contrast, individual data provided higher prognostic precision. We presented a novel approach to curb the need for individual data while retaining most of the differential information encoding individual patient disease trajectory.

Chronic motor impairments are a leading cause of disability after stroke. Previous studies have predicted motor outcomes based on the degree of damage to predefined structures in the motor system, such as the corticospinal tract. However, such theory-based approaches may not take full advantage of the information contained in clinical imaging data. The present study uses data-driven approaches to predict chronic motor outcomes after stroke and compares the accuracy of these predictions to previously-identified theory-based biomarkers. Using a cross-validation framework, regression models were trained using lesion masks and motor outcomes data from 789 stroke patients (293 female/496 male) from the ENIGMA Stroke Recovery Working Group (age 64.9{+/-}18.0 years; time since stroke 12.2{+/-}0.2 months; normalised motor score 0.7{+/-}0.5 (range [0,1]). The out-of-sample prediction accuracy of two theory-based biomarkers was assessed: lesion load of the corticospinal tract, and lesion load of multiple descending motor tracts. These theory-based prediction accuracies were compared to the prediction accuracy from three data-driven biomarkers: lesion load of lesion-behaviour maps, lesion load of structural networks associated with lesion-behaviour maps, and measures of regional structural disconnection. In general, data-driven biomarkers had better prediction accuracy - as measured by higher explained variance in chronic motor outcomes - than theory-based biomarkers. Data-driven models of regional structural disconnection performed the best of all models tested (R2 = 0.210, p < 0.001), performing significantly better than predictions using the theory-based biomarkers of lesion load of the corticospinal tract (R2 = 0.132, p< 0.001) and of multiple descending motor tracts (R2 = 0.180, p < 0.001). They also performed slightly, but significantly, better than other data-driven biomarkers including lesion load of lesion-behaviour maps (R2 =0.200, p < 0.001) and lesion load of structural networks associated with lesion-behaviour maps (R2 =0.167, p < 0.001). Ensemble models - combining basic demographic variables like age, sex, and time since stroke - improved prediction accuracy for theory-based and data-driven biomarkers. Finally, combining both theory-based and data-driven biomarkers with demographic variables improved predictions, and the best ensemble model achieved R2 = 0.241, p < 0.001. Overall, these results demonstrate that models that predict chronic motor outcomes using data-driven features, particularly when lesion data is represented in terms of structural disconnection, perform better than models that predict chronic motor outcomes using theory-based features from the motor system. However, combining both theory-based and data-driven models provides the best predictions.

BackgroundHuntingtons Disease (HD) is an inherited neurodegenerative disorder characterised by progressive cognitive and motor decline resulting from atrophy within basal ganglia networks. Although no disease-modifying therapies currently exist, several novel clinical trials are ongoing. Sensitive non-invasive imaging biomarkers are therefore essential for evaluating therapeutic effects. Soma and Neurite Density Imaging (SANDI), a multi-shell diffusion-weighted imaging model, estimates intracellular signal fractions arising from sphere-shaped soma that show promise as proxies for HD-related neurodegeneration. Although HD is rare, it offers a valuable model for understanding other neurodegenerative diseases due to its clear genetic cause and shared patterns of protein abnormalities. ObjectiveTo characterise HD-related microstructural abnormalities in the basal ganglia and thalami using SANDI and examine associations between SANDI indices, volumetric measurements, and motor performance. MethodsT1-weighted anatomical and multi-shell diffusion-weighted images (b-values: 200 s/mm{superscript 2}- 6,000 s/mm{superscript 2}) were acquired using a 3T Siemens Connectom scanner (300mT/m) in 56 HD individuals (MeanAge = 46.1, SDAge = 13.8, 25 females) and 57 healthy controls (MeanAge = 45.0, SDAge = 13.8, 31 females). HD participants completed Quantitative Motor (Q-Motor) tasks, including speeded and paced finger tapping, which were reduced to one principal component of motor performance. Following standard diffusion-weighted data preprocessing, SANDI and diffusion tensor models estimated apparent soma density, apparent soma size, apparent neurite density, extracellular signal fraction, fractional anisotropy, and mean diffusivity. The caudate, putamen, pallidum, and thalamus were segmented bilaterally, and micro-structural and volumetric indices were extracted and compared. Correlations between SANDI indices, Q-Motor performance, and volumetric measures were analysed. ResultsHD was associated with reduced apparent soma density (rrb = 0.32, p [&le;] 0.007) and increased apparent soma size (rrb = 0.45, p < 0.001) and extracellular signal fraction (rrb = 0.34, p [&le;] 0.003) in the basal ganglia, but not the thalami. These differences were more pronounced at HD-Integrated Staging System 0-1 than 2-3. No differences were found in apparent neurite density (rrb = 0.18, p = 0.17). HD-related increases in fractional anisotropy and mean diffusivity in the basal ganglia were replicated. Q-Motor component scores correlated negatively with apparent soma density and positively with apparent soma size and extracellular signal fraction. SANDI indices and age explained up to 63% of striatal atrophy in HD. ConclusionSANDI measures detected HD-related neurodegeneration in the striatum, accounted significantly for striatal atrophy, and correlated with motor impairments. Reduced apparent soma density and increased apparent soma size align with ex vivo evidence of medium spiny neuron loss and glial reactivity. SANDI shows promise as an in vivo biomarker and surrogate outcome measure for clinical trials of disease-modifying therapies for HD and other neurodegenerative diseases.

BackgroundTheory of Mind (ToM) deficits are well-documented in right-hemisphere stroke but remain understudied in post-stroke aphasia. Prior work suggests that performance on tasks assessing ToM may be relatively preserved in aphasia and dissociable from language impairment, but these findings are based largely on small studies. This study examined performance on nonverbal false-belief tasks in post-stroke aphasia, its relationship with aphasia severity, and whether vascular brain health, operationalized using cerebral small vessel disease (CSVD) markers, contributed to variability in performance. MethodsForty-four individuals with aphasia completed two nonverbal belief-reasoning tasks assessing spontaneous perspective-taking and self-perspective inhibition. Task accuracy served as the primary outcome. Linear regression models examined associations between task performance, aphasia severity (Western Aphasia Battery-Revised Aphasia Quotient), and CSVD markers, including white matter hyperintensities, cerebral microbleeds, lacunes and enlarged perivascular spaces in the basal ganglia and centrum semiovale. ResultsPerformance was heterogeneous across tasks, with reduced performance observed in 23% of participants on the Reality-Unknown task and 36% on the Reality-Known task. Aphasia severity was not associated with task accuracy. Greater cerebral microbleed count was associated with lower accuracy on both tasks, while greater basal ganglia enlarged perivascular spaces burden showed a more selective association with lower performance. ConclusionsPerformance on nonverbal false-belief tasks in aphasia is variable and not explained by aphasia severity alone. These findings suggest that apparent ToM-related difficulties in aphasia may be shaped by broader vascular brain health, supporting a more multidimensional framework for interpreting social-cognitive task performance after stroke.

The APOE-{varepsilon}4 allele is a genetic risk factor for late-onset Alzheimers disease (AD). Beyond cognitive decline, APOE-{varepsilon}4 affects motor function, reducing muscle strength and coordination, potentially through mitochondrial dysfunction and oxidative stress. This study examined the influence of the APOE-{varepsilon}4 allele on neuromuscular function in oral muscles involved in speech production, using surface electromyography (EMG); and assessed the predictive power of EMG measures in differentiating APOE-{varepsilon}4 carriers from noncarriers. Forty-two cognitively intact adults (16 APOE-{varepsilon}4 carriers, 26 noncarriers) completed speech tasks while EMG was recorded from seven craniofacial muscles. Seventy EMG features including amplitude, frequency, complexity, regularity, and functional connectivity were extracted. Statistical analyses assessed genotype effects, sex differences, and correlations with blood metabolic biomarkers. APOE-{varepsilon}4 carriers exhibited increased motor unit recruitment and synchronization, suggesting accelerated muscle fatigue. EMG-based measures outperformed cognitive tests in distinguishing carriers (AUC = 0.90) and correlated with metabolic biomarkers. Sex differences emerged, with female carriers showing reduced and male carriers showing increased functional connectivity. These findings highlight speech-based neuromuscular changes as potential early biomarkers of Alzheimers risk before cognition is affected.