Neuroimaging studies are essential for evaluating patients with drug-resistant focal epilepsy and determining their candidacy for epilepsy surgery. The past decade has seen the emergence of neuroimaging-based deep learning models, which have been developed to both detect epileptogenic lesions on MR imaging and predict post-surgical seizure outcome. Large, multi-center studies have demonstrated promise for epileptogenic lesion detection; however, neuroimaging-based surgical outcome prediction models remain exploratory. Translation of such models into routine epilepsy surgery clinical workflows will require transparent, interpretable, and prospectively validated model designs.

Epilepsy is characterized by widespread structural brain alterations extending beyond the epileptic zone, involving both cortical and subcortical regions. Importantly, the clinical manifestation of epilepsy, including seizure types, psychiatric comorbidities, and treatment responses, has been shown to differ between sexes. However, sex differences in structural alterations in epilepsy have been seldomly reported in neuroimaging studies, partly due to limited sample sizes and single-center designs. Here, we systematically investigated sex differences in common epilepsies and their related clinical variables using structural neuroimaging biomarkers in an international multi-center cohort of 1,253 epilepsy patients and 1,077 healthy controls. We studied cortical thickness and subcortical volume in two types of epilepsy: temporal lobe epilepsy (TLE) and genetic generalized epilepsy (GGE). Both male and female patients with TLE showed widespread cortical and subcortical thinning compared with controls. In GGE, when compared separately to controls, male patients showed only subtle structural alterations, whereas female patients exhibited more widespread structural alterations. Sex-stratified analyses revealed some variation in the extent and distribution of cortical thickness and subcortical volume alterations between male and female patients in both epilepsy cohorts. Yet, we did not find significant sex-by-diagnosis interaction effects in TLE and GGE. Similarly, no significant interaction effects were observed between sex and age of onset or disease duration in either patient group. Overall, although we observed some differences in regional cortical thickness and subcortical volume between male and female patients with epilepsy, we did not find significant sex-by-diagnosis interactions. Our findings indicate that sex differences in behavioral and clinical outcomes of epilepsy may involve biological or functional processes that require further investigation.

Diagnostic MRI evaluation of temporal lobe epilepsy (TLE) depends on the subjective visual interpretation of MRI images. These interpretations could be enhanced by quantitative artificial intelligence (AI) support tools. Humans often make sequential and conditional decisions during their radiological interpretations, such as whether an abnormality is present and, if present, characterizing the abnormality. It is not known whether it is superior to train AI to treat every decision separately in a similar step-wise manner or to train a model holistically on all decisions simultaneously. Here, we analysed three large epilepsy MRI datasets [n = 3676, 2320 people with epilepsy and 1356 healthy controls (HC)] to perform two tasks: (i) establish the presence of a TLE pattern on MRI and (ii) determine TLE pattern lateralization. We compared Step-wise models that independently classify TLE versus HC and lateralize patients as left TLE (L-TLE) or right TLE (R-TLE), against a simultaneous model trained to distinguish all three classes in a single step. To do this, 3D volumetric T1-weighted images were input into an EfficientNetV2 model multiple times to ensure reproducibility of results. Class prediction, model classification confidence and saliency maps were output for interpretability. Step-wise models outperformed the Simultaneous model on both tasks (both Ps < 0.001), with an average ∼2.8% accuracy increase for discriminating HC from TLE and an average 12.7% accuracy increase for distinguishing L-TLE from R-TLE. For both the Step-wise and Simultaneous models, important features discriminating TLE from HC included the known TLE limbic pattern involving the hippocampus, parahippocampal cortical regions, cingulate cortex and lateral temporal regions. However, there was less concordance between the Step-wise and Simultaneous models for the L-TLE versus R-TLE task (all Fisher's Zs > 10.5, Ps < 0.001); the Step-wise model focused less on subcortical regions such as the thalamus and hippocampus and focused more on distributed cortical pathology. Across the two Step-wise models, 95.1% of TLE patients had accurate classifications in either HC versus TLE and/or L-TLE versus R-TLE tasks. These results included 69.6% of patients being both correctly labelled as TLE and lateralized, 13.9% being correctly labelled TLE but lateralized incorrectly and 11.6% being lateralized correctly but not detected as TLE. These findings provide evidence that diagnostic tasks with simpler, Step-wise AI models may enhance diagnostic performance and interpretability in clinical workflows. Future AI clinical support tools can leverage this step-wise approach in the early identification of TLE-related structural patterns, supporting timely diagnosis and treatment decisions.

Extensive neuroimaging research in temporal lobe epilepsy with hippocampal sclerosis (TLE-HS) has identified brain atrophy as a disease phenotype. While it is also related to a complex genetic architecture, the transition from genetic risk factors to brain vulnerabilities remains unclear. Using a population-based approach, we examined the associations between epilepsy-related polygenic risk for HS (PRS-HS) and brain structure in healthy developing children, assessed their relation to brain network architecture, and evaluated its correspondence with case-control findings in TLE-HS diagnosed patients relative to healthy individuals. We used genome-wide genotyping and structural T1-weighted MRI of 3826 neurotypical children from the Adolescent Brain Cognitive Development (ABCD) study. Surface-based linear models related PRS-HS to cortical thickness measures, and subsequently contextualized findings with structural and functional network architecture based on epicentre mapping approaches. Imaging-genetic associations were then correlated to atrophy and disease epicentres in 785 patients with TLE-HS relative to 1512 healthy controls aggregated across multiple sites. Higher PRS-HS was associated with decreases in cortical thickness across temporo-parietal as well as fronto-central regions of neurotypical children. These imaging-genetic effects were anchored to the connectivity profiles of distinct functional and structural epicentres. Compared with disease-related alterations from a separate epilepsy cohort, regional and network correlates of PRS-HS strongly mirrored cortical atrophy and disease epicentres observed in patients with TLE-HS and were highly replicable across different studies. Findings were consistent when using statistical models controlling for spatial autocorrelations and robust to variations in analytic methods. Capitalizing on recent imaging-genetic initiatives, our study provides novel insights into the genetic underpinnings of structural alterations in TLE-HS, revealing common morphological and network pathways between genetic vulnerability and disease mechanisms. These signatures offer a foundation for early risk stratification and personalized interventions targeting genetic profiles in epilepsy.

Objectives: The severe cerebral atrophy patterns of progressive supranuclear palsy (PSP) create challenges in clinical neuroradiology practice and neuroimaging research as often, subcortical regions are not accurately segmented with existing atlases or require labour intense manual segmentation. We aimed to create a PSP specific neuroimaging atlas. Methods: Seven subcortical regions were manually segmented on Quantitative Susceptibility Mapping (QSM) images of 34 PSP Richardson’s Syndrome (PSP-RS) subjects and used to automatically segment five validation subjects using the ANTs Joint Label Fusion module. Accuracy was confirmed by manual segmentation of QSM images of the validation subjects as the gold standard. Results: The PSP atlas outperformed existing atlases with median dice coefficient and 95% Hausdorff Distance of 0.88 (0.84-0.90) and 1.13mm (1.13-1.60) compared to 0.69 (0.65-0.76) and 3.3mm (2.53-5.12) for Talaraich and Harvard-Oxford atlases. Conclusions: Our new PSP subcortical atlas outperforms currently available atlases and can be applied to multimodal datasets and T1-weighted images alone.

Objective:Exercise-assistance strategies are useful in allowing mobility-impaired patients to access the benefits of exercise. This trial is the first of the Reviver device, which facilitates exercise via a novel strength and balance training mechanism. The objective was to examine the effect of a 12-week Reviver intervention on symptoms of Parkinsonism, and to pilot the randomised controlled trial design, including randomisation and acceptance of the exercise intervention. Methods:This was a pilot, parallel-arm randomised controlled trial with assessor blinding. Participants (n=30: 22 with Parkinson's disease (PD) and 8 with atypical Parkinsonism conditions (AP)) were allocated to either experimental or control group. The experimental group received 24 sessions of 30 min on the Reviver over 12 weeks, the control group received their standard care. The Movement Disorders Society Unified Parkinson's Disease Rating Scale, and secondary outcomes (balance, gait, mobility, lower-body strength/coordination, tremor and grip strength) were acquired at endpoints the week prior to intervention commencement and the week after intervention termination. Recruitment progress, adherence to intervention, acceptability of intervention and adverse effects were also recorded. Results:For the PD cohort, lower-body strength/coordination (5× Sit-To-Stand; b(95% CI)=-3.02 (-5.16 to -0.89), p=0.013), gait (self-selected walking speed; b=12.48 (2.18 to 22.78), p=0.029, stride length; b=9.75 (0.99 to 18.52), p=0.043) and backward walking speed (b=14.25 (1.93 to 26.57), p=0.038) were improved by the intervention. No significant effect of intervention on the outcome variables was found in the AP cohort. No serious adverse events were recorded. Median adherence to treatment was 95.8%. Interpretation:This pilot trial indicates that the Reviver is a safe and well-tolerated exercise-assistance intervention. The Reviver showed some indications of benefit in our secondary measures for PD participants, but not in our primary outcome. This was a small sample, short duration pilot study, and further studies with larger samples and higher exercise volume are warranted to fully assess the safety and efficacy of the device.

Epilepsy is a heterogeneous neurological disorder affecting more than 70 million people worldwide, posing significant challenges for clinicians due to its complex etiology, diverse manifestations, variable treatment responses, and the inability to predict seizures or disease onset reliably. Despite advances in antiseizure medications, approximately 30% of patients remain treatment-resistant, highlighting the urgent need for therapies with antiepileptogenic or disease-modifying effects. To optimize and individualize strategies to predict and treat seizures and epilepsy, efforts to identify biomarkers of epilepsy risk, epileptogenesis, seizures, and therapy response are ongoing. This article reports key presentations and discussions from the 2023 Workshop on Neurobiology of Epilepsy (WONOEP XVII) in Kilkea, Ireland on novel epilepsy biomarkers and treatment strategies beyond the synapse and does not constitute a comprehensive review of biomarkers or treatment strategies. Much of the focus in epilepsy research has centered on identifying primarily neuronal processes or components. The 2023 WONOEP presentations discussed advances in plasma biomarkers for posttraumatic seizures and outcomes, perivascular spaces in posttraumatic epilepsy and poststroke epilepsy, imaging biomarkers of astrogliosis, and plasma microRNA biomarkers of intellectual disability and autism in tuberous sclerosis complex. Furthermore, research on immuno- and anti-inflammatory therapies and blood-brain barrier in drug-resistant focal epilepsies and infantile epileptic spasms syndrome was presented, as well as on the effects of antiseizure and cardioprotective medications on cardiac injury in temporal lobe epilepsy. The review also emphasizes the need for further interdisciplinary collaboration to accelerate the translation of these findings into clinical practice, ultimately improving outcomes and quality of life for people with epilepsy.

Enlarged perivascular spaces (PVS) are increasingly recognized as biomarkers of cerebral small vessel disease, Alzheimer's disease, stroke, and aging-related neurodegeneration. However, manual segmentation of PVS is time-consuming and subject to moderate inter-rater reliability, while existing automated deep learning models have moderate performance and typically fail to generalize across diverse clinical and research MRI datasets. We adapted MedNeXt-L-k5, a Transformer-inspired 3D encoder-decoder convolutional network, for automated PVS segmentation. Two models were trained: one using a homogeneous dataset of 200 T2-weighted (T2w) MRI scans from the Human Connectome Project-Aging (HCP-Aging) dataset and another using 40 heterogeneous T1-weighted (T1w) MRI volumes from seven studies across six scanners. Model performance was evaluated using internal 5-fold cross validation (5FCV) and leave-one-site-out cross validation (LOSOCV). MedNeXt-L-k5 models trained on the T2w images of the HCP-Aging dataset achieved voxel-level Dice scores of 0.88+/-0.06 (white matter, WM), comparable to the reported inter-rater reliability of that dataset, and the highest yet reported in the literature. The same models trained on the T1w images of the HCP-Aging dataset achieved a substantially lower Dice score of 0.58+/-0.09 (WM). Under LOSOCV, the model had voxel-level Dice scores of 0.38+/-0.16 (WM) and 0.35+/-0.12 (BG), and cluster-level Dice scores of 0.61+/-0.19 (WM) and 0.62+/-0.21 (BG). MedNeXt-L-k5 provides an efficient solution for automated PVS segmentation across diverse T1w and T2w MRI datasets. MedNeXt-L-k5 did not outperform the nnU-Net, indicating that the attention-based mechanisms present in transformer-inspired models to provide global context are not required for high accuracy in PVS segmentation.

Fluid transport in the neurovascular unit is essential for maintaining brain health through nutrient delivery and waste clearance. However, these systems are complex and the inter-dependencies between elements of these systems and how they may change through aging is not well understood. MRI outcomes provide insight into the underlying biological mechanisms of these systems in vivo, including water exchange rate through the neurovascular unit (BBB kw), enlarged perivascular spaces (ePVS), cerebral blood flow (CBF), free water (FW), and white matter hyperintensities (WMH). To explore the relationships between functional elements of the neurovascular unit, this study investigated relationships between these MRI measures using Bayesian mixed models, and their variation with chronological age or atrophy-related brain age (brainageR) using linear regression. In 132 non-clinical older adults (mean age = 67 years; 68% female), BBB kw positively associated with CBF (β^ = 0.08, 95% credible interval (CI) = [0.02, 0.15]). FW positively associated with both ePVS (β^ = 0.44, CI = [0.30, 0.63]) and WMH (β^ = 0.13, CI = [0.04, 0.21]). BBB kw, CBF and ePVS decreased with age, while FW and WMH increased (all p < 0.05). There were no associations with atrophy-related brain age (all p > 0.05). Relationships between FW, ePVS and WMH likely reflect interconnectivity of fluid regulation within different compartments, while the relationship between BBB kw and CBF indicates a link between neurovascular fluid flow and vessel function. While individual metrics of neurovascular integrity are associated with age, their inter-relationships appear stable, providing a baseline for future research in fluid transport and vascular health in neurodegenerative disease.

Individual imaging and fluid biomarkers provide insights into specific components of brain health, but integrated multimodal approaches are necessary to capture the complex, interrelated biological systems that contribute to brain homeostasis and neurodegenerative disease. Using data from the Brain and Cognitive Health (BACH) cohort study (N = 127; mean age = 67 years, 68

Approximately 40% of patients with drug-resistant epilepsy referred for surgical evaluation have no epileptogenic lesion on MRI (MRI-negative). MRI-negative epilepsy is associated with poorer seizure freedom prognosis and has therefore motivated the development of structural post-processing methods to "convert" MRI-negative to MRI-positive cases. In this article, we review the principles, advances, and challenges of voxel- and surface-based cortical morphometric MRI techniques in detecting the epileptogenic zone. The ground truth for the presumed epileptogenic zone in imaging studies can be classified into lesion-based (MRI lesion mask or histopathology) or epileptogenicity-based ground truth (anatomical-electroclinical correlations or resections that lead to seizure freedom). Voxel-based techniques are reported to have a 13%-97% concordance rate, while surface-based techniques have 67%-92% compared to lesion-based ground truths. Epileptogenicity-based ground truth may be more relevant in the case of MRI-negative cases; however, the sensitivity and concordance rate (voxel-based technique 7.1%-66.7%, and surface-based technique 62%) are limited by the reliance on scalp EEG and qualitative analysis of seizure-onset pattern. The use of stereo-EEG and quantitative EEG analysis may fill this gap to evaluate the correlation between cortical morphometry results and electrophysiological epileptogenic biomarkers of the epileptogenic zone and help improve the yield of structural post-processing tools. PLAIN LANGUAGE SUMMARY: Locating the epileptogenic zone (the brain area that is responsible for seizure generation) is important to diagnose and plan epilepsy treatments. An abnormal brain imaging (MRI) result can help clinical decision-making; however, around 40% of patients have normal MRI results (MRI-negative). We are reviewing the potential of two advanced MRI methods (voxel- and surface-based cortical morphometry) to localize the epileptogenic zone in the presence or absence of visible MRI abnormalities. We also describe the current challenge of applying the above methods in daily clinical practice and propose using advanced brain recording analysis to aid this translation process.

INTRODUCTION:Frontotemporal dementia (FTD) remains challenging to diagnose owing to the marked clinical heterogeneity associated with the disease. This heterogeneity stems from the complex interplay of various clinical phenotypes, genetic mutations and underlying neuropathologies, such as TDP-43 and tau proteinopathies. Currently, there is no single confirmed biomarker that can reliably diagnose disease, specifically disease stage, disease subtype and underlying neuropathology. Recent research has indicated that neuroimaging techniques hold the most promise for the discovery of FTD biomarkers. We propose a protocol for a systematic review and meta-analysis to identify MRI and fluorodeoxyglucose positron emission tomography (FDG-PET) biomarkers associated with clinical, genetic and pathological subtypes of FTD. We aim to address the following research questions: can regional MRI volumetry and FDG-PET hypometabolism differentiate (1) FTD patients from healthy controls; (2) sporadic cases of FTD from healthy controls; (3) genetic cases of FTD (MAPT, GRN, and C9orf72 mutations); and (4) underlying neuropathology, specifically discriminating between tau- and TDP-43-based FTD? METHODS:Literature searches will be performed across three databases: Ovid Medline, Ovid Embase and Web of Science. Publications that have fewer than five participants, are non-human-based, not written in the English language or contain unpublished data will be excluded. Two independent investigators will screen and subsequently evaluate which publications to include. Should any disagreements arise, a third investigator will settle the discrepancy. After the random-effects meta-analysis has been used to extract and pool the data, I2 analysis will be used to quantify heterogeneity. ETHICS AND DISSEMINATION:Ethics approval will not be required for this research. On completion, the systematic review and meta-analysis will be published in a peer-reviewed journal. PROSPERO REGISTRATION NUMBER:CRD42024545302.

Progressive supranuclear palsy (PSP) is a rare neurodegenerative disease with no current disease-modifying treatments approved. Longitudinal research and clinical trials for PSP are ongoing and require reliable measures that are sensitive to disease progression. Despite susceptibility to subjective limitations, clinical and cognitive assessments are the most used instruments in therapeutic trials in PSP. The objective of this review was to identify measures that have been studied longitudinally as measures of progression and are suitable for use as clinical trial endpoints. We reviewed the measures currently used as trial endpoints, identifying the clinical, cognitive, fluid and imaging measures that have previously been studied longitudinally, and discuss current diagnostic and emerging measures that are yet to be studied longitudinally but that may be sensitive to disease progression. We found that many fluid and imaging measures require further research to validate their use as longitudinal measures of change, including emerging measures that have not yet been studied specifically in PSP. We also summarize the sample size estimates required to detect changes in a two-arm, 52-week therapeutic trial and found that specific MRI volumes require the smallest sample sizes to detect change.

INTRODUCTION:Brain-derived tau (BD-tau) measures tau specifically from brain-derived sources and can differentiate Alzheimer's disease (AD) from other diseases. This study investigated BD-tau as a potential biomarker of treatment effect. METHODS:BD-tau and phosphorylated tau-217 (p-tau217) levels were measured after treatment with an anti-tau drug in AD and behavioral variant frontotemporal dementia (bvFTD) clinical trials, and the association with total tau (t-tau), p-tau181, and amyloid beta 42 (Aβ42) was examined. RESULTS:Cerebrospinal fluid (CSF) BD-tau decreased after treatment in the AD cohort; however, no change was seen in bvFTD or p-tau217 in either cohort. CSF t-tau and p-tau181 correlated with BD-tau in AD (r = 0.9113 and 0.7746, p < 0.0001) and bvFTD (r = 1.0 and r = 0.79, p < 0.05). CSF BD-tau did not correlate with serum or plasma BD-tau in bvFTD. DISCUSSION:CSF BD-tau shows potential as a biomarker of treatment effect in AD but not bvFTD. Further research is needed to investigate this effect in blood-based samples and in other neurodegenerative diseases. Trial registration: ACTRN12611001200976, ACTRN12617001218381. Highlights:Cerebrospinal fluid (CSF) brain-derived tau (BD-tau) levels decreased with sodium selenate treatment in patients with Alzheimer's disease (AD).CSF BD-tau levels did not change with sodium selenate treatment in bvFTD.Baseline CSF BD-tau correlated with CSF total tau (t-tau) and phosphorylated tau-181 (p-tau181) in AD and behavioral variant frontotemporal dementia (bvFTD).Baseline serum and plasma BD-tau levels did not correlate with CSF BD-tau in bvFTD.CSF p-tau217 did not change with sodium selenate treatment in AD or bvFTD.

BACKGROUND AND OBJECTIVES:Temporal lobe epilepsy (TLE) is commonly associated with mesiotemporal pathology and widespread alterations of gray and white matter structures. Evidence supports a progressive condition, although the temporal evolution of TLE is poorly defined. In this ENIGMA-Epilepsy study, we aim to investigate structural alterations in gray and white matter across the adult lifespan in patients with TLE by charting both gray and white matter changes and explore the covariance of age-related alterations in both compartments. METHODS:Mega-analysis of parcellated T1-weighted and diffusion MRI data across 18 international sites for patients with TLE was compared against healthy controls. We combined median-age split groupwise comparisons with cross-sectional sliding age-window analyses to explore gray (cortical thickness, subcortical volume) and white matter microstructure (fractional anisotropy, mean diffusivity) age-related changes. Five-year range age windows were constructed from mean z scores of all patients. Covariance analyses examined the coupled correlations of gray and white matter lifespan curves for each region. RESULTS:We studied 769 patients with TLE and 885 healthy controls across an age range of 17-73 years. Robust (pFDR < 0.05) gray matter thickness/volume decline (d < -0.20) was seen across a broad cortico-subcortical territory, extending beyond the mesiotemporal lobe throughout the adult lifespan in patients with TLE. White matter changes were also widespread across multiple fiber tracts with peak effects in temporolimbic fibers in fractional anisotropy (d < -0.3, pFDR < 0.05) and mean diffusivity measures (d > 0.3, pFDR < 0.05). Changes spanned the adult time window and effects exceeded typical aging-related processes in patients at the level of cortical thickness, subcortical volume, and diffusion measures, particularly in patients older than 55 years. Covariance analyses revealed strong associations across multiple white matter tracts, subcortical structures, and cortical regions within and beyond the temporolimbic system. DISCUSSION:This study highlights that patients with TLE exhibit more pronounced and widespread gray and white matter atrophy across the lifespan. The cross-sectional nature of our study limits definitive conclusions on whether the atrophy shown is progressive but emphasizes the importance of prompt diagnosis and intervention in patients. Collectively, our results motivate future longitudinal studies to clarify consequences of drug-resistant epilepsy.

OBJECTIVE:The glymphatic system is thought to be the brain's primary waste clearance system, responsible for eliminating soluble metabolites and proteins from the central nervous system. It consists of the cerebrospinal fluid, the interstitial fluid, and a conduit between the two, perivascular spaces (PVS). PVS and glymphatics may impact the pathophysiology of epilepsy and its associated neuropsychiatric comorbidities, potentially via reduced clearance of excitotoxic substances. This study investigates enlarged PVS burden in a large patient group with various types of epilepsy. METHODS:A total of 467 people with various types of epilepsy were recruited from the Hospital das Clínicas, University of Campinas, Brazil; 267 had temporal lobe epilepsy with hippocampal sclerosis (TLE-HS), 71 had TLE with no magnetic resonance imaging (MRI)-visible lesions, 65 had focal extratemporal epilepsy, and 64 had idiopathic generalized epilepsy. They were matched for age and sex with 473 healthy volunteers as controls. All participants were scanned with T1-weighted MRI, and a deep-learning algorithm, PINGU (Perivascular-Space Identification Nnunet for Generalized Usage), was applied to segment PVS. The volumes of PVS in the white matter (WM) and basal ganglia (BG) were calculated, and PVS volume fraction (PVS-VF) was used as a dependent variables in a general linear model, with the diagnostic group as the independent variable of interest. RESULTS:The epilepsy group, across all subtypes, had higher PVS-VF in the BG compared to controls (101%-140%, effect size = .95-1.37, p < 3.77 × 10-15). There was no difference in PVS-VF in the WM between the epilepsy groups and healthy controls, or between different epilepsy subtypes. Only the TLE-HS group had a PVS-VF asymmetry in the WM, with more PVS on the contralateral side, particularly in the temporal lobes. There was no association between PVS-VF and duration of illness. SIGNIFICANCE:There is an increase in PVS volume in the BG across broad subtypes of epilepsy, suggesting either a common mechanism of seizure generation or a common consequence of seizures.

Importance:A leading cause of surgically remediable, drug-resistant focal epilepsy is focal cortical dysplasia (FCD). FCD is challenging to visualize and often considered magnetic resonance imaging (MRI) negative. Existing automated methods for FCD detection are limited by high numbers of false-positive predictions, hampering their clinical utility. Objective:To evaluate the efficacy and interpretability of graph neural networks in automatically detecting FCD lesions on MRI scans. Design, Setting, and Participants:In this multicenter diagnostic study, retrospective MRI data were collated from 23 epilepsy centers worldwide between 2018 and 2022, as part of the Multicenter Epilepsy Lesion Detection (MELD) Project, and analyzed in 2023. Data from 20 centers were split equally into training and testing cohorts, with data from 3 centers withheld for site-independent testing. A graph neural network (MELD Graph) was trained to identify FCD on surface-based features. Network performance was compared with an existing algorithm. Feature analysis, saliencies, and confidence scores were used to interpret network predictions. In total, 34 surface-based MRI features and manual lesion masks were collated from participants, 703 patients with FCD-related epilepsy and 482 controls, and 57 participants were excluded during MRI quality control. Main Outcomes and Measures:Sensitivity, specificity, and positive predictive value (PPV) of automatically identified lesions. Results:In the test dataset, the MELD Graph had a sensitivity of 81.6% in histopathologically confirmed patients seizure-free 1 year after surgery and 63.7% in MRI-negative patients with FCD. The PPV of putative lesions from the 260 patients in the test dataset (125 female [48%] and 135 male [52%]; mean age, 18.0 [IQR, 11.0-29.0] years) was 67% (70% sensitivity; 60% specificity), compared with 39% (67% sensitivity; 54% specificity) using an existing baseline algorithm. In the independent test cohort (116 patients; 62 female [53%] and 54 male [47%]; mean age, 22.5 [IQR, 13.5-27.5] years), the PPV was 76% (72% sensitivity; 56% specificity), compared with 46% (77% sensitivity; 47% specificity) using the baseline algorithm. Interpretable reports characterize lesion location, size, confidence, and salient features. Conclusions and Relevance:In this study, the MELD Graph represented a state-of-the-art, openly available, and interpretable tool for FCD detection on MRI scans with significant improvements in PPV. Its clinical implementation holds promise for early diagnosis and improved management of focal epilepsy, potentially leading to better patient outcomes.

Perivascular spaces (PVSs) form a central component of the brain’s waste clearance system, the glymphatic system. These structures are visible on MRIs when enlarged, and their morphology is associated with aging and neurological disease. Manual quantification of PVS is time consuming and subjective. Numerous deep learning methods for PVS segmentation have been developed for automated segmentation. However, the majority of these algorithms have been developed and evaluated on homogenous datasets and high resolution scans, perhaps limiting their applicability for the wide range of image qualities acquired in clinical and research settings. In this work we train a nnUNet, a top-performing task driven biomedical image segmentation deep learning algorithm, on a heterogenous training sample of manually segmented MRIs of a range of different qualities and resolutions from 7 different datasets acquired on 6 different scanners. These are compared to the two currently publicly available deep learning methods for 3D segmentation of PVS, evaluated on scans with a range of resolutions and qualities. The resulting model, PINGU (Perivascular space Identification Nnunet for Generalised Usage), achieved voxel and cluster level dice scores of 0.50(SD=0.15) and 0.63(0.17) in the white matter (WM), and 0.54 (0.11) and 0.66(0.17) in the basal ganglia (BG). Performance on unseen “external” sites’ data was substantially lower for both PINGU (0.20-0.38 [WM, voxel], 0.29-0.58 [WM, cluster], 0.22-0.36 [BG, voxel], 0.46-0.60 [BG, cluster]) and the publicly available algorithms (0.18-0.30 [WM, voxel], 0.29-0.38 [WM cluster], 0.10-0.20 [BG, voxel], 0.15-0.37 [BG, cluster]). Nonetheless, PINGU strongly outperformed the publicly available algorithms, particularly in the BG. PINGU stands out as broad-use PVS segmentation tool, with particular strength in the BG, an area of PVS highly related to vascular disease and pathology.

The neural basis of functional/dissociative seizures represents a significant challenge for psychiatry and neurology. While abnormalities in neuroimaging have been observed, findings across studies remain inconsistent. This systematic review and meta-analyses examined neuroimaging abnormalities in functional/dissociative seizures, conducting separate meta-analyses of cortical thickness and whole-brain, seed-based, resting-state functional connectivity. A total of 22 studies were included in the review, of which seven and three were suitable for meta-analysis of structural and functional MRI data, respectively. Meta-analysis of cortical thickness from seven structural MRI studies (198 functional/dissociative seizures and 254 healthy controls) revealed a reduction in the right calcarine fissure/surrounding cortex (z = -3.08, p < 0.01) and right precentral gyrus (z = -3.112, p < 0.01). Additionally, the whole-brain seed-based resting-state functional connectivity analysis from three functional MRI studies (48 patients with functional/dissociative seizures and 64 healthy controls) demonstrated increased connectivity between seeds in the sensorimotor network (including posterior insula, paracentral lobule, postcentral gyrus, supplementary motor area and superior temporal gyrus) and areas of the ventral attention network, including left thalamus (z = 3.73, p < 0.01) and left putamen (z = 3.62, p < 0.01); and seeds in the default mode network (including ventral anterior insula, superior frontal gyrus and middle temporal gyrus) exhibited heightened connectivity with the left insula (z = 4.01, p < 0.01). These findings indicate structural and functional abnormalities in visual, sensorimotor, default mode, and attentional areas and networks associated with functional/dissociative seizures, highlighting potential neurobiological mechanisms. Further research is necessary to elucidate the complex interplay of clinical features and comorbidities, which will enhance our understanding of the pathophysiology of functional/dissociative seizures.

OBJECTIVE:The yield of voxel-based gray matter volume (GMV) quantification to identify the epileptogenic zone (EZ) remained moderate, which may be due to the choice of "ground truth." We explored whether GMV differences are associated with stereo-electroencephalography (SEEG)-defined epileptogenicity as potential EZ imaging biomarkers. METHODS:We included SEEG patients along with age- and sex-matched non-epilepsy controls. We performed a non-parametric voxel-based permutation inference between each SEEG patient's non-contrast 3T T1-weighted magnetic resonance imaging (MRI) scan and controls, resulting in pseudo-t-test maps reflecting GMV differences. We classified SEEG contacts based on their involvement in (i) the EZ, defined as contacts designated for radiofrequency thermocoagulation, and (ii) active irritative zones, defined as contacts generating the top 10% of spikes, fast ripples, and cross-rates of HFO*spikes for each patient. We then performed mixed-effects logistic regressions and performance analysis. RESULTS:We included 50 patients (median age 33.0 years, female 52.0%, MRI-negative 76.0%) and 51 controls (median age 37.0 years, female 56.9%). EZ: In general, increased GMV was associated with EZ contacts across the whole cohort (odds ratio [OR] 1.21, 95% confidence interval [CI] 1.11-1.31, p < .001) and the MRI-positive group (OR 1.45, 95% CI 1.28-1.64, p < .001), but not in the MRI-negative group (p = .621). Reduced GMV was associated with EZ contacts in mesiotemporal regions (OR .53, 95% CI .33-.84, p = .035) but the opposite in neocortical areas (OR 1.32, 95% CI 1.21-1.44, p < .001). Within MRI-visible lesions, increased GMV was positively associated with EZ contacts (OR 1.50, 95% CI 1.19-1.89, p = .005). Active irritative zone: In the MRI-positive group, increased GMV was linked to neocortical contacts exhibiting the top 10% of spikes (p = .040) and fast ripples (p = .008), but not in MRI-negative cases. SIGNIFICANCE:Local GMV differences were positively associated with EZ contacts in the whole cohort and MRI-positive patients but not in MRI-negative patients, with positive associations in the neocortex and a negative one in the mesiotemporal structures. Within MRI-visible lesions, increased GMV was associated with EZ contacts.