Biomarkers
Ke Xie et al. 2025
Patients with drug-resistant epilepsy often exhibit heterogeneous cognitive and neuroimaging profiles, posing challenges for accurate diagnosis and treatment planning. We developed computational frameworks that integrate multicenter, multimodal imaging with advanced normative modeling and artificial intelligence to detect brain alterations at the individual-patient level. These tools aim to localize epileptogenic zones, predict surgical outcomes, and inform personalized therapeutic strategies to improve clinical care.
听
Ella Sahlas et al. 2025
This study combined multimodal MRI and high-density EEG. Results revealed microstructural and morphological changes in areas of the brain generating epileptic spikes. Additional changes were found in areas with strong functional or structural connections to spike sources, but not in areas adjacent to these sources. These findings help us better understand brain structure-function relationships in epilepsy, and they point to imaging markers that could help identify epileptogenic tissue.
听
Ngo et al. 2024
The cerebrovascular system plays a critical role in maintaining whole-brain structure and function. However, in patients with pharmaco-resistant temporal lobe epilepsy (TLE), cerebral blood flow鈥攎easured using non-invasive arterial spin labelling magnetic resonance imaging鈥攊s significantly decreased, predominantly in fronto-temporal regions. These vascular alterations may contribute to the pathophysiology of TLE affecting and extending beyond mesiotemporal structures.
听
Ravnoor Gill et al. 2021
In patients with drug-resistant epilepsy, lesions are typically seen on MRI and surgical resection of this lesion can lead to seizure-freedom. However, many patients do not have a visible abnormality that can be seen with traditional methods and the surgical outcome is less favourable. Using a multicentre-validated deep learning algorithm, the team was able to detect MRI-negative focal cortical dysplasia with high sensitivity and specificity.
听
Zhengchen Cai et al. 2025
This study provides the first systematic investigation of hemodynamic response function (HRF) variability across the epileptic brain. We identified representative HRF shape classes and mapped the spatial distribution of HRF features across cortical and subcortical regions, revealing structured rather than random variability. HRF features were linked to underlying pathology in addition to anatomical location, showing graded differences across pathology types. These findings challenge conventional fMRI analyses in epilepsy that rely on canonical or predefined HRFs and support region- and pathology-informed modeling. By establishing an open-access whole-brain HRF library for epilepsy, this work lays the foundation for precision HRF modeling and may improve presurgical functional mapping, localization of epileptogenic zones and connectivity while advancing understanding of neurovascular alterations in neurological disease.
Epilepsy & Cognition
Donna Cabalo et al. 2024
This research explores a novel approach to profiling memory networks in both healthy individuals and patients with pharmaco-resistant epilepsies. By leveraging advanced data science techniques on functional neuroimaging data, the study employs a unique assessment of MRI scans acquired at high-field strengths (3T and 7T). The goal is to enhance our understanding of the functional organization of memory networks in the human brain, with a particular focus on how these networks may reorganize in patients with lesions in the mesiotemporal region and frontal lobe. Ultimately, this project aims to contribute to patient-specific surgical decision-making and the identification of novel biomarkers, which could improve the prognostication of surgical outcomes.
听
Qiongling Li et al. 2021
This study explored how the brain separates similar memories, a process called pattern separation, by comparing healthy people to those with temporal lobe epilepsy. Brain scans showed that healthy individuals used specific brain regions more effectively for memory, while patients showed disruptions, linking brain patterns to memory accuracy.
听
Lorenzo Caciagli et al. 2023
This study examined how epilepsy affects thinking and memory by comparing brain activity in people with frontal and temporal lobe epilepsy. Using brain scans and cognitive tests, our team found shared and distinct brain network disruptions in both types. These findings may help guide future treatments for cognitive problems in epilepsy.
Mechanisms
Raul Rodriguez-Cruces et al. 2025
People with drug-resistant temporal lobe epilepsy (TLE) may accumulate harmful tau proteins, which are misfolded proteins linked to brain cell damage and neurodegeneration. Using a specialized brain scan (PET) to detect abnormal tau deposits, we found increased tau in memory and thinking regions of TLE patients compared to healthy individuals. Higher tau levels were associated with longer disease duration and stronger brain connectivity in affected areas. Patients with more tau buildup also showed greater problems with memory and thinking. These findings suggest that tau accumulation may contribute to brain damage and cognitive decline in TLE.
听
Ngo et al. 2025
Temporal lobe epilepsy (TLE) is related to widespread brain alterations and a complex genetic architecture, but the transition from genetic risk factors to brain vulnerabilities remains unclear. Capitalizing on recent imaging-genetic initiatives, healthy developing children with elevated risk for TLE exhibited widespread thinning across the cortex, particularly in temporo-parietal and fronto-central regions. These patterns strongly mirrored cortical atrophy observed in patients with TLE with hippocampal sclerosis, suggesting a common pathway between genetic vulnerability and disease mechanisms.
听
Sara Larivi猫re et al. 2022
This large international study linked brain network changes in epilepsy to genetic risk factors. People with temporal or generalized epilepsy showed distinct brain connectivity patterns. These patterns matched the expression of epilepsy-related genes, offering insight into how genetics and brain structure interact鈥攁nd potentially guiding more targeted diagnosis and treatments.
Software
Behind some of the cutting edge epilepsy brain imaging research done at the Neuro are a set of freely available tools developed by the MICA Lab. These tools include micapipe (a fully automated toolbox for the processing and fusion of multimodal neuroimaging and connectomics data), BrainSpace (a toolbox for the mapping of spatial gradients in cortical organization), BrainStat (a Matlab/Python toolbox for the statistical analysis of univariate and multivariate surface/volume/parcel data and for multiscale neural contextualization), BigBrainWarp (integration of 3D histology with neuroimaging and other neurobiological modalities), Enigma Toolbox (a centralized, and continuously updated, repository of meta-analytical case-control comparisons across a wide range of disorders), and HippoMaps (an open data warehouse and toolbox for contextualization of hippocampal findings against post-mortem histology, in vivo imaging, and electrophysiology). The MICA Lab has also released open-access datasets including the (multimodal neuroimaging data acquired in 50 healthy control participants for study of multiscale brain organization) and the dataset (multimodal precision neuroimaging data acquired at ultra-high field strength of 7T).
