Category Ranking

98%

Total Visits

921

Avg Visit Duration

2 minutes

Citations

20

Article Abstract

Reconstructing brain sources is a fundamental challenge in neuroscience, crucial for understanding brain function and dysfunction. Electroencephalography (EEG) signals have a high temporal resolution. However, identifying the correct spatial location of brain sources from these signals remains difficult due to the ill-posed structure of the problem. Traditional methods predominantly rely on manually crafted priors, missing the flexibility of data-driven learning, while recent deep learning approaches focus on end-to-end learning, typically using the physical information of the forward model only for generating training data. We propose the novel hybrid method 3D-PIUNet for EEG source localization that effectively integrates the strengths of traditional and deep learning techniques. 3D-PIUNet starts from an initial physics-informed estimate by using the pseudo inverse to map from measurements to source space. Secondly, by viewing the brain as a 3D volume, we use a 3D convolutional U-Net to capture spatial dependencies and refine the solution according to the learned data prior. Training the model relies on simulated pseudo-realistic brain source data, covering different source distributions. Trained on this data, our model significantly improves spatial accuracy, demonstrating superior performance over both traditional and end-to-end data-driven methods. Additionally, we validate our findings with real EEG data from a visual task, where 3D-PIUNet successfully identifies the visual cortex and reconstructs the expected temporal behavior, thereby showcasing its practical applicability.

Download full-text PDF

Source
http://dx.doi.org/10.1109/TMI.2025.3594724DOI Listing

Publication Analysis

Top Keywords

brain source
8
brain sources
8
deep learning
8
source
5
brain
5
data
5
enhancing brain
4
source reconstruction
4
reconstruction initializing
4
initializing neural
4

Similar Publications

FDG PET Findings in Rare Brain Sodium Channelopathy Associated with SCN2A Gene Mutation.

Clin Nucl Med

September 2025

Department of Nuclear Medicine & PET/CT, Mahajan Imaging & Labs.

SCN2A gene mutations, which affect the function of the voltage-gated sodium channel NaV1.2, are associated with a spectrum of neurological disorders, including epileptic encephalopathies and autism spectrum disorders. Advanced imaging modalities such as magnetic resonance imaging (MRI) and positron emission tomography (PET) have been instrumental in elucidating the neuroanatomic and functional alterations associated with these mutations.

View Article and Find Full Text PDF

Motivation: Drug repositioning presents a streamlined and cost-efficient way to expand the range of therapeutic possibilities. Drugs with human genetic evidence are more likely to advance successfully through clinical trials towards FDA approval. Single gene-based drug repositioning methods have been implemented, but approaches leveraging a broad spectrum of molecular signatures remain underexplored.

View Article and Find Full Text PDF

Central nervous system (CNS) diseases, including neurodegenerative diseases, stroke, brain tumors, and others, result in poor quality of life and can cause substantial disability. Not all CNS diseases are amenable to surgical approaches, so drug development is important for disease treatment. Unfortunately, there are few drugs currently available for CNS diseases.

View Article and Find Full Text PDF

Neuroinflammation within the central nervous system (CNS) is recognized as a critical pathological process in meningitic Escherichia coli (E. coli) infection, leading to severe neurodegenerative disorders and long-term sequelae. Astrocyte reactivity plays a pivotal role in driving the neuroinflammatory cascade in response to pathological stimuli from peripheral sources or other cellular components of the CNS.

View Article and Find Full Text PDF

Epilepsy, a highly individualized neurological disorder, affects millions globally. Electroencephalography (EEG) remains the cornerstone for seizure diagnosis, yet manual interpretation is labor-intensive and often unreliable due to the complexity of multi-channel, high-dimensional data. Traditional machine learning models often struggle with overfitting and fail in fully capturing the highdimensional, temporal dynamics of EEG signals, restricting their clinical utility.

View Article and Find Full Text PDF