Connectivity-Based Simulations for Validating Connectivity-Informed Electrical Source Imaging
Poster No:
1341
Submission Type:
Abstract Submission
Authors:
Nino Herve1, Nicolas Roehri2, Yasser Alemán-Gómez3, Emma Depuydt4, Thomas Sanchez5, Patric Hagmann6
Institutions:
1UNIL / CHUV, Lausanne, Switzerland, 2University of Geneva, Geneva, Geneva, 3Centre hospitalier universitaire vaudois (CHUV), Lausanne, VT, 4Ghent University, Ghent, Belgium, 5CIBM / CHUV-UNIL, Lausanne, Switzerland, 6Lausanne University Hospital and University of Lausanne (CHUV-UNIL), Lausanne, Vaud
First Author:
Co-Author(s):
Introduction:
Connectome Spectrum Electromagnetic Tomography (CSET) is a novel EEG source reconstruction method that incorporates brain connectivity through the structural connectome (Rué-Queralt J., 2024). This dipole-distributed model regularizes the inverse problem by minimizing the L1 norm of the graph Fourier transform of brain activity, where the graph represents the brain's connectome. Recent studies show that brain signals are sparse in the graph Fourier domain, motivating this approach (Glomb K., 2020). By leveraging this sparsity, CSET provides a more physiologically informed prior compared to traditional methods, which often rely on smoothing or focalization to ensure a unique solution. To validate CSET, we introduce a novel EEG simulation framework with a dynamic source propagating through the brain's connectome, unlike conventional simulations using fixed point sources (Hauk O., 2022).
Methods:
We selected a random subject from the VEPCON dataset (OpenNeuro Dataset ds003505) as the template (Pascucci, 2022). This dataset includes T1-weighted MRI images, tractograms, and task-based EEG recordings. We reconstructed a source space (~5000 dipoles distributed uniformly over the cortical surface), a dipole-based connectome, a BEM head model, and an average reference lead-field matrix as described by Rué-Queralt et al. (2024). The brain surfaces were parcellated using a scale-4 atlas (Cammoun et al., 2012).
To generate the ground truth, a Mexican-hat signal (1s) was applied to dipoles within a selected brain region and propagated through the connectome (Figure 1). Realistic source data were simulated by adding Brownian noise (step size 0.1 nA) filtered with a high-pass filter (1 Hz) and white noise (0.1 µV standard deviation) after projection to the scalp. The source signal was scaled to achieve a targeted EEG signal-to-noise ratio (SNR), measured over 1s.
We conducted 100 simulations with different source locations uniformly distributed across the brain surface, applying both CSET (Rué-Queralt J., 2024) and sLORETA (Pascual-Marqui R.D., 2002) for source estimation. We evaluated source localization error (distance between maximum reconstructed activity and true source location) at both the 50% rising and peak phases and assessed the correlation with the ground truth at the peak phase (Figure 2).
To generate the ground truth, a Mexican-hat signal (1s) was applied to dipoles within a selected brain region and propagated through the connectome (Figure 1). Realistic source data were simulated by adding Brownian noise (step size 0.1 nA) filtered with a high-pass filter (1 Hz) and white noise (0.1 µV standard deviation) after projection to the scalp. The source signal was scaled to achieve a targeted EEG signal-to-noise ratio (SNR), measured over 1s.
We conducted 100 simulations with different source locations uniformly distributed across the brain surface, applying both CSET (Rué-Queralt J., 2024) and sLORETA (Pascual-Marqui R.D., 2002) for source estimation. We evaluated source localization error (distance between maximum reconstructed activity and true source location) at both the 50% rising and peak phases and assessed the correlation with the ground truth at the peak phase (Figure 2).
Results:
Figure 2A illustrates that the estimated source location is more accurate during the 50% rising phase than at the peak phase. Specifically, CSET reduced the localization error from 1.4 cm to 0.9 cm, while sLORETA improved from 1.1 cm to 0.5 cm. Figure 2B shows that CSET exhibits a higher correlation with the ground truth at the peak phase (0.39 vs. 0.33 on average). Figure 2C presents examples of the ground truth, CSET, and sLORETA at both the 50% rising and peak phases. These images emphasize that the source is clearly identifiable during the 50% rising phase, whereas at the peak phase, it has spread to its 1-hop neighbors. CSET more accurately captures this 1-hop neighborhood, while sLORETA becomes too blurry to resolve it clearly.
Conclusions:
This simulation incorporates connectivity as a key component, providing a novel framework for studying electrical source imaging with source propagation. CSET specifically leverages connectivity for more accurate reconstructions. Our results align with prior research suggesting that localizing sources during the peak phase is suboptimal, as the source has dispersed by that point (Göran L., 2003). Instead, the 50% rising phase offers a more accurate representation of source localization, a factor often underreported in traditional simulations. In this example, CSET closely approximates the ground truth at the peak phase, capturing one-hop neighborhood propagation more effectively than sLORETA (Figure 2B, second row). Future work could evaluate source reconstruction at different noise levels and examine CSET's performance when the connectome used for reconstruction is degraded from its true state.
Brain Stimulation:
Non-invasive Electrical/tDCS/tACS/tRNS
Modeling and Analysis Methods:
EEG/MEG Modeling and Analysis 1
Novel Imaging Acquisition Methods:
EEG 2
Keywords:
Computational Neuroscience
Electroencephaolography (EEG)
Source Localization
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
Which processing packages did you use for your study?
Provide references using APA citation style.
Glomb K. 2020, Connectome spectral analysis to track EEG task dynamics on a subsecond scale. NeuroImage, Volume 221, Pages 117-137
Göran L. 2003, Propagation of Interictal Epileptiform Activity Can Lead to Erroneous Source Localizations: A 128-Channel EEG Mapping Study. Journal of Clinical Neurophysiology 20(5):p 311-319
Hauk O. 2022, Towards an objective evaluation of EEG/MEG source estimation methods – The linear approach, NeuroImage, Volume 255
Pascual-Marqui. R.D. 2002, Standardized low resolution brain electromagnetic tomography (sLORETA): technical details. Methods & Findings in Experimental & Clinical Pharmacology 24D:5-12.
Pascucci, D. 2022, Source imaging of high-density visual evoked potentials with multi-scale brain parcellations and connectomes. Scientific Data 9:9
Rué-Queralt J. 2024, Connectome spectrum electromagnetic tomography: A method to reconstruct electrical brain source networks at high-spatial resolution. Hum Brain Mapping. 45(5).
Uutela K. 1999, Visualization of Magnetoencephalographic Data Using Minimum Current Estimates, NeuroImage, Volume 10, Issue 2, Pages 173-180