Identification of oscillatory brain networks with Hidden Gaussian Graphical Spectral models of EEG/MEG

Paz-Linares, Deirel; González-Moreira, Eduardo; Gonzalez, Ariosky Areces; Wang, Ying; Li, Min; Martı́nez-Montes, Eduardo; Bringas‐Vega, Maria L.; Valdés-Sosa, Mitchell; Valdés‐Sosa, Pedro A.

doi:10.36227/techrxiv.20254563.v1

Cited by 3 publications

(6 citation statements)

References 87 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As it is shown in (Paz-Linares et al, 2018), the unbiased estimator for HIGGS (𝚯 ̂𝜾𝜾 (𝑘+1) (𝜈)) follows the distribution:…”

Section: Figure 2 Structural Processing With Ciftify and Hcp Pipelinementioning

confidence: 97%

“…2.2 Ciftify-bcVARETA pipeline: Spatially resolved MEG/EEG source spectral networks mapping with HCP compatibility using sparse and surface-based structural priors. We incorporate novel methods designed to minimize the distortions of specific statistical quantities to be estimated by directly introducing mathematical priors (Paz-Linares et al, 2018;Gonzalez-Moreira et al, 2020). This style contrast with the conventional two-step approach used by HCP pipelines:…”

Section: Figure 2 Structural Processing With Ciftify and Hcp Pipelinementioning

confidence: 99%

“…Structured Sparse Bayesian Learning (sSSBL) method(Gonzalez-Moreira et al, 2020;Paz-Linares et al, 2023) and 4 (connectivity level analysis) the Hidden Gaussian Graphical State-Model (HIGGS) with connectivity regularization of the Hermitian Graphical LASSO (hgLASSO) method(Paz-Linares et al, 2018).…”

mentioning

confidence: 99%

See 2 more Smart Citations

Multimodal pipeline for HCP-compatible processing and registration of legacy datasets (MRI, MEG, and EEG)

Gonzalez¹,

Paz-Linares²,

Riaz³

et al. 2023

Preprint

View full text Add to dashboard Cite

<p>Extracting cortical features, which are the most relevant at characterizing structure and function for normal or abnormal brain conditions, would greatly benefit from multimodal neuroimage processing following the surface-based style. This style recognizes the natural definition space for such features due to the layered (surface-based) Cortex structural and functional organization. It may therefore be more sensitive and specific than the former volume-based style. The Human Connectome Project (HCP) multimodal pipelines render high-quality surface-based processing for some of the most consistently acquired neuroimaging modalities, with the quality too reliant on their precise acquisition requirements. Relevant international brain initiatives are espoused to develop an HCP-compatible neuroinformatic facility for the quality-ensured processing of international neuroimaging datasets, which may not follow the specific HCP acquisition requirements, also coined as legacy datasets. We appointed some initiatives to introduce multimodal pipelines in two HCP-compatible processing branches. a) Structural: forward- modeling with geometry (sources and head) and Lead Fields defined for legacy MEG, or EEG, in the HCP individual cortical space (Cifti) obtained from legacy MRI. Our pipeline (Ciftify-MEEG) leverages a more diverse neuroinformatic repository than the HCP structural or MEG pipelines. Ciftify-MEEG produces substantial processing illustrated here with EEG examples, incorporating alternative processing paths and corrections based on a quality control loop and upon qualitative and quantitative indicators. b) Functional: identifying spectral topographies and connectomes for the cortical oscillatory activity observed in the MEG or EEG frequency bands. We leverage a novel repository of Bayesian sparse inverse methods that target identifying the topographies and connectomes with actual statistical guarantees, the brain connectivity Variable Resolution Electromagnetic Tomographic Analysis (BC-VARETA). Our pipeline design for BC-VARETA, which we denominate Ciftify-bcVARETA, is integrated into the Ciftify- MEEG outputs and with Bayesian sparse priors structured upon Cifti space information. Ciftify-bcVARETA identification is less biased to forward models and more biased to the observations than the HCP pipeline illustrated here with topographies obtained for MEG and EEG legacy databases.</p>

show abstract

“…As it is shown in (Paz-Linares et al, 2018), the unbiased estimator for HIGGS (𝚯 ̂𝜾𝜾 (𝑘+1) (𝜈)) follows the distribution:…”

Section: Figure 2 Structural Processing With Ciftify and Hcp Pipelinementioning

confidence: 97%

Section: Figure 2 Structural Processing With Ciftify and Hcp Pipelinementioning

confidence: 99%

See 1 more Smart Citation

Multimodal pipeline for HCP-compatible processing and registration of legacy datasets (MRI, MEG, and EEG)

Gonzalez¹,

Paz-Linares²,

Riaz³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…In addition, we did not consider the vectorized (entry-wise) p-norm (Ding et al, 2006 ). The latter, which might be necessary to regularize off-diagonal entries in some cases (Paz-Linares et al, 2018 ), failed to ameliorate ill-condition or distortions.…”

Section: Novel Cesi Theory Leading To the Sssbl Approximationmentioning

confidence: 99%

“…However, if not impossible in a realistic cESI setup, such inverse solutions are N-P hard and therefore require Approximated Bayesian Computation (ABC) (Csilléry et al, 2010 ). This issue has been discussed in detail in the Bayesian literature (Dempster et al, 1977 ; Liu and Rubin, 1994 ; Daunizeau and Friston, 2007 ; Friston et al, 2007 ; Nummenmaa et al, 2007 ; Friston K. J. et al, 2008 ; Paz-Linares et al, 2018 ).…”

Section: Introductionmentioning

confidence: 99%

Minimizing the distortions in electrophysiological source imaging of cortical oscillatory activity via Spectral Structured Sparse Bayesian Learning

et al. 2023

Self Cite

View full text Add to dashboard Cite

Oscillatory processes at all spatial scales and on all frequencies underpin brain function. Electrophysiological Source Imaging (ESI) is the data-driven brain imaging modality that provides the inverse solutions to the source processes of the EEG, MEG, or ECoG data. This study aimed to carry out an ESI of the source cross-spectrum while controlling common distortions of the estimates. As with all ESI-related problems under realistic settings, the main obstacle we faced is a severely ill-conditioned and high-dimensional inverse problem. Therefore, we opted for Bayesian inverse solutions that posited a priori probabilities on the source process. Indeed, rigorously specifying both the likelihoods and a priori probabilities of the problem leads to the proper Bayesian inverse problem of cross-spectral matrices. These inverse solutions are our formal definition for cross-spectral ESI (cESI), which requires a priori of the source cross-spectrum to counter the severe ill-condition and high-dimensionality of matrices. However, inverse solutions for this problem were NP-hard to tackle or approximated within iterations with bad-conditioned matrices in the standard ESI setup. We introduce cESI with a joint a priori probability upon the source cross-spectrum to avoid these problems. cESI inverse solutions are low-dimensional ones for the set of random vector instances and not random matrices. We achieved cESI inverse solutions through the variational approximations via our Spectral Structured Sparse Bayesian Learning (ssSBL) algorithm https://github.com/CCC-members/Spectral-Structured-Sparse-Bayesian-Learning. We compared low-density EEG (10–20 system) ssSBL inverse solutions with reference cESIs for two experiments: (a) high-density MEG that were used to simulate EEG and (b) high-density macaque ECoG that were recorded simultaneously with EEG. The ssSBL resulted in two orders of magnitude with less distortion than the state-of-the-art ESI methods. Our cESI toolbox, including the ssSBL method, is available at https://github.com/CCC-members/BC-VARETA_Toolbox.

show abstract

EEG functional connectivity as a Riemannian mediator: An application to malnutrition and cognition

Lopez Naranjo,

Razzaq,

et al. 2024

Human Brain Mapping

Self Cite

View full text Add to dashboard Cite

Mediation analysis assesses whether an exposure directly produces changes in cognitive behavior or is influenced by intermediate “mediators”. Electroencephalographic (EEG) spectral measurements have been previously used as effective mediators representing diverse aspects of brain function. However, it has been necessary to collapse EEG measures onto a single scalar using standard mediation methods. In this article, we overcome this limitation and examine EEG frequency‐resolved functional connectivity measures as a mediator using the full EEG cross‐spectral tensor (CST). Since CST samples do not exist in Euclidean space but in the Riemannian manifold of positive‐definite tensors, we transform the problem, allowing for the use of classic multivariate statistics. Toward this end, we map the data from the original manifold space to the Euclidean tangent space, eliminating redundant information to conform to a “compressed CST.” The resulting object is a matrix with rows corresponding to frequencies and columns to cross spectra between channels. We have developed a novel matrix mediation approach that leverages a nuclear norm regularization to determine the matrix‐valued regression parameters. Furthermore, we introduced a global test for the overall CST mediation and a test to determine specific channels and frequencies driving the mediation. We validated the method through simulations and applied it to our well‐studied 50+‐year Barbados Nutrition Study dataset by comparing EEGs collected in school‐age children (5–11 years) who were malnourished in the first year of life with those of healthy classmate controls. We hypothesized that the CST mediates the effect of malnutrition on cognitive performance. We can now explicitly pinpoint the frequencies (delta, theta, alpha, and beta bands) and regions (frontal, central, and occipital) in which functional connectivity was altered in previously malnourished children, an improvement to prior studies. Understanding the specific networks impacted by a history of postnatal malnutrition could pave the way for developing more targeted and personalized therapeutic interventions. Our methods offer a versatile framework applicable to mediation studies encompassing matrix and Hermitian 3D tensor mediators alongside scalar exposures and outcomes, facilitating comprehensive analyses across diverse research domains.

show abstract

Identification of oscillatory brain networks with Hidden Gaussian Graphical Spectral models of EEG/MEG

Cited by 3 publications

References 87 publications

Multimodal pipeline for HCP-compatible processing and registration of legacy datasets (MRI, MEG, and EEG)

Multimodal pipeline for HCP-compatible processing and registration of legacy datasets (MRI, MEG, and EEG)

Minimizing the distortions in electrophysiological source imaging of cortical oscillatory activity via Spectral Structured Sparse Bayesian Learning

EEG functional connectivity as a Riemannian mediator: An application to malnutrition and cognition

Contact Info

Product

Resources

About