Influences of skull segmentation inaccuracies on EEG source analysis

Lanfer, Benjamin; Scherg, Michael; Dannhauer, Moritz; Knösche, Thomas R.; Burger, Martin; Wolters, Carsten H.

doi:10.1016/j.neuroimage.2012.05.006

Cited by 101 publications

(129 citation statements)

References 49 publications

Supporting

Mentioning

124

Contrasting

Order By: Relevance

“…Errors in skull segmentation are known to cause substantial issues in source localization [45]. The skull is actually composed of three layers: one layer of spongy bone encased by two layers of more compact bone, each with distinct conductivities [46].…”

Section: Discussionmentioning

confidence: 99%

A finite-element reciprocity solution for EEG forward modeling with realistic individual head models

et al. 2014

View full text Add to dashboard Cite

We present a finite element modeling (FEM) implementation for solving the forward problem in electroencephalography (EEG). The solution is based on Helmholtz's principle of reciprocity which allows for dramatically reduced computational time when constructing the leadfield matrix. The approach was validated using a 4-shell spherical model and shown to perform comparably with two current state-of-the-art alternatives (OpenMEEG for boundary element modeling and SimBio for finite element modeling).We applied the method to real human brain MRI data and created a model with five tissue types: white matter, grey matter, cerebrospinal fluid, skull, and scalp. By calculating conductivity tensors from diffusion-weighted MR images, we also demonstrate one of the main benefits of FEM: the ability to include anisotropic conductivities within the head model. Root-mean square deviation between the standard leadfield and the leadfield including white-matter anisotropy showed that ignoring the directional conductivity of white matter fiber tracts leads to orientation-specific errors in the forward model. Realistic head models are necessary for precise source localization in individuals. Our approach is fast, accurate, open-source and freely available online.

show abstract

Section: Discussionmentioning

confidence: 99%

A finite-element reciprocity solution for EEG forward modeling with realistic individual head models

et al. 2014

View full text Add to dashboard Cite

show abstract

“…The reference method is a combination of the BET and FAST tools commonly used for the creation of patient-specific conductivity models [5,12,13]. It takes the same inputs as the proposed HSA.…”

Section: Reference Method: Bet-fastmentioning

confidence: 99%

“…Nevertheless, it is laborious and time consuming. Two refinements that can be found in the literature are semi-automatic/interactive segmentation requiring user input [6,[8][9][10][11] and fully automatic segmentation followed by manual correction [5,12,13]. However, the need for user interaction or manual correction means that these approaches are still time consuming, subjective, tedious, and labor intensive.…”

Section: Introductionmentioning

confidence: 99%

“…In this paper, we present a new fully automatic method for head tissue segmentation from multimodal MR images suitable for EEG source localization. In common with several approaches in the literature [5,12,13], the method involves first partitioning the MR data into braintissue and non-brain-tissue sub-volumes and then independently segmenting each of these into multiple tissue classes. The idea behind this hierarchical segmentation approach (HSA) is that the detection of brain and non-brain tissue is a much simpler initial problem than the problem of segmenting the whole head into multiple tissue classes outright.…”

Section: Introductionmentioning

confidence: 99%

“…Each HSA instantiation is based on a different multi-tissue segmentation algorithm: the hidden Markov random field model and associated expectation-maximization (HMRF-EM) algorithm [17] (as implemented in the FAST in the FSL tools), the adaptive mean-shift (AMS) algorithm of Mayer and Greenspan [18], the improved Fuzzy c-means algorithm with spatial constraints (FCM_S), proposed in [19] and the simple k-means clustering algorithm [20]. The reference method BET-FAST is based on the BET [14,21] and FAST [17,21] tools commonly used [5,12,13] in the construction of realistic head models for EEG source localization. The BET tool is used with extended functionality to extract not only the brain-tissue mask but also the skin-and skulltissue masks (ideally both T1w and T2w images should be used although it is possible to use T1w only) and the FAST tool is employed to segment the brain tissues.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Unsupervised Segmentation of Head Tissues from Multi-modal MR Images for EEG Source Localization

et al. 2014

View full text Add to dashboard Cite

In this paper, we present and evaluate an automatic unsupervised segmentation method, hierarchical segmentation approach (HSA)-Bayesian-based adaptive mean shift (BAMS), for use in the construction of a patient-specific head conductivity model for electroencephalography (EEG) source localization. It is based on a HSA and BAMS for segmenting the tissues from multi-modal magnetic resonance (MR) head images. The evaluation of the proposed method was done both directly in terms of segmentation accuracy and indirectly in terms of source localization accuracy. The direct evaluation was performed relative to a commonly used reference method brain extraction tool (BET)-FMRIB's automated segmentation tool (FAST) and four variants of the HSA using both synthetic data and real data from ten subjects. The synthetic data includes multiple realizations of four different noise levels and several realizations of typical noise with a 20 % bias field level. The Dice index and Hausdorff distance were used to measure the segmentation accuracy. The indirect evaluation was performed relative to the reference method BET-FAST using synthetic two-dimensional (2D) multimodal magnetic resonance (MR) data with 3 % noise and synthetic EEG (generated for a prescribed source). The source localization accuracy was determined in terms of localization error and relative error of potential. The experimental results demonstrate the efficacy of HSA-BAMS, its robustness to noise and the bias field, and that it provides better segmentation accuracy than the reference method and variants of the HSA. They also show that it leads to a more accurate localization accuracy than the commonly used reference method and suggest that it has potential as a surrogate for expert manual segmentation for the EEG source localization problem.

show abstract

Effects of uncertainty in head tissue conductivity and complexity on EEG forward modeling in neonates

Azizollahi

Aarabi

Wallois

2016

Human Brain Mapping

View full text Add to dashboard Cite

In this study, we investigated the impact of uncertainty in head tissue conductivities and inherent geometrical complexities including fontanels in neonates. Based on MR and CT coregistered images, we created a realistic neonatal head model consisting of scalp, skull, fontanels, cerebrospinal fluid (CSF), gray matter (GM), and white matter (WM). Using computer simulations, we investigated the effects of exclusion of CSF and fontanels, discrimination between GM and WM, and uncertainty in conductivity of neonatal head tissues on EEG forward modeling. We found that exclusion of CSF from the head model induced the strongest widespread effect on the EEG forward solution. Discrimination between GM and white matter also induced a strong widespread effect, but which was less intense than that of CSF exclusion. The results also showed that exclusion of the fontanels from the neonatal head model locally affected areas beneath the fontanels, but this effect was much less pronounced than those of exclusion of CSF and GM/WM discrimination. Changes in GM/WM conductivities by 25% with respect to reference values induced considerable effects in EEG forward solution, but this effect was more pronounced for GM conductivity. Similarly, changes in skull conductivity induced effects in the EEG forward modeling in areas covered by the cranial bones. The least intense effect on EEG was caused by changes in conductivity of the fontanels. Our findings clearly emphasize the impact of uncertainty in conductivity and deficiencies in head tissue compartments on modeling research and localization of brain electrical activity in neonates. Hum Brain Mapp 37:3604-3622, 2016. © 2016 Wiley Periodicals, Inc.

show abstract

Influences of skull segmentation inaccuracies on EEG source analysis

Cited by 101 publications

References 49 publications

A finite-element reciprocity solution for EEG forward modeling with realistic individual head models

A finite-element reciprocity solution for EEG forward modeling with realistic individual head models

Unsupervised Segmentation of Head Tissues from Multi-modal MR Images for EEG Source Localization

Effects of uncertainty in head tissue conductivity and complexity on EEG forward modeling in neonates

Contact Info

Product

Resources

About