1: 3-D visualization of depth Electrodes in patient BI18 (a) Anterior view showing all electrodes with respect to the ventricles (colored in blue) (b) Same set of electrodes shown with respect to the ventricles (colored in blue) in a side view from the left (c) Left sided electrodes seen in a sagittal cut of the skull showing the left side of the skull as viewed from the right (d) Right sided electrodes seen in a sagittal cut of the skull showing the right side of the skull as viewed from the left (LA: Left Amygdala, LH: Left Hippocampus, LS: Left Supplementary Motor, LC: Left Cingulate, LO: Left Orbitofrontal, RH: Right Hippocampus, RO: Right Orbito-frontal, RS: Right Supplementary Motor, RC: Right Cingulate) 

Contexts in source publication

Context 1

... between the model and experimental data was identical at ~25 % for isotropic models ISO_II and ISO_III (see electrode LS from stimulation of RS12 drops from 21% for ISO_I to ~8% for both ISO_II and ISO_III (see Figure 7.16(e)) Similarly, for stimulation at RC12 (also near the inter-hemispherical CSF) the RDM at LS drops from a large 62% error generated by ISO_I to ~ 17% generated by ISO_III (see Figure ...

Context 2

... between the model and experimental data was identical at ~25 % for isotropic models ISO_II and ISO_III (see electrode LS from stimulation of RS12 drops from 21% for ISO_I to ~8% for both ISO_II and ISO_III (see Figure 7.16(e)) Similarly, for stimulation at RC12 (also near the inter-hemispherical CSF) the RDM at LS drops from a large 62% error generated by ISO_I to ~ 17% generated by ISO_III (see Figure ...

Context 3

... accurate fits at all electrodes sites (<10% RDM). Similar improvements were also found at stimulation sites RO45 and RO56 but the RDM across all depth electrodes was lowest for site LC12. Figure 7.20 shows the fractional anisotropy is highest at contact 1 on electrode LC and is also closest to region of high anisotropy (corpus callosum). We now take the case of dipole at LC12 to study the differences in the current field between the isotropic and anisotropic model. This change of current direction in the superior region of the radiata corona could explain the better fits at electrode site LS using the anisotropic model for stimulation at site ...

Context 4

... RO56 (see Figure 7.13(c)). RDM's calculated over measurement electrodes on the same side as the dipole source show improved fits using ANISO_WM_I when compared to their isotropic counterparts. For example, RDM's over right side electrodes dropped from ~16%(ISO_III)/~21%(ISO_I) to ~5% (ANISO_WM_I) for dipole at RO23; reduction from ~27%(ISO_III)/~37%(ISO_I) to ~13%(ANISO_WM_I) for dipole at RO45 and reduction from ~37%(ISO_I)/~37%(ISO_III) to ~15%(ANISO_WM_I) for dipole at RO56. Although these stimulation locations had a relatively low FA, these sites were embedded in the white matter. Electrode RO had contacts 2, 3 and 4 embedded in the white matter and is close to anterior region of the corona radiata. This suggests that white matter present in the pathway of the currents between the stimulation and recording site does also influence the intracranial forward ...

Context 5

... anisotropy ratio of 10:1 between the conductivity along and perpendicular to the white matter fiber direction has been reported in literature [96]. This ratio has been used for conducting simulation studies of the effects of anisotropic white matter in the brain on EEG topography [132]. Anisotropic model ANISO_WM_II that uses this anisotropic ratio has the worst performance of all the models. This is not a surprising result, since the 10:1 ratio can be considered as an upper bound on the anisotropy ratio for white matter fibers. The failure of ANISO_WM_II in providing good fits confirms We present some of the limitations of this study. Anisotropy at all locations in the brain such as white matter in the subcortical structure is not included (subcortical region is labeled as a single isotropic element) which is erroneous since anisotropy is present in subcortical white matter. This could explain the lack of improvements by ANISO_WM_I when dipoles were generated in subcortical areas using electrode LA (see Figure 7.14 (a)). Human grey matter also has a reported anisotropy of 2:1 [96] which is not taken into account in this model.. This chapter also contains the results from one subject. There could be inter-subject variability in reported numbers but we do not expect the major findings of the study to change. i.e. anisotropic WM is expected to provide better fits for iEEG. Considered as a whole, these results support our hypothesis that a detailed description of intracranial tissue including inhomogeneity and anisotropy is necessary to generate accurate intracranial forward solutions. Incomplete models could possibly lead to higher errors when certain details (such as anisotropy) in the model (ISO_III) are ...

Context 6

... iEEG forward solution obtained using the isotropic and anisotropic model is now compared to the experimental data. Figure 7.12 plots the average RDM between model and experiment data with averaging across all stimulation sites between model and experiment data. (left side ...