Article Figures & Data
Figures
- Figure 1:
Sources of metal in electrode implants. A, Tungsten electrodes. B, Silver paint to connect tungsten wires to gold pads on the circuit board. C, Solder to connect the Omnetics connector to the circuit board.
- Figure 2:
Schematic of micro-CT to MRH registration procedure.
- Figure 3:
Micro-CT images of implanted rat brains. A, 3D reconstructions illustrate 3D structure of electrode bundles and head implants. B, Examples of CT cross sections without (left) and with (right) artifact reduction applied. C, Eighty-eight micrometer voxels do not permit clear differentiation of individual wires within electrode bundles (2-4 individual wires are contained in each pictured bundle, despite their appearance as singular objects), but separate bundles are easily visualized (left), as are individual wires that are severely misplaced or bent (right).
- Figure 4:
Example of registration between individual CT scans and the MRH atlas. A, Coronal slice of MRH atlas. B, Corresponding slice in unregistered CT. C, Corresponding slice in registered CT.
- Figure 5:
Histological validation of CT to MRH atlas registration. Column 1, CT image. Column 2, Registered electrode mask from CT overlaid on MRH atlas. Column 3, Corresponding brain slice stained with cresyl violet. Row A, Orbitofrontal cortex bundle and top of insula bundle (both in correct location) overlaid on the b0 image from the MRH atlas. Row B, Anterior insula bundle (correct location) overlaid on the b0 image from the MRH atlas. Row C, Olfactory amygdala bundle (landed on the very lateral edge of intended area) overlaid on the gradient recalled echo image from the MRH atlas. In each row, the atlas contrast image that best highlighted the soft tissue architecture of the brain area around the electrode was chosen. For the examples illustrated in Rows A and C, the physical brain slices used for histology were not cut perfectly perpendicular to the anterior–posterior axis. Highlighting the advantages of 3D image volumes, the registered CT overlaid on the MRH was resliced to match the plane of the histology slice shown in Column 3 (as illustrated).
- Figure 6:
Results of electrode and head implant simulations. A, Schematic of simulations. B, Illustration of 977 fiducials on a glass brain separated into four labeled segments. C, Box plots of the mean and SD of Euclidean distance between all fiducials on the registered (after both rigid and nonrigid registration) nonimplanted brain and registered, simulated implanted brains across all 1000 simulations. Red line, median of all 1000 simulations. Box limits represent the 1st and 3rd quartiles. Whiskers represent the 5th and 95th percentiles. D, Percentage of 977,000 fiducials (the 977 fiducials that landed within the confines of the brain across 1000 simulations) whose Euclidean distance between the registered simulated implanted brain and nonimplanted brain was reduced, increased, or unchanged by the nonrigid registration step compared with rigid registration alone (pie charts). Histograms of the absolute value of the change in error (in Euclidean distance) caused by the nonrigid registration step (stacked bar charts; unchanged fiducials excluded) in the 977,000 fiducials across all imputations. Note that bin sizes increase after the dashed line.
- Figure 7:
Electrode bundles overlaid on segmentation atlas. A, Coronal brain slice of two amygdala bundles overlaid on the segmentation atlas and gradient recalled echo atlas image. The colors of the segmentation atlas are as follows: orange, amygdala; purple, hippocampus; yellow, caudate/putamen; blue, corpus callosum. B, 3D rendering of all bundles overlaid on a glass brain with the amygdala (orange) and hippocampus (purple) from the segmentation atlas shown.
Tables
Brain region Number of electrodes Coordinates (mm from bregma and top of brain) A/P M/L D/V Bilateral anterior cingulate 2 per hemisphere 2.5 ±.5 1.6 Bilateral orbitofrontal cortex 2 per hemisphere 3.7 ±2 3.8 Bilateral anterior insula 4 per hemisphere 2.2 ±4 4.5 Bilateral basolateral amygdala 4 per hemisphere -3 ±5.0 7.5 Bilateral olfactory amygdala 3 per hemisphere -1.4 ±3.2 8.8 - Table 2.
Goal and characteristics of statistical tests used to evaluate electrode configuration simulations
Goal Distribution of data Type of test Power of test aTest differences in fiducial movement between brain segments after rigid registration only Highly skewed right Kruskal–Wallis χ2 = 3.00e+05 bTest differences in fiducial movement between brain segments after rigid registration followed by nonrigid registration Highly skewed right Kruskal–Wallis χ2 = 2.33e+05 cTest differences in fiducial movement after rigid registration and rigid registration followed by nonrigid registration Highly skewed right Mann–Whitney U Z value = 30.4165
Rank sum = 1.01e+12
In this issue
