Table 1.
Litterature review, after 2005.
Fig 1.
Analysis pipeline describing dMRI, fMRI and EEG analysis.
Fig 2.
Templates of OR (green) and splenium (red).
The difference between deterministic + DTI tracking (paler) and HARDI + probabilistic (darker) was major, mainly in the Meyer’s loop. Probabilistic template was used.
Fig 3.
Templates (on the left) and fiber comparison with MDF (on the right).
Each streamline of the subject (white) is kept if it resembles the OR template (green), rejected otherwise (cases were it resembles the splenium template (red)).
Fig 4.
Cross section areas (CSAs) were defined as the number of voxels where passed at least 5 streamlines in each section. Mean and minimum CSA of sections 4-15, and mean and maximum CSA of sections 16-18 were measured.
Fig 5.
Blue: thalamic masks as defined by FSL. Green: voxels where FA ∈[0.1, 0.5]. Background: the FA map.
Fig 6.
Three subject’s final OR, axial view.
Table 2.
Cross-sectional areas (in number of voxels).
Table 3.
FA and AFD under the OR.
Fig 7.
Effect of the thresholding on the ORs.
In the thalamus: in blue, the final OR, voxels crossed by at least 5 fibers and in red, with a threshold on the z-score of 4. In the optic radiation: density map, in number of streamlines. On the left: z = 64. On the right: z = 69.
Table 4.
Position of the strongest density point, in ICBM space.
Table 5.
EEG results.
Fig 8.
Validation of LGN segmentation.
(left) The red/yellow overlay represents fMRI activity during the EC-EO task. Note the widespread activation in the visual cortex and more focal bilateral activation in the thalamus (LGN). (right) An enhanced view shows that our reconstructed streamlines converge nicely onto the fMRI activation sites (outline in white).
Table 6.
p-values for all combinations between EEG data and structural data.
Smallest value was 0.06.