Biomedical Engineering Reference
In-Depth Information
FIGURE 8.1
Significant auditory activation clusters (
0.01) with no motion correction (top
row), standard motion correction (middle row), and motion correction with spin history
correction (bottom row).
Z
2.3,
P
3
method described by Friston et al.
In all cases, statistical thresholding was
carried out using the same significance threshold.
Figure 8.1 shows results derived from data taken during periods of
auditory activation (a conversational radio program) interleaved with
periods of “silence.” Each row shows three consecutive slices from one of
the original 3D fMRI brain images, overlaid with significant activation
areas. The top row shows clear tell-tale signs of uncorrected motion
around parts of the ventricles and little activation in the auditory cortex
(this is the area shown as activated within the circles in the bottom row).
The middle row shows the improvement in reported activation, but much
“noise” remains—i.e., presumed activation well outside the auditory cortex.
The bottom row shows that the spin history correction has completely
removed this noise but has also greatly reduced the apparent amount of
“true” activation.
Figure 8.2 shows two consecutive slices resulting from an fMRI experiment
during which brief periods of painful heat were applied to the hand. There was
much less overall head motion during the experiment than in the previous
case; therefore there is less difference among the three sets of results. However,
it is still possible (particularly within the marked circles) to see areas of activation
