NIH NIfTI - Neuroimaging Informatics Technology Initiative
AFNI is a set of C programs for processing, analyzing, and displaying functional MRI (FMRI) data - a technique for mapping human brain activity. It runs on Unix+X11+Motif systems, including SGI, Solaris, Linux, and Mac OS X. It is available free (in C source code format, and some precompiled binaries) for research purposes.AFNI is free software, and comes with no warranty or guarantee. There is no assurance that it works as documented, that it does anything remotely useful, or that it won't cause actual damage to the structure of the Universe.
The conditions for use and redistribution of AFNI are listed here.
The software changes frequently; if you use the AFNI package, you should read the Latest News Page or the SSCC Blog regularly to keep up with recent changes.
下面是一个例子：
[img]http://afni.nimh.nih.gov/afni/about/images/1mm.gif[/img]
Left: 1x1x10 mm**3 data. Right: 1x1x1 mm**3 data
At the start of the animation, the 10 mm thick slab is viewed from above. The color overlay that then appears represents the functional activation, with red indicating signal changes under 15% and yellow signal changes over 20%. The background images are rendered as partly transparent so that the color overlay can "shine through".
As the animation progresses, the viewpoint swings around to the side, showing the function from different angles. The background image then fades out completely, and the activated volumes are left hanging in space while the viewpoint swings back to the original top-down orientation.
On the left, the activated regions are crudely shaped along the inferior-superior axis, due to the low spatial resolution in that direction.
The color shows that the estimated signal change due to activation is smaller in the low resolution dataset. This is probably due to partial-volume effects: not all of the volume of a 1x1x10 mm**3 voxel will be filled with "activated" tissue. Since the signal from a voxel is averaged over its entire contents, if only a small portion of the voxel has a large signal change, the net measured result is a small signal change. With smaller voxels, the regions that are not active will be pruned away and the observed signal changes will be larger.
A confounding problem occurs at high resolution when the activation is over a large region. In MRI, the intrinsic signal-to-noise ratio (SNR) declines as the voxel size shrinks. For a fixed number of image acquisitions and for a fixed statistical threshold, lower SNR means that only larger signal changes can be detected. This effect can be overcome by acquiring more images. This effect is also partly mitigated by the fact that much of the interfering "noise" in the detection of FMRI signal changes is not truly MRI noise but is physiological in origin. This type of "noise" will decline as the voxel size shrinks.
用户界面例子：
[img]http://afni.nimh.nih.gov/afni/about/images/screen_shots/afniscreen_01.gif[/img]
More Detail and software Download:
[url]http://afni.nimh.nih.gov/afni/[/url]
页:
[1]
