Hi Brandon,
unless there is a strong bias field in the image (a low frequency sigl
fluctuation) caused by receive/transmit characteristic of a coil, low
resolution and partial brain coverage should not be a problem. I have have
been able to successfully segmented hepatic/ovarian lesions in ADC maps
(manually pre-segmented from surrounding tissues). Provided you image is free
from strong bias field, you could have a look at image histogram to see, how
many significant intensities your image has.
Cheers,
Martin
On Thursday 12 June 2008 15:27:13 Brandon Whitcher wrote:
> Dear FSL users,
>
> I have a partial brain acquisition (T2-weighted at coarse resolution) and
> want to perform the usual 3-class tissue segmentation. I've run the data
> through FAST, but it produces a single CSF class that covers > 90% of the
> volume. Any tips/tricks on how to do this?
>
> thanks,
>
> Brandon
|