| T O P I C R E V I E W |
| syam.gadde |
Posted - Mar 16 2006 : 3:33:40 PM
There was a bug in bxh_eventstats that is only triggered when
using multiple runs that have different numbers of timepoints.
The nature of this bug is that it may have calculated the "data
range" time interval incorrectly for any run having different
numbers of timepoints than the first run, where the "data range"
time interval refers to the time interval between the first and
last volumes in the data. The impact is that events may have
been incorrectly excluded from the analysis if their epochs were
determined to be outside this incorrectly-calculated time interval.
Only those who ran bxh_eventstats on multiple 4-D time series
with different numbers of time points (e.g. run1 has 100 volumes,
run2 has 150 volumes, etc.) between Feb. 23 and March 16 should
be affected. I don't expect many people to have this problem.
If you fall into this category, check your <prefix>_LOG_EPOCHAVG.txt
file and look for entries that look like this:
Ignoring event at time 436.501 because the following ranges
fall outside data range: (0 - 453 secs)
Double-check that the "data range" specified looks correct.
This should normally be 0 to (N+1)*TR seconds where N is the
number of time points. |
|
|
