| Author |
 Topic  |
|
|
josh.bizzell
BIAC Staff
USA
118 Posts |
 Posted - Feb 04 2005 : 4:10:49 PM
|
A few updates have recently been made to tstatprofile2, which will individually be discussed in posts below this one. This new version has been temporarily put in \\Gall\Programs\User_Scripts\beta and will be pushed to BIAC Matlab tools on the next update.
Below, I’ve included a description of all the fields that might be in a TSP2 control file.
Common Fields:
Bins – List of the event-coded “bins” the user wishes to analyze. If unspecified, the function will analyze all bins found in the specified column(s) of the paradigm files.
BinsTest – If the user wants to do T-Tests between two different bins, this parameter must be set. For example, to run a T-Test comparing bins 1 and 2, the user would enter: BinsTest = [1,2].
PreEventPts – Number of time points in the epoch before the event.
PostEventPts – Number of time points in the epoch after the event.
Base1 – The time point in the epoch relative to the event onset that defines the start of the baseline. If the baseline starts five time points before the event, the user would enter Base1 = -5. The default value is 0, which represents the same time point as the event onset.
Base2 – The time point in the epoch relative to the event onset that defines the end of the baseline. This value cannot be less than Base1, and the default value is 0.
Threshhold – Pixel intensity that defines the boundary between background and brain.
CorrelationTemplate – File name and path of the hemodynamic response template.
OutPath – Path where all output analysis data will be stored.
ParadigmColumn – List of the columns in the paradigm file to analyze.
RunName – File name and path of each functional run to be analyzed.
Paradigm – File name and path of each paradigm file per run. If each run uses the same paradigm file, this parameter can be entered just once.
Required unless reading BXH files:
XPixels – Number of pixels in the X-plane.
YPixels – Number of pixels in the Y-plane.
ZPixels – Number of slices.
TimePts – Number of time points per run. If each run has the same number of time points, this parameter can be entered just once.
RunDataType – The input data type to be used by READMR. Default value is ‘volume’.
Optional Fields
SubjectNumber – The study code assigned by the scanner. This is used when naming the average and standard deviation volumes, and if it is not specified, a default filename will be used. (e.g. 41111.av will become grandAvg.img).
FilenameType – Describes how the output files will be named to reflect which bin they represent. FilenameType = Normal will use the integer bin codes to name output files, e.g. Bin01.T. FilenameType = Unique will use unique identifiers described in the paradigm files (see explanation below) or bin/column identifiers to name the output files; e.g. BinRed.T or BinB001C01.T. Default is Normal.
Standard – The bin code that represents your standard/regular time points when no “events” occur. This can be specified once for all runs, or per run. Default value is 0.
OutFlags - An eleven-element array that specifies which output files (see output file descriptions below) should not be written. A 0 tells the function to write the file, a 1 tells the function not to write the file. This option is used to save disk space. The order of the array is as follows:
[*.av *.sd *.avb *.sdb Avg*.img Var*.img StdDev*.img Zscore*.img *.N *.COR *.T]
Default is [0,0,0,0,0,0,0,0,0,0,0], all output files will be written.
NOPROGRESSBAR – Set to 1 if you don’t want the progress bar. Default is 0.
USEBRAINONLY – Set to 1 if only the “brain” voxels should be analyzed; saves memory and time. Default is 0.
BrainThresh – Usebrainonly automatically detects the threshold to define as “brain”. If you want to explicitly set this value, use the BrainThresh optional field.
OUTPUT3D – 4D TSVs will be written as a series of 3D files if this field is set to 1. Default is 0.
FilterTR – Set the TR(s) to do temporal smoothing/filtering for each run to be analyzed. This is described further in posts below.
FindBadPoints – Set to 1 to use an algorithm to automatically detect intensity spikes and remove them from analysis. Default is 0. This is described further in posts below.
BadPointThresh – Number of standard deviations from the mean to call a “bad point” in the intensity spike calculation. Default is 3.
|
|
|
josh.bizzell
BIAC Staff
USA
118 Posts |
Posted - Feb 04 2005 : 4:16:44 PM
|
Tstatprofile2 - Temporal Filtering
Each run can now be filtered across time with a low-pass, high-pass, band-pass or stop band butterworth filter (5th order). To use this, you will need to set the FilterTR field in the control file. The values set in this field will be either one integer value (low-pass, high-pass) or two integer values (band-pass, stop-band) which are in units of TR. For example, let’s say the TR of your functional run is 1.5sec. The sampling rate of that run is 1/1.5sec, or 0.667 Hz. Thus, if you set FilterTR = 5 180, you’ll get a band-pass filter with the pass band between 1/(180*1.5sec) to 1/(5*1.5sec), or 3.70kHz to 133.33kHz.
Examples on using the different types of filters:
Low-pass filter:
FilterTR = 180 l
Or
FilterTR = 180 L
High-pass filter:
FilterTR = 5 h
Or
FilterTR = 5 H
Band-pass filter:
FilterTR = 5 180 b
Or
FilterTR = 5 180 B
Stop-band filter:
FilterTR = 20 30 s
Or
FilterTR = 20 30 S |
 |
|
|
josh.bizzell
BIAC Staff
USA
118 Posts |
Posted - Feb 04 2005 : 4:18:40 PM
|
Tstatprofile2 - Intensity Spike Detection/Rejection
This version of tstatprofile2 can remove intensity spikes from functional data. The algorithm for spike detection will find the mean value of the functional run and if the mean of a volume at a certain time point is a predetermined number of standard deviations from this mean, that time point is considered a “bad time point.”
To use this functionality, set the to FindBadPoints flag to one.
FindBadPoints = 1
The default number of standard deviations for a “bad” time point is three, but the user can set it by using the BadPointThresh field in the control file.
BadPointThresh = 2
NOTE: If using the temporal filter and the spike detection, the filtering will be done first and the results will be passed to the spike detector.
|
|
|
| |
Topic |
|
|
|
Tstatprofile2 Updates
