• mkanalysis-sess.new

Clear message

Name

mkanalysis-sess.new

Synopsis

mkanalysis-sess.new -analysis [analysis] -TR [TR] -paradigm [paradigm.par] -designtype [design] [<options>]

Positional Arguments

None

Required Flagged Arguments

Optional Flagged Arguments

General

Event-Related and Blocked Design

Retinotopy and AB Blocked Design(Required)

Outputs

Description

mkanalysis-sess.new creates an 'analysis'. An analysis is a concept within the FsFast processing stream framework. The analysis is meant to be a placeholder for all the within-session processing options (eg, motion correction, spatial smoothing, event-type definition, etc). As such, the parameters apply to ALL sessions (ie, it does not need to be run separately for each session), and does not depend upon the data in any one session. This program should be run from your FsFast Study Directory. This program only configures the analysis; it does not analyze the data (see ["selxavg-sess"] or ["sfa-sess"]).

Detailed Descriptions of Options

analysis analysisname (Shortcut: -a)

This is the name given to the analysis. A folder called analysisname will be created in the Study Directory. Two files (analysis.info and analysis.cfg) will be created in this directory. The analysis parameters will be stored in these files. Contrast matrices (created by mkcontrast-sess) will also be stored in this directory. When the session data are analyzed with ["selxavg-sess"] or ["sfa-sess"] data, a folder called analysisname will be created in the functional subdirectory (FSD), and the results of the analysis will be placed there.

TR

This is the TR in seconds. It will be checked against the TR in each session when selxavg-sess is run. See also -TER.

paradigm parname (Shortcut: -p)

This is the name of the paradigm file as found in each functional run directory. For blocked and event-related designs, the paradigm file defines which event type was presented when. For periodic designs (retinotopy and AB blocked), it defines the phase and period of stimulation for each run. The name of the paradigm file must be the same across runs and sessions, though its contents can change. If event types are redefined (eg, post-hoc event sorting), then a new paradigm file (with a different name) should be created followed by creating a new analysis. For event and blocked designs, time=0 is relative to the time of acquisition of the first stored image.

designtype type (Shortcut: -dt)

This specifies the type of design for which the analysis is being created. Legal values are event-related, blocked, abblocked, and retinotopy. Event-related and blocked must be analyzed with ["selxavg-sess"]. abblocked and retinotopy must be analyzed with ["sfa-sess"]. abblocked is a block design in which only two conditions (or one condition and one null condition) are presented in a periodic fashion. Ie, each condition is presented for for the same duration in an alternating fashion.

fsd FSD

Functional subdirectory (FSD). This is the folder under the sessions directory in which the runs of raw functional data will be found (default is bold).

funcstem stem or motioncor

This is the bshort/bfloat stem of the functional volume to process. This volume must exist in each functional run. By default, the stem is f. If the -motioncor flag is included, then the stem fmc is used (this is the default output of ["mc-sess"]).

runlistfile filename (Shortcut: -rlf)

This is a file that should exist in each FSD that lists the runs to be processed in that session. The name must be the same across sessions but the contents of each file can vary. If an RLF is not specified, all runs found in the FSD will be processed.

inorm

Perform IntensityNormalization. This requires that ["inorm-sess"] be run on the volume specified by -funcstem (or whatever it defaults to). IntensityNormalization will force the grand within-tissue mean of the functional data to be 1000 for each run.

fwhm Full-Width-Half-Max

This functionality is somewhat obsolete. Consider using ["spatialsmooth-sess"] for 3D smoothing. -fwhm specifies that within-plane (ie, 2D) spatial smoothing using a Hanning kernel should be performed. The Full-Width-Half-Max (FWHM) indicates the extent of the smoothing and is specified in mm.

nskip N

Skip the first N time points in each functional run. It may be necessary to eliminate the first 4 or 5 time points in a functional run because the sample has not reached magnetic equilibrium. This is usually not necessary if the sample was scanned with discarded acquisitions (DISTACQS or DDAs). Note: the nskip does not affect the definition of time=0 in the paradigm file (see -paradigm). This functionality can also be implemented using a time-point exclude file (see -tpef).

tpef filename

This is the name of a time point exclude file (TPEF). The TPEF stores a list of times (in seconds) of time points to exclude from the analysis. The TPEF must be found in the run directory with the functional data to which it applies. Like paradigm files, the TPEFs must have the same name across subjects and runs though their contents can differ. Unlike paradigm files, it is not necessary to have a TPEF in each run; if no TPEF is found, no time points are excluded.

polyfit N

Drift removal. Fit and remove polynomial of order N. Default is 1 (i.e., 0=mean, 1=linear, and 2=quadradic).

nconditions Nc (Shortcut: -nc)

(For blocked and event-related designs). This is the number of conditions as found in the paradigm files. When ["selxavg-sess"] runs, it checks the number of conditions found the paradigm files against Nc and returns an error if they differ.

timewindow TimeWindow (Shortcut: -tw)

(For blocked and event-related designs). If no shape to the hemodynamic response is to be assumed (ie, -gammafit is not used), then this is the amount of time needed to capture the entire event and response, including a prestimulus window. Here, for blocked designs, the event refers to the entire block. If a shape to the hemodynamic response will be assumed (ie, -gammafit is used), then the time window should be long enough to capture the gamma function. In this case, there is no penalty for a long time window. If no time window is selected with -gammafit, an appropriate one will be selected automatically.

extreg extregstem

Use external regressors when doing estimation. The external regressors should be stored in the bfloat "volume" with stem extregstem. There should actually be only one file there (extregstem_000.bfloat). It should have one row, and the number of columns should be equal to the number of external regressors. The number of frames must be equal to the number time points. There should be an external regressor file in each run directory. By default, all the regressors will be used, but this can be changed with -nextreg. See also -mcextreg.

nextreg N

Use only N external regressors. See -extreg.

mcextreg

Use motion parameters as external regressors. This is the same as setting "-extreg mcextreg -nextreg 3". mcextreg is created when running ["mc-sess"] by orthognalizing the motion parameters. Only the top three components are used (unless changed with -nextreg). Note: this feature was incorporated into mc-sess on 7/1/02. If motion correction was run before this date, it will need to be re-run in order to create the mcextreg file.

prestim PreStimTimeWindow (Shortcut: -ps)

(For blocked and event-related designs). Start averaging PreStimTimeWindow seconds before the onset of the stimulus. This only applies to analyses in which the no shape to the hemodynamic response is to be assumed (ie, -gammafit is not used).

gammafit gfDelta gfTau

(For blocked and event-related designs). Assume that the hemodynamic response takes the shape of a gamma function. gfDelta is the hemodynamic delay (ie, the time until the onset of the response), and gfTau is the dispersion, which controls the rise and fall times. Both are specified in seconds. gfAlpha is the exponent for the gamma function (see -gammaexp); for BOLD, alpha=2. The gamma function takes the form:

• f(t) = ((t-delta)/tau)^alpha * exp(-(t-delta)/tau)

gammaexp gfAlpha

Exponent for the gamma function. Default is 2, which tends to fit the BOLD hemodynamic response better.

taumax Tau``Max

(For blocked and event-related designs). This instructs ["selxavg-sess"] to estimate and correct for global autocorrelation in the noise. It will assume that the autocorrelation function is zero beyond a delay of TauMax seconds.

mask maskid

This only applies when using -taumax. This is a mask (as found in bold/masks) that is used to limit the voxels used in the computation of the whitening fillter. Usually, one would run mkbrainmask-sess to create masks/brain, and then specify brain as maskid.

TER TemporalEstimationResolution

(For event-related designs). The TER is the Temporal Estimation Resolution. This is the resolution to which the hemodynamic response will be estimated (in seconds). It must be an integer divisor of the TR. The stimulus schedule (ie, paradigm file) must have been created with with the given TER in mind.

timeoffset toffset

(For blocked and event-related designs). The time to add to the times listed in each paradigm file. This is only useful if, for some reason, each paradigm file if off by the same amount. This may happen if there was a systematic delay between the beginning of the stimulus stream and the beginning of scanning.

ncycles N

(For retinotopy and AB blocked designs). For designs in which the stimuli are presented periodically, this is the number of cycles of stimulation presented in each run.

Example 1

mkanalysis-sess.new foo -i f -o out

description

Example 2

mkanalysis-sess.new foo -i f -o out -f fvalue

description

See Also

["othercommand1"], ["othercommand2"]

Links

FsFast

Methods Description

description
description

References

["References/Lastname###"]

Reporting Bugs

Report bugs to <analysis-bugs@nmr.mgh.harvard.edu>

Author/s

DougGreve