Inhaltsverzeichnis
Creation of Reference Datasets
Based on motion phases, defined in single trials, mean- and standard-deviation can be determined for a set of reference subjects. Typically one or more files in the „.d3d“-format are used as input and one „.d3d“-file, including the time-normalized and averaged data, is written into the projects „Reference Data“ folder, also in the „.d3d“-file format.
The reference data can be visualized as gray bands, behind the single trials, in the Plot Diagrams components.
Wizzard procedure
Use the “Group mean data” wizzard. You can find it in the Tools main-menu.
In the wizzards first panel you have to select your input data.
If the paths to your input data are defined in a Session Set file, you have to check the corresponding box.
In the other case, it is assumed, that your input data is saved in the “output” folder of your main project in “.d3d”-files.
Two-step procedure (deprecated)
This procedure is not recommended but it explains, what is done internally in the automated process configuration based procedure.
- Step:
You have to select some input .d3d files each representing the data of a single session of a subject. It has the following hierarchically structure:
<subject>/<session>/<group>/<phase>/<original trial name>_<phase number> <subject>/<session>/<group>/<phase>/<original trial name>_<phase number> <subject>/<session>/<group>/<phase>/meanStd ... <group>/meanStd
The phase numberst starts with „0“. At least the last line „5“ must be included. The other lines are optional but needed for determination of coefficients of multiple correlations (CMC) to describe the „qualtity“ of the normal data. Input files with this structure can be exported from a gait data sheet to the output folder of the project.
- Step:
The normal data file can then be created from selected input files via the context sensitive „create normal data“ action. The normal data file is automatically written in the normalData folder of the project.
The hierarchically structure of the output is the same as for the input files. The default behavoir is that parts of the hierarchy (of the input files) of the form
<subject>/<session>/<group>/<phase>/meanStd
are averaged inclusive determination of the standard deviation. This results in the output of the form:
<norm>/<group>/<phase>/meanStd
If optional data is available this is simple copied to the output.
The normal data determination can be configured in the process configuration file as following example shows:
<NormalData phases="LStride Rstride" frames="101" basedOnSessionMeans="true" sessionSet="abc.set" groups="normal/>
Attribute Name | Default | Description |
---|---|---|
phases | LStrides RStrides | blanc seperated list of phase type names |
frames | 101 | Number of frames all of the phases are resampled to. |
basedOnSessionMeans | true | If set to false, all single trials are averaged. If not set or set to true the sessions means are averaged. |
sessionSet | File, which includes a set of volunteers measurement sessions (inter subject) saved in the session set file format. | |
groups | all defined groups | A blanc seperated list of calc group names. For each normal data are determined. |
intraSubjectSessionSet | If CMC value should be calculated an additional set of sessions with measurement repetitions of the same subject is needed. |
If such a definition is available the normal data file can be determined by a single activation of the action „create normal data“ in the main menu or the context menu of the project node.
Reference Gait data
For gait data the normal data is determined for both sides together with the consequence that the events and phases must be renamed in a side independend manner before averaged.
Mapping of the events:
Original names | Description | Side independend name |
---|---|---|
RHS |
Graphical user interface (GUI)
The GUI presents the normal data as gray bands based on the data of the corresponding group (same group name). This default behavoir can be overwritten by setting a normal data group name explicitly in the <CalcGroup> by the attribute „normaldatagroup“.
Typically the normal data variables have the same names than the trials data variables. But if this it not the case, in the sheet definition explicit variables names can overwrite the default. See the <Norm>-element in the following example labelset fragment as an example.
Example Labelset fragment
<Diagram name="RPelvisAngle_X" column="0" row="0" unit="deg" diagramDescriptionDown="post" diagramDescriptionUp="ant" plane="s" diagramType="xt" diagramTitle="Becken Kippung"> <Norm name="NORMPelvisAngles" component="0" mean="NORMPelvicTilt" std="NORMStdPelvicTilt"/> <Label name="RPelvisAngles" component="0" side="right"/> <Label name="LPelvisAngles" component="0" side="left"/> </Diagram>