Gait- and Motionlab Heidelberg Tools

Eine ältere Version des Dokuments wurde geladen! Beim Speichern wird eine neue Version des Dokuments mit diesem Inhalt erstellt.

======Projects======
A folder with subfolders which includes a set of xml-files to describe what to do with a dataset (session, a set of trials, typcially with c3d-files) is called a "Project". There are some projects specific for the Heidelberg Motionlab.
=====PiGMo10 Vicon PiG based=====

This is a clone of what the PiGMo10 "modelscript" in **MoMo** calculates, based on the Vicon PiG model.

So the complete pipeline from the raw data, to the sheets up to the exported hd-asc files, used by **HeiDataProVIT**, is possible and can substitute **MoMo**.

Extensive testing and finetuning is needed, to make the pipeline more robust for daily usage. Also some parameters MoMo saves in the hd-asc files are missed. It has to be checked if they are needed and if so determining of these parameters must be implemented.

The lower body kinematics angles of this project are described [[http://nimue-platform.motion-science.org/doku.php?id=analysistools:conventionalgaitmodels|here]].
=====PiGMo10 independent=====

Additional to the above this also includes a Vicon PiG kinematics clone. This makes the project completly independent from Vicon codebase, for full flexibility. The project is in experimental status. There are some rotation offsets comparing with the ViconPiG which are not yet understood.

=====PiG with functional centers/axes=====

This project adds functional center/axes determination to the default Vicon PlugInGait model. It works together with the old **Vicon Workstation** and also with new Vicon **Nexus system**.

Additional to the SCoRE/SARA functional method implementation of Vicon, several mathematical methods are available and the phases, which are used as input for the estimation, can be controlled by setting of events in the dynamic calibration trials.

====Eclipse entries====
To process the data, specific dynamic calibration trials must be done and its specific eclipse entries must be set:

====Modelparameter files====
Modelparameter-files are needed to read parameters from clinical measurements, which are used to define offset angles in the kinematics model.

There is an important limitation in the mechanism to work with these files: You can have an mp-file only for each session and not for each group. That is why the mp-files are found only, if they are saved in the session and not in the group. Details of working with .mp-files and its limitations you can find [[http://nimue-platform.motion-science.org/doku.php?id=glue:vicon&#mp-files|here]].

Used parameters are:

<code>
$LKneeWidth = 103 
$LAnkleWidth = 71
$LTibialTorsion = -15
$RKneeWidth = 104 
$RAnkleWidth = 71
$RTibialTorsion = -15
</code>
=====PiG with functional knee joint axes only =====

This project adds functional knee joint axes determination (SARA) only to the default Vicon PlugInGait model. It works together with the old Vicon Workstation system and allows to control the usage of the motion phases, which are used for the estimation by setting of events.

In the gait trials the lateral knee markers are projected onto the functional knee joint axes and the knee joint center is determined based on the clinical measurement of the knee width. Than the femur coordinate system is constructed equivalent to the original PiG. Keep in mind that this coordinate system does not have the functional axis as one of its axes. The axes are defined by cross products to achive orthoganal systems and have the longitudinal femur axes included.

The proximal tibia (anatomical) coordinate systems are defined by PiG virtual markers.

==== Preconditions ====
  * It is assumed, that the Vicon PiG is processed for the static calibration trial and all gait trials, before processing with this model. Vicon PiG processing is not needed for the dynamic calibration trials.
  * Additional markers on the shank are needed. Have a look at the follwing markerset picture
  * Specific dynamic calibration movements with knee flexion; events to define the phases used by the estimation.
  * Specific eclipse entries. Have a look at the following table and screenshot.

==== Markerset ====

{{ :markerset.png?400 |}}
====Eclipse entries====
To process the data, specific dynamic calibration trials must be available and its specific eclipse entries must be set:

^ Trial ^ Name ^ Value ^ Description ^
^ 1 |  CLASS | Static_Cal | Static standing trial with knee alignment device. |
^ 2 |  CLASS | HipFlex/Ext | Flexion/Extention of the hip, which should cause also flexion/extention of the knee; last is used for knee-axis estimation; so it is important, that in this trial also the knee is flexed at minimum of 30 degrees RoM. |
^ all trials | SIDE | Side R, Side L, bds | Hip/Flex/Ext trials are typically unilateral, gait trials are typically bilateral |
^ 3-n | CLASS | Dynamic or Dynamic+Kinetics | Trial with gait, Strides defined by events. |
{{ :eclipse_functionalpig_54662.png?800 |}}

====Modelparameter files====
Modelparameter-files are needed to read parameters from clinical measurements, which are used to define offset angles in the kinematics model.

There is an important limitation in the mechanism to work with these files: You can have an mp-file only for each session and not for each group. That is why the mp-files are found only, if they are saved in the session and not in the group. Details of working with .mp-files and its limitations you can find [[http://nimue-platform.motion-science.org/doku.php?id=glue:vicon&#mp-files|here]].

Used parameters are:

<code>
$LKneeWidth = 103 
$RKneeWidth = 104 
</code>

==== Output ====

In the default configuration the output is written into ".glx"-files in the hd-asc file format besides the corresponding input c3d-files. These files are written only for the gait-trials.

Hip- and knee-angles are saved to its corresponding PiG angles with the same names but based on the functional knee joint axes.

^ Name ^ Description ^
^ R/LKneeAngles | Cardan angles describing the knee motion based on functional knee joint axes corresponding to the equal named original PiG angles.  |
^ R/LHipAngles | Cardan angles describing the hip motion based on functional knee joint axes corresponding to the equal named original PiG angles  |

You can change the output behavoir in the calcml.xml file in the subfolder "jobs" of the project. Have a look at the following definition of processing gait-trials in this file:

<code xml>
<CalcGroup name="dynamic">
            <identification>
              <entry key="CLASS">Dynamic|Dynamic\+Kinetics</entry>
              
            </identification>
            <process>
                <entry key="POINTS">true</entry>
                <entry key="EVENTS">true</entry>
            </process>
            <input labelset="Input.lml">
                 
            </input>
            <output labelset="Output.lml" dir=""
                    handlerclass="de.orat.motionDataConverter.MotionDataOutputHandler">
            </output>
            
        </CalcGroup>
</code>

In this code snipset the the output is done by the "MotionDataOutputHandler" and the timeseries to be written are defined in the "Output.lml" file. This results in a file with the suffix ".d3d". To activate the output into hd-asc files youy have the comment out the "<output>"-tag and substitute it by the one which find at the ende (commented out). This focussed on the the output labelset "Output_reduced.lml". Inside this labelset there is the mimetype of hd-asc set.

<Labels name="all" basefile="true" fileSuffix="glx" mimeType="text/x-hd-asc">
       ...
</code>
==== Process Log ====
The logging output is shown in a seperate window typically shown at the buttom of the application. If this window is not shown you can reopen it by the corresponding action in the Windows-main-menu.

The following picture shows a part of the logging output of processing a session. The messages in black are only for information. The blue ones are warning. If there are red ones, something is totally wrong.

You can see Intervals of frames defined by start- and end-frame enclosed in "[ ; ]". In these intervals the calculation cauld not be done. Typically the origin of this are unlabled markers. You should go to Vicon Worstation to check this.

The information about the differences between PiG and functional PiG, given as an direction cosine angle between the corresponding two knee joint axes is print in blue but typically there is nothing wrong. But you should always have a look at it. If the differences is hight you should think about if this can be correct.

Zusammenfassung

Hinweis: Durch das Bearbeiten dieser Seite geben Sie Ihr Einverständnis, dass Ihr Inhalt unter der folgenden Lizenz veröffentlicht wird: CC Attribution-Share Alike 4.0 International

Gait- and Motionlab Heidelberg Tools

Benutzer-Werkzeuge

Webseiten-Werkzeuge

Seiten-Werkzeuge