Wiki Documentation Site

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.
====== Metric Interquartile======

=====Overview=====

The **Metric Interquartile** command is used to find the range of the middle 50% of datapoints within a set. A quartile is a statistical measure that divides a data set into four equal parts, with each part representing 25% of the observations. There are three quartile values: the first quartile (Q1), which marks the 25th percentile; the second quartile (Q2), which is the median; and the third quartile (Q3), which marks the 75th percentile. The **interquartile range** can be defined as Q3-Q1 and can be helpful when analyzing the variability of data and/or identifying outliers.

Read more [[https://en.wikipedia.org/wiki/Interquartile_range|here]]. 

=====Pipeline Command=====

The command can be found in the **Pipeline Workshop** under **Metric** as so:

<code>
Metric_Interquartile
! /RESULT_METRIC_FOLDER=PROCESSED
/RESULT_METRIC_NAME=
! /APPLY_AS_SUFFIX_TO_SIGNAL_NAME=FALSE
/SIGNAL_TYPES=
! /SIGNAL_FOLDER=ORIGINAL
! /SIGNAL_NAMES=
! /COMPONENT_SEQUENCE=
/EVENT_SEQUENCE=
/EXCLUDE_EVENTS=
! /GENERATE_MEAN_AND_STDDEV=TRUE
! /GENERATE_MEAN_AND_STDDEV_ACROSS_SUBJECTS=FALSE
! /APPEND_TO_EXISTING_VALUES=FALSE
;
</code>  

====Command Parameters====
The following table shows the command parameters seen above and their descriptions:

|**RESULT_METRIC_FOLDER**|**The name of the result signal folder**|
|**RESULT_METRIC_NAME**|**The name of the result signal**|
|**APPLY_AS_SUFFIX_TO_SIGNAL_NAME**|**Specify the metric name to be the ORIGINAL signal plus a SUFFIX**|
|**SIGNAL_TYPES**|**Specify the signal type**|
|**SIGNAL_FOLDER**|**Specify the origin folder**|
|**SIGNAL_NAMES**|**Specify the Signal to be used**|
|**COMPONENT_SEQUENCE**|**Specify the Signal components to be used (e.g. X + Y + Z or 0 + 1 + 2 etc)**|
|**EVENT_SEQUENCE**|**A list of events (separated by "+" signs). For example, LHS+RTO**|
|**EXCLUDE_EVENTS**|**If this event occurs before the first and last event, do not computed a metric**|
|**GENERATE_MEAN_AND_STDDEV**|**Generate the mean and standard deviation of this metric**|      
|**GENERATE_MEAN_AND_STDDEV_ACROSS_SUBJECTS**|**Generate the mean and standard deviation of this metric across ranges and files**|                                                                                                                                                                        
|**APPEND_TO_EXISTING_VALUES**|**Add these metric values to an existing metric**|  

====Dialog=====

The command can be edited in a text editor or in a dialog form. To edit in the dialog pop up form either click on the **Edit** button in the pipeline workshop or double-click on the pipeline command. The dialog is shown below:

{{:interquartile_dlg.png}}

The dialog box allows you to assign values to the command parameters outlined above.

==== Example: COFP Range for balance trial ====
Here the **Metric Interquartile** command is used to analyze center of foot pressure data during a standing trial and compare the ranges computed for a subject's left and right sides. 

The command for the right side looks like this:

<code>
Metric_Interquartile
! /RESULT_METRIC_FOLDER=PROCESSED
/RESULT_METRIC_NAME=COFP_RIGHT_INTER
! /APPLY_AS_SUFFIX_TO_SIGNAL_NAME=FALSE
/SIGNAL_TYPES=COFP
! /SIGNAL_FOLDER=ORIGINAL
/SIGNAL_NAMES=FP3
 /COMPONENT_SEQUENCE=ALL
/EVENT_SEQUENCE=START+END
/EXCLUDE_EVENTS=
! /GENERATE_MEAN_AND_STDDEV=TRUE
! /GENERATE_MEAN_AND_STDDEV_ACROSS_SUBJECTS=FALSE
! /APPEND_TO_EXISTING_VALUES=FALSE
;
</code>

COFP is a 2 dimensional signal, with the X and y components representing the medial/lateral and anterior/posterior directions respectively. The resulting metrics contain an x value and a y value indicating the variation in pressure for each direction during the trial. A larger range indicates greater variation in the COFP value during the trial and a less consistent balance point.  

The results for this trial show a larger range for the right side than left and a greater range in the Y direction than X for both sides.

{{:cofp_left.png}}{{:cofp_right.png|}}

====Example: Comparing Interquartile Ranges for joints====
In this example **Metric Interquartile** is used to compare the left and right ankle angles of subjects running on a treadmill. 

First, **Automatic Gait Events** is used to define key events that will define the following commands:

<code>
Automatic_Gait_Events
! /FRAME_WINDOW=8
! /USE_TPR=TRUE
! /TPR_EVENT_INSTANCE=1
;
</code>

Next, the **Interquartile** command is used to identify the middle 50% range of ankle angle values for both sides between toe and heel strikes.

<code>
Metric_Interquartile
! /RESULT_METRIC_FOLDER=PROCESSED
/RESULT_METRIC_NAME=RANKLE_INTER
! /APPLY_AS_SUFFIX_TO_SIGNAL_NAME=FALSE
/SIGNAL_TYPES=LINK_MODEL_BASED
! /SIGNAL_FOLDER=ORIGINAL
/SIGNAL_NAMES=RAnkleAngle
! /COMPONENT_SEQUENCE=
/EVENT_SEQUENCE=RHS+RTO
/EXCLUDE_EVENTS=
! /GENERATE_MEAN_AND_STDDEV=TRUE
! /GENERATE_MEAN_AND_STDDEV_ACROSS_SUBJECTS=FALSE
! /APPEND_TO_EXISTING_VALUES=FALSE
;

Metric_Interquartile
! /RESULT_METRIC_FOLDER=PROCESSED
/RESULT_METRIC_NAME=LANKLE_INTER
! /APPLY_AS_SUFFIX_TO_SIGNAL_NAME=FALSE
/SIGNAL_TYPES=LINK_MODEL_BASED
! /SIGNAL_FOLDER=ORIGINAL
/SIGNAL_NAMES=LAnkleAngle
! /COMPONENT_SEQUENCE=
/EVENT_SEQUENCE=LHS+LTO
/EXCLUDE_EVENTS=
! /GENERATE_MEAN_AND_STDDEV=TRUE
! /GENERATE_MEAN_AND_STDDEV_ACROSS_SUBJECTS=FALSE
! /APPEND_TO_EXISTING_VALUES=FALSE
;
</code>

The results for this trial show that the subjects on average exhibited a larger interquartile range for flexion of their left ankles than right.

{{:lankle_inter.png|}}{{:rankle_inter.png|}}