Inspect3D Tutorial: Export Results
| Language: | English • français • italiano • português • español |
|---|
This tutorial will show you how to export the results of your analysis. This allows users to report specific signals, metrics, or events from a data set and to share results from analysis conducted within Inspect3D.
Data
This tutorial uses overground walking data from four subjects. The subjects walked at three different speeds: slow, normal and fast. The data was processed using a pipeline with automatic event detection and link model based item calculations (most notably the knee joint angles). [Four Subjects Walking Data Set].
Load and query the CMO library
1. As always, the library must first be loaded and our queries defined.
1.1 A thorough explanation of loading a library and defining data can be found in Tutorial 1.
1.2 This tutorial uses a library with overground walking data at several speeds and six queries. These queries extract the knee flexion angle for the right and left sides for each of the three walking speeds throughout a complete gait cycle.
2.To export data to a text file, click 
Export to open the Export Results dialog.
Customize the export
1. The 'Definition' section (see 1) is used to define which results are to be exported. The options here are Signals, Metrics, or PCA Results.
2. The "Options' section (see 2) is used to specify how the results will be exported. The format for the text file and the precision with which to export numerical quantities can both be set and a preview of the export generated.
Definition
1.1 In the definition section, the user selects which signals to export. Data are only available from queries, metrics and events found in queries, or results from any PCA calculations. If the user wishes to export a certain signal from the CMO library they must first define that signal using a query (See Tutorial 1 for explanation of signal definitions) or perform PCA on multiple pre-defined signals.
1.2 The three types of data can be exported using this dialog are signals, metrics, or PCA results.
1.3 The check boxes that allow you to decide what parts of the data to export changed based on the export type in the combo box. Regardless of the export type, the first 5 rows (or 5 columns) describe the signals being exported (header information). To preview the export format, after any signals are selected, the user can press 'Generate Preview'.
2. Select the check boxes of the data you wish to export.
Signals
The Signals option exports previously defined waveforms. The user controls the Number of Points that will be used to normalize the data; the default setting is 101. The types of data that can be exported include both individual and group data as well as raw and processed data. The five options are: including the header information for each export type:
1. Group Mean: Exports the normalized mean for the defined group (a group is a query). The exported file will contain the normalized mean for each group. The normalized mean for each subject is taken, and mean is taken across the subject means.
- The headers will be:
- Row 1: CMO library path
- Row 2: Signal Name_MEAN
- Row 3: P2D
- Row 4: Group Name
- Row 5: Signal Component
2. Group Standard Deviation: Exports the standard deviation for the group using the mean subject curves as the input data (ie. N = the number of subjects/cmo files contained within the group). The normalized mean for each subject is taken, and the standard deviation is taken across the subject means.
- The headers will be:
- Row 1: CMO library path
- Row 2: Signal Name_STD_DEV
- Row 3: P2D
- Row 4: Group Name
- Row 5: Signal Component
3. Subject Mean: Exports the normalized mean data for each subject in each of the groups.
- The headers will be:
- Row 1: CMO file name
- Row 2: Signal Name
- Row 3: P2D
- Row 4: Group Name
- Row 5: Signal Component
4. Subject Raw: Exports the normalized data for each subject in each of the groups.
- The headers will be:
- Row 1: CMO file name/C3D file name
- Row 2: Signal Name_StartFrame_EndFrame
- Row 3: P2D
- Row 4: Group Name
- Row 5: Signal Component
4.1 Events (Normalised): This option becomes visible when the Subject Raw option is selected, and exports the normalised events that occur within this sequence (intermediate events). The data is normalized to the percentage of the signals.
- The headers will be:
- Col 1: CMO file name/C3D file name
- Col 2: Signal Name_StartFrame_EndFrame_EventName
- Col 3: EVENTS
- Col 4: Group Name
- Col 5: Signal Component.
5. Subject Raw (Original): Exports the time based (un-normalized) data for each subject in each of the groups.
- The headers will be:
- Row 1: CMO file name/C3D file name
- Row 2: Signal Name_StartFrame_EndFrame
- Row 3: P2D
- Row 4: Group Name
- Row 5: Signal Component
5.1 Events (original): This option becomes visible when the Subject Raw option is selected. This options exports the normalised (to 101 points) events that occur within this sequence. The headers will be:
| Row 1: | Group | Subject | "Raw" | "Raw Events" | "Raw Original" | "Raw Events Original" |
|---|---|---|---|---|---|---|
| Row 2: | CMO Library Path | CMO Name | C3D Name | C3D Name | C3D Name | C3D Name |
| Row 3: | Group_Mean/Std_Dev | Subj_Mean/Std_Dev | SignalName_Frames | EventName | SignalName_Frames | EventName |
| Row 4: | P2D | P2D | P2D/Metric | Events | Derived/Metric/P2D | Event_Frames |
| Row 5: | Group Name | |||||
| Row 6: | Signal Component(s) | |||||
Metrics
The second type of data that can be exported is called 'Metrics' under the Export type combo box. This export option is useful if certain aspects of a waveform, such as the minimum or the maximum, are desired and not the entire waveform. When selecting this export type certain options become available to the user.
The first option specifies the data to be used to calculate the results, being either queries or subjects. If 'queries' is selected, then the resulted metrics are calculated for each query, using the mean waveform for the query as the input data. If the 'subjects' option is selected, then metrics are calculated for each subject, for each query.
After selected which data to export from, the user can specific the desired metrics to export. Here there are three options and sub-options, which are described as follows. Included in this description are the header formats for each export option:
1. Mean: Calculates and exports the mean value(s). The header format for the queries are as follows: Col 1: CMO lib path, Col 2: Signal Name_MEAN, Col 3: P2D, Col 4: Group Name, Col and 5: Signal Component. The header format for the subjects are as follows: Col 1: CMO file name, Col 2: Signal Name_MEAN, Col 3: P2D, Col 4: Group Name, and Col 5: Signal Component.
1.1 Std. Dev.: Calculates and exports the standard deviation about the selected metric. The header format for the queries are as follows: Col 1: CMO lib path, Col 2: Signal Name_MEAN_STD_DEV, Col 3: P2D, Col 4: Group Name, and Col 5: Signal Component. The header format for the subjects are as follows: Col 1: CMO file name, Col 2: Signal Name_MEAN_STD_DEV, Col 3: P2D, Col 4: Group Name, and Col 5: Signal Component.
2. Min: Calculates and exports the minimum value. If 'queries' is selected then the mean waveform is computed for the entire query and the minimum value for this signal is computed. If 'subjects' is selected then the mean waveform for each subject is computed, and the minimum value(s) for this waveform will be computed i.e. The minimum of each subject's mean waveform, not the average of individual waveforms minimum value.
2.1 Std. Dev.: Calculates and exports the standard deviation about the minimum point. If 'queries' is selected then the position of the minimum value in the waveform is computed. The mean waveforms for each subject are then computed. The value for each subject at the position of the minimum computed initially is then found. These data are used as the input for the standard deviation calculation (ie. N = number of subjects). If 'subjects' is selected, then the same process is followed, but rather than using the index of the minimum for the query waveform, the mean subject waveforms are used, and the position of their minimum(s) will be computed. This index is then applied to the raw data for the subject (ie. N = number of traces per subject).
2.2. Event Point: Calculates and exports the position (index) of the minimum value(s). If 'queries' is selected, then a single index is exported for each group. If 'subjects' is selected, then an index for each subject within each query is exported.
3. Max: Calculates and exports the maximum value. The same rules as described in the Min export option also apply here.
4. Add Metrics Columns: Adds and removes additional columns of metrics to the export.
PCA Results
The final export options allows the user to export results from the PCA calculations. These export options allow all plots generated in Inspect3D to be recreated in a separate application. Effectively, this is a raw data export option for all PCA results. The header format is highlighted within the figures describing each export option.
1. Subject Scores: This exports the mean subject scores for each group (query). The number of values exported is equal to the number of PCs selected.
2. Group Scores: This exports the mean group scores for each group. The number of values exported is equal to the number of PCs selected.
3. Group Score Std. Dev: This exports the standard error for each group. The number of points is equal to the number of PCs used in the analysis.
4. Variance Explained: This exports variance explained (as a percentage of the total variance) by each principle component.
5. Eigenvectors: This exports eigenvectors for each PC. The number of points exported is 101 and the number of signals depends on the number of PCs used in the analysis.
Options
1. Set the Export Format to Standard ASCII, Transposed ASCII, SPSS ASCII, or P2D and click Generate Preview.
1.1 The options section of the export dialog is used to determine the direction of the ASCII export and preview the export.
1.2 The export format has three options. The Transposed ASCII option, which is seen on the right, exports the signals in a column centric format. The Standard ASCII option, which is seen in the examples used previously in this tutorial, exports the signals in a row centric format. The selection of either option depends on the analysis being performed on the exported signals. The SPSS ASCII, allows the exported file to be read by SPSS. Finally the P2D format is a legacy file format from the MOVE3D software application.
1.3 When the Generate Preview button is clicked the table will show a preview of the signals to be exported, which is controlled by the selections made in the Definition section. This preview is identical to the format of the file that will be exported.
2. Click on the Browse button and select where you want to save your data. Once a path is defined click on the Export button to export the results.
Recap
This tutorial provides an overview of using Inspect3D as an export tool for a .cmo library. It describes the three export options: Signals, Metrics, and PCA Results. Using these options, the user has improved flexibility to export only the desired signals from a data set to be further analyzed.