IN THIS ARTICLE
Custom metrics add an important new dimension to your ProKnow data in that they allow you to record and analyze the important metrics you care about, metrics that generally can't be extracted or calculated from the original DICOM data.
Accessing Custom Metrics
To access your organization's custom metrics, click on the ProKnow icon in the top left corner of the page, and select Custom Metrics under Organization Settings. Just to the right of the main navigation are vertical tabs for the other Organization Settings. The remainder of the view is devoted to the area where the custom metrics may be defined. If you do not yet have custom metrics, this space will contain a box titled, "No Custom Metrics." Otherwise, your custom metrics will be displayed in the table.
Creating Custom Metrics
Note: You must have the Create Custom Metrics permission to define custom metrics for your organization.
- Click the Add custom metric... link.
Choose a Name, the Type, and the Context for your custom metric.
When choosing a name, use words that will be clear to the people using your system. A good custom metric name is clear while being as concise as possible.
Choose a type that suits the data you wish to record and analyze. Numbers are best if you want to analyze your data in a histogram or scatterplot across a population of patients. Text and Choices can both be used for grouping data across collections. Choices, however, is best when the possible set of values is fixed to a limited set of values, while text is best if you would like to allow free-form text entry.
The context specifies the level at which the custom metric data will be defined. For example, for a metric that specifies whether the contours of a structure set were generated by an auto-segmentation algorithm, you would likely wish to specify the context of Structure Set. Another way to determine the appropriate context of a custom metric is to think about the "highest" point in the hierarchy where the custom metric will have a unique value. For instance, a custom metric titled "Dose Algorithm" that indicates the type of dose algorithm used (e.g., Pencil Beam, Collapsed Cone, or Monte Carlo) would be inappropriate to store at the Plan context since multiple doses, generated by different algorithms, may exist under a single plan (i.e., the value is not unique to a plan).
- Press the Create button to create the custom metric.
Editing Custom Metrics
Note: You must have the Update Custom Metrics permission to update custom metrics for your organization.
CAUTION: Editing a custom metric context is an irreversible action, so use caution. Any custom metrics data associated with the previous context will be removed.
- Double-click the custom metric you wish to edit or select the row and press the Edit button located in the toolbar.
- Modify the field values as needed.
- If you have elected to change the type of the custom metric and upgrading the existing values is a non-trivial operation, you may be asked to specify how existing values should be handled. Please note that the option to upgrade existing values is not possible if you are also making changes to the context of the custom metric. This is because altering the context of a custom metric will result in the removal of any existing custom metric data (so there will be no data to upgrade).
From To Trivial Upgrade Text Number No Text Choices No Number Text Yes Number Choices No Choices Text Yes Choices Number No
- If you have elected to change the context of the custom metric you will be asked to confirm your change. Once you've read and understood the confirmation message, check the confirmation checkbox.
- Press the Save button to save your changes.
Deleting Custom Metrics
Note: You must have the Delete Custom Metrics permission to delete custom metrics for your organization.
CAUTION: Deleting a custom metric is an irreversible action, so use caution.
- Select the custom metrics you wish to delete using the checkbox for each custom metric row.
- Press the Delete button located in the toolbar.
- Once you've read and understood the confirmation message, check the confirmation checkbox, and press the Delete button.
Importing Custom Metrics
Constructing the CSV:
At minimum, your spreadsheet should have Name, Type, and Context values for a custom metric. We recommend labeling these columns as Name, Type, and Context. Next, add columns for each of the custom metric choices you wish to import labeling as Choice 1, Choice 2, Choice 3, etc. In the example below, we have a spreadsheet containing four custom metrics with different types and contexts.
- Acceptable Custom Metric Types are Number, Text, and Choices.
- Acceptable Custom Metric Contexts are Patient, Study, Image Set, Structure Set, Plan, and Dose.
|Name||Type||Context||Choice 1||Choice 2|
|Metric 2||Text||Image Set|
Note: You must have the Create Custom Metrics permission to import custom metrics for your organization.
- Press the Import button located in the toolbar.
- Begin by selecting the CSV file containing the custom metrics data. Press the Select File button to choose a file or drag and drop a file from your file system. Then press Next to continue to the next step.
- Associate each column in the spreadsheet with the appropriate custom metric data. Columns that match metric details such as Name, Type, and Context will be automatically associated. Otherwise, to associate a column, choose your column from the select box. Press Upload when finished to continue.
- To view a detailed results report, click on the "Click here to download a results report" link which will download a CSV file containing detailed information on all metrics imported (and any that may have failed to import). Press Close to exit the wizard.
Downloading Custom Metrics
- Select the custom metrics you wish to download using the checkbox for each custom metric row or select all custom metrics using the top checkbox.
- Press the Download button to initiate the download.
- Locate the CSV file in your downloads folder.