Page History

The This example includes information about how to represent imaging device information in the event that the research requires this type of data. When the requirements of the study data include the CMR device manufacturer, software and version information, and the specific settings for each subject, how to model this type of information is included in this example. The requirements of the study determine what data is needed. This examples shows how to do it, if it is desired.The PR dataset shows the procedure of cardiac magnetic resonance imaging using the device associated with SPDEVID ABC001 and the PRREFID of 12345678. This information shows what CMR device is used for the procedure and Procedures (PR) dataset includes the SPDEVID variable to show what device was used and the PRREFID variable to represent the accession number or procedure reference identifier associated with the specific procedure for subject 301. The SPDEVID variable is was used to relate records by the device (in this case, the CMR machine). The Device Identifier (DI) domain represents the device information details, the Device Properties (DO) domain represents characteristics of the device that do not vary between subjects. The Device-In-Use (DU) domain represents settings that are set on a device when it is used and may vary between participants.link this PR record to the device records.

The Cardiovascular System Findings This example cardiovascular system findings (CV) domain represents physiological and morphological findings related to the cardiovascular system so it holds the assessments from the parametric mapping. The example CV dataset below shows how to represent longitudinal relaxation time, transverse relaxation time, and extracellular volume findings from CMR . The timing of the contrast specificity is noted using the CVTPT to note if the CVTEST was done before or after contrast. CVTPTREF associates the timepoints in CVTPT with "contrast administration." In the example below, row 9 shows an unreadable CMR using the CVSTAT variable imaging. This example includes records for an unreadable image and an unperformed test. For these records, CVSTAT was populated with "NOT DONE" and the CVREASND variable shows was populated with the reason the procedure result was not done, provided or in this case the result not provided which was a non-evaluable image. A similar approach may be used if there is a desire to record that the CMR was not obtained at the specific visit (omitting the poor-quality information) and why in the dataset records. The non-standard variable CVOIQ=Cardiovascular Overall Image Quality is used to represent if the image for the evaluation is "non-evaluable", "evaluable", "evaluable Image - sub-optimal", or "evaluable image - optimal"procedure was not done. An NSV was used to represent image quality.

The example Procedure Agents (AG) dataset shows details for the Gadolinium gadolinium-based contrast that was used for the procedure. In this example, the researchers did not collect the time of the contrast.

The Device Identifiers (DI) domain represents is used to represent the device identifier information that describes the device used to produce the image that was the basis of the interpretation recorded in the CV domain. Characteristics in DI information details. Characteristics recorded in a DI dataset are those necessary to identify each device to the level of granularity necessary for the study (e.g., to the model level if knowing the actual unit is not necessary, to the serial number level if there is a need to distinguish among units).

Fixed properties of devices are represented in the Device Properties (DO) domain. The sponsor chose to keep the software version constant throughout the study. DO should contain properties that are important for interpreting the data.

Changeable properties and parameters of devices are represented in the Device-in-Use (DU) domain. These settings could can be linked to the record records in the CV domain dataset by their shared SPDEVID and DTC values. For brevity, only 1 subject's parameters are shown.  For more information on relating records, see SDTMIG v3.4 Section 8.2, Relating Peer Records. 

The RELREC dataset is used to describe the relationship between 2 or more records in different domains. For more information on relating records see SDTMIG v3.4, Section 8.2.