mi2b2 Client for RPDR

Getting Started

The Research Patient Data Registry (RPDR) allows the research community to obtain access to clinical data, with the appropriate IRB protocol. This clinical data however does not include Radiology material. The mission of the mi2b2 workbench is to couple the clinical information coming from RPDR to the Radiology material belonging to the same patient cohort of the specific IRB-approved project. It serves as the intermediate layer between the user and the PACS of an institution or a department.

This page provides an overview of the process a user will have to go through from the moment the RPDR query is submitted, until the final delivery of the Medical Imaging material via the new mi2b2 workbench.

RPDR tools and the new mi2b2 workbench

Using the RPDR Query tool as well as the RPDR Data Acquisition Engine, the user obtains aggregate numbers of patients with user-defined characteristics, and can receive more detailed medical information about the queried patient cohort. The new release provides the user with a personalized mi2b2 workbench, directly configured to include the queried cohort information. It is delivered in an folder along with the encrypted RPDR data results. Waiting for the RPDR results to become available usually takes between 2 to 4 weeks, however this time depends on the amount of data requested and can extend to even 6 weeks.

The folder contains the executable mi2b2 workbench application, and can be launched by simply double clicking the mi2b2 workbench icon for the user’s operating system of choice.

The user then logs in with their MGB username and password to their customized mi2b2 workbench. After entering the project-specific passphrase delivered with the RPDR query results, the user accesses the patient information. Consult the Example Workflow of the Reviewing Patients page for a guided description of the procedure.

Using the mi2b2 workbench: The five stages

The process of searching for and finally acquiring the Radiology material associated with the RPDR-queried cohort consists of five steps, which are introduced in this page and thoroughly described on the linked webpages. The following schematic summarizes the information presented in this page, with an estimate of the time of completion of each phase of the process.

I. Reviewing the Patients

The goal of the user in the first step should be to select (making use of the filtering tools) the patients of interest. By clicking on the Get Image Study Information for Selected Patients button, the user is guided to the next level of the process.

II. Requesting studies

The second step consists of the communication of the mi2b2 workbench with the PACS specific to the project (either Brigham & Women’s or Mass General Hospital for the moment). The user previously asked for more information about a subset of patients: This step delivers the queried study information back to the user, displaying it in the main panel of this tab. More details about a study, for example the Date or the Modality are now available. It is important to note that while this important detailed study information is made available by this step, the corresponding medical imaging data has not yet been retrieved.

The user’s goal at this stage is to determine, using this additional detailed information, which studies are of actual interest and will need to be pulled from the PACS, where they are permanently housed, and which not. By clicking on the Move to Image Repository button, the user designates the set of studies that the mi2b2 workbench will request from the PACS.

III. Downloading studies

This is the rate-limiting step of the entire mi2b2 workflow. The user has simply to wait for the requested studies to be successfully retrieved from the PACS into the mi2b2 workbench repository. When that happens, their status shifts from Scheduled to Completed.

The time it takes for the studies to arrive to the mi2b2 repository will depend on the PACS downtimes (mi2b2 usually accesses PACS at its lowest clinical use times, which is typically overnight), PACS maintenance, its access policy etc. In addition, there is a limit to the number of studies that can be retrieved on a daily basis. In total, this step takes at least a couple of days and its time of completion evidently depends on the number of studies requested. As soon as any of the studies are Completed, the user can click on the see Transferred Images button and be directed to the next tab.

IV. Cache management

All the studies that have been retrieved from the PACS since the user started using the mi2b2 workbench and submitting requests are now hosted in the project-specific mi2b2 repository. It is important to clarify that the user still does not have access to the DICOM files of the retrieved studies. These need to be downloaded to a local machine in order for them to be parsed by specific Imaging software or to be visualized in the Image Viewer.

The objective of this stage for the user is to select (making use of the filters, once more) the studies that he or she wants to transfer to a local machine. This is achieved by selecting the studies in the main panel, and clicking the Download Studies to My Computer button. There is the option for the user to download the studies in an encrypted form as well - more details are provided in the page dedicated to the Cache Management.

V. Image Viewer

The mi2b2 workbench has succeeded in finding, requesting, retrieving and finally downloading to a local machine the Imaging Material the user indicated in the first two steps. The images that the user chose to download are now in the user’s local storage space and can be viewed by this tab (Image Viewer) of the mi2b2 workbench or parsed by other DICOM software (e.g. Mango).