CTSC Stephan Voss, CHB

Back to CTSC Imaging Informatics Initiative

Mission

Description of big picture, goal(s) of project

Participants

• PI, clinicians, IT staff, CT investigator, etc
• staff missing

Data

COG (Children's Oncology Group) phase I consortium imaging/clinical trial ADVL0712.

• 25 patients (with multiple imaging studies per patient, including CT, MRI and PET imaging)
• In addition to DICOM files, much data exist on various Excel spreadsheets, in file folders, and as email attachments, in LA and Boston. Currently, the only way to integrate all of the data available about a patient is to manually assemble all of these pieces of information. This is a common scenario for multi-institutional clinical trials.

Data Transfer

Diagnostic imaging and meta data reside at the COG Phase I Image Center in LA, and can be easily pulled to a local workstation at Children¹s Boston via the COG Grid.

Work Flow

Osirix is the preferred image viewer. Other tools provided by COG.

Data Management Need

• All of the non-protected DICOM header info should be extractable and searchable (not just study name and protocol ID, but other data such as FDG dose and pt weight)
• The studies must be referenced to therapy surveillance dates, i.e. post-treatment week 4, off-study scan, etc
• The surveillance period and actual imaging studies dates may not coincide for various practical reason related to patient care. We are currently logging this information on spreadsheets
• The institutions are usually asked to submit the local radiologist’s report along with the images. These reports come in various formats, ranging from electronic reports included with the image files to PDFs. These are currently sent by email.
• Measurements are made to determine treatment response; there is no way of posting those annotated images used to determine response back to the Grid. Posting the measurements and the annotated images used to make the assessment would be VERY useful when disagreements arise between central reviews and institutional assessments.
• Genetic and pathology data.
• As a research study progresses and new information is learned, supplemental data is often required (by the FDA, for example). It is very common for us to go back to institutions and request additional clinical data based on our interim analyses. We would need a “placeholder” for such future data.
• launch image analysis tools – such as Simon, Karl, and others have been developing, DCE-MRI for example – and use XNAT to access the appropriate imaging data hosted on the GRID and subsequently post the analyses back to the GRID. Again, these are currently manual operations.