IT Innovation Contest

A team-based contest for creative IT solutions

ImageFlow – A Medical Imaging Data Portal

Proposal Status: 

ImageFlow – A Medical Imaging Data Portal

 

Description:

Medical Imaging data plays an important role in translational research projects across many departments and disciplines. However, there are numerous barriers to accessing this data in a form that is convenient for researchers, imposes minimal impact on clinical systems, and complies with various institutional and legal requirements.

 

Imaging data from clinical and research modalities, including MR, CT, and Ultrasound devices are sent to the institution’s Picture Archiving and Communications System (PACS). Data can be extracted from the PACS, but the process of identifying candidate studies, obtaining approvals, scheduling data transmission with minimal impact on clinical operations, and de-identifying data is more onerous than necessary. Due to the steep learning curve associated with the DICOM protocol there are a number of pitfalls to properly de-identifying data sets, and it can be an overwhelming task to identify the DICOM viewing and editing tools appropriate for a given project.

 

We propose streamlining this process by developing a web-based system that would address these needs in a convenient and secure fashion.

 

Deliverables:

Web portal with the following capabilities:

  • Integrated with centralized authentication / authorization to enforce role based access to imaging data
  • Search function for identifying studies of interest
  • Interface for scheduling the transmission of imaging data sets
  • Robust data set de-identification capabilities
  • Integration with MyResearch and/or Integrated Data Repository where practical
  • Information resource providing recommendations on effective tools for conducting research utilizing medical images in the UCSF computing environment

 

Impact on UCSF’s Mission:

  • Improve efficiency of researchers utilizing medical imaging data, particularly those with limited exposure to the field of radiology or the DICOM standard
  • Shorten cycle of translational research projects
  • Enhance utility of existing data repositories with the addition of medical imaging data
  • Improve efficiency of staff currently manually processing image extraction and post-processing requests

 

Team Members:

 

Maxwell Cheong: Technologist

Mark Day: Visionary / subject matter expert

David Luth: Technologist

Vivek Swarnakar : Visionary / subject matter expert

 

We are actively recruiting end-users interested in this project, as well as other technologists that have experiencing working with central UCSF data repositories, including MyResearch and the Integrated Data Repository

Comments

Very nice proposal! Clinical systems are very often not amenable to research needs, and putting a front end to PACS to make the information de-identified and discoverable will certainly move the research aims of UCSF forward. Integration with MyResearch and/or IDR will certainly make the system more valuable and usable. Hopefully this will facilitate research involving Radiology images. Parenthetically, PACS systems in the future may also store whole slide imaging data, in which case it would enable matching abilities for Pathology images (once digital workflow becomes commonplace in Pathology - currently used for educational, not diagnostic, purposes).

Commenting is closed.