RSNA tackles imaging informatics
CHICAGO, Nov. 26—During a Monday morning refresher course, some of the best minds in imaging informatics provided an overview of current challenges and future directions at the 93rd annual meeting of the Radiological Society of North America (RSNA).

Display systems under the microscope

Aldo Badano, MD, with CDRH/NIBIB Laboratory for the Assessment of Medical Imaging Systems, reviewed the current state of the display market, reminding the audience that LCDs now dominate the landscape. The challenge for imaging informatics professionals is to accurately assess display systems to provide radiologists the optimal start point for image interpretation. Poor displays can increase errors and variability and lengthen reading time. Viewing images at an off-angle also can impact error rates and increases interpretation time. Other current trends and innovations in the radiology arena like content overload, tiled systems, color and portable display devices bring additional challenges, said Badano. 
Standards and guidelines such as AAPM TG 18 are insufficient, said Badano. He closed with a call for additional standards and assessment methodologies.

Modality advances and PACS

Katherine Andriole, PhD, of Brigham and Women’s Hospital, Center for Evidence-Based Medicine at Harvard Medical School, discussed implications of modality advances for PACS.
Initially, the “senior” digital development, computed radiography (CR), did not impact PACS workflow significantly as its workflow mimicked film-based processes.

The transition to digital radiography (DR), however, does change workflow by initiating immediate reading at exposure. Andriole foresees a bright future for CR and DR, predicting that devices will continue to get faster and smaller, decrease in cost and offer value-added tools for image intense specialties such as orthopedics.
CR and digital mammography can stress PACS workflow, said Andriole. CR offers some advantages, mainly by streamlining the transition and offering a lower initial cost. One of digital mammography’s main challenges comes in the form of file size, which can range from 48 to 160 megabytes depending on scanning resolution. “Digital mammography entails very large datasets to transmit, archive and display,” noted Andriole. Other challenges in digital mammography include workstation limitations. Specifically, workstations are limited by each vendor’s acquisition device, which creates workflow issues. Plus, traditional PACS workstation lack specialty requirements for mammography.
The era of multidetector CT may make computed tomography a misnomer, said Andriole. Computerized volumetric imaging may be a more accurate term. Multidetector CT brings potential for new clinical applications, greater diagnostic accuracy and improved workflow and productivity; however, radiologists may need to transition from the 2D to 3D reading model to realize the potential of current and next-generation CT solutions. CT’s functional cousin—PET/CT—carries related challenges such as evolving DICOM standards, large datasets, isotropic voxels and multi-modality acquisition.
Andriole closed with an analysis of the image overload problem. Healthcare sees more patients per day with more imaging exams per patient and more images per study. A modest estimate of a radiologist’s daily workload—26,000 images—takes more than 7 hours to review if the physician reads one slice per second in axial mode. “The model must change,” asserted Andriole. What’s more, sites like Brigham and Women’s generate more than a terabyte (TB) of imaging data per month, adding another element to the problem. Every imaging site requires a robust storage plan.
Yet Andriole believes radiology is up to the task. The Society for Imaging Informatics in Medicine (SIIM) Transforming the Radiological Interpretation Process (TRIP) initiative is designed to find new solutions to the data overload challenge. The multi-faceted solution will include compression, larger and faster archival devices and integration of advanced image processing tools like CAD, she said.