Defining the future of image interpretation

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon

“Radiologists’ needs and expectations are changing, and PACS must evolve to better meet radiologists’ needs,” opined Eliot Siegel, MD, professor and vice chair at University of Maryland during the Imaging Informatics Administration Symposium yesterday at SCAR 2006 in Austin, Texas.
   
“The workstation is the scalpel of the radiologist,” Siegel said. In addition to clinical interpretation, the radiologist’s job also entails determining the best way to process images, whether or how to compress images and maintaining optimal monitor and workstation performance.     
   
Radiologists must archive images for at least five years, share images with other healthcare providers and optimize quality of acquired and displayed images. Effective performance of these radiology functions requires a high level of cooperation between IT and radiology, said Siegel.
   
The other, well-documented change in radiology is the exploding volume of data, said Siegel, with image volume increasing approximately 10 percent annually. “In 1994, the average radiologist read 2,000 CT images daily and had two seconds to read each image,” noted Siegel. Today, that number stands at 80,000, leaving radiologists with a mere 0.45 second per image. The phenomenon is not limited to CT and includes MRI and other modalities, said Siegel.
   
The upshot, Siegel said, is the clear need to change image interpretation to cope with information overload. Radiology has evolved from film to static soft-copy interpretation to dynamic soft-copy interpretation and stack mode and linked stack mode, which takes advantage of the visual system’s ability to detect motion and increases the radiologist’s speed and accuracy. The next — and necessary phase — Siegel said is volumetric navigation, which separates acquisition and display processes and allows the radiologist to navigate images as volume.
   
This evolution in image interpretation impacts PACS. “PACS that have not shifted to volumetric imaging can not keep up with radiology reading needs,” confirmed Siegel. The shift and data explosion also dictates the need for server-side rather than client-side rendering.
   
Volumetric image review, however, can not provide a complete solution. The PACS workstation must be redesigned to facilitate improved workflow and image review. Radiologists need more intuitive controls, said Siegel, who concluded the session with a sneak peek at a futuristic workstation enabling touch-screen navigation of images to enable more rapid, less fatiguing data set organization.