Advanced viz, image transmission critical to dealing with info overload
CHICAGO—Attendees of the 94th annual meeting of the Radiological Society of North America (RSNA) gathered in McCormick Place to hear a panel of experts discuss the reasons behind the explosion of images for diagnostic interpretation and to potentially gain insight into what tools and methods could help alleviate some of the information overload, as the hustle and bustle of the second day tapered off.
The Special Focus Session, “Image Overload: Dealing with It” featured presentations from Bradley Erickson, MD, PhD, Mayo Clinic in Rochester; Minn.; Paul Chang, MD, University of Chicago; and Geoffrey D. Rubin, MD, Stanford University in California, CT and vascular studies.
The session was designed to help attendees understand why multidetector isotropic volumetric CT and advanced MR sequences have created an explosion in the number of images to be reviewed for diagnostic interpretation; learn how technological innovations combined with psycho-visual science have enabled creation of new visualization tools; understand how to use these tools safely and effectively for diagnostic accuracy in the neuro-axis, abdomen and cardiovascular systems.
Erickson said that many in the industry question whether computers, “that got us into this trouble, can get us out of this mess,” adding that radiologists typically are seeing close to 30,000 images per day, which translates into one image per second.
So there are a number of strategies out there to deal with the image overload problem, however the MR problem is a little bit different and he attributes it to the parameter space explosion within MR. “While there has been some increase in spatial res[olution], the larger increase has been in parameter space,” he said.
He suggested image alignment as one strategy to combat the overload. Image alignment makes it easier to compare studies and sequences, and reduces busy work—however it does not reduce the amount of information, he noted. Two other strategies he offered as a way to deal with image overload were change detection and segmentation, which he said is a technology that radiology needs to leverage more.
While he presented attendees with an overview of the advanced visualization tools available to aid in interpreting the mountains of image data sets, Chang tackled the IT side of the issue and, focusing on digital image management.
“There is a significant IT informatics challenge when it comes to these extremely large datasets,” Chang said. “It’s going to get worse, not better, with respect to this challenge. How do you get these images to those workstations and how do you do it in a way that doesn’t frustrate radiology with respect to firewalls and performance?”
He said that there is a movement away from the hierarchical storage models for PACS, toward on-demand archive. However, he noted that storage is not the problem—the fundamental problem with managing these large image data sets is the transmission to modalities and 3D workstation or to the server of the client-server environment.
The Special Focus Session, “Image Overload: Dealing with It” featured presentations from Bradley Erickson, MD, PhD, Mayo Clinic in Rochester; Minn.; Paul Chang, MD, University of Chicago; and Geoffrey D. Rubin, MD, Stanford University in California, CT and vascular studies.
The session was designed to help attendees understand why multidetector isotropic volumetric CT and advanced MR sequences have created an explosion in the number of images to be reviewed for diagnostic interpretation; learn how technological innovations combined with psycho-visual science have enabled creation of new visualization tools; understand how to use these tools safely and effectively for diagnostic accuracy in the neuro-axis, abdomen and cardiovascular systems.
Erickson said that many in the industry question whether computers, “that got us into this trouble, can get us out of this mess,” adding that radiologists typically are seeing close to 30,000 images per day, which translates into one image per second.
So there are a number of strategies out there to deal with the image overload problem, however the MR problem is a little bit different and he attributes it to the parameter space explosion within MR. “While there has been some increase in spatial res[olution], the larger increase has been in parameter space,” he said.
He suggested image alignment as one strategy to combat the overload. Image alignment makes it easier to compare studies and sequences, and reduces busy work—however it does not reduce the amount of information, he noted. Two other strategies he offered as a way to deal with image overload were change detection and segmentation, which he said is a technology that radiology needs to leverage more.
While he presented attendees with an overview of the advanced visualization tools available to aid in interpreting the mountains of image data sets, Chang tackled the IT side of the issue and, focusing on digital image management.
“There is a significant IT informatics challenge when it comes to these extremely large datasets,” Chang said. “It’s going to get worse, not better, with respect to this challenge. How do you get these images to those workstations and how do you do it in a way that doesn’t frustrate radiology with respect to firewalls and performance?”
He said that there is a movement away from the hierarchical storage models for PACS, toward on-demand archive. However, he noted that storage is not the problem—the fundamental problem with managing these large image data sets is the transmission to modalities and 3D workstation or to the server of the client-server environment.