CHICAGO--Advanced visualization, or any technology that creates images, diagrams or information to communicate a message, can be employed in various forms and techniques by radiologists, but is not limited to just radiology, said Adam Flanders, MD, at the annual meeting of the Radiological Society of North America (RSNA) on Monday.
Flanders and colleagues presented two Practical Informatics sessions in regard to radiology. Flander’s presentation focused on the progression of advanced visualization and its benefit to the medical field.
Flanders noted computer graphics as the foundation for advanced visualization, stating that the development of computer graphics paved the way for this technology to be utilized by radiologists today.
“We can all look at tables and numbers and have a very hard time with them, but we have great visual perception and cognition for interpreting visual displays and making sense of it in a much more rapid time,” said Flanders. “[Advanced visualization] really takes advantage of that fact.”
Advanced visualization can also be utilized in many applications besides radiology, as Flanders noted, such as weather maps and car dashboards as important tools that also take advantage of human perception and cognition of visual images.
In the radiological perspective, however, Flanders said, “[advanced visualization] is really the concept of extracting data from existing data sets. “
According to Flanders, volumetric interpretation is the kingpin – images can now be looked at in a different way than how they were acquired. The ability to extract more information from the study than what may have been previously seen exists as well.
Cropping and the addition of filters were noted advancements in visualization. Flanders believed that the ability to bring out a small section of an image and enhance the edges while reducing background noise makes for images that “are not your grandfather’s x-rays.”
Flanders spoke about three visualization techniques that are currently in use by radiologists today. Perhaps the most popular, according to Flanders, is multiplaner reformatted images (MPR), which allow data sets to be viewed in multiple planes. Maximum intensity projection (MIP) is a 2-D technique that is taken for granted, said Flanders, as it is very reproducible and is not operator-dependant.
Despite the fact that the method uses 2-D only, Flanders said that the technique gives an illusion of depth when the images are rotated. Volumetric rendering is a method that refers to stacking high resolution images together to create a completely different data set -- typically producing very dramatic images, said Flanders.
Computer-aided detection (CAD), a decision support tool, is thought by some as a technique that can be used in place of the radiologist, “but it is really here to help,” said Flanders.
Deemed “a cognitive grease pencil that watches over [radiologists],” by Flanders, he stressed that CAD plays a complementary role designed to automatically mark any abnormalities in an image for further review by a radiologist.
While the FDA has approved CAD for mammography and chest imaging, Flanders noted that the system suffers from “automation bias,” indicating that a radiologist could become less vigilant in studying an image if he or she knows a system is set in place to note suspicious areas may have been missed.
Looking into the future, Flanders concluded that the natural evolution of interpretation requires finding a novel way to view data without losing precision or accuracy. “We are not going to grow new hands or feet, so we must find more innovative ways to grow-either with interfaces with PACS or new visualization techniques.”