At the 1999 Symposium for Computer Applications in Radiology (SCAR) meeting, software developer LizardTech turned some heads with the medical debut of its image compression package, MrSID (multiresolution seamless image database). Like the dynamic
At the 1999 Symposium for Computer Applications in Radiology (SCAR) meeting, software developer LizardTech turned some heads with the medical debut of its image compression package, MrSID (multiresolution seamless image database). Like the dynamic transfer syntax (DTS) technology developed at the University of Pittsburgh in the late 1990s (and subsequently commercialized by Stentor), MrSID is a Web-based product designed to improve the speed with which images can be stored, retrieved, and viewed over the Internet.
Unlike Stentor and subsequent competitors, however, LizardTech has been unable to gain a foothold in the medical market. The company has yet to gain FDA clearance or introduce a commercial version of MrSID for medical images, and it still claims only a single beta site: the University of California, San Francisco.
Even so, LizardTech continues to charm the radiology community. At the 2000 SCAR meeting last June, Dr. Ronald Arenson, chair of radiology at UCSF, showed an auditorium full of radiologists that a small handheld computer could adequately display a chest image and provide instant access to related patient information contained within a hospital database. That demonstration, which garnered oohs and aahs from the audience, was enabled by LizardTech, working with UCSF to refine medical image applications of the company's wavelet-based software.
Like some of its competitors-namely, RealTimeImage-LizardTech has found commercial success outside of the medical field. As the exclusive licensee of patented technology developed at Los Alamos National Laboratory, LizardTech originally targeted the geospatial, publishing, and criminology markets. While recognizing MrSID's potential for medical imaging, the company found that the first and even second generations of the product had critical limitations.
"Because MrSID was originally developed for satellite imagery and fingerprints, it was designed to deal with only eight bits of gray," said Fred Beuthal, who is now overseeing LizardTech's medical and wireless business development. "The next generation of product development has focused on providing the ability to handle 32 bits of color."
In fact, LizardTech believes MrSID v. 3 is well suited for transferring and viewing medical images, especially over low-bandwidth Internet connections. Similar to the image compression and streaming capabilities offered by Stentor, RTI, and ImageMedical (now part of Avreo), LizardTech's software allows users to receive a low-resolution version of an image and then randomly zoom in on regions of interest at higher resolution. Images of any size can be sent from platform to platform without compromising resolution and contrast, and additional image pixels can be transmitted on demand.
"The real value of this technology is in low-bandwidth wireless and DSL/ISDN applications," Beuthal said.
Study Assesses Lung CT-Based AI Models for Predicting Interstitial Lung Abnormality
September 6th 2024A machine-learning-based model demonstrated an 87 percent area under the curve and a 90 percent specificity rate for predicting interstitial lung abnormality on CT scans, according to new research.