USC Norris Comprehensive Cancer Center: Discovering Hidden Data With PACS-embedded AV Tools
Having a tool kit of advanced visualization tools embedded in the PACS of the Keck Hospital of the University of Southern California (USC) has touched every aspect of the practice of Vinay Duddalwar, MD, FRCR. As abdominal-imaging section chief and director of the USC Norris Comprehensive Cancer Center (NCCC) imaging department, Duddalwar reports that there no longer is a problem getting reformatted studies into the PACS. This enables surgeons to view 3D reconstructions on 70-inch screens in the operating room as they work. In his personal practice, Duddalwar enjoys the ease of obtaining tumor volumes for renal- and liver-resection patients, a formerly time-consuming process that now takes minutes, with the embedded fusion and registration tools at his fingertips—not at the workstation across the room. A supplemental—and perhaps more promising—benefit, as the NCCC radiologists explore the tool set, is being able to get at what Duddalwar calls a study’s hidden information. This has already generated a number of new clinical applications, with presumably more yet to be discovered. “We’ve always found that there is a lot more information on imaging—be it CT or MRI—that is almost hidden,” he explains. “We don’t really use all of the information (such as some aspects of applied functional information) that is available, but difficult to extract.”
“We are using the advanced visualization tools in order to extract more and more information that actually makes a difference in the way the surgeons approach a surgery.”
—Vinay Duddalwar, MD, FRCR USC Norris Comprehensive Cancer CenterOne of the key clinical applications of advanced visualization that Duddalwar and colleagues have explored, since taking delivery of the 3D Clinical Application Suite embedded in the Synapse® PACS from FUJIFILM Medical Systems USA, Inc, a year ago (with enhancements in September 2012) has been working with surgeons on their approach to renal masses. At this world center for partial nephrectomy, USC surgeons take a laparoscopic approach to most resections, and the radiologists work to provide what Duddalwar describes as a roadmap of what they will see during the surgery. “In fact, we try to give them almost a trial run with our visualizations, so that if they dissect this patient, this is what they are going to see,” he explains. “We work very closely with them.” What’s more, Duddalwar and his team are providing surgeons with prognostic information intended to anticipate problems that might be encountered in a renal resection and that might help predict functional outcomes for the patient after surgery; these successes have been shared by Leslie et al at urology meetings in 2012.¹,² “We are measuring the volume of the remaining kidney, and we discuss, with the urologist, what the margin of resection would be or could be; we then try to define how much of a functional loss the patient will have,” Duddalwar explains. The tools used in this application primarily are volume rendering and fusion. The radiologists use the Synapse fusion tool to fuse multiple phases of imaging, such as the vascular and the arterial phases of the study, which are then fused with the tumor to understand the tumor’s relationship with the venous and arterial systems. “We are using this information to predict that, for example, we are going to be taking a slightly longer time with this surgery,” Duddalwar says. Additional Applications The advanced visualization tools also are being used to differentiate among various types of renal masses and how they respond to selected chemotherapies. “This is a work in progress, and we are applying it in various ways,” Duddalwar notes. “Some of these patients with tumors get different chemotherapies, and we are analyzing the way that the tumor changes, using the advanced visualization tools. Predominantly, we are monitoring the changes in density and perfusion.” A pixel-mapping tool is being used to characterize fat in tumors. Patients are referred to NCCC with imaging studies from many other locations, so tumor tracking can be challenging. The radiologists at NCCC are beta testing a new Synapse tool called Response Evaluation Criteria in Solid Tumors (RECIST) tracker, which assists them in monitoring tumor progression using the RECIST measurements. “It is often difficult to be very objective about these things, but we do use those tools to register the old scans to the new scans, to be certain that this is the lesion that we are following, that this is the change that is happening, and (during a period of six months) that there is a growth of 30% or a decrease in size of 20%, and so forth,” Duddalwar says. The tool that has proven indispensible to the oncologic imagers is the fusion tool. “Some of these patients have lesions that are seen on a PET scan that may have been done elsewhere, and we need to know that they are the same lesions seen on a CT exam here,” Duddalwar explains. “We fuse all of these studies; the fusion tool is used pretty much throughout the whole day.” Impact on Workflow Having the tools embedded in Synapse has had a marked impact on radiologist workflow, Duddalwar says. “Previously, we had to go to a stand-alone workstation to process the images—and then, it was difficult to transfer them back, either to the PACS or into a format that our surgeons could use,” he explains. Radiologists are both processing the images themselves and working with technologists in the 3D laboratory to cross-check and collaborate on how to process certain images that will be transferred to Synapse (to be available, for instance, to surgeons in the operating rooms). “They have all of these vascular maps, which we’ve devised in formats after multiple consultations with them,” he explains. “They actually project these on giant 70- inch screens in the operating room to have them in front of them while they are operating.” Radiologists at NCCC have been working with the tools for a year, but took delivery, in September, of an enhanced and improved version of the advanced visualization tool set that is more intuitive than the previous version. “It’s a matter of working with the development people to see if there is a better way of achieving what we want,” Duddalwar says. “Sometimes, they learn from us, and sometimes, we learn from them; it’s a bidirectional process. There are lots of tools that we haven’t explored completely. Our work is driven primarily by what the clinical questions are.” Over the course of the past year, with ready access to advanced visualization, the USC radiologists are using the tools “pretty much all of the time,” Duddalwar says. “The more my colleagues and I get comfortable with the set of tools, the more applications we find, and the more we are using them.” Cheryl Proval is the editor of Radinformatics.com.