IT, Integration & PET/CT

The advent of PET/CT has been a boon to radiology and oncology. The hybrid scanners combine CT’s exquisite attention to anatomic detail with PET’s unparalleled ability to analyze functional information. This unique combination facilitates more accurate staging of many types of cancers. Consequently, many facilities are budgeting for and deploying hybrid scanners instead of standalone PET systems. The approach is a wise decision from clinical and economic standpoints; however, PET/CT is not without challenges. The site that thoughtfully and proactively considers and addresses these challenges can better optimize its investment and derive the maximum diagnostic and clinical benefits from PET/CT.

Many PET/CT challenges involve IT. PACS integration has been a slow process. Traditionally, PET/CT interpretation required the radiologist to use a separate, specialized workstation, which has a negative impact on workflow.

The good news is vendors are paying attention and developing new systems and applications to overcome integration woes, enabling radiologists to ramp up PET/CT workflow. Other potential challenges relate to the hefty file sizes of PET/CT images and can be addressed up front via network and storage planning.


PET/CT PACS integration in evolution

“The primary IT challenge is that most PACS workstations do a miserable job at displaying nuclear medicine studies, and PET/CT studies are nuclear medicine exams on steroids,” asserts Patrick Barr, MD, director of Southwest Center for Molecular Imaging in Dallas, Texas. The practice invested in a PACS server for storage three years ago when it deployed Philips Medical Systems Gemini PET/CT scanner. The server is integrated with the PACS at Presbyterian Hospital in Dallas and with the main campus of the imaging center.

Most PACS workstations can provide a basic, quick view of PET/CT images, but the process can be time-consuming and fail to meet the demands of clinical diagnosis. For example, Southwest Center for Molecular Imaging uses its PACS to send screen snapshots, but receiving clinicians cannot window or level the images or easily link PET and CT studies to scroll at the same slice thickness. “Right now, data are not presented in a clinically useful manner on the PACS workstation,” states Barr. For example, oncologists cannot view PET/CT studies side by side or fused images. Consequently, clinicians are forced to trek to Barr’s office to view studies on the Philips Extended Brilliance PET/CT Workstation. “The real issue is to design workstations to provide clinicians with access to PET/CT images that is as easy as accessing a chest x-ray,” says Barr.

The workstation issue affects radiologists as well as other clinicians and specialists. In the traditional configuration, radiologists must rely on a specialized PET/CT workstation rather than the ubiquitous PACS workstation to read PET/CT exams. The dual-workstation workflow is not as streamlined as a single, one-stop-shop workstation arrangement.

“The tide is turning,” asserts James Busch, MD, radiologist with Diagnostic Radiology Consultants in Chattanooga, Tenn., whose practice relies on Siemens Medical Solutions Biograph 6 and Biograph 16 PET/CT scanners in conjunction with syngo PACS and syngo 3D task card. The 3D application is “syngoized.” In a nutshell, syngoization means images can be accessed within PACS, enabling the physician to interpret PET/CT images from anywhere a PACS workstation is located rather than a specialized workstation.


PET/CT in the mega-enterprise

Memorial Sloan-Kettering Cancer Center in New York City is a high-volume PET/CT site, performing 30 to 35 scans daily. Other factors further complicate the environment. Every patient at the center has one or multiple prior PET/CT examines that must be reviewed for comparison. Finally, Memorial Sloan-Kettering relies on PET/CT solutions from multiple vendors including GE Healthcare and Siemens Medical Solutions. “Each unit required the radiologist to use its specialized workstation. With a large practice and many radiologists in nuclear medicine, it was a challenge to cope with all of the workstations,” admits Peter Kijewski, PhD, medical physicist.

Several months ago, Memorial Sloan-Kettering Cancer Center equipped a nuclear medicine reading room with nine PACS workstations that included the GE Advantage Windows Suite. This application contains features for reading PET/CT and Volume Viewing and Analysis software. Additional workstations were installed in radiologist offices. The goal, says Kijewski, was to enable radiologists to sit at any workstation to read PET/CT and eliminate the workflow-busting side trips to separate, specialized workstations. Transferring PET/CT reading to PACS workstations also added workflow functions that include reading worklists, integrated automatic speech recognition for exam reporting, presentation of current and prior studies with a single key stroke, and access to related studies performed on other modalities.

The advantages of the new arrangement are significant, says Timothy Akhurst, MD, radiologist. “We are able to do away with multiple workstations and cope with images from any vendor’s imaging system. Advantage Windows can incorporate data from anywhere, and I can read [images] in my office,” sums Akhurst. The newfound efficiency does not require a sacrifice in terms of image interpretation; Akhurst says Advantage Windows is as accurate as the dedicated workstations in terms of image analysis.


Mega-files

Another challenge of PET/CT is the size of the data sets generated by the scanner. A single scan can generate up to 3,000 images, and a PET/CT case can yield a file up to 200MB (compressed). The large file size can stress the network and compromise the instant retrieval necessary for streamlined workflow. “Radiologists can’t wait three to five minutes for a PET/CT study to load. If a radiologist reads 30 cases daily, that’s 90 minutes of wait time,” states Akhurst.

To speed image access, Memorial Sloan-Kettering upgraded its network to a gigabit Ethernet connection to the PACS storage device. PET/CT studies load in seconds with the upgrade, somewhat appeasing radiologists.

Opening multiple prior PET/CT studies for comparison remains a challenge. Viewing three or four priors can entail 600 to 800MB of data. “This may be beyond the capabilities of current workstations,” admits Akhurst. In fact, Advantage Windows does not readily allow simultaneous viewing of multiple PET/CT exams. Users must generate a single data set at a time. Akhurst bypasses the restriction by loading four to six exams on each panel of the dual-monitor PACS configuration on the same workstation; the third monitor runs Advantage Windows Suite and displays the current PET/CT exam. “It’s a very useful technique. In the case of a lymphoma patient, I can review seven or eight prior PET coronals and compare them to the current exam on the third monitor,” explains Akhurst.

The paradigm is likely to change in the future. Souped-up workstations may better handle multiple large data sets. At the same time, new scanners under development will generate much larger files and may require a more substantial network and other IT upgrades.

Busch points to another issue to tackle, reminding facilities that it is necessary to develop a plan for saving images. A minimalist approach is one possibility, but saving just raw PET and raw CT data requires fusion software on the back end. “Not all software programs are created equally and may not yield diagnostic-quality images. Be sure to check that the software yields diagnostic-quality results,” says Busch.

Diagnostic Radiology Consultants goes beyond the minimalist approach, saving diagnostic CT images and fused coronal images in addition to raw CT and raw PET data. A gigabit network infrastructure to the database facilitates rapid loading of exams. Busch offers some final advice. “If the site is just starting out with PET/CT, it’s important to avoid underestimating the size of the images and to purchase expandable storage to meet ongoing demands.”


Conclusion

PET/CT has entered the radiology mainstream and led to improved staging and optimized oncology treatment. These hybrid nuclear medicine scanners, however, do present a few challenges. PACS integration, for example, is a work in progress. Vendors are in various stages of developing and delivering solutions to address the integration issue. Busch advises, “The practice needs traditional PACS and specialized applications that can be integrated into PACS in order to optimize PET/CT.” A thorough analysis of the possibilities of viewing clinically useful PET/CT studies on the PACS workstation is recommended.

PET/CT’s bulky file sizes bring other challenges and can stress the network and storage capacity of the unprepared site. Sites can avoid problems by completing a realistic estimate of storage and network capacity required to handle PET/CT images and evaluating back-end fusion software if it is necessary as part of the archiving plan. 

 

Peeking into the future of PET/CT

The need for true PET/CT PACS integration is apparent and seems to be on the horizon. The PET/CT field will gain from other upcoming developments. Scanners are evolving, and researchers continue to explore new tracers or biomarkers. This month, Health Imaging & IT visits with David Townsend, PhD, director of the Cancer Imaging and Tracer Development Research Program and professor of medicine and radiology at the University of Tennessee Medical Center in Knoxville, to discuss the future of PET/CT imaging. How are PET/CT scanners changing? What clinical applications will the new systems support? PET/CT scanners are still evolving. Vendors are developing new instruments with higher performance such as the Siemens Medical Solutions Biograph TruePoint PET/CT with an extended 21.6-cm axial field of view that we operate here at the University of Tennessee. The extended field of view makes better use of the radiation emitted from the patient. Following the FDG injection and uptake period, radiation is emitted from the whole body and a PET/CT scanner images only a small part of the body at a time and thus “wastes” radiation that could be used for imaging. The new TrueV extended field-of-view option makes much more efficient use of the radiation emitted from the patient. Sixty-four-slice CT scanners also are having a significant impact in cardiology. Cardiology is undoubtedly an up-and-coming PET/CT application area that requires the higher performance of the 64-slice CT; however, 15-cm axial coverage generally suffices for cardiac studies, and therefore, the TrueV option may be less important for cardiology than for imaging cancer. How does this impact hospital and imaging centers in the PET/CT market? It’s important to match the instrument to the intended tasks. Oncology imaging doesn’t require 64-slice CT, and cardiac imaging doesn’t really need the extended field of view. The right PET/CT instrument for the task is not necessarily either the most expensive or the most recent device. What about radiotracer development? What types of new tracers will hit the market? Biomarker development is the major objective of the next few years. FDG [F-18 fluorodeoxyglucose] is an excellent, generic tracer for whole-body surveys of cancer. It is not necessary to know exactly what type of cancer the patient has when imaging with FDG, which is, of course, both a strength and a disadvantage. Some tumors, such as prostate cancer, take up only low levels of FDG. Also, FDG can be taken up by inflammatory and infectious processes. We’d like to see a tracer that distinguishes malignancies from other nonmalignant processes, and we also need a good biomarker for prostate cancer. Are there other new clinical applications in oncology? Of course, in the longer term, we need outcome studies that identify the role of PET and PET/CT in clinical management. Currently, we know that PET/CT is extremely useful for staging disease. We are taking the next step and using PET/CT to monitor therapy as it is being administered. This allows us to change or discontinue therapy that isn’t working. Another role for PET/CT is in actually defining treatment plans such as imaging hypoxia in radiation therapy planning. Pinpointing hypoxic areas of tumors that are more radiation resistant will allow radiation oncologists to boost the radiation dose to these areas, potentially reducing the likelihood of recurrent disease.