The steady rise in imaging data produced in radiology departments shows no sign of slowing, thanks to advances in scanner hardware and increasing patient throughput. Thinking ahead for expected growth has become more important than ever when planning a PACS implementation.
The steady rise in imaging data produced in radiology departments shows no sign of slowing, thanks to advances in scanner hardware and increasing patient throughput. Thinking ahead for expected growth has become more important than ever when planning a PACS implementation.
Hospitals in Germany and Belgium have addressed large-scale data handling and storage issues by developing local approaches to managing the growing daily digital data burden. Radiologists at the 1500-bed University Hospital Essen in Germany acquired a PACS in April 2002. Short-term storage capacity has already been multiplied more than sevenfold to address demand, according to Dr. Uwe Fronz, a researcher in Essen's department of diagnostic and interventional radiology.
"Our short-term RAID storage holds at least six months of data online," he said. "We first set it as 1.2 terabytes, but, just recently, we increased it to 9 TB because 1.2 TB was too small. For long-term archiving, we use a tape-based system holding 27 TB of data, and for longer-term storage, we are introducing an application service provider archive."
The PACS is fed by imaging data from four CT scanners, three MR scanners, and eight radiography systems, including angiography. These imaging modalities together accounted for 600 examinations per day during a two-week period in September 2003, resulting in 45,035 images stored (average 74 images per study), Fronz told delegates at the European Congress of Radiology, held in Vienna in March. A repeat of the survey in January 2004 revealed an increase in productivity to 640 examinations per day and around 55,000 images stored (85 images per study).
"We see a clear improvement in number of examinations per day over time, and we are also producing more images per study. This is due to greater use of MRI and multislice CT," he said. "Radiology yesterday was film-based and limited. Radiology today for us is digital and efficient."
Essen radiologists can view imaging data on one of 25 diagnostic workstations, each equipped with 3D postprocessing functionality. Web browser technology distributes images and reports outside the radiology department.
Strict security measures have been set in place. Radiologists log in with a personal smartcard. The system's main server and backup server are located in different buildings, and two databases are maintained separately.
For the future, Fronz hopes implementation of a nationwide state health card system, which is due in 2006, will give professionals and patients secure access to medical imaging data. He would also like to see greater use of wireless tools and remote access solutions, interhospital connectivity, and shared data storage solutions.
Shared storage is already a fact of life at University Hospital Leuven in Belgium, according to Dr. Erwin Bellon, the hospital's PACS manager. But at this 1900-bed institution, sharing means greater internal connectivity rather than regional networking.
Although the storage demands of PACS are well known, it is not the only storage-hungry medical IT application, Bellon said. Departments of ophthalmology and pathology at Leuven generate data-rich medical images on a daily basis, in common with their colleagues in cardiology and radiology. Functional measurements from examinations must also be stored.
"All this makes storage one of the most important cost factors in hospital informatics now," he said. "We must consider storage consolidation to achieve economies of scale."
The Leuven PACS, which is shared between the hospital's radiology and cardiology departments, started out with relatively conventional storage architecture and limited RAID. IT managers then opted to hold all images online to take advantage of a hospital-wide storage system. The PACS currently uses 12 TB of the available 35 TB central storage.
"This online technology changes so fast that every two years you can get further capacity for the same price. This means you can increase capacity gradually and capital investment can be spread over time," he said.
Following the switch to online-only storage, the 18 months' worth of digital imaging data already held on optical disk jukeboxes were transferred across. This migration took just a few weeks and was completed without disrupting clinical routine.
The storage system is connected directly to the central network, and has eliminated the need for prefetching, he said. Consolidating the hospital's storage needs has produced a more flexible solution and made it easier to accommodate changes in departmental storage requirements. Daily management is surprisingly easy. Problem-solving sometimes necessitates discussions with both the PACS vendor (Agfa) and the central storage vendor (Network Appliance), but this has not caused any difficulties.
Bellon has not ruled out a future move back to a more conventional form of long-term data storage, as long as the system can serve all of the hospital's data-generating departments. Choice of an online-only solution was prompted by the desire to be application-free, and this is now a guiding tenet for all future purchases.
"For every new information system our hospital buys, we now demand it must be hooked to all the clinical data and the central storage system," he said.
Leading Breast Radiologists Discuss the USPSTF Breast Cancer Screening Recommendations
May 17th 2024In recognition of National Women’s Health Week, Dana Bonaminio, MD, Amy Patel, MD, and Stacy Smith-Foley, MD, shared their thoughts and perspectives on the recently updated breast cancer screening recommendations from the United States Preventive Services Task Force (USPSTF).
Multicenter CT Study Shows Benefits of Emerging Diagnostic Model for Clear Cell Renal Cell Carcinoma
May 15th 2024Combining clinical and CT features, adjunctive use of a classification and regression tree (CART) diagnostic model demonstrated AUCs for detecting clear cell renal cell carcinoma (ccRCC) that were 15 to 22 percent higher than unassisted radiologist assessments.
CT Study: AI Algorithm Comparable to Radiologists in Differentiating Small Renal Masses
May 14th 2024An emerging deep learning algorithm had a lower AUC and sensitivity than urological radiologists for differentiating between small renal masses on computed tomography (CT) scans but had a 21 percent higher sensitivity rate than non-urological radiologists, according to new research.