Volumetric image review offers potential, challenges

Advances in image-acquisition technologies are creating oceans of imaging data, leading to serious image-interpretation and management problems. To chart a course through the data deluge, Dr. Eliot Siegel advocates volumetric navigation.

"Volumetric navigation and multiplanar and 3-D imaging are going to allow us to take the process of image review and separate it from the manner in which they were acquired and reconstructed," said Siegel, who is professor and vice chairman of information systems at the University of Maryland School of Medicine in Baltimore. "CT scanners, for the most part, acquire the images in the axial plane. However, especially with isotropic, thin-section imaging, once the volume of patient data is acquired, the radiologist is ... free to navigate the data in an unlimited number of ways."

Siegel discussed interpretation strategies for large imaging data sets during a SCAR University session at the Symposium for Computer Applications in Radiology, held earlier this month in Boston. The concept of volumetric navigation has been accelerated by the rapid transition to multi-detector CT scanners, he said.

At the Baltimore VA Medical Center (BVAMC), researchers found that the stack-mode reading approach is inadequate for reviewing the 300 to 500 images acquired during a routine CT of the chest, abdomen, or pelvis, and is certainly not adequate for the 1600 to 2000 images generated during a CT angiography "runoff" study. This is also the case for complicated MR procedures such as functional studies, and even general radiography studies have become more complex, Siegel said.

To cope with image overload, radiologists today are adopting tactics such as acquiring multidetector scanner images using thin collimation and then reconstructing the images that are sent to the PACS network using much thicker (such as 5 or 8-mm) sections, Siegel said.

This approach reduces the number of images sent to the PACS network, Siegel said. Additional reconstruction or rendering can then be performed by the technologist on a dedicated CT workstation.

This model is not satisfactory, however. It takes a lot of additional technologist time and too much archive, network, and workstation memory space. In addition, radiologists are constrained to review routine images in the axial plane at a pre-determined, usually relatively thick-plane of section -- negating the added value and power of the new generation of scanners, Siegel said.

"Radiologists should have the flexibility from case to case to determine whether the images should be reviewed in the sagittal or coronal or any oblique plane in any slice," he said. "Volumetric navigation using an advanced workstation frees the radiologist from the previously imposed shackles of conventional CT, which has been limited to just axial sections."

While there's not much data in the literature evaluating the clinical value of advanced workstation tools, several reports have suggested that a maximum intensity projection (MIP) algorithm increases the ability to detect pulmonary nodules, Siegel said.

BVAMC recently collected data from pilot studies studying a 3-D/multiplanar workstation. According to Siegel, the researchers discovered that in 19% of the patients, routine sagittal review of body CT studies led to what were deemed to be significant findings in the spine, which were not commented upon initially from the axial images.

Making use of workstation auditing tools, BVAMC researchers have also discovered that when given access to sagittal and coronal planes for CT of the thorax, radiologists increase their use of the sagittal and coronal planes, and migrate towards increased use of the coronal.

Three-dimensional imaging, once dismissed as little more than "eye candy" for referring clinicians and patients, can be very useful in providing a general survey of an area and portraying anatomic structures such as the ribs, Siegel said.

Challenges and barriers

Volumetric image navigation also leads to some interesting issues, including the possibility that the problem of image volume overload might being traded for clinical image content overload. At BVAMC, thoracic and abdominal subspecialists have asked whether they are now responsible to detailed reports of the musculoskeletal system, the spine, and individual vessels now seen on a routine "body" CT study, Siegel said.

"How should these exams be billed, given that a single acquisition can now generate the equivalent of many subspecialty studies?," he asked. "And should subspecialists in an academic environment such as angiographers and neuroradiologists overread these cases?"

It can be difficult to integrate these advanced workstation tools into the PACS workflow, given that they are typically built as features of dedicated modality workstations for technologists. And in most departments, such high-end workstations are expensive, they are only rarely networked to each other, and comparison studies are rarely available, he said.

At Siegel's institution, the BVAMC employs an enterprise-wide 3-D/multiplanar system that provides clinicians with PC-based access to advanced processing using a client/server model. About a dozen PCs can share the server at a time.

"The system has been enthusiastically received by our medical and surgical colleagues, who have learned to use the workstation pretty effectively in their clinical areas," Siegel said. "Enterprise-wide access to 3-D and multiplanar images has led to increased utilization of CT imaging by the vascular surgeons, orthopedic surgeons, podiatrists, and others. Non-radiologists in particular seem to appreciate and adopt more rapidly to the intuitive perspective provided by 3-D and multiplanar images."

Perhaps the biggest challenge to the adoption of volumetric navigation has been the lack of integration with PACS workstation software; it's simply not practical for radiologists to walk over to a dedicated 3-D/multiplanar workstation for each case. This problem is being addressed, however, though, as several PACS vendors are working on adding volumetric navigation capabilities, according to Siegel.

The rest of the transformation

The transformation of radiological interpretation will continue to evolve at a rapid pace. Although image navigation and enhancement will continue to improve, the next major phase will focus on decision support tools such as computer-aided detection and more sophisticated integration with electronic medical records (EMR), Siegel said

CAD programs will come into routine use in the next few years, especially for the detection of lung nodules and new cancers, he said. Other modalities such as complex new anatomical and functional MRI sequences, cardiac catheterization, and many other types of studies will surely create new challenges, Siegel said.

"Although our continuing trip will be take us into new territory along a yet to be determined route, we believe that the radiology community will be able to stay in the driver's seat as long as we keep our eyes and our minds open to change and as long as we continue to allocate time and energy and resources to investigate the roads that lie just around the bend," he said.

The Society for Computer Applications in Radiology (SCAR) is studying these issues as part of its TRIP (Transforming the Radiological Interpretation Process) initiative.

By Erik L. Ridley
AuntMinnie.com staff writer
June 27, 2003

Related Reading

SCAR debuts TRIP initiative, May 12, 2003

Copyright © 2003 AuntMinnie.com

Page 1 of 775
Next Page