REPORTER'S
NOTEBOOK

REPORTER'S NOTEBOOK

Random notes and RSNA observations

By Brenda Tilke and John C. Hayes

Crowds are waiting in line to view GE's new MRI system. The company located a portion of its display in an upstairs venue accessible only by a staircase. The tactic proved successful in controlling access and ensuring an attentive audience.

*** The GE exhibit notwithstanding, word is that traffic on the exhibit floor is lighter than in years past, and that attendees are spending fewer days overall at the RSNA meeting, averaging one or two days rather than three to five.

*** Picker's name switch to Marconi apparently caught some of its employees by surprise and without the basic currency of the meeting: business cards. No problem, one decided. He hand-wrote a version on a lined piece of notebook paper. A lack of business cards is not an unusual problem at the RSNA meeting, but it's usually more common at the end of the week than at the beginning.

*** Despite an exhibit hall crammed with thousands of images from every modality, the image that seems to concern some radiologists most is their own. At several presentations, radiologists lamented the poor view other clinicians seem to hold of their profession. At one session on how to sell PACS to their colleagues and hospital administrators, a speaker complained, "The biggest problem with pitching PACS is that the other doctors say 'What do we need PACS for? It's just an expensive way for our lazy radiologists to get even lazier.'"

*** Paper news releases are out and CD-ROM news releases are in. And, in a new twist, companies are putting promotional materials on a CD-ROM the size of a credit card. It fits in a normal CD drive.

*** Easily the most ear-pleasing exhibit was put together by the European Congress of Radiology to promote its March 5-10 meeting in Vienna (www.ecr.org). The ECR's booth features a pianist playing classical music on a grand piano. Low-tech, but high-class.

*** Digital radiologists who think they can return to film any time should consider what's happening at the University Health Network in Toronto.

"We don't give films to anyone inside or outside the institution," said Gregory Couch, manager of medical imaging information systems, during an InfoRAD presentation. "The only way to get films is to go to the chief radiologist and beg, and he always says no."


Taking oncological imaging to the next level

By James M. Brice

Before I attended the opening session of this year's RSNA meeting, my notion of radiology's role in cancer diagnosis revolved mainly around anatomic imaging. Although I was aware of the metabolic powers of PET and MRS, I expected 95% of cancer-related imaging to involve finding, localizing, and characterizing the size and shape of cancers. I believed that follow-up imaging consists of variations on these themes.

I discovered while listening to Dr. Samuel Hellman, a professor of oncology at Chicago University, and Dr. Hedvig Hricak, chief of abdominal radiology at the University of California, San Francisco, that this view is as misguided as the statement that baseball boils down to batting a pitched ball. There is much more to cancer imaging, including fulfilling the diagnostic needs of the oncologist and exploiting the diagnostic tools that are or will be at the radiologist's command.

Hellman wants not only an anatomic evaluation of tumors found on initial screening, but an evaluation of their genetic progression as well. That means providing data showing how likely the tumor is to metastasize. This information includes the tumor's dimensions, volume, density, genetic makeup (such as the presence of inherited oncogenetic mutations), invasiveness, and angiogenesis. Hellman would prefer to have the lesion's location defined down to the cellular level. In addition to detecting the presence of metastases, the oncologist needs to know their number, size, location, and their phenotype and genotype.

The diagnostic imaging tools that Hricak would employ to respond to these demands are as varied as some of the requirements themselves. She recommends 3-D volumetric imaging as the tool to anatomically evaluate tumors and give Hellman the precise localization and measures of progression he seeks. To avoid biopsy, she anticipates the introduction of optical coherence imaging to generate a definitive, noninvasive diagnosis.

The noninvasive characterization of tumor biology may not be as futuristic as it seems. Researchers at Sloan-Kettering Cancer Institute are already using PET and I-124-FIAU to identify tumor genotypes, Hricak said. FDG PET and MR spectroscopy can also help the radiologist examine a tumor's biochemistry for clues about its probable progression. In some cases, a fusion of these two exams is illuminating, according to Hricak. Research from UCSF is demonstrating the value of a combined display of gray-scale MRI and color mapping of MRS measures of choline relative to creatine to identify the presence and extent of prostate cancer.

Hellman's and Hricak's visions are radical departures from the traditional orientation toward anatomical radiology. Yet, ultimately, they will provide the functional information essential to creating a more evidence-based approach to cancer therapy.