Diagnostics & Imaging Week

CHICAGO – Adults may recall the View-Master toy from their youth, with its circular cards filled with pictures that you popped into the system and then advanced with a lever on the side. Viewing the cards with a good light source in front of it brought wonderful 3-D images that captured the imagination of many a child.

That is essentially what the images produced by a stereoscopic digital mammography system look like, according to lead researcher David Getty, PhD, in his presentation during the annual meeting the Radiological Society of North America (Oak Brook, Illinois) titled “Improved Accuracy of Lesion Detection in Breast Cancer Screening with Stereoscopic Digital Mammography.”

“The image just leaps out at you,” Getty said as he discussed interim results of an ongoing clinical trial.

Getty is a division scientist at BBN Technologies (Cambridge, Massachusetts). BBN and Planar Systems (Beaverton, Oregon) developed the stereoscopic system used in the study, conducted at the Breast Imaging Center of Emory University (Atlanta) to help radiologists see the internal structure of the breast.

In the procedure BBN’s Stereoscopic Digital Mammography (SDM) system, a mammography system that provides a 3-D, in-depth image of the breast is viewed by a radiologist on a Planar StereoMirror display.

During a press conference on the study, Getty passed out slides embedded in stereo handheld viewers that displayed five different samples of what the system can see. The clarity and depth of the images were indeed impressive.

Getty, who is also a researcher in the field of medical imaging and a fellow in the American Association for the Advancement of Science and the American Psychological Society, conducted the preliminary study of the SDM technology with Carl D’Orsi, MD, director of Breast Imaging at Emory’s Winship Cancer Institute and a professor of Radiology and Hematology/Oncology at Emory and primary clinical investigator of the SDM trial.

Getty described how the system works.

Stereoscopic digital mammography acquires two digital radiographs of the breast, separated by about 8 degrees. The mammographer can then fuse the images at the workstation to view the breast in 3D. In this study, researchers use a full-field digital mammography unit slightly customized to acquire images in pairs.

“The goal in viewing a stereo image like this is that the left hand image has to be directed to one eye only and the right hand image has to be directed to the other eye. That is, each eye has to get only one of those two images and that’s the basis then for the fusion that the brain does to enable you to see depth.” Getty said.

Getty said he has been developing the project for 12 years.

“Standard mammography is one of the most difficult radiographic exams to interpret,” he said. “In a 2-D image of the breast, subtle lesions may be masked by underlying or overlying normal tissue and can be missed. Normal tissue located at different depths can align to mimic a lesion, leading to false-positive findings.”

As of July 2007, 1,093 patients at risk for developing breast cancer were enrolled in the study. Each patient received a full-field standard mammogram and a full-field stereoscopic digital exam. Images were interpreted independently by different radiologists. In all, 259 suspicious findings were found by both procedures. Additional diagnostic tests, including biopsy, were also administered.

Of the 259 suspicious findings, 109 were true lesions. Standard mammography missed 40 of the 109, while the stereoscopic exam missed just 24, a 40% reduction. Standard mammography missed 20 of 41 calcifications while stereoscopic missed only four —an 80% reduction.

The findings seem to suggest that stereo digital mammography could lead to earlier detection of cancer, said Getty.

“A small percentage of the additional lesions missed by standard mammography but detected by stereoscopic mammography will turn out to be cancerous,” he said.

There were 150 false-positives among the 259 suspicious findings. Traditional mammography detected 103 and stereo mammography 53.

“Stereo digital mammography reduced false positives by 49%,” said Getty. “This could have a significant impact by cutting in half the number of women who are needlessly recalled for additional diagnostic work-ups, resulting in reduced costs and reduced patient anxiety.”

D’Orsi, who attended the press conference, told Diagnostics & Imaging Week that while the false-negative reduction had not quite achieved statistical significance, he believes it will do so by the time the expected 1,500 patient total is achieved. Thus far, he noted that 1,300 patients have actually been entered into the database.

“The significance achieved already [with the false-positives} will probably get more significant, and the other area, the false-negatives, which are approaching significance may actually at the end of the trial show significance,” he told D&IW. He said that the nearly 50% reduction in false-positives already achieved in the trial is well above the 15% reduction that the trial designers had judged to be a statistically significant number when it was initiated.

Other hospitals interested in participating in such research could easily update their equipment at a relatively low cost, said Getty. For the Emory study, for instance, researchers modified a collimator to improve the image.

Other hospitals seeking to acquire the stereoscopic equipment may soon get help from manufacturers. Getty said three manufacturers – Hologic (Bedford, Massachusetts), Siemens Medical Solutions (Malvern, Pennsylvania) and GE Medical Systems (Waukesha, Wisconsin) are creating breast tomosynthesis machines with motorized movements.

Mammograms from these machines will reconstruct a stack of image slices through the breast very much like a breast CT, he said. The radiograph source in the equipment goes through a smaller arc and not around the whole body.

Getty said he now looks to improve standard scrolling through the stack of image slices by giving radiologists a stereo view through the volume of slices. If this technology works, he said, radiologists could interactively move the stereo point-of-view around to look at all areas. For example, when viewing images of dense or fatty breasts, the radiologist could render some areas “invisible” in order to check suspicious areas.