News Story

Specimen Radiography of MRI-Guided Breast Biopsy Enables Immediate Assessment of Surgical Margins
Focused on MRI-Guided Breast Biopsy?the ATEC Emerald Finds its Niche
MR Software Provides Multiple Views in Single Acquisition
Product Showcase: Sunnex Lights Up the MRI Surgical Suite
Running the Numbers
Prostate Coil for 3T Enables Targeted Radiation Therapy

Specimen Radiography of MRI-Guided Breast Biopsy Enables Immediate Assessment of Surgical Margins

In an attempt to identify abnormalities on specimen radiography previously seen only on MRI, Basak Erguvan-Dogan, MD, and her fellow researchers at the University of Texas MD Anderson Cancer Center (Houston) recently developed a procedure for verifying surgical margins of specimens obtained via MR-guided needle localization of breast lesions.

The procedure, reported on in the August issue of the American Journal of Roentgenology,¹ indicates that radiologists can help confirm if an MRI-guided breast biopsy successfully removed the lesion—before leaving surgery.

“By taking x-rays of the lesion specimen, then slicing it up and taking additional x-rays, we can determine if the lesion has been removed or if additional tissue needs to be excised while the patient is still in the operating room,” says Erguvan-Dogan, who believes that patients, surgeons, and pathologists benefit from careful evaluation of the surgical specimen. “Immediate confirmation of removal of the lesion detected on MRI is made possible with this method. The pathologist is able to identify the malignant or benign lesion without having to process the entire surgical specimen. The surgeon can confirm that the surgical excision is successful and that the surgical margins are clear.”

The process applied during the study requires performing a specimen radiograph after surgical excision of the lesion attached to the needle-guidewire system. The tissue is oriented by the surgeon, the pathologist inks the tissue, and then a radiograph of the entire specimen is obtained.

“That radiograph is reviewed, and then the specimen is sliced like a loaf of bread. The slices are then radiographed, which allows for a three-dimensional assessment of the margins,” Erguvan-Dogan explains. “The whole-specimen radiograph and the sliced-specimen radiographs are reviewed by a breast imager, who consults with the pathologist and the surgeon.”

Based on the specimen radiograph and pathologic evaluation of the specimen, the consultation delivers a decision on whether or not additional tissue should be excised.

Whole-specimen and sliced-specimen radiography were performed in 10 patients for the study. In all five malignant cases, sliced-specimen radiographs showed the lesion in question, helped the pathologist correctly identify the lesion while the patient was still in the operating room, and helped the surgeon obtain negative surgical margins, according to Erguvan-Dogan.

“Although the study population is low, we were able to identify 82% of all MRI lesions and 100% of all malignant lesions on specimen radiography,” she notes. “Borderline malignant-benign lesions, such as atypical ductal and lobular hyperplasia, were identified on specimen radiography, whereas only 42% of benign lesions were identified on specimen radiography.”

Erguvan-Dogan believes specimen radiography is a reliable, cost-efficient alternative to repeat contrast-enhanced MRI in the confirmation of lesion removal after surgery, with the additional advantage of enabling immediate assessment of surgical margins and confirming the integrity of the localization wire.

Dana Hinesly is a contributing writer for Medical Imaging.

Reference

1. Erguvan-Dogan B, Whitman GJ, Nguyen VA, et al. Specimen radiography in confirmation of MRI-guided needle localization and surgical excision of breast lesions. AJR. 2006;187:339?344. Available at: www.ajronline.org/cgi/content/abstract/187/2/339. Accessed September 6, 2006.

MR Software Provides Multiple Views in Single Acquisition

Although MR technology boasts impressive 3D imaging capabilities, the magnets can be slow and, in many cases, require additional scans to obtain the desired views.

“With much of traditional MR, if you want to see the sagittal view, you’d have to do a sagittal acquisition,” explains Walter

F. Block, associate professor of biomedical engineering, medical physics, and radiology at the University of Wisconsin-Madison, who has developed a software program to combat this issue. “Not so in our case. You can take one exam with isotropic resolution and reformat it any way you want.”

Currently being used on knees, Block’s new technique, known as VIPR-SSFP, separates the fat and water signal, providing image contrast between bone and the cartilage surface. The conventional data-acquisition methods spend half of their scan time suppressing the signal from fat, instead of imaging cartilage. But the new approach exploits the difference in resonant frequencies between fat and water. During the scan time, then, the technique maximizes each component of the image, so that a technologist can view any aspect.

“Typically, in T2-weighted images, you’d spend several seconds per experiment, and you might need hundreds or thousands of them to make a complete set of images, which would result in a long scan time,” Block says. “Until now, people have known how you could do it faster, but you couldn’t get high resolution—which is what you need to see smaller cartilage defects. As others are developing therapies now to slow or stop cartilage degradation, we want to be able to see smaller defects and the effects of therapy over shorter time intervals.”

All manufacturers are working toward creating higher-powered reconstruction engines, which will be useful with this method. He says that it requires more power to do the reconstruction quickly, because after capture, the data must be interpolated from a radial data space to a Cartesian space.

According to Block, the primary reason this type of high-resolution, non-Cartesian 3D imaging hasn’t been performed successfully to date is because it requires substantial software calibration to obtain reliable, robust images from any system on which it is used. Now, Block’s data-acquisition technique calibrates the system very quickly and capitalizes on recent hardware advances that, coupled with a clever way of maintaining a high-level MR signal throughout the scan, will speed an MRI session. However, speed isn’t necessarily the ultimate goal.

Rather than using the conventional approach, which rasters horizontally to gather MR data, Block’s technique acquires the body’s signals radially. “Typically, radiologists don’t shorten a scan time; they go for a higher resolution, so the standard of what they can see improves,” Block says.

Patented through the Wisconsin Alumni Research Foundation, the technique also can be applied to other areas of the body requiring detailed images. “We started with the knee because it is simple and static and there have been a large number of exams done,” Block explains. “But we’re currently investigating ways of applying the software to T2-rated breast imaging to look at small structures in tumors that are consistent with either benign or malignant lesions.”

High-resolution 3D images are important not only from diagnostic and clinical standpoints, but also to help patients better understand their health conditions, Block says. “If you could actually look at a 3D model from different perspectives, you’d have a much better chance of making sense of the pain you are feeling, your doctor’s diagnosis, and your treatment options,” he adds.

The algorithms were built on a platform from GE Healthcare (Waukesha, Wis), but they are applicable to virtually any MR system above 1.5T. Block is currently collaborating with researchers at Stanford University, the Mayo Clinic, and the universities at Michigan State and Toronto to conduct a multi-center study of the technique.

Dana Hinesly is a contributing writer for Medical Imaging.

Running the Numbers

280% more accurate than ultrasound and mammography combined is a new software for MRI in detecting breast cancer. The software has been developed at the University of Kansas Hospital (Kansas City) by William Smith, MD, and Marc F. Inciardi, dedicated breast radiologists in the hospital’s breast imaging center. When breast cancer is diagnosed, patients undergo an MRI to further determine the size and scope of the cancer. And with this software, Smith and his team can see at the millimeter level. “We are now realizing just how bad we?ve been,” Smith said in a press release. “It’s hard to believe we can do this; [it’s] like we are looking with a whole different set of sensors.”