MR analysis tool color-codes breast tumors to identify cancer

Contrast-enhanced MR is useful for detecting masses in the breast but less adept at determining whether the kinetic cycle of contrast media, if properly interpreted, can be used to characterize tissues. Breast imaging enhancement software developed by 3TP Imaging Sciences can uncover that information.

The Southhampton, NY, company is preparing a marketing blitz in September for its 3TP (three time point) software package, which the company claims can distinguish malignant from benign tumors. The name refers to the three time points during which data are obtained for the analyses. The software supports expert interpretations, according to Jonathan Read, vice president of marketing for 3TP Imaging Sciences.

"Our software provides an indication of benign tissue versus malignant tissue," Read said.

Clinical studies of radiologists assessing solid breast lesions with 3TP software have demonstrated sensitivity of 96% and specificity of 82%. The key is in coding the pixels to indicate changes in intensity. These changes, due to the washing of contrast media into and out of extracellular space, are characteristic of the type of tissue. The kinetics of this cycle indicate the density of the capillaries, permeability of the blood vessels, and nature of the extracellular space-three measures that the company has linked algorithmically to the presence or absence of cancer.

Radiologists seeking to interpret MR breast images without the aid of 3TP software must try to glean patterns indicative of cancer from gray-scale images. The software quantitates these changes, producing color-coded maps that are overlaid on the MR image. Red corresponds with malignancy, blue with benignancy.

The colors help interpreters identify cancerous lesions, regardless of their level of expertise. For interpreters experienced with breast MR, 3TP software boosts productivity. The algorithm is automatically applied, eliminating the need to manually assign a region of interest for analysis. Results are presented with pixel by pixel analyses.

Building on 10 years of research done by Prof. Hadassa Degani of the Weizmann Institute, the company has optimized the appropriate data thresholds and identified exactly the three times in the wash-in/washout cycle that deliver the most appropriate data.

"We put a lot of work into making the software as simple as possible for the radiologist by taking out as many variables as we could," Read said.

The software has the potential to increase referrals by aiding communications, according to the company. Color-enhanced images are easier for referring physicians and patients to understand.

Because the software requires high-resolution images, the company recommends its use on data obtained at 1T or higher. It is compatible with any MR scanner or workstation using DICOM 3.0. A breast coil typically is required.

The software cleared the FDA in June last year (SCAN 8/3/03). The commercial version was released in May 2004. It is licensed for $45,000 upfront, and users then pay a fee for each patient assessed.

The installed base of about 10 sites could expand in the near future, as the company launches a marketing campaign next month to raise awareness of the product, reaching out to potential customers through advertising, direct mail, and scientific meetings. 3TP will depend initially on its five-person direct sales force, but it plans to develop distribution agreements with other companies, including OEMs. Its campaign will focus on how the product will increase productivity, enhance diagnostic confidence, and improve relations with the referral base, particularly among breast surgeons.

"The software clearly defines for the surgical community where their patients have problems," Read said. "A colorized image is much easier to interpret than a standard gray-scale image."

The primary emphasis in the months ahead will be on breast imaging. But 3TP Imaging Sciences has another application in mind that addresses the identification and differentiation of prostate cancer. The prostate application also uses the wash-in/washout contrast model to distinguish cancerous from benign tissue. But it relies on a different algorithm with its own unique data thresholds and time points and MR protocol. The package has already gained clearance from the FDA but is still in beta testing.

Current plans are to begin marketing the product later this year, possibly at the RSNA meeting. Pricing has not been set, but Read expects it will be in line with the charges established for the breast imaging package.

The prostate product may be more challenging to market than the breast imaging tool. Radiologists are not usually in the diagnostic loop for patients suspected of having prostate cancer. Typically, this is the province of general practitioners and urologists, he said. Surgeons and radiation oncologists handle therapy.

3TP offers other postprocessing algorithms that correct for motion and generate volumetric images. Motion correction is critical to ensure that the pixels corresponding to the three time points are aligned correctly, according to Read. Some referring physicians believe the 3D images are needed to properly analyze the data.

The breast and prostate analysis tools are the first steps toward a family of clinical products. The company has already begun looking ahead to applications addressing the lung and the female genitourinary tract for cervical and ovarian cancer. The lung analyses could be carried out on CT data, Read said.