Medical Device Daily

CHICAGO — What do bomb detection technology and healthcare imaging technology have in common?

Guardian Technologies (Herndon, Virginia) execs attended the 2006 annual meeting of the Radiological Society of North America (RSNA; Oak Brook, Illinois) to explain just that.

The company is taking its core digital imaging detection technology, 3i (Intelligent Imaging Informatics), which is now being evaluated in the labs of the U.S. Transportation Security Administration , and applying it to medical uses. Guardian was on hand to describe how its technology that can detect the material inside a bomb or other container can also detect lesions in the brain, for example, that could indicate the progression of multiple sclerosis.

“The uniqueness about this is that we’ve learned that every material has a very unique substance signature or fingerprint,” Vice President of Business Development Richard Borrelli said in an interview with the ABC network program “Ahead of the Curve,” which the company was showing at its RSNA booth.

With other technologies that exist currently, “thousands and thousands of these algorithms can be analyzed simultaneously,” he told the program. And it can be used for the human body as well, he said.

“As you might imagine, these organic materials [in bombs] have a unique signature — so does all the material in your body,” he said. “Lesions, fluid, liquid in your body all have a unique signature as well. Now, we’re training these algorithms, with some minor modifications [to detect] things like multiple sclerosis in the brain or for blood in the brain where it’s not supposed to be.”

The company’s technology uses mathematic equations and algorithms “to really extract information that’s unseen by the naked eye in these pixels of information.”

These algorithms are expected to give radiologists “the ability to see something that was not quite as clear before — and seeing it more sharply and clearly.” For example, the healthcare applications are designed to determine whether a material in the body is a solid lesion or a fluid-filled lesion.

“Hopefully, as we progress the technology we’ll be able to tell the radiologists whether this corresponds to a material that looks like a lesion — like a cancer,” Borrelli said on the program.

Ideally, if the technology works the way the company hopes, it would be able to look at tissue very early in a disease state to detect a disease earlier than is now possible.

The company said that Signature Mapping clinical results were derived from five pilot studies that have been conducted through an agreement with the Image Processing and Informatics Laboratory of the University of Southern California (USC; Los Angeles).

The research has covered detection and quantification of multiple sclerosis lesions; normal pressure hydrocephalus; acute intracranial hemorrhage; tuberculosis; and the quantification, tracking, clarification and detection in dense breast tissue.

As an example, Borrelli told Medical Device Daily that Guardian was able to train its algorithms to detect multiple sclerosis in the brain and “do it accurately and with more specificity than the radiologists that were helping us train the algorithms.”

“So, we were able to dependably say, ‘Yes, this algorithm is capable of detecting multiple sclerosis and then quantifying it and determining the size of the lesions, which is actually more important in multiple sclerosis,” Borrelli said, “because it’s not such a difficult area to detect, but it’s a very difficult area to measure and quantify.”

For example, more lesions — or lesions that are larger in size — equate to a patient whose disease is progressing.

With acute intracranial hemorrhage, it’s possible, particularly in trauma cases, Borrelli said, for there to be blood seepage in this area that is sometimes missed by the radiologists or emergency room physicians. Still, it may be present, which would indicate a serious problem and require the patient to be kept at a treatment center.

“As we begin to understand the capabilities of the technology, we can apply it to many areas of the body and help diagnosticians get more information out of the image they create,” he said.

The next step for the company and the technology is to meet with the FDA to discuss what the company expects its claims to be and “what is going to be required in order to demonstrate safety and efficacy,” Borrelli said, acknowledging that there is “still a fair amount to do on the clinical side.”

Guardian said this technology essentially will be used as an adjunctive technology to the imaging devices already used by radiologists to help supplement their optical diagnosis.

“We don’t necessarily think of ourselves as a competitive technology,” Borrelli told MDD. “We would be happy to work with R2Technologies (Sunnyvale, California) or i-CAD (Nashua, New Hampshire) to help them build a better product.”

Guardian reported on Nov. 8 that, in accordance with a purchase agreement with a group of institutional accredited investors, it successfully completed the first of two closings of a private placement of its securities.

The company said investors have agreed to purchase in the aggregate, and before deduction of certain fees and expenses of the offering, $5,150,000 of securities, $2,575,000 of which were purchased at the first closing and $2,575,000 of which will be purchased upon effectiveness of a registration statement to be filed by the company with regard to shares underlying the securities.