Devices Oncology Radiology

 News Release - April 3, 2007

Signature Mapping(TM) CAD Technology to Enhance the Performance of Breast Cancer Detection

Signature Mapping(TM) Is The Next Step In The Technology Evolution Of Computer-Aided-Detection (CAD) Of Breast Cancer

HERNDON, Va.--(HSMN NewsFeed)--Guardian Technologies International, Inc. (OTCBB:GDTI ), a leading provider of high-performance security and healthcare solutions, announced today that the clinical survey results employing their Signature Mapping(TM) CAD technology have demonstrated that it is capable of accurately detecting and characterizing tumors in screening mammograms. Most importantly, the study found that Signature Mapping(TM) CAD is a highly sensitive tool for accurately detecting lesions in dense breast tissue that are undetectable with the human eye. In one clinical case using a multi-year longitudinal study, Guardian's technology indicated the presence of a developing lesion four years before it was diagnosed and confirmed by biopsy as being a cancerous tumor.

The American Cancer Society estimates that 212,000 women will be diagnosed with breast cancer. It is the most commonly diagnosed cancer in American women and about 41,000 women will die from breast cancer this year. The challenge to healthcare providers is to detect the existence of breast cancer in the early stages, when it can be managed and the survival rate is higher. Robert A. Smith, Director of Cancer Screening for the American Cancer Society was recently quoted in a Bloomberg article as stating, "For the foreseeable future, the control of breast cancer will depend mostly on the early detection, careful diagnostic evaluation and therapy." Current mammography imaging technologies frequently encounter difficulty distinguishing between cancerous and non-cancerous (benign) lesions; nor can they accurately identify and measure the extent of diffuse and multi-focal breast cancer. Furthermore, normal breast tissue can conceal breast cancer, especially with dense breast tissue. Statistics show that between 10-15% are missed (false negatives). Alternatively there can be abnormalities that look like a cancer, but are actually normal tissue (false positives) that result in unnecessary biopsies. These avoidable biopsies (over one million per year) cause fear and anxiety and are very expensive.

Guardian's research was conducted using clinical data and images provided by the Image Processing and Informatics Laboratory at the University of Southern California and Howard University. This initiative focused on identification, clarification, and clinical interpretation of both dense and fatty breast tissues. Low ranges of sensitivity for breast cancer detection have been directly attributed to the challenges of interpreting these types of images. The Signature Mapping(TM) algorithms employed in the study demonstrated that the process was effective in locating and characterizing these tumor types. In addition, it was very successful at identifying extremely small micro calcifications, tumors that were very diffuse, or were embedded in the chest wall or outer perimeter of the breast itself.

The work performed by Guardian's scientists has resulted in the development of a rich set of clinical tools that address currently unresolved medical image analysis challenges. The tools provide capabilities for improving image clarification, visualization, feature extraction, tissue characterization and quantification, and disease detection, monitoring and staging. For example, algorithms have been created to clarify dense breast tissue, subtract difficult-to-see-through fatty tissue, and determine lesion sizes and volumes. The tools were designed to integrate seamlessly into a radiologist's workflow. The intent is to reduce diagnostic analysis time while simultaneously being able to facilitate a more accurate diagnosis by the radiologist.

Additionally, Guardian's preliminary results using Signature Mapping(TM) CAD for identifying multiple sclerosis lesions in the brain, based on analysis of MRI images, have been extremely promising. Based on these preliminary results, Guardian has undertaken the process of extending the above capabilities and technologies for use in MRI breast imaging. The company believes that its Signature Mapping(TM) CAD can improve MRI detection sensitivity for women at high risk for breast cancer.

The Guardian website now includes information and images which will help illustrate the capabilities of this and other Guardian imaging technologies. Please visit www.guardiantechintl.com/CAD to review the results of the clinical analyses.

About Guardian Technologies International, Inc.

Guardian Technologies International, Inc. (OTCBB:GDTI ) employs high-performance imaging technologies and advanced analytics to create integrated information management products and services. Guardian's technologies are focused on two primary markets: Homeland Defense and Healthcare Solutions. Guardian's homeland defense security technologies provide highly accurate threat-item detection at the image-scanning level, and the company's Healthcare Solutions for radiology improve information and process flow with state-of-the-art CAD technology. For additional information, please visit Guardian's, web site at www.guardiantechintl.com.

This news release contains forward-looking statements that involve risks and uncertainties. Forward-looking statements in this document and those made from time to time by Guardian through its senior management are made pursuant to the safe harbor provisions of the Private Securities Litigation Reform Act of 1995. These forward-looking statements reflect the Company's current views with respect to the future events or financial performance discussed in this release, based on management's beliefs and assumptions and information currently available. When used, the words "believe", "anticipate", "estimate", "project", "should", "expect", "plan", "assume" and similar expressions that do not relate solely to historical matters identify forward-looking statements. Forward-looking statements concerning future plans or results are necessarily only estimates and actual results could differ materially from expectations. Certain factors that could cause or contribute to such differences include, among other things, in particular, the size and timing of contract awards, performance on contracts, performance of acquired companies, availability and cost of key components, unanticipated results from audits of the financial results of the Company and acquired companies, changing interpretations of generally accepted accounting principles, outcomes of government reviews, developments with respect to litigation to which we are a party, potential fluctuations in quarterly results, dependence on large contracts and a limited number of customers, lengthy sales and implementation cycles, market acceptance of new or enhanced products and services, proprietary technology and changing competitive conditions, system performance, management of growth, dependence on key personnel, ability to obtain project financing, general economic and political conditions and other factors affecting spending by customers, the unpredictable nature of working with government agencies and other risks, uncertainties and factors including those described from time to time in Guardian's filings with the Securities and Exchange Commission, including without limitation, Guardian's Form 10-K for the year ended December 31, 2005 and its quarterly reports on Form 10-Q. Guardian expressly disclaims any obligation to update any forward-looking statements.

Source: Guardian Technologies

Issuer of this News Release is solely responsible for its content.
Please address inquiries directly to the issuing company.