Health Discovery Corporation Announces Approval and Issuance of Multiple Patent Applications Covering SVM Methods

Business Wire, July 14, 2005

SAVANNAH, Ga. -- Health Discovery Corporation (OTCBB:HDVY) today announced that over the past several weeks it has received notices from patent offices in four different countries regarding the acceptance or issuance of five different patent applications covering aspects of HDVY's support vector machine (SVM) technology.

The U.S. Patent and Trademark Office has issued a notice of allowance of HDVY's claims covering the use of SVMs in computer-aided image analysis. The application of SVMs to image analysis has important implications in a wide range of technologies, including, for example, medical applications, such as x-ray interpretation, digital mammography and circulating tumor cell evaluation, security and law enforcement uses, such as face, fingerprint, and handwriting recognition, and defense applications, such as satellite image analysis. The covered methods use a dedicated subsystem, or SVM, to classify each different characteristic or feature of interest within a digitized image. After successful classification of the individual features within the different subsystems, an SVM evaluates the results from the combined subsystems to generate a conclusion based on all relevant data in the image.

The Australian Patent Office issued two patents to HDVY in June. The first, Australian Patent No. 780050, covers a tiered arrangement of SVMs that provides for the analysis of multiple data sets, such as distinct data types, to produce a single output. Any number of tiers can be used to optimize the analysis results. One application of this inventive technology is in breast cancer diagnosis, in which a diagnosis can be performed using a combination of mammography image data and gene expression data. A counterpart of this patent has already issued in the U.S. as Patent No. 6,658,395, and both the European Regional and Norwegian Patent Offices have in the past month issued notices of allowance of the claims of the European and Norwegian counterparts of this patent.

The second issued patent, Australian Patent No. 779635, discloses and claims the use of recursive feature elimination (RFE) in support vector machines (SVMs). RFE is a process by which key features of input data can be identified as being more significant for use in classifying the data using a support vector machine. Feature selection methods are critical in reducing processor run time when analyzing the high volumes of data, such as that encountered with gene expression data and related information for discovery of biomarkers. By eliminating those features that are either duplicative or non-critical to the successful classification of the data, accuracy is improved and considerable time savings are achieved during processing. RFE-SVMs have been well accepted and widely adapted to both commercial and research applications. Examples of uses of RFE-SVMs include gene expression analysis for biomarker identification and medical diagnostics, predicting antisense oligonucleotide efficacy, and evaluating beef cattle quality, to provide just a few examples. Claims in a counterpart of this application have previously been approved by the U.S. Patent Office.

With the issued patents and allowed patent applications described above, Health Discovery Corporation now holds the exclusive rights to 13 issued U.S. and foreign patents covering uses of SVMs for discovery of knowledge from large data sets. The issued patents cover methods and systems for pre-processing of data to enhance knowledge discovery using SVMs, analysis of data using multiple support vector machines and for multiple data sets, and providing SVM analysis services over the Internet. HDVY's pending U.S. and foreign patent applications cover numerous improvements to and applications of SVMs including computer-aided image analysis using SVMs, with particular application to diagnosis using medical images, methods of feature selection for enhanced SVM efficiency and biomarkers for colon cancer, prostate cancer and renal cancer discovered with these methods, and use of SVMs for analysis of spectral data, such as mass spectrometry data used for protein analysis.

Stephen D. Barnhill, M.D., Chairman and CEO of Health Discovery Corporation commented "The Issuances and Notices of Allowance of these pioneer patents being granted internationally and in the US continues to strengthen our position as a world leader in the use of support vector machines." Dr. Barnhill continued "We are actively and aggressively pursuing licensing agreements with companies internationally for the use of our technology in biomarker discovery in the fields of prostate cancer, colon cancer and leukemia, image analysis in the fields of circulating tumor cells and digital mammography in breast cancer, fingerprint and face detection in the field of homeland security and Internet uses in the fields of spam email and search engines. In addition, we will continue to use these patents in our own internal biomarker discovery program."

Health Discovery Corporation is a systems biology-oriented biomarker and pathway discovery company, which provides all aspects of First-Phase Biomarker Discovery(SM). Health Discovery Corporation was established to provide pharmaceutical and diagnostic companies with a broad range of analytical and decision support solutions to produce more effective and cost effective diagnostic and drug discovery tools. Founded in September 2003, the Company is headquartered in Savannah, Ga.