Health Discovery Corporation Receives Notice of Allowance of Landmark Patent Application Covering Recursive Feature Elimination in SMVs for Data Mining

Business Wire, Feb 28, 2006

SAVANNAH, Ga. -- Health Discovery Corporation ("HDC") announced today that it has received a Notice of Allowance from the United States Patent and Trademark Office in response to a request for continued examination of its patent application covering the use of recursive feature elimination (RFE) in support vector machines (SVMs).

RFE is a powerful tool for achieving feature reduction because it takes into account dependencies between features and, therefore, is considered more sophisticated than other methods that evaluate and eliminate features independently. This technology has gained wide acceptance in both academia and commercial applications and is utilized by many top tier institutions and corporations. Some examples of academic success with the use of RFE include The University of California for diagnostic markers of cardiovascular illness, Columbia University for genomic analysis, MIT for epileptic seizure detection using EEGs, University of Oviedo, Spain, for assessing beef quality, University of Trento, Italy, for molecular profiling of DNA microarray data, classification of heart arrhythmias and digital mammogram analysis, Vanderbilt University for mass spectroscopy data, Stanford University for gene microarray analysis and many others. Additionally, corporate scientists from Intel (Nasdaq:INTC), Microsoft (Nasdaq:MSFT), Hoffman Laroche (SWX:ROG), and others have incorporated RFE into their research on data mining, gene expression, and other machine learning applications.

The Patent Office had previously approved the application (www.healthdiscoverycorp.com/pr/march1_05.html) however, HDC subsequently filed a request for continued examination in April 2005 to submit additional claims and to ensure that all claims received a rigorous examination to confirm that the claimed RFE method was novel and non-obvious over the prior art. The allowed claims encompass the application of the RFE-SVM method to all data types, not limited to biological data analysis, but include claims directed to the use of RFE-SVM for identifying patterns in biological data and for use in diagnosis, prognosis and treatment of a disease.

"This Notice of Allowance for our RFE technology is very significant for our Company," said Stephen D. Barnhill, M.D., HDC Chairman and CEO. "We made the strategic decision to file additional claims to this application and by doing so we've made this very important patent even stronger. The importance of this technique has been demonstrated and published by leading scientists around the world. Now that we have received this Notice of Allowance we can begin our patent enforcement strategy for this patent."

The RFE method was first reported in 2002 in an article entitled "Gene Selection for Cancer Classification Using Support Vector Machines," co-authored by HDC Scientific Team members Dr. Isabelle Guyon, co-inventor of SVMs, Dr. Vladimir Vapnik, recent Humboldt prize winner for the development of the theory behind SVMs and Dr. Stephen Barnhill. The article describing the initial discovery was published in the journal Machine Learning and is widely cited in subsequent publications recognizing the significance of the RFE method within the SVM field (www.healthdiscoverycorp.com/pdf/GENESEL.PDF).

In addition to the allowed patent application, HDC now holds the exclusive rights to 17 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, providing SVM analysis services over the Internet, use of SVMs for digital image analysis, and new kernels for machine learning. HDC'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 data generated in mass spectrometry.

Savannah-based Health Discovery Corporation (OTCBB:HDVY) is uniquely positioned in the field of pattern recognition technology. Through the application of its patent protected technology, Health Discovery is a biology-oriented biomarker and pathway discovery company providing all aspects of First-Phase Biomarker Discovery(sm). The Company's pattern recognition technology also holds significant application potential in other sizable commercial markets such as diagnostics, Internet search and spam, homeland security, financial futures, and other areas where analysis of complex data is required.

This news release contains "forward-looking statements" within the meaning of Section 27a of the Securities Acts of 1933 and Section 21E of the Securities Exchange Act of 1934. Although the Company believes that the expectations reflected in such forward-looking statements are reasonable, it can give no assurance that such expectations will prove correct.