Biological agent detection and identification using pattern recognition

摘要

This paper discusses a novel approach for the automatic identification of biological agents. The essence of the approach is a combination of gene expression, microarray-based sensing, information fusion, machine learning and pattern recognition. Integration of these elements is a distinguishing aspect of the approach, leading to a number of significant advantages. Amongst them are the applicability to various agent types including bacteria, viruses, toxins, and others, ability to operate without the knowledge of a pathogen's genome sequence and without the need for bioagent-specific materials or reagents, and a high level of extensibility. Furthermore, the approach allows detection of uncataloged agents, including emerging pathogens. The approach offers a promising avenue for automatic identification of biological agents for applications such as medical diagnostics, bioforensics, and biodefense.