MRI statisticians have been applying pattern recognition techniques (including statistical methods such as principal component analysis, nearest neighbor analysis, cluster analysis, discriminant analysis) to chemical and environmental data in support of the Environmental Protection Agency (EPA), the Department of Veterans Affairs (VA), and state research for many years. The focus of our studies has been on PCDDs and PCDFs, either in human adipose tissue, fish, foods, and/or environmental media such as soils and sediments. Our research includes:

• Applying multivariate statistical analysis techniques to dioxin and furan levels in fish, sediment, and soil samples collected near resource recovery facilities statewide to assist the State of Connecticut in assessing the impact of the emissions of these compounds on the environment
• Using multivariate statistical techniques to analyze a VA/EPA database of human adipose tissue levels of PCDDs and PCDFs to evaluate relationships between chemical and demographic parameters and military/civilian status of individuals
• Performing multivariate statistical analysis of airborne emissions of PCDDs and PCDFs from municipal waste
• Providing statistical data analysis and interpretation for California Air Resources Board projects to determine levels of PCDDs and PCDFs in the food chain and to determine current body burdens of PCDDs and PCDFs in California residents
• Applying neural networks to interpret flame images and produce a signal for a combustion control system

Contact:

John S. Stanley,
Ph.D.
Group Vice President,
Business Development
(816) 753-7600,
ext. 1160