Disease Pattern Identification using Multivariate Analysis
Alzheimer's Disease and other forms of dementia each cause a characteristic patterns of glucose
metabolism decline. However, these patterns have overlap, and are complex. It is difficult, and even
prohibitive, to correctly identify disease based upon visual patterns, particularly in early stages of disease.
Abiant is applying multivariate analysis approaches (NPAIRS, developed by Dr. Stephen Strother) that can
identify the key patterns differentiating disease, and the ranking of an individual relative to those patterns.
We have applied these methods to over 300 subjects from the Alzheimers Disease Neuroimaging Initiative
database, and are incorporating this approach into our diagnostic development.
|Copyright 2006-2010 Abiant, Inc.
|Abiant, Inc. 480 West Center Street, Suite 202, Grayslake, Illinois 60030
At left is the pattern of glucose
metabolism decline (blue) that
differentiates Normals (NL), Mild
Cognitive Impairment (MCI) patients,
and Alzheimer's Disease (AD)
patients. The circles and triangles
represent each individual patient's
ranking relative to the disease
pattern. The higher the position, the
more similar to the pattern.
At right, NPAIRS was used to
identify the pattern of glucose
metabolism associated with
disease progression in 8 patients
whose symptoms became more
severe over time. As subjects
declined from baseline (first
vertical set of points) to 12 months
and 24 months (last set), glucose
metabolism decreased in the
regions shown in blue.