This Article  • References  • Full text PDF  • Reply to article  • Send to a friend  • Save in My Folder  • Save to citation manager  • Permissions  Citing Articles  • Citing articles on Web of Science (1)  • Contact me when this article is cited  Related Content  • Similar articles in this journal  Social Bookmarking   What's this?

Clinical Application of Statistical Discriminant Analysis

Charles M. Poser, MD; David L. Sylwester, PhD; Bonson Ho; Adam Alpert

Arch Neurol. 1975;32(5):308-314.

Abstract
Statistical discriminant analysis of the amino acid composition of serum and cerebrospinal fluid (CSF) proteins provides an objective method for distinguishing between normal controls and patients with multiple sclerosis (MS). This method also results in a high degree of specificity in separating MS patients from those with other diseases of the nervous system.

The CSF protein serine residue is highly correlated with the CSF IgG and holds promise for a more sensitive diagnostic test for MS than the currently used CSF IgG. Finally, the serum/CSF protein serine ratio seems to correlate best with clinically determined degree of activity for the disease, the most active cases having the lowest ratio. These results suggest that investigation of the amino acid composition of serum and CSF protein in multiple sclerosis and, possibly, in other diseases might lead to the development of clinically useful tests of diagnosis and degree of activity of MS.

Author Affiliations
From the departments of neurology, and epidemiology and environmental health, College of Medicine, University of Vermont, Burlington, Vt.

Footnotes
Accepted for publication Aug 20, 1974.

Reprint requests to Department of Neurology, Medical Center Hospital (DU), Burlington, VT 05401 (Dr. Poser).

CiteULike    Connotea    Del.icio.us    Digg    Reddit    Technorati    Twitter     What's this?