Premium
Discriminant factor analysis of 31 p nmr spectroscopic data in myopathies
Author(s) -
Morvan D.,
Jehenson P.,
Duboc D.,
Syrota A.
Publication year - 1990
Publication title -
magnetic resonance in medicine
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.696
H-Index - 225
eISSN - 1522-2594
pISSN - 0740-3194
DOI - 10.1002/mrm.1910130205
Subject(s) - linear discriminant analysis , mitochondrial myopathy , abnormality , nuclear magnetic resonance , discriminant , pathology , medicine , chemistry , artificial intelligence , biochemistry , computer science , mitochondrial dna , physics , psychiatry , gene
Discriminant factor analysis (DFA) enables one to distinguish among diagnostic groups using diagnostic variables. It provides discriminant functions that are linear combinations of the diagnostic variables and that optimally separate diagnostic groups. It was used to enhance the accuracy of 31 P NMR data in the diagnosis of myopathies. DFA allowed a good separation of normal subjects, congenital neuromuscular disorders with type I fiber predominance, and McArdle's diseases. It elicited an unexpected homogeneity of idiopathic rhabdomyolyses, the metabolic origin of which is unknown, and suggested that the abnormality could involve the mitochondrial oxidative metabolism in some of them. In mitochondrial myopathies, an expected heterogeneity is shown by DFA. It may allow an automatic diagnosis of some myopathies from 31 P NMR data or guide biochemists by comparing biochemical features of a new patient to those of previously established groups.