Premium
Noninvasive diagnostic assessment of brain tumors using combined in vivo MR imaging and spectroscopy
Author(s) -
Galanaud Damien,
Nicoli François,
Chinot Olivier,
ConfortGouny Sylviane,
FigarellaBranger Dominique,
Roche Pierre,
Fuentès Stéphane,
Le Fur Yann,
Ranjeva JeanPhilippe,
Cozzone Patrick J.
Publication year - 2006
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.20886
Subject(s) - in vivo magnetic resonance spectroscopy , magnetic resonance imaging , medicine , univariate , linear discriminant analysis , radiology , nuclear medicine , multivariate analysis , contrast (vision) , proton magnetic resonance , multivariate statistics , nuclear magnetic resonance , computer science , mathematics , artificial intelligence , statistics , physics
To determine the potential value of multimodal MRI for the presurgical management of patients with brain tumors, we performed combined magnetic resonance imaging (MRI) and proton MR spectroscopy (MRS) in 164 patients who presented with tumors of various histological subtypes confirmed by surgical biopsy. Univariate statistical analysis of metabolic ratios carried out on the first 121 patients demonstrated significant differences in between‐group comparisons, but failed to provide sufficiently robust classification of individual cases. However, a multivariate statistical approach correctly classified the tumors using linear discriminant analysis (LDA) of combined MRI and MRS data. After initial separation of contrast‐enhancing and non‐contrast‐enhancing lesions, 91% of the former and 87% of the latter were correctly classified. The results were stable when this diagnostic strategy was tested on the additional 43 patients included for validation after the initial statistical analysis, with over 90% of correct classification. Combined MRI and MRS had superior diagnostic value compared to MRS alone, especially in the contrast‐enhancing group. This study shows the clinical value of a multivariate statistical analysis based on multimodal MRI and MRS for the noninvasive evaluation of intracranial tumors. Magn Reson Med, 2006. © 2006 Wiley‐Liss, Inc.