Voxelwise multivariate analysis of multimodality magnetic resonance imaging

• Published in: Human brain mapping Vol. 35; no. 3; pp. 831 - 846
• Main Authors: Naylor, Melissa G., Cardenas, Valerie A., Tosun, Duygu, Schuff, Norbert, Weiner, Michael, Schwartzman, Armin
• Format: Journal Article
• Language: English
• Published: New York, NY Blackwell Publishing Ltd 01.03.2014; Wiley-Liss; John Wiley & Sons, Inc; John Wiley and Sons Inc
• Subjects: Aged; Alzheimer Disease - pathology; Alzheimer Disease - physiopathology; Alzheimer's disease; Anisotropy; Biological and medical sciences; Brain - anatomy & histology; Brain - pathology; Brain - physiology; Brain - physiopathology; Case-Control Studies; Cerebrovascular Circulation - physiology; Computer Simulation; Data Interpretation, Statistical; diffusion tensor imaging; Diffusion Tensor Imaging - instrumentation; Diffusion Tensor Imaging - methods; Female; Fundamental and applied biological sciences. Psychology; Humans; Investigative techniques, diagnostic techniques (general aspects); Magnetic Resonance Imaging - instrumentation; Magnetic Resonance Imaging - methods; Male; Medical sciences; Middle Aged; Models, Statistical; multimodality imaging; multiple comparisons; Multivariate Analysis; Nervous system; Perception; perfusion weighted magnetic resonance imaging; Psychology. Psychoanalysis. Psychiatry; Psychology. Psychophysiology; Radiodiagnosis. Nmr imagery. Nmr spectrometry; Random Allocation; Spin Labels; structural magnetic resonance imaging; Vision; Nervous system diseases; Radiodiagnosis; Alzheimer disease; structural magnetic resonance imaging; Multimodal imaging; Multivariate analysis; Nuclear magnetic resonance imaging; Cerebral disorder; Diffusion tensor imaging; multimodality imaging; Perfusion imaging; Central nervous system disease; Degenerative disease; perfusion weighted magnetic resonance imaging; Alzheimer's disease; multiple comparisons; Comparative study; diffusion tensor imaging; multivariate analysis
• ISSN: 1065-9471; 1097-0193
• DOI: 10.1002/hbm.22217

Most brain magnetic resonance imaging (MRI) studies concentrate on a single MRI contrast or modality, frequently structural MRI. By performing an integrated analysis of several modalities, such as structural, perfusion‐weighted, and diffusion‐weighted MRI, new insights may be attained to better understand the underlying processes of brain diseases. We compare two voxelwise approaches: (1) fitting multiple univariate models, one for each outcome and then adjusting for multiple comparisons among the outcomes and (2) fitting a multivariate model. In both cases, adjustment for multiple comparisons is performed over all voxels jointly to account for the search over the brain. The multivariate model is able to account for the multiple comparisons over outcomes without assuming independence because the covariance structure between modalities is estimated. Simulations show that the multivariate approach is more powerful when the outcomes are correlated and, even when the outcomes are independent, the multivariate approach is just as powerful or more powerful when at least two outcomes are dependent on predictors in the model. However, multiple univariate regressions with Bonferroni correction remain a desirable alternative in some circumstances. To illustrate the power of each approach, we analyze a case control study of Alzheimer's disease, in which data from three MRI modalities are available. Hum Brain Mapp 35:831–846, 2014. © 2013 Wiley Periodicals, Inc.