VBM

Voxel-based morphometry (VBM) is a neuroimaging analysis technique that enables voxel-wise comparison of local brain tissue concentrations or volumes, typically derived from structural magnetic resonance imaging (MRI). The method aims to detect systematic differences in brain anatomy between groups or in relation to behavioral measures.

VBM involves preprocessing steps such as segmentation of MRI scans into gray matter, white matter, and cerebrospinal fluid; spatial normalization to a standard template; and smoothing of the data. In gray matter studies, voxel values reflect tissue density or probability, and can be modulated to reflect actual volume by accounting for local expansion or contraction (modulated versus unmodulated VBM). Statistical analyses are conducted with a general linear model, allowing covariates like age, sex, and total intracranial volume to control for confounds. Advanced normalization approaches such as DARTEL improve alignment across subjects. Results are interpreted as regions showing differences in tissue concentration or volume.

Applications of VBM include investigations of development and aging, neurodegenerative diseases (for example Alzheimer’s disease or Huntington’s disease), and psychiatric conditions (such as schizophrenia or mood disorders). It is widely used to characterize patterns of atrophy or morphological change and to relate structural differences to cognitive or clinical measures.

Limitations include sensitivity to preprocessing choices, reliance on accurate segmentation and normalization, and the potential for smoothing to blur anatomical boundaries. Interpreting gray matter density versus true volume can be challenging, and cross-sectional designs limit causal inferences. Related methods include tensor-based morphometry and surface-based morphometry, with common software in use such as SPM, FSL, and CAT12.