Potential and limitations of diffusion MRI tractography for the study of language

• The strengths and limitations of diffusion MRI tractography are reviewed.
• Tractography suffers from an abundance of false positive and false negative results.
• Resolution, algorithmic details, and validation are problems in tractography.
• Tractography can be used for segmentation and network analysis.
• Care must be taken in the interpretation of neuroanatomical results.

Diffusion magnetic resonance imaging (MRI) is a tremendously promising tool for imaging tissue microstructure, and for inferring large scale structural connectivity in vivo. However, the sensitivity of the technique is highly dependent on methodological details. Acquisition parameters, pre-processing steps, reconstruction models, and statistical analysis all affect the final sensitivity, specificity, and accuracy of a study. In the case of fiber pathway reconstruction in the central nervous system, termed tractography, false positive and false negative results abound, and interpretation of results must take into account the potential shortcomings of the techniques used. This article will review the strengths and limitations of different types of diffusion MRI tractography analysis, and highlight what one can realistically hope to learn from such imaging studies of the human brain.