Diffusion MRI

Diffusion-weighted and Diffusion Tensor MR imaging

出版元 Denis Hoa


After reading this chapter, you should be able:

  • To describe the diffusion phenomenon and the different types of diffusion (free, restricted, anisotropic)
  • Explain diffusion-weighting and the sequences used in diffusion MRI
  • Describe the relationship between T2, diffusion-weighted image and apparent diffusion coefficient as well as the T2 remanence effect
  • List the artifacts that can appear in diffusion MRI
  • Develop the basic principles of diffusion tensor and its interest
  • Present the main applications of diffusion and diffusion tensor MRI


  • Diffusion MRI explores the micromovements of water molecules. The diffusion of these molecules may be free (as in cerebrospinal fluid) or restricted (by cell membranes, macromolecules, fibers…). These molecules can move in all spatial directions (isotropic diffusion) or in a specific manner in a given direction (anisotropic diffusion) as in nerve fibers.
  • Diffusion weighting derives from applying diffusion gradients to each side of a 180° pulse. The higher the b-factor the greater the diffusion weighting. The b-factor depends on the characteristics of the diffusion gradients (amplitude, duration, spacing). The imaging sequence used is generally a SE-echo planar type with parallel reconstruction technique.
  • These sequences place stress on the gradients, which are a source of several artifacts of the image distortion type. Added to this is the high sensitivity of the echo planar to magnetic susceptibility artifacts.
  • The diffusion-weighted image has an associated T2 weighted part. Acquisition must be repeated with gradients oriented in each of the 3 directions in space. With 2 acquisitions with different b-factors (typically b = 0 and 1000 s/mm2), it is possible to calculate the apparent diffusion coefficient (ADC) to remove the T2. This is useful namely in cases of lesion with a T2 hypersignal, appearing as a hypersignal on diffusion-weighted imaging (differentiation between T2-shine-through and diffusion restriction).
  • To study diffusion anisotropy and the direction of diffusion in the voxels, the number of diffusion directions to acquire must be multiplied (at least 6 for the diffusion tensor model and up to several hundred for the other models). Thanks to better angular resolution and depending on the model, it is possible to deduce the preferred diffusion directions and use these to reconstitute the nerve fiber trajectory (fiber tractography). These algorithms are still being developed and the diffusion tensor model has limitations (difficulties with fiber-crossings, divergent or convergent bundles…).
  • The applications of Diffusion MRI in neuroradiology are increasingly widespread, chiefly in ischemic strokes, but also in tumoral, infectious, and inflammatory brain pathologies. Fiber tractography is beginning to be applied in pre-neurosurgical check-ups.


  1. Luypaert, Boujraf. Diffusion and perfusion MRI: basic physics. European journal of radiology. 2001 Apr;38(1):19-27.
  2. Huisman. Diffusion-weighted imaging: basic concepts and application in cerebral stroke and head trauma. European radiology. 2003 Oct;13(10):2283-97.
  3. Bammer. Basic principles of diffusion-weighted imaging. European journal of radiology. 2003 Mar;45(3):169-84.
  4. Habas. [Basic principles of diffusion tensor MR tractography]. Journal de radiologie. 2004 Mar;85(3):281-6.
  5. Le Bihan, Poupon. Artifacts and pitfalls in diffusion MRI. J Magn Reson Imaging. 2006 Sep;24(3):478-88.
  6. Hagmann, Jonasson. Understanding diffusion MR imaging techniques: from scalar diffusion-weighted imaging to diffusion tensor imaging and beyond. Radiographics. 2006 Oct;26 Suppl 1:S205-23.
  7. Kremer, Oppenheim. [Diffusion MRI: technique and clinical applications.]. Journal de radiologie. 2007 Mar;88(3 Pt 2):428-43.
  8. Oppenheim, Ducreux. [Diffusion tensor imaging and tractography of the brain and spinal cord.]. Journal de radiologie. 2007 Mar;88(3 Pt 2):510-20.