Diffusion Kurtosis Imaging
Diffusion Kurtosis Imaging (DKI) · Also known as: DKI, non-Gaussian diffusion, diffusion kurtosis
Diffusion Kurtosis Imaging (DKI) is an advanced diffusion MRI technique that quantifies non-Gaussian diffusion of water molecules, providing detailed information about tissue microstructure beyond conventional diffusion tensor imaging. Introduced by Jensen and colleagues in 2005, DKI detects deviations from normal Gaussian diffusion, revealing information about cellular compartmentalization and fiber heterogeneity.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
DKI is appropriate for detecting subtle microstructural changes, when sensitivity to fiber organization is desired, and for clinical assessment of white matter disease. Use DKI for detecting early pathology where standard FA changes are absent. Avoid DKI if high b-value acquisition is unavailable or when standard DTI metrics suffice.
Strengths & limitations
- Detects microstructural changes (myelination, axonal density) not captured by standard DTI
- More biologically specific than FA; kurtosis reflects compartmentalization and restriction
- Sensitive to early pathological changes; can detect disease before visible on conventional MRI
- Complements DTI; provides independent information about tissue organization
- Requires high b-values (>1000 s/mm²) and many gradient directions; longer acquisition times
- More complex acquisition and analysis than standard DTI; prone to noise at high b-values
- Tissue-specific interpretation unclear; kurtosis changes can reflect multiple microstructural factors
- Computational expense higher than DTI due to tensor fitting complexity
Frequently asked
What does kurtosis measure mechanistically?
Kurtosis quantifies non-Gaussian behavior of water diffusion, reflecting compartmentalization and restriction. High kurtosis indicates water is constrained (high restriction); low kurtosis indicates freer diffusion. Mechanisms include cell membranes, myelin, fiber packing, and tortuosity. Interpretation requires complementary histology or biophysical modeling.
What b-values do I need for DKI?
Minimum two b-values with high b-value >1000 s/mm². Standard protocols use b=0, 1000, 2000, 3000 s/mm². Higher b-values improve kurtosis estimates but reduce signal-to-noise. 30–64 gradient directions per b-value typical. Total acquisition time ~10–15 minutes for clinical DKI.
How does DKI relate to NODDI?
Both measure non-Gaussian diffusion but with different approaches. DKI is model-free (pure measurement of kurtosis). NODDI uses biophysical modeling assuming intra-axonal and extra-axonal water compartments. DKI is simpler but less interpretable; NODDI offers more specific metrics (axonal density, dispersion) but requires model assumptions.
Can DKI replace DTI?
Not as a complete replacement. DKI provides complementary information (microstructural detail); DTI provides different metrics (FA, MD). Best practice: acquire both DTI and DKI in clinical protocols; analyze together for comprehensive microstructural assessment.
Sources
- Jensen, J. H., Helpern, J. A., Ramani, A., et al. (2005). Diffusional kurtosis imaging: the quantification of non-Gaussian water diffusion by magnetic resonance imaging. Magnetic Resonance in Medicine, 53(6), 1432–1440. DOI: 10.1002/mrm.20508 ↗
- Lu, H., Jensen, J. H., Topgaard, D., & Helpern, J. A. (2018). Effective medium theory of apparent diffusion coefficient in fibrous media. Magnetic Resonance in Medicine, 81(5), 3245–3260. link ↗
How to cite this page
ScholarGate. (2026, June 3). Diffusion Kurtosis Imaging (DKI). ScholarGate. https://scholargate.app/en/neuroimaging/diffusion-kurtosis-imaging
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Magnetisation Transfer RatioNeuroimaging↔ compare
- NODDINeuroimaging↔ compare
- Tract-Based Spatial StatisticsNeuroimaging↔ compare