Regional Homogeneity
Also known as: ReHo, regional synchronization
Regional Homogeneity (ReHo) is a measure of synchronization between a voxel and its spatial neighbors in resting-state fMRI. Introduced by Zang and colleagues in 2004, ReHo quantifies local within-cluster activity coherence, reflecting the degree to which brain regions exhibit synchronized spontaneous activity at rest.
Key highlights
- Detects local functional organization without assuming connectivity structure
- Simple to compute and interpret; directly reflects neighborhood synchronization
- Sensitive to disease-related changes in local brain organization
- Complements connectivity-based measures; ReHo captures local while connectivity captures global organization
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
ReHo is useful for detecting local functional disruption in clinical populations, when interest is in regional organization rather than long-range connectivity, and for whole-brain studies. Avoid ReHo if long-range network organization is the primary focus (use connectivity analysis instead).
Strengths & limitations
- Detects local functional organization without assuming connectivity structure
- Simple to compute and interpret; directly reflects neighborhood synchronization
- Sensitive to disease-related changes in local brain organization
- Complements connectivity-based measures; ReHo captures local while connectivity captures global organization
- Kernel size (neighborhood definition) arbitrary; sensitivity unclear
- Limited interpretability; high ReHo indicates synchronization but not functional meaning
- Sensitive to temporal filtering and motion artifacts; baseline preprocessing choices influence results
- Voxel-wise measure; lacks information about which brain networks are synchronized
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
How is ReHo computed using Kendall concordance?
Kendall rank concordance (W) measures agreement among multiple rankings. For ReHo, the 'rankings' are time series values at each timepoint: rank the voxel and its 26 neighbors at each TR, then compute W. W=1 means perfect agreement (all series rank identically); W near 0 means random agreement. Mean W across all timepoints is ReHo at that voxel.
What kernel size should I use for ReHo?
Standard is 3x3x3 voxels (27 voxels total). This captures local organization without large spatial smoothing. Alternative kernel sizes (5x5x5, 7x7x7) capture broader neighborhoods. No consensus exists; report kernel size used. Sensitivity to kernel size suggests results should be validated with multiple kernels.
How is ReHo different from local correlation?
ReHo (Kendall concordance) is rank-based and robust to outliers. Local correlation is Pearson-based. ReHo is preferred for resting-state fMRI due to robustness; Pearson correlation is standard elsewhere. Results are usually similar, but ReHo is more conservative with extreme values.
Can ReHo be used for clinical diagnosis?
ReHo shows promise as a biomarker—clinical populations often have characteristic ReHo patterns. However, single-subject prediction accuracy is modest (~60–70%); ReHo is most useful for identifying group differences. Research biomarker status; clinical utility remains unproven.
Sources
- 1.Zang, Y. F., He, Y., Zhu, C. Z., et al. (2004). Altered baseline brain activity in children with ADHD revealed by resting-state functional MRI. Brain and Development, 26(7), 429–439.
- 2.Yang, Y., Raine, A., Han, C. B., et al. (2007). Localizing brain abnormalities in ADHD: a meta-analysis of neuroimaging studies. Neuroscience & Biobehavioral Reviews, 31(4), 508–515.
You have read it. What now?
Cite this page
ScholarGate. (2026, June 3). Regional Homogeneity. ScholarGate. https://scholargate.app/neuroimaging/regional-homogeneity