๐งญ Workshop Week 4 Outline
- Differential Tractography
- Cross-sectional analysis
- Longitudinal analysis
- Correlational Tractography
- Cross-sectional analysis / group comparison
- Longitudinal analysis
๐ง Session 1: Differential Tractography
Adds an additional termination criterion:
โ Tracking stops when the signal difference exceeds a defined threshold
๐งช Hands-On Exercises
๐ Example Dataset:
- [OpenNeuro (disease)][ds001378] SCA2 study
๐งฌ Longitudinal Analysis
- Compare follow-up vs. baseline for patients
- Compare follow-up vs. baseline for controls
- Compute false discovery rate:
(Findings in controls) / (Findings in patients)
๐ Cross-Sectional Analysis
- Build a population average template
- Download control subjectโs qsdr.fz file: select matching โ
*control*ses-01*qsdr.fz
- [Tractography (Batch)] โ [Create Population Average]
- Download control subjectโs qsdr.fz file: select matching โ
- Compare patient vs. average and control vs. average
- Compute false discovery rate:
(Findings in controls) / (Findings in patients)
๐ง Session 2: Correlational Tractography
Identifies fiber tracts that correlate with a variable or condition
โ Based on Spearman correlation across subjects
๐ Prepare the Database
Aggregate metrics (e.g., FA, QA) from: QSDR reconstructed files (*.qsdr.fz
)
๐ Example Dataset: SCA2 study
๐งช Analysis Types
๐ Cross-Sectional Analysis
- Correlate diffusion metrics with variables across subjects
- FDR v.s. tract length
๐งฌ Longitudinal Analysis
- Changes in patients using unfiltered changes
- Changes in patients versus healthy control using filtered changes
๐งฌ Post-hoc Analysis
- Segment tracts and compute associated values