Connectometry Analysis

Use --action=cnt to initiate [Step C3 Connectometry].

Examples

1. Use a multiple regression model with three variables (0:SEX, 1:BMI, 2:AGE) to study how BMI (the second variable) affects the brain connection in the CMU 60 connectometry database.

dsi_studio --action=cnt --source=CMU60.db.fib.gz --index_name=qa --effect_size=0.3 --demo=CMU60.txt --variable_list=0,1,2 --voi=1

2. Use ROI and ROA to limit connectometry analysis on subjects with specific demographics: scanner=1 and group not equal to 3.

dsi_studio --action=cnt --source=study.db.fib.gz --demo=demo.csv --effect_size=0.2 --roa=excluded_mni_region.nii.gz --select="scanner=1,group/3" --variable_list=2,4,5 --voi=5

---

Core Functions

Parameter	Description
source	Specify the db.fib.gz file.
index_name	Specify which diffusion metrics to analyze (e.g. –index_name=dti_fa)
demo	Assign the path to the demographic CSV or text file.
variable_list	Assign the study variables for regression. Use commas to include multiple variables. Example: --variable_list=0,1,2 includes the first three variables.
voi	Specify the “variable of interest” for multiple regression. Use --voi=0 for the first variable, --voi=1 for the second, etc. For longitudinal analysis, use --voi=longitudinal.

---

Parameters

| Parameter | Default | Description | |——————–|———————-|———————————————————————————| | t_threshold or effect_size | 2.5 for t threshold, or 0.3 for effect size Assign the T or the effect size threshold for correlational tracking. | | length_threshold| 20 voxel distance | Specify the minimum tracking length for correlations. | | fdr_threshold | 0 | Set a nonzero FDR threshold (e.g., 0.05) to enable FDR control. | | permutation | 2000 | Specify the number of permutations used in analysis. | | thread_count | Hardware max | Define the number of threads for computation. | | exclude_cb | 1 | Use --exclude_cb=1 to exclude the cerebellum from analysis. | | no_tractogram | 1 | Set to 1 to avoid generating 3D tractogram images (output files are still generated). | | normalize_iso | 1 | normalize qa or rdi values by iso to provide more robust results | | tip_iteration | 16 | Number of pruning iterations to reduce noise. | | region_pruning | 1 | remove fragmented findings using the tract coverage region | —

Optional Functions

| Parameter | Description | |———————|———————————————————————————| | select | Select subjects for analysis. Example: --select=Gender=1,Age>20 filters subjects with Gender=1 and Age > 20. | | seed, roi, roa...| Define regions to limit the tracking areas (see --action=trk for details). | | output | Specify the prefix for all output files. Defaults to the demographics file name with the study variable appended. | —

Updates Based on Code Features

Features Found in Source Code:

• Multiple Regression: The class group_connectometry_analysis handles the regression model and adjusts QA values based on statistical models.
• FDR Analysis: The function calculate_FDR computes the false discovery rate to identify significant findings.
• Exclusion of Cerebellum: The method exclude_cerebellum removes cerebellum-related data when exclude_cb=1.
• SPM Generation: Statistical Parametric Mapping is computed via calculate_spm and results are saved using save_result.
• Report Generation: The generate_report method creates a detailed HTML visualization of findings.