I am a PhD student researching Alzheimer's Disease (AD) modeling. I am attempting to batch-process DTI data to extract raw fiber count matrices (116x116) between brain regions using the AAL116 atlas. My goal is to export these as .csv or .mat files for network analysis.
I have followed the pipeline but encountered several issues:
1. Batch Processing (Tractography - Batch)Steps Completed: I successfully performed "DICOM to SRC" (Step B1c) to generate .sz files, and "Run Reconstruction" (Step B3) using the GQI method (Step T2b) to generate .gqi.fz files.
The Issue: In Step B4 (Fiber Tracking), when selecting the .gqi.fz files, the AAL116 template does not appear under "Select Atlas -> human." Even though I manually downloaded the AAL116 NIfTI template and its corresponding .txt file and placed them in the atlas folder, they still do not show up in the batch interface.
Since the batch failed, I tried the individual interface:
Tracking: I loaded a .gqi.fz file, set the max seed to 100,000, and ran Fiber Tracking.
Atlas Loading: I used "Regions -> Open MNI Region..." to successfully import my AAL116 template, which correctly generated the 116 regions.
Matrix Export Issue: When I navigated to Tracts -> Connectivity Matrix..., I encountered several problems:
Connectivity Matrix: It generated a whole_brain_ROIs_cg.txt file (which seems to contain spatial coordinates rather than a matrix).
.mat file: The exported .mat file appears to be corrupted and cannot be opened in MATLAB.
Network Measures: It produced a .txt file containing topological metrics (e.g., small-worldness, efficiency), but not the 116x116 fiber count matrix I need.
Missing CSV: There is no apparent option to export the matrix directly as a .csv file.
Is it possible to batch generate 116x116 raw fiber count matrices (in .mat or .csv format) for multiple subjects using a custom AAL116 atlas?
Am I missing a specific step in the UI, or is this a known limitation of the current version?
Is there a comprehensive PDF User Manual available? I could only find functional descriptions and a Wiki on the official website.
Any guidance on how to correctly extract these matrices would be greatly appreciated as it is critical for my research on AD modeling.