Differential Tractography - Low R2 Values

[Tyler Diorio]

Hello,

Our group is trying to generate differential tractograms for longitudinal patient data (2017-2019), where we have Axial DTI with 2mm isotropic spatial resolution. When using the pipeline outlined on the DSI Website as well as in the respective video, we are achieving R2 values of 0.45-0.6 for our patient set. 

We are aware that Dr. Frank Yeh mentioned that these low R2 values (<0.75) may be caused by registration issues.  Upon inspection of our images in external software (ITK Snap), it seems that the longitudinal images are not well aligned before importing into DSI Studio (see figure below):

Figure: Longitudinal DTI scans for a single patient, from 2017 (Blue) and 2019 (Green). 

Question: Are there are methods that we can utilize to increase our R2 values or understand why they're so low?

Thank you for your help,

Tyler

[Frank Yeh]

Hi Tyler,

    Sorry for my late reply (the thread went to SPAM folder).

    

Question: Are there are methods that we can utilize to increase our R2 values or understand why they're so low?

    It seems that there are image quality issues, and you can check out this video to see if some of them can be fixed: https://www.youtube.com/watch?v=stL4GMeTC1I

Best regards,

Frank