• 
• 
• 
• 
• 
• 
• 

Tensor Based Morphometry (TBM) For Longitudinal Data

Written by Owen Phillips . Email if you have any questions.

Tensor based morphometry requires all your study subjects to be registered together so that an atlas can be generated. See http://www.loni.ucla.edu/twiki/bin/view/CCB/SIGPNS_Summer2009_BG_CS_2 for more information. This protocol is set up to allow a longitudinal analysis. It will generate one jacobian image for each subject who was scanned at two time points. Right click the following link to download: TBM_Longitudinal_MDT_v9

Inputs:

1. Raw Data Time 1
2. 1st subject - this is just used to set the histograms for the rest of the subjects.
3. Raw Data Time 2
4. *note, this pipeline includes extra modules like Bias Field Corrector, this can be cut out of the pipeline if it is not needed.
   

Outputs:

1. Longitudinal Jacobians
2. 4-step MDT
3. KL_MI_deform 2

Analysis: You can download the pipeline for TBM Longitudinal Analysis by right clicking: One_Sample_t-test

Inputs:

1. Input Design Matrix: See http://wiki.stat.ucla.edu/socr/index.php/SOCR_EduMaterials_AnalysesCommandLineVolumeMultipleRegression for instructions on how to make this.
2. Mask

Outputs:

1. P-value output volume
2. T-statistic
3. FDR corrected P-value

• Site Map ∗
• BMAP Home