Multivariate Tensor-based Subcortical Morphometry System

Quick Start Guide


          The Quick Start Guide to the Multivariate Tensor-based Subcortical Morphometry System (mTBM) covers the fundamental usages. For the more detailed description of the system features, please see the User Guide. The system can only be used under Linux Operating System.


2.1 Executable Code and Software

  • FSL toolbox
  • Executable Code and python scripts in folder mtbm
  • Example in folder example

2.2 Download

2.3 Setup

3.Run FSL toolbox

3.1 Introduction

          The command can segment the hippocampus from the MRI image.

3.2 Command

          >> run_first_all –s L_Hipp,R_Hipp –i <input image with its directory> -o <output name and its directory>

3.3 Input and Output

  • Input: MRI image (.img/.hdr)
  • Output: Hippocampus image segmented from input image

NOTE: The output-file name should be the same with the input-file name and put them in the same folder. The next program will copy the output files automatically.

You can get the detailed description about the command here. 

4.Run mTBM package

4.1 Introduction

          The software will generate a mesh for the input hippocampus image and calculate feature for each vertex in the mesh.

4.2 Usage

         >> python /*directory to the file*/ <input image file with its directory> <output directory>

4.3 Input and Output

  • Input: original image (only input the name of original MRI with its extension and the software would get the other files generated by last step)
  • Output: mesh file with features for each point. The type is .m and you can view the file by using the Matlab file.
  • filename_Lhippo_60k_std_par_flowed_jfeat.m
  • filename_Rhippo_60k_std_par_flowed_jfeat.m

NOTE: DO NOT use relative directory for the input and output.

5.Statistical Group different

5.1 Introduction

The software applied Hotelling’s T test to evaluate the overall significance of experimental result and give a significance p-map with p-values.

5.2 Usage

  • TBM: python /**/HippoUCF/Rainier/UCFGroupDifference/ <Excel file> <input mesh directory> <output directory> <repeat number> <signfic> <group 0 label> <group 1 label> <string>
  • MTBM: python /**/HippoUCF/Rainier/UCFGroupDifference/ <Excel file> <input mesh directory> <output directory> <repeat number> <signfic> <group 0 label> <group 1 label> <string>
  • MADMTBM: python /**/HippoUCF/Rainier/UCFGroupDifference/ <Excel file> <input mesh directory> <output directory> <repeat number> <signfic> <group 0 label> <group 1 label> <string>

5.3 Input and Output


Excel file (.xls) which contains the file names of the mesh got from last step and their group label.

  • Input mesh fold directory
  • Output mesh folder directory
  • Repeat number (usually 10,000 or 15,000)
  • Statistical threshold (0.05)
  • The label of first group
  • The label of second group
  • String of experiment (TBM, MTBM, MADMTBM)


  • A file of p-map (outputname_mvgd_pmap.m)
  • A file with the significance (sigval_MAD.txt). The first line is the significance value if the FDR in the second line is zero. The second line is the significance value after using the theory of false discover rates (FDR).
  • Distribution functions of P values with y=x and y=20x.


  • REMEMBER to put the left-hippocampus files and right-hippocampus into different directories. You need to run the program for left hippocampus and right hippocampus separately.
  • In the excel file, you don’t need to write the full name of your file but just use the first several characters before symbol, ‘_’.