This software package implements MABMIS: Multi-Atlas Based Multi-Image Segmentation – an algorithm for effective atlas-based groupwise segmentation.
0. Download packages
Software package for download at
- NITRC MABMIS download page:
- IDEA Lab MABMIS download page:
1. Software manual: download.
This software package implements MABMIS: Multi-Atlas Based Multi-Image Segmentation  – an algorithm for effective atlas-based groupwise segmentation, which has been published as:
Hongjun Jia, Pew-Thian Yap, Dinggang Shen, "Iterative Multi-Atlas-Based Multi-Image Segmentation with Tree-Based Registration", accepted for NeuroImage.
The package is available free to the public for the academic research purpose at http://bric.unc.edu/ideagroup/free-softwares/.
This software is distributed WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
2. System requirements
This software is developed using ITK and has been tested on Windows XP (32-bit) and Linux (64-bit, kernel version 2.6.18-194.el5).
The required input is a set of 3D MR atlas image and another set of target intensity images to be segmented (in Analyze format with paired .hdr and .img files) with a text file (.txt) listing all header file (.hdr) names. The output is the set of segmentation images corresponding to all target intensity images.
3. Image pre-processing
The images should be in Analyze format with paired header and image files. All images should be preprocessed so that they have been affinely aligned together.
4. Software package
Package for win-32bit includes:
MABMIS_GUI_main.exe, MABMIS_command.exe, MSVCP80.dll, MSVCR80.dll, MSVCM80.dll, tcl85.dll, tk85.dll, a tcl/tk library folder: lib, and vcredist_x86.exe
Package for linux-64bit includes:
Test data for MABMIS includes 8 atlases with labels (AA ~ AH, extracted from LONI) and 6 images to be labeled (UA, UB, VA, VB, WA and WB, extracted from ADNI). Please input 8 in the field of number of atlases and select all_ids.txt as the IDs file.
5. Software document
Please read lastest MABMIS software manual.
 Hongjun Jia, Pew-Thian Yap, Dinggang Shen, "Iterative Multi-Atlas-Based Multi-Image Segmentation with Tree-Based Registration", accepted for NeuroImage.
7. Contact us
For any questions or bug reports, please email firstname.lastname@example.org.
8. Update history
20110728 Update NITRC download webpage
20110725 Update packages
20110502 Upload files