A key question in working with rodents in biomedical research is the following: how different are my mice? One solution that we propose is to use imaging; it provides whole animal or whole organ views of anatomy or tissue microstructure, it can potentially be carried out in-vivo, and it's parallelizable for high-throughput studies. In some cases simply looking at the resulting images is sufficient to ascertain abnormalities; more often, however, detailed analyses are necessary in order to extract and quantify phenotypic differences. This is where MICe-build-model comes in: its goal is to use image registration to bring all scans from a particular study into correspondence in order to derive measures about anatomical variation or changes in tissue intensities. Some example uses include:
- Aligning MRI brain scans of mouse models of human disease in order to find regions of atrophy or abnormal shape.
- Aligning lower resolution MRI scans in order to bring extra information - such as perfusion maps, tissue microstructure information (i.e. DTI), etc. - into correspondence.
- Aligning whole embryo 3D optical scans to look for developmental abnormalities.
- Aligning CT scans of skulls to identify cranio-facial abnormalities.
The rest of these pages describe MICe-build-model usage and provide links to further resources. MICe-build-model is freely available to the academic research community - if you are interested in using the software please contact us (see http://www.mouseimaging.ca/about/contact.html for contact info) to discuss terms.
- Algorithm overview
- How to run MICe-build-model
- MBM.py cookbook
- What to do with MICe-build-model output
- Creating the initial model
- Changing the parameters for the non linear stages (-nlin-protocol)
- Disk cleanup after a registration
MICe-build-model is currently being redesigned for greater flexibility and ease of use. More information can be found here.