Introduction
This is an interface that can automatically assess the quality of rat MCAO procedure. It is based on positron emission tomography (PET) imaging, 3D scale-invariant feature transform (SIFT) and support vector machine (SVM).
Download
Here are the related data and files for users to download and useļ¼
Data: (1) Training data (2) Test data (3) Validation data (4) TTC staining photos (5) BC0011.img/hdr(for manual align)
MATLAB files: (1) PET_template.img/hdr (2) target.img/hdr (3) brainmask.img/hdr (4) keypoints.mat (5) Preprocessing_automatic_script (6) SVM_model.mat
Example
 |
 |
 |
 |
 |
 |
Examples of six PET scans taken after MCAO that show variation of intensity patterns in the injured right hemisphere. First row: regions indicated by the arrows have hyper-uptake pattern. Second row: regions indicated by the arrows have hypo-uptake pattern. Third row: no obvious difference can be visualized between the two hemispheres. These images have been spatially normalized but not smoothed.
User instruction
You have different options to use our method:
- You can roughly align your raw images to the exemplary image we provide. Then upload your images to the interface in the Demo section. After a few minutes, you will get the assessment result. This is the simplest one.
- You can also follow our procedure in the paper to carry out preprocessing and SIFT feature extraction yourself. Then download the SVM_model.mat and carry out classification on your own data. In this option, you do not need to upload the data onto our server.
- You can also download our data and combine with your data to train your own classification model.
Environment configuration
Please download MATLAB, statistical parametric mapping and 3D SIFT to use this tool.
If you intend to use our online interface directly, you can use any software that has rigid-body transformation of your images such as SPM, MIRTK, etc.