"Seeing is believing"; "百聽不如一見"; "百聞は一見にしかず"
All idioms above describe how important an image is compared with hundreds of words. We would like reconstruct CT images from their respective sinogram (raw data). Assuming readers are already familiar with Radon Transform, this repository attempts to reconstruct a CT sinogram using the following two popular methods:
- Algebraic Reconstruction Technique (ART) -- colloquially called Iterative Method
- Filtered Back Projection (FBP)
The following table compares each reconstructed image with the original version. Each image has a size of 128 x 128 pixels.
Original Image | ART Reconstructed | FBP Reconstructed |
---|---|---|
ART | FBP | |
---|---|---|
Speed | Slow | Fast |
Resolution | Poor | Good |
Contrast | Good | Poor |
Noise | Low | High |
- Algebraic Reconstruction Technique (ART)
- Using Matlab
- Using Python
- Filtered Back Projection (FBP)
- Using Matlab
- Using Python
- Walkthrough with Jupyter Notebook
- Documentation using DoxygenPy
- Build modules and packages (pip)