I am trying to implement a 3D MRI geometric distortion analysis tool.
My inputs are:
- CT scan (3D) slices of the MRI phantom which has rods that intersect forming a grid
- MRI (3D) slices of the same phantom
The algorithm will first start by aligning the external fiducials on the phantom surface.
Next the algorithm will find all intersection point of the rods in the CT image and then on the MR image.
Finally the algorithm will compare intersection points to calculate distortion.
Please see image
I am looking for general pointers as to how to implement this. I am OK with coding all of it but just wanted to throw this out there to see who has any ideas first.