trying to figure out where to get started with this:
I have a few hundred CT images where certain three-dimensional features in the image (anatomy) are moving in a correlated fashion with a set of radio-opaque markers. These anatomic features can rotate, translate and deform and the markers can move together or sometimes in relation to each other. The position and motion of the markers are indicative of the position and motion of the anatomy in which they are embedded in.
I'd like to develop a model whereby given the positions and motion of the markers I can then predict the position and motion of the full anatomy in which they are embedded and use this for segmentation.
What deep-learning software and techniques are suited for this type of problem?