/diy/ - Do It Yourself

OP here. Spent a couple hours building a (simple) simulated CT scanner and a reconstruction algorithm and I think the results are pretty good. Pic related. I definitely need to try this with some real world datasets, and also adjust it to account for the fact that I'll likely be using a machine with cone beam geometry rather than parallel beams (and my simulation & reconstruction here assumes parallel beams), but to me it's an encouraging result so far. Thanks to all anons who posted helpful information.