As a research project to do during high school, I am planning to code something that uses AI to recognize “defects” in MRI scans. What is the best way to go about coding this?
Hi, @Arjun_Krish!
For MRI scans I’m assuming you would like to do some kind of semantic segmentation, so I would go for encoder-decoder or UNet type of architectures. I recommend you the segmentation models pytorch package, which I worked with. I’m sure there has to be an equivalent to tensorflow out there.
Would you recommend any course that would teach me how exactly to code the AI part for visual recognition in python? I know how neural networks and machine learning and the rest of AI work conceptually, but I have no idea where to start and code this.
I’m pretty sure the docs have all you need and well explained. Anyway, there are plenty of medium and towarddatascience articles
1 Like