Currently a masters student in physiology, in a non-technical lab. But would like to learn the technical skills required for understanding and applying deep learning (etc). I often see DL/ML/AI applied to fields such as radiology, dermatology, cardiology, and it would be great to not only understand this kind of research, but also be an active participant in the future.

As someone with a limited technical/quantitative background (currently learning R for a biostatistics course), what would be the ideal approach I should take to learn DL/ML/AI?

From my search it seems that there are two ways of approaching the learning roadmap/pathway:

(1) Practical approach: learn the right tools and apply them to the problem at hand. (2) Comprehensive approach: understand the mathematics behind DL/ML/AI so that you have a greater understanding of the tools.

I think many researchers in medicine, especially if they're coming from clinical backgrounds as opposed to Comp Sci backgrounds, use this practical approach. This approach would be more feasible for someone like me, whereas learning all the required mathematics would probably take several years of study (on top of my other responsibilities).

Ultimately, what steps should I take? I think Step 1 would be to learn Python. I am already learning R, but I see Python used more frequently.