Disclaimer: I know very little about both of the fields in question.
My question is pretty simple:
What's the relation between differential Galois theory and D-modules over algebraic curves?
Differential galois theory can't subsume D-modules obviously since the latter contains also information about behavior at singularities. So, in particular, is differential Galois theory a "natural subset" of D-module theory over curves? If so then in what precise sense? If not why? Still are theirthere any methods/algorithms/ideas from differential Galois theory which are useful in solving problems aboutfor studying D-modules over curves?
Finally, are there any sources discussing this?