I wish I could help you, but I am only familiar with Kant's account of practical philosophy--and only a fraction of that even. I understand the transcendental idealism from the perspective of the categorical imperative and Kant's brand of deontological ethics. I can tell you that it does have to do with the a priori synthetic, because Kant was all about reason independent of experience, and judgment distinct from the subject.

I'm trying to read the Prolegomena to Any Future Metaphysics right now for whatever reason (it is more readable than the Critique of Pure Reason and occasionally Kant shows hints of stylistic brilliance) and it seems to me that transcendental idealism is basically Kant's saying that for any synthetic reasoning to take place, there must be a first part, a subject from which the predicate is based; and a priori reasoning, that is knowing something about the world before observation of it. In the Prolegomena, taking as his starting point the idea that math and physics have theorems that are both synthetic and a priori, what he says is that the human idea of space and time are the transcendental basis that makes such reasoning possible.

I think.

I don't think you can say that the mind 'creates space and time'. It is more that space and time are the forms that the mind imposes on sense data. In order for the mind to be able to impose space and time on sense data, it must have knowledge of space and time apart from sense data (said knowledge is thus a priori in that it is apart from experience and synthentic in that it is expansive).

Transcendental idealism is the idea that space and time are not properties of the noumena (the world apart from experience) per se, but are only properties of phenemona as the mind creates it. Thus, space and time are transcendentally (apart from outside world) ideal (in the mind).

When studying Kant, always keep his 'Copernican Revolution' in mind, this has always helped me.