That's the same thing.
Yes she does. Just because we can't observe them now doesn't mean they are not there. In principle, using the same technology for mind uploading, one could do a precise enough brain scan to find and identify the neural correlates. It's all encoded in the brain state, there is no immaterial magic or anything causing the behaviours.
In principle, the technology that is hypothetical from science fiction could make this possible?
Beyond that, I think your brain-scan hypothesis assumes that there's a single set of neural elements that can completely define a behavior.
Let's assume for this that you're suggesting that a clearly-bounded and well-defined set of discrete elements can be observed to produce a particular behavior, rather than a single element. I think that's a reasonable assumption.
Now I would like to challenge that approach. I don't see a reason to assume a set of defined elements. The end-state (bipolar disorder or gender dysphoria, or any other behavioral disorder) isn't binary, for one thing. They exist in varying degrees. Based on behavior, I'd guess that my mom and my aunt are both very slightly bipolar, where my sister is middling bipolar. I've met people who are extremely bipolar. That suggests that the set of elements aren't digital but are analogue, for one thing.
But beyond that, I doubt it's a well-defined set to begin with. I think it's more likely to be a very loose cluster of possible contributors. Further more, I suspect that the cluster in question can express in multiple ways, not all of which would fit the definition of a specific behavioral disorder.
So rather than it being a one-to-one relationship (which you seem to assume, and which is likely assumed by the brain-scan hypothesis) I suspect it's a many-to-many relationship.
One thing to bear in mind is that thoughts, conditioning, and behavior are affected by neural infrastructure... but they also affect the infrastructure itself. It's a feedback loop between perception, cognition, and structure. That's part of why neural plasticity is such a wrench in the works when it comes to behavior (psychology and psychiatry as opposed to neurology).
It's pretty well demonstrated that if a person is told a thing enough times, they can come to believe it to be true, even when it's not - especially when those things are qualitative as opposed to quantitative. A child told over and over that they're dumb will believe that they are dumb, regardless of any objective measure of their intelligence. This works for self-talk as well, which is why a lot of visualization techniques are effective.