Consider creating a "Volume" class

A new "Volume" class could wrap the brain volume data and hold additional metadata like

- metadata from the download
- path on disk
- the non-zero slices
- etc.

It would also provide a standard interface for volume data that would assert a certain d-type, resolution, order of the axes, range of the values etc.

The issue does not name any files, tests, or entry points. Start by locating the code that loads and represents brain volume data, then determine the required metadata and data invariants before proposing the Volume interface; done means those responsibilities and validation rules are agreed and covered by tests.

Written by the indexing model from the issue text.