If you are happy to be supported in a vat and talk to people via them analysing your MRI scans that's fine, but for any quality of life I think we are talking seriously impossible.
Why through MRI? There are plenty of developments in artificial eyes, ears, and limbs. The brain has I/O ports and the ability to adapt to to what's connected to them. There's no reason why senses and limb control can't be superior via artificial equipment. Brains could probably adapt to multiple eyes and different inputs (seeing sounds, radio waves, etc). Colonists inside Europa would have bodies optimized for activities in a vast, dark under-ice environment. Humans could even colonize Jupiter. Maybe with the right bodies, it would be an exciting and beautiful world to live in.
Artificial connections to the brain are, AFAIK, a solved problem. Reliable and superior (and size and cost) are just a matter of refinement of technology. There are already people self-experimenting with body modification (I forget the term they use), so that's going to happen.
A brain in a box would lose some pleasures, such as enjoying food, but that's just pleasure provided via natural I/O connections, and there's no reason why they couldn't be replaced by superior pleasurable inputs. Addictions to such inputs would be a problem to be faced.
Artificial vision superior to natural eyes is pretty much guaranteed in the not-too-distant future. I think that will be solved before all the eye diseases and injuries can be solved. Likewise, artificial limbs, with superior comfort and capabilities might be possible before bone and joint and blood vessel problems can be solved by medical research. The former problem is just a matter of refinement of technology, while the latter has to deal with interactions with other biological systems and still-unknown interactions. Inject the latest "knee cartilage stem cell treatment" and 20 years later find out that it's causing spleen cancer or whatever. The body simply presents too much complexity to model completely.
Which will come first: restoring full function after spinal damage, or connecting to the spinal nerves directly for I/O to artificial limbs complete with senses?
I think this is a question that someone should investigated properly, doing all the number-crunching for rates of advances in medicine vs artificial support, including rates of successes per dollar invested. It's too complex and too important to just make assumptions, such as "a brain in box can only interact via MRI scans".