There is a lot of infrastructure around the MRI in order to support liquid helium storage, cycling, and inter-device pathways. It isn't just what you see in the room.
It might still be a large machine, but a bunch of bottlenecks disappear. With that, it is only a matter of time until a startup develops a much cheaper, smaller, and more efficient device.
Philips is a big player making the assumption that the next generation of MRI will be smaller, cheaper, and more widely available. But, I can confirm smaller players are operating on that assumption as well.
It might still be a large machine, but a bunch of bottlenecks disappear. With that, it is only a matter of time until a startup develops a much cheaper, smaller, and more efficient device.