Let's say for a moment that the mind is of unknown substance, and that the state of a particular brain region is emergent from it. That is certain states of the mind will cause certain states of this brain region, but the states of this region will not cause any changes in the mind. The states of this region are an epiphenomenon of the mind.
Question: Has this or a mechanism for it been discussed?
Now normally Epiphenomenalism holds that the mind is an epiphenomenon of the brain. But given the brain can have input only regions this means that the above scenario can obtain while at the same time the mind is emergent from the brain(as a whole)... The only concession the epiphenomenalist need to make is that the mind is of substance that is poorly understood.
Question: Has the body-mind problem been attacked in this way before?
Another way to state it is: The mind may well be an epiphenomenon of the brain, but the brain can have a specialized sensory organ the states of which is epiphenomenal to the mind. Note that the mind can be of a different substance(dualism) or purely emergent(non-dualim).
The use of "epiphenomenalism" here is meant to show that the mechanism of mind-to-brain causation may be subtle and difficult to observe by normal means (because the cause-effect relationship is unidirectional and requires observing of the inner workings of the mind). Ultimately though it suffers the same criticism Descartes' "pineal gland" do, as being a postulated point of interaction without a theory of the mechanism nor any observation to support it.
Moreover the "brain region" may not be homogeneous, and may be distributed and fractional. As far as I can see our best hope of finding it would be to map the whole brain and finding neurons for which the function is unaccounted. Even then it may be that axons or dendrites are the receptors, or sub-molecular structures serving to change threshold potential...