I've found it useful to think in terms of how such structures might have evolved. The beginning state is always something very simple: the first nerve. It must go from A to B. Perhaps from a food sensing area to a muscle. Sense food, stop moving, something like that. Organisms could evolve multiple copies of that structure: N nerves going from A(n) to B(n). At some point the next invention arises: some logic gate type thing that branches and/or combines nerves. So now sensing food can make movement stop and also digestion begin. Or food response can be attenuated during the day time. And so on. There's no reason why you'd have exactly one of these "nerve processor" nodes. There's also no reason why several different kinds of them and different instances might evolve. They're useful. Then at some point the next discovery is made: by combining several nerve processor nodes in close proximity, with their own inter-node connections, you can perform more complex processing. You've now got a proto-brain. But again there's nothing to say you have exactly one of these things, or that they're exact replicas of each other.