A Theoretical Aside
So this is new for me, I am explicitly going to talk about a paper of mine that was just accepted for publication. Among other things I have been learning in my foray into writing a blog is that it is critical that we be able to speak about our own research, along with what interests us most about the research others are doing.
I am going to talk about a paper entitled, “Embracing Complexity: Using the Attribute Model to Elucidate the Role for Distributed Neural Networks Underlying Spatial Memory Processes” that will appear in the inaugural issue of OA Neurosciences.
Now instead of my normal linkout, I have decided I am going to start posting my own papers on a Github.io site so it is static, so the link to the HTML of this paper is here.
So this was a somewhat odd paper for me to write. My goal in this paper was to express in as concise a way as possible that the models that we are using to test brain function are grossly over-simplified. Now I am not just talking about Squire’s old Declarative/Nondeclarative parsing or the Cognitive mapping theory from O’Keefe and Nadel. I am referring to the models that we use today to test brain function.
My impetus to write this theoretical review was that I am watching as the field moves away from the ablative lesion work I am trained in, even away from palce cell recording, and into optogenetics, DREADDs, and increasingly more advanced methods of data acqisition to study brain function. However, despite these advances, the theories being tested using the state of the art techniques are often best described as gross oversimplifications. This is alright, I guess, in the context of designing a question to test in the first place, but I think it has led to a lack of subtlety in our study of systems neuroscience.
I see papers where there is an inactivation of some sort in the hippocampus and the authors proudly state that they have “disrupted” spatial memory. Great. Can anyone please tell me that this actually means? I set out to answer this question.
So my paper uses an old-ish model developed by my undergraduate mentor, Ray Kesner. It is called the Attribute model and is really just a logical extension of the Atkinson-Shiffrin memory model. In this model, all sensory inputs enters the brain and is processed by every region that the information passes through. As such, there is no hippocampus=space, amygdala=fear, formation of boxes. There is an emphasis on what type of information is processed where and how did that region process it. The resulting data can be used to design experiments to more concisely disrupt brain function and ask very specific questions about brain function.
So my use of this model was to extend this model to its logical conslusion so far as “spatial memory” is involved. So instead of focusing on the hippocampus (a surprise to anyone that knows me), I spread the focus across most of the brain and what contributions to spatial processing each region provides.
What was exciting in writing this review was that there was a lot of work I had overlooked involving thalamic, rostral cortical, and retrosplenial contributions to spatial memory.
Before I drone on and on I will end this post herre, but I would love to see this model applied by labs other than my own to ask very specific questions about spatial memory and related behaviors.