Friday, July 31, 2020

The Memory of Places Long Gone

I had a dream last night about my old inpatient psychiatric unit. I was moving up and down the hallways like I used to do looking for people to interview. It always seemed like I had to track people down in order to talk with them. In my dream, I was in and out of familiar rooms and talking with familiar staff. I have known some of the staff people for over 20 years. The place was congested, noisy, and malodorous. There were the associated tensions and anxiety. Everything was in the exact spatial order that I remember it being in - including all of the disarray in my old office. I eventually woke up from that dream somewhat anxious and decided that I was done sleeping.

As I drank a first cup of coffee, I thought about the various meanings in the dream and developed three or four different threads. I am used to analyzing my own dreams and the dreams of others so I had all my theories finalized by my 2nd cup of coffee. At that point I started to think about the visuospatial memory necessary to create that dream. Dreams are certainly fascinating if you just pay attention to the content and the emotional tone but to be they are more fascinating when you consider how much computational power it takes to create a Technicolor dream of your real life experience. I have asked neuroscientists to speculate about this for years now and nobody has come up with any ideas.  I came up with a few of my own based on very simplistic extrapolations of the bandwidth of single neurons and simplistic bundles of neurons. One of the problems is that bandwidth and information calculations are valid only for typical electronic devices. Bandwidth and biological systems has been talked about for 20 years but there are no straightforward or consistent calculations.

A first approximation might be a two-dimensional photograph. Everyone these days is used to seeing photographs and an estimate of how much information is in that picture. But when you are walking through a building, you are essentially walking through an infinite number of two-dimensional photographs with multiple soundtracks and additional sources of information. Even if a measure of the information it takes to portray that data in real time and space was possible the next question is - what happens when it is represented in the brain?

I pulled up Facebook and noticed that someone had posted a photo of my old high school. As I looked at the front of the school I located all the windows where my classes were held. I was in that building for the last six years of school because in our town it contained grades seven through 12. It took me a few minutes to look at the windows and think about significant events that occurred in those classes. When I think about those years, there is a lot of confusion and anxiety. I did not have a very positive experience in middle school or high school. I was one of the most obedient kids in the history of the world and yet I managed to have contentious relationships with teachers. Back in those days if you were on the wrong side of the teachers – you were basically on the wrong side.  Parents didn’t question teacher authority and I knew better than to bring that information back home. But just like my dream I was focused on the structure this building. I still remember the creaky wooden floors and stairwells. I remember the bad acoustics and echoes. I remember where my locker was on the first floor. I remember what the place smelled like. I remembered how to go to the ground floor and take the tunnel to a connected building where I had some regular classes but also mechanical drawing and technical graphics. In other words, I still had the memory of the three dimensional space - long after that space had ceased to exist. I made a short list of other structures where that was also true including my old inpatient unit. That building was demolished about three years ago.

If you think about it visuospatial memory is a fascinating area of study. It takes more than remembering a list of words or calculation or how to play a musical instrument. The data have to be stored and retrieved in a way that makes spatial sense. When I think about the way we assess this memory clinically, the tests are generally crude and two-dimensional. Real live visuospatial memory is much different. As an example I can visualize being in an English class at the northwest corner of my high school. I know the exit I would have to take and the stairwell to get out of the building. I know the directions to get to my locker and I can run through all of this traveling in my head. At some level that is a declarative memory function because I know the information I want, I can retrieve it, and I can do an operation on it and in this case imagine myself traveling according to that information. Why my brain elects to access visuospatial information while I am dreaming is not clear and I don’t think anyone knows the answer to that question.

Early information from lesion studies of parietal cortex was involved in visuospatial memories especially the right parietal cortex.  That was largely because patients with lesions in this area could not complete the basic construction tasks on bedside cognitive screening. Parietal cortex is important in the subjective experience of memory.  For example patient with bilateral parietal cortex lesions have similar recall as controls but they have less confidence in their recall (2).  One of the most well-known studies of visuospatial recall and brain structural changes was a study of London cab drivers.  They are required to acquire a significant amount of visuospatial information about London geography and be tested on it. In the initial study (3) the authors showed that acquiring this information was associated with greater cross sectional area of the right posterior hippocampus on MRI scan.  They suggested: “These data are in accordance with the idea that the posterior hippocampus stores a spatial representation of the environment and can expand regionally to accommodate elaboration of this representation in people with a high dependence on navigational skills.”  In a recent study of typical (non-expert) taxi drivers navigational skills did not seem to correlate with right posterior hippocampal volume (5).  One of the explanations that the authors had for this issue is that the posterior hippocampus is only one part of the structures necessary for the navigational map.

Since this thread of research on the hippocampus, the area of spatial learning has become a lot more complex.  As an example, there are now three different types of spatial learning (6).  Landmark knowledge is specific for objects encountered in the environment.  Route knowledge involves environmentally dependent decisions that have to occur at specific times to move through the environment.  Survey knowledge is a detailed representation of the relationship between all objects in the environment. In reference 6, the authors look at testing all of these types of spatial learning in a large group of people who are navigating a nuclear facility and looking at the impact that maps of varying complexity have on their learning that environment.

As a psychiatrist, one of my concerns is what the neural substrate of spatial learning is for both my own thought experiments but also the patients I see with known brain lesions or who have clear deficits in spatial navigation.  One of the most useful reviews I could find (7) looked at a summary of the available data and proposed a navigational network that is really a rough sketch.  In a fairly complex paper the authors suggest that the occipital place area (OPA), retrosplenial complex (RSA) and parahippocampal place area (PPA) may function for landmark recognition in the context of a larger network involving the hippocampus, parahippocampus, retrosplenial cortex, and entorhinal cortex.  From a clinical standpoint many of these structures are compromised in dementias like Alzheimer Disease with severe associated problems in spatial maps and orientation.   

The story of spatial mapping and memory is an incomplete one at this time, but the details are building. At some point in the near future we will have a better idea of what the brain substrate is for this memory and the various kinds of real or imagined navigation. How it is activated and placed in dreams will probably take additional research effort.  And at some point, I really would like to see a practical approach to estimating the information content and flow necessary to accomplish these tasks. 


George Dawson, MD, DFAPA




1:  Rosen ML, Stern CE, Devaney KJ, Somers DC. Cortical and Subcortical Contributions to Long-Term Memory-Guided Visuospatial Attention. Cerebral Cortex (New York, N.Y. : 1991). 2018 Aug;28(8):2935-2947. DOI: 10.1093/cercor/bhx172.

2:  Simons JS, Peers PV, Mazuz YS, Berryhill ME, Olson IR. Dissociation between memory accuracy and memory confidence following bilateral parietal lesions. Cereb Cortex. 2010;20(2):479-485. doi:10.1093/cercor/bhp116

3:  Maguire EA, Gadian DG, Johnsrude IS, et al. Navigation-related structural change in the hippocampi of taxi drivers. Proc Natl Acad Sci U S A. 2000;97(8):4398-4403. doi:10.1073/pnas.070039597

4:  Maguire EA, Woollett K, Spiers HJ. London taxi drivers and bus drivers: a structural MRI and neuropsychological analysis. Hippocampus. 2006;16(12):1091-1101. doi:10.1002/hipo.20233

5:  Weisberg SM, Newcombe NS, Chatterjee A. Everyday taxi drivers: Do better navigators have larger hippocampi?. Cortex. 2019;115:280-293. doi:10.1016/j.cortex.2018.12.024

 6: Stites MC, Matzen LE, Gastelum ZN. Where are we going and where have we been? Examining the effects of maps on spatial learning in an indoor guided navigation task. Cogn Res Princ Implic. 2020 Mar 20;5(1):13. doi: 10.1186/s41235-020-00213-w. PMID: 32198712; PMCID: PMC7083990.

 7: Epstein RA, Patai EZ, Julian JB, Spiers HJ. The cognitive map in humans: spatial navigation and beyond. Nat Neurosci. 2017 Oct 26;20(11):1504-1513. doi: 10.1038/nn.4656. PMID: 29073650; PMCID: PMC6028313.

 8:  Bellmund JLS, Gärdenfors P, Moser EI, Doeller CF. Navigating cognition: Spatial codes for human thinking. Science. 2018;362(6415):eaat6766. doi:10.1126/science.aat6766

 9:  He Q, Brown TI. Environmental Barriers Disrupt Grid-like Representations in Humans during Navigation. Curr Biol. 2019;29(16):2718-2722.e3. doi:10.1016/j.cub.2019.06.072

 10:  Eichenbaum H. The role of the hippocampus in navigation is memory. J Neurophysiol. 2017;117(4):1785-1796. doi:10.1152/jn.00005.2017



No comments:

Post a Comment