A perspective piece with Bruce McNaughton on how Environmental enrichment can serve as a biological model for developing AI models capable of forward knowledge transfer. #neuroAI #neuroscience
@elduvelle_neuro Very Good point! In two environments, I do have a reward site within 50cm of teleportation (lap start), whereas in the other environment, it is further (125cm) away.. let me compare the replay in two environments and get back to you.
@elduvelle_neuro I'm recording from awake head-fixed mouse while they're exploring N=3 VR environments presented in an interleaved order. I see awake replay at the lap start (potentially another salient/surprise location especially in head-fixed VR with random environment order). Although, more replay events are near reward sites.
I’ll be presenting a poster about #HippocampalReplay at #SFN2023!
The session: https://www.abstractsonline.com/pp8/#!/10892/session/1900
The poster: Hippocampal replay does not reflect trajectory planning in a spatial planning task
The date: November 15, 2023, 8:00 AM - 12:00 PM
The location: Walter E. Washington Convention center, Washington D.C., WCC Halls A-C
Will you be there?!? There is also an online version!
Teaser video ⬇️
#Hippocampus #PlaceCells #FlexibleNavigation #NeuroRat #Neuroscience
1/2
@elduvelle sorry for the delayed response. Attending a course in woods hole. I've never coded phy plugin before. But, I've seen plugins code from others. I'm sure we can figure it out.
@WorldImagining @elduvelle @tyrell_turing @tdverstynen
One of the complexities is whether SWRs can occur in the middle of theta. Personally, I have never seen anything that I would call an SWR in theta. But Csicsvari has argued that SWRs also occur in the middle of theta. (Jozsef has said as much at his posters in SFN meetings in the past. I don't know what his current take is.) Loren Frank has argued that SWRs and theta cycles are two edges of a continuum. (Loren has said this in person - I don't know if he's said it in print and I don't know what his current belief is.)
My current belief is that these are very different processes, that SWRs do not occur during theta, and that people often mislabel high-gamma (140-180 Hz) signals that do occur in theta cycles as SWRs because they stretch their SWR bandwidths too low.
Note: belief = best story so far (belief in a Bayesian sense. :)
@elduvelle line 10 of the rezToPhy2.m script. The waveform templates (default length=61, ops.nt0 samples) are pre-padded with zeros (1+20 (ops.nt0min)), making total length 1+20+61=82. The ops values are present in preprocessDataSub.m script. Hope this helps.
@elduvelle that would be 82 samples. So, for a 30KHz recording, it would be about 2.7ms
HIPPOCAMPAL HISTORY TOUR PART 11: Mike Hasselmo
#Hippocampus #HippocampusHistory #HippocampusGurus
Today's tour expands to colleagues from the generation that followed most of those who've been featured up to now. Mike Hasselmo gets us started - a friend whose work I've followed and admired for several decades. Here's what Mike has to say:
What got you interested in the hippocampus?
As an undergraduate, I was interested in how the complexity of thought could be described in terms of the principles of physics. I started out as a linguistics major, but I remember going up to Noam Chomsky after a talk and asking him if he could recommend a book that describes how neurons mediate language function. Such a book does not exist even today. However, I came across O'Keefe and Nadel's The Hippocampus as a Cognitive Map in the 6th floor library at William James Hall. Their book made me realize that memory function was a complex cognitive process that could be understood in terms of neurophysiological circuits. I wrote a letter to John O'Keefe to ask to work in his lab (he politely wrote to say there was no room), and then I joined the Rolls lab at Oxford in 1984 because I wanted to do unit recording in hippocampus. Edmund encouraged me to work on inferotemporal cortex instead, but I learned about hippocampal modeling there (and met people like Larry Weiskrantz and Richard Morris). I went as a post-doc to Jim Bower's lab at Caltech partly because he was modeling the piriform cortex as an associative memory in a manner very similar to region CA3 of the hippocampus. It was only once I was an Assistant Professor in 1991 that I could finally start projects focused on the hippocampus (initially studying the cholinergic presynaptic inhibition of glutamatergic transmission that I had studied in the piriform cortex).
Aside from your own work, what findings about hippocampus (and related brain parts) in the past 50 years most excited you, and why?
The finding of theta phase precession was particularly exciting. I remember Matt Wilson telling me about his Hippocampus paper with Bill Skaggs and Bruce McNaughton over lunch during the 1996 Computational Neuroscience conference in Cambridge. I knew about the O'Keefe and Recce, 1993 paper, but the full excitement of the finding sank in then. To me, theta phase precession is still one of the most intriguing phenomena in brain function, as it provides such clear evidence for temporal coding by spike times. I have always found rate coding to be very unsatisfying as a framework for coding the complexity of experience. Theta phase precession is exciting as it so clearly links dimensions of behavior to a temporal code in individual neurons. Also, because it involves a precession of membrane potential relative to field potential, it indicates that membrane potential dynamics are essential to the code. I still think we have not fully explored the ramifications of this phenomenon.
Another exciting finding was the dramatic hippocampal fMRI activation during encoding of complex novel stimuli (Stern et al., 1996). Previous studies of encoding and retrieval in verbal memory tasks had seen little hippocampal activity, so this study was striking for the robustness of hippocampal activity.
And of course the discovery of both grid cells and boundary vector cells were exciting as they both showed such elegant quantitative properties of neurons that appear to contribute to place cell firing properties.
Can you relate one personal story about interactions with colleagues that most exemplifies the world of hippocampal research?
To me, the world of hippocampal research is exemplified by the friendliness and enthusiasm of the researchers. I remember first meeting Howard Eichenbaum after a talk that he gave when I was a junior faculty member. He gave me priceless advice to switch from submitting NIA R01 grants to submitting applications to the Computational Neuroscience program run by Dennis Glanzman at NIMH. I am grateful for the support Howard and others gave me, and I have sought to be similarly supportive of my junior colleagues. I don’t have specific stories, but I have many fond episodic memories of meeting researchers for the first time and feeling the excitement of shared interest in understanding the hippocampus. I remember meeting Gyuri Buzsaki for the first time at a conference in Paris and discussing his two stage model of memory. (I also remember asking him at dinner about what something was on the menu and when it turned out to be kidney, Gyuri actually exchanged dinners with me!) I remember the first time I met Lynn Nadel at a meeting at the Arizona Inn in Tucson. I remember Bruce McNaughton coming to my poster at SFN and our discussion of facilitating synapses in different lamina of dentate gyrus. I remember the excitement of meeting younger researchers and connecting over our shared interests. I remember the first time I met David Redish at the Computational Neuroscience meeting, the first time I met Neil Burgess at the NeurIPS meeting, the first time I met Colin Lever at UCL, and the first time I met Randy O’Reilly at the student lunch when I spoke at Carnegie Mellon. I also have fond memories of the first time that I met my future grad students such as Brad Wyble and Lisa Giocomo and Mark Brandon and Jim Heys. There are so many memories that I can’t list them all. My associative network of knowledge about the hippocampus is intertwined with my memories of conversations with the researchers that shared that knowledge with me.
What would tell a young researcher interested in the hippocampus to focus on now?
I would tell young researchers to focus on linking membrane potential dynamics to network function. That has always been my ultimate goal and I'm afraid that the field has drifted away from this goal rather than embracing it. The third wave of neural network research has distracted us from the central question of how the complexity of membrane potential dynamics of neurons mediates complex function. The new label of deep learning doesn’t change the fact that the current models use essentially the same backprop algorithms used in the second wave of neural networks in the late 1980s. With the recent focus on RNNs and DNNs, I see less focus on essential neurophysiological properties such as spike frequency adaptation and presynaptic inhibition and rebound spiking and bursting and dendritic dynamics than twenty years ago, but I still feel these neuronal dynamics are essential to understanding the network dynamics of brain circuits. We aren't going to understand the brain with ReLu units in DNNs or RNNs. We need theory and experiments that link membrane conductances to network function. My hope is that development of voltage imaging in population of neurons will finally provide the data to make this link.
HIPPOCAMPAL HISTORY TOUR, PART 3
#hippocampus #HippocampusGurus
#HippocampusHistory
Today we hear from Gyuri Buzsaki, whose career reminds us that some very important neuroscience happened in what was then behind the so-called iron curtain separating west from east in Europe. Endre Grastyan, who Gyuri mentions, was the head of a very important lab in Pecs, Hungary, and published a number of critical papers in the 1950s and 60s. He visited McGill when John O'Keefe and I were grad students and we were all very impressed. Perhaps we will hear more about Grastyan down the road. For now, here is what Gyuri had to say about my questions. I hope readers will note, once again, the incredible sense of community and sharing that existed in our field in the early days.
1. What got you interested in the hippocampus?
My dear mentor, Endre Grastyan was the first investigator to record from the hippocampus of a freely moving animal. He recognized that theta is much faster (5-6 Hz) when the cat orients towards and approaches a novel signal, compared to the slow rhythm in the anesthetized rabbit (Green and Arduino, 1954). Thus, I ‘inherited’ the hippocampus, theta oscillations and all their problems from Endre.
2. Aside from your own work, what findings about hippocampus (and related brain parts) in the past 50 years most excited you, and why?
Place cells of course. However, for many years place cell research remained a bit of phenomenology. Although I have seen them myself and proudly showed them to the visitors of my lab, my real interest in place cells started with John O’Keefe’s second discovery, the phenomenon of phase precession. Now, time (i.e., phase) entered into the place field literature and my interests in theta and other oscillations and behavioral correlates of pyramidal cell firing inevitably converged. My fascination was really about this problem: how can the spike phase of theta inform the rest of the brain about the animal’s position when theta is internally generated thus it’s phase varies randomly every time the animal enters the place field. Eventually, this was the path for our formulation of internally generated neuron assembly sequences that could be related to episodic memory. Perhaps I should mention another fascinating aspect of the hippocampus that bugs me a lot but never devoted enough time to address it. What make the hippocampus special is the granule cell. There is no such cell type in the neocortex. Granule cells keep dividing postnatally and the only cell type known whose life depends on circulating hormones (steroids). The latter was discovered by Bob Sloviter yet nobody really followed up on this. Why is there such hormone-neuron connection?
3. Can you relate one personal story about interactions with colleagues that most exemplifies the world of hippocampal research?
Although back then I took it for granted, in retrospect I am amazed how kind people were toward me in my younger days. John O’Keefe, Phil Best, John Disterhoft, Chuck Ribak, Fernando Lopes da Silva and others visited us behind the Iron Curtain to see our set up and methods. And when I eventually could leave Hungary, I was invited to give departmental seminars as if I was somebody. Here is a story my wife and I will always remember. I drove up from San Antonio, TX (where I did my first of many postdocs with Eddie Eidelberg) to NYC to pick up my wife who was eventually allowed to visit me. I had enough money to buy her a ticket to NYC but not all the way to Texas. On the way back from New York, I visited several neuroscience lab and we were always hosted by neuroscientist colleagues and their families wherever we stopped. The great Mort Mishkin invited me to give a seminar and arranged to pay me an honorarium of $150. For this large sum of money, we could get our first Peking duck dinner, buy enough gas to drive back to Texas and visit the famous jazz places in New Orleans. On our trip to California, I gave two departmental seminars at UCI and Jim McGaugh put us up in the Laguna Beach Hotel (I cannot afford it today!) and invited us for a pool party in his house. Looking for a new postdoc, my wife and I drove to Canada. Brian Bland hosted me for two weeks in Calgary and our families became good friends. I called up Case Vanderwolf telling him that I need a job and he hired me instantly (after we drove the Trans-Canada highway from Calgary to London, Ont). And the list goes on. I could never repay the generosity of these and many other future colleagues for their support.
4. What would tell a young researcher interested in the hippocampus to focus on now?
Whatever their interests are. Remember, the most important question in science is always YOUR question.
Studying neural correlates of learning and memory.
Neuroscience PhD student, McNaughton lab
@UCIrvine