bioRxiv Science⌕ Search

Biology subjects

Kaster, M.

Publications and source records attributed to Kaster, M..

2 recordsLinked to original sources

Rewiring the Human Brain: On the Fabric of Associative Thinking

Concept cells are neurons in the medial temporal lobe that represent an abstract concept, such as a familiar person, in a context-independent way. They are activated by heterogeneous and potentially multi-modal sensory input related to the concept, such as viewing a photo of the person, reading the persons name, or hearing the persons voice. Learning the concept implies connecting the cortical cells that are triggered when such features are perceived with the cells of the concept, resulting in an associative memory engram. Traditional models explain the formation of such an engram with Hebbian learning through synaptic plasticity, the strengthening of existing synapses in response to co-stimulation. However, the low edge density of the brain suggests that direct connections in the form of preexisting synapses between these cells are relatively unlikely, rendering such constructs inefficient and volatile. Instead, it is plausible that more persistent concept engrams rely on structural plasticity, involving the creation of synapses de novo. Yet, it has still been unclear how neurons can project their axons across such a considerable distance, finding a target they do not know in advance. In this paper, we simulate the formation of such structural engrams in the connectomes of healthy subjects, offering a model hypothesis of how such engrams can be formed on a structural level. Based on our model, we further demonstrate how activating a concept can trigger related concepts through overlapping sensory associations, in a process we call a percept-concept loop.

neuroscience↗

Building a realistic, scalable memory model with independent engrams using a homeostatic mechanism

Memory formation is usually associated with Hebbian learning, using synaptic plasticity to change the synaptic strengths but omitting structural changes. Recent work suggests that structural plasticity can also lead to silent memory engrams, reproducing a conditioned learning paradigm with neuron ensembles. However, this work is limited by its way of synapse formation, enabling the formation of only one memory engram. Overcoming this, our model allows the formation of many engrams simultaneously while retaining high neurophysiological accuracy, e.g., as found in cortical columns. We achieve this by substituting the random synapse formation with the Model of Structural Plasticity (Butz and van Ooyen, 2013). As a homeostatic model, neurons regulate their activity by growing and pruning synaptic elements based on their current activity. Utilizing synapse formation based on the Euclidean distance between the neurons with a scalable algorithm allows us to easily simulate 4 million neurons with 343 memory engrams. These engrams do not interfere with one another by default, yet we can change the simulation parameters to form long-reaching associations. Our model paves the way for simulations addressing further inquiries, ranging from memory chains and hierarchies to complex memory systems comprising areas with different learning mechanisms.

neuroscience↗