Summary: A new software framework incorporates dendritic properties and mechanisms into large-scale neural network models.
FORTH-IMBB researchers have developed innovative computational tools to unravel the role of dendrites; complex neuronal brain structures with a key role in information processing.
The new software allows the incorporation of important dendritic properties into neural network models. This work has important applications both in understanding brain function and in the field of artificial intelligence.
Deciphering the secrets of the brain is considered to be one of the most important scientific endeavors of the 21st century. A better understanding of the mechanisms underlying brain function will contribute to research on the treatment of neurological diseases as well as the field of artificial intelligence, which has become an integral part of our society.
In a recent publication, in Nature Communications, the team of Dr. Panayiota Poirazi, a Research Director at the Institute of Molecular Biology and Biotechnology (IMBB) of the Foundation for Research and Technology-Hellas (FORTH), presents a new software framework that allows the incorporation of dendrites and their key mechanisms into large-scale neural network models.
Dendrites are the branched extensions of nerve cells that morphologically resemble the branches of a tree (hence their name). Their main function is to receive information from other neurons (in the form of electrical or chemical signals) and transmit it to the main body of the cell.
For decades since their discovery, their role in information processing remained unknown, due to technological limitations in research.
But recent studies reveal that dendrites have a wealth of mechanisms that allow them to perform complex mathematical calculations independent of the main neuron. At the same time, dendrites are equally important for the plasticity of the nervous system, i.e., the ability of the brain to change and adapt to its environment. This process plays a dominant role in complex brain functions, such as learning, memory, decision making and cognition.
Although we now largely understand the contribution of dendrites to the behavior of a single neuron, their implications at the level of networks or entire brain regions remain unexplored.
A small number of studies correlate dendritic complexity with various cognitive markers, and dendrites are known to decline during aging or in neurodegenerative diseases such as Alzheimer’s.
In addition, research on AI has already benefited from the use of dendritic mechanisms as a source of inspiration for the development of new improved and more efficient algorithms.
However, many open questions remain and Dr. Poirazi’s team hopes that the tools they are developing will facilitate the work of those attempting to understand the role of dendrites in brain function.
This study introduces a new software framework that allows even naive users to build neuronal models with dendrites in a simple and efficient manner, thus minimizing computational complexity. Such computational models help to elucidate the role of dendrites in complex brain functions while also facilitating their incorporation in neuromorphic devices, a type of neuro-inspired artificial intelligence architecture.
This effort was led by Michalis Pangalos, Ph.D. candidate in the Department of Biology of the University of Crete, in collaboration with Dr. Spiros Chavlis, postdoctoral researcher at IMBB, under the supervision of Dr. Poirazi.
About this AI and neuroscience research news
Author: Press Office
Contact: Press Office – FORTH
Image: The image is in the public domain
Original Research: Open access.
“Introducing the Dendrify framework for incorporating dendrites to spiking neural networks” by Michalis Pagkalos et al. Nature Communications
Introducing the Dendrify framework for incorporating dendrites to spiking neural networks
Computational modeling has been indispensable for understanding how subcellular neuronal features influence circuit processing.
However, the role of dendritic computations in network-level operations remains largely unexplored. This is partly because existing tools do not allow the development of realistic and efficient network models that account for dendrites.
Current spiking neural networks, although efficient, are usually quite simplistic, overlooking essential dendritic properties. Conversely, circuit models with morphologically detailed neuron models are computationally costly, thus impractical for large-network simulations.
To bridge the gap between these two extremes and facilitate the adoption of dendritic features in spiking neural networks, we introduce Dendrify, an open-source Python package based on Brian 2. Dendrify, through simple commands, automatically generates reduced compartmental neuron models with simplified yet biologically relevant dendritic and synaptic integrative properties.
Such models strike a good balance between flexibility, performance, and biological accuracy, allowing us to explore dendritic contributions to network-level functions while paving the way for developing more powerful neuromorphic systems.