Loading…
CNS*2020 Online has ended
Welcome to the Sched instance for CNS*2020 Online! Please read the instruction document on detailed information on CNS*2020.
Back To Schedule
Sunday, July 19 • 7:00pm - 8:00pm
P85: The Neuron growth and death model and simulator

Log in to save this to your schedule, view media, leave feedback and see who's attending!

Feedback form is now closed.
P85 Virtual Meeting room is  in following.

Attend the Zoom meeting 
https://zoom.us/j/8437029563

Meeting ID: 843 702 9563
Password: 3psgxH

You can see our  2-mins teaser on https://www.youtube.com/watch?v=Cprz2-DRCAQ&feature=youtu.be

if you have any problmes, please contact the following speaker.
Yuko Ishiwaka (yuko.ishiwaka@g.softbank.co.jp)

---
Yuko Ishiwaka
, Tomohiro Yoshida, Tadateru Itoh

Some kinds of characteristics of neurons depend on morphology. There are many neuron types in a brain and the functions of each cell type are varied. For example, auditory cells that receive sound stimulus from the external world and pyramidal cells that relate to thinking and memory have different functions. A bushy cell which is one of the sensory neurons of auditory treats tempo of sounds, therefore the immediate responses are required for producing action potential and short time refractory period. On the other hand, pyramidal cells that mainly exist in the hippocampus and amygdala treat memory and emotion, therefore the producing action potential is slower than sensor neurons and the refractory period is longer. Hodgkin and Huxley (H-H) equations can calculate action potentials based on ion channels. H-H does not consider morphology, however, on actual cell membranes, the number of exiting ion channels and locations are based on shapes of neurons. How quick or slow soma can produce action potentials are depend on how narrow and how many ion channels on the producing area. Therefore, we assume that there are strong relationships between cell shapes and characteristics of action potentials. Expanded H-H equations can adapt to the quickness of producing action potential by adding axon hillock parameters.

Connectivity between neurons is also important. Geometry varies according to cell types. Purkinje cells which are one of the inhibitory neurons have complex branches of the dendritic arbor. On the other hand, Pyramidal cells which are one of the excitatory neurons and multipolar types neurons have one axon and many dendrites but the complexity of geometry is simpler than Purkinje cells. These differentials of geometry cause differences in connectivity.

In this paper, we propose a new neuron growth and deal model and simulator considered neuron morphology and connectivity between multi cell types. In our model, a characteristic of a growth cone is applied to neuron growth and treated as a navigation system, an L-system is adapted for creating the geometry of each neuron, and Life game is embedded for a cell division rule.

We also adopt glial cells for neuron growth not only stimulus from other neurons. In our model, each neuron receives the energy for growing from contacted astrocytes which are one of the glial cells. The direction of growth of the growth cones has determined by set goal areas for far, and during growing, growth cones try to contract near oligodendrocytes to obtain myelin around their axons. A cell division rule for Oligodendrocytes follows life game rules. The glial cells are treated as obstacles.

In our simulation system, a user can create various types of neurons, set the goals for both dendrites and axons, create connections between various functions and geometries of neurons with growth rules and add injections such as IPSP or EPSP on purpose to calculate action potentials.

In conclusion, we show simulation results of our proposed model. In our simulator, variety of geometry can be produced automatically based on expanded L-system, variety and flexible connectivity can be also produced based on our proposed new neuron growth and death model. Furthermore, we added two types of glial cells for growth and goal rules and also treat as obstacles. Our proposed model and simulator is quite flexible to simulate cell geometry, action potentials, cell connections in each brain region.

Speakers
avatar for Yuko Ishiwaka

Yuko Ishiwaka

Senior Research Scientist, Advanced Technology Promotion OfficeIT & Network UnitInformation Technology Division Technology Unit Softbank Corp.
I have got my Ph.D. by a multiagent system. I was an assistant at Hakodate National College of Technology and moved to Hokkaido University as an associate professor. I am working at SoftBank Corp. as a researcher. My research interest is machine learning inspired by neuroscience to... Read More →



Sunday July 19, 2020 7:00pm - 8:00pm CEST
Slot 07