Parametric Anatomical Modeling: A method for modeling the anatomical layout of neurons and their projections
-
1
Ruhr-University Bochum, Department of Psychiatry, Germany
Biological neural networks are likely to be described by a low-dimensional parameter space. Those parameters include, for example, the 3d-shape of neuron layers, the neurons' spatial projection patterns, spiking dynamics and plasticity rules. Many studies generate artificial neural networks that match some of the properties of biological networks to study their computational properties. While spiking dynamics and plasticity rules can comparatively easily be expressed in a programming language and have been subject of extensive research, defining and investigating anatomically more realistic connection and latency properties for large-scale artificial neural networks remains a challenging task. In particular, a method is missing that allows to translate anatomical data from e.g. histological images and tracer studies into an encoding that can generate more realistic network architectures.
We present a new method, called Parametric Anatomical Modeling (PAM), to fill this gap. The basic idea behind PAM is that neural networks are computed based on neuronal, synaptic and intermediate layers, which are generated from anatomical data. Using a set of mapping techniques, complex connection patterns between those layers can be easily defined (fig. 1). For example, any location on one layer can be mapped on another layer based on euclidean distance, a normal vector or topological similarity. Connections between neurons are computed by mapping pre- and post-synaptic locations on a synaptic layer and applying connectivity kernels on the surface of the synaptic layer. A central feature of PAM is that distances between neurons (which may affect transmission delays) can be computed by combining spatial distances on the surface of layers and spatial distances between layers. Thereby, complex connection and distance patterns between layers can be expressed and be used as template for generating small-scale and large-scale network architectures. PAM is implemented as a Python tool and integrated in the 3d-modeling software Blender. We provide a set of add-ons and Python modules that amend the functionality of Blender to generate and relate anatomical layers to each other and to create neural networks for the network simulator NEST. These tools along with example files and video tutorials will be freely available.
On a 3d-model of the hippocampus and entorhinal cortex, we demonstrate the benefits of PAM and show how PAM can help to uncover the relationship between the form and function of the hippocampus. Models created by PAM can also serve as an educational tool to visualize the 3-d connectivity of brain regions. The low-dimensional, but yet biologically plausible, parameter space renders PAM particularly suitable for analysing allometric and evolutionary factors in networks and for modeling the complexity of real networks with comparatively little effort.
Keywords:
Hippocampus,
3D model,
consolidation,
evolution,
Blender,
anatomical modeling
Conference:
Neuroinformatics 2014, Leiden, Netherlands, 25 Aug - 27 Aug, 2014.
Presentation Type:
Demo, to be considered for oral presentation
Topic:
Large-scale modeling
Citation:
Pyka
M and
Cheng
S
(2014). Parametric Anatomical Modeling: A method for modeling the anatomical layout of neurons and their projections.
Front. Neuroinform.
Conference Abstract:
Neuroinformatics 2014.
doi: 10.3389/conf.fninf.2014.18.00020
Copyright:
The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers.
They are made available through the Frontiers publishing platform as a service to conference organizers and presenters.
The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated.
Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed.
For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions.
Received:
04 Apr 2014;
Published Online:
04 Jun 2014.
*
Correspondence:
Dr. Martin Pyka, Ruhr-University Bochum, Department of Psychiatry, Bochum, Germany, martin.pyka@gmx.de