Computational Life Sciences
cls
EM-NEURON

 

 

Neuron stimulation in the hippocampus of a rat by a point electrode

Functionalized high-resolution head model for EM-neuron simulations. 

 

First Head Model Functionalized for EM-Neuron Simulations

Background

The Virtual Population 3.0 constitutes the next milestone in computational anatomical phantoms. Imminent challenges are models with functionalized anatomies, i.e., that capture the dynamics of physiological processes. A novel functionalized high-resolution head model and an EM-neuron simulation environment for in silico research was developed by a joint effort of IT’IS and the Office of Science and Engineering Laboratories at the U.S. Food and Drug Administration. The new head model enables investigation of EM-neuron interaction-based applications and exposure safety, and of devices such as neuro-prosthetics, deep-brain stimulators (DBS), and neuro-motoric incapacitation devices.

Past Achievements

  • Anatomical head model: Magnetic resonance imaging (MRI) data from two volunteers was generated, including high (0.5 mm isotropic) resolution T1- and T2-weighted images of the entire head, vasculature imaging, as well as slabs that cover the eye and ear regions with a dedicated nerve contrast sequence. Diffusion tensor imaging (DTI) of the brain was acquired and registered to the other image data, and fiber tracking was performed. Segmentation of one head model based on the image data has been achieved, and the Morel stereotactic atlas of the human thalamus has been registered to the corresponding data set. The current model consists of 142 structures, including many deep brain structures, a large number of differentiated muscles, and the various skull layers.
  • Electromagnetic (EM) solver: In addition to existing EM solvers, a dedicated anisotropic low frequency (LF) finite element method (FEM) solver has been developed, for which the existing FEM solver framework was further modularized, and support for anisotropy has been added. The FEM solver is parallelized to allow high performance computing, e.g., it is able to run on clusters and super-computers. The solver has been validated extensively and it has been integrated into the Sim4Life platform.
  • EM-NEURON coupling: Integration of the widely employed NEURON software into the Sim4Life platform has been achieved, and interfacing based on Python scripting as well as a custom dialog tool has been realized. Coupled simulations of EM-induced neuronal activation can be performed, and existing neuron dynamics models from the NEURON model database can now easily be integrated.
  • Applications:
    Head model – EM: The head model has been used in combination with EM solvers to investigate transcranial alternating current stimulation and compare different electrode montages. Furthermore, the head model has also been used for DBS-related EM modelling.
    EM-neuron interaction: Electrode-based activation of a rat hippocampus neuron described in the literature has been reproduced to validate the EM-NEURON coupling, including determination of the activation threshold. The SENN model, which is the basis for various safety guidelines, has been implemented in the platform developed to allow safety-relevant activation thresholds to be determined for a wide range of exposure scenarios involving complex anatomical models for R&D and for regulatory submissions. As a relevant extension of the SENN model, local thermal effects can now be considered. An in-depth study on gradient coil switching-induced neuron activation has been performed, in which the impact of the neuron model, inhomogeneity, temperature, and other parameters were investigated. Experiments are also being performed on retinal ganglion cell stimulation and on neuroprosthetic stimulation of the rat sciatic nerve.

Next Challenges

  • We are working to integrate complex models of DBS-activated neurons and other specialized subregions into the functionalized head model.
  • Although the solver has already been extensively validated, we are committed to pursue further validation of the solver results with the functionalized head model.

Reference

Iacono MI, Neufeld E, Akinnagbe E, Wolf J, Oikonomidis IV, Sharma D, Wilm BJ, Wyss M, Pruessmann KP, Jakab A, Makris N, Cohen ED, Kuster N, Kainz W, Angelone LM. “MIDA: A Multimodal Imaging-based Detailed model of the Anatomy of the human head and neck”. PLoS ONE 10(4): e0124126. doi:10.1371/journal.pone.0124126.