Human Neocortical Neurosolver (HNN)
Type: Software,
Keywords: Modeling tool, Neural model, Neural pathology, Neuronal network, Electrophysiology, Academic software, BRAIN Initiative, Electroenchephalography, Magnetoencephalography
Resource ID: RRID:SCR_017437
Human Neocortical Neurosolver software tool: A user friendly software tool to test and develop hypothes on the circuit mechanism underlying EEG/MEG data
Magneto- and electro-encephalography (MEG/EEG) are the two methods to non-invasively record human brain activity with millisecond temporal resolution. MEG and EEG provide reliable markers of healthy brain function and disease states. However, the difficulty of relating these macroscopic signals to the underlying cellular- and circuit-level neural generators is a major, fundamental limitation that constrains using MEG/EEG to reveal novel principles of information processing or to translate the findings into new therapies for neural pathologies. To address this problem, we built the Human Neocortical Neurosolver (HNN, https://hnn.brown.edu). HNN is a user-friendly software tool designed to help researchers, and clinicians interpret the cellular and network origins of MEG/EEG data. HNN’s core is a detailed, mechanistic neural model including canonical features of a layered neocortical circuit, with layer-specific thalamocortical and cortico-cortical drive. HNN’s model is uniquely designed to account for the biophysical origin of the electrical currents generating MEG/EEG with enough detail to connect to the underlying cellular-level activity. HNN provides a user-friendly graphical user interface so that researchers can work interactively between model and data without needing to alter the underlying mathematical model or the open-source code. Tutorials on how to simulate the most commonly measured signals, including event related potentials and brain rhythms (alpha, beta, gamma), are provided. Researchers can compare simulated signals to recorded data and easily manipulate the model parameters to develop and test alternative hypotheses for the neural origin of their signals. Micro-scale features, including layer-specific responses, cell spiking activity, and somatic voltages, can be visualized and used to guide validation of model predictions with a variety of invasive and non-invasive methods. HNN is being developed with best practices in open source software design and is also distributed with a python interface and corresponding python tutorials. The ability of HNN to associate signals across scales makes it a unique tool for translational neuroscience research.
* User friendly software platform
* Open source tool
* Hypothesis development and testing tool
* A software bridge between ‘macroscale’ extracranial recordings and the underlying cellular and circuit-level activity in the brain
* Can be used for translational neuroscience research
* Based on the biophysical origin of the electrical currents generating MEG/EEG, connecting to the underlying cellular-level activity
* Template circuit represents a generalizable, canonical cortical column with a scalable number of neurons and adjustable parameters
* Adaptable with the recent genetic mutations and phenotype data in animal model systems
* Run on high performance computers through the Neuroscience Gateway Portal (www.nsgportal.org) and Amazon Web Services (https://aws.amazon.com)
* Tutorials, community-sharing resources, example datasets and interactive workflow available
* To translate human EEG/MEG recordings into circuit-level activity
* To develop and test hypotheses on the circuit mechanism underlying the EEG/MEG data in an easy-to-use environment
* To identify the neural origins of EEG/MEG signals and to study the brain activity
* To interpret the cellular and network origins and the neural underpinnings of EEG/MEG data and changes in these signals in correlation with behavior or neuropathology
* To study the impact of non-invasive brain stimulation on circuit dynamics measured with EEG
* To study the nerual origin of evoked responses and brain rhythms in health and disease
* Novel prediction on the origin of transient neocortical beta oscillations, supported by laminar recordings in mice and monkeys
* Interpretation of circuit difference is children with Autism compared to healthy developing children
* Simulation of circuit mechanisms of sensory evoked responses from somatosensory and auditory cortex
* Simulation of circuit mechanics of acute pain
* Graphical user interface
* No coding necessary
* Installation on all major platforms Mac, PC, and Linux
* Free to use
* Computational neural model that simulates the electrical activity of the neocortical cells and circuits generating the primary electrical currents underlying EEG/MEG recordings
* Useful for generating novel and testable hypotheses on the circuit origin of some of the most commonly measured signals
* Can be expanded to include other cell types or circuits
* Open-source code
* Developed with best practices in open source software design, including documented API, continuous integration, unit testing, and contributing guidelines
* Available through a python interface with tutorials
* Not designed to simulate the activity across the entire cortex, currently only a single brain area
* Complex parameter optimization
* Currently, compatible only for template neocortical column model
* Basic knowledge of Neuron/Python required for advanced applications needing code expansion
* Currently, designed to study source localized signals as opposed to sensor level activity
* Operating System (OS)-specific Docker installation: Windows, MacOS, CentOS, Ubuntu, Amazon, Oscar, Neuroscience Gateway
* Neymotin et al. 2020, Human Neocortical Neurosolver (HNN), a new software tool for interpreting the cellular and network origin of human MEG/EEG data, eLife 9:e51214
* https://hnn.brown.edu
* https://elifesciences.org/articles/51214
* https://zenodo.org/record/3829558#.X2k_fC2ZNZ0
* https://hnn.brown.edu/index.php/publications/
* Runs in the NEURON simulator with the Python interpreter
* All source-code, model parameters, and associated data are provided in a permanent public-accessible repository on github (https://github.com/jonescompneurolab/hnn)
Dr. Jones won Biomag2020 Mid Career Award for the HNN software development
Stephanie R Jones, Associate Professor
Department of Neuroscience, Brown University
TEAM / COLLABORATOR(S)
Stephanie Jones (Principal Investigator)
Sam Neymotin (Research Scientist)
Dylan Daniels (Research Technician)
Blake Caldwell (Postdoc)
Carmen Kohl (Postdoc)
Matti Hamalainen, co-PI (MGH)
Michael Hines, co-PI (Yale)
Blake Caldwell (Brown)
Dylan Daniels (Brown)
Christopher Moore (Brown)
Sam Neymotin (Nathan Kline Institure)
Mainak Jas (MGH)
Noam Peled (MGH)
Ted Carnevale (Yale)
Robert McDougal (Yale)
Amitava Majumdar (SDSC)
Kenneth Yoshimoto (SDSC)
Subhashini Sivagnanam (SDSC)
Salvador Dura-Bernal (SUNY Downstate)
Matteo Cantarelli (Metacell)
FUNDING SOURCE(S)
NIH NIBIB BRAIN Award 5-R01-EB022889-02
NIH NIBIB BRAIN Award Supplement R01EB022889-02S1