fagan2888 / grfnnmotor

An oscillatory neural network model of motor dynamics during continuous periodic movement

authors:

Wisam Reid
Center for Computer Research in Music and Acoustics
Stanford University
[email protected]

Iran Roman
Stanford Neuroscience Graduate Program
Center for Computer Research in Music and Acoustics
Stanford University
[email protected]

abstract:

Brain oscillations are of functional relevance to the healthy functioning of the brain. In the motor system, desynchronization of oscillatory activity in the beta band ( 20Hz) is observed through periodic somatosensory and/or auditory stimulation. This desynchronization occurs at a rate equal to the period of stimulation, and resynchronization of oscillatory activity anticipates the next stimulus. The computations underlying motor function can be explained by networks of neural oscillators carrying out non-linear transformations of stimuli, reflected by desynchronization of oscillatory activity in the motor system. We developed a neural model of nonlinear oscillators that transforms stimuli into characteristic oscillatory activity of the beta band. Our model is built using a canonical model of nonlinear oscillators, captures even-related desynchronization in the beta band, and is able to anticipate the next period of stimulation through resynchronization. Additionally, our model provides a dynamical-systems perspective of computations and stimuli transformations in the motor system.

---

The Code

Fixed Point Analysis and Amplitude Vector Plane Plots

[Folder] analysis/

  scripts in this folder utilize functions in lib/

autonomousOscAnalysis.m

  Autonomous oscillator parameter analysis:      
  Plots the amplitude vector feild for ranging oscillator parameters
  
  Current Total Runtime: ~70 seconds for 9 images in the 
  12-16-2016 research report

Models and Demos

[Folder] models/

  Models and prototypes 

[Folder] demos/

  1. Demo for McClelland Lab Meeting
  2. Bursting Oscillator 3D animation 

[Folder] images/

  Matlab Images 

Functions

[Folder] lib/

  type "help name_of_file" in the Matlab terminal for more information 
        1. getFP [Passes 66 Unit Tests]
              * tests/Fixed_Points            
        2. getSS [Passes the same 66 Unit Tests] 
              * tests/Fixed_Points
        3. getOscParamRegime [Passes 10 Unit Tests] 
              * tests/Param_Regime
        4. plotAmplitudeVectorFeild
        5. plotFP

Unit Tests

[Folder] tests/

        Tests were made from figures in Kim & Large 2015
        To run tests use  
              1. runFixedPointUnitTests
              2. runOscParamRegimeUnitTests
              
         * run from inside the tests/ folder          

---