Real-time Visual Neuron Simulator

Detail

Language

English

Title

Real-time Visual Neuron Simulator

Free Keywords

spatiotemporal receptive field, simple cell, inseparable space-time receptive field, separable space-time receptive field, Poisson process, digital video, FireWire, iSight, VNS, ReceptiveField

Description

This software has been developed primarily for testing/debugging/calibration of electrophysiological recording/visual stimulation system without using live animal. It is also useful for class-room demonstrations of visual receptive fields, and student laboratories.

Date

Jan 18, 2008

Last Modified Date

Jan 18, 2008 00:29:40

Created Date

Mar 1, 2004 03:51:32

Contributor

Izumi OHZAWA (ohzawa)

Item Type

Tool

Change Log(History)

Jan 18, 2008	Source package runs with Xcode 3.0 added
Feb 8, 2007	Modified; Index, Related to.
Feb 15, 2006	Modified; Language, Preview, Tool type, Developer, Readme, Rights.

Tool type

Program

Developer

Saito Y

Suginohara H

Ohzawa I

Preview

Tool file

VNS0.1a.zip

Type	: application/x-zip
Size	: 3.4 MB
Last updated	: Jan 18, 2008
Downloads	: 226

Total downloads since Mar 1, 2004 : 263

Readme

REAEME.txt

Visual Neuron Simulator: ReceptiveField version 0.1
----------------------------------------------------
Copyright 2003, 2004 Y. Saito, H. Suginohara, and I. Ohzawa
Ohzawa Lab. Graduate School of Frontier Biosciences, Osaka University
(See License.txt in the same directory which contains this file.)

Download Sites:
---------------
  http://www.bpe.es.osaka-u.ac.jp/ohzawa-lab/resources/
  http://platform.visiome.org/  (Search with "ReceptiveField" or "VNS")

Requirements:
-------------
MacOS X system (10.3 or later, and G4/G5 CPU recommended)
iSight or equivalent DV (FireWire) video camera

Description:
------------
ReceptiveField (version 0.1) is a real-time visual neuron simulator that generates action potentials for a single simulated visual neuron based on video camera input on a MacOS X system (10.3 or later).

The software has been developed primarily for testing and debugging electrophysiological recording and visual stimulation systems without a live animal.  It is also useful for class-room demonstrations and student laboratories.

Version 0.1 allows real-time simulations of spatio-temporally linear neurons with a half-rectifying power-law output nonlinearity, such as X-type retinal ganglion cells, LGN neurons and V1 simple cells (both separable and inseparable XT receptive fields).  For V1 simple cells, space-time characteristics may be set by modeling receptive fields with functions emprically known to approximate experimental data.  It is also possible to load receptive field data from electrophysiological experiments, or synthetic data.  Currently, receptive fields of LGN and X retinal ganglion cells must defined by loading spatio-temporal data computed from concentric center-surround RF models.


How it works:
-------------
Data from the central 20x20 pixels (indicated by a red square in VNSannotated.jpg) in the 320x240 pixel video input from a DV (FireWire) camera, such as iSight, (at 30 frames/sec, reduced from raw 640x480 to 320x240 in the DV API) are saved in a pipeline buffer of approximately 660ms (21 frames) in memory.  The dot product between this video data and the XYT space-time RF data is computed every video frame.  The resulting signal may be thought to represent a total PSP (post-synaptic potential) of a spatiotemporally linear neuron.  This signal is then rectified and processed through a power-law nonlinearity that represents the transfer properties of the action potential generation mechanism.  Finally, this signal controls the mean firing rate of a Poisson generator.  Action potential wave forms are generated from a waveform template and sent to audio output ports via MacOS X CoreAudio API.

The application has been tested with Apple iSight video camera, but other camcoders with DV output may also work.


Useful Tips and Notes:
----------------------
1. You may wish to turn off autolaunching of iChatAV application in its Preferences setting.

2. iSight DV camera from Apple and DV API/driver has a pipeline delay of approximately 200ms.
This latency sets the lower limit for the latency of temporal response of simulated visual neurons.

3. RF_inseparable.txt file contains XYT RF data from a real neuron.  This may be loaded by "Set RF" feature of the simulator.

4. The spatial extent of the RF (regardless of functionally synthesized or loaded from a file) occupies a tiny 20x20 pixel area of the 320x240 video preview area.  This area is approximately indicated by a red square in VNSannotated.jpg.

5. There is a movie for sinusoidal drifting grating for demo as "drift_grating.avi".  Point your DV camera to this movie to demonstrate modulated responses, orientation and direction selectivities of a simple cell.


To Do List:
-----------
Displaying RF border and orientaion/direction in video preview
Function based setting of X retinal ganglion, and LGN neuron.
Complex cells by motion energy model
PSP waveform display
Firing rate meter
Binocularity (split camera field)
MT neuron (based on Simoncelli et al.)

**There is no plan for a Windows version.  Sorry.

References:
-----------
1. Anzai A, Ohzawa I, Freeman RD. Neural mechanisms for processing binocular information I. Simple cells. J Neurophysiol. 1999 Aug;82(2):891-908. PMID: 10444685

2. Ohzawa I, DeAngelis GC, Freeman RD. Encoding of binocular disparity by simple cells in the cat

Rights

License for:
Visual Neuron Simulator: ReceptiveField version 0.1

Index

/ Public / Visiome 2004 / Visual System / Visual Pathway / V1/Area 17/Striate Cortex
/ Public / Visiome 2004 / Visual System / Visual Pathway / V1/Area 17/Striate Cortex / Stimulus Specificity
/ Public / Visiome 2004 / Visual System / Visual Pathway / V1/Area 17/Striate Cortex / Cell Type / Simple Cell
/ Public / Visiome 2004 / Visual System / Visual Pathway / Receptive Field
/ Public / Visiome 2004 / Tools & Techniques / Simulator/Language / Original Program
/ Public / Visiome 2004 / Tools & Techniques / Analysis / Neural Signal Analysis / Reverse Correlation
/ Public / Tool

Related to

Item summary

Spatiotemporal organization of simple-cell receptive fields in the cat's striate cortex. I. General characteristics and postnatal development.
DeAngelis GC , Ohzawa I , Freeman RD
J Neurophysiol 1993 ;69 (4) :1091-117 [PMID: 8492151]

Receptive-field dynamics in the central visual pathways.
DeAngelis GC , Ohzawa I , Freeman RD
Trends Neurosci 1995 ;18 (10) :451-8 [PMID: 8545912]

Functional micro-organization of primary visual cortex: receptive field analysis of nearby neurons.
DeAngelis GC , Ghose GM , Ohzawa I , Freeman RD
J Neurosci 1999 ;19 (10) :4046-64 [PMID: 10234033]

Simple cell responses to drifting gratings: spatial frequency tuning
Ohzawa I . Freeman RD

Encoding of binocular disparity by simple cells in the cat's visual cortex.
Ohzawa I , DeAngelis GC , Freeman RD
J Neurophysiol 1996 ;75 (5) :1779-805 [PMID: 8734580]

Real-time Visual Neuron Simulator 1999 for Windows
Program

Spatiotemporal receptive field organization in the lateral geniculate nucleus of cats and kittens.
Cai D , DeAngelis GC , Freeman RD
J Neurophysiol 1997 ;78 (2) :1045-61 [PMID: 9307134]