What TAP is about ...
Why we thought TAP might help you/us ...
How TAP works ...
Where to get TAP ...
A To Do List and Some Known Problems
Credit Where it is due (References)
Recent Changes and Corrections
Theoretical Analysis Platform (TAP) is aGraphical User Interface written in Matlab, which provides an interface for the modelling of brain-behavior dynamics.
The interface defines two levels, a behavioral level and a brain level. The behavioral level provides two excitator models which capture the intrinsic dynamics of human limbs. These may be coupled so that their behavior is not independent, contributing to extrinsic dynamics components. External stimuli, visual, auditory or tactile, may be read in and provide external information for the excitator system. These elements together define a platform, which allows the modelling of human bimanual or unimanual dynamics, discrete or rhythmic, under the impact of environmental information.
Complementary the brain level, incorparates the activation of the corresponding cortical areas (motor, visual, auditory, tactile) during movement execution. The forward solution of the problem is calculated for the electroencephalogram (EEG)and the magnetoencephalogram (MEG) for every time point on the human scalp.
In summary, the user provides an external environment (by providing stimulus time series) and defines task conditions (by setting the parameters of the excitators and couplings). Given this frame work, TAP will provide you with the resulting movement dynamics and corresponding brain dyanmics (EEG, MEG). How wrong, how right is it? Well, run your experiment and compare it to these predictions. Or just play around and find interesting dynamics in brain and/or behavior.
Right now TAP is in a very preliminary (primitive?) stage of development.But still we thought it would be better to bring out a basic yet functional version of the program, since in any endeavour of this kind the feedback from the user community makes a pronounced difference.
Why we thought TAP might help you/us ...
TAP brings under one rather simple to use interface the calculation and plotting of the Behaviour and Transfer functions and also the EEG and MEG forward solutions. So if you are some on who is interested in our model and would like to test against it your stimulus sequences and calculate and plot the tranfer functions and forward solutions for it and would like to do all this with the minimum possible hassle then TAPcould be a useful tool for you.
Also we all like a nice toy, dont we!!!
The current implementation of the interface consists of two GUI 's. A main window which consists of a set of ui (user interface) controls and axes and a second window which can be called from the main window(by clicking on the 'Set Ode Parameters' button.).
All the plotting functions are in the main interface.You can also set the variables and parameters for the spherical head forward solutions for EEG and MEG from the main window of the interface itself.
The second interface is for setting the different parameters for the integration of the coupled Excitators. You can also read in user defined stimulus vectors to pass to the integrator from this interface.After you download the zip file,unzip the files to a directory (preferably a seperate directory named tap).Add this directory to you matlab path.Now you can run TAP by typing
>>Interface0p2
at the the command window.This will bring up the main window of TAP.
The main interface window consists of a set of ui controls on the panel to your right and four axes on the left. The top two axes are for plotting the Behaviour and Transfer functions. Before you can plot you need to perform the integration of the Excitator.The integration is done using the second interface.To bring up the second interface click on the 'Set OdeParameters' button.This will bring the second window up.For instructions on how to use the second interface scroll down . After you have done the Intergration you are now ready to plot the Behaviour and Transfer fucntions.You can do this plotting by clicking on the 'Plot Transfer Fcns' and 'Plot Behavior' buttons for the corresponding plot.
Next to these push buttons there is a toggle button labeled 'Seperate Axes'. All the plots in the TAP can be transferred on to new figures by clicking on the corresponding axes.There are two modes of viewing these plots on new figures.For more on modes of axes seperation scroll down.
Below these set of buttons you can see a text area where you can specify a particular time point at which you want to calculate the the EEG and MEG forward solutions.
To calculate the EEG and MEG forward solutions you need to enter the radius of the sperical head and the number of points on the suface of the sphere.You also need to specify the dipole moments and the dipole source locations.Both of these should be n x 3 matrices where n is the number of dipoles.These variables should be loaded on to the Matlab workspace before you enter their names in the text areas provided for them.After you have done so if you click on the 'Plot EEG' or 'Plot MEG' button it will calculate the corresponding EEG and MEG forward solutions. If you want to see any of the axes in more detail click on the axes you want to enlarge/seperate.This can be done in any of the two modes of axes seperation .Setting Parmeters and variables in the 'OdeParameters' window
Clicking the 'Set OdeParameters' button in the main window brings up the 'OdeParameters' window where you can set parameters for the integration and also read in your stimulus vectors which can be passed to the integrator.
The interface has been divided in to two logical halves.The upper half allows you set the parameters for the ode solver like the maximum integration step, noise fucntion, noise strength, the time span to integrate over and the initial conditions.The lower half allows you to set the parameters for the actual excitator equations. It also allows you to read in user defined stimulus vectors of arbitrary lengths(both vectors should be of same length).
To do this you have to check the 'Read User Defined Vectors' checkbox and then enter the names of two vectors that have already been loaded in the Matlab workspace.If the variables are found in the workspace TAP will read it and pass it to the integrator.
At the bottom of the 'OdeParameters' window there are two buttons labeled 'Reset to Default' and 'Submit'.If you click on the 'Reset to Default' button all the parameters are set to the original default values.Note that this will also clear the 'Read User Defined Vectors' checkbox .The 'Submit' button will submit the values you have entered and do the integration of the coupled Excitators.There are two modes of axes seperation in TAP.
The first mode is when toggle button is not selected.When in the first mode(Toggle button not selected), TAP will plot whatever is present in the axes at the instant you have clicked in a new figure window and just that.The first mode will work even if you dont have any thing in the axes.
The second mode is when the toggle button is selected(The color of text on the button changes to white). In the second mode depending on which axis you click TAP will plot differently on to a new figure window.
If you click on the Behavior level axes the new figure that pops up will also have the Phase Space Diagram plotted as a subplot on the new window. If you click on either the MEG or EEG axes while the toggle button is selected it will plot the forward solutions on top of a canonical head.(The data for plotting this is included as mat file along with the distribution of TAP.) The second mode will only work if you have already plotted the in main widow axes.Where to get TAP...
A slightly modified and hopefully more userfriendly version of TAP is available for download now.We are calling it TAP 0.2 .The earlier version of TAPused to run only on the new MATLAB release (Release13 ).But the new version of TAP will also run on Release12.
You can download the zip version for windows here( TAP 0.2.zip 581 KB)
After you download the zip file,unzip the files to a directory (preferably a seperate directory named tap).Add this directory to you matlab path.Now you can run TAP by typing
>>Interface0p2
at the the Matlab command window.This will bring up the main window of TAP. To know more about how the interface works read How TAP works section.A To Do List and Some Known Problems
The to do list is indeed a long one .And it is precisely here that we would like to have your feedback and suggestions the most.
Some of the problems mentioned earlier, like the correction of EEG forward solution has been done Check the recent changes section. We are still working on various efficiency issues like faster versions of both EEG and MEG forward solutions.Please email your suggestions to tap@ccs.fau.edu
Recent Changes and Corrections
We have new version of TAP out. Why?? Well for starters, the old version of TAP will only work with the new version of MATLAB(Release13).But the new version of TAP will also work with the older version of MATLAB(Release12).
The problems with EEG forward solutions have been corrected.You now have to provide the dipole moments instead of the orientations of the dipoles.
Also you can specify a particular time point for which you want to calculate the EEG and MEG forward solutions.This feature was not enabled in the earlier version of the program.Last Updated on 11/21/2002
Credit where it is due (References)
Scientific References:
Forward solution:
V.K. Jirsa, K.J. Jantzen, A. Fuchs, J.A.S. Kelso: Spatiotemporal forward solution of the EEG and MEG using network modelling, IEEE Transactions on Medical Imaging, 21, 5, 493-504 (2002)
V.K. Jirsa, K.J. Jantzen, A. Fuchs, J.A.S. Kelso: Neural field dynamics on the folded three-dimensional cortical sheet and its forward EEG and MEG, Information Processing in Medical Imaging, Springer Lecture Notes
in Computer Science, 286-299 (2001)Brain dynamics:
V.K. Jirsa, J.A.S. Kelso: Spatiotemporal pattern formation in continuous systems with heterogeneous connection topologies, Phys. Rev. E 62, 6, 8462-8465 (2000)Movement dynamics:
V.K. Jirsa, J.A.S. Kelso: The Excitator as a Minimal Model for discrete and rhythmifc movement generation, Soc. f. Neurosc. abstarcts 855.7 (2002)
P. Fink, P. Foo, V.K. Jirsa, J.A.S. Kelso: Local and Global Stabilization of Coordination by Sensory Information, Exp. Brain Res. 134, 9-20 (2000)
V.K. Jirsa, P. Fink, P. Foo, J.A.S. Kelso: Parametric Stabilization of Biological Coordination: A Theoretical Model, Journ. Biol. Phys. 26, 85-112 (2000)
Back to top
Last Updated on 11/21/2002
