Technical Information About the BrainMaster Unit

The BrainMaster is a clinical quality, FDA-registered, EEG training system intended for clinical work and research in EEG biofeedback.  It is designed for "production" EEG training.  It automatically manages protocols and settings for each trainee, with automatic numbering and storage of session data, statistical analysis, data export, internet-enabled data management and "Session Librarian," and many other integrated system functions.  Since its inception in 1995, BrainMaster has added more than 250 specific user-requested improvements.

The BrainMaster has differential amplifiers.  All standard EEG systems use differential amplifiers.  Each amplifier has an active input and a reference input.  The output is based on the difference between the two electrodes measured.  A single channel uses two electrodes, plus a ground electrode, for 3 electrodes.

 
Ours is a 1, 2, or 4 channel system, so you can have 1, 2, or 4 separate actives, 1, 2, or 4 references, and a ground, for a total of 3, 5, or 9 sensor leads. If you want, you can share the references between the channels, to reduce the number of reference sensors required, and use a "common reference".
 
Differential amplifiers have very high differential gain, and very low common-mode gain. The EEG wave accurately reflects the exact voltage difference between the active and reference leads.  Then, the computer processes the EEG using various signal processing methods.  One metric that is computed in real time is the signal amplitude.  Our system measures the amplitude of the signal in 8 or more frequency component bands, and you can uptrain or downtrain any of the components in any combination.
 
The digital filters are 1st to 10th order Butterworth, with adjustable sidebands.  The response is what you expect mathematically, which is response within 1 to 4 cycles of the wave you are measuring, depending on the filter order. The filters update 30 times per second, to the display, for a 33 millisecond maximum screen latency, and a typical filter response of 20-200 milliseconds, depending on the wave. All NF systems work according to similar principles.  However, the details and quality of the signal processing may vary depending on the knowledge and skill used in the mathematical and software implementation.  For example, BrainMaster system uses quadrature filters with an "implicit envelope detection" method that provides amplitude information up to 200 milliseconds faster than conventional peak-to-peak detection methods.
 
There are extensive additional calculations in the BrainMaster, including real-time "quadrature" methods that extract information quickly and accurately.  These methods are used for all filter-based amplitude, coherence, phase, similarity, and comodulation calculations.
The BrainMaster also includes Fast Fourier Transform (FFT) processing.  This is used for spectral displays, spectral correlation coefficient, peak frequency, percent energy, and other calculations.  The system also includes controls for various system parameters such as filter damping factors, sustained reward criteria, refractory periods, autothresholding parameters, and session control parameters.

BrainMaster units can be adapted for HEG and hand temperature as well, using optional battery-powered adapters.  Either pIR (passive IR / Carmen) or nIR (near IR / Toomim) types can be used.
 
The BrainMaster system has many built-in protocols.  These include standard alpha, beta, smr, squash, and other protocols.  They also include protocols for peak frequency, percent energy, variability, coherence, phase, comodulation, spectral correlation, and other types of training.  Most protocols include support for sounds, multimedia, midi music, mp3 files, DVD's, CD's, and Flash games for feedback. 

Developers can develop their own screens and games in C++, Visual Basic, Delphi, Forth, or Flash.  The programming tools are provided without cost to developers who desire to develop BrainMaster add-on software.

The standard built-in protocols include 1, 2, and 4 channel training.  They include all the standard types of training found in the literature.  You can move any of the filter bands anywhere you like, so customization is easy.

The recent development of the Live Z Score DLL training capability brings QEEG-based EEG analysis to the real-time training environment.  The new Z Score training is the subject of several research projects, and is documented in plenary talks, workshops, grant proposals, and upcoming publications.

Attached Files

    Comments

      Related Entries