Block Diagram of CFAM1. Not shown is the calibration source, also controlled by the 6808 8bit microprocessor. Reproduced from "Monitoring Cerebral Function" Prior P F and Maynard D E, pub Elsevier, 1986.
After amplification the EEG signal is passed through an asymmetric bandpass filter. This is similar to that used in the CFM but has a wider bandwidth of 1-27Hz. Over the range 2 to 22Hz it produces increased amplification with frequency at a rate of 12dB/decade. This compensates for the underlying attenuation with increased frequency of EEG signals. Without the asymmetry, EEG frequency analysis would be dominated by by the high amplitude lower frequency content of EEG waves. The appropriate slope of 12dB/decade was established earlier (Maynard 1969) The filter output is converted to a root square measurement on a logarithmic scale. The transforms EEG amplitude fluctuations to an approximately Gaussian distribution. The filter output is also separated into the conventional EEG frequency bands beta, alpha, theta and delta by means of a modified zero crossing and integration technique. Delta band and lower theta band activity are dealt with separately from higher frequencies. An amplitude compression technique is used to retain accuracy at lower signal amplitudes whilst retaining the ability to analyse large signals
The frequency analysis is calibrated against a random noise signal, attenuated at -12dB/decade, to give equal measurements in each frequency band. The analysis is given in terms of the percentage activity in each frequency band and is independent of overall signal amplitude. Frequency analysis can be performed on signals of as little as 1uV peak to peak. Amplitude distribution is computed for signals in the range 0.63 to 220uV peak to peak, with a second range of 1.9 to 660uV peak to peak being available for cortical recording.
Note that the activity in each frequency band is given in uv rms per Hz. Thus it is independent of the width of the frequency band. The percentage measurements are taken from the uV/Hz measurements.
Programme cards with various pre-set system configurations could be inserted.
Two channel Averaged Evoked Potentials could be computed.
Rewritten from "The cerebral function analysing monitor: initial clinical experience, application and further development", D E Maynard and J L Jenkinson, Anaesthesia, 1984, Volume 39, pages 678-690.Published by Blackwell Science.
