The name is absent

133

StimuIusSignaILeakage

When the switching board is set to record, the stimulus waveform of that
channel shows up on the coinciding channel of the Cerebus system as if it were being
recorded from the electrode. The amount of signal that "leaks" through varies from
channel to channel, with channels 17,19, and 24 being particularly problematic. Table 1
below shows the level that a IkHz 5V_p∣< sinusoidal input "leaks" onto each recording
channel.

Although this "leakage" shows up on other channels with varying amounts of
attenuation (except on channels 17,19, and 24, where a leaked signal can show up
more strongly than on the channels the stimulus is actually applied), the pattern seems
to be based on the channel layout on the stimulus PCB rather than the microelectrode
layout. Also the signal does not show up on neighboring channels that aren't connected
to the Stim Project, and non-sinusoidal waveforms such as square waves are heavily
distorted. This indicates that the problem is not caused by current leaking across the
switches and into the electrode, but is instead a byproduct of the cross-talk issue
discussed in the previous section. However, the Stim Project has an output status signal
(labeled STAT on the mainboard) that signals whether the stimulus board is in stimulus
or recording mode; and since virtually all cortical stimulators have a trigger function this
signal can easily be used to gate the input waveform and eliminate the problem.

<<   141   142   143   144   145   146   147   >>

More intriguing information