One idea to estimate the linear response of the membrane is the following:
Inject one-step pulses of random amplitude, with long intervals between pulses such that there is no current through the electrode (this is however an approximation), like in DCC.
On intervals without current injection, estimate an autoregressive model: x(k) = F.x(k-1) (x is a vector with x0= V), or x(k) = sum a(m) x
For this, one could use Kalman filter + EM. Or maybe something simpler.
One idea to estimate the linear response of the membrane is the following:
For this, one could use Kalman filter + EM. Or maybe something simpler.