In Arry-380 price order to remove the artifact, we assumed that the artifact would not significantly change between non-expressing tissue and expressing tissue. The distance between the ferrule and the electrodes was fixed during construction (Figures 1J,K), and assuming the light-scattering properties of cortical and hippocampal tissue are similar, photo-induced artifacts would largely be the same within the two regions. Furthermore, electrical coupling between the ribbon cable and the LED stimulation input signal would not be expected to differ between the cortex

and hippocampus. Thus, to remove the artifact signal offline, we subtracted the mean artifact recorded in the cortex – where there was no ChR2 expression – from the LFP recording in the hippocampus (Figure ​Figure8B8B). As the neurophysiologic response was much larger amplitude than the artifact, little appreciable change in spectrographic power was noted (Figure ​Figure8B8B, bottom). While the artifacts in the LFP were readily identifiable from the underlying neurophysiologic signal, the single-unit responses proved difficult to resolve. While common median referencing was employed to attempt to improve the signal to noise ratio of the action potentials (Rolston et al., 2009a),

it remained difficult to distinguish true single-units from artifacts. This is demonstrated in (Figures 8C–F), wherein a unit believed to be real, and a unit believed to be an artifactual response, are presented. The first detected unit (Figures 8C,D) had a basal firing rate preceding the stimulus

that increased during the stimulation epoch in successive trials. The second detected unit (Figures 8E,F) also increased its firing rate during the stimulus, and appeared to be largely locked to stimulus onset. However, the latter unit failed to be detected outside of the stimulation epoch, and despite the favorable appearance of its waveform, appeared to have been consequent Drug_discovery to high-pass filtering of the stimulation artifact on this electrode. Without an accompanying intracellular waveform, or a tetrode-based identification scheme, it remains very difficult to clearly define a unit in this fashion. This is particularly a problem if the unit only appears during stimulation, and is locked to the stimulation frequency. CLOSED-LOOP STIMULATION We used NeuroRighter for closed-loop stimulation of MS in which the hippocampal theta-rhythm was used as a control signal to trigger the stimulation of the MS. The control system was implemented using a dynamic link library (DLL) based on the NeuroRighter application programming interface (API; Newman et al., 2013). The API contains a set of tools for interacting with NeuroRighter’s input and output streams.