PMID- 20975212
OWN - NLM
STAT- MEDLINE
DCOM- 20110425
LR  - 20110105
IS  - 1741-2552 (Electronic)
IS  - 1741-2552 (Linking)
VI  - 7
IP  - 6
DP  - 2010 Dec
TI  - Firing rate control of a neuron using a linear proportional-integral controller.
PG  - 066004
LID - 10.1088/1741-2560/7/6/066004 [doi]
AB  - Some electrophysiology experiments require periodically firing neurons. One 
      example is when measuring a neuron's phase response curve (PRC) where a neuron is 
      stimulated with a synaptic input and the perturbation in the neuron's period is 
      measured as a function of when the stimulus is applied. However, even regular 
      spiking cells have considerable variations in their period. These variations can 
      be categorized into two types: jitter, which characterizes the rapid changes in 
      interspike intervals (ISIs) from spike to spike, and drift, which is a slow 
      change in firing rate over seconds. The jitter is removed by averaging the phase 
      advance of a synaptic input applied at a particular phase several times. The 
      drift over long time scales results in a systematic change in the period over the 
      duration of the experiment which cannot be removed by averaging. To compensate 
      for the drift of the neuron over minutes, we designed a linear 
      proportional-integral (PI) controller to slowly adjust the applied current to a 
      neuron to maintain the average firing rate at a desired ISI. The parameters of 
      the controller were calculated based on a first-order discrete model to describe 
      the relationship between ISI and current. The algorithm is demonstrated on 
      pyramidal cells in the hippocampal formation showing ISIs from the neuron in an 
      open loop (constant applied current) and a closed loop (current adjusted by a 
      spike rate controller). The advantages of using the controller can be summarized 
      as: (1) there is a reduction in the transient time to reach a desired ISI, (2) 
      the drift in the ISI is removed allowing for long experiments at a desired 
      spiking rate and (3) the variance is diminished by removing the slow drift. 
      Furthermore, we implemented an auto-tuning algorithm that estimates in real time 
      the coefficients for each clamped neuron. We also show how the controller can 
      improve the PRC estimation. The program runs on Real-Time eXperiment Interface 
      (RTXI), which is Linux-based software for real-time data acquisition and control 
      applications.
FAU - Miranda-Dominguez, O
AU  - Miranda-Dominguez O
AD  - Department of Biomedical Engineering, University of Minnesota, Minneapolis, MN 
      55455, USA.
FAU - Gonia, J
AU  - Gonia J
FAU - Netoff, T I
AU  - Netoff TI
LA  - eng
PT  - Journal Article
PT  - Research Support, Non-U.S. Gov't
DEP - 20101026
PL  - England
TA  - J Neural Eng
JT  - Journal of neural engineering
JID - 101217933
SB  - IM
MH  - Algorithms
MH  - Animals
MH  - Electric Stimulation
MH  - Electronics
MH  - *Electrophysiological Phenomena
MH  - Electrophysiology/*instrumentation
MH  - Female
MH  - Hippocampus/cytology/physiology
MH  - Male
MH  - Models, Neurological
MH  - Neurons/*physiology
MH  - Patch-Clamp Techniques
MH  - Pyramidal Cells/physiology
MH  - Rats
MH  - Rats, Long-Evans
MH  - Synapses/physiology
EDAT- 2010/10/27 06:00
MHDA- 2011/04/26 06:00
CRDT- 2010/10/27 06:00
PHST- 2010/10/27 06:00 [entrez]
PHST- 2010/10/27 06:00 [pubmed]
PHST- 2011/04/26 06:00 [medline]
AID - S1741-2560(10)66304-5 [pii]
AID - 10.1088/1741-2560/7/6/066004 [doi]
PST - ppublish
SO  - J Neural Eng. 2010 Dec;7(6):066004. doi: 10.1088/1741-2560/7/6/066004. Epub 2010 
      Oct 26.