Header Image
Schoenbaum Lab
Abstracts & Presentations
Members
Movies
Photo Gallery
Publications
Research
University of Maryland
School of Medicine

Department of Anatomy & Neurobiology
20 Penn St
HSF-2, Room S251
Baltimore, MD  21201

Office: HSF-1, Room 280K
Phone: 410-706-3814
Cell: 443-722-6746
Fax: 410-706-2512
University of Maryland School of Medicine
Noise in Electrophysiology
Introduction

Electrical noise is present in all electrophysiological recordings. Unique technical challenges are presented by the freely-behaving preparation used in behavioral neurophysiology. The electrodes usually consist of fine wires which are quite different from the conventional saline-filled pipette used in-vitro. In addition, because the preparation is chronically implanted and behaving, the recording system must be designed to minimize artifacts associated with movement. Knowledge of basic electrical principles is essential for an understanding of electrical noise. Thus, our discussion of begins with some electrical concepts.
 
 

Electrical Concepts

In an electrical circuit, current, voltage and resistance are all related as expressed by Ohm's Law,

V = IR

where V represents voltage (in Volts), I represents current (in Amperes) and R represents resistance (in Ohms) with polarity and direction as shown (Figure 1). Current flows from postive (+) to negative (-).


Figure 1

ohmslaw


In an extracellular electrophysiological recording, changes in ionic concentration cause small currents to flow through the electrode. The electrode is connected to a preamplifier with a very high input impedance. Impedance (Z) is measured in Ohms, like resistance. However, impedance is a complex quantity that describes how a circuit element responds to both steady-state and time-varying signals. Ohm's law also applies to impedances as shown in the equation:

V(t) = I(t) Z

and also as depicted in Figure 2.



Figure 2

impedance



Connecting the electrode to a preamplifier with a high input impedance allows small currents induced at the electrode tip to be seen as measurable voltages (50-300 mV for spontaneous brain activity). A simplified equivalent circuit of an electrode / preamplifier recording system is shown (Figure 3). When recording from freely-behaving animals, the preamplifier is often implemented on a headstage and generally has unity gain (Av = 1). For more about the headstages see movement artifact, below.

Figure 3

preamp


Types of Noise

High input impedance preamplifiers and high impedance electrodes are necessary for electrophysiological recording. The small tip diameter required for proper unit isolation results in electrodes with high impedances. Unfortunately, high impedances also allow small currents that are unrelated to brain function to compete with the signals of interest. Signals that interfere with measurement of neural activity are called noise. Generally, noise signals can be characterized as either intrinsic or external. Intrinsic noise is generated in the resistive components of your system by the random movement of charge carriers and is known as Johnson noise.  In the figure above, the resistive portion of each impedance contributes to Johnson noise. In general, higher resistive impedances generate more noise. In the laboratory, Johnson noise is controlled by appropriate electronic (bandpass) filtering and minimization of electrode impedance (Metal Microelectrodes for Electrophysiology in Behaving Animals).

External noise is introduced through connections between elements of your recording system. The important elements and connections between them are shown (Figure 4). All of these connections are prone to noise induced by electromagnetic radiation from various sources called inductive pickup. Common sources for inductive pickup are computer monitors and light fixtures. In general, inductive pickup can be avoided by using shielded cables to make connection between different components. It is also helpful to keep the cable lengths as short as is practical. In particular, the length of the electrode connection should be kept to a minimum because of the high impedances involved (more about this later when we discuss movement artifact).


Figure 4

cables



In addition, when electrical components make voltage measurements from different reference points a condition called a ground loop may result in a noisy recording.  To avoid ground loops, all elements of the recording system should use a single point as a reference (Figure 5A) or the preparation should be electrically isolated (Figure 5B).  An additional consideration for the freely-behaving preparation is that any conductive surface accessible inside the recording chamber can inadvertantly provide an additional path to ground. Thus, it is adviseable to paint the interior of the recording chamber to make it non-conductive.

Figure 5A

grounding

Figure 5B

isolation



Movement artifact is a type of noise specific to the freely-behaving preparation.  Flexing of the cable connecting the animal to the recording equipment results in changes in capacitance.  This change in capacitance causes the flow of charge (current) through the cable.  Under high impedance recording conditions, particularly in the electrode - preamplifier connection, this movement-induced current can be many times greater than the electrophysiological signal.  To minimize movement artifact, the electrode - preamplifier connection should be made as short and as rigid as possible by mounting the preamplifier directly to the skull of the preparation (called a headstage). Thus, the preamplifier presents a low impedance to the flexing cable which minimizes the movement-related artifact.  The headstage effectively converts the high impedance signal from the electrode into a low impedance signal at the level of the cable (Headstage Amplifiers for Electrophysiology in Behaving Animals).


Web Content from:  Schoenbaum, G.  Olfactory Learning and the Neurophysiological Study of Rat Prefrontal Function.  In: CRC Series: Methods and Frontiers in Neuroscience.  Edited by S.A. Simon and M.A.L. Nicolelis, CRC Press, NY, 2000.



This web page coauthored by Kevin B. Austin, Ph.D., Eclectic Engineering Studio, www.EclecticStudio.com

Please send comments and suggestions regarding this site to Dr. Geoffrey Schoenbaum, Director of the Lab.