This unit covers Descriptions of most commonly employed biofeedback modalities: EEG (III-A), Sources of artifact (III-B), EEG patterns and their behavioral correlates (VI-C), and Potential effects of prescribed and nonprescribed drugs (VI-E).
Students completing this unit will be able to discuss:
The International 10-20 system is a standardized procedure for placement of 21 recording and one ground electrode on adults. The 10-20 system received its name because electrode sites are separated by 10% or 20% of the distance between two corresponding anatomical landmarks. Four important landmarks are the nasion (depression at the bridge of the nose), inion (bony prominence on the back of the skull), and left and right preauricular points (slight depression located in front of the ear and above the earlobe). The vertex (Cz) is the intersection of imaginary lines drawn from the nasion to inion and between the two preauricular points.
The 10-20 system assigns recording electrodes a letter and subscript. The letters represent the underlying region and include: Fp (frontopolar or prefrontal), F (frontal), C (central), P (parietal), O (occipital), and A (auricular). A subscript of z represents a midline (central axis from nasion to inion) placement. Numerical subscripts range from 1-8 and increase with distance from the midline. The 10-20 system assigns odd-numbered recording electrodes on the left and even-numbered electrodes on the right side of the head. Two reference electrodes are usually placed on the earlobe (Fisch, 1999).
The International 10-20 system calculates the distance from the nasion to the inion and from the left preauricular notch to the right preauricular notch. The 19 active electrode positions are found taking either 10% or 20% of these distances. Two reference electrodes are usually placed on the earlobes (Andreassi, 2000).
The American Clinical Neurophysiology Society published guidelines for expanding the 10-20 system to 75 electrode sites. The expansion of the 10-20 system allows clinicians to define the sites midway between two 10-20 sites commonly used in clinical practice, better localize epileptiform activity, increase EEG spatial resolution, and improve detection of localized evoked potentials. The modified combinatorial system replaces inconsistent designations (T3/T4 and T5/T6) with consistent ones (T7/T8 and P7/P8). These replacement sites are depicted by black circles with white lettering in the diagram below.
The modified combinatorial system, which is also called the
10-10 system, locates electrodes at
every 10% along medial-lateral contours and adds new contours. Each
electrode site is an intersection between a medial-to-lateral
coronal line (designated by letters) and longitudinal
sagittal line (designated by
As with the 10-20 system, letters represent the underlying region and include: N (nasion), Fp (frontopolar or prefrontal), AF (anterior frontal), F (frontal), FT (frontotemporal), FC (frontocentral), A (auricular), T (temporal), C (central), TP (temporal-posterior temporal), CP (centroparietal), P (parietal), PO (posterior temporo-occipital or parieto-occipital), O (occipital), and I (inion). FT and FC lie along the second intermediate coronal line, TP and CP along the third, and PO along the fourth.
A subscript of z represents a midline (central axis from nasion to inion) placement. Numerical subscripts range from 1-10 and increase with distance from the midline. The modified combinatorial system assigns odd-numbered recording electrodes on the left and even-numbered electrodes on the right side of the head (Fisch, 1999).
Once a site is identified and marked, site preparation is a three-step process as follows:
Silver-silver chloride and gold disc electrodes are recommended because of low electrode noise. Clinicians should use identical electrodes to minimize artifact produced by imbalanced impedance. The electrodes should be secured using collodion glue for recordings longer than 30 minutes. Pastes are acceptable when recordings are shorter than 30 minutes or involve young infants in isolettes.
Neurofeedback therapists monitor the EEG using monopolar or bipolar recording methods.
The monopolar method uses one active and one reference electrode. The active electrode is placed over a site that is an EEG generator like O2. The reference electrode is placed over sites like earlobes that are not sources of EEG and are electrically less active. The reference can actually be a combination of two earlobe electrodes. Both earlobes can be electrically connected to produce a linked-ears reference (Ray, 1990).
The bipolar method uses two active electrodes and one reference electrode to detect differences in electrical potential between the actives.
The active electrodes can be placed over two cortical sites or one cortical site and the second earlobe to preserve phase-synchronous signals (EEG signals whose peaks and valleys coincide). The reference electrode is placed over a site like an earlobe that is not an EEG source.
A derivation is a combination of electrodes used in a single amplifier channel. A montage combines several derivations to detect localized or global EEG activity. Software allows montage reformatting, which permits data analysis where an electrode site can be referenced to other sites or combinations of sites.
A referential montage uses earlobe or mastoid process references. Since these sites usually detect minimal electrical activity, electrical potentials are attributed to the active electrode.
A bipolar montage compares signals from two adjacent electrodes at 10-20 system coordinates inserted into inputs 1 and 2. This montage allows good detection of localized EEG activity, but poor detection of widely distributed EEG activity, since it subtracts synchronous EEG activity. Analysis of asymmetry is only adequate. This montage permits excellent detection of electrode artifact.
A common electrode reference montage combines derivations in which the same reference electrode (distant from the EEG signal source) is inserted into input 2 of each amplifier while a different electrode is placed in input 1 for each derivation. This montage allows poor detection of localized EEG activity and adequate detection of widely distributed EEG activity. Analysis of asymmetry is good when reference electrodes are positioned symmetrically. This montage permits only adequate detection of electrode artifact due to reference contamination by high amplitude (EKG and EMG) artifacts.
An average reference montage combines derivations that add all remaining 10-20 electrodes in each amplifier's input 2 and compares each active electrode in input 1 against this reference signal. This montage allows only adequate detection of localized and widely-distributed EEG activity and electrode artifact, due to reference contamination by artifacts like eye blinks. Analysis of asymmetry is only adequate due to reference contamination.
A Laplacian montage combines derivations that add all remaining10-20 electrodes (weighted by their distance from the single active electrode) in each amplifier's input 2 and compares each electrode in input 1 against this reference signal. This references the active electrode to an average signal from surrounding electrodes. This montage allows good detection of localized EEG activity, but poor detection of widely distributed EEG activity, since it subtracts synchronous EEG activity. Analysis of asymmetry is only adequate. This montage permits only adequate detection of electrode artifact (Fisch, 1999; Thompson & Thompson, 2003).
An analog-to-digital (A/D) converter samples the EEG signal at a fixed sampling interval. An A/D converter's resolution is limited by the smallest signal amplitude it can sample. An A/D converter should be able to resolve amplitudes as small as 0.5 μV.
A bit number is the number of voltage levels that an A/D converter can discern. Lubar and Gunkelman (2003) consider an A/D converter resolution of 14 bits, which can discriminate among 16,384 voltage levels, acceptable. They prefer a 16-bit resolution, which can can discriminate among 65,536 voltage levels. Lower A/D converter resolutions overemphasize small voltage increases.
According the Nyquist theorem, an A/D converter's sampling rate should be at least twice the highest frequency component you intend to sample. For visual inspection of the EEG, a sampling rate of 128 samples per second is acceptable, and rates of 500-1000 samples per second are preferred. Sampling at slower rates results in aliasing, where a slower beat frequency is produced by a signal near the sampling rate (Lubar & Gunkelman, 2003; Teplan, 2002; Thompson & Thompson, 2003).
After the EEG signal is amplified by a differential amplifier and then a single-ended amplifier, it is filtered and then amplified by a second single-ended amplifier. Filters exclude unwanted EEG frequencies to detect activity of clinical interest and minimize artifact and distortion.
Electroencephalographs may use three kinds of filters for analog and digital filtering: low frequency, high frequency, and notch filters. The NeXus-32, shown above, uses carbon-coated cables and active shielding to control 60-Hz artifact instead of a 60-Hz notch filter.
A low frequency filter suppresses slow wave activity. It is also called a high pass filter because it does not change the amplitude of higher frequencies.
A high frequency filter suppresses fast wave activity. It is also called a low pass filter because it does not change the amplitude of lower frequencies.
A notch filter suppresses a narrow band of frequencies produced by line current (60-Hz artifact).
Analog filters contain analog circuits designed using components like capacitors, resistors, and operational amplifiers.
Digital filters use digital processors, like a digital signal processing (DSP) chip, to exclude unwanted frequencies. First, an analog-to-digital (ADC) converter samples and digitizes the analog signal, representing signal voltages as binary numbers. Second, a DSP chip performs calculations on the binary numbers. Third, a digital-to-analog converter (DAC) may transform the sampled, digitally-filtered signal back to analog form.
Two main methods of digital filtering are Fast Fourier Transformation (FFT) and finite impulse response (FIR). FFT and FIR methods enjoy four advantages over analog filters. First, a clinician can retrospectively adjust the filter settings as he or she reviews the EEG record since digital filters are programmable. Second, digital filters can be designed to minimize phase distortion (displacement of the EEG waveform in time). Third, digital filters are stable over time and across a range of temperatures. Fourth, digital filters accurately process low-frequency signals .
Below is a BioGraph ® Infiniti EEG 2D FFT display. Frequency is displayed on the X axis and amplitude on the Y axis.
Below is a BioGraph ® Infiniti EEG 3D FFT display. Frequency is displayed on the X axis, amplitude on the Y axis, and time on the Z axis.
EEG signals may be described by their frequency, amplitude, power, phase, coherence, comodulation, and amplitude asymmetry. Frequency is the number of cycles completed each second. The higher the frequency, the shorter the wavelength.
This is an example of a typical eyes closed "alpha" response showing the onset of normal 10-Hz rhythmic EEG activity in an eyes-closed condition measured from Pz. The top BioTrace+ /NeXus-10 spectral display shows a frequency range from 1-21 Hz (X-axis) with amplitude in microvolts (Y-axis). Time progresses from the back of the display to the front. The center display is a horizontal vernier or bar graph showing the median frequency within the 1-21 Hz range. Note that it tends to center around 10 Hz. The bottom raw EEG oscilloscope display of the same sample data shows frequencies from 2-45 Hz. This movie was generously provided by John S. Anderson.
Delta (0.5-3.5 Hz)
Theta (4-7 Hz)
Alpha (8-13 Hz)
Low beta (13-21 Hz)
High beta (20-32 Hz)
Signal strength is the energy contained within the EEG signal, and is represented by amplitude (μV) or power (picowatts). Picowatts = microvolts2 x 6.65 (Thompson & Thompson, 2003). High amplitude or power means that a large number of neurons are depolarizing and hyperpolarizing at the same time.
Phase refers to the degree to which the peaks and valleys of EEG waveforms coincide. Phase is determined by the speed of signal transmission, which is measured in milliseconds, between networks of cortical neurons (La Vaque, 2003).
The waveforms shown below are 90 degrees out of phase.
Porges and Bohrer (1990) developed coherence analysis, which can measure the correlation between two EEG signals and compare their arrival times. Coherence means that the EEG activity of two recording sites is correlated. The correlation between two EEG signals at their component frequencies is squared to produce values ranging from 0 - 1. Two EEG signals can exhibit high coherence at one frequency and low coherence at others. Coherence analysis can also determine whether one signal arrives earlier than another. This time delay is termed the phase angle (Stern, Ray, & Quigley, 2001). When EEG signals are in phase, strong coherence suggests that processing by these cortical regions is coordinated and that they are functionally connected (La Vaque, 2003; van Beijsterveldt, Molenaar, de Geus, & Boomsma, 1998). When strong coherence occurs between in-phase signals generated in different brain hemispheres, this provides evidence of interhemispheric communication.
Comodulation measures the cross-correlation between the "variation in spectral density estimates" from all recording sites at specific frequency bands and during specified conditions (e.g., eyes closed). This index of temporal synchrony ranges from -1 to +1. The higher the cross-correlation between sites, the greater their functional connection. Comodulation analysis allows clinicians to perform within-subject comparisons as well as comparisons with a normative database (Sterman & Kaiser, 2001).
Amplitude asymmetry refers to the difference in signal strength of a frequency band at two EEG sites. For example, Davidson (1995) trained depressed subjects to reduce alpha asymmetry detected from F3 (left frontal) and F4 (right frontal) sites.
An epoch is an EEG signal sampling period, and an epoch number (like epoch 52) is used to index a location in a session's record. One second is the most common sample when quantifying the EEG. A clinician should obtain at least 60 epochs of artifact-free data for valid assessment (La Vaque, 2003).
An EEG time constant is the period that an EEG signal is averaged before a value is displayed. Lubar (1989) recommends a 0.5-second delay between detection and display of EEG signals to preserve the contingency between EEG events and feedback.
A spectral analysis plots EEG frequency (X axis) against signal amplitude (Y axis).
This BioTrace+ / NeXus-10 clinician screen shows raw EEG, EEG artifact and spectral analysis for signal quality assessment.
To create a compressed spectral array (CSA), the EEG record is divided into epochs that are each analyzed using Fast Fourier Transformation (FFT) analysis. The power spectrum for each epoch is layered above each other (the earliest epoch is at the top). A three-dimensional display is created using hidden line suppression so that each epoch’s spectrum is separated from the others. The CSA can be arranged to correspond to scalp sites where single EEG channels were sampled.
The Quantitative EEG (QEEG) calculates average EEG voltages within frequency bands. This information provides the basis for the topographical mapping of spectral (EEG frequency) averages like BEAM (brain electrical activity mapping, shown below) and compressed spectral array graphs.
Clinicians report ratios between the amplitudes of frequency bands like theta and beta. For example, in Lubar's ADHD protocol, electrode placement is determined by where the theta-beta ratio is highest.
Gunkelman (2001) advocates performing a QEEG to customize clinical interventions. This can minimize sessions wasted on trial-and-error selection of placement sites and frequency tuning, and provide clinical consultation, which may detect undiagnosed "occult" conditions like early dementia, epilepsy, metabolic or toxic encephalopathy, or tumor. Neurotherapy interventions based on QEEG assessment may achieve higher success rates than conventional treatments (p. 3).
In EEG topography, multiple EEG channels are simultaneously sampled, Fast Fourier Transformation (FFT) analysis is performed, and the cortical surface is mapped with color displays of EEG amplitude and frequency. This technique can display evoked, or event-related, potentials in addition to EEG signals.
This BioTrace+ / NeXus-10 screen is used by the clinician to observe the quality of the raw EEG signals. It also features a special Spectrogram display.
Sixty-Hz artifact has a fundamental frequency of 60 Hz and harmonics at 120 Hz, 180 Hz, and 240 Hz. This is the main artifact in EEG and EMG recording. A BioGraph ® Infiniti display of 60-Hz artifact is shown below in red. Note the cyclical fluctuation in voltage.
Sources include power outlets, fluorescent lights, and electrical equipment. You can minimize 60-Hz artifact by equipment location, skin preparation (low and balanced impedances), careful bandpass selection, using a 60-Hz notch filter, carbon-coated cables with active shielding, and selection of a well-designed electroencephalograph with high differential input impedance and common-mode rejection.
Bridging artifact is produced by a short circuit between adjacent electrodes due to excessive application of electrode paste, or a client who is sweating excessively or with a wet scalp. Bridging artifact can cause adjacent electrodes to produce identical EEG recordings. To prevent this artifact, apply electrode paste carefully and instruct clients to arrive with dry hair (Thompson & Thompson, 2003).
Drowsiness artifact may resemble frontal theta or "thalpha," and sometimes may generate spike-like transients. You should suspect drowsiness when these changes are associated with reduced occipital alpha. When you detect drowsiness artifact during a training session, suspend recording and instruct your client to move her hands and legs to increase wakefulness. To avoid this artifact, ask your clients to retire early and sleep for 9 hours if possible (Thompson & Thompson, 2003).
To control drug artifact, ask clients to list the medications they are currently taking and to refrain from ingesting recreational drugs (Fisch, 1999; Thompson & Thompson, 2003).
EKG artifact is signal contamination by the R-wave of the electrocardiogram when monitoring the upper torso or scalp. The frequency range for EKG artifact is .05-80 Hz, and it contaminates the delta, alpha, and beta bands. This is also called ECG artifact. Since this artifact is detected by multiple electrodes, it can create the appearance of greater coherence than is present.
You can detect EKG artifact by inspecting chart recorder, data acquisition, or oscilloscope displays of the raw EEG waveform. EKG artifact appears as a wave that repeats about once per second (Thompson & Thompson, 2003). Below is a BioGraph ® Infiniti EKG artifact display.
You can control EKG artifact using the following steps: (1) use an EEG with a 40-Hz high-pass filter and a differential amplifier; (2) digitally subtract the ECG complex; (3) use a non-cephalic neck-chest reference; and (4) maintain low and balanced skin-electrode impedance (Fisch, 1999).
Electrode pop artifact produces a sudden large deflection in at least one channel when an electrode abruptly detaches from the scalp. You can diagnose this problem by checking impedance and the integrity of electrode cables (Thompson & Thompson, 2003).
In dry environments, friction against carpet or clothing during movement can produce electrostatic artifact. Electrostatic artifact can contaminate recordings and damage the entire data acquisition system.
You can control electrostatic artifact using the following steps:(1) ask clients and staff to avoid nylon clothing and hair spray; (2) avoid nylon carpeting; (3) use a humidifier; (4) spray carpets with an anti-static product; and (5) Install grounded, anti-static pads.
EMG artifact is interference in EEG recording by volume-conducted signals from skeletal muscles. The frequency spectrum for this artifact ranges from 2-1,000 Hz. While strong contraction can contaminate all frequency bands, the beta rhythm is most affected by this artifact. EMG artifact may create the appearance of greater beta activity than is actually present.
Below is a BioGraph ® Infiniti EMG artifact display.
Thompson (2003) observed that EMG artifact is readily detected because it affects one or two channels, particularly at T3 and T4 at the periphery, and less often at O1, O2, Fp1, and Fp2.
You can identify EMG artifact by by visually inspecting the raw signal displayed on a chart recorder, data acquisition system, or oscilloscope.
Lubar (1989) recommended two procedures to control EMG artifact. First, use coincidence detection so that we only record EEG signals (14 Hz) when EMG signals at a predetermined frequency (70 Hz) are absent. Second, only record EEG signals (40 Hz) when identical frequency (40 Hz) EMG signals are absent from the trapezius and temporalis muscles.
EOG artifact is produced by corneal-retinal potentials generated by eye blinks, eye flutter, and eye movements, and contaminates the EEG and EMG records.
Eye blink produces artifact because there can be a 100 mV potential difference between the cornea's positive charge and the retina's negative charge. When the eyelid touches the cornea, it generates a high amplitude slow wave (Fisch, 1999; Thompson & Thompson, 2003).
The BioTrace+ /NeXus-10 display below shows repetitive eye blink artifact. Pay particular attention to the lower oscilloscope or line graph tracing and note the very high-amplitude, intermittent, isolated slow-wave events that result from the rotation of the eyeball. The artifact is the result of differing electrical charges of the vitreous and aqueous humors of the eye, producing an electrical discharge as the eyeball rotates, either through movement or eye blinking. This movie was generously provided by John S. Anderson.
Eye flutter may be confused with frontal intermittent delta, which can signal brain injury or learning disability. You can rule it out through careful observation of your client.
Eye movements also produce deflections in the EEG record as the positively-charged cornea moves toward an electrode in your montage.
An upward eye movement will create a positive deflection at Fp1, while a downward eye movement may create a negative deflection. In a longitudinal sequential montage, artifact is typically seen at frontal sites (Fp1-F3 and Fp2-F4). A left movement may produce a positive deflection at F7 and a negative deflection at F8 (Thompson & Thompson, 2003).
Below is a BioGraph ® Infiniti eye movement artifact display.
You can identify EOG artifact by visually inspecting the raw signal displayed on a chart recorder, data acquisition system, or oscilloscope. EOG artifact appears as spikes in the EEG record.
You can control EOG artifact using the following steps: (1) maintain low and balanced skin-electrode impedance; (2) use two dedicated EOG channels (vertical and horizontal) to identify this artifact; (3) manually exclude trials contaminated by EOG; (4) use appropriate computerized time-domain or frequency-domain correction methods (Gratton, 1998).
Evoked potential artifact can contaminate the EEG record with transients that often appear in multiple channels. Thompson (2003) recommends rejecting an epoch when its amplitude is 50% greater than that of background activity. While evoked potentials increase recording variability and reduce its reliability, they minimally affect averaged data.
Sudden limb and electrode cable movements can alter skin-electrode impedance, and EEG signal voltage.
This artifact can produce high-frequency and high-amplitude voltages identical to EEG and EMG signals. While the delta rhythm is most affected by this artifact, it may also contaminate the theta band (Thompson & Thompson, 2003).
Below is a BioGraph ® Infiniti cable movement artifact display.
You can control movement artifact using the following steps: (1) secure the electrode cable with tape to the client and chair; (2) secure the EEG electrode to the scalp; (3) shorten the electrode cable; and (4) instruct the client to restrict movement.
Radio frequency artifact radiates outward like a cone from the front of televisions and computer monitors. You can control radio frequency artifact using the following steps: (1) move equipment away from high-frequency artifact sources (radiology equipment), placing biofeedback equipment behind or to the side of monitor and never closer than two feet in front (Montgomery, 2004); (2) use a 40-Hz high-pass filter, differential amplifier, differential input impedance over 1 megohm, and fiber optic electrode cables; (3) maintain low and balanced skin-electrode impedance; and (4) install a Faraday cage.
Fried (1987) reported that hyperventilation produces low-frequency, high-amplitude theta activity by constricting cerebral blood vessels. Diaphragmatic breathing increases the incidence of coherent alpha.
Sweating reduces electrode contact with the scalp and generates large-scale up and down EEG line movements in several frontal channels. Also, sympathetic cholinergic artifact involves increased eccrine sweat gland activity, and usually appears as isolated 1-2 Hz slow waves of 1-2 s duration at frontal and temporal sites. This artifact is often elicited by abrupt, unexpected stimuli (Thompson & Thompson, 2003).
Since the base of the tongue is positive and the tip is negative, this creates a dipole that can discharge voltage with movement. Tongue movement contaminates the delta band. The entire EEG record detected from frontal or temporal sites may gradually move up or down. Clinicians may confuse this artifact with frontal intermittent rhythmic delta activity (FIRDA), which can signal brain lesions. Tongue and swallowing artifact is often a problem with dystonia clients (Thompson & Thompson, 2003).
An EEG impedance test uses an impedance meter to run an AC current through active-reference electrode pairs to measure the quality of skin-electrode contact. This test can be performed manually or automatically using a circuit within a data acquisition system.
A conservative guideline is that impedance values for disc electrodes should not exceed 5 Kohms and should be balanced within 1 Kohm to ensure acceptable common-mode rejection and high sensitivity, and to permit valid comparisons between homologous sites located in different hemispheres.
Imbalanced impedance can create the appearance of false hemispheric asymmetry or symmetry. Slew artifact is the "shifting of voltages to one side" when there are imbalanced impedances between linked ears (Lubar & Gunkelman, 2003; Thompson & Thompson, 2003).
Ferree et al. (2001) contend that clinicians can obtain valid EEG recordings without scalp abrasion using modern electrodes and electroencephalographs that feature high input impedance (200 Mohm) and effective digital filters to control 60-Hz artifact. EEG recording without scalp abrasion is increasingly attractive due to increased concern about the transmission of infections like Creutzfeldt-Jacob Disease, hepatitis-C, and HIV. The authors report that increasing skin-electrode impedance from below 10 Kohms (scalp abrasion) to 40 Kohms (no abrasion) does not affect the amplitude within EEG frequency bands and that a 20 Kohm imbalance between impedances only slightly increases 60-Hz artifact when using modern amplifiers.
State-of-the-art data acquisitions systems like the I-330-C2+, NeXus-32 and ProComp Infiniti provide automated measurement of skin-electrode contact.
I-330C2+ with Physiolab Software
NeXus-32 with BioTrace+ Software
ProComp Infiniti with Infiniti Software
An EEG tracking test confirms whether an EEG responds to changes in brain function caused by verbal instructions. Example: when recording from the occipital lobe, ask the client to close his or her eyes (alpha amplitude should increase) and then focus on a nearby object (alpha amplitude should decrease).
Now that you have completed this module, identify the recording montages you most often use and summarize their advantages and disadvantages.
Andreassi, J. L. (2000). Psychophysiology: Human behavior and physiological response. Hillsdale, NJ: Lawrence Erlbaum and Associates, Inc.
Evans, J. R., & Abarbanel, A. (1999). Introduction to quantitative EEG and neurofeedback. San Diego: Academic Press.
Ferree, T. C., Luu, P., Russell, G. S., & Tucker, D. M. (2001). Scalp electrode impedance, infection risk, and EEG data quality. Clinical Neurophysiology, 112, 536-544.
Fisch, B. J. (1999). Fisch and Spehlmann's EEG primer (3rd Ed.). New York: Elsevier.
Gratton, G. (1998). Dealing with artifacts: The EOG contamination of the event-related brain potential. Behavior Research Methods, Instruments, & Computers, 1(30), 44-53.
Gunkelman, J. (2001). QEEG based neurofeedback protocol design. Presentation at the Marriott Hotel, Woodland Hills, California.
La Vaque, T. J. (2003) Neurofeedback, neurotherapy, and quantitative EEG. In D. Moss, A. McGrady, T. Davies, and I. Wickramasekera (Eds.), Handbook of mind-body medicine for primary care. Thousand Oaks, CA: Sage Publications, Inc.
Lubar, J., & Gunkelman, J. (2003). Neurometrics, neurotherapy, and clinical practice. Workshop presented at the 34th annual Association for Applied Psychophysiology and Biofeedback convention, Jacksonville, Florida.
Peavey, B. (2003). Effects of drugs on biofeedback. Short course presented at the 34th annual Association for Applied Psychophysiology and Biofeedback convention, Jacksonville, Florida.
Porges, S. W., & Bohrer, R. E. (1990). The analysis of periodic processes in psychophysiological research. In J. T. Cacioppo & L. G. Tassinary (Eds.), Principles of psychophysiology: Physical, social, and inferential elements (pp. 708-753). New York: Cambridge University Press.
Sterman, M. B., & Kaiser, D. (2001). Comodulation: A new QEEG analysis metric for assessment of structural and functional disorders of the central nervous system. Journal of Neurotherapy, 4(3), 73-83.
Stern, R. M., Ray, W. J., & Quigley, K. S. (2001). Psychophysiological recording (2nd ed.). New York: Oxford University Press.
Teplan, M. (2002). Fundamentals of EEG measurement. Measurement Science Review, 2(2), 1-11.
Thompson, M., & Thompson, L. (2003). The biofeedback book: An introduction to basic concepts in applied psychophysiology. Wheat Ridge, CO: Association for Applied Psychophysiology and Biofeedback.
van Beijsterveldt, C. E. M., Molenaar, P. C. M., de Geus, E. J. C., & Boomsma, D. I. (1998). Genetic and environmental influences on EEG coherence. Behavior Genetics, 28(6), 443-453.