LRIHHGEDFN &DVH 6WXGLHV 8VLQJ /LYH = 6FRUH 7UDLQLQJ …brainm.com/software/pubs/brain/BiofeedbackCaseStudy2.pdf · EEG Biofeedback Case Studies Using Live Z-Score Training and a Normative

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

This article may be used for research, teaching and private study purposes. Any substantial orsystematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply ordistribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae and drug dosesshould be independently verified with primary sources. The publisher shall not be liable for any loss,actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directlyor indirectly in connection with or arising out of the use of this material.

http://www.informaworld.com/smpp/title~content=t792306937

http://dx.doi.org/10.1080/10874200903543963

http://www.informaworld.com/terms-and-conditions-of-access.pdf

EEG Biofeedback Case Studies Using Live Z-ScoreTraining and a Normative Database

Thomas F. Collura, PhDJoseph Guan, MM.ED, PhD

Jeffrey Tarrant, PhDJohn Bailey, PhDFred Starr, MD

ABSTRACT. This article summarizes clinical results using a neurofeedback approach that hasbeen developed over the last several years and is seeing increasing clinical use. All participantsused a form of live Z-score training (LZT) that produces sound and video feedback, based on acomputation using a normative database to produce multiple targets. The client receives simplefeedback that reflects a complex set of relationships between amplitude and connectivitymetrics. Changes in the EEG are readily seen that conform to the reinforcement parametersbeing used in relation to the live Z-scores. In addition, over multiple sessions, QEEG dataare seen to change significantly, generally on a path toward overall remediation. In thisseries of case studies LZT is seen to effectively address EEG abnormalities in a structuredfashion and to facilitate normalization of the EEG. In individual cases, specific changes areobserved, related to the initial conditions, and the brain’s ability to respond with appropriatechanges. Overall, LZT is found to be a relatively efficient form of neurofeedback that can bedemonstrated to be effective in a variety of clinical scenarios.

KEYWORDS. Biofeedback, live Z-score training, multivariate proportional training, neuro-feedback, normative database, QEEG

INTRODUCTION

This article discusses the technical back-ground, and initial clinical results obtained,in an implementation of live Z-score based

training (LZT) in an EEG biofeedbacksystem. This approach makes it possible tocompute, view, and process normative Z-scores in real time as a fundamental elementof EEG biofeedback. Although employing

Thomas F. Collura is affiliated with BrainMaster Technologies, Inc., Bedford, OH.Joseph Guan is affiliated with the Brain Enhancement Center, Singapore.Jeffrey Tarrant is affiliated with the Columbia Neurofeedback Center, Columbia, MO.John Bailey is affiliated with Allina Medical Clinic, Northfield, MN.Fred Starr is in practice in Nashville, TN.Address correspondence to: Thomas F. Collura, PhD, BrainMaster Technologies, Inc., 195 Willis Street,

Suite #3, Bedford, OH 44146 (E-mail: [email protected]).Dr. Collura has a financial interest in BrainMaster Technologies, Inc. Certain of the methods described here

are patent pending in the United States, Canada, and Europe. We acknowledge the following for their contribu-tions to the case studies in this article: Doerte Klein, PhD; Penijean Rutter, MA; Nancy Wigton, MA; HarryKerasidis, MD; Charles R. Stark, MD; and Jonathan Walker, MD.

Journal of Neurotherapy, 14:22–46, 2010Copyright # Taylor & Francis Group, LLCISSN: 1087-4208 print=1530-017X onlineDOI: 10.1080/10874200903543963

22

Downloaded By: [WNEU Journal of Neurotherapy] At: 20:30 28 July 2010

the same type of database as conventionalQEEGpostprocessing software, LZT softwareis configured to produce results in real time,suiting it to live assessment and training,rather than solely for analysis and review.

The Z-scores described here are based ona published database and computed usingthe same software code that exists in theanalysis software, when used in ‘‘dynamicJTFA’’ (joint time-frequency analysis) mode.The database includes more than 600 people,ages 2 to 82. The system computes real-timeZ-scores using joint time-frequency analysisrather than using the Fast Fourier Trans-form, which is more commonly used forobtaining postprocessed results. As a result,Z-scores are available instantaneously, with-out windowing delays, and can be used toprovide real-time information.

Initial LZT implementations have used asingle Z-score, or a small number ofZ-scores, for example, ‘‘all coherences,’’ todevelop the feedback. In our work, we havecome to use generally all available Z-scoresin the training, providing an effective bound-ary around the EEG activity, within whichthe trainee learns to put their EEG.

All of the cases described here use a spe-cific form of Z-score training that hasevolved over several years (Collura, 2008a,2008b, 2009; Collura, Thatcher, Smith,Lambos, & Stark, 2009). Using this method,up to 248 simultaneous Z-scores are trainedat once, using a single metric that reflectsthe instantaneous state of all of the Z-scores.The method makes it possible to target parti-cular Z-scores for normalization, whileavoiding overtraining ‘‘outliers’’ and whilealso giving the brain sufficient freedom tochoose a path of self-regulation that is notlimited to ‘‘training to the norm.’’

METHODS

The concept of using Z-scores to providebiofeedback in real time was proposedby Thatcher (1998, 1999, 2004). Colluraand Thatcher (2006) discussed details andimplications of a practical design approach.The first reported implementation with clini-cal results were reported by DuRousseau

(2007), Smith (2008), Stark (2008), Wigton(2008), and Collura et al. (2009). Thesereports included six case studies with docu-mented QEEG and clinical benefits, andemployed the Lifespan Database reportedby Thatcher (1998) as a means of computingZ-scores in real time. Based on these scores,feedback variables were computed andreflected to the user in the form of soundsand graphic feedback of the type normallyused for conventional neurofeedback. Sincethese reports were published, a number ofclinicians have adopted the use of four chan-nels of what we now call LZT in their prac-tices. This article compiles a set of casestudies that were submitted upon request, asa means for disseminating clinical findings,as well as psychometric and QEEG data thatare available. All of the cases in this articleused the same approach to LZT training,which is described in more detail next. Thisarticle also discusses possible relationshipsbetween the specific training algorithms used,and the EEG changes that were observed.

Figure 1 shows the Live Z-score Text dis-play panel that is used by the practitioner.All 248 Z-scores are displayed for the four-channel montage. The display updatescontinually, and the color of the text indi-cates whether the Z-score is currently withinor outside of the standard boundaries of1.0, 1.5, and 2.0 standard deviations. Clini-cians learn to watch this screen, to quicklyidentify deviant scores, and to watch for pat-terns in time and space, as the brain adjusts tothe training parameters being presented. Thecolor coding rules for this text display are notadjustable and do not depend on the trainingcriteria. They are therefore a consistent rep-resentation of the client’s brain state; changesin this screen reflect objective changes in theEEG, and this screen can be relied upon togive a dependable ‘‘reading’’ of the client’sbrain. The screen becomes, in a sense, anavigational panel that guides the assessmentand treatment in real time. This approachcompresses the usual hours, days, or weeksrequired to get a quantitative EEG assess-ment into a fraction of a second and performsthe assessment continually. The differencebetween watching live Z-scores and readinga conventional QEEG report is similar to

Clinical Corner 23


the difference between watching a baseballgame or a golf match compared with readingthe statistics after the fact. They both give avaluable indication of the condition and stateof the subject, but the live representation allowsthe practitioner to understand how the num-bers got to be what they are, in terms that aredynamic and neurophysiologicallymeaningful.

While observing the LZT Text display,clinicians set up reward training by specifyingthree key parameters: the upper and lower

bounds of the Z-score ‘‘target window’’ andthe percentage of Z-scores that must bewithin these bounds to achieve a reward.The use of the percentage of Z-scores is anonobvious yet critical step and has pro-found impact on the training. Its value hasbeen found through repeated clinical treat-ment sessions and by observing the LZT Textdisplay during the training process.

Figure 2 shows two additional panels ofthe training screen. The upper area shows

FIGURE 2. Live Z-score display panels used to control feedback.

FIGURE 1. Live Z-score text display (four channels! 248 Z-scores). Note. Z-scores are colored to show whenthey are above normal range (yellow, orange, red) or below normal range (green, turquoise, blue).

24 JOURNAL OF NEUROTHERAPY


the current value of the ‘‘PercentZOK’’value, which is the percentage of Z-scoresmeeting the training criteria at the currentmoment. In addition, the percentage of timethat the trainee meets the conditions, beingthe ‘‘percent reward’’ is also shown. Theupper and lower Z-score targets are thenshown. The lower area shows the progressof the training, as the PercentZOK variableis being monitored and trained. There is athreshold that it must meet, to get a reward,and there is also the percentage of time thatthis is being met, shown in a trend graph.The clinician watches all of these values clo-sely and watches the trend graph throughouttraining. This gives the clinician controlof the variables that determine how thefeedback is produced.

This method employs a Z-score ‘‘target’’that is expressed in standard deviations (e.g.,"1.5 to 2.0 SDs) and produces rewards whena specified percentage of all Z-scores meet thecriterion. It does not require all Z-scores tobe within the target window. The percentageof Z-scores achieved thus becomes a

‘‘proportional’’ feedback variable, ratherthan a simple ‘‘on=off’’ feedback signal.

One might consider widening the targetwindow to accommodate all Z-scores.However, this was observed in early studiesto provide the brain with too much freedomin which to operate. For example, usinga wide-enough window to accommodatehighly deviant amplitudes would allow otherparameters to move from a normal to anabnormal range, while the EEG continuedto meet the overall training condition. Thismotivated the approach that allows someZ-scores to remain outside the target rangeyet effectively be ignored.

Figure 3 shows the important relationshipbetween target size and percentage of Z-scores required, in various training scenarios.This concept was not obvious when workbegan, and it is not necessarily obvious toother developers of real-time Z-score systems.Naively, one might think that by allowing asubset of Z-scores to provide the feedback,it is no different than simply ignoring someset of scores, or setting an upper range where

FIGURE 3. Relationship between size of LZT targets, and the use of ‘‘Percent ZOK’’ to establish reward cri-teria. Note.Within this model, clinicians can choose which aspects of training to emphasize, and which to vary.

Clinical Corner 25


Z-scores are allowed to go. But neither ofthese methods produces the same effect.When one overtly ignores Z-scores, then thepractitioner has decided what is importantand what is not. If an upper band of Z-scoresis rewarded, this would produce a ‘‘vortex,’’or attractor, that would pull certain scoresinto the abnormal range.

By specifically allowing a percentage ofany of the Z-scores to be out of the targetrange, the brain is allowed to decide how itmeets the strategy of normalization. Extremeoutliers are effectively ignored, but whichscores they are may change from momentto moment. Finally, by selecting the targetsize and position, it is possible to ‘‘comb’’through the tangle of Z-scores, and give thebrain information relating to a certainboundary of its function, and allow it tolearn from different regions of its function.

Different practitioners give differentemphasis to the use of this range of rewardstrategies. One clinician (NW) emphasizesworking at the low end of the reward range,using 25% to 40% feedback rates. Otherpractitioners work at higher levels of reward,

up to and above 90% reward. Generally, it isfound that adjusting these values during thesession is valuable, and provides importantflexibility to the training.

Figure 4 shows the LZT review and selec-tion screen. This is used to select Z-scores forgraphical analysis and to search for deviantZ-scores to include on the report. By specify-ing the condition for Z-scores to be viewed,the system then selects those that have metthat condition and allows them to be viewedand graphed. This is useful when checkingZ-scores after a session, to determine howthey have changed. As in the training displaypanel, patterns in the deviant Z-scoresare visible evident, so that combinations ofamplitude or connectivity scores that tendto ‘‘go out’’ together, or show definitepatterns, are readily recognized.

Figure 5 shows a typical graphical sum-mary of progress. This is the change in themost deviant Z-scores during a 40-min dem-onstration session. The change in Z-scoreswithin the session is clearly evident. Thisgraph is extremely useful in seeing the brain’sprogress during a session and between

FIGURE 4. Live Z-score selection screen, used for review of session data.



sessions, by watching which scores move invarious directions.

Figure 6 shows the progress of the Multi-variate Proportion value during the session.This is an important aspect, as it shows thatthe trainee has ‘‘acquired the task,’’ in oper-ant conditioning terms, and is able to improvetheir performance. This improvement provesthat learning has occurred, and it is an impor-tant aspect of LZT training.

This approach allows the brain to developits own strategy, as the rewards are achievedwhen a criterion is reached, and the brain isable to compute its own ‘‘cost function’’ tooptimize rewards. In some of the studiesshown here, certain Z-scores remained out-side the normal range during training,reflecting the fact that the brain was adopt-ing specific mechanisms to cope with thereward strategy. This amounts to the brain

FIGURE 5. Live Z-score changes plotted across 40min, one session with a naı̈ve trainee.

FIGURE 6. Progress of multivariate proportion (% of Z-scores meeting criteria) plotted across 40min, onesession with a naı̈ve trainee.

Clinical Corner 27


discovering dynamics that allow it to reducethe overall index of abnormality whileallowing certain features to remain outsidenormal, and to function as coping or com-pensatory mechanisms. This is significant,as it avoids the pitfall of ‘‘training to an out-lier’’ that may result when all Z-scores arerequired to meet the training targets.

RESULTS

Practitioners who adopted the LZTapproach and had been instructed on itsuse were invited to submit case studies thatillustrated their experience with the tech-nique. Not all cases were submitted, so thisarticle presents selected successful cases thatwere submitted. The details for each case,as well as clinical, behavioral, psychometric,and QEEG changes, when available, aresummarized in Table 2.

Application of Training Protocol

All of the reported cases used a form ofwhat we refer to as ‘‘multivariate compositetargeting,’’ in which a number of Z-score tar-gets are continually assessed in a particularway, and used to produce the feedback. Allof the reported cases used this capability inthe form of the ‘‘Percent ZOK’’ algorithm inLZT. In all cases, there were individual differ-ences in the precise strategy and control meth-odology used. The protocol provides sufficient

freedom for the clinician to determine the nat-ure and extent of information presented to thebrain, relative to the current state of the mul-tiple Z-scores. This is to be distinguished tomore simple ‘‘training to the mean’’ that maybe assumed to be used, if one is not knowledge-able of the relevant details.

Various practitioners differ in their preciseuse of the controls and options within theLZT paradigm. Although all practitionersmake use of target size and percentageof Z-scores as control variables, the exactprocess used varies. Some practitionersemphasize adjusting target size and allowingthe brain to learn the various levels of taskdifficulty, whereas others focus on the per-centage of Z-scores as the key variable. How-ever, as the values are interrelated, there isalways a dynamic interplay in changing one,the other, or both. The LZT approach allowsthe practitioner to emphasize different aspectsof the Z-score training, based on the observedclinical and electrophysiological changes.

Summary of Cases

Details for all cases are presented in thesecond table. This section summarizes specificresults including Z-score statistics, QEEGmaps, and relevant psychometric results,which illustrate the clinical and electrophysio-logical changes which were observed.

Three cases (‘‘S,’’ ‘‘C,’’ and ‘‘Z’’) were pre-sented by one clinician (JG), who uniformly

FIGURE 7. Percentage of amplitude Z-scores outside #1.0 SDs for 3 participants ‘‘S,’’ ‘‘Z,’’ and ‘‘C,’’ 10sessions each.



used 10 sessions of four-channel LZTtraining on each of three patients. Figure 7summarizes the percentage of amplitudeZ-scores (absolute power, relative power,power ratios) that were outside the nominalrange of #1.0 SDs. This illustrates therepeatable reduction in deviant Z-scoreswhile revealing some difference betweenparticipants. It is interesting that the parti-cipants who presented with the most deviantZ-scores on the outset showed the leastnumber of deviant scores after training.

Figure 8 shows the percentage of connec-tivity Z-scores (coherence, phase, asymmetry)outside the range #1.0 SDs for the samethree participants. Although uniformreduction is seen in two participants, anactual increase is seen in one. It is shownlater that this reflects a compensatory mech-anism, in which the brain evidently allowssome scores to become deviant, to achievegreater overall normalization. Note that thisphenomenon likely depends on the abilityof the training software to allow some scores

FIGURE 8. Percentage of connectivity Z-scores (asymmetry, coherence, and phase) outside #1.0 SDs, 3participants, 10 sessions each.

FIGURE 9. Number of amplitude Z-scores outside target range as a function of target size, participant ‘‘S.’’Note. The number of Z-scores outside the narrow range 1.0 actually rises, whereas the overall distributionpulls strongly within the 1.5 and 2.0 ranges.

Clinical Corner 29


FIGURE 10. Number of connectivity Z-scores outside target range as a function of target size, participant ‘‘S.’’Note. Although many scores begin outside the 2.0 and 2.5 SD ranges, all Z-scores fall within the range 1.5 SDafter 10 sessions.

FIGURE 11. Pre- and posttreatment QEEG maps for participant DQ.



to remain outside normal bounds whilecontinuing to reward for training.

Figure 9 shows the number of amplitudeZ-scores outside specified target ranges, forone participant.Note that, at the outset, a greatmany scores are outside even the larger targetsof 2.5, 3.0, and larger targets.After training, thenumber of scores outside of the larger rangeshas reduced dramatically. At the same time,the number of scores outside the range #1.0has increased, as a result of ‘‘packing’’ of thescores into the narrower ranges. This reflectsthe ability of the brain to achieve overall nor-malization while having room to move at thelower limits of the training targets.

Figure 10 shows the number of connectiv-ity (coherence, phase, asymmetry) Z-scoresoutside of specified target sizes. Thereduction of scores outside the range #1.5is essentially complete, as no Z-scores arefound outside this range after the training.This demonstrates the brain’s ability to pro-cess dozens of Z-scores in a single trainingparadigm, and effectively normalize all ofthem, while performing a simple training.

Figure 11 shows pre- and post-QEEGmaps for participant ‘‘DQ’’ presented byclinician JT. Improvements are visible inboth the power and the coherence maps,after 39 sessions.

FIGURE 12. Pre- and posttreatment maps (26 sessions) for participant TB. Note. Top pair: Absolute powermaps, Bottom pair: Coherence maps.

Clinical Corner 31


Figure 12 shows pre- and post-QEEGmaps from participant ‘‘TB’’ (clinician JT).The normalization of power as well ascoherence maps is visibly evident after 26sessions. It is interesting to note examplesof ‘‘overshoot’’ in which one value movesout of normal range, whereas others normal-ize. For example, a slight excess of right deltaand theta appears, whereas the excess of highbeta and alpha normalizes. Also, there is aslight alpha hypocoherence that appears,along with the significant normalization ofother coherences.

Figure 13 shows pre- and post-QEEGmaps, and Figure 14 shows post-QEEG

maps after 20 sessions, for participant‘‘Norb’’ (clinician F8) using a Laplacianderivation. All evident power and asym-metry abnormalities are seen to effectivelynormalize. Although the pre maps show sig-nificant delta and theta excess, and a deficitof high frequencies, the post maps are essen-tially normal, with only a very slight betaexcess on the left, which may be musclerelated.

Figure 14 shows IVA$Plus StandardScales Analysis for ‘‘Norb.’’ The Full ScaleResponse Control Quotient rises from 94 to101, whereas the Full Scale AttentionQuotient rises from 62 to 95.

FIGURE 13. Pre- (left) and posttreatment (right) QEEG maps after 20 sessions, patient Norb.

FIGURE 14. Pre- and posttreatment IVA$Plus, patient Norb, 20 sessions.



Figure 15 shows pre- and post-QEEGmaps for participant ‘‘44YOM’’ (clinicianNW), after 25 sessions. Visible trendsinclude a reduction in coherence and phaseabnormalities, and some improvement inrelative power.

Figure 16 shows IVA$Plus results fromparticipant ‘‘44YOM.’’ Full-Scale ResponseControl Quotient rises from 29 to 94,whereas Full Scale Attention Quotient risesfrom 0 to 96.

Figure 17 shows the progression ofparticipant ‘‘12YOM,’’ clinician PR, with eyesopen, after 20 and after 40 sessions. This pro-gression shows definite compensatorymechanisms at work, as the post-20 sessionmaps show interesting adaptations to thetraining. In absolute power, areas that weredeficient in delta are normalized in 20sessions, whereas surrounding areas exhibit adelta excess. This suggests a global reregula-tion mechanism wherein surrounding areasadjust their function, as a compensation forthe normalization of other areas. Similarly,whereas hypocoherence across frequency

bands is seen to remediate, there is a signifi-cant hypercoherence in high beta thatemerges, again likely as a compensating mech-anism. The final condition characterized byoverall normalization in the presence of hyper-coherent beta, suggests amechanism involvingcortico-cortical binding, as a strategy towardproducing the requisite overall normalization.

Figure 18 shows the progression ofeyes-closed QEEG maps for the same clientas Figure 17. A rather different pattern ofnormalization is evident, suggesting that thebrain adopts a different strategy to keepingitself normalized, depending on whether theeyes are open or closed. In this case, the finalQEEG shows essential normalization, alongwith a phase deficit (phase-locking) in alphaacross the head. This suggests a stronger-than-normal thalamocortical binding, inwhich thalamic activity is controlling corticalrhythms with excessively tight timing. Inother words, rather than having several some-what independent alpha generating processes(e.g., occipital, frontal, temporal), the brain isdominated by a single alpha pacemaker.

FIGURE 15. Pre- (left) and posttreatment (right) QEEG maps, 25 sessions, patient 44YOM.

Clinical Corner 33


Figure 19 summarizes the progress ofparticipant ‘‘12YOM.’’ In our analysis anddiscussions with clinicians, observed changes

in Z-scores may indicate compensatingmechanisms in which areas surrounding theoriginal delta excess change their state, as a

FIGURE 16. Pre- (left) and posttreatment (right) IVA Standard Scales Analysis, 25 sessions, subject 44YOM.

FIGURE 17. Overview of client 12YOM progress in Eyes Open condition, after 20 and 40 sessions.



way of containing the abnormal areas. Whenthere is a worsening of an EEG deviation, wehypothesize that it may represent a com-pensatory mechanism. For example, whenthere is an area with a deficit of deltaactivity, surrounding areas may exhibitexcess delta during the normalization pro-cess, even as the previously abnormal areasnow become normal. Temporarily, thesechanges look like a cap or a wrapper aroundthe areas which in fact are normalized in the20-session maps. So the brain has adopted astrategy to lower its mean scores, by allowing

the surrounding areas to provide acontaining medium for the abnormalactivity, as the brain globally produceshigher delta amplitudes. The global natureof the delta is indicated by the newlyemerged hypercoherence after 20 sessions.

After 40 sessions, when most all amplitudeand connectivity measures are normalized, weare still left with very specific EEG abnor-malities that again seem to be coping or com-pensating mechanisms. As an example, thisclient had extreme phase synchrony in alphawith eyes closed and extreme beta coherence

FIGURE 18. Overview of client 12YOM progress in Eyes Closed condition, after 20 and 40 sessions.

FIGURE 19. Summary of changes EC and EO.

Clinical Corner 35


with eyes open. We believe that thebrain may be setting up its own bindingmechanisms, again, to maximize the globalnormality. In this case, where the remainedphase synchrony and coherence abnormali-ties, it was nonetheless noted that the clientexperienced significant clinical benefits uni-formly from treatment. To quote the clin-ician,

Changes in delta absolute power andcoherence over the course of the train-ing indicate some interesting possibili-ties for future research. There may bea mid-training phase that prioritizesallocation of cortical resources for thepurpose of reorganizing neural connec-tivity. Hypercoherence could be a mani-festation of increased thalamocorticalactivity which necessitates a temporarydiversion of energetic resources toimprove the efficiency of the interactivepathways between the thalamus and thecortex. (Rutter, 2009)

Figure 20 shows pre- and post-QEEGmaps from participant SonjaK taken beforeand after her 16th session, after having had15 previous weekly sessions (clinician DK).These maps thus document single-sessionresults. The Beta and High Beta powerexcesses are trained into the normal range.An extreme amount of hypercoherence indelta and theta is reduced during the singlesession. Specific hypo-phase (phase-locking)in alpha is also slightly reduced.

Figure 21 shows pre- and post-QEEGmaps from participant ‘‘NW’’ (clinician JT)after 38 sessions. Remediation of powerabnormalities and coherence deficits is visi-bly evident. An insignificant amount ofhypercoherent posterior alpha remains.

Figure 22 shows pre- and post-QEEGmaps from participant TA after 20 sessions(clinician CRS). The remediation of coher-ence in beta and high beta is striking. Thereis also a visible reduction in theta absolutepower, which was targeted using conven-tional thresholded feedback in addition tothe LZT feedback, in a combined protocol.

FIGURE 20. Pre- and posttreatment maps before and after one session (Session 16) for SonjaK.



Commonalities in Clinical Results

The overall numbers of studies with out-come reporting, and the numbers showingimprovement, are summarized in Table 1.Detailed specific observations are summar-ized in Table 2.

Of the 19 cases reporting presentingsymptoms, they were 7 Cognitive and Affect-ive Problems, 5 ADD=ADHD, 3 AutisticDisorder or ASD, 2 Behavioral Problems, 1Cerebral Palsy, and 1 TraumaticHead Injury.

All respondents reporting clinical out-come identified clinical improvement duringthe treatment sessions, as noted in Table 2.

QEEG Results

When QEEG data are available, allrespondents show visible improvement in

QEEG maps relative to the normative data-base used for analysis. In most cases, theNeuroGuide ANI (‘‘Lifespan’’) databasewas included in the analysis. In some cases,the QEEG normalization is dramatic, result-ing in essentially normal QEEG maps afterthe training. In other cases, we see eitherremaining abnormalities or newly emergentdeviations that may reflect compensatingmechanisms. The availability of pre- andpost-QEEG maps is found to be of consider-able value in monitoring and assessing theprogress of LZT training. Whereas LZTtraining can be viewed as an automatic guid-ance mechanism for feedback, EEG andQEEG analyses provide important infor-mation guiding the placement of sensors,choice of protocols, and management ofanticipated and observed clinical and beha-vioral changes. In some of the cases, indicessuch as the NeuroGuide Traumatic Brain

FIGURE 21. Pre- and posttreatment QEEG maps for participant NW, 38 sessions. Note. Top pair: AbsolutePower maps, Bottom pair: Coherence Maps.

Clinical Corner 37


Index or Predicted IQ are reported, andshow improvements.

Abreactions/Negative Side Effects

In this clinical series, no abreactions to theLZT training were noted. In earlier clinicalwork, one initial mechanism identified forabreaction occurred when the target windowrequired for Z-scores was excessively wide(e.g., #.3.0 SDs). In one case (not reported

here), an individual presenting with excessiveEEG absolute power was trained with a widewindow. The result included an ‘‘overshoot,’’in which excessively high power Z-scores werefound to become excessively low, in effectfinding another limit at which to function. Itbecame necessary, through the software, toprovide separate upper and lower limitvalues, so that the Z-scores could be trainedwith an upper limit of 3.0 standard deviationsbut a lower limit of "1.0 standard deviations.Training with this modification eliminated

FIGURE 22. Pre- (left) and post- (right) QEEG maps, 20 sessions, participant TA.

TABLE 1. Overall studies with outcome reporting and the numbers showing improvement.

No. Reported Visible Improvement

Total cases reported 24Reporting presenting symptoms 22 22Reporting clinical=behavioraloutcome

23 23

Pre- and posttreatment LZT data 10 10Pre- and posttreatmentQEEG data

12 12

Pre- and posttreatment IVA data 5 4

Note. LZT! live Z-score training; IVA! Integrated Visual and Auditory Performance Test.



TABLE2.Summarize

dca

sedetails,includingclinical,behavioral,psy

chometric,andQEEG

changes.

Study#

CLIN

IDPrese

ntin

gPreviousTreatm

ent

Preassessment

#LZTSessionsandSite

sResu

ltEEG

Change

1JG

Z17YO

male

seve

rely

autistic

w=little

speech

,frequenttantrums,

short

attentio

nsp

an,doesnot

listento

instructions

LZT:highamplitude

generally,co

herence

deficits

10se

ssionsF7F8T5T6

more

clarity

insp

eech

and

longerse

ntence

s,talksmore

appropriately,less

andless

tantrumsandmuch

calm

er,

more

willingto

carryout

instructionsandtriesall

dishesthatthemaid

cooks

Reductionin

aberrant

Z-Sco

res:

1st

session:36

Z-sco

res>1.0

SD

during

session.10th

session:13

Z-sco

res>1.0

SD.

Reductionin

both

amplitudeand

connectivity

Z-Sco

res

2JG

C7YO

female

cerebral

palsy,

speaks

twoto

threewords,

limite

dgross

andfin

emotor

skills(can’twalk

yet’’,

shortattentio

nsp

an,

dreamy

Othmerwideinhibitsite

specific

6mo.Slow

but

steadyprogress

LZT:highamplitude

generally,co

herence

deficits

10se

ssionsF7F8T5T6

more

resp

onsive

tothe

environment,sp

eakingmore

with

longerse

ntence

s,reciproca

tetwo-w

ay

communicatio

n,interacts

more

with

otherch

ildrenin

school,demonstratesmore

strength

inhermove

ments

Reductionin

aberrant

Z-Sco

res:

1st

session:35

Z-sco

res>1.0

SD

during

session.10th

session:27

Z-sco

res>1.0

SD.

Reductionprimarily

inamplitudeZ-Sco

res

3JG

Shaan

52YO

female

with

occasionalmemory

lapse

s,difficulty

inrememberingnew

learning

LZT:highamplitude

generally,co

herence

deficits

10se

ssionsF7F8T5T6

able

toremembermost

ofthe

stepsoftheChinese

classicaldancingandis

able

toabso

rbmore

ofhear

learningoftheKorean

language.Now

whensh

etake

sco

ffeesh

edoesnot

have

aheadach

eanym

ore

Reductionin

aberrant

Z-Sco

res:

1st

session:60

Z-sco

res>1.0

SD

during

session.10th

session:31

Z-sco

res>1.0

SD.

Reductionin

both

amplitudeand

connectivity

4JT

DQ

9YO

male,behavioral

problemsatschooland

home,incidents

of

verbalandphysical

aggression,inability

toco

ntroltemper,

stuttering

place

don1=2

day

attendance

,individual

behavioralplan,

vestibularand

proprioce

ptivetraining

QEEG

map:significa

nt

lack

ofalphaactivity,

significa

ntexcess

of

delta

activity

frontally,

mixedhyp

er-

and

hyp

o-coherence

39se

ssionsF3F4P3P4,

FzPzC3C4,F3F4Cz

PzT3T4FzCzF3F4

P3P4alphaenhance

addedwith

emphasison

parietalviathresh

olds

speech

improve

dsignifica

ntly

after15se

ssions,

much

less

stutteringataroundse

ssion

17,less

argumentativeand

stoppedtempertantrums

afterse

ssion22,hadstopped

takingmedicatio

nsatse

ssion

26,se

emedhappierand

friendlieratse

ssion27.

improve

denergyandaffect,

more

cognitive

flexibility,able

tohandle

transitio

nsin

sessionandin

office,no

longerstutteredafter39

sessions.

planto

return

tofull

dayattendance

BASCbehaviorassessment

system

resu

ltssh

ow

improve

ment14of28

scoresim

prove

d2or

more

standard

deviatio

ns,

pre-andpost-Z-Sco

res

show

changes

significa

ntly

inhighbeta

frontally,andin

reductionsin

hyp

ercoherence

s

5JT

TB

11

YO

male,

refusa

lto

complete

schoolwork,

easily

distracted,

unorganizedwork

area,

IVA

shows

diagnosis

of

ADHD,CombinedTyp

e.

QEEG

map:significa

nt

excess

ofhighbeta

activity

focu

sedin

parietalregions,

26se

ssionsT3P3Fp2P4

thenFp1Fp2FzCz

manydeviatio

nsvisibly

cleared

upmidwayse

ssionsasearly

asse

ssion2.After7th

session,‘‘hasbeenve

ry

individualizedbehavior

questionnaireprovidedto

motherandch

ecklists

scoredandave

ragedas

(Con

tinu

ed)

39


TABLE2

(contin

ued)

Study#

CLIN

IDPrese

ntin

gPreviousTreatm

ent

Preassessment

#LZTSessionsandSite

sResu

ltEEG

Change

flataffect,‘‘head

twitching’’,

avo

idance

of

physicalactivities,

physicalawkw

ardness

=clumsiness,andso

cial

immaturity

significa

nt

hyp

ercoherence

inall

bands,

esp

ecially

beta

andhighbeta

cooperativein

schoolthis

week’’,‘‘tryingnew

thingsat

swim

lessons’’,‘‘only

notice

ticsoccasionally,after

session13,‘‘finally

rodea

bike!"

pre-andpost-treatm

ent

scores.

IVA

perform

ance

testsim

prove

d,3of6

scoresim

prove

dmore

than2standard

deviatio

ns

6FS

Norb

21YO

female,cu

rrently

on

aca

demic

leave

from

college‘‘requiredto

get

mentalhealth

treatm

ent

togoback

toschool’’

difficulty

makingfriends,

socializing,moderate

sxofdepression.

Engrossedin

inner

‘‘fantasy

life"

diagnose

dwith

ADHD,

place

donstim

ulant.Not

currently

taking

prescribedmedicatio

n(adderall).Currently

takingce

lexa

20mg.

Previouslytookproza

k20mg

QEEG

map:excess

delta

andtheta

left

centrotemporal,

anorexia,bingepurging

type,remission.ADD

by

report.MDE:moderate,

partialremission,mTBI

byreport.SocialPhobia

from

mother:Idose

eim

prove

mentin

B.Seemsto

befollowingthroughbetter

andco

mpletin

gtasks.

Has

discu

ssedwith

methe

reco

mmendatio

nsaboutpast

rigity.Agreethatsh

eneeds

tolearn

toalte

rherplansand

stillaccomplishtasks.

She

hasco

mealongwayandI

seeagreatdealof

improve

ment.organizatio

nis

betterandco

nfid

ence

isup.

QEEG

mapsh

ows

significa

ntnorm

alizatio

n,

amplitudesnow

norm

al.

TBIdiscrim

inantreduce

dfrom

porobability

of97.5%

to70.0%.IVAplusscores

pre-andpost

show

improve

mentin

attentio

nsu

bscales

7FS

bers1

54YO

CF

(anesthesiologist),

diagnose

dwith

ADD,

inattentivetype,

GeneralizedAnxiety

Disorder,Recu

rrent

MajorDepression.

Complaints

relatedto

inattentio

n,fatig

ue,

worry,

and‘‘feelingblue"

Methylphenidate

96mg

daily,Lexa

pro

20mg

daily.Disco

ntin

uedand

wash

edoutpriorto

QEEG.

IVA$PlusQ

Sco

res;

QEEG

EC

andEO

showeddecrease

dpowerat1Hz,

frontal

low

powerF3andF4.

low

power9–10Hzin

all

leads.

Increase

dpower

ove

rT4,T6from

13–15Hzand24–30Hz,

maximaldeviatio

nat

28Hz.

20

sessions

conve

ntio

nal

training,monopolarto

reduce

fruntaltheta,10

sessionsalpha

asymmetrytraining.20

LZTse

ssionsEO

Fz=

T3=T

4=P

zusingDVD

playe

r

IVA$Plus

QSco

res

show

improve

mentof10%

to20%

inGlobalResp

onse

Control,

GlobalAttentio

n,and

Audito

ryResp

onse

Control

andAttentio

nSubscales.

Hyp

eractivity

indexincrease

dfrom

95to

109.Im

prove

ment

inove

rallattentio

n,

occupatio

nalfunctioning,

increase

dproductivity,better

organizatio

n,greaterjob

satisfaction.decrease

inBeck

Depressionfrom

25at

base

lineto

14atpost

LZT.

Changespersistedat3mo.

follow-up.Lowered

Methylphenidate

to70mg,

disco

ntin

uedLexa

pro

QEEG

showsgeneraltrend

toward

norm

alizatio

n,

changesof1.0–2.0

SD

notedove

ralltraining

sites.

Low

Powerat

9–10Hzco

mpletely

norm

alized.

8FS

math1

IVA$PlusQ

Sco

res

IVA$Plus

QSco

res

show

improve

mentof10%

to30%

inGlobalResp

onse

Control,

andAudito

ryResp

onse

Controla

ndVisualP

rudence

.

40


(Con

tinu

ed)

Hyp

eractivity

indexincrease

dfrom

69to

86

9FS

pase

1IVA$PlusQ

Sco

res

IVA$PlusQ

Sco

ressh

ow

improve

mentof10%

to20%

inGlobalResp

onse

Control,

GlobalAttentio

n,and

Audito

ryResp

onse

Control

andAttentio

nSubscales.

Hyp

eractivity

index

decrease

dfrom

95to

50

10FS

skeg1

56YO

CF,guidance

counse

lor,difficulties

with

comprehension,

earlymorningwaking

ove

rse

veralye

ars.

Complaints

relatedto

hyp

eractivity,audito

ryproce

ssing,audito

rymemory.

long-actingstim

ulant,

SSRI,and

non-benzo

diaze

ine

sleepaid.Disco

ntin

ued

andwash

edoutpriorto

QEEG.

IVA$PlusQ

Sco

res

consistentwith

proce

ssingprudence

andaudito

ryco

mprehension

problems.

Global

abnorm

alitiesove

rF3,

F4,T3,T4,C3,C4,P3,

P4,O1,O2.Slow

wave

spredominate

ove

rT3

andO2between1–3Hz,

fast

activity

ove

rO2in

15–20Hzrange.Loreta

identifiesBrodman

areas10,30and

anteriorcingulate

abnorm

alities.

9LZTse

ssionseye

sopen,

30–40minutes,

P3=P

4=

O1=O

2usingDVD

and

complextonereward.

$1SDin

delta

atF3,F4,

C3,C4,T3,T4.Also$1

SD

inbeta

andgamma

atT3andT4.Move

toF3=F

4=P

3=P

4andF3=

F4=T

3=T

4to

target

proce

ssingand

attentio

n.

20

LZT

sessions

completed.

IVA$PlusQ

Sco

resmostly

steady,

show

improve

mentof

10%

to20%

inAudito

ryPrudence

,Audito

ryComprehension.

Hyp

eractivity

indexminim

al

changefrom

109to

111.

Subjectivereport

improve

mentin

audito

ryproce

ssing,audito

rymemory,

ove

rallattentio

n.Noted

improve

ments

inoccupatio

nalfunctioning,

increase

dproductivity,better

organizatio

n,greaterjob

satisaction.Sleepim

prove

dmost

ove

rall,

nolonger

requiringsleepmedicatio

n.

Generaltrendtoward

norm

alizatio

nvisible

inQEEG,most

evidentin

low

frequency

range

(1–3Hz)

inleftfrontal,left

temporal,andoccipita

lleadsin

mid

tohigh

frequencies(16–22Hz).

Decrease

inco

herence

notedbetweenP3O1and

P4O2.

11JB

11YOF

11YO

female,depression

andanxiety,irritability,

low

moticatio

n,high

forgetfulness

&disco

ganizedattentio

n,

poorschool

perform

ance

nse

ssions

F3=F

z=P3=P

zF3=F

p1=P

3=P

zF3=F

4=

C3=C

4

could

only

tolerate

DVD

feedback

ofstop=g

o,but

wantedto

train

for40

minutes.

Did

notwantto

stop

NFtraining

able

toreduce

LZT

target

size

from

$="

2.0

SD

to$1.5="

1.2

SD

12JB

10YOM

10YO

male,pervasive

deve

lopmentaldelay

(6weeks

premature,3

weeks

inneonatalunit),

ADHD

symptoms,

inattentio

n,

hyp

eractivity,ca

n’t

handle

transitio

ns;

lack

ofco

operativeness,

problemsin

school;

difficulty

with

sleep

19se

ssionssitesbase

donQEEG,parental

reportofsymptoms&

subjectiveresp

onse

totraining

Improve

mentco

nsistently

noticedin

2days

post

training

andbytheendlastedforthe

fullweekwith

very

noticeable

improve

mentin

postivie

moodandleve

lof

cooperativeness

&noreports

ofproblemsatschoolforth

elast

twomonthsoftraining

parents

tracked

improve

mentonscales

for2–4-6

days

post

training.

(Con

tinu

ed)

41


TABLE

2(contin

ued)

Study#

CLIN

IDPrese

ntin

gPreviousTreatm

ent

Preassessment

#LZTSessionsandSite

sResu

ltEEG

Change

13JB

12YOM

12YO

male,depression&

ADHD

with

hairpulling

onleftsideatC3;

complete

shutdownsat

schoolwith

oppositio

nal

refusa

lto

dowork;not

doinganyhomework

medicatio

nhadbeen

ineffective&

wasatrisk

ofbeingse

ntto

aresidentia

lprogram

QEEG

showsfrontal

slowingatF3

1se

ssion

F3=F

4=C

3=C

3;

addedlow

alpha

coherence

reward

sound=v

isualF3F4&

C3=C

4,thenaddedlow

beta

coherence

reward

sound=v

isualforF3=C

3

single

sessionpositive

resp

onse

notedafterse

ssion

haslasted-parents

and

teach

ers

rate

itat75%

improve

ment=noneedfor

residentia

l.

norm

alizedF3

beta1=g

ammabyfirst

norm

alizingthelow

coherence

s;

14JB

14YOM

14YO

male,wantedto

be

offmedicatio

n,wanted

totrain

his

brain,poor

schoolperform

ance

,not

doingch

ores,

rather

irritable

and

argumentative

5se

ssions

F3=F

4=P

3=P

4with

addedreward

for

norm

alizingbeta

coherence

F3=F

4

reports50%

improve

mentin

sleeppattern,in

homework

completio

nandin

doing

choreswith

outparental

promptin

g.Hereports75%

improve

mentin

sports

perform

ance

and75%

improve

mentin

beingless

irritable,andproneto

anger

oroppositio

nalin

resp

onse

sto

parents.

15JB

17YOM

17YO

male,Asp

ergers

with

seve

redepression=

Anxiety.

6se

ssionstraditional

neurofeedback

(1–6),10

sessionsOthmer(14–

23)

23se

ssionstotal,LZT

for

sessions7–13

complete

symptom

reliefwith

both

depressionandanxiety

low

andmuch

improve

dschoolperform

ance

and

socialrelatio

ns.

16JB

10YOM2

10YO

male,ADHD;IQ

inthe70s,

distractible,

doesn

’tfin

ishtasks,

poor

memory,im

pulsive,

emotio

naloutbursts,

easily

frustrated,misse

sso

cialcu

es,

problems

with

reading&

comprehension

use

sAVE

andCaptains

Logathome

QEEG:

dim

inished

beta

activity,loca

lizedin

left

parietal&occipita

lareas;

more

pronounce

dEEG

abnorm

alities

undertask

17

sessions

C3=C

4=P

3=

P4C3=F

3=F

z=C4C3=

C4=T

5=F

z$C4SMR

Fp1=F

p2=F

3=F

4C3=

C4=T

5=F

z$bipolar

T3=F

p1&T4=P

4

Reductionin

aberrant

Z-Sco

res

17NW

44YOM

44YO

male,ADHD,

bipolar,occasional

anxiety

symptoms,

rarely

hasmanic

episodes,

multiple

blows

tothehead,rece

ntly

stoppedworking

variouspsych

otropic

medicatio

nssince

2001,

lack

ofadequate

symptom

reso

lutio

nfrom

medicatio

ns

QEEG

maps:

significa

nt

amplitudeand

connectivity

deviatio

ns

IVA:50$behavioral

symptoms=

functioning

issu

es

25se

ssionsF3=C

3=P

z=F4

F3=F

4=P

z=C4P3=P

4=

Cz=PzF3=F

4=C

z=Pz

beginningofsymptom

reso

lutio

natse

ssion5.able

tofunctionadequately

after

medicatio

ntitratio

n.After2

monthsand25se

ssions

clientsu

ccessfully

titratedoff

ofallmedicatio

nsandnew

QEEG

take

n.Reportsbeing

able

toove

rallfunctionbetter

thanbefore

NF,eve

nwith

out

medicatio

n.betterable

tofocu

sandfunctionin

QEEG:EO

mapssh

ow

significa

ntnorm

alizatio

n,

reductionin

amplitude

excesses,

andin

hyp

ercoherence

s;so

me

hyp

oco

herence

remains.

EC

mapssh

ow

significa

nt

norm

alizatio

n,so

me

residualhyp

ercoherence

.After9more

sessions,

clientreturnedto

work

andfelthis

improve

ment

42


(Con

tinu

ed)

business.nolongersu

pport

workingdiagnosisofADHD

wasdirectly

dueto

NF

trainingandthatheno

longerneededto

contin

ue

with

NF

18PR

12YOM

12YO

male,autistic

disorder,delayin

deve

lopmentofve

rbal

andnon-verbal

communicatio

n,lack

of

socialoremotio

nal

reciprocity,stereotyped

andrepetitivemotor

manners,im

pairedfin

emotor,Tourettes-like

physicalsp

asm

s,and

high-pitched

voca

lizatio

ns,

failure

todeve

loppeer

relatio

nsh

ips

QEEG

maps:

EC:bilateral

excess

ofhighbeta,

alphadeficitce

ntrally,

delta

excess

rt.Frontal,

delta

deicitce

ntrally,

broadfrontoce

ntral

hyp

oco

herence

s.

20

sessions

F3=F

4=C

3=

C4

verbalizatio

nsch

angedfrom

primarily

promptedand

time-delaye

dto

spontaneous

andreal-tim

e.Spontaneous

displays

ofaffection.

Decrease

inrepetitive

behaviors

andve

rbalizatio

ns.

Increase

dmotoricandve

rbal

self-regulatio

n.Increase

dvo

luntary

interactionwith

peers

inso

cialandschool

environments,im

prove

dsleeppatters,anddecrease

dnocturnalenuresis

QEEG

mapssh

owed

compensa

tion

mech

anismsatlow

and

highfrequency

after20

sessions,

‘‘containing’’

deficitareaswith

neighboring

compensa

tory

amplitude

excesses,

and

connectivity

‘‘ove

rshoot’’.

Furthernorm

alizatio

nse

enafter40se

ssions.

Evidence

ofsp

ecific

thalamic

andco

rtical

adaptivemech

anisms

dueto

operant

conditioning

19WLSam

7YO

male

rapid

learner,

easily

excite

dand

aggressivewith

other

children.Canbe

classifiedasAD=H

D

QEEG

maps:

EO:Delta

andtheta

excess

frontally,highbeta

excess

ove

rPz,

global

hyp

ercoherence

inall

bands,

particularlydelta

andhighbeta

21se

ssionsF3F4P3P4

visible

improve

mentin

behavior,notable

changein

interactionwith

NFstaff,

more

positiveand

cooperative,schoolreports

improve

dbehaviorand

reductionin

aggression

QEEG

mapssh

ow

significa

ntnorm

alizatio

n.

Phase

andrelated

measu

resremain

outside

norm

s

20DK

SonjaK

13

YO

female,diagnose

dICD

10earlych

ildhood,

F94.0

electivemutism,

now

F93.2

emotio

nal

disorderwith

anxiety,

F83.2

disca

lculia.

Emotio

naldiagnostics:

emothionalstress,total

blockingwith

‘‘open’’

tasks.

Markedso

ciala

nd

emotio

naltendency

towith

draw

successfulergotherapy,

waldorf-sch

ool7

thgrade

poormath,no

medicatio

n.HAWIK

3markedly

below

ave

rage,assumeddue

toanxiety

disorder

QEEG

maps:

significa

nt

excess

ofdelta

and

theta

diffuse

ly,andhigh

beta

occipita

l

15weeklyse

ssionsT3T4

C3C4Fp1Fp2O1O2F3

F4T5T6C3T3T5P3

Fp1F4T4P4(address

mirrorneurons;

empathy)

high

compliance

dim

inished

feelings

of

being

dominated,betterability

tostandupto

classmates,

beginssh

opping,meetin

gwith

friends.

Reports

increase

dalertness,

awake

ness

aftereach

session.Thinks

ofNFdisplay

‘‘dolphins’’duringschool

tests,

increase

dreaso

ning.

Improve

dse

lf-assurance

,math,so

cialskills,

open-m

indedness,ve

rbal

communicatio

n

QEEG

mapssh

ow

norm

alizatio

nduring1

session(session16),both

amplitudeand

connectivity.T3-C

4hyp

ercoherence

reduce

d,

Theta

T3,C4,T4,

hyp

ercoherence

fast

wave

snorm

alized,T5-P3,

prefrontalDelta

,ove

rall

frontalslowing

norm

alized.IVA,TOVA,

CPTforthco

ming

21JT

NW

9YO

female

moodsw

ings,

impulseco

ntrol

problems,

conce

ntratio

ndifficulties.

18mgConce

rtaplus

GABA

QEEG

maps:

excess

amplitudes,

hyp

oco

herence

theta

andalphaposterior

temporal

38se

ssions4-channel26

sessionsC3C4P3P4

11se

ssionsF3F4CzPz

atse

ssion27teach

erreported

significa

ntim

prove

ments

inse

lfco

ntrol.Atse

ssion33,

parents

indicatedthat

remainingproblemswere

due

BASCbehaviorassessment

system

resu

ltssh

ow

improve

ment,pre-and

post-Z-Sco

ressh

ow

changes.

24z-scores

(Con

tinu

ed)

43


TABLE

2(contin

ued)

Study#

CLIN

IDPrese

ntin

gPreviousTreatm

ent

Preassessment

#LZTSessionsandSite

sResu

ltEEG

Change

tofamily

dyn

amicsandold

behavioralhabits

-family

soughttherapy

show>1SD

change.

QEEG

mapssh

ow

essentia

lnorm

alizatio

n22DSTA

14YO

male,

preadolescentLD,

conce

ntratio

nproblems,

impulsivebehavior

Seve

ralADD

medicatio

ns

previouslyw=lim

ited

benefits,

unacceptable

sideeffects.

10mg

Abilify

for2.5

yrw=no

sideeffects,

some

increase

dco

nce

ntratio

n

20se

ssionsLZT

w=a

mplitudetrainingto

lower3–8Hz,

raise

16–19,andlower

22–30Hz.

self-reportedclinical

improve

mentin

allsymptoms

QEEG

mapssh

ow

light

reductionin

theta

z-scoresaswellas

20–30Hzz-scores.

16–19Hzrangerise

sto

%"1.6

SD.Essentia

llyco

mplete

coherence

norm

alizatio

nand

moderate

theta

abso

lute

powerim

prove

ments

visibly

evident

23HKHAR9

23YO

female

traumatic

headinjury

(atage13)

with

chronic

residual

cognitive

deficits

and

sleepdisturbance

.Short

term

memory

andword

retrieva

lproblems,

impulsivity,difficulty

conce

ntratin

g

Partialrighttemporal

lobectomyand

ventriculostomy.

AdderallXR

20mgand

Provigil200mgdaily,

helpedlittle

with

hyp

ersomnolence

.Paxil

20mgsince

age16.

Imipramine50mgthen

20mgdaily

Glasg

ow

Comascale

3on

admissionto

ER,in

comafor6.5

weeks.

Base

lineQEEG

shows

excess

Beta

andHigh

beta,mostly

leftfrontal.

Loreta

confirmsleft

frontalandtemporal

invo

lvementat9–25Hz

20

sessions

F3=C

3=T

4=

C430minutesonce

=weekso

metim

estwice=

week.

Counse

ling,

Vita

min

B2added

after7se

ssions,

significa

nt

improve

mentreportedby

patie

ntandmotherin

memory

andenvironmental

awareness.After20

sessions,

reportedsignifica

nt

improve

mentin

sleep

patterns,

able

tofallasleep

faster,andreturn

tosleepif

shewoke

upin

middle

of

night.increase

dawareness

ofco

gnitive

deficitand‘‘lost

time’’.

INcrease

dawareness

offuture,andability

toplan

foreduca

tionalactivities.

StartedB2,headach

es

decrease

d.

Post-treatm

ent

QEEG

forthco

ming.Contin

uing

weeklyse

ssions,

reports

steadyim

prove

mentof

memory

andco

gnition

24JW

NK

6YO

F,ch

ronic

anxiety,

difficulty

with

anger

control,reading

difficulty;left-handed

noprevioustreatm

ent

QEEG

showsexcess

slow

1–10HzatT5,O2,T3,

T6,T4,F8,Fp1,Fp2.

Excess

21–30HzatF3,

C3,P3,O1,Pz,

P4.

Coherence

deficitdelta

5L,1R,theta

3L,1R,

andalpha1R.

53se

ssions3x=

week,

variousplace

ments

No

longeranxious.

No

further

angeroutbursts.Readingat

gradeleve

l.Im

prove

dsh

ort-term

memory.

Improve

dfocu

singability.

AllQEEG

abnorm

alities

norm

alized.

Note.QEEG

Maps:

19-channelEEG

reco

rdingsanalyze

dusingNeuroGuide(AppliedNeurosc

ience

).Othmerrefers

tothelow-frequency

bipolartrainingperOthmer(2008).

LZT!LiveZ-Sco

retext

displays

onlivesc

reenoronstatisticalsummaries;

QEEG!quantitativeEEG;IVA!IntegratedVisuala

ndAudito

ryPerform

ance

Test;SSRI!

serotonin

re-uptake

inhibito

rs;NF!neurofeedback

;TOVA!Test

ofVariablesofAttentio

n.

44


the overshoot phenomenon and allowed thefeedback to be more effectively targeted.

Once the use of separate upper andlower Z-score limits became the practice,abreactions or negative side effects havenot been reported from the field, where thegeneral clinical population is involved.Results from our early clinical reports andtrials thus far suggest that well-targetedEEG normalization does not appear to havesignificant downside risk, when the pretreat-ment EEG is initially clearly abnormal. Mildabreactions have been observed by oneauthor (TFC) during training demonstra-tions on some individuals, in which theprominent EEG deviations appeared topotentially represent coping or compensa-tory mechanisms, and the Z-score traininghas the effect of reducing the deviations.Such deviations may be in amplitude or inconnectivity, or in both. For example,chronic pain sufferers may present with glo-bally decreased alpha power. We hypothe-size that this may reflect a state of tensionand abnormal activation of the cortex repre-senting some kind of coping mechanism, andthus, when alpha activity is uptrained, it mayresult in an increased experience of pain asthe coping mechanism is reduced. Whetheror not this is an abreaction is a matter of ter-minology, as restoration of sensory aware-ness, including pain, might be used as apath toward self-regulation and recovery.

As another example, clients with chronicanxiety may exhibit excess alpha, whichmay be a coping mechanism, or may simplyreflect their individual state of activation,particularly where emotional control andregulatory centers are involved. Again,downtraining the alpha, which is an acti-vation procedure, may result in increasedperception of anxiety, even as the EEGnormalizes. Another example arises whennormal, functional, achieving adults presentwith what the Z-score software interprets as‘‘excess SMR.’’ This excess is not necessarilyabnormal, may simply reflect an above-average ability to sit still and remain motion-less, or may reflect the normal onset ofdrowsiness. This variant is in fact commonlyseen in clinical professionals, for whom still-ness and attentiveness are traits that are

cultivated and nurtured, and in trainingworkshops, in which drowsiness may appearunder normal circumstances. In some indivi-duals in these circumstances, downtrainingthe SMR has been seen to result in a feelingof irritation and uneasiness, secondary to theactivation of the trained areas.

DISCUSSION AND CONCLUSIONS

It has been seen that the LZT trainingused here is capable of inducing brainchanges that are specific and profound,particularly with regard to whole-brainactivation and connectivity. Using this tech-nique in conjunction with QEEG and beha-vioral data, it is possible to demonstrateclinical effects that are well correlated withobjective measures, and support the claimthat this approach is an important additionto clinical practice.

It has been found that four-channel LZTtraining is sufficient to resolve globalconnectivity issues and that it can effectivelytarget abnormalities visible on the LOR-ETA, and to resolve them. This is likelybecause the brain has limited degrees of free-dom, and in order to bring a predominanceof parameters into the normal range, otherparameters must also normalize. That is,when sufficiently constrained, the brain can-not conspire to ‘‘circumvent’’ the training,and produce untoward effects.

Nonetheless, when using the MVPapproach, it is found that the brain is pro-vided with information that is particularlyvaluable. By ignoring ‘‘outliers,’’ the braincan concentrate on fundamental mechan-isms, without being distracted by details thatmay confound the training. If only somefraction of Z-scores are required to fallwithin a target for rewards, then the trainee’sEEG is given a large dynamic range withinwhich to function. By using smaller targets,and allowing some Z-scores to remain out-side the defined range, the brain is providedwith options, that it appears to be preparedto use to best advantage.

With LZT training, the brain is exploringits dynamic range, and this is key to theeffectiveness. It is broadening its functional

Clinical Corner 45


repertoire, and is finding a trajectory andpath toward normalization that is not astraight line through state space. It is a cir-cuitous path, but it is a path that the brainseems to be equipped to navigate. Everyperson may respond differently, but LZTtrainees receive concise information withwhich to develop and implement a strategytoward self-regulation.

Again, because the technique using MVPallows the extreme deviations to beuntouched, you are allowing the brain to cre-ate its own strategy toward normalization. Itis significant that clients may have clinicalbenefits uniformly through the treatment. Itis also important to vary the target size andthe percentage of Z-scores required, so thatthe brain has full information to explorethese boundaries, without requiring fullnormalization from the start.

We have found that four channels is actu-ally very effective at localizing and trainingthe entire brain, and some of our publishedresults show the whole-head EEG beingessentially normalized, as a result ofjudicious choice of the four channels. Mostoften, montages such as F3=F4=P3=P4 orF3=F3=C3=C4 are used. There is also a‘‘big box’’ that can be used, which is F7=F8=T5=T6. When four channels are well cho-sen, the brain does not have a lot of room tomove around. We do not generally see pro-blems due to the fact that the four channelshave missed anything. It is possible to ident-ify certain montages that appear to isolatefunctional hubs and subsystems. Theseprovide additional focus and meaning forplacements for four-channel training. Forexample, posterior integration issues associa-ted with stress or aging are well addressed byusing C3=C4=P3=P4.

We believe that assessing the client’sclinical signs and ‘‘complaints’’ is essentialto planning and carrying out the LZT train-ing. It is possible to more flexibly addressvarious brain areas quickly when the chan-nels can be quickly changed, as with aMINI-Q, or when using a full 19-channelcap with LZT training. However, the econ-omy and convenience of applying fourmonopolar leads provides benefits of sim-plicity. It is fair to say that four-channel

LZT training is being proven and that it isa robust and effective method.

REFERENCES

Collura, T. F. (2008a, April). Whole-head normaliza-tion using live Z-scores for connectivity training(Part 1). NeuroConnections, 12–18.

Collura, T. F. (2008b, July). Whole-head normaliza-tion using live Z-scores for connectivity training(Part 2). NeuroConnections, 9–12.

Collura, T. F. (2009). Neuronal dynamics in relation tonormative electroencephalography assessment andtraining. Biofeedback, 36, 134–139.

Collura, T. F., & Thatcher, R. W. (2006, April 29).Real-time EEG Z-score training, realities and pro-spects. Bedford, OH: BrainMaster Technologies,Inc, and Applied Neurosciences, Inc.

Collura, T. F., Thatcher, R. W., Smith, M. L.,Lambos, W. A., & Stark, C. A. (2009). EEG bio-feedback training using live Z-scores and a norma-tive database. In T. Budzynski, H. Budzynski, J.Evans, & A. Abarbanal (Eds.), Introduction toquantitative EEG and neurofeedback (2nd ed., pp.103–141). Amsterdam: Elsevier.

DuRousseau, D. (2007, September). Simultaneoustraining of slow cortical potentials & coherence inchildren with AD=HD using realtime Z-score neuro-feedback: A preliminary study. Paper presented atthe annual meeting of the International Societyfor Neurofeedback and Research.

Rutter, P. (2009, April). Reconnecting our lost childrento the world – Z-score training and autistic disorder.Paper presented at the 40th annual meeting of theAssociation for Applied Psychophysiology and Bio-feedback, Albuquerque, NM.

Smith, M. L. (2008, April). A father finds a solution:Z-score training. NeuroConnections, 22–25.

Stark, C. (2008, April). Consistent dynamic Z-scorepatterns observed during Z-score training sessions.NeuroConnections, 37–38.

Thatcher, R. W. (1998). EEG normative databases andEEGbiofeedback. Journal ofNeurotherapy, 2(4), 8–39.

Thatcher, R. W. (1999) EEG database guided neu-rotherapy. In J. R. Evans & A. Abarbanal (Eds.),Introduction to quantitative EEG and neurofeedback(pp. 72–93). San Diego, CA: Academic.

Thatcher, R. W. (2004) Z-score EEG biofeedback—Technical foundations. St. Petersburg, FL: AppliedNeurosciences, Inc.

Wigton, N. (2008, August–September). Does Z-scoreNF work better than non Z-score NF? Paperpresented at the 16th Annual International Societyfor Neurofeedback and Research Conference, SanAntonio, TX.



Documents

LRIHHGEDFN &DVH 6WXGLHV 8VLQJ /LYH = 6FRUH 7UDLQLQJ …brainm.com/software/pubs/brain/BiofeedbackCaseStudy2.pdf · EEG Biofeedback Case Studies Using Live Z-Score Training and a Normative