Journal of Neurotherapy: Investigations in , Neurofeedback and Applied Neuroscience The Neurophysiology of Dyslexia: A Selective Review with Implications for Neurofeedback Remediation and Results of Treatment in Twelve Consecutive Patients Jonathan E. Walker MD a & Charles A. Norman PhD b a Neurotherapy Center of Dallas , TX b Department of Educational Psychology , University of Alberta , Canada Published online: 08 Sep 2008.

To cite this article: Jonathan E. Walker MD & Charles A. Norman PhD (2006) The Neurophysiology of Dyslexia: A Selective Review with Implications for Neurofeedback Remediation and Results of Treatment in Twelve Consecutive Patients, Journal of Neurotherapy: Investigations in Neuromodulation, Neurofeedback and Applied Neuroscience, 10:1, 45-55

To link to this article: http://dx.doi.org/10.1300/J184v10n01_04

PLEASE SCROLL DOWN FOR ARTICLE

© International Society for Neurofeedback and Research (ISNR), all rights reserved. This article (the “Article”) may be accessed online from ISNR at no charge. The Article may be viewed online, stored in electronic or physical form, or archived for research, teaching, and private study purposes. The Article may be archived in public libraries or university libraries at the direction of said public library or university library. Any other reproduction of the Article for redistribution, sale, resale, loan, sublicensing, systematic supply, or other distribution, including both physical and electronic reproduction for such purposes, is expressly forbidden. Preparing or reproducing derivative works of this article is expressly forbidden. ISNR makes no representation or warranty as to the accuracy or completeness of any content in the Article. From 1995 to 2013 the Journal of Neurotherapy was the official publication of ISNR (www. Isnr.org); on April 27, 2016 ISNR acquired the journal from Taylor & Francis Group, LLC. In 2014, ISNR established its official open-access journal NeuroRegulation (ISSN: 2373-0587; www.neuroregulation.org).

THIS OPEN-ACCESS CONTENT MADE POSSIBLE BY THESE GENEROUS SPONSORS

CURRENT CONCEPTS IN NEUROTHERAPY

Articles appearing in “Current Concepts” advance hypotheses, descriptions, and reviews of techniquesimportant to clinicalneurotherapy. The techniquesdescribed are not necessarilysup- ported by clinical research, and opinions expressed regarding the effectiveness or efficacies of these techniques are solely those of the authors.

The Neurophysiology of Dyslexia: A Selective Review with Implications for Neurofeedback Remediation and Results of Treatment in Twelve Consecutive Patients

Jonathan E. Walker, MD Charles A. Norman, PhD

ABSTRACT. Dyslexia is a common and important problem in all industrial societies, with a prev- alence rate of five to ten percent, for which no consistently effective treatment is available. Recent advances in imaging (morphometric MRI, functional MRI, PET, regional cerebral blood flow), as well as in neurophysiology (evoked potentials, QEEG, event-related desynchronization, coher- ence studies, magnetic source imaging, reading difference topography) have clarified our under- standing of the normal circuitry involved in reading and differences seen in individuals who have trouble learning to read. These studies have important implications for the use of neurofeedback to help dyslexic individuals learn to read more easily. First, we obtain a QEEG and a reading differ- ence topograph. We then train down any abnormalities that are significantly increased and train up any abnormalities that are significantly decreased. Increasing 16-18 Hz activity at T3 (left mid-temporal area) has also proved quite helpful in improving reading speed and comprehension. These combined approaches have been helpful in all cases of dyslexia we have treated, dramati- cally so in some cases. Each of the 12 individuals treated improved by at least two grade levels af- ter 30 to 35 sessions. doi:10.1300/J184v10n01_04

Jonathan E. Walker is affiliated with the Neurotherapy Center of Dallas, TX. Charles A. Norman is affiliated with the University of Alberta, Canada, Department of Educational Psychol- ogy. Address correspondence to: Jonathan E. Walker, 12870 Hillcrest Road, Suite 201, Dallas, TX 75230 (E-mail: [email protected]). Journal of Neurotherapy, Vol. 10(1) 2006 Copyright © 2006 ISNR. All rights reserved. doi:10.1300/J184v10n01_04 45 46 JOURNAL OF NEUROTHERAPY

KEYWORDS. Dyslexia, QEEG, neurofeedback, coherence training

INTRODUCTION the method and timing of testing. Baseline data is rarely available. Many studies indicate an as- Dyslexia is a common problem (Watson & sociation of word comprehension (encompass- Williams, 1995) and recent studies (Bhatngar, ing phonemic, lexical, and semantic represen- Mandybur, Buckingham, & Andy, 2000; tations) with the left superior temporal . Leisman, 2002; Simos, Breier, et al., 2002) Lesions restricted to this area tend to produce have shown that there is considerable perfor- pure word deafness, leaving lexical, semantic, mance variability on various tasks used to re- orthographic,phonological,andsyntacticknowl- fine the diagnosis. Most dyslexics perform edgeintact(Tanaka,Yomodori,&Mori,1987). poorlyontasksofprocessingandremembering Lesions of the left (ante- symbols and some also perform poorly on vi- rior to T3) are associated with deficits in spell- sual processing tasks, so-called double dys- ing (Roeltgen & Heilman, 1984) and verbal lexia. Rapid naming may also be affected and short-term memory (Vallar & Shallice, 1990). anotherclassificationdescribesproblemsinsu- Lesions affecting the left middle temporal perficialversusdeepbraintractsinvolvedinthe gyrus and the (near T3) reading process (Damasio & Geschwind, and the inferior temporal gyrus (inferior to T3) 1984).Dyslexiaisatermprimarilyusedbyneu- frequently disrupt performance on semantic rologists; educators describe reading difficulty tasks (Alexander, Hiltbrunner, & Fischer, 1989). instead. Traditional tests for dyslexia include Lesions of the (anterior to psychometric testing or computerized testing O1) produce memory encoding difficulties ac- (Greene, 1996). companying semantic task performance (Rudge Past research has shown that the left superior & Warrington, 1991). temporal gyrus plays an important role in inte- grating auditory, visual, perceptual and mem- Cortical Stimulation and Language ory inputs in order to accomplish fluent read- ing.Adultpatientswithlesionsinthisarea(e.g., Bhatnagar et al. (2000) reviewed previous strokes) have deficits in auditory comprehen- studies involving cortical stimulation and sion and a general disruption in language pro- cessing (Strub & Black, 1986). Evoked poten- added three cases of their own, using flexible tial studies have also shown abnormalities in grids of electrodes to stimulate the brain after this area in dyslexic individuals (Duffy, the patients were out of surgery and in a more Denckla, Bartels, Sandini, & Kiessling, 1980). natural environment. There was considerable Most studies using quantitative EEG have not variation between the three patients. Stimula- found a consistent abnormality in the resting tion of multiple sites in the frontal, temporal, eyes-closed state. In a recent study Walker and and parietal lobes was associated with difficul- Norman (2004) found a linear increase in beta ties in comprehension and expression of lan- activity at T3 (roughly overlying the superior guage. The Bhatnagar et al. (2004) suggest that sylvian gyrus) with progressive increases in there is a multi-layered arrangement. Some reading difficulty. This area may be vitally im- language functions are overlaid in a common portant in reading comprehension, especially peri-sylvian region that is involved in to se- as reading difficulty increases. quencing of motor movements, verbal mem- ory,andphonemicidentification.Itistheorized Lesion Deficit Correlation that the anterior serves syntactic functions while the posterior temporal lobe is Damasio and Geschwind (1984) reviewed associated with naming functions and the fron- studies of lesion deficits in understanding dys- tal lobe is associated with motor speech func- lexia.Unfortunately,there are numerous lesion tions. Multiple frontal, parietal, and temporal variables (extent, location, and distribution sites were implicated in errors of syntax overlap) and detection of deficits varies with construction. Current Concepts in Neurotherapy 47

Morphometric Studies erentially activated by letter strings versus faces or visual textures (Puce, Allison, Asgari, Foster, Hynd, Morgan, and Hugdahl (2002) Gore, & McCarthy, 1996). Some studies sug- have reviewed studies showing that 60 to 70 gest this may be involved in differentiating fa- percentofnormalindividualshaveanasymme- miliarlettersfromnon-sense(unfamiliar)char- try of the , the left being acters (Pugh et al., 1996). Stimulation by larger. Many dyslexicpersons have eithersym- orthographic versus non-orthographic activa- metric plana or the right is larger. The superior tion also occurs in the left inferior frontal area surfaces of the temporal lobes are symmetric in (F7;Fiez&Petersen,1998)andtheleftanterior dyslexic individuals, but larger on the left in insula (deep temporal, no surface representa- controls (Kushch, Gross-Glenn, Jallad, & tion; Price, Wise, & Frackowiak, 1996). Lubs, 1993). Pennington et al. (1999) found A hypothesis suggesting that dyslexics have people with reading disabilities to have smaller a specific impairment within the visual magno- insulae (deep temporal) and anterior superior cellular pathway (Stein & Walsh, 1997) has neocortices (roughly F7/F8) but larger retro- been largely discredited (Amitay, Ben-Yehudah, callosal cortex (anterior to O1) when compared Banai, & Ahissar, 2002). Over-activation of tocontrols.Theseabnormalitiesoccurredbilat- Broca’s area in dyslexics may represent in- erally. It should be noted that the anterior supe- creased effort in performing language related rior cortex includes Broca’s area. Decreased tasks (Shaywitz et al., 1998). Pugh et al. (2000) graymatterdensitywasfoundinthelefttempo- summarized the neuroimaging literature on ral lobe involving superior (above T3), middle normal reading in dyslexia. They propose that (T3), inferior (below T3), and mesial (no sur- in normally developing readers, a dorsal left face representation) structures in dyslexics posterior temporo-parietal system predomi- (Brown et al., 2001). Studies on the corpus cal- nates at first and is associated with the analytical losum have found variable and inconsistent ab- processing necessary for learning to integrate normalities. Subjects with reading difficulty the orthographic features with phonological exhibited decreased diffusion and anisotropy and lexical-semantic features of printed words. bilaterallyinthetemporal/parietalwhitematter Later, a posterior ventral occipital/temporal (Klingberg et al., 2000). These abnormalities circuit comes to constitute a fast word-identifi- may correlate with fewer connections between cation system underlying fluent word recogni- various language specialized areas. tion in skilled readers. These two posterior systems are functionally disrupted in develop- Functional MRI Studies mental dyslexia. Reading disabled persons demonstrate higher reliance than normals on Posterior ventral areas of the superior tem- bilateral frontal inferior (F7, F8) and right pos- poral gyrus (above T3) are preferentially acti- terior hemisphere (P4) regions to compensate vated by speech sounds and speech versus si- for left posterior deficits. lent presentations, more on the left than on the A remediation program focused on auditory right in fMRI studies (Binder et al., 1977). processing and oral language training im- These posterior ventralsuperior temporalareas proved oral language and reading performance are probably involved more with phoneme rec- and also increased activation toward normal in ognition and grapheme-phoneme translation the lefttemporo-parietalcortexand leftinferior than with lexical-semantic processes. Differ- frontalgyrus of dyslexics(Templeetal.,2003). ences between words and non-words are de- This study suggests that hypoactivation of tected in neighboring regions, including the these areas is a crucial determinant of dyslexia. temporal poles (F7/F8), the middle temporal However, several other areas showed an in- gyrus (T3), the inferior temporal gyrus (below crease in activity after remediation (bilateral T3) and the (between T5 and P3; cingulate gyrus, left hippocampal gyrus, left Binder & Price, 2002). , right pre-/posterior cing- Primary visual areas are not differentially ulate, right parietal-occipital , and bilat- activated by words or pseudo-words. The left eral hypothalamus). Furthermore, no correla- lateral extrastriate (lateral to O1) region is pref- tion was found (these areas are not near the 48 JOURNAL OF NEUROTHERAPY

surfaceareasof the10 /20 system)betweenim- Electrophysiological Imaging proved reading scores and activation of the left of Brain Function in Dyslexia temporo-parietal region or the inferior frontal region. Nobre and McCarthy (1995) reviewed evoked potential studies in normal word recog- Positron Emission Tomography Studies nition. is often impaired in (PET) dyslexia, and evoked potential studies are only used for studying discrete events. Quantitative Gross-Glenn, Duara, Barker, and Loewenstein EEG, on the other hand, may be used in combi- (1991) have made a case for involvement of the nationwithrecallofreadmaterialstodetermine cerebellum in developmental dyslexia, the which areas of the brain are involved in both so-called “dyslexia automatization hypothe- word recognition and comprehension. Most sis.” They found classical signs of cerebellar QEEGstudieshaveshown noconsistentdiffer- dysfunction in dyslexic children, including ence between dyslexics and controls in the dystonia,discoordination,dysequilibrium,and eyes-closed state as expected, but some changes have been seen with activation (i.e., reading decreased muscle tone. PET studies indicated during acquisition of the QEEG). Flynn and decreased activation in the right cerebellar Deering (1989) found no difference between hemisphereandvermisin80percentofthesub- dysphonetic dyslexics (phonologically im- jects when learning a button-press sequence paired) and control subjects, but dyseidetic (vi- compared to controls. They proposed that the sually impaired) children showed a marked in- reduced quality of an articulatory representa- creaseinlefttemporalandparietalthetaactivity tion might lead to impaired sensitivity to onset, during reading and spelling tasks. Flynn, rhyming, and the phonemic structure of lan- Deering, Goldstein, and Rahbar (1992) failed guage, leading to deficits in phonological to confirm the theta activation but found re- awareness and in naming speed (the “double duced beta activity bilaterally in the dyseidetic deficit” hypothesis). Other PET studies in dys- children and decreased right parietal/occipital lexics have shown bilateral hyperactivation of beta activity in the dysphonetic children. the lingual gyrus (medial occipital) with rela- Walker and Norman (2004) hypothesized that tive right frontal (F4) hypo-activation (Hagman easy reading would not require activation of et al., 1992). Another study found a bilateral in- “reading areas,” but progressively difficult crease in metabolismin the medial temporal re- reading would. This proved true at T3 (roughly gions (not projecting to the cortical surface; overlying the ), but no Rumsey et al., 1992). Other studies showed other areas were activated by difficult reading. failure of activation in left posterior temporal This suggests that training the T3 area to acti- (near T3) and inferior parietal (between P3 and vateduringdifficultreadingtasksmightbeuse- ful in making children with dyslexia easier to T5) regions in dyslexic adults (Paulesu et al., teach and help them to catch up with their 1996). Other studies supported a disconnection classmates, using intensive remedial instruc- between posterior and anterior language areas tion. (Horwitz, Rumsey, & Donohue 1998; Rumsey Event-related desynchronization has also et al., 1999). beenused tostudydyslexia.RipponandBruns- wick (2000) found a global absence of task-re- Regional Cerebral Flow lated alpha amplitudereduction for both a read- ing task and a picture completion task. There An important study points to the left angular was also a marked asymmetry (right greater gyrus (between T5 and P3) as the most likely thanleft)inbetaactivityintheparietal/occipital site of functional impairment in dyslexic adults regions, again in both tasks, in the dyslexic (Ramsey et al., 1999). Higher blood flow is as- group. Dyslexics had an increase in frontal sociatedwithbetterreadingskillincontrolsand thetawith the phonologicaltasks, whereas con- lower blood flow is associated with poor read- trol subjects exhibited a decrease. Both dys- ing in dyslexics. lexicandcontrolsubjectsshowedatask-related Current Concepts in Neurotherapy 49

reduction in frontal theta with the visual task. Magnetic Source Imaging (MEG) Improvements in cognitive skill have been as- sociated with reduction in theta, and Orekhova, Simos, Fletcher, et al. (2002) found de- Stroganova, and Posikera (1999) have sug- creased activation of the left temporo-parietal gested that higher levels of theta may be evi- region in dyslexics during a visual pseudo- denceof less activetask engagement(i.e., more word rhyme-matching task. After an intensive “idling”). 80-hour intervention that produced significant Another approach has been to look at coher- improvement in reading skills their scores on ence between brain areas in dyslexia. Leisman basic word reading tests increased into the nor- mal range. They observed increased activation (2002) found that controls have significantly in the left posterior superior temporal gyrus of greater coherence in the 1 to 30 Hz range be- the dyslexic children. Salmelin, Service, Kiesila, tween the hemispheres at several homologous Uutela,andSalonen(1996)describedimpaired sites compared to dyslexics. Dyslexics demon- processing of word forms in adultdyslexicsub- stratedsignificantlygreatercoherenceat1to30 jects.Aleftinferiortemporaloccipitalarea(be- Hz within hemispheres compared to normal. low T3) was differentially engaged in the nor- He suggested that developmental dyslexia may mal subjects as early as 180 milliseconds after represent a functional hemispheric disconnec- word presentation. Dyslexic subjects either tion syndrome. Pugh et al. (2000) found a dis- failed to activate this area or showed a slowly ruption in functional connectivity between the increasing late response. Letter-string specific leftangulargyrusandrelatedoccipitalandtem- responses peaking around 150 milliseconds af- poral sites, but only on tasks requiring phono- ter presentation in the left inferior occipital logic assembly and not on print tasks. Evans temporal cortex in fluent readers were unde- (1996) also found decreased coherence be- tectable (no surface representation) in dyslexic tween multiple left posterior hemisphere sites subjects.Childrenwithdyslexiaexhibitedlittle or no activation of the left posterior superior (P3 > T5 > T3 > O1) in dyslexic subjects, with temporal gyrus (close to T5) compared to con- several different combinations being noted in trols. Activation was increased in the right pos- 70 percent of the dyslexic subjects. terior superior temporal gyrus compared to Duffy et al. (1980) and Duffy, Denckla, control subjects. After a two-month interven- McAnulty, and Holmes (1988) are the only tion that produced significant improvement in group that have reported quantitative evoked readingskills,activityintheposteriorleftsupe- potentials (QEPs) in dyslexic subjects. They rior temporal gyrus increased by several orders found decreased responsiveness to auditory of magnitude, of magnetic flux, in each partici- stimuli (non-verbal) in the middle posterior pant (Simos, Fletcher, et al., 2002). portions of the left (between T3 and T5) tempo- rallobes.Whenaverbaltaskwasusedasastim- Suggestions for Neurofeedback Training ulus (count “TYKE” versus “TIGHT” presen- Based upon Cortical Stimulation and tations), more posterior regions were affected Imaging Studies including a prominent right posterior occipital/ temporal region. They found that they could Moststudiesindicateacentralroleforthesu- discriminate between anomic, dysphonemic, perior temporal gyrus and certain adjacent ar- and global types using several tasks and found eas (in and around T3) in normal reading com- prehension with impairment in one or more of differences in both QEEG and QEPs using a these areas in persons with dyslexia. Beta 2 ac- discriminant involving several such tasks. No tivity (15-18 Hz) is an indicator that the area is significant differences were found in the left engaged in the reading process (Walker & Nor- parietal/occipitalareas,suggestingthatalldys- man, 2004). With an easy learning task, this lexics are impaired in this area. Differences area will only generate slightly more beta 2 ac- found in the subtypes may reflect compensa- tivity. With progressively more difficult read- tory activation of areas not normally required ing tasks more beta 2 activity will be generated for reading rather than a pathological change. by that area, but if the task becomes too diffi- 50 JOURNAL OF NEUROTHERAPY

cult, the area may disengage and stop generat- tion of anxiety and would need to be down- ingbeta2activity(Walker&Norman,2004). trained. Then coherence abnormalities are By training such a critical brain area to func- trainedtonormalization.Thefirst12patientsin tion normally (i.e., train that area to be quiet at our clinic reported an improvement of at least rest, become active with an easy learning task, two grade levels, as judged by their reading and to become increasingly more active with teachers, in reading speed and comprehension more difficult learning tasks) reading skills using this protocol (30 to 35 ten-minute ses- should improve. Higher frequency beta (21-30 sions each). Results are presented in Case I, Hz) may well be dysfunctional and an indica- Case II and Table 1.

TABLE 1. Effect of Neurofeedback in Improving Reading Level in 10 Additional Cases

Pre-Neurofeedback Post-Neurofeedback Case Age Grade Reading Grade Neurofeedback Protocols Reading Grade Level (5 sessions each) Level 31610 9 ↓ 2-7 Hz/↑ 12-15 Hz at FP2 12 ↓ 1-8 Hz plus ↓ 18-30 Hz at OZ ↓ coherence of beta at P3/O1 ↓ coherence of beta at FP2/O2 ↑ coherence of delta at F3/O1 ↑ coherence of theta at C4/P4 ↑ coherence of delta at F4/O2

4104 1 ↓ 2-7 Hz/↑ 15-18 Hz at FP1 4 ↓ 1-5 Hz at O2 ↓ 1-8 Hz at PZ while reading ↑ coherence of beta at F4/O2 ↓ coherence of theta at F3/P3

5115 2 ↓ 1-4 Hz at O2 5 ↓ 1-5 Hz at F8 ↑ 15-18 Hz at T3 while reading ↓ 10-12 Hz at O1

6 7 1 Pre-K ↓ 4 Hz at F7 2 ↓ 5-6 Hz at F8 ↓ 29-30 Hz at FZ ↓ coherence of beta at O1/T3 ↓ 1-10 Hz at F3 while reading ↓ 2-7 Hz/↑ 15-18 Hz at C3

7115 4 ↓ 2-7 Hz at F3 plus F4 10 ↓ 2-7 Hz at O1 plus O2 Current Concepts in Neurotherapy 51

Pre-Neurofeedback Post-Neurofeedback Chart Age Grade Reading Grade Neurofeedback Protocols Reading Grade Level (5 sessions each) Level ↑ coherence of beta at F7/F4 ↑ coherence of theta at C4/T4 ↓ coherence of delta at O1/F3 ↓ 2-7 Hz/↑ 15-18 Hz at PZ ↑ coherence of theta at PZ/O2 ↓ coherence of theta at T3/C3

8126 5 ↓ 2-7 Hz/↑ 15-18 Hz at T5 7 ↓ 2-7 Hz/↑ 15-18 Hz at O2 ↓ 2-7 Hz/↑ 15-18 Hz at O1

993 2 ↓ 2-7 Hz/↑ 12-15 Hz at C4 4 ↓ 2-7 Hz/↑ 15-18 Hz at O2 ↓ 2-7 Hz/↑ 15-18 Hz at O1 ↓ 2-7 Hz/↑ 15-18 Hz at P4

10 10 5 4 ↓ 1-4 Hz at CZ 7 ↓ 1-2 Hz/↑ 9-11 Hz at O1 ↓ 1-11 Hz at FP1 while reading ↑ coherence of beta at T3/T4 ↓ coherence of alpha at F1/T3 ↓ coherence of delta at T5/P3 ↑ coherence of theta at C3/T3

11 10 3 2 ↓ 2-7 Hz/↑ 15-18 Hz at C3 4 ↓ 1 Hz/↑15-18 Hz at T4 (gifted and talented) ↓ 1-10 Hz at F3 while reading ↓ 30 Hz at F8 ↓ coherence of delta at F7/T5 ↓ coherence of alpha at O1/T3 ↓ coherence of theta at F3/T5

12 7 2 1 ↓ 1-8 Hz/↑ 15-18 Hz at CZ 3 ↓ coherence of alpha at F7/F3 ↑ coherence of delta at CZ/PZ ↑ coherence of alpha at CZ/C4 ↑ 11-12 Hz at O2 52 JOURNAL OF NEUROTHERAPY

Future efforts to improve our treatments will termittent depression. He was in a special include enhancing our understanding and use school, in the ninth grade, but reading at a fifth of reading difference topographies. These are grade level. His initial QEEG revealed an in- obtained by mapping baseline eyes open data crease in the relative power of theta at FP1, F3, and subtracting that data from maps obtained F4, FZ, and CZ. There was an elevated theta/ while reading moderately difficult material. beta ratio at FZ and CZ. Reading difference to- We can then train the person while reading to pography revealed a diffuse increase in fre- decrease slow or “idling” rhythms (4-10Hz) quencies 1 to 10 while reading, especially and increase faster rhythms (15-18Hz) which marked in the occipital areas in the 1-2 Hz bins are more appropriate for paying attention and (normal readers usually decrease the amount of learning. 1-10 Hz diffusely while reading). Coherence During analysis of the QEEG, we use raw analysis revealed increases in the coherence of singleHz differencemapsbetweenreadingand delta at F8/T6 and at F4/O2. eyes open conditions. These maps show how He completed 30 sessions of neurofeedback the EEG changes when the person reads. We with 5 sessions of each of the following proto- compute these difference maps with NeuroRep cols: (Hudspeth, 2004) by subtracting the eyes open raw numbers from the task raw numbers. If the Based on task voltage is higher than the eyes open base- 1. Decrease 1-7 Hz Eyes open QEEG line, the remainder will be positive and it will at CZ. indicate how much the EEG voltage increased the when person read. If the remainder is nega- 2. Decrease 2-7 Hz Eyes open QEEG tive, the maps will show now much the voltage and increase decreasedwhilereading.WeuseNeuroRepbe- 15-18 Hz at CZ. cause all the single Hz topographs are created 3. Decrease 1-10 Hz Readingdifference withthesamecolorscaleso itis easytolookata at O1 while reading. topography page of topographs and see where the largest 4. Increase 11-12 Hz Eyes open QEEG changes happen. The results from NeuroRep at O1. giveonlyabsolutedifferences.Inordertogetan 5. Decrease coherence Eyes open QEEG idea of percent change, we compare the change of delta at F8/T6. in voltage in the difference topographs to the eyesopenbaseline.Forexample,a2micro-volt 6. Decrease coherence Eyes open QEEG changeinthesingleHzdifferencesisimportant of delta at T4/O2. ifthebaselineEEGvoltageis2.7uV.However, a 2 uV difference is not very important if the Following the 30 neurofeedback sessions, baselinevoltageis7uV.NeuroGuide(Thatcher, he was reading at grade level (Grade 10). He 1998) can make single Hz difference maps as was no longer having memory problems and well. It has the advantage of calculating the was not depressed, by self-report. A repeat maps as percent change maps. However, QEEG showed normalization of the reading NeuroGuide scales each map with a different difference topograph with a decrease in fre- color scale so we cannot look at a page and see quencies 1 to 10 in the reading condition. The where the biggest changes occur. SKIL (Sterman coherenceabnormalitieswerealsonormalized. & Kaiser, 2005) may also be able to calculate There was still mild slowing (in the theta band) and create different topographs as well. at F4, C3, C4, P4, O1, O2, and PZ. A new find- ing was an increase in beta (15-18 Hz) Two Cases Reports Illustrating fronto-polar,frontalandcentralregionsbilater- the Application of These Findings ally. No new symptoms were noted. to Persons with Dyslexia Case II. This 9-year-old girl had the chief com- plaints of dyslexia, spelling difficulty and bad Case I. A 15-year-old boy presented with a handwriting. She was in the fourth grade, but chief complaint of dyslexia. He also com- was reading at first grade level and hated read- plainedofshort-termmemorydifficultyandin- ing. The initial QEEG showed an elevated Current Concepts in Neurotherapy 53 theta/beta ratio, maximal at C3/P3. Reading involved in reading and of the abnormalities difference topography showed a diffuse in- leading to dyslexia. QEEG is the only modality crease in the 1 to 10 Hz frequencies in the read- that indicates the state of connectivity between ing condition. There was an increase in the these critical areas. We describe an approach to intrahemispheric coherence of alpha at T4/T6, correcting the abnormalities of magnitude and a decrease of beta coherence at CZ/PZ and an coherence which characterize dyslexic states. increase in delta coherence at F1/F7. She com- Our preliminary results suggest that this ap- pleted 53 neurofeedback sessions as follows: proach is likely to result in improved reading Based on skills over relatively short time periods (a few 1. Decrease 2-8 Hz Eyes open QEEG weeks to a few months). and increase 15-18 Hz at CZ (10 sessions). REFERENCES 2. Decrease 2-8 Hz Eyes open QEEG Alexander, M. P., Hiltbrunner, B., & Fischer, R. S. and increase (1989). Distributed anatomy of transcortical sensory 15-18 Hz at P3 . Archives of Neurology, 46, 885-892. (10 sessions). Amitay, S., Ben-Yehudah, G., Banai, K., & Ahissar, M. (2002). Disabled readers suffer from visual and audi- 3. Decrease 1-10 Hz Reading difference tory impairments but not from a specific magno- at O1 while reading topographies cellular deficit. Brain, 125, 2272-2285. (10 sessions). Bhatnagar, S. C., Mandybur, G. T., Buckingham, H. W., 4. Decrease coherence Eyes open QEEG & Andy, O. J. (2000). Language representation in the : Evidence for cortical mapping. Brain of alpha at T4/T6 & Language, 74, 238-259. (5 sessions). Binder, J. R., Frost, J. A., Hammeke, T. A., Cox, R. W., 5. Decrease coherence Eyes open QEEG Rao, S. M., & Prieto, T. (1997). Human brain lan- of beta at CZ/PZ guage areas identified by functional magnetic reso- (5 sessions). nance imaging. Journal of Neuroscience. 17, 353-362. Binder, J. R., & Price, C. J. (2002). Functional neuro- 6. Decrease coherence Eyes open QEEG imaging of language. In R. Cabeza & A. Kingstone of delta at FP1/F7 (Eds.), Handbook of functional neuroimaging of cog- (5 sessions). nition (pp. 202 - 224). Boston MA: MIT Press. Brown, W. E., Eliez, S., Menon, V., Rumsey, J. M., 7. Increase 11-12 Hz Eyes open QEEG White, C. D., & Reiss, A. L. (2001). Preliminary evi- at O2 (8 sessions). dence of widespread morphological variations of the brain in dyslexia. Neurology, 56, 781-783. Followingthetrainingshewasreadingabout Damasio, A., & Geschwind, N. (1984). The neurologi- grade level (Grade 6), as judged by her reading cal basis of language. Annual Review of Neurosci- ence, 7, 127-147. teacher. She was reading for pleasure for the Duffy, F. H., Denckla, M. B., Bartels, P. H., Sandini, G., first time. Her handwriting and spelling were & Kiessling, L. S. (1980). Dyslexia: Automated di- much improved. agnosis by computerized classification of brain elec- The follow up QEEG revealed normaliza- trical activity. Annals of Neurology, 7, 421-428. tion of all of the abnormalities trained. Reading Duffy,F.H.,Denckla,B.D.,McAnulty,G.B.,&Holmes, difference topographies revealed the normal J. A. (1988). Neurophysiological studies in dyslexia. In F. Plum (Ed.), Language, communication, and the decrease in 9 to 10 Hz activity in the reading brain (pp. 60, 149-170). New York: Raven Press. condition, though the occipital 4 to 7 Hz fre- Evans, J. R. (1996). Quantitative EEG abnormalities in a quencies and F7 and F8 1 to 2 Hz frequencies sample of dyslexic persons. Journal of Neurotherapy, were still increased during reading (better, but 2 (1), 1-5. not normal). No new abnormalities were seen. Fiez, J. A., & Petersen, S. E. (1998). Neuroimaging Our experience with 10 additional patients is studies of word reading. Proceedings of the National Academy of Sciences of the United States of America, summarized in Table 1. 95, 914-921. Recent advances in neuroimaging have led Flynn, J. M., & Deering, W. M. (1989). Subtypes of dys- toanimprovedunderstandingofthebrainareas lexia: Investigation of Boder’s system using quanti- 54 JOURNAL OF NEUROTHERAPY

tative neurophysiology. Developmental Medicine and Pennington B. F., Filipek, P. A., Lefly, D., Churchwell, Child Neurology, 31, 215-223. J., Kennedy, D. N., Simon, J. H., et al. (1999). Brain Flynn, J. M., Deering, W. M., Goldstein, M., & Rahbar, morphometry in reading-disabled twins. Neurology, M. H. (1992). Electrophysiological correlates of dys- 5, 723-729. lexic subtypes. Journal of Learning Disabilities, 25, Price, C. J., Wise, R. J., & Frackowiak, R. S. (1996). 133-141. Demonstrating the implicit processing of visually Foster, L. M., Hynd, G. W., Morgan, A. E., & Hugdahl, presented words and pseudowords. , K. (2002). Planum temporale asymmetry and year 6, 662-670. advantage in dichotic listening in developmental dys- Puce, A., Allison, T., Asgari, M., Gore, J .C., & McCar- lexia and Attention-Deficit/Hyperactivity Disorder thy, G. (1996). Differential sensitivity of human vi- (ADHD). Journal of the International Neuropsy- sual cortex to faces, letter strings, and textures: A chological Society, 8, 22-36. functional magnetic resonance imaging study. Jour- Greene, J. F. (1996). Psycholinguistic assessment: The nal of Neuroscience, 16, 5205-5215. clinical base for identification of Dyslexia. Topics in Pugh, K. R., Mencl, W., Einar, J., Annette, R., Katz, L., Language Disorders, 16, 45-72. Frost, S. J., et al. (2000). Functional neuroimaging Gross-Glenn, K., Duara, R., Barker, W., & Loewenstein, studies of reading and reading disability (develop- D. (1991). Positron emission tomographic studies mental dyslexia). Mental Retardation & Develop- during serial word-reading by normal and dyslexic mental Disabilities Research Reviews, 6, 207-213. adults. Journal of Clinical and Experimental Neuro- Pugh, K. R., Shaywitz, B. A., Shaywitz, S. E., Consta- psychology, 13, 531-544. ble, R. T., Skudlarski, P., Fulbright, R. K., et al. Hagman, J. O., Wood, F., Buchsbaum, M. S., Tallal, P., (1996). Cerebral organization of component processes Flowers, L., & Katz, W. (1992). Cerebral brain me- in reading. Brain, 119, 1221-1238. tabolism in adult dyslexic subjects assessed with Rippon, G., & Brunswick, N. (2000). Trait and state positron emission tomography during performance EEG indices of information processing in develop- of an auditory task. Archives of Neurology, 49, mental dyslexia. International Journal of Psycho- physiology 36 734-739. , , 251-265. Horwitz, B., Rumsey, J. M., & Donohue, B. C. (1998). Roeltgen, D. P., & Heilman, K. M. (1984). Lexical agraphia: Further support for the two-system hypoth- Functional connectivity of the angular gyrus in nor- esis of linguistic aphasia. Brain, 107, 811-827. mal reading and dyslexia. Proceedings of the Na- Rudge, P., & Warrington, E. K. (1991). Selective im- tional Academy of Sciences of the United States of pairment of memory and in splenial America, 95, 8939-8944. tumours. Brain, 114, 349-360. Hudspeth, W. J. (2004). NeuroRep EEG Report Soft- Rumsey, J. M., Andreason, P., Zametkin, A. J., Aquino, ware Manual. Reno, NV: , Inc. T., King, A. C., Hamburger, S. D., et al. (1992). Fail- Klingberg, T., Hedehus, M., Temple, E., Salz, T., Gabrieli, ure to activate the left temporoparietal cortex in dys- J. D., Moseley, M. E., et al. (2000). Microstructure of lexia. An oxygen 15 positron emission tomographic temporo-parietal white matter as a basis for reading study. Archives of Neurology, 49, 527-534. ability: Evidence from diffusion tensor magnetic res- Rumsey. J. M., Horwitz, B., Donohue, B. C., Nace, K. onance imaging. , 25, 493-500. L., Maisog, J. M., & Andreason, P. (1999). A func- Kushch, A., Gross-Glenn, K., Jallad, B., & Lubs, H. tional lesion in developmental dyslexia: Left angular (1993). Temporal lobe surface area measurements on gyral blood flow predicts severity. Brain & Lan- MRI in normal and dyslexic readers. Neuropsy- guage, 70, 187-204. chologia, 31, 811-821. Salmelin, R., Service, E., Kiesila, P., Uutela, K., & Leisman, G. (2002). Coherence of hemispheric function Salonen, O. (1996). Impaired visual word processing in developmental dyslexia. Brain and Cognition, 48, in dyslexia revealed with magneto-encephalography. 425-431. Annals of Neurology, 40, 157-162. Nobre, A. C., & McCarthy, G. (1995). Language-related Shaywitz, S. E., Shaywitz, S. E., Pugh, K. R., Fulbright, field potentials in the anterior-mediated temporal R. K., Constable, R. T., Mencl, W. E., et al. (1998). lobe: Effects of word type and semantic priming. Functional disruption in the organization of the brain Journal of Neurosciences, 15, 1090-1098. for reading in dyslexia. Proceedings of the National Orekhova, E. V., Stroganova, T. A., & Posikera, I. N. Academy of Sciences of the United States of America, (1999). Theta synchronization during sustained an- 95, 2636-2641. ticipatory attention in infants over the second half of Simos, P. G., Breier, J. I., Fletcher, J. M., Foorman, B. the first year of life. International Journal of Psycho- R., Bergman, E., Fishbeck, K., et al. (2002). Brain physiology, 32, 151-172. activation profiles in dyslexic children during non- Paulesu, E., Frith, U., Snowling, M., Gallagher, A., word reading: A magnetic source imaging study. Morton, J., Frackowiak, R. S., et al. (1996). Is devel- Neuroscience Letters, 290, 61-65. opmental dyslexia a disconnection syndrome? Evi- Simos, P. G., Fletcher, J. M., Bergman, E., Breier, J. I., dence from PET scanning. Brain, 119, 143-157. Foorman, B. R., Castillo, E. M., et al. (2002). Dys- Current Concepts in Neurotherapy 55

lexia-specific brain activation profile becomes nor- Thatcher, R. W. (1998). EEG normative databases and mal following successful remedial training. Neurology, EEG biofeedback. Journal of Neurotherapy, 2 (4), 58, 1203-1213. 8-39. Stein, J., & Walsh, V. (1997). To see but not to read: The Vallar, G., & Shallice, T. (1990). Neuropsychological magnocellular theory of dyslexia. Trends in Neuro- impairments of short-term memory. New York: Cam- science, 20, 147-152. bridge University Press. Sterman, M. B., & Kaiser, D. A. (2005). SKIL Topo- Walker, J. E., & Norman, C. A. (2004, August). Normal metric Software. Los Angeles, CA: Sterman-Kaiser adult readers recruit increasing beta power at T3 as Imaging Laboratory. reading difficulty increases. Presentation at the 12th Strub, R., & Black, F. W. (1986). Neurobehavioral dis- International Society for Neuronal Regulation Con- orders: A clinical approach (pp. 253-311). Philadel- ference, Ft. Lauderdale, FL. phia, PA: F. A. Davis. Watson, C. W., & Williams, D. M. (1995). Information Tanaka, Y., Yomodori, A., & Mori, E. (1987). Pure processing patterns in specific reading ability. Jour- word deafness following bilateral lesions. Brain, 110, nal of Learning Disabilities, 28, 216-231. 381-403. Temple, E., Deutsch, G. K., Poldrack, R. A., Miller, S. RECEIVED: 06/30/04 L., Tallal, P., Merzenich. M, M., et al. (2003). Neural deficits in children with dyslexia ameliorated by be- REVISED: 12/01/04 havioral remediation: Evidence from functional MRI. 04/22/04 Proceedings of the National Academy of Sciences of 01/25/05 the United States of America, 100, 2860-2865. ACCEPTED: 08/23/05 doi:10.1300/J184v10n01_04