Wake Forest University

Wake Forest University

Gallaudet University

Supported by the NIH and NSF

Georgetown University

What Brain Imaging Tells us about Reading Acquisition and Intervention

Guinevere EdenCenter for the Study of Learning

Reading is a Cultural Invention

First use of alphabet - 1800 B.C.

Divination Characters on Oracle BonesShang Dynasty 1400-1100 BC

Bolger et al. (2005)

Alphabetic Logographic

Reading has to be Learned – in Different Ways

Dehaene et al. (2010)

Learning to Read Changes the Brain

Petersson et al. (2007)

LDAH

Brain Imaging

Franz Gall (1758-1828) Phrenology

10 Second Cortical AnatomyTouch

Vision

Hearing

Motor

Dorsal pathway(magnocellular):motionlocationmental rotationspatial relationships

Ventral pathway (parvocellular):formcolorobject identificationface identity

Parallel Visual Pathways

Where

What

Left Hemisphere: Right Hemisphere:Word Form Area Face/Object Area

Kanwisher et al, 1997

Face Processing

Gauthier et al., 1999

Left Hemisphere: Right Hemisphere:Word Form Area Face/Object Area

● Reading Acquisition ● Precocious Reading ● Reading Disability

Learning to Read

• Phases of reading acquisition (Ehri, 1992)

– Pre-alphabetic- visual– Partial alphabetic- phonological cues– Full alphabetic- decoding– Consolidated Alphabetic- chunking,

analogy• Phonological processing abilities are

critical (Wagner and Torgesen, 1987)

Research Questions

• What is the neural basis of visual word

processing in healthy children?

• How does the neural basis of word

processing change during schooling?

• What is the relationship between these

neural systems and phonological skills?

Reading in Typical Children Simos et al., 2001

MEG (MSI)12 Children (8-15 years)16 Adults (23-28 years)Real word recognition

8 year old 28 year old

Increasing left inferior frontal gyrus

Lateralization of infero-temporal cortex

No change in left temporoparietal cortex

The neural basis of reading

• Left inferior frontal gyrus

• Left temporo-parietal cortex

• Left infero-temporal cortex





OrthographyDirect Lexical Access





Cross-modal integrationPhonological assembly

Semantics





SemanticsPhonological assembly

Center for the Study of Learning, Georgetown University

Implicit Word Processing

- =Adapted from Price et al, 1996

41 normal subjects

6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23Male Female

Task PerformanceWords vs. False Fonts

Accuracy Reaction Time

-15

-10

-5

0

5

10

15

20

6 8 10 12 14 16 18 20 22

Age (Years)

Acc

ura

cy D

iffe

ren

ce (

% c

orr

ect)

R2= .00 n.s.

-200

-150

-100

-50

50

100

150

200

6 8 10 12 14 16 18 20 22

Age (Years)

Rea

ctio

n T

ime

Dif

fere

nce

(m

s)

R2= .03 n.s.

0

6- 9.4 yn=13

9.4- 18 yn=13

20- 23 yn=15

Implicit Reading Activity

Turkeltaub et al. Nature Neuroscience, 2003

Developmental Changes in Activity

p< .001, peak p< .0001Reading Composite

fMR

I

Samuel Orton1925

“In the process of early visual education… the storage of memory images of letters and words occurs in both hemispheres…. the process of learning to read entails the elision from the focus of attention of the confusing memory images of the nondominant hemisphere”

Phonology and Reading

• Types of phonological processing (Wagner & Torgesen, 1987)

– Phonological awareness (LAC)– Phonological naming (RAN)– Working memory (Digit Span)

• Subtypes of dyslexia are associated with these types of phonology

• Are these abilities associated with different aspects of reading?

19981998 20002000

National Reading Panel, 2000: Five Component Skills Essential for reading

Phonemic AwarenessPhonemic Awareness

PhonicsPhonics

FluencyFluency

VocabularyVocabulary

Comprehension strategiesComprehension strategies

Identifying words accurately and fluently

Constructing meaning once words are identified

Phonological Awareness

Lindamood Auditory Conceptualization Test (LAC)

“Show me /p/ /t/ /p/”

“If this says ‘eth’,show me ‘ith’

Phonological Naming

Rapid Automatized Naming Test (RAN)

s a o d o p a p d os d a o a p s p d so p s p d o s a o pa d o p s p a s d sp o s d s p o a o d

s a o d o p a p d os d a o a p s p d so p s p d o s a o pa d o p s p a s d sp o s d s p o a o d

Working Memory

Digit Span

“3 8 2 4”“7 4 6 2 5”

“9 2 3 6 1 8”“5 3 8 2 7 4 6”

“2 5 4 3 2 8 9 4”

Correlations with Phonology

Phonological AwarenessPhonological RecodingWorking Memory

p< .005, peak p< .0005

Turkeltaub et al. Nature Neuroscience, 2003

Conclusions

Young Readers Phonology

Young readers activateleft temporoparietalcortex, related to phonological awareness

Conclusions

Young Readers Phonology

Reading Acquisition

Young readers activatetemporoparietal cortex, related to phonologicalawareness

Reading acquisition=

Right posterior cortex(nonlinguistic visual)

Left frontal & temporal(phonology, semantics)

Reading in Children Aged 6 to 9Chinese and American

Data provided by LH Tan Turkeltaub et al., 2003

The neural basis of precocious reading acquisition: fMRI case study

of hyperlexic reading

Hyperlexia

• Developmental disorder of communication (usually autism spectrum)

• Extremely precocious reading learned very early without explicit instruction

• Reading scores above expectation, with comprehension commensurate with verbal ability

• Incidence ≈ 2 / 10,000 (Burd et al., 1985, Yeargin- Allsopp, 2003)

Ethan

• 10-year-old boy• Disordered

– expressive/receptive language (first word at 3.5y)– social interaction– motor coordination

• Pervasive Developmental Disorder- Not Otherwise Specified

• Early intense interest in text• Precocious reading

Ethan’s Reading Scores

Age

5y-11m

9y-9m

Word I.D.

Age eq.

8y-10m

15y-1m

WordAttackAge eq.

9y-4m

16y-11m

GORTPassage

Age eq.

10.3

14.9

GORTComp.Age eq.

<7.9

12.1

Dyslexic vs. Hyperlexic Reading

Words Dyslexic Hyperlexic

11.1 yr 7.0 yr

therapeutic

bouquet

trivialities

Hyperlexia HypothesesLeft Hemisphere

Phonological AdvantageWelsh et al., 1987

Right HemisphereVisual Advantage

Cobrinik, 1982

Methods

• Same fMRI methods as cross sectional study

• Compared Ethan to two control groups– Age Matched (n=9)– Reading Matched (n=8)

Ethan- Implicit Reading

P< .005Turkeltaub et al., Neuron 2004

Ethan vs. ControlsLeft Hemisphere

AgeMatched

(n=9)

ReadingMatched

(n=8)

Correlations with Phonological Awareness

p< .005, peak p< .0005LAC

fMR

I

Ethan vs. ControlsRight Hemisphere

AgeMatched

(n=9)

ReadingMatched

(n=8)

Reading Acquisition

Turkeltaub et al. Neuron, 2004

Hyperlexia HypothesesLeft Hemisphere

Phonological AdvantageWelsh et al., 1987

Right HemisphereVisual Advantage

Cobrinik, 1982

Conclusions

• In contrast to single hemisphere theories, Ethan demonstrated both– Hyper-activity in left hemisphere

phonological areas– Increased activity in right hemisphere

visual areas

• Left temporoparietal cortex is hyper-active in hyperlexia

Reading Research

BasicResearch

TranslationalResearch

Policy and Education

The International Dyslexia Association / NICHD Research

Definition of Dyslexia

Dyslexia is a specific learning disability that is neurological in origin. It is characterized by difficulties with accurate and / or fluent word recognition and by poor spelling and decoding abilities...



… These difficulties typically result from a deficit in the phonological component of language that is often unexpected in relation to other cognitive abilities and the provision of effective classroom instruction…



… Secondary consequences may include problems in reading comprehension and reduced reading experience that can impede growth of vocabulary and background knowledge.

Developmental Dyslexia

Research•Behavioral Evidence•Anatomical and Physiological

evidence•Genetic evidence

Practice•Early identification•Remediation


• 15% of the population is Learning Disabled

• 75-80% or these are Dyslexic

• 5 to 12 % school aged children in the US have basic deficits in reading (IQ discrepant model versus low achievement)

• Males and females a not equally affected


• 74% of poor 3rd graders were still impaired when tested in high school

• “Compensated” adult dyslexics still demonstrate measurable underlying problems

• 60% of poor adult readers have undetected or untreated LD

• 50% of juvenile delinquents have undetected LD





Behavioral Evaluation of Dyslexia

Measurement:

• Single Word Reading

• Phonemic Awareness

• Automatic Naming Speed

• Sequential Working Memory

Behavioral Evaluation of Dyslexia

Measurement:

Single Word Reading

• Phonemic Awareness

• Automatic Naming Speed predictive

• Sequential Working Memory

Behavioral Manifestations of Developmental

Dyslexia• Phonological Processing

– Sound segmentation and manipulation

– Rapid name retrieval– Verbal working memory

Behavioral Manifestations of Developmental

Dyslexia• Phonological Processing

– Sound segmentation and manipulation

– Rapid name retrieval– Verbal working memory

• Sensorimotor Processing– Visual motion processing– Motor coordination– Auditory processing

Ramus, TINS 2004

Eden et al., Nature 1996

Visual Motion Perception: Controls versus Dyslexics

Demb et al.,1997 Proc.Nat.Acad.Sci.

Ramus, TINS 2004

Ramus 2004• Genetically driven focal cortical

abnormalities disrupt processes in the left hemisphere (phonological processing)

• Under certain hormonal conditions these disruptions propagate to the thalamus, leading to sensory impairment

• These disruptions also extend to posterior parietal cortex and cerebellum





Anatomical VariationsLeftLeft

RightRight

II

IIII

IIIIII

IVIV

VV

VIVI

Galaburda et al. 1985

Neural Basis of Dyslexia: Structural MRI Studies

Findings reported only once

Findings reported more than once

Eckert et al. 2004

Left Hemisphere White Matter Tracts and Dyslexia

• Reduced fractional anisotropy (FA) in L arcuate fasciculus.

• Positive correlation of L arcuate fasciculus FA with PA.

• Positive correlation of L inferior fronto-occipital fasciculus FA with orthographic processing.

Vandermosten et al. 2012

White Matter Studies of DyslexiaChildren

• Compared to Typical Readers– Reduced FA in left IFG (Rimrodt et al., 2010)– Reduced FA in left temporo-parietal WM

(Rimdodt et al., 2010)– Reduced FA of inferior-fronto occipital fasciculus

(Rollins et al., 2009)– Reduced FA of inferior longitudinal fasciculus

(Rollins et al., 2009)– Increased posterior corpus collosum (Hasan et

al., 2012)– Reduced left superior longitudinal fasciculus FA

(Carter et al., 2009)– Abnormal orientation of right superior longitudinal

fasciculus (Carter et al., 2009)

• Correlations – Positive for reading speed and left IFG FA

(Rimrodt et al., 2010)– Positive for posterior corpus collosum mean

diffusivity with word reading and comprehension (Hasan et al., 2012)

– Positive for superior corona radiata and single/pseudoword reading (Odegard et al., 2009)

– Positive for left temporo-parietal WM with reading scores (Niogi and McCandliss, 2006)

Adults• Compared to Typical Readers

– Reduced FA of left arcuate fasciculus (Vandermosten et al., 2012)

– Reduced FA in bilateral temporo-parietal WM (Klingberg et al., 2000)

– Reduced FA bilateral fronto-temporal WM (Steinbrink et al., 2008)

– Reduced FA left temporo-parietal WM (Steinbrink et al., 2008)

• Correlations– Positive for phonemic awareness and

speech perception with FA of left arcuate fasciculus (Vandermosten et al., 2012)

– Positive for left temporo-parietal WM and reading score (Klingberg et al., 2000)

Flowers, Wood, & Naylor, 1991

Typical Readers

Neurobiological Basis of Reading

Typical Readers Dyslexic Readers

Neurobiological Basis of Reading

fixate repeat delete

+ rat rat

rat at

fixation vocalization vocalization +

phonological manipulation

Task

Stimulus

Response

Processes

Phoneme Deletion TASK

Typical Readers: Deletion versus Repetition

left right

Dyslexic Readers: Deletion versus Repetition

left right

Group Comparison: Controls > Dyslexics

left right

Eden et al., Neuron 2004

Controls > Dyslexics

Dyslexia across cultures: same or different?

Same brain region less active in dyslexics during reading tasks in all countries

Paulesu et al., 2001





Genetic Basis of Dyslexia

• 75 - 100% concordance in monozygotic twins

• 35 - 40% concordance in first degree relatives

• Chromosomes 1, 2, 6, 15 and 18 are implicated in various studies





Early Identification

Predictive Assessment of Reading (PAR)

Source: Wake Forest University

• Letter and word recognition (correctly pronouncing printed letters and words)

• Phonemic awareness (recognition of specific sounds in a word)

• Naming speed (speed of naming of letters, digits or common objects that are already known)

Early Identification


Source: Wake Forest University

Accuracy of PAR Prediction

20

40

60

80

100

120

140

160

40 60 80 100 120 140 160

PAR Predicted Reading Score

Ac

tua

l WJ

-III

Re

ad

ing

S

co

re

Early Identification of Dyslexia

Comprehensive Test of Phonological Processing (CTOPP)

Early Literacy Screening (NCLD)

Fox in a Box


Phonological Awareness Literacy Screening (PALS)

Test of Word Reading Efficiency (TOWRE)

Test of Phonological Awareness (TOPA)

Texas Primary Reading Inventory

Source: Parenting a Struggling Reader, Hall & Moats





Remediation Approaches • Phonology and other language structures are explicitly and systematically taught

• Large amount of practice given & small group or one-one

• Use of enhancing techniques (multi-sensory techniques linking listening, speaking, reading, and writing)

Examples of Multisensory Remediation Approaches

Orton-Gillingham Sonday System

Alphabetic Phonics Lindamood-Bell

Slingerland Phonographix

Spaulding

Wilson Language

LANGUAGE!

Project Read

June and Samuel Orton

http://www.interdys.org/

Research and Science Policy

No Child Left Behind Act, Reading First, 2001

• Requires methods that are used in the classroom to teach children how to read to be based upon valid scientific findings

• Based on findings from the National Reading Panel: Meta-analysis of scientific studies on reading that met certain criteria

Institute of Educational SciencesUS Department of Education

Guidelines on how to evaluate whether an educational intervention is supported by rigorous evidence

Randomized controlled trials

+

Effective in two or more settings

= Strong Evidence

Pre-post studies do not comprise “strong” or even “possible evidence” - often produce erroneous results

Institute of Educational SciencesUS Department of Education

Coalition for Evidence-Based Policy

“There are a vast array of educational interventions that claim to improve educational outcomes and to be supported by evidence…

...introduced with great fanfare as being able to produce dramatic gains…

…yielding little in the way of positive and long lasting changes”

-

=

)(

After BeforeIntervention Intervention

Dyslexia-specific brain activation profile becomes normal following successful remedial training

Simos, Fletcher, et al. Neurology,2002

Pretesting Posttesting





Interval Between Measurements

P-Pretest Pre Post 1 year 2 year

75

80

85

90

95

LIPS

EP

Growth in Total Reading Skill Before, During, and Following Intensive

InterventionS

tand

ard

Sco

re

Torgesen, et al., 2001

Study Design• Assignment of individuals into different interventions

• Groups are equal in reading measures prior to the intervention

• Compare the two groups after intervention

Group1 Group 2

Pre intervention

A B

Post intervention

June and Samuel Orton

Adult Phonological Intervention Study

Subjects:• 20 Adults from Orton Center, recruited through Wake Forest University

Intervention:• 112.5 hours of Lindamood-Bell (over 8 weeks)

Before and after measures:

• Behavior: reading, phonemic awareness

• Physiology (fMRI): phonemic segmentation

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

16.0

18.0

20.0

Visual Imagery (SI)

Phonemic Awareness (TAAS)

Per

cen

t C

han

ges

Non-Intervention Group Intervention Group

*p < .05

*

*

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

16.0

18.0

20.0

Visual Imagery (SI)

Phonemic Awareness (TAAS)

Per

cen

t C

han

ges

Non-Intervention Group Intervention GroupNon-Intervention Group Intervention GroupNon-Intervention Group Intervention Group

*p < .05

*

*

Skills Targeted by Intervention

Visual Imagery (SI) Phonemic Awareness (TAAS)

-5.0

0.0

5.0

10.0

15.0

20.0

25.0

30.0

Non-Word Reading (WJWASS)

Phonemic Transfer Index (DST)

Per

cen

t C

han

ge

s


**

*p < .005

-5.0

0.0

5.0

10.0

15.0

20.0

25.0

30.0

Non-Word Reading (WJWASS)

Phonemic Transfer Index (DST)

Per

cen

t C

han

ge

s


**

*p < .005

Skills Supporting Reading

Non-Word Reading Phonemic Transfer Index(WJWASS) (DST)

-4.0

-2.0

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

Real Word Reading (WRAT)

Reading Accuracy (GORT)

Reading Rate (GORT)

Reading Comprehension(GORT)

Sta

nd

ard

Sco

re C

han

ges


*

*p < .05

-4.0

-2.0

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

Real Word Reading (WRAT)

Reading Accuracy (GORT)

Reading Rate (GORT)

Reading Comprehension(GORT)

Sta

nd

ard

Sco

re C

han

ges


*

*p < .05

Oral Reading Skills

Real Word Reading Reading Accuracy Reading Rate Reading(WRAT) (GORT) (GORT) Comprehension (GORT)

-

=

)(

After BeforeIntervention Intervention

Intervention No Intervention

Before

After

ANOVA Group x Day: Increases in Activity Following Intervention

left rightEden et al., Neuron 2004

ANOVA Group x Day: Increases in Activity Following Intervention

left rightEden et al., Neuron 2004

Conclusion

• After phonological intervention adults with dyslexia show increased activation in the left and right hemispheres.

• The right hemisphere areas are similar to those in the left hemisphere involved in phonological processing in good readers.

Increases in Activity Following Intervention Children

Left Inferior Frontal Gyrus: increases in children (average 9 years) N=32

Skills Targeted by Intervention

*** p<0.001, ** p<0.01, * p<0.05

***

**

Reading Skills

*** p<0.001, ** p<0.01, * p<0.05

***

***

***

Skills Supporting Reading

*** p<0.001, ** p<0.01, * p<0.05

*

What else changes in the brain?

What’s the matter GRAY MATTER?

Draganski et al., Nature 2004

Research Design

• 11 dyslexic children: reading intervention followed by no instruction

• Anatomical scans obtained at each time point for analysis of gray matter volume

Reading Intervention No Intervention

SCAN1 SCAN 2 SCAN 3

8 weeks 8 weeks

Behavioral Changes After Intervention

Gray Matter Volume Increases After Intervention

Right Cerebellum

Left Fusiform/Hippocampus Right

Hippocampus

Left Precuneus

Percent Change in Gray Matter Volume

•Regions know to be involved in the processing of information from multiple sensory modalities are also involved in reading.

• The neurobiological representation of reading is established early on.

Overall Summary

•Dyslexic individuals show under- activity in these regions, especially parietal cortex.

•Brain activity changes following intensive remediation. Adults and children exhibit different patterns.

•Structural changes are also observed following intervention.

Overall Summary

Who will have the greatest reading gains after the intervention?

Those who show brain activity in inferior parietal cortex prior to the intervention. Activity here is predictive of the amount of reading improvement.





John AgnewKate CappellEmily CurranEmma ColeIain DeWittErin EinbinderLynn GareauKaren JonesDaniel KooAnthony KrafnickJoe MaisogMartha MirandaAlison MerikangasCorinna MooreEileen NapolielloOlumide OluladeJenni RosenbergPeter TurkeltaubRobert TwomeyJohn VanMeter

Wake Forest UniversityLynn FlowersFrank WoodDebi Hill

Gallaudet UniversityCarol LaSassoKelly Crain

Supported by NSF, NICHD, NIDCD, NIMH

The 63rd International Dyslexia Association Annual Conference

http://www.interdys.org/

http://csl.georgetown.edu

Wake Forest University

Documents

neural systems

right hemisphere

alphabetic logographicreading

temporal cortexno change

phonological skills

process of learning

healthy children

confusing memory images