Stimulating language: Insights from TMS Joseph T. Devlin MSc Neuroscience, Language & Communication 16 November 2011.

Stimulating language: Insights from TMS

Joseph T. Devlin

MSc Neuroscience, Language &Communication

16 November 2011

Lack of animal models

“I say, is that a banana?”

Neurology of language

Intra-cortical mappingStrokes & Disease

MRI studies of brain structure

PET and MRI studies of brain function

Non-invasive methods

Magnusson & Stevens (1911-1912)

Magnetic stimulation elicited phosphenes

Magnetic stimulation

Magnetic stimulation

Barker, Jalinous & Freeston (1985). Lancet, 11, 1106-7.

Transcranial magnetic stimulation

TMS coil current

8kA

Magnetic field pulse2.5T

Rate of change of magnetic field

30kT/s

Induced tissue current

15mA/cm2

Induced electric field

500v/m

How does it work?

It’s fun -- really!

Speech arrest with TMS

Semantic decision(Synonyms?)

Sentencecompletion: Meaning

Categorisation(man-made?)

Meaning preferentially engages anterior, ventral Broca’s area

Broca’s area re-visited

Broca’s area re-visited

Phonological decision(Homophones?)

SentenceCompletion: Rhymes

Two syllables?

Sounds of words (or sentences) preferentially engage posterior, dorsal parts of Broca’s area

Common activations

BUT… both sound and meaning engage all of Broca’s area relative to low level baselines

Word pairs SentencesSingle words

Two possibilities

1. Necessary processing

quantitative difference

2. Incidental processing

qualitative difference

Semantic Phonological

rTMSNone

Rostral Caudal

Semantic

Phonological

Gough et al (2005). J Neuroscience

rTMSNone

rTMSNone

rTMSNone

Subdividing Broca’s area

Is there a double dissociation in LIFC for semantic and phonological processing?

Are both areas engaged by both types of

processing?

TMS Results

Rostral: -52, 35, -7 4 × 6cm

Caudal: -52, 15, 8 2 × 3cm

Mean distance in cortex of 2.3cm apart

Sites on scalp separated by 3.5cm, on average

MNI coordinates Relative to C-T line

Anatomic localisation

*

Single pulse TMS

Devlin et al (2003). J of Cognitive Neuroscience

NoTMS

Motor evoked potentials

Functional connectivity

Seyal et al. (1999). Clin Neurophysiol, 110(3), 424-429.

MEP magnitude in hand during reading

Before

+

Functional connectivity

Seyal et al. (1999). Clin Neurophysiol, 110(3), 424-429.

MEP magnitude in hand during reading

Before After

officer

Implications

Evolutionary link?

Or inexplicable link between hand gestures and language(most refined in Italian speakers)?

Actions and motor cortex

Buccino et al. (2005). Brain Res Cogn Brain Res, 24(3), 355-363.

He turned the key.

He kicked theball.

He forgot thename.

Speech comprehension

Watkins et al. (2003). Neuropsychologia, 41(8), 989-994.

Somatotopy of speech

Results

D’Ausilio et al. (2009). Current Biology, 19, 381-5.

Disrupting speech perception

Meister et al. (2007). Current Biology, 17, 1692-6.

TMS Results

Recovery from aphasia

L R

peri-lesionalactivation

Contralateral activation

Stimulating IFG in patients

Thiel et al. (2006). Brain Lang. 98(1): 57-65.

No effect

Pre-morbid differences?

Knecht et al. (2000). Brain, 123 ( Pt 1), 74-81.

Lateralisation in 324 normal adults by fTCD

Left Right

Laterality affects susceptibility

Knecht et al. (2002). Nat Neurosci, 5(7), 695-699.

Theraputic TMS?

Naeser et al. (2005) Neurocase, 11(3), 182-193; Naeser et al. (2005) Brain Lang, 93(1), 95-105

Long term enhancement

Relation to other methods

Walsh and Cowey (2000). Nat Revs Neurosci.

PET and TMS

Paus et al. (1997). J Neurosci.

EEG and TMS

Ilmoniemi et al. (1997). NeuroReport

Designing experiments for TMS

Design considerations

1. Type of stimulation

2. Choosing control conditions

3. Targeting stimulation

4. Choosing parameters

5. Ethical considerations

Design considerations

1. Type of stimulation

2. Choosing control conditions

3. Targeting stimulation

4. Choosing parameters

5. Ethical considerations

Choosing a type of TMS

On-line stimulation occurs while the subject performs a task and the effects last for approximately the duration of stimulation.

Eg: Virtual lesionsChronometricsFunctional connectivity

Off-line stimulation occurs without a task and the length of effect is typically measured in minutes.

Eg: 1Hz stimulationTheta burst

Repetitive or chronometric?

Repetitive stimulation typically involves trains of three or more pulses evenly spaced in time

• Effect lasts approx. duration of stimulation

• Don’t need to know exactly when to stimulate

• Lots of pulses

Chronometric studies use either single or paired-pulses to examine the processing time course in a region

• Requires far more trials!!!• Subjects tolerate stimulation

better• How to best order trials?

Single pulse TMS

NoTMS

Ordering timing trials

No TMS

Design considerations

1. Type of stimulation

2. Choosing control conditions• Control sites• Control tasks• Control stimuli• Sham stimulation

3. Targeting stimulation

4. Choosing parameters

5. Ethical considerations

Control site: Vertex

Choosing another control site

Control task(s)

Sham TMS…

…is a sham

Design considerations

1. Type of stimulation

2. Choosing control conditions

3. Targeting stimulation• Functional localizers• Anatomically guided: MRI based stereotaxy• Heuristics

4. Choosing parameters

5. Ethical considerations

Ventral occipito-temporal cortex

Area sensitive to reading words

Functionally localize w/ fMRI

Rostral siteTask: Same category?

potato+

turnip

Caudal siteTask: Rhyme?

vein+

pane

41 ms* 52 ms*

Functionally localize w/ TMS

Identifying corresponding positions on the subject and subject’s MRI scan for registration

Frameless stereotaxy

Rostral: -52, 35, -7 4 × 6cm

Caudal: -52, 15, 8 2 × 3cm

Mean distance in cortex of 2.3cm apart

Sites on scalp separated by 3.5cm, on average

MNI coordinates Relative to C-T line

Scalp coordinates

International 10-20 system

Design considerations

1. Type of stimulation

2. Choosing control conditions

3. Targeting stimulation

4. Choosing parameters

5. Ethical considerations

Choosing parameters

• Stimulation intensity / duration / rate

• Inter-stimulation interval

• Type of coil

• Type of stimulation / stimulator

• Accessibility

• Number of trials per condition?

• Number of subjects in a study?

• Analysis methods?

Design considerations

1. Type of stimulation

2. Choosing control conditions

3. Targeting stimulation

4. Choosing parameters

5. Ethical considerations

Ethics of TMS

Although the risk is small, it is always present, so there is an obligation on the experimenter to always consider the value of a given experiment

• How can you minimize risk & discomfort?

• What is the minimal stimulation necessary?

• Is the TMS information clear and consent

informed?

• Are subjects always screened?

• Are the experimenters safety trained?

• Are emergency procedures clear & in place?

• Would YOU do this experiment?