Your SlideShare is downloading. ×
Upcoming SlideShare
Loading in...5

Thanks for flagging this SlideShare!

Oops! An error has occurred.

Saving this for later? Get the SlideShare app to save on your phone or tablet. Read anywhere, anytime – even offline.
Text the download link to your phone
Standard text messaging rates apply



Published on

  • Be the first to comment

  • Be the first to like this

No Downloads
Total Views
On Slideshare
From Embeds
Number of Embeds
Embeds 0
No embeds

Report content
Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

No notes for slide
  • That is, whether it has to do with input or output
  • Could be sound system organization or the mental representation of phonemes
    Output has been described as a problem with motor programming or prearticulatory sequencing, but either way this neurological in nature and not an issue with oral musculature
    There is a debate whether the deficit is at the input or output level
    We can test if it’s at the input level – i.e. a representational deficit – using an MMN / EEG approach Shriberg, Aram and Kwiatkowski, 1997
    The prearticulatory sequrencing deficit refers to motor speech programming processes( Neuromuscular) whereas the others refer to cognitive linguistic or phonological or planning( Neurological)
    Shriberg and Aram (1997)n have proposed that five primary theoretical perspectives dominate the DAS literature
    Input processing deficit- suggested by Tallal 7 colleagues: auditory temporal or perceptual memorial processes: That children with DAS are less able to analyze temporal patterns as stimuli increased in rate. OR that they have “ waeker auditoy memory traces”
    Organizational processing disorder: suggests that the deficits stem from an underdevopled representations of the language phonemes or words. Dogil, Mayer and Vollmer (1994) proposed that adults with AOS have overspecified speech sounds, rather than underspecified speech sounds based on the non-linear framework.
    Output processing deficit: They have difficulty selecting and retrieving the appropriate phonemes to produce speech. many studies showed evidence against this perspective.
    Motor programming disorder: They have difficulty in motor planning speech productions. .compared to unplanned spontaneous productions. There difficult in motor control oif speech productions. I is apparent in planning complex movements during speech and other motor activities
    Prearticulatory sequencing disorder: difficulty to voluntarily progrma necessary movements for speech at the rearticulatory level in the absence of neuromascular impairements. The child knows what he or she want to say, he/she identifies the correct phonemes, yet despite normal physical structures (mouth, lios), she is for some reason unable to translate the instrcutions to her vocal articulators to make the correct souds.
    Assessment process
    Diagnosis: for families
    A clinician’s theoretical perspective regarding the etiology of DAS will dictate the assessemnt protocol used by the clinician and the treatment approach used by her/him.
    There are reporting in the literarture that there is 80% rate of false positive dignosis ( meaning that 80% are miscoreectly diagnosed with DAS)
    Uptil now, most of the traertments conducted to children with suspected DAS focus on motor speech eventhough ity is clear that the deficits layt higher ( at least to me
  • If it is neurological in nature, it makes sense to look at it with neurophysiological measures
  • In this study, we focused on a specific ERP, MMN
    Sharma et al looked at MMN within and between sound categories spanning a phonetic boundary – bababababa with different VOTs, vs ba da
    Mismatch negativity (MMN)- preattentive negativity peaking at 100-200 ms after stimulus onset in frontol-central sensors elicited by infrequent acoustic or phonetic changes in a sequence of a repetitive auditory stimulus or features of the auditory stimulu (Dehaene-Lambertz,1997; Näätänen et al, 1993)
    MMN reflects auditory discrimination as it indexes the underlying neural process (Näätänen, 1992; Näätänen & Winkler, 1999).
    MMN has been found to be elicited/enhanced for pre-existing traces in the brain (Cowan et al, 1993) . Hence, MMN can be a useful neurophysiological tool to examine the suggestion of overspecification of speech sounds in CAS.
    They found response to the acoustic differences but no greater response to the phonetic difference – but we have some contra evidence to that, as other researchers have also found
    Naatanen (2000) : MMN data indicate that each sound, both speech and non-speech, develops its neural representation corresponding to the percept of this sound in the neurophysiological substrate of auditory sensory memory. The accuracy of this representation, determining the accuracy of the discrimination between different sounds, can be probed with MMN separately for any auditory feature or stimulus type such as phonemes. Furthermore, MMN data show that the perception of phonemes, and probably also of larger linguistic units (syllables and words), is based on language-specific phonetic traces developed in the posterior part of the left-hemisphere auditory cortex. These traces serve as recognition models for the corresponding speech sounds in listening to speech.
  • Minimal pair phonemic
  • Deviant
    These sounds were pre-existing recordings which we adapted so that they were matched for length and frequency. The VOT for /pa/ is 50 msec, and the VOT for the deviant /ba/ is 5 msec. The overall duration of both sounds is 500 msec, and we manipulated the sounds digitally so that the voiced part of each is identical in terms of frequency and intensity. In the experiment, 290 stimuli: 17% deviants, very typical for an oddball paradigm.
  • Age difference is n.s.
  • So we subjected our data to a standard analysis paradigm, the steps are shown here and we can go through them later for anyone who is intertested. But mainly, it shows the processes of segmentation of the continuous EEG recoridng into sections of data related to the onset of each stimulus. We averaged together the responses to each condition in the experiment, and that gives us an Event Realted Potential, or ERP. Looking at the ERP over the appropriate sensors for registering the MMN, this is what we found.
  • Red = average response to the standard stimulus /pa/
    Blue= average response to the deviant stimulus /ba.
    Black = difference wave, we get by subtracting standard minus the deviant
    If you like right here, the average response to the deviant is more positive than the standard
    and right here the average response to the deviant is more negative than the average response to the standard. This is our MMN,
    Significant: 0-100, 100-200, 300-400
    P1 effects often go hand in hand with MMN effects – P1 is an early sensory response reflecting very preliminary categorization
    Significant negativity 300-400ms
  • Phonemic condition: No significant negativity to phonemic deviants
    A significant positivity to the standards might reflect P responding in the group
    But there is NO SIGNIFICANT MMN here – abnormal response, totally contrasts with the control group
  • 1. The use of ERP may help determine whether there are differences in the perceptual and cognitive processes involved in the discrimination of native and nonnative contrasts for children with CAS
    2. This is one way we can get information about the relationship between phonetic perception and language outcomes for children with CAS; behavioral measures don’t let us directly examine the organization and representation of speech sounds
  • Transcript

    1. Childhood Apraxia of Speech (CAS): Neurophysiological evidence of phonological involvement Reem Khamis Dakwar, Melissa Randazzo & Karen Froud ASHA 2008
    2. Background • No single validated list of diagnostic features of CAS (Shriberg et al., 1997) • Approximately 50 different features used by SLPs in the field (Forrest, 2003) • It is debated in the literature whether CAS is specifically a linguistic impairment (Crary, 1984) or a motor planning impairment (Edwards, 1984)
    3. Theoretical Perspectives • Input processing deficit • Organization or Representation • Output processing deficit • Motor programming or Prearticulatory sequencing deficit Shriberg, Aram and Kwiatkowski, 1997
    4. EEG • Noninvasive method for measuring electrical potentials generated by brain activity • High temporal resolution ( useful for looking at language) • Experiments can be done without overt participation from subjects ( useful for working with children) • EEG is a measure of continuous brain activation (voltages) • To look at event-related activations (ERPs), we average together segments of the continuous recording time-locked to stimuli
    5. Mismatch negativity • A negative deflection in the ERP • The brain’s “automatic change-detection response” (Näätänen et al., 1997) • Present large numbers of a frequent stimulus (ta ta ta ta ta) • Plus an infrequent stimulus (“oddball”) (ta ta ta ta da ta ) • MMN: negativity peaking around 150-250 milliseconds after the oddball is presented • MMN is completely automatic, does not require (conscious) attention • Reflects formation of traces in auditory sensory memory (Näätänen, 2001) • Can use this mismatch to study language- specific phonetic representations (Conboy et al, 2008) Picture adapted from Sharma et al., 1993
    6. Aim of the study • To compare ERPs to standard and deviant syllables between children with CAS and typically developing controls • Standard: /pa/ • Phonemic deviant: /ba/
    7. pa Stimuli: standard VOT: 50msec Duration: 500msec
    8. ba Stimuli: phonemic contrast VOT: 5msec Duration: 500msec
    9. Pilot study participants CAS • n=5 (1 male, 4 females) • Mean age = 6.8 years • All right-handed • Diagnosis of CAS by neurologist or SLP Control • n=5 (3 males, 2 females) • Mean age = 7.3 years • All right-handed • 1 control was a sibling of a CAS participant
    10. Experiment • Continuous EEG recordings were made while children watched an age appropriate movie of their choosing • Sounds were presented through earphones at a comfortable listening level • 128-channel Geodesic Sensor nets, digitally sampled via a high- impedance amplifier at 250Hz, impedances maintained below 40kΩ
    11. Data analysis procedures Raw EEG data Select sections of data based on onset of stimuli Remove noise from data Detect and remove artifacts Average over all remaining trials for each participant Combine averaged results from other participants May re-reference, baseline correct, remontage, average channels together For each condition in the experiment: segmentation Noise reduction Notch filter High / low pass filters Bad channel replacement Reject artifact trials Averaged Event Related Potential (ERP) Grand- averaged Event Related Potential (ERP) Look at time course Look at ERP distribution Look at time course Look at ERP distribution Statistical comparison with other conditions
    12. Results: Controls ** ** ** p <0.05
    13. Results: Apraxic group ** p <0.05 **
    14. Discussion • Phonological representation in CAS – Pilot results provide preliminary evidence from a neurophysiological standpoint that there is phonological involvement in CAS • EEG is an effective method for investigating neural correlates of phonological representation • Early phonetic perception and language outcomes • Need for further studies – Longitudinal studies of children with CAS – Subgroups of children with mild vs severe CAS – Other differences in speech perception – e.g. allophonic contrasts
    15. Thank you Neurocognition of Language Lab Teachers College, Columbia University 525 West 120th Street, New York, NY 10027 /
    16. References Crary, M.A. (1984). A neurolinguistic perspective on developmental dyspraxia, Communicative Disorders, 9 (3), 33-49. Edwards,M. (1984). Disorders of Articulation. New York: Springer-Verlag. Forrest, K. (2003). Diagnostic criteria of developmental apraxia of speech used by speech-language pathologists. American Journal of Speech-Language Pathology, 12, 376-380. Sharma, A., Kraus, N., McGee, T., Carrell, T., and Nicol, T.(1993). Acoustic versus phonetic representation of speech as reflected by the mismatch-negativity event related potential. Electroencephalographic Clinical Neurophysiology, 88 (1), 64-71/ Shriberg, L.D., Aram, D.M., and Kwiatkowsi, J. (1997). Developmental Apraxia of Speech: I. Descriptive and theoretical perspectives. Journal of Speech, Language, and Hearing Research, 40, 273-285.