Essay 36 Translational Display of! Neurophysiologic Investigations } in Communication Sciences and Disorders Reem Khamis-Dakwar As a new scholar in the field of communication sciences and disorders, it is unimaginable not to be touched and influenced by the prolific academic work of Dr. Singh. Even though I did not have the honor to meet him in person, I feel like his work has left remarkable contribution to our field. I am honored to be part of a project to commemorate and celebrate his rich fruitful life. CSD to elucidate brain processes underpinning speech-language disorders and language learning in specific communication disorders is exemplified by describing several clinically driven studies conducted at the Neurophysiology of Speech and Language Pathology Lab (NSLP Lab). Introduction Neurophysiologic investigation in Communication Sciences and Disorders (CSD) refers to studies on the nature of a specific communication disorder, language acquisition given a specific communication disorder, as well as the assessment and intervention effects for a specific communication disorder. In this essay, event-related potential (ERP) methodology will be introduced. Furthermore, the benefit and need for clinically driven neurophysiologic investigations in Neurophysiologic Investigations in CSD Cognitive Neuroscience, CSD, and EEG: Introduction Cognitive neuroscience refers to the scientific disciplines concerned with the study of higher cognitive functions in humans, such as memory and language, and their underlying neural bases. It aims to understand how cognitive functions, and their manifestations in behavior and subjective experience, arise
268 Translational Speech-Language Pathology and Audiology from the activity of the brain (Rugg, 1997, p.1). The field of CSD encompasses several scientific disciplines concerned with typical human communication, speech, and language development, communication disorders, and therapy. There are relatively limited clinically driven neurophysiologic studies in CSD to date. In this essay, we maintain the fields of cognitive neuroscience and CSD can inform each other and we exemplify the incorporation of EEG/ERP methodology to relate the observed communicative behaviors to their underlying brain mechanisms in the study of CSD. EEG and Language Correlates Several neuroimaging techniques can be utilized to investigate brain activity. These techniques can be divided into yielding either hemodynamic or electrophysiological measures. Hemodynamic measures, such as those resulting from fmri, identify brain activity resulting from metabolic changes, including blood flow. Electrophysiologic measures, such as those obtained in EEG and MEG, reflect cortical activity in different brain areas derived by recording electromagnetic activity across the scalp. Generally speaking, hemodynamic measures have superior spatial resolution (i.e., can more accurately measure the place of an examined brain activity), whereas measures of electromagnetic activity have better temporal resolution (i.e., can more accurately measure the timing of an examined brain activity). For example, using fmri techniques spatial information can be obtained on a millimeter scale, but temporal resolution can only be recorded on a second scale. In contrast, EEG can provide temporal information on a millisecond range (Luck, 2005) but spatial resolution within a wider millimeter scale. Hence, studies utilizing hemodynamic brain imaging techniques are better able to more accurately localize specific areas of the brain associated with a variety of language and/or cognitive functions, whereas studies utilizing electrophysiologic brain imaging techniques are more appropriate for identifying task-related neural activity at the millisecond level. Relative to these respective advantages, electrophysiologic measures are thought to be more effective for evaluating the subprocesses involved in language processing (Friederici, 2002). The ideal investigative approach would be to incorporate both imaging techniques, hemodynamic and electrophysiological, to obtain information with the highest temporal and spatial resolution. This essay is focused on introducing the use of EEG/ERP method in CSD (see Ward, 2010 for a review of imaging techniques). Electroencephalography (EEG) refers to the continuous measurement of electrical activity in the brain. To acquire EEG data, in the NSLPlab, we use a HydroCel Geodesic Sensor Net (HCGSN) from Electrical Geosdesic Inc. The net enables a quick and safe way to place 32 electrodes on the scalp of participants. Nets are available in different sizes allowing for a snug but comfortable fit for each participant. In the NSLPLab we are using HCGSN nets with 32 electrodes; however, the HCGSN nets are available in densities of 32, 64, 128, and 256 channels. Figure 36 1 shows a medium 32-channel HCGSN net as viewed from the front. Each pedestal in the net contains an electrode embedded within a sponge, which decreases any potential discomfort a participant may feel. In addition, the wires connecting individual electrodes form a net enabling the whole collection of electrodes (the net) to be quickly fitted. During the EEG recording session, participants wear the HCGSN net which is connected to an amplifier. The amplified electrical activity is recorded via EEG acquisition software. Figure 36 2 shows the data acquisition setup. On the right side of the picture, you
Translational Display of Neurophysiologic Investigations in Communication Sciences and Disorders 269 of interest, so that we get the brain responses only for the event of interest (the experimental condition). For example, if we are examining an individual s brain responses to English phonemic contrasts, and we are presenting two types of stimuli (/pa/ and /ba/), one standard step in analyzing the EEG data in this type of study is segmenting the EEG recordings that are linked to the presentation of the event of interest in the study, in this case the presentation of /pa/ and /ba/ as shown in Figure 36 3. We also average the recorded segmented responses over all the different presentations for the same condition to derive the ERP. There are several types of ERP components indicative of different cognitive processes. In this essay we will focus only on introducing the main correlates/erp components associated with language. Figure 36 1. Medium-sized HCGSN (32 channel). can see the participant wearing a HydroCel net (Electrical Geodesics, Inc.) with the 32 electrodes embedded in sponges. On the left side of the picture, you can see the data display of the amplified electrical activity from each of the 32 channels as it is recorded during an EEG experiment. In EEG studies of language, we are less interested in the brain s ongoing electrical recordings (see Figure 36 2) as might be tested in a clinical setting, but more in examining the brain s responses as they are elicited with respect to a specific experimental event presentation (e.g., hearing a phonemic contrast). To do that, an offline analysis of the recorded data is performed to derive the event related potential (ERP) which is the average electrical activity correlated with the specific type of stimulus presentation. The analyses to derive the ERP responses begins with segmentation of the EEG recordings into epochs which are segments of time linked to the event Mismatch Negativity (MMN) Phonological Representation and Processing Correlate MMN refers to a negative preattentional frontocentral ERP component peaking around 150 to 250 milliseconds elicited by the presence of an oddball sound in a sequence of repetitive auditory stimuli. In the MMN task an automatic prediction of the central auditory system is violated, such as in presenting a deviant /ta/ within a train of a frequent stimulus /da, da, da, da/. Experimental studies reveal that the MMN component is associated with sensory-memory updating and change/ rule violation detection (Näätänen, Kujala, & Winkler, 2010). Several studies revealed that MMN responses to deviants that constitute language-specific phonological contrasts were greater in native speakers of that language than in control participants who do not speak that particular language (for a review see
Electical activity recording Figure 36 2. A participant in an EEG study wearing a HCGSN 32-channel net. Participant wearing a hydrocel net 270
Translational Display of Neurophysiologic Investigations in Communication Sciences and Disorders 271 Segment before and after pa was presented Segment before and after pa was presented Segment before and after ba was presented Figure 36 3. Data segmentation for /pa/ and /ba/. Näätänen, Kujala, & Winkler, 2010). MMN is very commonly used in language studies due to its preattentional aspect, where MMN can be elicited while children are sleeping or watching a silent movie, as well as due to its robustness as a phonological correlate. Early Left Anterior Negativity (ELAN) and P600: Morphosyntactic Correlates ELAN refers to an enhanced negativity over the left anterior brain regions, peaking around 100 to 200 ms after the onset of a syntactic phrase structure violation, indexing first pass parsing processes (Hahne & Friederici, 1999). Another syntactic processing correlate is the P600, also referred to as LPC (Late Positive Complex). P600/LPC refers to a positivity detected over centroparietal regions around 500 msec after recognition of a syntactic anomaly (e.g. Friederici, Hahne, & Saddy, 2002). P600 can be elicited for several types of grammatical violations as well as for lexical codeswitching between L1 and L2 (Jackson, Swainson, Cunnington, & Jackson, 2001; Moreno, Federmeier, & Kutas, 2002), or between two language varieties in diglossic situations (e.g., Khamis-Dakwar & Froud, 2007). Hence, in addition to being considered as a correlate of grammatical processing, it is also considered to index reanalysis of syntactic and semantic information in situations carrying a high
272 Translational Speech-Language Pathology and Audiology cognitive load, especially when it appears in conjunction with N400 effect (Bornkessel- Schlesewsky & Schlesewsky, 2008). N400 Lexical and Semantic Correlate N400 refers to a negativity detected over centroparietal regions around 400 ms after onset of a stimulus, and is considered to index lexical and semantic processing (Lau, Philips, & Poppel, 2008). It has been found to be sensitive to meaning manipulations and to the presentation of an unexpected word at both the word and sentence level in different modalities (e.g., spoken and written language), as well as with other kinds of anomalies, such as mismatches in phonology, or unexpected faces, color patches, and pictures (for a review see Kutas, Van Petten, & Kluender, 2006 or Lau et al., 2008). Clinically Driven Neurophysiologic Studies in CSD at the NSLP Lab There are only a few known clinically driven ERP investigations in CSD. Many of these studies have focused on examining the language representation and processing in a specific communication disorder such as dyslexia or stuttering (e.g., Cuadrado & Weber-Fox, 2003; Datta et al, 2010; McAnally & Stein, 1997; Weber-Fox & Hampton, 2008; Weber- Fox & Neville, 2001). Fewer studies have examined the brain changes occurring across time as language is acquired in a specific communication disorder or during the implementation of a specific treatment approach. Building on the existing ERP research and the power of the ERP method for examining the different subprocesses of language processing as indexed by the classic language ERP correlates, we argue that further incorporation of ERP studies in the field of CSD is beneficial. There is a need to expand ERP studies not only to enhance our understanding of the nature of specific communication disorders, but also to understand the course of language learning in specific communication disorders and the different SLP therapy effects. Toward that end, the NSLP Lab in Adelphi University was established in 2008. As the director of the lab, part of my mission is to inform colleagues and students within the Adelphi University community about the ERP method and to encourage them to make use of the EEG system in the NSLP Lab to address their clinically based inquiries. Research at the NSLP Lab focuses on examining the neural underpinnings of language learning in specific sociolinguistic situations (such as Arabic diglossia), specific communication and language disorders (such as apraxia of speech), as well as during rehabilitation (such as in accent modification therapy) using event-related potentials (ERPs). Lexicosemantic Deficits in Children Undergoing Insulin Treatment for Diabetes Difficulties in vocabulary development have been reported in children with insulindependent diabetes mellitus (IDDM) (Northam, Anderson, Wether, Warne, Adler, & Andrewes, 1998). Using a word-picture semantic processing task, we examine differences in the brain responses of children with IDDM after initial and long-term exposure to insulin treatment. This study aims to clarify whether reported deficits in vocabulary development in this population reflect changes in neural organization for linguistic processing,
Translational Display of Neurophysiologic Investigations in Communication Sciences and Disorders 273 which have potential clinical implications in terms of the need for preventive service delivery for children with diabetes, beginning at the earliest stages of insulin therapy. Language Development in Children Who Are Internationally Adopted (CWIA) The effects of immersing an internationally adopted (IA) child in the new language of the environment (LE) whereas ceasing exposure to the language of adoption (LA, or first language), are not fully understood. A recent meta-analytic examination of language skills in CWIA revealed that IA children show poorer language performances during the school-age years (and beyond) as opposed to their nonadopted peers (Scott, Roberts, & Glennen, 2011). To evaluate whether later language difficulties might be related to differences in phonemic system organization, an investigation of auditory MMN responses from five CWIA (3 5 years old) adopted from China and exposed to LE for at least 2 years, and five matched nonadopted monolingual English speaking children was conducted (Khamis-Dakwar & Scott, in preparation). Words were presented in a randomized order in passive listening oddball paradigms under three conditions: (1) Chinese only phonemic feature (/maai5/tone 5 (buy ) & /maai6/tone 6 (sell), based on tonal differences present in Mandarin and not in English; (2) Englishonly phonemic feature not represented in Mandarin (/mad/ vs. /mat/); and (3) phonemic contrast evident in both languages (/mad/ and /man/). MMN was derived from 32 channel EEG recordings by averaging and montaging to frontocentral sensors and subtraction of averaged standard responses from averaged deviant responses within each condition. Enhanced MMN responses were observed for tonal differences in CWIA children, but not for the control group. The English phonemic contrasts were associated with delayed MMN responses for the CWIA group, but not the control group. These preliminary findings reveal differences in the phonological representations in CWIA children that were not identifiable based on behavioral testing of language and phonological performance. Therapy Effects A growing body of anecdotal evidence reveals positive effects of using yoga in speech therapy with children and adults (see Kaley-Isley, Peterson, Fischer, & Peterson, 2010 for a review). Reported effectiveness of incorporating yoga intervention techniques in SLP treatment can be underlined by direct effect on the language processing system or indirect effect on general cognitive skills. In a pilot study, we examined the effects of yoga intervention using ERP on an 11-year-old child presenting with a language delay and ADHD. EEG data were recorded while the participating child was presented with two sets of stimuli in a picture-word paradigm while practicing a previously learned relaxation technique developed as a part of yoga-based therapy for that child. The first set of stimuli represented a congruous condition composed of 25 pairs of auditory word representations and matching picture presentations (congruous condition). The second stimuli set represented an incongruous condition in which 25 pairs of auditory words did not match the visual representation on a display (incongruous condition). The study findings reveal that the implementation of the yoga-based relaxation techniques during exposure to the lexical task was correlated with a P300 enhancement indexing attentional resource allocation, as compared with
274 Translational Speech-Language Pathology and Audiology ERP responses elicited from a child serving as a control participant (i.e., matched for typical language development for gender, age, socioeconomic status, and region of residence). Viewed from a neurophysiologic perspective, these pilot results showing a P300 enhancement correlated with yoga therapy, rather than an N400 enhancement (an index of lexical processing) suggests that the reported positive effects of incorporating yoga for this child are likely due to general cognitive enhancement during linguistic processing. Conclusion Since the establishment of the NSLPlab, my mission has been to open the lab for colleagues and students to make use of ERP method to enhance their understanding of a specific clinical question of interest. My intention here has been to provide examples of studies developed and conducted in the lab in collaboration with colleagues and students that exemplify the power and range of potential benefits to be gained from incorporating the ERP method in CSD studies. References Bornkessel-Schlesewsky, I., & Schlesewsky, M. (2008). An alternative perspective on semantic P600 effects in language comprehension. Brain Research Reviews, 59(1), 55 73. Cuadrado, E., & Weber-Fox, C. (2003). Atypical syntactic processing in individuals who stutter: Evidence from event-related brain potentials and behavioral measures. Journal of Speech, Language, and Hearing Research, 46, 960 976. Datta H., Shafer, V. L, Morr, M. L., Kurtzberg, D., & Schwartz, R. G. (2010.) Brain discriminative responses to phonetically similar long vowels in children with SLI. Journal of Speech Language Hearing Sciences, 53(3), 757 777. Friederici, A. D. (2002).Towards a neural basis of auditory sentence processing. Trends in Cognitive Science, 6, 78 84. Friederici, A. D., Hahne A., & Saddy, D. (2002). Distinct neurophysiological patterns reflecting aspects of syntactic complexity and syntactic repair. Journal of Psycholinguistic Research, 31(1), 45 63. Hahne, A., & Friederici, A. D. (1999).Electrophysiological evidence for two steps in syntactic analysis:early automatic and late controlled processes. Journal of Cognitive Neuroscience, 11, 194 205. Jackson, G. M., Swainson, R., Mulin, A., Cunnington, R., & Jackson, S. R. (2004). ERP correlates of receptive language-switching tasks. Quarterly Journal of Experimental Psychology, 57A(2), 223 240. Kaley-Isley, L. C., Peterson, J., Fischer, C., & Peterson, E. (2010). Yoga as a complementary therapy for children and adolescents: A guide for clinicians. Psychiatry, 7(8), 20 32. Khamis-Dakwar, R., & Froud, K. (2007). Lexical processing in two language varieties: An event related brain potential study of Arabic native speakers. In Mughazy, M. (Ed.) Perspectives on Arabic Linguistics (pp. 153 166). Amsterdam: John Benjamins. Khamis-Dakwar, R., & Scott, K. Neural correlates of language acquisition in internationally adopted children. Poster presented at Cognitive Neuroscience Symposium. San Fransisco, April 2011. Kutas, M., Van Petten, C., & Kluender, R. (2006). Psycholinguistics electrified II: 1994 2005. In M. Traxler & M. A. Gernsbacher (Eds.), Handbook of psycholinguistics (2nd ed., pp. 659 724). New York, NY: Elsevier. Lau, E. F, Phillips, C., & Poeppel, D. (2008)A cortical network for semantics: (De)constructing the N400. Nature Reviews Neuroscience, 9(12), 920 933. Luck, S.L. (2005). An introduction to the eventrelated potential technique. Cambridge MA: Massachusetts Institute of Technology. McAnally, K. I., & Stein, J. F. (1997). Scalp potentials evoked by amplitude-modulated tones in
Translational Display of Neurophysiologic Investigations in Communication Sciences and Disorders 275 dyslexia. Journal of Speech, Language, and Hearing Research, 40, 939 945. Moreno, E. M., Federmeier, K. D., & Kutas, M. (2002). Switching languages, switching palabras (words): An electrophysiological study of codeswitching. Brain and Language, 80, 188 207. Näätänen, R., Kujala, T., & Winkler, I. (2010). Auditory processing that leads to conscious perception: A unique window to central auditory processing opened by the mismatch negativity and related responses. Psychophysiology, 48, 4 22. Northam, E. A., Anderson, P. J., Wether, G. A., Warne, G. L., Adler, R. G., & Andrewes, D. (1998). Neuropsychological complications of IDDM in children 2 years after disease onset. Diabetes Care, 21, 379 384. Rugg, M. D. (1997). Cognitive neuroscience. Hove, East Sussex, UK: Psychology Press. Scott, K., Roberts, J. A., & Glennen, S.(2011). How well do children who are internationally adopted acquire language? A meta-analysis. Journal of Speech, Language, and Hearing Research, 54, 1153 1169. Ward, J. (2010). The student s guide to cognitive neuroscience. Hove, East Sussex, UK: Psychology Press. Weber-Fox, C., & Neville, H. J. (2001). Sensitive periods differentiate processing for open and closed class words: An ERP study in bilinguals. Journal of Speech, Language, and Hearing Research, 44, 1338 1353. Weber-Fox, C., & Hampton, A. (2008). Stuttering and natural speech processing of semantic and syntactic constraints on verbs. Journal of Speech, Language, and Hearing Research, 51(5), 1058 1071.