Communicating through speech is an important part of everyday life, and losing that ability can be devastating. Millions of patients around the globe have lost the ability to hear or speak due to auditory cortex deficits. Doctor’s ability to help these patients has been hindered by a lack of understanding of the speech processing mechanisms in the human auditory cortex. This dissertation focuses on enhancing our understanding of the mechanisms of speech encoding in human primary and secondary auditory cortices using two methods of electroencephalography (EEG) and electrocorticography (ECoG). Phonemes are the smallest linguistic elements that can change a word’s meaning. I characterize EEG responses to continuous speech by obtaining the time-locked responses to phoneme instances (phoneme-related potential). I show that responses to different phoneme categories are organized by phonetic features, and each instance of a phoneme in continuous speech produces multiple distinguishable neural responses occurring as early as 50 ms and as late as 400 ms after the phoneme onset. Comparing the patterns of phoneme similarity in the neural responses and the acoustic signals confirms a repetitive appearance of acoustic distinctions of phonemes in the neural data. Analysis of the phonetic and speaker information in neural activations reveals that different time intervals jointly encode the acoustic similarity of both phonetic and speaker categories. These findings provide evidence for a dynamic neural transformation of low-level speech features as they propagate along the auditory pathway, and form an empirical framework to study the representational changes in learning, attention, and speech disorders. Later in this dissertation, I use ECoG neural recordings to explore mechanisms of speech communication in real-world environments that require adaptation to changing acoustic conditions. I explore how the human auditory cortex adapts as a new noise source appears in or disappears from the acoustic scene. To investigate the mechanisms of adaptation, neural activity in the auditory cortex of six human subjects were measured as they listened to speech with abruptly changing background noises. Rapid and selective suppression of acoustic features of noise in the neural responses are observed. This suppression results in enhanced representation and perception of speech acoustic features. The degree of adaptation to different background noises varies across neural sites and is predictable from the tuning properties and speech specificity of the sites. Moreover, adaptation to background noise is unaffected by the attentional focus of the listener. The convergence of these neural and perceptual effects reveals the intrinsic dynamic mechanisms that enable a listener to filter out irrelevant sound sources in a changing acoustic scene. Finally, in the last chapter, I introduce the Neural Acoustic Processing Library (NAPLib). NAPLib contains a suite of tools that characterize various properties of the neural representation of speech, which can be used for characterizing electrode tuning properties, and their response to phonemes. The library is applicable to both invasive and non-invasive recordings, including electroencephalography (EEG), electrocorticography (ECoG) and magnetoecnephalography (MEG). Together, this dissertation shows new evidence for dynamic and adaptive processing of speech sounds in the auditory pathway, and provides computational tools to study the dynamics of speech encoding in the human brain.

The body of work contained in this dissertation consists of seven studies investigating conversational skills in people with aphasia (PWA). The predominant focus is on establishing reliable measures of language skills in unstructured conversation in PWA. Overall, ten measures are investigated, and much of the work is concerned with establishing interrater reliability and test-retest stability. These measures are needed to determine generalization of aphasia intervention to conversation, to inform treatment decision-making, and to develop future interventions that have the capacity to improve language abilities at a conversation-level. The initial work focused on microlinguistic skills (i.e., word and sentence-level language; Leaman & Edmonds, 2019a; 2019c), and then evolved to include macrolinguistic skills (discourse-level language) with a focus on global coherence (Leaman & Edmonds, in press) and topic initiation (Leaman & Edmonds, 2020). In addition, questions emerged regarding: a) the relationship of language production in monologue and in conversation (due to the predominance of monologue testing, as opposed to conversation, in clinical environments; b) normative data for the measures in monologue and in conversation; c) the sensitivity of the measures as treatment outcome measures. Research questions regarding items a and b are addressed in the novel research conducted for the dissertation (reported in the last two manuscripts in this document, i.e., Dissertation Studies 1 and 2 (DS1 and DS2)), and item c is addressed in Obermeyer et al., (in press). In addition, a related outcome of this research is a methodology, The Conversation Collection Protocol (CCP). The CCP was developed to consistently collect unstructured conversations that would have similar interactional features that could be used as language samples. The protocol is based on conversational interactions in typical speakers, and is primarily informed by the Conversation Analysis literature (for an overview see Schegloff, 2007). The CPP was piloted to train SLP conversation partners to use typical, familiar, social, adult-style interactions during the conversations (rather than traditional therapy or instructional behaviors) in the first study (Leaman & Edmonds, 2019a). The protocol was further developed prior to data collection for the dissertation studies. In this development phase, the systemized training protocol was expanded to include excerpted readings from literature regarding conversational interaction, and a post-training quiz for the partners. In addition, a session fidelity protocol was developed and implemented. Use of the CCP in all of the studies contributed to achieving similarity in the SLP partners’ interactional styles across conversational dyads, allowed fostering of social conversations which were desired (i.e., as opposed to interview-style conversations often used in the literature), and promoted the PWA to direct their own communication decisions and topics of discussion which in typical therapy interactions may be drastically limited by the clinician (Simmons-Mackie & Damico, 1999). The CCP resulted in high session fidelity (98-99%) across the 27 SLPs who participated in the two dissertation studies. The CCP training also resulted in a corpus of conversations that are similar in content and complexity (measured by mean length of utterance and type-token ratio), with similar topics and an equitable distribution of topic-initiating utterances between the PWA and the partners (Leaman & Edmonds, 2019a; DS1). This research agenda is motivated by a clinical need and vision for a dramatic shift in aphasia intervention, which moves away from structured, decontextualized therapy tasks and towards use of everyday conversation as the primary vehicle of intervention. Prerequisite to development of such an intervention is development of outcome measures capable of capturing real-world changes in conversation. Without such measures, it is not possible to determine whether treatment has the intended effects on conversation. Because conversation is a complex, multi-modal, and contextually-bound phenomenon, treatment that improves everyday conversation could potentially affect many aspects of communication. Change in conversation can be realized by treatment focused on verbal skills, nonverbal skills, compensatory strategies, participation, and/or partner training, and ideally should combine all of these communication parameters. Currently, measures and scales exist for each of these areas, except for in the area of language ability in conversation. It is this clinical and research gap, the lack of reliable measures to evaluate language in its most commonly used context, conversation, that fuels this line of research. The publications, in press manuscripts, and two manuscripts resulting from the dissertation research are presented in their order of publication. Conclusions, clinical implications, and future directions are presented in each. However, in brief summary, the primary findings of this body of work are: 1. Reliability and Stability of Language in Conversation Measurement of language production skills in PWA can be accomplished with a high-level of reliability and stability for all measures investigated except for: a) the measure of behavioral manifestations of lexical retrieval (LEXoth; Leaman & Edmonds, 2019a); and b) referential cohesion (REF), which demonstrated variability that precluded test-retest stability in two studies (Leaman & Edmonds, 2019a; DS1). The clinical implication is that language in unstructured conversation, for certain measures, is reliable and stable. 2. Language Production Relationships in Monologue and Conversation Language production in monologue does not tend to parallel language production in unstructured conversation, thus performance during monologue therapy tasks cannot reliably predict conversational language production for most measures investigated (DS2). Consequently, language findings based on a monologue task (this research used a story narrative monologue) cannot be extrapolated for understanding of conversational language skills for most of the measures investigated. Thus, if the desired outcome of treatment is impact at a conversation level, evaluation of the effectiveness of the intervention to achieve this aim cannot be estimated with use of a story narrative monologue language sample. Further, development of intervention relevant to language needs in conversation cannot be adequately developed based on a story narrative monologue. Further investigation is needed regarding the relationship between single picture description tasks and conversation in terms of language production skills. 3. Topic Initiation Mechanisms in PWA and Their Partners PWA often use similar mechanisms to alert listeners that a new topic is being initiated as their communication partners without aphasia, such as waiting for an old topic to end, or using a marker like “oh, and by the way…”. In interactions between individuals without communication disorders, these mechanisms are often layered and used simultaneously. However, the findings of this research demonstrate that as aphasia severity increases, individuals use fewer simultaneous mechanisms to introduce topics. In addition, for people with moderate to severe aphasia, the fewer topic initiation mechanisms they use, the less successful they are during their topic initiating utterances (Leaman & Edmonds, 2020). This work provides a broader analysis of topic initiating behaviors in PWA with a larger sample size (n=10) than had been previously available. In addition, it established a needed methodology for locating the beginning and end of topic locations in unstructured conversation. This ability to reliably locate topics within conversation is also key to the subsequent research regarding global coherence in conversation, which depends on analysis of each utterances coherence to the overall topic being discussed (Leaman & Edmonds, in press; DS1). An important clinical implication suggests that explicit teaching PWA to use simultaneous methods of topic initiation may have therapeutic benefit to support a greater level of successfulness when they initiate new topics during conversation. Further, teaching both PWA and their regular partners about mechanisms of topic initiation may facilitate improved awareness of these mechanisms with positive therapeutic effect in conversation. 4. Sensitivity of Linguistic Measures as Post-Treatment Outcomes Evidence of stability and sensitivity of linguistic measures in conversation is provided in an intervention case study (Obermeyer et al., in press). As a case study, this research suggests preliminary evidence that a discourse-level intervention (Attentive Reading and Constrained Summarization -Written) can affect change in conversation, and that measures investigated in the research presented here may be sensitive to such change. 5. Development and Use of the CCP to Train SLPs as Conversation Partners Although not addressed as a research question, the CCP appears to be effective as a systematic method to collect unstructured conversations suitable for language analysis. Further, the CCP training is brief (less than an hour), and a large group of SLPs (27) demonstrated learning and adherence to the protocol, as evidenced by high session fidelity and resulting conversations that are similar in terms of vocabulary use frequency, mean length of utterance, type-token ratio, and even distribution of topic-initiating turns between the PWA and their partners, and similar topical content across the conversations (Leaman & Edmonds, 2019a; DS1). Further, in the dissertation research over 90 conversations were collected using the CCP training (some were not analyzed due to subsequent ineligibility of the participant), and no conversation resembled an interview or traditional didactic therapy interaction. The clinical implication is that SLPs can be efficiently and effectively trained as conversation partners to collect unstructured (social) conversational samples for the purpose of assessment. Next steps in this line of research are detailed in the conclusion of each of the seven articles and manuscripts. In addition, a summary of the findings and future directions based on the entire body of work are included in the Epilogue of this dissertation.

The consistent increase of cultural diversity and immigration within the United States over the last fifty years has contributed to a societal shift towards a growing bilingual population. The growth of this population has generated a need to evaluate current assessment and treatment plans for bilingual post-stroke aphasia patients within the United States to ensure that these individuals are receiving effective healthcare. This study aims to investigate the current knowledge gap surrounding appropriate methods of assessing and treating bilingual post-stroke aphasia patients within the United States and suggest potential approaches based on existing research. In order to synthesize information regarding current methods of assessing and treating bilingual post-stroke aphasia patients and to suggest areas for future research, a review of previously published literature was conducted. To illustrate the association between bilingualism and approaches to healthcare, potential and previously studied assessment and treatment plans for bilingual post-stroke aphasia patients within the United States were evaluated based on the likelihood of their success in a physical clinical setting. Because minimal research currently exists concerning intervention in bilingual aphasic adults, SLPs in the United States are forced to provide services without the knowledge necessary to provide efficacious healthcare to this population. As a result, there is currently a critical need for the development of consistent, culturally relevant assessments and treatment approaches for bilingual post-stroke aphasia patients and for large-scale empirical studies to be conducted in the United States that examine the validity of these assessment and treatment protocols.

<p><i>Purpose</i>. Sentence production is impaired in many persons with aphasia (PWA). Structural priming, a speaker’s tendency to re-use a previously heard sentence structure, has been shown to facilitate sentence production in PWA. Man et al. (2019), however, found that PWA showed significant priming only in transitive sentences but not in dative sentences when these two different types of sentences were presented in an alternating manner within a session [Man, G., Meehan, S., Martin, N., Branigan, H., Lee, J. (2019). Effects of Verb Overlap on Structural Priming in Dialogue: Implications for Syntactic Learning in Aphasia. <i>Journal of Speech, Language, and Hearing Research, 62</i>, 1933-1950]. This study sought to examine whether presenting transitive vs. dative stimuli in a blocked format would yield more consistent priming effects in PWA.</p><p><i>Methods. </i>Twelve PWA and twelve healthy older adults (HOA) completed a dialogue-like priming task, where participants took turns describing pictures with the experimenter. Importantly, each participant received two blocks of transitive and dative priming. In addition, we repeated verbs between prime and target items for half of each block to test if lexical overlap boosts priming, i.e., lexical boost. We measured how often the participant re-used the same syntactic structure they heard the experimenter produce previously when they described their own picture. </p><p><i>Results. </i>HOA showed significant priming and lexical boost in the transitive block and significant priming in the dative block, replicating Man et al. (2019). PWA, showed near significant priming in the transitive block. Importantly, the priming effect became significant when the verb was repeated between prime and target, indicating lexical boost. However, PWA failed to show priming in the dative block. </p><p><i>Discussion.</i> Using a blocked stimulus design only modulated lexically-mediated priming in transitives for PWA, different from Man et al. (2019). Findings suggest that while it is feasible to use structural priming to ameliorate sentence production deficits in PWA, the presentation of target stimuli would likely not influence outcomes.</p><p></p><p></p>

No description available.

Previous research suggests that attentional deficits could be the underlying cause of language impairments in people with aphasia (PWA) and that behavioral engagement ratings can be an accurate way to measure attention to specific tasks. Previous research also suggests that PWA have lower levels of behavioral engagement than neurologically healthy adults. Participants in the present study included 9 PWA and 18 neurologically healthy adults. This was an exploratory study investigating the relationships and differences between behavioral engagement and physiological measures, perceived arousal, and naming accuracy and response time in PWA and neurologically healthy adults. Participants completed a confrontational naming task while physiological measures (heart rate, heart rate variability, and skin conductance) were taken simultaneously. Subsequent video footage was used to rate participants' behavioral engagement (i.e., how engaged the participant was in the naming task). In general, PWA had lower behavioral engagement ratings of attention than neurotypical adults. Significant correlations were found between behavioral engagement ratings of attention, naming response time, and naming accuracy. No statistical significance was found between behavioral engagement ratings of attention and heart rate, heart rate variability, and skin conductance. Further research is needed to support these findings.

No description available.

Typical symptoms of aphasia such as comprehension problems and expressive language difficulties may affect the ability to take part in social contexts. Therefore, it is important to find ways to facilitate communication for people with aphasia. CIRCA is a communicationpromoting web service, designed for people with dementia and healthcare providers. Since people with aphasia show similar linguistic and communicative difficulties as people with dementia, it is possible that CIRCA would work as a communication support for people with aphasia as well. The aim of the present study was to investigate CIRCA as group activity for people with aphasia, in comparison to traditional group activities. Five participants with aphasia were recruited. The data collection involved four sessions with the use of CIRCA and four sessions where traditional activities were conducted, each session lasting for about 30 minutes. The sessions were videotaped, transcribed and analyzed according to methods inspired by Conversation Analysis, CA. The authors of this study chose to analyze the use of gestures, laughter, communication supporting behaviour and utilization of the available material in the various activities. The selected sections consisted of commonly used conversational phenomena that were considered to shape and characterize the conversations during traditional group activities, and when using CIRCA. Interviews about the participants' experiences of the conversations were conducted during the various activities. The selected conversational phenomena occurred frequently during all sessions as a type of interactive resources. What primarily characterized CIRCA as a group activity, in comparison to the traditional group activities, was that both pointing and iconic gestures were used to convey or clarify a message; that laughter served as a way of expressing mutual understanding about the material in CIRCA; that guesses in conversations when using CIRCA more often succeeded as communicative help behaviour; and that the images and music in CIRCA were used as a support for expression and understanding in conversation. All participants had a positive attitude toward the use of CIRCA.

Afasi innebär impressiva och expressiva språkliga svårigheter, vilket kan medföra svårigheter att vara delaktig i sociala sammanhang. Det är därför viktigt att hitta sätt att stödja kommunikationen för denna patientgrupp. CIRCA är en kommunikationsfrämjande webbtjänst, framtagen för personer med demens och deras vårdgivare. Då personer med afasi uppvisar liknande språkliga och kommunikativa svårigheter som personer med demenssjukdom, bör CIRCA även kunna användas som kommunikationsstöd för personer med afasi. Föreliggande studies syfte var att undersöka CIRCA som gruppaktivitet för personer med afasi, i relation till traditionella gruppaktiviteter. Fem deltagare med afasi rekryterades. Datainsamlingen bestod av fyra samtal då CIRCA användes och fyra då traditionella aktiviteter genomfördes, där varje session pågick under ungefär 30 minuter. Samtalen videofilmades, transkriberades och analyserades utifrån en CA-inspirerad metod. I föreliggande studie valdes att analysera användningen av gester, skratt, kommunikativa hjälpbeteenden och nyttjande av materialet i aktiviteterna. Utvalda utdrag i studien bestod av vanligt förekommande samtalsfenomen som ansågs forma och karaktärisera samtalen under traditionella gruppaktiviteter och vid användning av CIRCA. Intervjuer genomfördes kring deltagarnas upplevelser av samtalen under de olika sessionerna. Utvalda samtalsfenomen i studien förekom ofta under samtliga samtal och fungerade som olika typer av interaktionella resurser. Det som framförallt utmärkte CIRCA som gruppaktivitet, i relation till de traditionella gruppaktiviteterna, var att både pekning och ikoniska gester nyttjades för att förmedla eller förtydliga ett budskap; att skratt fungerade som ett sätt att uttrycka ömsesidig förståelse kring materialet i webbtjänsten; att gissningar i samtal vid användning av CIRCA oftare lyckades som kommunikativt hjälpbeteende; samt att bilderna och musiken i CIRCA nyttjades som ett stöd för uttryck och förståelse i samtal. Samtliga deltagare var positiva till användandet av CIRCA. / Typical symptoms of aphasia such as comprehension problems and expressive language difficulties may affect the ability to take part in social contexts. Therefore, it is important to find ways to facilitate communication for people with aphasia. CIRCA is a communicationpromoting web service, designed for people with dementia and healthcare providers. Since people with aphasia show similar linguistic and communicative difficulties as people with dementia, it is possible that CIRCA would work as a communication support for people with aphasia as well. The aim of the present study was to investigate CIRCA as group activity for people with aphasia, in comparison to traditional group activities. Five participants with aphasia were recruited. The data collection involved four sessions with the use of CIRCA and four sessions where traditional activities were conducted, each session lasting for about 30 minutes. The sessions were videotaped, transcribed and analyzed according to methods inspired by Conversation Analysis, CA. The authors of this study chose to analyze the use of gestures, laughter, communication supporting behaviour and utilization of the available material in the various activities. The selected sections consisted of commonly used conversational phenomena that were considered to shape and characterize the conversations during traditional group activities, and when using CIRCA. Interviews about the participants' experiences of the conversations were conducted during the various activities. The selected conversational phenomena occurred frequently during all sessions as a type of interactive resources. What primarily characterized CIRCA as a group activity, in comparison to the traditional group activities, was that both pointing and iconic gestures were used to convey or clarify a message; that laughter served as a way of expressing mutual understanding about the material in CIRCA; that guesses in conversations when using CIRCA more often succeeded as communicative help behaviour; and that the images and music in CIRCA were used as a support for expression and understanding in conversation. All participants had a positive attitude toward the use of CIRCA.

L’aphasie est un symptôme fréquent après un AVC et a un impact majeur social, économique, médical et psychologique sur les patients. Des études récentes ont tenté avec peu de succès de rechercher des critères pronostiques cliniques précoces de récupération d’une aphasie. L’enjeu de cette possibilité de prédiction est un enjeu majeur clinique et scientifique et peut influencer la prise en charge ré-éducative décidée dès les premiers jours après l’AVC. De plus, l’étude clinico-physiologique de la récupération du langage permettrait de mieux comprendre les mécanismes de plasticité cérébrale mis en jeux. Tout d’abord, nous allons nous intéresser chez le sujet sain à l’implication du cortex moteur dans des tâches de perception, renforçant l’idée que cette structure anatomique jouerait un rôle plus étendu que celui auquel elle a été reléguée pendant de nombreuses années, puis nous allons explorer des facteurs prédictifs de la récupération de l’aphasie, les facteurs langagiers dans un premier temps et des facteurs électrophysiologiques, notamment via la TMS explorant l’intégrité du cortex moteur, et leur contribution dans la prédiction de la récupération. / Considering the high incidence of post-stroke aphasia and its significant social and economic impact, better understanding the mechanisms of language recovery in order to predict patient’s outcome and to optimize rehabilitation is a clinical and scientific challenge. Here we aimed to study whether the motor cortex is involved in speech and language perception, suggesting this structure could play a crucial role. Then, we investigated whether some language features could contribute to the prognosis of aphasia recovery. Finally, we investigated whether the anatomofunctional evaluation of the corticomotor pathway using TMS could improve the prediction of post stroke aphasia recovery.