Cynthia Thompson, Ph.D. - US grants

The efficacy of a linguistic-specific approach for treatment of sentence production deficits in agrammatic, Broca's aphasic patients who evince difficulty with producing (and comprehending) "complex" sentences -- sentences where noun phrases (NPs) have been moved out of their canonical positions is examined in this research. In consideration of aspects of theoretical linguistics (i.e., Government Binding theory (Chomsky, 1981, 1986)) and the linguistic, psycholinguistic, and neurolinguistic literature on sentence processing, this treatment was designed explicitly for training complex sentences by controlling linguistic properties known to underlie sentence formation and by emphasizing linguistic principles used across different sentence types. Exploiting the underlying linguistic representation of sentences, subjects will be trained to appreciate the thematic role assignments of sentential NPs and to operate the required movement to eventually derive surface forms of Wh questions, object clefts, and passive sentences (i.e., considering "Move Alpha" rules -- either Wh- or Np-Movement (Chomsky, 1981)). In a series of experiments, single- subject experimental designs are utilized to study the relation among trained and untrained sentences while carefully controlling lexical and syntactic properties. The following experimental questions are posed: 1) Does treatment of particular Wh-movement structures (e.g., "What" questions) result in generalized production of untrained structures relying on similar Wh-movement (e.g., "Who" questions)?; 2) does treatment of sentences derived from Wh-movement (e.g., Wh-questions) generalize to untrained sentences (e.g., object clefts) that are very different in their surface realizations, yet are also derived from Wh-movement?; 3) does training structures derived from Wh-movement generalize to untrained structures derived from NP-movement, and vice versa?; 4) does generalization from more to less complex sentences relying on either Np- or Wh-movement result from this training (with complexity defined in terms of the number of NPs in the surface form of sentences and in terms of the inherent complexity of the verb's representation)?; 5) does treatment affect aspects of spontaneous discourse?; 6) does production treatment influence comprehension?; and (7) how does 'linguistic-specific' treatment compare with other treatments for sentence production deficits? This research will potentially result in new and efficacious treatment for aphasic individuals with sentence production deficits. If our hypotheses are correct concerning the need to consider the underlying representation of aberrantly produced sentences, then generalization across sentences relying on the same mechanisms but with different surface forms should be noted. Further, the extent to which learning and generalization patterns follow predictions made based on linguistic theory, will provide information concerning the value of such theories for predicting breakdown and recovery patterns in aphasia.

The efficacy of a linguistic-specific approach for treatment of sentence production deficits in agrammatic, Broca's aphasic patients who evince difficulty with producing (and comprehending) "complex" sentences -- sentences where noun phrases (NPs) have been moved out of their canonical positions is examined in this research. In consideration of aspects of theoretical linguistics (i.e., Government Binding theory (Chomsky, 1981, 1986)) and the linguistic, psycholinguistic, and neurolinguistic literature on sentence processing, this treatment was designed explicitly for training complex sentences by controlling linguistic properties known to underlie sentence formation and by emphasizing linguistic principles used across different sentence types. Exploiting the underlying linguistic representation of sentences, subjects will be trained to appreciate the thematic role assignments of sentential NPs and to operate the required movement to eventually derive surface forms of Wh questions, object clefts, and passive sentences (i.e., considering "Move Alpha" rules -- either Wh- or Np-Movement (Chomsky, 1981)). In a series of experiments, single- subject experimental designs are utilized to study the relation among trained and untrained sentences while carefully controlling lexical and syntactic properties. The following experimental questions are posed: 1) Does treatment of particular Wh-movement structures (e.g., "What" questions) result in generalized production of untrained structures relying on similar Wh-movement (e.g., "Who" questions)?; 2) does treatment of sentences derived from Wh-movement (e.g., Wh-questions) generalize to untrained sentences (e.g., object clefts) that are very different in their surface realizations, yet are also derived from Wh-movement?; 3) does training structures derived from Wh-movement generalize to untrained structures derived from NP-movement, and vice versa?; 4) does generalization from more to less complex sentences relying on either Np- or Wh-movement result from this training (with complexity defined in terms of the number of NPs in the surface form of sentences and in terms of the inherent complexity of the verb's representation)?; 5) does treatment affect aspects of spontaneous discourse?; 6) does production treatment influence comprehension?; and (7) how does 'linguistic-specific' treatment compare with other treatments for sentence production deficits? This research will potentially result in new and efficacious treatment for aphasic individuals with sentence production deficits. If our hypotheses are correct concerning the need to consider the underlying representation of aberrantly produced sentences, then generalization across sentences relying on the same mechanisms but with different surface forms should be noted. Further, the extent to which learning and generalization patterns follow predictions made based on linguistic theory, will provide information concerning the value of such theories for predicting breakdown and recovery patterns in aphasia.

The efficacy of a linguistic-specific approach for treatment of sentence production deficits in agrammatic, Broca's aphasic patients who evince difficulty with producing (and comprehending) "complex" sentences -- sentences where noun phrases (NPs) have been moved out of their canonical positions is examined in this research. In consideration of aspects of theoretical linguistics (i.e., Government Binding theory (Chomsky, 1981, 1986)) and the linguistic, psycholinguistic, and neurolinguistic literature on sentence processing, this treatment was designed explicitly for training complex sentences by controlling linguistic properties known to underlie sentence formation and by emphasizing linguistic principles used across different sentence types. Exploiting the underlying linguistic representation of sentences, subjects will be trained to appreciate the thematic role assignments of sentential NPs and to operate the required movement to eventually derive surface forms of Wh questions, object clefts, and passive sentences (i.e., considering "Move Alpha" rules -- either Wh- or Np-Movement (Chomsky, 1981)). In a series of experiments, single- subject experimental designs are utilized to study the relation among trained and untrained sentences while carefully controlling lexical and syntactic properties. The following experimental questions are posed: 1) Does treatment of particular Wh-movement structures (e.g., "What" questions) result in generalized production of untrained structures relying on similar Wh-movement (e.g., "Who" questions)?; 2) does treatment of sentences derived from Wh-movement (e.g., Wh-questions) generalize to untrained sentences (e.g., object clefts) that are very different in their surface realizations, yet are also derived from Wh-movement?; 3) does training structures derived from Wh-movement generalize to untrained structures derived from NP-movement, and vice versa?; 4) does generalization from more to less complex sentences relying on either Np- or Wh-movement result from this training (with complexity defined in terms of the number of NPs in the surface form of sentences and in terms of the inherent complexity of the verb's representation)?; 5) does treatment affect aspects of spontaneous discourse?; 6) does production treatment influence comprehension?; and (7) how does 'linguistic-specific' treatment compare with other treatments for sentence production deficits? This research will potentially result in new and efficacious treatment for aphasic individuals with sentence production deficits. If our hypotheses are correct concerning the need to consider the underlying representation of aberrantly produced sentences, then generalization across sentences relying on the same mechanisms but with different surface forms should be noted. Further, the extent to which learning and generalization patterns follow predictions made based on linguistic theory, will provide information concerning the value of such theories for predicting breakdown and recovery patterns in aphasia.

Project Summary The overarching goal of this project is to develop effective language treatment protocols for individuals with stroke-induced agrammatic aphasia. Building on theories of linguistic representation and the neurocognitive mechanisms of sentence processing in healthy speakers, this project investigates the processes that support sentence production and comprehension (and treatment-induced recovery) in agrammatic aphasia, using a series of online processing experiments as well as an experimental treatment study. This project uses primarily online methods to examine thematic integration processes in sentence production and comprehension. Thematic integration is the incremental mapping between thematic roles (e.g., Agent, Theme) and sentence constituents (e.g., subject, object) ? i.e., the processes of comprehending or expressing ?who did what to whom.? These processes are essential for successful sentence production and comprehension but appear to be impaired in agrammatic aphasia. A series of four sentence comprehension experiments, using eyetracking and ERP, test the hypothesis that impaired thematic prediction contributes to impairments in thematic role assignment and sentence comprehension accuracy. Four sentence production experiments, using eyetracking and priming paradigms, examine the effects of impaired thematic planning processes on sentence production ability. Studies also examine the relation between comprehension and production abilities, testing the hypothesis that thematic integration deficits are seen across domains in agrammatic aphasia. Further, an experimental treatment study compares the effectiveness of two novel extensions of Treatment of Underlying Forms (TUF; Thompson & Shapiro, 2005), which train thematic integration using a TUF-Structural approach, which emphasizes verb-based structural processing and a TUF- Word approach, focused on word-by-word incremental processing. Consistent with previous work, both versions of TUF are expected to support improved sentence production, including generalization to simpler linguistically-related structures; however, TUF-Word is expected to foster stronger gains in comprehension, due to its emphasis on training incremental (word-by-word) thematic processing. In addition, this project investigates the neurocognitive mechanisms of treatment-induced language recovery using eyetracking and sentence production paradigms, as well as the effects of language treatment on functional communication ability. The results of this work will contribute to the understanding of sentence processing in aphasia and in healthy adults, the factors that contribute to effective treatment for sentence-level impairments, as well as the neurocognitive mechanisms of treatment-induced recovery in agrammatic aphasia.

DESCRIPTION (provided by applicant): The goal of this project is to develop and test a computer-automated treatment for individuals with aphasia using a virtual clinician. We propose to combine the research findings resulting from a program of work focused on treatment of post-stroke agrammatism, for which there are considerable available efficacy data, i.e., Treatment of Underlying Forms (TUF) (Thompson et al. 2003), with a well-developed and tested interactive computer system that enables face-to-face communication with an intelligent animated agent developed at the Center for Spoken Language Research (CSLR) at the University of Colorado. Although the components for this work are in place, they have never been combined and thus represent innovative research. We propose to develop a "virtual aphasia clinician" that will be capable of delivering stimuli for aphasia treatment. The automated system will be evaluated for its potential independent use by aphasic individuals to further their progress in language recovery beyond what might be expected to result from the limited clinical contact available under current healthcare policy. If successful, this project will lay the groundwork for a next step - to develop speech recognition software permitting the virtual clinician to evaluate aphasic users' verbal responses. In Phase I of the project we will develop the system using presently available TUF protocols for training syntactically complex sentences. Phase 2 will implement and experimentally assess the effectiveness of the automated TUF-tutor (TUF-T) protocols developed in Phase 1. We propose to compare traditional, clinician-delivered TUF to two TUF-T conditions, one with and one without clinician mediation. The working hypothesis is that TUF-T will be more effective than, or equally effective as, clinician-administered TUF when time and clinician-cost restraints, as well as language improvement are considered. If successful, this work will result in new directions for the management of speech and language disorders that incorporate and take advantage of advances in computer technology.

[unreadable] DESCRIPTION (provided by applicant): The goal of the work described in this proposal is to provide a neuroimaging-based examination of the mechanisms of aphasia treatment and recovery. We propose to use functional magnetic resonance imaging together with structural and perfusion MR imaging as tools for tracking, recording and evaluating brain behavior changes linked to treatment-based changes in four experiments. We propose to systematically examine both behavioral and (correlated) neural effects of treatment while manipulating several critical patient-based and treatment-based variables that have been hypothesized to be relevant to aphasia recovery. The patient-related variables we will study include site and extent of lesion, in patients with anterior vs. posterior lesions, in early post-stroke vs. late post-stroke (chronic) stages of recovery. We propose to study the effects of treatment (vs. placebo-treatment) for these patients with a focus on both comprehension and production deficits in sentence level processing, comparing: two different treatment-intensity schedules (intensive vs. extended) and two different complexity-of-treatment conditions (training complex-to-simple vs. simple-to-complex structures). Our focus on this latter variable is based on our previous work showing that the former (complex-to-simple training) results in greater efficacy of treatment and generalization effects as compared to the latter. Additionally, we propose behavioral studies that extend examination of this 'complexity-of treatment' hypothesis into new domains of language processing and new populations (posteriorly lesioned patients with Wernicke's aphasia). Overall, we propose to employ (and replicate) treatment methods with previously demonstrated behavioral efficacy so as to provide a solid basis for examining (via neuroimaging) brain-based changes associated with a range of critical patient-based and treatment-based variables during treatment. Additionally, we propose to examine theoretically motivated extensions of the complexity method to new domains of language treatment. [unreadable] [unreadable]

DESCRIPTION (provided by applicant): The goal of this project is to develop and test a computer-automated treatment for individuals with aphasia using a virtual clinician. We propose to combine the research findings resulting from a program of work focused on treatment of post-stroke agrammatism, for which there are considerable available efficacy data, i.e., Treatment of Underlying Forms (TUF) (Thompson et al. 2003), with a well-developed and tested interactive computer system that enables face-to-face communication with an intelligent animated agent developed at the Center for Spoken Language Research (CSLR) at the University of Colorado. Although the components for this work are in place, they have never been combined and thus represent innovative research. We propose to develop a "virtual aphasia clinician" that will be capable of delivering stimuli for aphasia treatment. The automated system will be evaluated for its potential independent use by aphasic individuals to further their progress in language recovery beyond what might be expected to result from the limited clinical contact available under current healthcare policy. If successful, this project will lay the groundwork for a next step - to develop speech recognition software permitting the virtual clinician to evaluate aphasic users' verbal responses. In Phase I of the project we will develop the system using presently available TUF protocols for training syntactically complex sentences. Phase 2 will implement and experimentally assess the effectiveness of the automated TUF-tutor (TUF-T) protocols developed in Phase 1. We propose to compare traditional, clinician-delivered TUF to two TUF-T conditions, one with and one without clinician mediation. The working hypothesis is that TUF-T will be more effective than, or equally effective as, clinician-administered TUF when time and clinician-cost restraints, as well as language improvement are considered. If successful, this work will result in new directions for the management of speech and language disorders that incorporate and take advantage of advances in computer technology.

DESCRIPTION (provided by applicant): The overarching goal of this project has been, and continues to be, to develop treatment paradigms for individuals with stroke-induced morphosyntactic language deficits, i.e., agrammatic aphasia. This work makes use of mutually supportive representational (linguistic) and processing accounts of language, as well as psycholinguistic and neurolinguistic research findings, as the basis for understanding language breakdown, selecting training targets, and predicting recovery patterns. Patterns of language (re)learning and generalization provide blueprints for clinical protocols and, in turn, address the utility of this translational approach for studying language disorders. Training involves the use of metalinguistic tasks, which exploit linguistic properties and constructs involved in building grammatical sentences. This approach, termed Treatment of Underlying Forms (TUF), has proved to be successful in our past work, showing that linguistically related structures recover in parallel and that maximal recovery results from training more complex rather than simple structures. Four sets of experiments are proposed for the next grant cycle: Set 1 extends our sentence work from Wh- to NP-movement structures. In Set 2, we continue to examine the relation between and among grammatical morphemes, motivating complexity hierarchies by morphological, rather than syntactic, theories and data. We also experimentally examine the relation between syntax and morphology in Set 3 experiments by training syntactic structures and testing generalization to related grammatical morphemes, and vice versa. Finally, Set 4 experiments focus on verb argument structure, testing recovery of canonical sentence forms by controlling the number and type of arguments selected by the verb. The processing mechanisms that support recovery are studied as part of each set of experiments by tracking eye movements during language comprehension and production tasks and the neural correlates of recovery are examined using functional magnetic resonance imaging (fMRI). Both methods used in the previous cycle revealed important processing patterns as well as, in the case of fMRI, pre- to post-training effects on activation patterns. We, therefore, propose to continue these efforts in this continuation proposal. PUBLIC HEALTH RELEVANCE The overarching goal of this project is to develop treatment paradigms for stroke-induced agrammatic aphasia, a language deficit affecting sentence production (and comprehension) ability. Linguistic theory as well as psycholinguistic and neurolinguistic research findings are used as the basis for understanding language deficit patterns, selecting treatment targets, and predicting recovery patterns. The processing mechanisms that support recovery also are studied by tracking eye movements during language comprehension and production tasks and the neural correlates of recovery are examined using functional magnetic resonance imaging (fMRI).

In non-human primates, groups often engage in between group aggression. Despite the integral role these social interactions play in models predicting and explaining primate behavior, our knowledge of why individuals choose to engage in this risky behavior remains somewhat limited. In general, aggression between groups is attributed to defense of resources or defense of mates and infants. This project will investigate the factors underlying aggression during between-group encounters in a territorial South American primate: the white-faced saki monkey. This arboreal primate occupies small home ranges which are actively defended from neighboring groups along home range boundaries. In order to comprehensively test the function of this aggression, this project will gather data on between-group encounters among white-faced saki groups at Brownsberg Nature Park, Suriname. Variation in aggression and individual participation in these interactions will be assessed in terms of dependence upon food abundance, female reproductive state, and within group social bonds. Such data will provide valuable information on a little known primate species as well as serve as a case study to help develop models of primate social behavior. The broader impacts of this study include collaboration with the National Herbarium of Suriname and the training of a Surinamese student. This project will also integrate lines of evidence from often distinct aspects of primatology (feeding behavior, social behavior, and endocrine analysis) in order to comprehensively examine a behavioral trait. This doctoral dissertation research project will contribute to the academic training of a female graduate student.

DESCRIPTION (provided by applicant): The goal of this project is to investigate how linguistically grounded intensive language treatment for individuals with aphasia after stroke can be enhanced by cortical stimulation, performed with implanted electrodes. We hypothesize that cortical stimulation, administered during intensive behavioral language treatment, will strengthen and sustain learning effects. Many speakers with Broca's aphasia present with agrammatism, i.e. difficulty with comprehending and producing sentences, particularly sentences with complex syntax. Research into behavioral treatment for aphasia is progressively leading to more refined methods of intervention. However, although language improvement in speakers with aphasia can still be shown many years after their stroke, there does seem to be a natural, neurophysiological boundary as to what is possible to achieve with behavioral therapy alone. In the current research project, therefore, we will assess the potential to enhance and sustain the beneficial effects of language treatment, by stimulating core brain areas that are known to potentially support language processing of the kind that is trained, simultaneous to behavioral intervention. We expect a program of intensive and focal language treatment, in combination with cortical stimulation of crucial language areas, to lead to significant and long-lasting improvement of language functions in agrammatic aphasia, in excess of the improvement seen after treatment without cortical stimulation. This novel research brings together a multidisciplinary team of established researchers in the areas of neurosurgery and cortical stimulation, neurophysiology, neurolinguistics and treatment for aphasia and holds promise to enhance language recovery in aphasia, raising the existing cap on language improvement in aphasia. Ten participants with agrammatic aphasia will receive 64 sessions of intense language treatment, focused on sentence building processes and grammatical markers. Prior to the treatment phase, five of these participants will be implanted with a cortical stimulation device, which will be switched on during language behavioral treatment. Functional imaging studies are used to determine optimal implant localization, as well as to investigate effects of treatment and stimulation on neural plasticity. An extensive battery of language tests will be used to assess treatment effects and their maintenance in the two subject groups.

DESCRIPTION (provided by applicant): This Clinical Research Center (P50) proposal is focused on providing a neuroimaging-based examination of language recovery in individuals with chronic aphasia resulting from stroke. The goal of the Center is to bring together the expertise of clinical investigators with mutual scientific interest and complementary expertise in understanding the cognitive and neural correlates of stroke-induced language recovery, the effects of language treatment on these processes, and cognitive and neural factors (biomarkers) of language/brain recovery. The proposal includes three subprojects, which examine language recovery in three language domains, with experts in each serving as PI: Subproject 1. The neurobiology of recovery of spoken naming in aphasia, Swathi Kiran (Boston University) and David Caplan (Harvard, MGH); Subproject 2. The neurobiology of recovery of written naming in acquired dysgraphia, Brenda Rapp (Johns Hopkins); and Subproject 3. The neurobiology of recovery of sentence processing in agrammatism, Cynthia Thompson (Northwestern). Combining the expertise and efforts of the PIs on this project will allow us to study aphasia recovery in a comprehensive manner. At the heart of the approach is a common set of cognitive and neuroimaging measures, including functional magnetic resonance imaging (fMRI), structural and perfusion imaging, and diffusion tensor imaging (DTI), which we will use as tools for identifying, monitoring and evaluating cognitive-brain changes at several time points, coinciding with treatment application and spanning a one-year period. Cross-project data will be deposited into a centralized Neuroimaging and Data Analysis Core (Core B) where they will be analyzed by an expert team of neuroimaging specialists and statisticians, led by Todd Parnsh (PI; neurophysicist at Northwestern), using state-of-the art neuroimaging acquisition and analysis methods. This will allow us to examine the effects of treatment designed to improve specific aphasic language deficits and to answer questions relevant to the relation between language and brain variables and treatment-induced versus natural recovery in chronic aphasia. The work will highly impact clinical intervention practices for individuals with aphasia, providing importan information relevant to the effects of treatment and prognosis for recovery, and contribute to knowledge about brain plasticity and the reorganization of language functions in the adult brain.

Using scent to communicate is common in primates and even can occur in humans. Yet research on primate olfaction (sense of smell) lags behind studies of other senses such as vision and hearing. Understanding how scents communicate information, what reactions they produce, and how this information is chemically communicated, can improve our understanding of primate behaviors and motivation. In order to measure and characterize primate scents, researchers have used laboratory methods that require large, high-cost equipment and sample preservation during shipment from a field site. In this high-risk project, the investigators will test a portable field method for characterizing the chemical composition of scents in real time. If the field method is feasible, it will allow new avenues of research that could advance the understanding of olfactory communication in primates and other mammals. This project will promote international collaboration between US and Brazilian researchers, and support student training at a US institution with a high number of first generation college students.

Laboratory-based gas chromatography-mass spectrometry (GC-MS) has been the standard method for quantifying the chemical composition of primate scents. In this high-risk project, the investigators will explore the potential and feasibility of a portable, in-field GC-MS method to characterize the chemical composition of scent marks for wild common marmoset monkeys in Brazil. Development of a field method would mitigate the expense and challenges associated with long-distance transport of samples, and allow for data collection on how scent cues change over time in their original context and the resulting reaction by recipients of the scent mark.

Abstract Agrammatic aphasia (agrammatism), an acquired language disorder, affects the ability to produce and comprehend sentences and significantly impacts the ability to communicate. The choice of the most efficacious and cost-effective treatment for specific language impairments (including agrammatism) is a major challenge for clinicians, as is the ability to formulate prognostic statements about the effects of treatment, based on individuals' language and neural profiles. Advances in understanding the effects of treatment for sentence processing deficits are therefore important for providing optimal intervention. The present proposal builds on previous work showing that Treatment of Underlying Forms (TUF, [1-2]), a metalinguistic approach focused explicitly on improving sentence processing abilities in agrammatic aphasia, results in strong acquisition and generalization effects (see [1], for review), changes in on-line sentence processing as identified in visual world eyetracking studies [3-6], and adaptive neuroplasticity as measured by shifts in BOLD signal derived from functional neuroimaging [7-8]. The overarching goal of the proposed studies is to refine TUF to maximize its efficacy and boost treatment outcomes to ultimately increase its clinical use. We propose a set of studies that test the effects of early treatment components of TUF to identify those that are most critical for successful outcomes (Aim 1, Study 1), compare the relative effects of TUF and Verb Network Strengthening Treatment (VNeST) on offline, online and neural processing (Aim 2, Study 2), and examine the effects of noninvasive neurostimulation (i.e., high-definition, transcranial direct current stimulation (HD- tDCS)) as an adjunct to TUF (Aim 3, Study 3). In addition, we will use multimodal neuroimaging and behavioral data to examine the factors underlying successful treatment outcomes and develop algorithms for predicting responsiveness to treatment based on these variables (Aim 4, Study 4).