Kirrie Ballard | The University of Sydney (original) (raw)

Papers by Kirrie Ballard

International Journal of Language & Communication Disorders, Feb 15, 2021

BACKGROUND To date, studies have not explored whether a dual diagnosis of aphasia plus apraxia of... more BACKGROUND To date, studies have not explored whether a dual diagnosis of aphasia plus apraxia of speech (AOS) versus aphasia alone (APH) affects the response to language-based naming treatments. AIMS To compare the effects of semantic feature analysis (SFA) treatment for individuals with APH versus aphasia plus AOS, and to test if the presence of AOS impacted the effects of treatment. METHODS AND PROCEDURES A non-randomized experimental group study was conducted to explore the treatment, generalization and maintenance effects between the AOS and APH groups. Participants included nine individuals with aphasia and 11 with concomitant aphasia and AOS. Dependent measures included lexical accuracy, number of sound-level distortions, and lexical stress and syllable segmentation errors. OUTCOMES AND RESULTS Both groups showed significantly improved naming accuracy of trained items for up to 2 months post-treatment. Improvement on naming accuracy of untrained items post-treatment, both semantically related and unrelated to trained items, was lower in magnitude. That this may have been due to effects of repeated probing (which included target repetition) or regression to the mean cannot be excluded. There was a tendency for the AOS group to respond slightly better to treatment than the APH group overall, which was not correlated with aphasia severity. Also, measures of phonetic accuracy and fluency improved for both groups, with no main effect of group. Treatment effects did not generalize to formal measures of (untrained) picture naming or expression of correct information units in discourse in a story retelling task. CONCLUSIONS AND IMPLICATIONS Findings indicate that individuals with aphasia plus AOS can gain equivalent benefits in word retrieval and production from the language-based SFA treatment as individuals with aphasia alone. This may be, in part, due to the tendency for SFA to incorporate principles of practice that are known to support motor learning in AOS, such as high intensity, random stimulus presentation and variable practice. Findings provide further support for high-intensity practice and use of self-generated features to facilitate maintenance of effects. What this paper adds What is already known on the subject SFA treatment is the most common intervention for word-finding difficulties for individuals with aphasia. AOS is a common concomitant disorder to aphasia. However, it is not clear whether the effects of language-based SFA treatment are mitigated by the presence of AOS, which tends to respond well to treatments focused on articulatory-kinematic aspects of speech movement. What this paper adds to the existing knowledge This study compares the effects of SFA in a group of individuals with aphasia alone and a group with similar severity of aphasia but with concomitant AOS, ranging from mild to moderate-severe. Overall, AOS did not have a negative effect on response to the treatment. What are the potential or actual clinical implications of this work? Individuals with aphasia plus AOS can be expected to benefit to a similar degree from SFA as people with aphasia alone. It is likely that the use of practice principles of high intensity, random stimulus presentation and varied practice are important components of the protocol.

Aphasiology, Sep 28, 2017

ABSTRACT Background: McNeil and colleagues argued that individuals with pure apraxia of speech (A... more ABSTRACT Background: McNeil and colleagues argued that individuals with pure apraxia of speech (AOS) have low variability of speech error type and error location within repeated multisyllabic words, compared to individuals with conduction aphasia. While this concept has been challenged, subsequent studies have varied in the stimuli and tasks used. Aims: Our aim was to re-examine the variability of segmental errors, as well as lexical prosodic errors, using the same stimuli and tasks as used by McNeil and colleagues in a sample of individuals with AOS plus aphasia or aphasia alone. This sample is considered to be clinically relevant given the high concomitance of these disorders. Methods & Procedures: Participants were 20 individuals with stroke-related AOS plus aphasia and 21 with aphasia alone (APH), with diagnosis based on expert judgments using published criteria. Three consecutive repetitions of 10 polysyllabic words were elicited and variability of error type, error location, and durational stress contrast was measured. Outcome & Results: Errors were significantly more variable in type and more consistent in location within word for the AOS group than the APH group. The AOS group showed a greater number of errors overall, were less likely to improve production over the three repetition trials, and produced no clear difference in vowel duration across the first two syllables (i.e., durational stress contrast) across repetitions. The measure of durational stress contrast was a stronger predictor of AOS presence than the measures of error variability. Conclusions: The divergence of our findings from previous work likely reflects the more complex profile of the AOS plus aphasia cases in the current study. While durational stress contrast was sufficient to predict diagnostic group, error variability measures were significantly associated with AOS and can contribute to developing targeted intervention goals.

Cortex, 2017

The majority of logopenic variant primary progressive aphasia (lv-PPA) cases harbour Alzheimer pa... more The majority of logopenic variant primary progressive aphasia (lv-PPA) cases harbour Alzheimer pathology, suggesting that lv-PPA constitutes an atypical presentation of Alzheimer's disease (AD). However, even if caused by Alzheimer pathology, the clinical manifestations of lv-PPA differ from those observed in the typical or amnestic AD presentation: in lv-PPA, aphasia is the main feature while amnestic AD is characterised by impaired episodic memory. Anomia or impaired naming, however, is present in both AD presentations. Whether these presentations share anatomical and mechanistic processes of anomia has not been fully investigated. Accordingly, we studied naming performance and its relationship with regions of brain atrophy in 23 amnestic AD and 22 lv-PPA cases with presumed underlying Alzheimer pathology. Both AD groups displayed some degree of anomia and impaired word comprehension but these were particularly severe in lv-PPA and accompanied by a range of linguistic deficits, comprising phonological substitutions, superordinate semantic paraphasias and abnormal single-word repetition. Analysis of cortical thickness revealed that anomia was correlated with thinning in left superior temporal gyrus in both groups. In amnestic AD, however, anomia was also associated with thinning in right inferior temporal regions. Single-word comprehension (SWC), by contrast, was associated with cortical thinning involving bilateral fusiform gyri in both groups. These findings suggest that anomia in both amnestic AD and lv-PPA results from the involvement at multiple steps of word processing, in particular, semantic and lexical retrieval; in addition lv-PPA patients display a more marked involvement of phonological processing.

International Journal of Speech-Language Pathology, May 4, 2023

Interspeech 2022, Sep 18, 2022

International Journal of Stroke, Aug 1, 2014

The Journal of Nuclear Medicine, Aug 6, 2015

Alzheimer disease is the cause of up to one-third of cases of primary progressive aphasia or cort... more Alzheimer disease is the cause of up to one-third of cases of primary progressive aphasia or corticobasal syndrome. The primary objective of this study was to determine the accuracy of 18 F-FDG PET metabolic imaging for the detection of Alzheimer disease in patients with primary progressive aphasia or corticobasal syndrome. Methods: A cohort of patients (n 5 94), including those with an expert clinical diagnosis of logopenic (n 5 19), nonfluent (n 5 16), or semantic (n 5 13) variants of primary progressive aphasia, corticobasal syndrome (n 5 14), or Alzheimer disease (n 5 24), underwent 18 F-FDG metabolic and 11 C-labeled Pittsburgh compound B (11 C-PiB) amyloid PET brain imaging. 18 F-FDG PET scans interpreted with Neurostat and 3D-SSP displays were classified as revealing Alzheimer disease or "other" by interpreters who were unaware of the clinical assessments and 11 C-PiB PET results. 11 C-PiB PET imaging was considered to be the diagnostic reference standard, with a threshold standardized uptake value ratio of 1.5 being indicative of Alzheimer disease pathology. To address possible bias from subgroup selection for the Alzheimer disease binary classifier, we calculated both conventional and balanced accuracies. Results: Diagnoses of Alzheimer disease based on 18 F-FDG PET resulted in 84% accuracy (both conventional and balanced). In comparison, diagnoses based on clinical assessments resulted in 65% conventional accuracy and 67% balanced accuracy. Conclusion: Brain 18 F-FDG PET scans interpreted with Neurostat and 3D-SSP displays accurately detected Alzheimer disease in patients with primary progressive aphasia or corticobasal syndrome as focal-onset dementias. In such diagnostically challenging cohorts, 18 F-FDG PET imaging can provide more accurate diagnoses, enabling more appropriate therapy.

International Journal of Speech-Language Pathology, Jul 11, 2018

Purpose: A systematic search and review of published studies was conducted on the use of automate... more Purpose: A systematic search and review of published studies was conducted on the use of automated speech analysis (ASA) tools for analysing and modifying speech of typically-developing children learning a foreign language and children with speech sound disorders to determine (i) types, attributes, and purposes of ASA tools being used; (ii) accuracy against human judgment; and (iii) performance as therapeutic tools. Method: Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were applied. Across nine databases, 32 articles published between January 2007 and December 2016 met inclusion criteria: (i) focussed on children's speech; (ii) tools used for speech analysis or modification; and (iii) reporting quantitative data on accuracy. Result: Eighteen ASA tools were identified. These met the clinical threshold of 80% agreement with human judgment when used as predictors of intelligibility, impairment severity, or error category. Tool accuracy was typically580% accuracy for words containing mispronunciations. ASA tools have been used effectively to improve to children's foreign language pronunciation. Conclusion: ASA tools show promise for automated analysis and modification of children's speech production within assessment and therapeutic applications. Further work is needed to train automated systems with larger samples of speech to increase accuracy for assessment and therapeutic feedback.

Perspectives of the ASHA Special Interest Groups, Dec 17, 2020

Purpose: "Children are the hidden victims of the COVID-19 pandemic" (United Nations Children's Fu... more Purpose: "Children are the hidden victims of the COVID-19 pandemic" (United Nations Children's Fund, 2020). Timely and effective speech intervention is important to reduce the impact on children's school achievement, ability to make friends, mental health, future life opportunities, and government resources. Prior to the coronavirus disease (COVID-19) pandemic, many Australian children did not receive sufficient speech-language pathology (SLP) services due to long waiting lists in the public health system. COVID-19 restrictions exacerbated this issue, as even children who were at the top of lengthy SLP waiting lists often received limited services, particularly in rural areas. To facilitate children receiving speech intervention remotely during the COVID-19 pandemic, evidence from randomized controlled trials regarding three technological solutions are examined: (a) Phoneme Factory Sound Sorter (Sound Start Study), (b) Waiting for Speech Pathology website, and (c) Apraxia World. Conclusions: For the first two technological solutions, there were similar gains in speech production between the intervention and control groups, whereas, for the third solution, the average magnitude of treatment effect was comparable to face-to-face SLP therapy. Automated therapy management systems may be able to accelerate speech development and support communication resilience to counteract the effects of the COVID-19 restrictions on children with speech sound disorders. Technology-based strategies may also provide a potential solution to the chronic shortage of SLP services in rural areas into the future.

Alzheimers & Dementia, 2010

Objective: To disentangle the clinical heterogeneity of nonsemantic variants of primary progressi... more Objective: To disentangle the clinical heterogeneity of nonsemantic variants of primary progressive aphasia (PPA) and to identify a coherent linguistic-anatomical marker for the logopenic variant of PPA (lv-PPA). Methods: Key speech and language features of 14 cases of lv-PPA and 18 cases of nonfluent/ agrammatic variant of PPA were systematically evaluated and scored by an independent rater blinded to diagnosis. Every case underwent a structural MRI and a Pittsburgh compound B (PiB)-PET scan, a putative biomarker of Alzheimer disease. Key speech and language features that showed association with the PiB-PET status were entered into a hierarchical cluster analysis. The linguistic features and patterns of cortical thinning in each resultant cluster were analyzed. Results: The cluster analysis revealed 3 coherent clinical groups, each of which was linked to a specific PiB-PET status. The first cluster was linked to high PiB retention and characterized by phonologic errors and cortical thinning focused on the left superior temporal gyrus. The second and third clusters were characterized by grammatical production errors and motor speech disorders, respectively, and were associated with low PiB brain retention. A fourth cluster, however, demonstrated nonspecific language deficits and unpredictable PiB-PET status. Conclusions: These findings suggest that despite the clinical and pathologic heterogeneity of nonsemantic variants, discrete clinical syndromes can be distinguished and linked to specific likelihood of PiB-PET status. Phonologic errors seem to be highly predictive of high amyloid burden in PPA and can provide a specific clinical marker for lv-PPA. Neurology® 2014;82:1–8

International Journal of Speech-Language Pathology, Nov 23, 2022

Brain communications, 2021

Aquired apraxia of speech is a disorder that impairs speech production, despite intact peripheral... more Aquired apraxia of speech is a disorder that impairs speech production, despite intact peripheral neuromotor function. Its pathomechanism remains to be established. Neurodegenerative lesion models provide an unequalled opportunity to explore the neural correlates of apraxia of speech, which is present in a subset of patients diagnosed with non-semantic variants of primary progressive aphasia. The normalized pairwise variability index, an acoustic measure of speech motor programming, has shown high sensitivity and specificity for apraxia of speech in cross-sectional studies. Here, we aimed to examine the strength of the pairwise variability index and overall word duration (i.e. articulation rate) as markers of progressive motor programming deficits in primary progressive aphasia with apraxia of speech. Seventy-nine individuals diagnosed with primary progressive aphasia (39 with non-fluent variant and 40 with logopenic variant) and 40 matched healthy controls participated. Patients were followed-up annually (range 1-6 years, median number of visits ¼ 2). All participants completed a speech assessment task and a high-resolution MRI. Our analyses investigated trajectories of speech production (e.g. pairwise variablity index and word duration) and associations with cortical atrophy in the patients. At first presentation, word duration differentiated the nonfluent and logopenic cases statistically, but the range of scores overlapped substantially across groups. Longitudinally, we observed progressive deterioration in pairwise variability index and word duration specific to the non-fluent group only. The pairwise variability index showed particularly strong associations with progressive atrophy in speech motor programming brain regions. Of novelty, our results uncovered a key role of the right frontal gyrus in underpinning speech motor programming changes in non-fluent cases, highlighting the importance of right-brain regions in responding to progressive neurological changes in the speech motor network. Taken together, our findings validate the use of a new metric, the pairwise variability index, as a robust marker of apraxia of speech in contrast to more generic measures of speaking rate. Sensitive/specific neuroimaging biomarkers of the emergence and progression of speech impairments will be useful to inform theories of the pathomechanisms underpinning impaired speech motor control. Our findings justify developing more sensitive measures of rhythmic temporal control of speech that may enable confident detection of emerging speech disturbances and more sensitive tracking of intervention-related changes for pharmacological, neuromodulatory and behavioural interventions. A more reliable detection of speech disturbances has relevance for patient care, with predominance of progressive apraxia of speech a high-risk factor for later diagnosis of progressive supranuclear palsy or corticobasal degeneration.

The 21st International ACM SIGACCESS Conference on Computers and Accessibility

Automatic speech recognition (ASR) technology can be a useful tool in mobile apps for child speec... more Automatic speech recognition (ASR) technology can be a useful tool in mobile apps for child speech therapy, empowering children to complete their practice with limited caregiver supervision. However, little is known about the feasibility of performing ASR on mobile devices, particularly when training data is limited. In this study, we investigated the performance of two low-resource ASR systems on disordered speech from children. We compared the open-source PocketSphinx (PS) recognizer using adapted acoustic models and a custom template-matching (TM) recognizer. TM and the adapted models significantly out-perform the default PS model. On average, maximum likelihood linear regression and maximum a posteriori adaptation increased PS accuracy from 59.4% to 63.8% and 80.0%, respectively, suggesting that the models successfully captured speaker-specific word production variations. TM reached a mean accuracy of 75.8%

Ns = nonsignificant; *** p<0.001; 1 All flu... more Ns = nonsignificant; *** p<0.001; 1 All fluent English speakers premorbidly, by self or family report; 2 3 logopenic and 3 nonfluent participants reported cataracts, one reported a macular tear; polysyllable words were repeated after the examiner and all patients were able to read the Grandfather Passage (large font) unassisted; the number of affected cases is balanced across patient groups; 3 Pittsburgh compound B positron emission tomography scanning using cut-off of 1.5 for neocortical standardized uptake value ratios; 4 Cases of nonfluent PPA without clinical diagnosis of AOS are included, as acoustic measures may be more sensitive than perceptual judgments to motor speech behaviors <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0089864#pone.0089864-Ballard2" target="_blank">[38]</a>; 5 Frontotemporal Lobar Degeneration-modified Clinical Dementia Rating Scale with score >3 indicating impairment <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0089864#pone.0089864-Knopman1" target="_blank">[45]</a>, data available for 12 controls; 6 Lower limit of normal performance is 28/30 for 60–69 year olds and 27/30 for 70–79 year olds <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0089864#pone.0089864-Crum1" target="_blank">[69]</a>, data available for 12 controls.</p

The Journal of Nuclear Medicine, 2015

Interspeech 2021, 2021

Here we present AusKidTalk [1], an audiovisual (AV) corpus of Australian children's speech collec... more Here we present AusKidTalk [1], an audiovisual (AV) corpus of Australian children's speech collected to facilitate the development of speech based technological solutions for children. It builds upon the technology and expertise developed through the collection of an earlier corpus of Australian adult speech, AusTalk [2,3]. This multi-site initiative was established to remedy the dire shortage of children's speech corpora in Australia and around the world that are sufficiently sized to train accurate automated speech processing tools for children. We are collecting ~600 hours of speech from children aged 3-12 years that includes single word and sentence productions as well as narrative and emotional speech. In this paper, we discuss the key requirements for AusKidTalk and how we designed the recording setup and protocol to meet them. We also discuss key findings from our feasibility study of the recording protocol, recording tools, and user interface.

Journal of Communication Disorders, 2020

One of the key principles of motor learning supports using knowledge of results feedback (KR, i.e... more One of the key principles of motor learning supports using knowledge of results feedback (KR, i.e., whether a response was correct / incorrect only) during high intensity motor practice, rather than knowledge of performance (KP, i.e., whether and how a response was correct/incorrect). In the future, mobile technology equipped with automatic speech recognition (ASR) could provide KR feedback, enabling this practice to move outside the clinic, supplementing speech pathology sessions and reducing burden on already stretched speech-language pathology resources. Here, we employ a randomized controlled trial design to test the impact of KR vs KP feedback on children's response to the Nuffield Dyspraxia Programme 3, delivered through an android tablet. At the time of testing, ASR was not feasible and so correctness of responses was decided by the treating clinician. Method: Fourteen children with CAS, aged 4-10 years, participated in a parallel group design, matched for age and severity of CAS. Both groups attended a university clinic for 1-hr therapy sessions 4 days a week for 3 weeks. One group received high frequency feedback comprised of both KR and KP, in the style of traditional, face-to-face intensive intervention on all days. The other group received high frequency KR + KP feedback on 1 day per week and high frequency KR feedback on the other 3 days per week, simulating the service delivery model of one clinic session per week supported by tablet-based home practice. Results: Both groups had significantly improved speech outcomes at 4-months post-treatment. Posthoc comparisons suggested that only the KP group showed a significant change from pre-to immediately post-treatment but the group difference had dissipated by 1-month post-treatment. Heterogeneity in response to intervention within the groups suggests that other factors, not measured here, may be having a substantive influence on response to intervention and feedback type. Conclusion: Mobile technology has the potential to increase motivation and engagement with therapy and to mitigate barriers associated with distance and access to speech pathology services. Further research is needed to explore the influence of type and frequency of feedback on motor learning, optimal timing for transitioning from KP to KR feedback, and how these parameters interact with task, child and context-related factors.