Loading...
48 results
Search Results
Now showing 1 - 10 of 48
- Speech intelligibility of Parkinson’s disease patients evaluated by different groups of healthcare professionals and naïve listenersPublication . Carvalho, Joana; Cardoso, Rita; Guimarães, Isabel; Ferreira, JoaquimABSTRACT Introduction: Speech intelligibility, how well a listener comprehends the speaker’s message, is related to the listener’ expertise and type of the message conveyed. There is no evidence about speech intelligibility in different groups of healthcare professionals and naïve listeners. Objectives: This study is the first to understand if there were differences in the speech intelligibility of Parkinson’s Disease (PD) patients by different experienced and naïve listeners, according to the speech stimuli and dysarthria severity. Materials and methods: Randomly digitised audio-files (50 words and 50 sentences) of 10 PD patients, one without dysarthria and 9 with different dysarthria severities (3 each: mild, moderate and severe dysarthria), were collected from a database of 60 PD patients’ audio-files. A jury panel was formed by five different listeners groups including 10 speech and language therapists, 10 neurologists, 10 PD relatives, 12 PD patients, and 10 people from the general population. The jury panel transcribed single words and sentences from the audio recordings, the percentage correctly understood was calculated and the results were compared between the groups. Results: Multiple comparisons showed significant speech intelligibility differences between healthcare professionals and naïve listerners in words (highest effect size, n2 = 0.7) and sentences (the highest effect size: n2 = 0.6). Pairwise comparisons revealed that those significant differences were specifically in words with moderate and severe dysarthria and sentences with all severity levels of dysarthria. Conclusion: The groups of healthcare professionals who work with dysarthria are more likely to understand the PD patients’ speech than the groups of naïve listeners.
- Speech and voice response to a Levodopa challenge in late-stage Parkinson’s diseasePublication . Fabbri, Margherita; Guimarães, Isabel; Cardoso, Rita; Coelho, Miguel; Guedes, Leonor Correia; Rosa, Mário M; Godinho, Catarina; Abreu, Daisy; Gonçalves, Nilza; Antonini, Angelo; Ferreira, JoaquimBackground: Parkinson’s disease (PD) patients are affected by hypokinetic dysarthria, characterized by hypophonia and dysprosody, which worsens with disease progression. Levodopa’s (l-dopa) effect on quality of speech is inconclusive; no data are currently available for late-stage PD (LSPD). Objective: To assess the modifications of speech and voice in LSPD following an acute l-dopa challenge. Method: LSPD patients [Schwab and England score <50/Hoehn and Yahr stage >3 (MED ON)] performed several vocal tasks before and after an acute l-dopa challenge. The following was assessed: respiratory support for speech, voice quality, stability and variability, speech rate, and motor performance (MDS-UPDRS-III). All voice samples were recorded and analyzed by a speech and language therapist blinded to patients’ therapeutic condition using Praat 5.1 software. results: 24/27 (14 men) LSPD patients succeeded in performing voice tasks. Median age and disease duration of patients were 79 [IQR: 71.5–81.7] and 14.5 [IQR: 11–15.7] years, respectively. In MED OFF, respiratory breath support and pitch break time of LSPD patients were worse than the normative values of non-parkinsonian. A correlation was found between disease duration and voice quality (R = 0.51; p = 0.013) and speech rate (R = −0.55; p = 0.008). l-Dopa significantly improved MDS-UPDRS-III score (20%), with no effect on speech as assessed by clinical rating scales and automated analysis. conclusion: Speech is severely affected in LSPD. Although l-dopa had some effect on motor performance, including axial signs, speech and voice did not improve. The applicability and efficacy of non-pharmacological treatment for speech impairment should be considered for speech disorder management in PD.
- The BioVisualSpeech corpus of words with sibilants for speech therapy games developmentPublication . Cavaco, Sofia; Guimarães, Isabel; Ascensão, Mariana; Abad, Alberto; Anjos, Ivo; Oliveira, Francisco; Martins, Sofia; Marques, Nuno; Eskenazi, Maxine; Magalhães, João; Grilo, Ana MargaridaAbstract: In order to develop computer tools for speech therapy that reliably classify speech productions, there is a need for speech production corpora that characterize the target population in terms of age, gender, and native language. Apart from including correct speech productions, in order to characterize the target population, the corpora should also include samples from people with speech sound disorders. In addition, the annotation of the data should include information on the correctness of the speech productions. Following these criteria, we collected a corpus that can be used to develop computer tools for speech and language therapy of Portuguese children with sigmatism. The proposed corpus contains European Portuguese children’s word productions in which the words have sibilant consonants. The corpus has productions from 356 children from 5 to 9 years of age. Some important characteristics of this corpus, that are relevant to speech and language therapy and computer science research, are that (1) the corpus includes data from children with speech sound disorders; and (2) the productions were annotated according to the criteria of speech and language pathologists, and have information about the speech production errors. These are relevant features for the development and assessment of speech processing tools for speech therapy of Portuguese children. In addition, as an illustration on how to use the corpus, we present three speech therapy games that use a convolutional neural network sibilants classifier trained with data from this corpus and a word recognition module trained on additional children data and calibrated and evaluated with the collected corpus.
- Hierarchical classification and system combination for automatically identifying physiological and neuromuscular laryngeal pathologiesPublication . Cordeiro, Hugo; Fonseca, José; Guimarães, Isabel; Meneses, CarlosObjectives. Speech signal processing techniques have provided several contributions to pathologic voice identification, in which healthy and unhealthy voice samples are evaluated. A less common approach is to identify laryngeal pathologies, for which the use of a noninvasive method for pathologic voice identification is an important step forward for preliminary diagnosis. In this study, a hierarchical classifier and a combination of systems are used to improve the accuracy of a three-class identification system (healthy, physiological larynx pathologies, and neuromuscular larynx pathologies). Method. Three main subject classes were considered: subjects with physiological larynx pathologies (vocal fold nodules and edemas: 59 samples), subjects with neuromuscular larynx pathologies (unilateral vocal fold paralysis: 59 samples), and healthy subjects (36 samples). The variables used in this study were a speech task (sustained vowel /a/ or continuous reading speech), features with or without perceptual information, and features with or without direct information about formants evaluated using single classifiers.Ahierarchical classification system was designed based on this information. Results. The resulting system combines an analysis of continuous speech by way of the commonly used sustained vowel /a/ to obtain spectral and perceptual speech features. It achieved an accuracy of 84.4%, which represents an improvement of approximately 9% compared with the stand-alone approach. For pathologic voice identification, the accuracy obtained was 98.7%, and the identification accuracy for the two pathology classes was 81.3%. Conclusions. Hierarchical classification and system combination create significant benefits and introduce a modular approach to the classification of larynx pathologies.
- Undernutrition in institutionalized elderly patients with neurological diseases: comparison between different diagnostic criteriaPublication . Miranda, Diana; Cardoso, Rita; Gomes, R; Guimarães, Isabel; Abreu, Daisy; Godinho, C; Pereira, P; Domingos, Josefa; Pona, N; Ferreira, JoaquimObjectives: To determine and compare the frequency of undernutrition in institutionalized elderly patients with neurological diseases at admission using different nutritional assessment tools. Design: Cross-sectional observational study. Setting: One long-term care institution specialized in neurodegenerative diseases. Participants: 92 Elderly people (aged ≥ 65 years) with at least one neurological condition. Measurements: Mini Nutritional Assessment (MNA), body mass index (BMI), mid-arm (MAC) and calf circumferences (CC) were used for nutritional status assessment. Presence and severity of dysphagia, polypharmacy and feeding difficulties were also assessed. Results: According to MNA, 77.1% of the participants were undernourished at admission. BMI identified 46.8%, MAC identified 44.6% and CC identified 22.8% of undernourished participants. Undernutrition was more frequent in Alzheimer’s disease, stroke and dementia syndromes. 63% had dysphagia for at least one food consistence and most of these patients were malnourished. MNA revealed best concordance with BMI and MAC than with CC. BMI and feeding difficulties were the major risk factors for undernutrition. Conclusion: Undernutrition prevalence in institutionalized elderly with neurological diseases at admission is high. Nutritional assessment tools revealed low concordance between them.
- Projeto SIENHA – Rede europeia estratégica e inovadora para a educação no envelhecimento saudávelPublication . Martins, Maria Elisabete; Alves Lopes, António; Santos, Hugo; Rosado, Maria Da Lapa; Guimarães, Isabel; Paquete, Patrícia; Almeida, Patrícia; Lopes, António M F; Carrillo, Elena; Vilaró, Jordi; Tuomi, Sirpa; Jarvinen, Sari; Kozakiewicz, Mariusz; Kedziora-Kornatowska, Kornelia; Hoxha, Adnan; Pekaric, Jasmin; Gruneberg, ChristianIntrodução e Objetivos: Os desafios em saúde relacionados com o envelhecimento da população, que a Europa enfrenta, obrigam a mudança das medidas e políticas sociais e de saúde nos seus estados-membros. O paradigma de Envelhecimento Saudável (ES) exige a inovação dos ambientes educacionais, sociais e de saúde e a preparação dos seus profissionais, nomeadamente os Fisioterapeutas, para a atuação com foco na promoção e da saúde e bem-estar ao longo do ciclo de vida. Assim torna-se essencial promover o ES no espaço europeu através do desenvolvimento de competências dos estudantes, profissionais de saúde e de serviço social. Material e Métodos: Com esse objetivo estabeleceu-se um projeto de cooperação académica multilateral, com a duração de 36 meses, com a participação de sete Instituições de Ensino Superior (IES), financiado pelo programa ERASMUS+, constituído pelos seguintes produtos intelectuais: identificação de um quadro de competências profissionais em ES; desenvolvimento de um currículo em ES; desenvolvimento de um toolkit de inovação e investigação em ES. Resultados: Prevê-se assim no período de vigência do projeto, a criação de um currículo educativo, centrado no envelhecimento saudável ao longo da vida, dirigido à comunidade educativa das IES parceiras, onde se incluem os cursos da ESSAlcoitão. Os conteúdos desenvolvidos podem ser adaptados a vários níveis de educação para assegurar a sustentabilidade e impacto. Tem como objetivo a longo prazo o desenvolvimento de um mestrado conjunto pelas IES parceiras. Conclusões: O projeto SIENHA poderá contribuir assim para: (i) a expansão do conhecimento sobre o envelhecimento saudável; (ii) O desenvolvimento contínuo de competências de estudantes, professores e profissionais; (iii) Inovação, desenvolvimento e a utilização das tecnologias para suportar a promoção da saúde na Europa; (iv) O desenvolvimento da investigação aplicada em contexto internacional, e a transferência efetiva da teoria na prática; (v) O intercâmbio de conhecimentos entre os setores sociais e os cuidados de saúde europeus para o desenvolvimento de práticas de qualidade.
- Speech handicap index: cross-cultural adaptation and validation in European Portuguese speakers with oral and oropharyngeal cancerPublication . Guimarães, Isabel; Sousa, Ana Raquel; Gonçalves, FilomenaIntroduction: To date, no valid outcome measure has been developed in European Portuguese (EP) to assess the opinion of the patients with oral and oropharyngeal cancer about the speech impact in his/her related quality of life. Objectives: This study aimed to cross-culturally adapt the speech handicap index (SHI) in EP and contribute to its validation in oral and oropharyngeal cancer patients. Materials and methods: A cross-sectional study was carried out in two phases: (i) SHI translation, back-translation, expert analysis, and pilot study; (ii) application of the EP version of the SHI (EP-SHI) to 95 patients with oral and oropharyngeal cancer and 83 gender- and age-matched healthy individuals. A total of 23 randomly chosen patients were asked to re-fill the EP-SHI questionnaire after 3 weeks. The psychometric properties (feasibility, practicability, reliability, and validity) were assessed. Results: The EP-SHI version brought together expert consensus and the acceptability of 75% of the participants in the pilot study. The EP-SHI questionnaire is feasible (no missing data and floor effect < 7%), practical (an average of 5.5 min to complete), possesses internal consistency reliability (Cronbach’s alpha > 0.90), test-retest reliability (interclass correlation coefficient, ICC > 0.90), significantly strong convergent validity with EP-voice handicap index (VHI) (r^o>0.94), distinguishes patients from healthy speakers and within the patients’ group according to individual speech rating and glossectomy surgery. Conclusions: The EP-SHI is a culturally relevant, valid and reliable patient reported outcome measure (PROM), and therefore, it is appropriate to be recommended for used with oral and oropharyngeal cancer patients.
- Protocolo de Avaliação Orofacial: revisão e validação da versão 2 (PAOF-2) em crianças dos quatro aos nove anosPublication . Guimarães, Isabel; Teixeira, Paulo; Raimundo, Ana Filipa; Miguel, Susana; Nobre, Helena; Ascensão, MarianaObjetivos: Descrever o processo de revisão do protocolo de avaliação orofacial (PAOF) e determinar a praticabilidade, fidedignidade e validade da segunda versão (PAOF-2). Métodos: Estudo metodológico, observacional e transversal que envolveu duas fases, revisão e validação. A fase de revisão incluiu a análise dos conteúdos do PAOF (dimensões, itens, escala de avaliação e forma de registo) seguido de validação de conteúdo em grupo focal e pré-teste. Na fase de validação, o PAOF-2 foi aplicado, por terapeutas da fala, a crianças com idades entre os quatro e os nove anos. Foram analisadas as propriedades clinimétricas de praticabilidade (tempo de aplicação), fidedignidade (consistência interna e acordo intra- e inter-examinadores) e validade (construto, convergente e discriminativa). Resultados: A revisão resultou num instrumento com duas dimensões ‘Estrutura’ e ‘Mobilidade’, 47 itens, escala de cinco pontos e folha de registo. A validade de conteúdo do PAOF foi analisada por 13 terapeutas da fala. A validação realizada em 312 crianças indicou praticabilidade adequada (15 minutos de tempo médio de aplicação), consistência interna excelente a boa (Alfa de Cronbach entre 0,94 e 0,77), excelente fidedignidade intra- e inter-examinadores (Coeficiente de correlação intraclasse >0,75), validade de construto (a análise fatorial exploratória explica 71,4% da estrutura interna do PAOF-2 com 47 itens), boa a moderada validade convergente entre o PAOF-2 e o original e validade discriminativa na dimensão ‘Mobilidade’ de acordo com a idade. Conclusão: O PAOF-2 é um instrumento prático, fidedigno e válido com potencialidade para instrumento de avaliação orofacial de crianças entre os quatro e os nove anos de idade.
- An investigation of the voice handicap index with speakers of portuguese: preliminary dataPublication . Guimarães, Isabel; Abberton, EvelynAlthough the Voice Handicap Index (VHI) has been recognized as a useful tool for quantifying American English speakers’ perceived consequences of voice disorder, it has not yet been widely applied in the clinic to study dysphonic patient populations, or tested with a normally speaking population. The purpose of this investigation was to obtain information about the VHI in a group of Portuguese speakers with voice complaints compared with an age- and gender-matched group of speakers without voice complaints. The results demonstrate that speakers with voice complaints have overall VHI total scores significantly higher than speakers without voice complaints (p < 0.001). This is also true for allVHIsubscores in the emotional (p < 0.001), functional (p < 0.05), and physical (p < 0.001) domains. So, the assumption that a group with voice complaints has higher voice handicap impact than the matched control comparison group is justified for Portuguese speakers.
- Sibilant consonants classification with deep neural networksPublication . Anjos, Ivo; Marques, Nuno; Grilo, Ana Margarida; Guimarães, Isabel; Magalhães, João; Cavaco, SofiaAbstract. Many children su ering from speech sound disorders cannot pronounce the sibilant consonants correctly. We have developed a serious game that is controlled by the children's voices in real time and that allows children to practice the European Portuguese sibilant consonants. For this, the game uses a sibilant consonant classi er. Since the game does not require any type of adult supervision, children can practice the production of these sounds more often, which may lead to faster improvements of their speech. Recently, the use of deep neural networks has given considerable improvements in classi cation for a variety of use cases, from image classication to speech and language processing. Here we propose to use deep convolutional neural networks to classify sibilant phonemes of European Portuguese in our serious game for speech and language therapy. We compared the performance of several diferent arti cial neural networks that used Mel frequency cepstral coefcients or log Mel lterbanks. Our best deep learning model achieves classi cation scores of 95:48% using a 2D convolutional model with log Mel lterbanks as input features.