Navigation auf uzh.ch
Navigation auf uzh.ch
Meike Brockmann Bauser & Lei He
Summary
Classic source-filter theory of speech production prescribes that vocal fold vibrations and vocal tract resonances are independent processes. However, the relationship between these two mechanisms are more nuanced. Articulatory movements, on the one hand, change the vocal tract transfer function, where the glottal signal is sieved in the frequency domain. On the other hand, different articulatory setups should affect the muscles responsible for phonation as well. In various studies it has been observed that an open vowel tends to have a lower fundamental frequency as compared to a closed vowel. This project aims to untangle the relationship between articulation and phonation more systematically using both electromagnetic articulography and electroglottography among vocally healthy speakers. Moreover, the description of the naturally large variability among healthy speakers will be helpful for targeting pathologies in speech production.
This study is part of the project «Untangling the relationship between voice and face: A cross-modal approach to talker identity».
Further Information: Meike Brockmann-Bauser & USZ
Meike Brockmann Bauser & M.F. De Paula Soares
Summary
Main aim of this project is to investigate speech related influencing factors on voice onset time (VOT) in individuals with hyperfunctional voice disorders during a variety of speech tasks. On the acoustic dimension, VOT expresses the temporal duration of the phonatory trigger signal during stop consonant production. Previous studies have suggested a delay in VOT with increased speaking fundamental frequency, however opposite results were reported on male and women’s speech. Fundamental frequency control is related to multiple interaction factors, such as laryngeal muscle stiffness, vertical larynx position, and stiffness of extrinsic laryngeal muscles. These factors have been reported to be unbalanced in patients with hyperfunctional dysphonia. In accordance with this, recent studies have confirmed the relationship between decreased VOT duration in patients with hyperfunctional voice disorders, vocal nodules and moderate or severe perceptual dysphonia. Thus, VOT may be an acoustic measure with the potential to indicate the increase of laryngeal tension in patients with so-called primary and secondary hyperfunctional voice disorders. In the present work, the interrelation between speaking fundamental frequency and intensity, speech token and syllable stress with VOT will be investigated in women and men with primary and secondary hyperfunctional dysphonia.
This project study is linked to a broad project called “Study of the behavior of the acoustic measures of the Praat Program in the vocal assessment” witch focuses on the clinical application of the open source software Praat in vocal assessments of Brazilian Portuguese (BP) speakers.
Further Information: Meike Brockmann-Bauser & USZ
Meike Brockmann-Bauser
Summary
Instrumental acoustic and electroglottographic analysis techniques have been recommended for standard voice diagnostics to objectively detect and rate voice dysfunction. However, the currently recommended acoustic measures including jitter, shimmer, HNR and CPPS significantly improve in phonations with higher voice sound pressure level (SPL) in both healthy and disordered voices. This minimizes the diagnostic accuracy and usefulness of these measures when applying current standards in clinical assessments.
The observed changes in louder phonations have been explained with different vocal fold tonus and closure patterns. However, to date this has not been conclusively investigated under uninfluenced phonatory conditions. Electroglottographic (EGG) assessments provide a non-invasive examination of vocal fold movements and functionality. Recent signal analysis innovations facilitate the modelling and categorization of EGG waveform types in relation to voice fundamental frequency and SPL. Therefore, main aim of the present work is to investigate changes in the electroglottographic waveform in different loudness conditions and vowels in healthy and pathologic voices, and to compare these results to standard diagnostic acoustic and electroglottographic indices. Further, it will be assessed, if voice pathology recognition is improved, when EGG assessment results are controlled for speaking voice SPL.
Further Information: Meike Brockmann-Bauser & USZ
Birgit Kleim, Sebastian Olbrich, Elisabeth Stark, Guido Seiler
submitted
Alexis Hervais-Adelman
Content and Goals
Hearing disorders are a source of disturbances in the sound we perceive. These lead to difficulties in understanding speech because the sound signal reaching the speech brain system is impaired. For people with such disorders, the compensatory mechanisms to 'reconstruct' the meaning of the interlocutor's speech are all the more crucial.
When we are confronted with acoustically dulled speech, the brain recruits more cognitive resources to decode the content. Several studies have shown that this process involves, among others, the speech production system. The link between production and perception has been postulated since Aristotle, but its existence and nature remain controversial. The aim of this project is to clarify this link, not only for theoretical interest, but also to develop cognitive interventions to improve language comprehension in people with hearing impairments.
Scientific and Social Context
The World Health Organization estimates that more than 5% of the world's population suffers from disabling hearing loss, and that one-third of people over the age of 65 are affected by disabling hearing loss. Hearing impairment is a growing and increasingly costly problem for society - approximately 1.1 billion young people between the ages of 12 and 35 are at risk of hearing loss from recreational noise exposure. The annual global cost of untreated hearing loss is $750 billion. Thus, the development of new interventions using the latest theories and methods may have important benefits not only for the quality of life of those affected by hearing loss, but also for society.
Further Information: SNF
Basil Preisig
Content and Goals
Auditory stimuli to which we direct our attention (e.g. the speech signal of our conversation partner) are amplified in the brain by "top-down" control signals, and those which we do not pay attention to (e.g. ambient noise in the background) are suppressed.
There is evidence that these neural processes of attention control are impaired in hearing loss and tinnitus. However, it is still unclear exactly which sub-processes are impaired: The amplification of target stimuli or the suppression of interfering sounds.
The aim of the project is to use electroencephalography (EEG) to identify neuronal markers of these attentional subprocesses in people with hearing loss and tinnitus.
Subsequently, the significance of the identified markers for attentional control will be investigated in detail with the help of non-invasive electrical brain stimulation and neurofeedback.
Scientific and Societal Context
Hearing loss and associated complaints such as tinnitus are the fourth leading cause of disability worldwide and are estimated to cost over $750 billion annually.
In particular, socially relevant situations, such as rounds of conversations with multiple participants, are challenging for people with hearing loss and tinnitus.
However, the current generation of hearing aids is not yet able to selectively amplify those acoustic signals on which the hearing aid user wants to focus.
The results of this research project can provide important foundations for the development of future therapeutic brain stimulation and neurofeedback interventions. Furthermore, the results can provide important impulses for the development of attention-guided hearing aids.
Further Information: SNF
Heiko Hausendorf
Summary
When people encounter, meet or otherwise come together and thus enter into an interaction, this is fundamentally associated with a space of perception, movement and action tailored to the purposes of their respective get-together. Without having to think or even talk about it for long, as a rule, the participants create a common 'interaction space'. In the modern institutions of society, this production of interaction space is supported in a distinctive way by architecture in the sense of built, designed and equipped spaces. The visible and accessible expression of this are the functional buildings of modern society and its organizations. In them, specific forms of institutional interaction have found their built home: Thus the 'lecture' takes place as a matter of course in the 'lecture hall', the 'church service' in the 'church', and the 'ticket sale' at the 'ticket counter'. We speak of interaction architectures in these and similar cases because the spaces in question represent architectural forms of solving repeatedly similar interactive problems. As such, they are expressions and components of institutional communication and, like these, are subject to change as products of social evolution. This change is the focus of the present project: both at the counter and in the church hall, and - with some limitations - also in the lecture hall, we have for some time been able to experience how long-established interaction architectural specifications are being withdrawn (and in some cases actually dismantled), so that transitions occur from the counter to the service center, from the church hall to the open space, and from the lecture hall to the multimedia hub. By examining the aforementioned settings and the interactions that take place there, the project aims to test the thesis that the interaction-architectural changes are accompanied by tendencies towards the de-differentiation, de-structuring and de-hierachization of institutional communication, which can be understood as a process of 're-figuration'.
Further Information: SNF
Heiko Hausendorf
Summary
The COVID-19 pandemic is leading to fundamental changes in the communication
between physicians and patients - also in psychotherapy: many therapies no longer take place no longer face-to-face, but via video telephony. What impact does this have on the
effects on the therapeutic relationship? What problems arise, how are they solved? Which ones remain? How is 'presence' created and a relationship established under these new circumstances?
These questions are investigated by an interdisciplinary team from psychiatry/psychotherapy and linguistics with the aim of supporting therapists and patients by providing them with specific communicative strategies to help them adjust to the new situation.
Further Information: Heiko Hausendorf
Volker Dellwo
Summary
In the past, voice recognition has been studied almost exclusively on the side of the recipient, i.e., under the question of how good is the recognition performance of human listeners or machines under certain circumstances. In the present project, we investigate whether, and if so how, speakers modify their voice to make it more or less recognizable. In a first step, we are building a large database of about 500 speakers from the greater Zurich area in which speakers speak under different situational circumstances (e.g., speaking to children, to elderly people, or in a formal situation). This database will be used as a reference. Following the data collection, we will test in different experiments how speakers change their voice when they want to make themselves more recognizable and when they want to hide their vocal identity. We hypothesize that speakers will follow reference when adapting their voices, i.e., they will tend to adapt their vocal features to the vocal average to avoid being recognized, or deviate from the average to become recognizable.
The results are of importance both for understanding the whole process of voice recognition in humans, but also for improving the recognition performance of computer systems.
Further Information: SNF
Nikola Biller-Andorno
Summary
The Swiss Database of Individual Patient Experiences provides a systematic and methodologically collection of individual interviews about real-life experiences with health issues as Alzheimer diseases or multiple sclerosis. DIPEx Switzerland is a group of researchers from the Institute of Biomedical Ethics and History of Medicine (University of Zurich) and the Zurich University of Applied Sciences – School of Health Professions (ZHAW). The project is part of the international DIPEx network with 14 countries, led by UK. DIPEx works with a well-established and rigorous qualitative research methodology developed by the University of Oxford in 2000, and which is documented in the Health Experience Research Group (HERG) manual. This qualitative study is based on narrative and semi-structured interviews. Video- and/or audio-or text sequences will be presented online.
We conduct narrative interview studies of people’s experiences of health issues and provide resources for information and support that are freely available for patients, family members, clinicians in training as well as teachers and researchers on dipex.ch. The project aims at harnessing the power of patient voices to stimulate improvement. A systematic collection of evidence offers the chance to remedy lack of patient representation and to include low threshold exposure to patient experience into teaching materials that complement bedside teaching. We provide opportunities for interprofessional and interdisciplinary collaboration with other health professionals and aim at stimulating and establish the dialogue between health and the arts and design within the Medical Humanities.
Further Information: Institute of Biomedical Ethics and History of Medicine & dipex.ch
Viktor von Wyl & Milo Puhan
Rationale
Multiple sclerosis (MS) is a severe neurodegenerative health condition characterized by both extensive and highly variable corporal and mental symptomatology. In addition to the extensive symptomatology, the unpredictable and often idiomatic nature of MS is, in and of itself, a substantial source of distress for persons with MS. Everyday coping with symptom management, sudden exacerbations, and the general unpredictability of MS symptomatology often represent an overwhelming challenge impacting on quality of life and mental health. However, standard treatment protocols for MS are typically based on the practitioner’s perspective and center on specific MS symptoms and disease progression. Understanding how persons with MS themselves perceive the progression of their disease as part of their personal, evolving life story may shed light on novel avenues for tailoring MS treatment to individual needs and optimize support systems. The present research thus examines which life events persons with MS perceive as crucial in terms of their disease across their life course, how such key events resonate emotionally in hindsight, how they coped with them, and what they would have needed.
Database
“My life with MS” is a survey which has been developed in a joint effort by scientists, health care professionals, and persons with MS. The survey is designed to assess how persons with MS perceive the progression of their disease as part of their evolving life story in terms of, e.g., key events, emotional impact, and unmet needs. Participants are being recruited from the Swiss MS Registry which is an ongoing longitudinal patient-centered survey in Switzerland funded by the Swiss MS Society. The sample currently comprises about 950 persons diagnosed with MS. Moreover, the study’s embedding into the Swiss MS registry survey allows linkage of text data with the longitudinal quantitative registry assessments for validation purposes and deeper insight.
Methods
The present study implements natural language processing methods to preprocess and subsequently extract meaning from large-scale text data. Specifically, we use language models and topic modelling in Python for text translation, determination of latent topic clusters, and sentiment determination (e.g., using the Python ‘transformers’ library). Further, we link text data to longitudinal registry data using quantitative modelling techniques.
Significance and expected impact
The innovative nature of the present study concerns the extraction of individual perspectives of persons with MS at a large scale using natural language processing methods and their subsequent combination with longitudinal registry data. By inviting registry participants to tell their unique story in their own words, the study gives voice to persons with MS and captures a myriad of unique perspectives. Given the richness of data and the methodological novelty, the present study might spark new hypotheses and serve as a precedent for future research into this matter.
Further information: Viktor von Wyl
https://issuu.com/schweizerischemsgesellschaft/docs/forte_2020_02_de
Anke Maatz & Yvonne Ilg
in cooperation with Henrike Wiemer
Background
Talking about the experience of mental illness plays a central role in psychiatric-psychotherapeutic diagnostics and treatment. Talking about mental illness outside the clinical context, e.g. with friends and employers, has recently also been promoted as it is thought to foster destigmatization, social inclusion and personal recovery. At the same time, it is recognized that talking about mental illness is difficult and often thwarted. Linguistic challenges that lie with verbalizing and communicating about mental experience are crucially involved here: For example, there is no standard vocabulary for these experiences that are essentially private, often new and bewildering. Also, language use is highly regulated by social norms incl. stigma and taboo, which easily impede communication. To exploit the positive potential of talking about mental illness, systematic insights about how we can talk about mental illness despite these challenges are needed.
Aims
The project combines psychiatric and linguistic expertise to provide theoretical insights and practical know-how on the verbalization and communication of mental illness experience. It identifies communicative strategies - both verbal and non-verbal - that enable communication on mental illness.
Methods
Face-to-face conversations on mental illness with different stakeholders (service users, relatives, professionals) and in different contexts (clinical and non-clinical) are video recorded. For all interlocutors, their experience of the conversation is assessed using qualitative and psychometric measures. The analysis combines qualitative (content analysis, conversation analysis) and quantitative (statistical analyses of psychometric data) methods in an interactional paradigm. Thereby, communicative patterns are identified and their suitability in mastering the above-mentioned communicative challenges is assessed. A participatory design is employed i.e.‘experts by experience’ are involved in all stages of the research process.
Impact
The results serve as building blocks for a theory of communication on mental illness in clinical and non-clinical settings. The results are also made available as a communicative resource for professionals, service users, their relatives and the public, they can inform professional training and mental health campaigns. The project thus contributes to the destigmatization of mental illness, fosters social inclusion and personal recovery.
Further Information: Anke Maatz & Yvonne Ilg
Marilia Carvalho Sampaio, Jörg E. Bohlender & M. Brockmann-Bauser
Summary
PRAAT is a free software measuring acoustic characteristics of the human voice, and is widely applied in scientific research. To date the evidence base for the clinical application to characterize and diagnose voice disorders is limited. One of the obstacles of using PRAAT in clinical routines is the lack of reliable reference values of the vocal measures provided by PRAAT. Thus, the normalization and validation of these acoustic parameters may offer an important contribution to their clinical use, and may facilitate better data comparability to research studies.
In a cross sectional study recordings of voice patients with a variety of voice disorders will be assessed using PRAAT. Normative values for irregularity parameters such as jitter and shimmer and harmonicity parameters such as Harmonics-to-Noise Ratio will be derived considering known influencing factors including profession, gender, age, vocal sample, vocal intensity, type of vowel and fundamental frequency. Further, their accuracy to detect pathology and the relation to perceptual assessments and subjective voice findings will be investigated.
Collaboration
Independent research project between the Phoniatrics and Speech Pathology Research Group, Department of Phoniatrics and Speech Pathology, Clinic for Otorhinolaryngology, Head and Neck Surgery, University Hospital Zurich in cooperation with Prof. M. Carvalho Sampaio, Prof. M. L. Masson, Prof. M. F. De Paula Soares, Federal University of Bahia, Dept. of Speech-Language and Hearing Sciences, Salvador-BA, Brazil.
Funding from the Federal University of Bahia, Dept. of Speech-Language and Hearing Sciences, Salvador-BA, Brazil.
Further Information: USZ
Sophie Witt
Body representations? Scenic understanding? Staging and expressive gestures? The subject of the SNSF PRIMA project is the area of overlap between theatre and medicine: while the self-description of psychosomatics as a medical discipline still makes frequent use of theatre metaphors today, tragedy, conversely, has invoked affects since antiquity, i.e. those mostly violent emotions that can be defined as holistic psychophysical moods. The project takes this overlap as the starting point for a historical and systematic investigation of the connection between the psychosomatic and the theatrical.
Further Information: Webseite & SNF
Nathalie Giroud
Summary
Switzerland as one of the countries with the oldest population on earth faces tremendous challenges dealing with the health care and societal costs of the high prevalence of Alzheimer’s dementia (AD). With the predicted increases in AD-prevalence in the next few years in Switzerland and around the world, researchers are at a crucial point understanding that AD-research needs to broaden the focus from AD-treatment to AD-prevention. In this line of thinking, a recent key paper (Livingston et al., 2017) has argued that a significant decrease of AD-prevalence may be achieved by identifying and treating modifiable AD-risk factors.
Dr. Giroud’s research program will identify novel ways to assess and characterize conditions associated with a modifiable risk factor which has been estimated to have the highest potential impact on AD-risk reduction, namely age-related hearing loss, which is highly prevalent among older adults. Using neuroimaging techniques with high temporal resolution such as the electroencephalogram (EEG), neural processes during the auditory processing of complex sounds (e.g. human speech carrying linguistic information) will be investigated in real-time in older adults who are at risk for developing AD or who have already received an AD-diagnosis. The results will provide key insights into basic scientific questions regarding the neurobiology of aging and the neurobiology of language as well as outline methods for measuring hearing impairment for clinical purposes.
Further Information: Nathalie Giroud & ICL
