The effects of acoustic modification of fundamental frequency, intensity and duration on the perception of stress in dysarthria speech

Ijitona, Tolulope and Lowit, Anja and Soraghan, John (2017) The effects of acoustic modification of fundamental frequency, intensity and duration on the perception of stress in dysarthria speech. In: Royal College of Speech and Language Therapists Conference 2017, 2017-09-27 - 2017-09-28.

Full text not available in this repository.Request a copy


Stress marking plays an essential role in conveying meaning and drawing listener's attention to specific parts of a message. Past research has shown that healthy speakers mark stress using three main acoustic cues; pitch, intensity and duration. We are also aware that dysrathric speakers experience problems in manipulating these acoustic cues when marking stress. The relationship between acoustic cues and listeners perception is vital in the development of a computer-based tool that aids the Speech and Language Therapists (SLTs) in providing effective treatment to dysarthric speakers. The aim of this study is, therefore, to investigate the acoustic cues deficiencies in dysarthric speech and the potential compensatory techniques needed for effective treatment. In this study, we investigate the relationship between acoustic features of dysarthric stress production and what is perceived by listeners. We also investigate which acoustic cues (or combination thereof) are the most significant and what degree of change needs to be achieved in order to mark stress effectively to listeners. In this study, we perform perceptual experiments on dysarthric speech which involves acoustic modification of stress marked sentences from 10 speakers with Ataxic dysarthria. Each speaker produced 30 sentences using the 10 Subject-Verb-Object-Adjective (SVOA) structured sentences across three stress conditions. These stress conditions are stress on initial (S), medial (O) and final (A) target words respectively. The sentences were perceptually scored by 5 untrained listeners based on the location of the stress target. We then identified sentences where listeners were not able to identify the target word and chose 15 of them across the 3 sentence conditions for our study. To effectively measure the deficiencies in Dysarthria speech, the acoustic features (pitch, intensity and duration) of the target words, in the selected sentences, were modified incrementally in accordance to the degree of change observed in age and gender matched healthy control group. The modified sentences were replyed back to 50 listeners to evaluate what degree and combination of modifications resulted in correct identification of the stress target. We present the patterns of deficiency observed in dysarthric speech when marking stress as well as the effects of modifying the acoustic cues on the ability of the listeners to identify the stress target. We will also report on the effects of combining these acoustics features modifications (fundamental frequency, intensity and duration). The results of this study will provide information and guidance for future studies on stress production with regards to focus of analysis. This is will also provide an evidence base for intervention guiding Speech and Language Therapists (SLTs) on how best to address the deficits in stress marking when treating their patients. With regards to other speech disorders, the results of this study can also offer some guidance in developing acoustic tools for monitoring and managing of patients during and after treatment.