Adaptation of wavelet transform analysis to the investigation of biological variations in speech signals

Abstract

The purpose of this study was to adapt wavelet analysis as a tool for discriminating speech samples taken from healthy subjects across two biological states. Speech pressure waveforms were drawn from a study on effects of hormone fluctuations across the menstrual cycle on language functions. Speech samples from the vowel portion of the syllable 'pa', taken at the low-and high-hormone phases of the menstrual cycle, were extracted for analysis. Initial analysis applied Fourier transforms to examine the fundamental and formant frequencies. Wavelet analysis was used to investigate spectral differences at a more microbehavioural level. The key finding showed that wavelet coefficients for the fundamental frequency of speech samples taken from the high-hormone phase had larger amplitudes than those from the low-hormone phase. This study provided evidence for differences in speech across the menstrual cycle that affected the vowel portion of syllables. This evidence complements existing data on the temporal features of speech that characterise the consonant portion of syllables. Wavelet analysis provides a new tool for examination of behavioural differences in speech linked to hormonal variation.

Vocal Folds Stroboscopic Image Processing for Otolaryngology

“…Its application in bioengineering is very extended [26], and more specifically in applications of medical images like: ultrasound images, tomography images or magnetic resonance images. The capture methods of these images and, therefore, its characteristics are very different to the vocal cord images that are object of study in this chapter.…”

Acoustic investigation of speech pathologies based on the discriminative paraconsistent machine (DPM)

“…The same problem occurs even when modern technologies such as video-laryngoscopy and stroboscopic light are used as some speak-ers may present a reflex action in their supra-glottal cavity, causing wrong assessments. Therefore, non-invasive acoustic analysis of human voices represents an important tool for physicians during the pre-diagnosis of larynx diseases, implying that discrete-time processing of recorded voice signals [4,5] is useful to detect intrinsic characteristics that define normal and pathologically-affected voices, as argued in articles [6][7][8].…”