Matching Items (32)
Filtering by
- Status: Published
Description
Speech is known to serve as an early indicator of neurological decline, particularly in motor diseases. There is significant interest in developing automated, objective signal analytics that detect clinically-relevant changes and in evaluating these algorithms against the existing gold-standard: perceptual evaluation by trained speech and language pathologists. Hypernasality, the result of poor control of the velopharyngeal flap---the soft palate regulating airflow between the oral and nasal cavities---is one such speech symptom of interest, as precise velopharyngeal control is difficult to achieve under neuromuscular disorders. However, a host of co-modulating variables give hypernasal speech a complex and highly variable acoustic signature, making it difficult for skilled clinicians to assess and for automated systems to evaluate. Previous work in rating hypernasality from speech relies on either engineered features based on statistical signal processing or machine learning models trained end-to-end on clinical ratings of disordered speech examples. Engineered features often fail to capture the complex acoustic patterns associated with hypernasality, while end-to-end methods tend to overfit to the small datasets on which they are trained. In this thesis, I present a set of acoustic features, models, and strategies for characterizing hypernasality in dysarthric speech that split the difference between these two approaches, with the aim of capturing the complex perceptual character of hypernasality without overfitting to the small datasets available. The features are based on acoustic models trained on a large corpus of healthy speech, integrating expert knowledge to capture known perceptual characteristics of hypernasal speech. They are then used in relatively simple linear models to predict clinician hypernasality scores. These simple models are robust, generalizing across diseases and outperforming comprehensive set of baselines in accuracy and correlation. This novel approach represents a new state-of-the-art in objective hypernasality assessment.
ContributorsSaxon, Michael Stephen (Author) / Berisha, Visar (Thesis advisor) / Panchanathan, Sethuraman (Thesis advisor) / Venkateswara, Hemanth (Committee member) / Arizona State University (Publisher)
Created2020
Description
Markov Chain Monte-Carlo methods are a Bayesian approach to predictive statistics, which takes advantage of prior beliefs and conditions as well as the existing data to produce posterior distributions of relevant parameters. This approach, implementable through the JAGS packaging in R, is promising for its impact on the diagnostics space, which is a critical bottleneck for pandemic planning and rapid response. Specifically, these methods provide the means to optimize diagnostic testing, for example, by determining whether it is best to test individuals in a certain locale once or multiple times. This study compares the expected accuracy of single and double testing under two specific conditions, a general and Icelandic test case, in order to ascertain the validity of MCMC methods in this space and inform decisionmakers and future research in the space. Models based on this platform may eventually be tailored to the priors of specific locales. Additionally, the ability to test multiple regimes of real or simulated data while maintaining uncertainty widens the pool of researchers that can impact the space. In future studies, ensemble methods investigating the full range of parameters and their combinations can be studied.
ContributorsSuresh, Tarun (Author) / Naufel, Mark (Thesis director) / Panchanathan, Sethuraman (Committee member) / Harrington Bioengineering Program (Contributor) / Barrett, The Honors College (Contributor)
Created2020-05