Collaborative areas and projects
1. Bioacoustic instrumentation:
The search for better tools to quantify and analyze body sounds has been a central topic of BARLAB for many years. Among others, past projects have explored several issues in the design of air-coupled microphones for lung sound recordings, explored transducers for acoustic guidance systems, and signal enhancement methods in the framework of auscultation under very noisy conditions. More recent projects are aimed at the exploration of the response of light-weight accelerometers when used to capture skin vibrations, as well as novel aeroacoustic flow sensors, both with applications in speech production.
2. Numerical modeling:
In order to better understand the phenomena associated with the transmission of body sounds, we have developed and continue working on a number of numerical models of sound transmission in biological tissues and organs. Our latest efforts are devoted to developing time and frequency domain based models of acoustic transmission in the subglottal system that can be integrated within the framework of acoustic coupling with the vocal tract and the non-linear glottal interactions observed during speech production.
3. Development of clinical applications:
Body sounds contain significant information related to the underlying complex physiological phenomena in which they are generated. Transforming this information into clinical tools is one of the ultimate goals of the laboratory. In some of our past projects we have developed systems such as acoustical guidance and position monitoring system for endotracheal tubes, respiratory airflow monitoring using structure-borne recordings in the external ear, methods for acoustic imaging of the human chest, among others. Our current efforts are devoted to investigate the acoustic coupling between supraglottal and subglottal systems by means of simultaneous measurements of skin acceleration, high-speed videoendoscopy, oral volume velocity, intraoral pressure, electroglottography, and radiated acoustic pressure during voice production. This acoustic coupling has been shown to adversely affect the continuous monitoring of vocal function that uses measurements of neck skin acceleration. To overcome this problem, a model-based inverse filtering scheme that considers the effects of acoustic coupling is being developed and tested. Some of these efforts have been also translated into a new method for auditory biofeedback in Parkinson’s disease patients (see note below).
Current notes and news
- Voice monitoring for PD patients: Parkinson's can leave its victims afflicted with certain speech motor problems. We have developed a device that could help Parkinson's sufferers articulate their thoughts more audibly by exploiting the Lombard effect, a reflex in which people automatically speak louder in the presence of background sound. The device includes an earpiece that automatically plays ambient noise whenever a person speaks by means of a light weight neck accelerometer. See full article in Scientific American and Purdue news.
- Matías Zañartu (PhD candidate from ECE) received the Best Student Paper Award in Speech Communication at the Acoustical Society of America Meeting in Portland for the poster entitled "An impedance-based inverse filtering scheme with glottal coupling" presented in the special session on source-filter interactions in biological sound production. The paper was co-authored by Julio Ho from BME at Purdue, Daryush Mehta from MIT-Harvard, Robert Hillman from MGH/MIT-Harvard, and George Wodicka from Purdue. The project is part of the collaboration between Prof. Wodicka and Prof. Robert Hillman, Research Director of the Center for Laryngeal Surgery & Voice Rehabilitation at the Massachusetts General Hospital in Boston, MA . See spotlight in ECE.