STTR Phase I: Speaklear: A Telehealth Application for Evaluating and Treating Pathalogical Speech

Project: Research project

Description

Abnormal speech production is often symptomatic of an existing (or emerging) neurological disorder and often goes untreated until it directly impacts communicative ability. Currently, patients will visit speech-language pathologists for in-person treatment and evaluation. The evaluation is subjective and the treatment is behavioral and repetitive. In person visits, however, are expensive and impossible for those in rural areas or with ambulatory problems. Furthermore, subjective assessment can vary from listener to listener (and even for the same listener at different time points). The proposed product, speaklear, allows patients to remotely record speech and video samples and provides a variety of derived calculations, novel and traditional, to assess the integrity of speech. These objective tools provide access to expertise in rural areas with the goal of reducing health disparities in the field. Objectives and Methods: Exploit recent technical advances in the emerging field of personalized telemedicine, in conjunction with the increasingly widespread use of tablets in clinical practice,
to remotely track abnormal speech production in at-risk individuals and augment human perceptual evaluation through modern data analytics. Realizing this vision requires the development of an advanced signal processing methodology and a suite of mobile applications that implements it in scalable fashion.
Key Words: Speech language pathology, objective speech evaluation, telemedicine applications, predictive algorithms, Mobile Applications, Computational Models, Feature Extraction, Sparsity.
Subtopic SH3: From Data to Decisions.
StatusFinished
Effective start/end date9/29/157/2/17

Funding

  • NSF-ENG: Division of Industrial Innovation & Partnerships (IIP): $121,000.00

Fingerprint

Telemedicine
Pathology
Feature extraction
Signal processing
Health