Our previous studies have generated the first short- and long-term experimental efficacy data for speech treatment for individuals with Idiopathic Parkinson disease (IPD), documenting acoustic, aerodynamic and physiologic changes accompanying treatment. Despite these major steps in establishing the experimental efficacy of speech treatment for individuals with IPD, the """"""""real world"""""""" treatment of speech remains an unmet need for the vast majority of these individuals. At least 89% of individuals with IPD have disordered speech, but only 3-4% receive speech treatment. This is a serious problem, as individuals with IPD who have received Lee Silverman Voice Treatment (LSVT) therapy benefit in many ways: they not only produce more intelligible and expressive speech, but they often report dramatic improvements in their self esteem and quality of life. We thus propose to conduct research to demonstrate the feasibility 6f a novel solution to make LSVT training more accessible to over one million individuals--the development and testing of a fully automated computer-based speech therapy system, in which voice training is conducted by a virtual speech therapist--and animated computer character that models the key perception and production behaviors of a human therapist. While development of the proposed system may seem futuristic and ambitious at first blush, perceptive animated agents are now a reality, and language training applications using these agents have produced significant gains in speech and language skills of profoundly deaf children. In addition, under a seed grant from the Coleman Foundation, we have developed an initial version of an LSVT therapist that models well the facial expressions, gestures and utterances of a human LSVT therapist. It is thus feasible to develop and assess the potential value of an automated training system that uses a perceptive animated agent to conduct voice training with individuals with IPD. The proposed work builds upon our well established foundation of experimental efficacy data and the successful application of perceptive animated agents in language training tasks to accomplish two specific aims: (1) design and test a fully automated voice training system in which a virtual therapist conducts LSVT sessions with individuals with IPD; and (2) explore the feasibility of using this system as either a substitute or adjunct to LSVT training with human clinicians. In addition to assessment procedures used in previous LSVT efficacy studies, the proposed work will provide new insights into LSVT treatment through sophisticated analyses of audio and video data collected during sessions and pre and post treatment, that employ computer vision and speech signal processing algorithms to analysis of facial movements, gestures and speech data. ? ?

Agency
National Institute of Health (NIH)
Institute
National Institute on Deafness and Other Communication Disorders (NIDCD)
Type
Exploratory/Developmental Grants (R21)
Project #
5R21DC006078-02
Application #
6885764
Study Section
Special Emphasis Panel (ZRG1-MESH (02))
Program Officer
Shekim, Lana O
Project Start
2004-04-15
Project End
2007-03-31
Budget Start
2005-04-01
Budget End
2007-03-31
Support Year
2
Fiscal Year
2005
Total Cost
$205,957
Indirect Cost
Name
University of Colorado at Boulder
Department
Psychology
Type
Other Domestic Higher Education
DUNS #
007431505
City
Boulder
State
CO
Country
United States
Zip Code
80309
Mahler, Leslie A; Ramig, Lorraine O; Fox, Cynthia (2015) Evidence-based treatment of voice and speech disorders in Parkinson disease. Curr Opin Otolaryngol Head Neck Surg 23:209-15