Noise-augmented Automatic Speech Recognition for Speech Treatment in Parkinson's Disease

Overview

The goal of this clinical trial is to learn whether noise-augmented feedback using automatic speech recognition improves intelligibility and social participation in individuals with Parkinson's disease. The main questions the investigators will address are: Does noise-augmented feedback lead to increased intelligibility, as measured by transcription accuracy scores? Does a self-managed speech protocol lead to increased social participation, as measured by the Communicative Participation Item Bank and the Communicative Effectiveness Index questionnaires? Participants will: Receive 2-4 initial speech assessments in-person in our lab. Engage in speech treatment sessions at home using their phone and computer four days a week for four weeks. Four of those sessions will be directly monitored online by a speech-language pathologist. The rest of the sessions are completed by the participant. Come to the lab for two post-treatment assessments following completion of treatment. Come to the lab six months after treatment completion for an additional assessment.

The goal of this clinical trial is to learn whether noise-augmented feedback using automatic speech recognition improves intelligibility and social participation in individuals with Parkinson's disease. Data collection will be face-to-face at a convenient time for participants. The screening and assessment sessions will take place in the Communication, Technology, and Language Diversity Lab (Room 952a), on the 9th floor of the 528 Building at Teachers College, Columbia University (525W 120th St, New York, NY 10027). Treatment sessions will be conducted independently by the participant in the participant's home. Four out of the 16 treatment sessions will be supervised online (via Zoom) by a licensed speech and language pathologist. The study involves 5-7 assessment sessions (including one 6-month follow-up) and 16 treatment sessions ( = 1 month). Each participant is expected to be involved in the study for 7-8 months. Participants will first attend a screening session to confirm eligibility, which will involve a hearing screening, a brief cognitive screening, and a brief depression screening. If the screening is successful, participants will complete two or four assessment sessions before initiation of the treatment protocol (the number of assessment sessions depends on their group allocation: immediate treatment or deferred/delayed treatment). Screenings will only be completed once, during their first in person visit on the Teachers College campus. In the first assessment session participants will be asked to complete voice recordings and three questionnaires. Participants will complete the same speech tasks and questionnaires in the rest of the assessment sessions. The number of assessment sessions (5 versus 7) depends on whether they are assigned to the immediate treatment group or the delayed treatment group. Group assignments are random, and participants will always receive the treatment protocol. Each assessment session is estimated to take 60-90 minutes to complete. Participants will complete the treatment sessions on their own using your phone and computer. Participants will be instructed to find a quiet place at home and practice their exercises at a time when they feel ok and not too tired. They will log in onto our web-based app, Understand Me for Life, to start their speech session. App instructions will be provided during the second assessment session for familiarization purposes. Each session will contain 20 short sentences (between 3 and 7 words in length). Participants will be asked to try to complete 5 sessions in 1 hour (a total of 100 sentences), taking as many breaks as they may need. Participants will be instructed to complete 1-hour of speech exercises 4 days a week for 4 weeks. The total number of days to participate in the treatment study is 16.