Back

A comparison of in-person and telehealth treatment modalities using the SpeechVive device

Covert, R.; Snyder, S.; Lambert, A.; Spremulli, M.; Brown, B.; Dwenger, K.; Malandraki, G.; McDonough, M.; Brosseau-Lapre, F.; Huber, J. E.

2025-01-17 rehabilitation medicine and physical therapy
10.1101/2025.01.16.25320611 medRxiv
Show abstract

Telehealth is increasing popular as a treatment option for people with Parkinson disease (PD). The SpeechVive device is a wearable device that uses the Lombard effect to help patients speak more loudly, slowly, and clearly. This study sought to examine the effectiveness of the device to improve communication in people with PD, delivered over a telehealth modality as compared to in-person, using implementation science design. 66 people with PD were enrolled for 12 weeks with 34 choosing the in-person group and 32 in the telehealth group. Participants were assessed pre-, mid-, and post-treatment. Participants produced continuous speech samples on and off the device at each timepoint. Sound pressure level (SPL), utterance length, pause frequency, and total pause duration were measured. Psychosocial surveys were administered to evaluate the effects of treatment on depression, self-efficacy, and participation. The in-person group increased SPL when wearing the device while the telehealth group did not. Both groups paused less often while wearing the device. Utterance length increased post-treatment for the telehealth group, but not for the in-person group. An increase in communication participation ratings in the telehealth group, but not the in-person group, was the only significant change in the psychosocial metrics. The in-person group showed similar treatment effects as previous studies. The device was not as effective in the telehealth group. One limitation was data loss due to recording issues that impacted the telehealth group more than the in-person group.

Matching journals

The top 12 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.