A Personalized Voice Restoration Device for Patients With Laryngectomy

Not Applicable

Not yet recruiting

• Conditions: Aphonia; Dysphonia
• Interventions: Device: Surface Electromyography

• Registration Number: NCT05479643
• Lead Sponsor: Weill Medical College of Cornell University

Brief Summary

The investigators will conduct a pilot experiment for a novel and personalized method for voice restoration using machine learning applied to surface EMG (sEMG) signal from articulatory muscles of the face and the neck allowing recognition of silent speech. The investigators predict that the use novel personalized method for voice restoration will be feasible and successful for patients.

Detailed Description

This is a prospective pilot study evaluating the feasibility of a personalized voice restoration device and patients' experience with it. Study participation will include a one-time visit where subjects will read passages and phrases. Acoustic and signal data will be captured. Machine learning will be applied to the data to classify words. Subjects will also participate in a qualitative interview about their experience with voice restoration devices.

Study Timeline

• Study First Submitted: 2022-07-25
• Study First Submitted QC: 2022-07-28
• Study First Posted: 2022-07-29
• Status Verified: 2025-01
• Last Update Submitted: 2025-01-22
• Last Update Posted: 2025-01-23
• Study Start: 2025-12-01
• Primary Completion: 2026-12
• Study Completion: 2026-12

Recruitment & Eligibility

• Status: NOT_YET_RECRUITING
• Sex: All
• Target Recruitment: 100

Inclusion Criteria

Group A: Healthy Volunteers

1. Adult subjects, 18 or older
2. Without any voice impairments

Group B: Subjects with Aphonia or Dysphonia

1. Adult subjects, 18 or older
2. Documentation of severe dysphonia and/or aphonia, or a GRBAS score > 0 (GRBAS is a scale that can be used to assess voice quality of subjects who do not have a recorded history of dysphonia or aphonia. The GRBAS scale evaluates for grade, roughness, breathiness, asthenia, and strain).

Exclusion Criteria

• Group A: Healthy Volunteers

1. Voice impairment

Group B: Subjects with Aphonia or Dysphonia

1. Subjects whose face muscles are entirely paralyzed

Study & Design

• Study Type: INTERVENTIONAL
• Study Design: PARALLEL

Arm && Interventions

Group	Intervention	Description
Patients with Aphonia or Dysphonia	Surface Electromyography	Participants with Aphonia or Dysphonia will be asked to recite phrases with sEMG attached to articulatory muscles.
Healthy Volunteers	Surface Electromyography	Healthy Volunteer will be asked to recite phrases with sEMG attached to articulatory muscles.

Primary Outcome Measures

Name	Time	Method
Subject experience with voice restoration devices: Qualitative	10 minutes after the time of intervention	Subject will be asked to complete a semi-structured interview about their experience with voice restoration devices. Categories of response include voice/communication history, experience with previous voice restoration devices, experience with our device, and suggestions for improvement.
Percentage of Words and Numbers Correctly Identified	"During Procedure/use of Surface ElectroMyoGraphy"	Accuracy of sEMG signal will be determined by subjects reciting common phrases. Subjects will recite the "Rainbow Passage" 10 times. Subjects will be asked to count from 1-10 once. Subjects will recite the alphabet once.

Trial Locations

Locations (1)

Weill Cornell Medicine; 🇺🇸 New York, New York, United States