Project
# | Title | Team Members | TA | Documents | Sponsor |
---|---|---|---|---|---|
79 | Voice Dosimeter for Voice Therapy |
David Gong Jaden Li Michael Rizk |
Chi Zhang | proposal1.pdf |
|
# Voice Dosimeter for Voice Therapy Team Members: - David Gong (dsgong3) - Jaden Li (sizhel2) - Michael Rizk (rizk2) # Problem The US societal costs of voice-related teacher absenteeism and treatment expenses alone have been estimated to be as high as 2.5 billion dollars annually. Such absenteeism could be prevented if teachers were able to measure how much they were using and straining their voices every day. Furthermore, there are currently no commercially available voice dosimeters. Some devices were available in the past, but they cost thousands of dollars. A low-cost and widely available voice dosimeter would allow for clinical and research use of voice-related problems and voice therapy options. This project will be conducted in collaboration with graduate student Charlie Nudelman and Professor Pasquale Bottalico at the College of Applied Health Sciences. Their group had previously developed a DIY voice dosimeter using a contact microphone and a portable audio recorder. They are still using this device today, but there are a few improvements they would like to see. # Solution The device that Charlie and Pasquale are using is bulky and requires a wired connection. It is impractical for patients to wear daily and collect data for a long period of time. We aim to create a cheaper and more comfortable voice dosimeter that is capable of recording data for long periods of time without recharging while the data can be uploaded to another device wirelessly. # Solution Components ## Sensor System Accelerometer: Low power and need bandwidth of at least 3kHz (BMA580) Medical Tape Silicone enclosure ## Microcontroller Microcontroller: SOC with BLE low power consumption (NRF52832-QFAB-R) ## Data Processing and User Interface The recorded data is collected and uploaded to a device through bluetooth. This data is then processed to extract information about sound pressure level, fundamental frequency, and cepstral peak. Finally, this data can be viewed either through a website or an app. ## Power Battery: enough power to last 8 hours given MCU, power consumption estimate - Either a small rechargeable battery, or a button cell battery that can be replaced PMIC: TPS65720EVM-515 # Criterion For Success - Design a device that costs less than $200 - Accurately measure within some uncertainty compared to a microphone measurement in a quiet room: sound pressure level within 2dB, fundamental frequency within 5Hz, cepstral peak within 2dB - Reject external noise and voices by at least 20 dB compared to the wearer’s voice. - At least 8 hours of battery life ## Secondary Objective - Cloud integration |