Components Include:

• Teensy Microcontroller

• TL072 inverting operational amplifier

• AOM microphone

• Voltage Dividers

• Utilized Spectral Gating filter to reduce noise

Filtered Audio Stethoscope Audio for Comparison

Swallow 1: SNR = 30.12 dB Swallow 1: SNR = 24.6 dB

Swallow 2: SNR = 22.78 dB Swallow 2: SNR = 15.1 dB

Swallow 3: SNR = 27.68 dB Swallow 3: SNR = 17.1 dB

• Lightweight, productive, easy access

• 6.28 ounces 

• snap fit lid

• mic and power chord outlets

• latex rubber tubing

• URSA sticky circles for adhesives

• Utilizing Penetration-Aspiration Scale

  • PAS scoring: 1-2 = low risk, 3-8 = high risk

  • Dataset: ~600 swallows from 45 patients

  • Model: Random Forest classifier

  • Pipeline: Inputs audio (multiple swallows), age, and gender

• Inputs

  • Name, Age, Biological Sex, Food texture type, Trial #

• Features

  • Start and stop recording buttons

  • Record background noise button

  • Analyze recording files button

  • Quit button

• Easy to use for both patients and clinicians