ECE 445 Spring 2016 ELECTRONIC STETHOSCOPE ARRAY Groupe

ECE 445 Spring 2016 ELECTRONIC STETHOSCOPE ARRAY Groupe 70 Robin GUIGNY & Fatima Zohra HASNAOUI

INTRODUCTION 2/5 medical misdiagnosis & underdiagnosis per year Traditional stethoscopes only provide qualitative diagnosis Find a way to provide a more accurate diagnosis

Objectives Have a quantitative assessment of lung activity Capturing and processing the lung signal to identify symptoms Easy-to-read output for the doctor

Overall System Patient Sound capture&treatment Signal analysis and diagnosis

Table of contents 1. 2. 3. 4. 5 6. 7. Block diagram Microphone Filter Bluetooth module Headphone amplifier Power supply Software

Block diagram

Choosing the microphone Sensitivity and frequency response Cost Interferences Power

Microphones comparison MEMS Sensitivity & freq response Condenser >100 Hz Cost $400 Interferences Power Very bad Acceptable Good Fiber optic Final choice!

Choosing the filter High attenuation in the stopbands Flat passband Easy design for multiple modifications Requirements: • Attenuation < 6 d. B in the passband [50 Hz ; 2500 Hz] • Attenuation > 12 d. B in [0 - 30 Hz] and 6000 Hz +

Filter design Butterworth filter : maximally flat in the passband. 3 rd order : to obtain desired attenuation in the stopband. Sallen-Key design : easy to implement and modify.

Identification Butterworth Q=1/√ 2 2 nd cutoff frequency : f=2500 Hz

Schematics

LTSpice Bode diagram

Results

Attenuation < 6 d. B passband Vs/Ve > 1/2 at 2500 Hz and at 50 Hz ✔

Attenuation > 12 d. B Vs/Ve < 1/4 for 6000 Hz+ ✔ and [0 – 30 Hz]

Bluetooth module choice Analog input Resolution > 8 bits

Numato’s GPIO BT Module 10 bits resolution 3, 3 V ADC range RN-42 2. 4 GHz Bluetooth module Command through Hyper. Terminal

Headphone amplifier design Able to drive low-impedance output found in headphones (9Ω to hundreds of Ω). Requirement Provide >0, 5 m. W to 9Ω earphones (96 d. BSPL/m. W) for the loudest expected sound. This corresponds to 67 m. V output.

Cmoy audio amplifier

Results X 100 amplification ≈70 m. V Pk-Pk for loudest sound ->93 d. BSPL : tolerated for 8 hours

Power unit Relatively small in-use time (estimated <30 min/day) No particular need for rechargeable battery. Able to provide +/-9 V to all the opamps Requirement The battery must allow 10 hours of use (last about a month)

Power supply schematics

Battery life 2 x 9 V alkaline batteries + 3 V coin cell battery Battery life calculation : LT 082 opamps : 2, 5 m. A Absolute max for BT module : 95 m. A

Software overview Aim : visualize lung sounds and spot symptoms How? According to lung sounds features (FFT, shape, length) Two main abnormal sounds are linked to lung diseases: ü Wheezes ü Crackles Fine crackles Coarse crackles

Lung sounds features Sounds Max peak frequency Waveform Duration Normal 200 Hz • Breath in = Cycle*2/3 • Breath in = breath out + 11 d. B 2 s<cycle length<10 S Wheezes 400 Hz Sinusoid >80 ms Fine Crackle 650 Hz Dampened wave deflection Around 5 ms Coarse Crackle 350 Hz Dampened wave deflection Around 15 ms Focus of the software

Steps Pick the sound Press a button Compute fft Find max peak’s frequency (f>200 Hz) Conclusion

Flowchart

User interface Scroll bar

Example

Requirements Latency : 2 to 10 s 95% success

Verifications Latency : ok ✓ 95% success rate : ok on lung sounds files provided by medical research articles -> need for a bigger sample

Issues Bluetooth : not able to connect the bluetooth module Analog identification can be possibly implemented if we take into account localization : A) Fine crackles B) Wheezes C) Coarse crackles Lung sound preferential spots

Conclusion Hardware Software Link between the two?

Further work Bluetooth transmission between hardware and software Add microphones (multiple inputs) Take microphone localization into account Try with other type of filters (FIR. . ) Design stethoscope head for better acoustic clarity

Thank You!