Content uploaded by Shuenn-Yuh Lee
Author content
All content in this area was uploaded by Shuenn-Yuh Lee on Jun 01, 2019
Content may be subject to copyright.
An Intelligent Stethoscope with ECG and Heart
Sound Synchronous Display
Yu-Jin Lin
1
, Chen-Wei Chuang
1
, Chun-Yueh Yen
1
, Sheng-Hsin Huang
1
, Peng-Wei Huang
1
, Ju-Yi Chen
2
, and
Shuenn-Yuh Lee
1
1
Department of Electronic Engineering, National Cheng Kung University, Tainan, Taiwan
2
Division of Cardiology, Department of Internal Medicine, National Cheng Kung University Hospital, College of Medicine,
National Cheng Kung University, Tainan, Taiwan
Email: jerry71520@gmail.com, ieesyl@mail.ncku.edu.tw
Abstract—This study presents an intelligent stethoscope
that can visualize heart sound signals and can simultaneously
measure human’s electrocardiogram (ECG) and heart sounds.
The proposed stethoscope can be used in hospitals and
telemedicine through an internet of things (IoT) system and a
cloud database. The proposed stethoscope includes three parts,
namely, a front-end device for ECG and heart sound
measurement, a smart device application (APP), and a cloud
server. The ECG-measuring device is designed for single lead
measurement and has low power consumption and IoT-based
design, which can send real-time ECG data to the smart device
APP. Simultaneously, the heart-sound-measuring device
combined with a traditional stethoscope head is used to
measure heart sound signals. This device includes an analog
front-end circuit and a microphone to filter environmental
noises and to record heart sound signals. Heart sound signals
are transmitted to the APP by using a Bluetooth Low Energy
module. The smart device APP can display synchronized and
real-time signals, including ECG and heart sounds. Meanwhile,
those signals are recorded in smart devices and are uploaded to
the cloud server, where doctors and users can diagnose and
monitor healthcare anytime. The cloud server can store
previous signals and can realize telemedicine through a web
user interface. The proposed intelligent stethoscope is applied
on human trials in the National Cheng Kung University
Hospital.
Keywords—electrocardiogram, heart sound, synchronous
display, artificial intelligence of things, wearable device,
smartphone application, cloud server.
I. I
NTRODUCTION
oday, young doctors cannot precisely distinguish first
heart sound (S1) and second heart sound (S2) by using a
traditional stethoscope because auscultation relatively relies
on their experience. Heart murmurs or other abnormal heart
sounds cannot be easily identified when doctors cannot
distinguish S1 and S2.
Doctors can diagnose several cardiac diseases, such as
atrial or ventricular septal defect, valvular stenosis, valve
regurgitation, valve cleft, arteriovenous malformation, and
hypertension, with the discrimination of heart murmurs.
According to statics [1], that prevalence of valvular heart
diseases increases with age, which is from 3.7% in 18–74
years old to 13.3% in 75 years and older group. Moreover,
the incidence of congenital valvular defect is approximately
10% of congenital heart disease [2]. However, only
experienced doctors can well diagnose the above symptoms
by auscultation. Heart sound auscultation is a subjective
method, and heart sound patterns are difficult to verbally
describe. Doctors who are not cardiologists and trained
medical students face an immense challenge when they
conduct heart sound auscultation.
The visualization and quantization of heart sound signals
can provide doctors a specific diagnosis reference. This
concept reduces the burden of cardiologists and aids medical
students and inexperienced doctors in heart sound
auscultation training.
II. S
YSTEM
O
VERVIEW
Although several digital stethoscopes have been proposed
[3] [4] [5], all of them lack electrocardiogram (ECG)
measurement, in which they cannot confirm the accuracy of
heart sound signals. In other words, the accuracy of S1 and
S2 cannot be guaranteed by only displaying heart sound
signals. The pattern of heart sounds cannot be defined when
the locations of S1 and S2 are not identified. To address this
problem, this study combines a synchronized ECG signal
with a heart sound signal. The peak value of sound near the
R peak of ECG is labeled as S1, and the peak value of sound
near the T wave of ECG is labeled as S2, as shown in Fig. 1.
Heart sounds can be easily identified based on the
synchronized ECG signal.
To verify the suitability of the proposed intelligent
stethoscope, a traditional stethoscope head is adopted, where
the stethoscope line is replaced by an analog front-end device
with a microphone and a Bluetooth Low Energy (BLE)
module. Furthermore, this study proposes a system where
ECG and heart sound signals can be independently measured
and can be simultaneously displayed on a smartphone, as
shown in Fig. 2. The proposed system can collect ECG and
heart sound signals and can also realize telemedicine based
on internet of things (IoT) and cloud server.
A. Heart sound measuring device
Fig. 3 shows the system block of a heart-sound-
measuring device based on IoT. The system block includes a
traditional stethoscope head, an analog front-end circuit with
microphone, a micro-controller unit (MCU), and a BLE
module. A 3M
TM
Littmann® Master Cardiology
TM
Stethoscope’s head is used to ensure sound source quality.
The analog front-end includes an electret microphone, a
T
Fig 1. Association of ECG and Heart Sound
transimpedance amplifier (TIA), a second-order low pass
filter, and a low-dropout (LDO) regulator.
The electret microphone is used due to its sensitivity,
environmental noise interference, and heart sound frequency.
First, heart sounds are mainly produced by the heart’s valves
and are transmitted through human’s thoracic cavity. To
amplify small heart sounds, stethoscope heads are designed
with a special diaphragm structure that can enhance the
vibration of heart sounds.
However, enhanced heart sounds are still relatively
lower than that of normal sounds in our daily life. First, an
electret microphone, such as an electromagnetic
microphone, is adopted rather than a common microphone
to accurately collect heart sounds. Second, a tight tube is
used to connect the stethoscope head and electret
microphone to maintain clean and original heart sounds. The
junction between the microphone and tube is complete
sealed to prevent the microphone from collecting
environmental noises. Third, heart sound frequency is set to
20–200 Hz [5]. Compared with microelectromechanical
system microphones, the electret microphone does not
require a high-pass filter in circuit design. The TIA can
transform the output current from the microphone into
voltage, which is easily operated by the following signal
processing and is recorded by using an analog-to-digital
converter (ADC). The low-pass filter applies a second-order
Butterworth filter with the cutoff frequency of 200 Hz. The
circuit is difficult to realize on a small printed circuit board
(PCB) when the filter order is high. Moreover, the values of
resistors and capacitors should be large because the cutoff
frequency is set at 200 Hz, which is lower than that of many
applications. On this basis, a high-order filter may cause
signal distortion at low-frequencies (approximately 20–100
Hz). As previously mentioned, a second-order filter is used.
A 3.3 V LDO is used for the power management of the
entire system.
The MCU is an ARM Cortex M4, and heart sound
signals are recorded by using a 12-bit ADC with sample rate
of 1 kHz. Subsequently, heart sound signals are added with
timestamps and are encoded by MCU. Data are transmitted
to smart devices by using the BLE module.
B. ECG masuring device
12……... 19
Time
Stamp
Verification
Code
0
Verification
Code
18
ADC Data
Fig. 5. Packet format with ECG and heart sound synchronization
Fig. 6. APP structure in the smart device
Fig. 3. System block of heart sound measuring device
Fig. 4. System block of ECG measuring device
Smart Phone APP
User Interface
Bluetooth
Cloud Server
4G
Wi - Fi
Analog Front End
Circuit
ARM
Cortex - M4
BLE 5.0
Module
Heart Sound Measuring Device
ECG Measuring Device
BLE 5.0
Module
ARM
Cortex - M4
Analog Front End
Circuit
WEB
User Interface
Fig 2. System block of intelligent stethoscope
Fig. 7. Cloud server structure
The ECG-measuring device is an IoT-based design, as
shown in Fig. 4. This device includes an analog front-end
circuit, an MCU, and a BLE module. The analog front-end
circuit consists of level shift units, pre-amplifiers, a
differential amplifier, a low-pass filter, a high-pass filter, and
an LDO.
Level shift units are used to control ECG signals to a
positive level. The differential amplifier behind the pre-
amplifier is used to remove the common mode noise from
the human body and to increase the signal-to-noise ratio
(SNR). High-pass and low-pass filters are adopted to remove
undesired noises caused by motion artifact and
electromyogram signal.
Considering power consumption and PCB size, the
ready-made operational amplifier with ultra-low power
consumption and a second-order filter are used in this study.
The MCU is an ARM Cortex M4, and ECG signals are
recorded by using a 12-bit ADC with sample rate of 1 kHz.
Furthermore, ECG signals are added with timestamps and
are encoded by the MCU. Data are transmitted to smart
devices by using the BLE module.
C. Encoding method of heart sound and ECG signals
This study presents an encoding method with timestamps
to synchronize ECG and heart sound signals, as shown in
Fig. 5. Each packet is encoded in 20 bytes, which includes
the first and last bytes for verification, the second byte for
timestamp, and the remaining 17 bytes for recording the
output voltage from ADC. The safety and accuracy of signals
are considered in the encoding method. Moreover, the
proposed method is designed to collect ECG and heart sound
signals when the two devices are connected in the APP.
D. Smart device’s APP
The smart device’s APP is a user interface based on an
iOS system, as shown in Fig. 6. The APP has five main
functions, namely, decoding, digital filtering, signal display,
data storage, and uploading.
A decoder is used to decode the data transmitted from the
MCU. The decoder aligns the data based on the timestamps
after verifying the verification code, and the data are
processed by the digital filter. A 50th-order finite impulse
response filter is implemented to compensate the insufficient
rejection of the analog front-end filter, which can completely
remove noises. Real-time ECG and heart sound signals are
shown by APP’s displaying function. The recording function
can record the data through screen recording or raw data
storage. Users can separately record heart sound signals at
four auscultatory sites, namely, aortic, pulmonic, tricuspid,
and mitral areas. Users can observe the visualized heart
sound signals and can hear heart sounds through headphones
or speakers from the smart phone. Meanwhile, all the data
are encrypted and uploaded to the cloud database via mobile
internet or Wi-Fi.
E. Cloud server
Fig. 7 shows the cloud server that acts as a database and a
telemedicine platform. The cloud server includes a decoder,
MySQL, a web user interface, and an artificial intelligence
(AI)-based analyzing system for big data.
The decoder is used to decode the data transmitted from
the APP. MySQL is used to store and manage the historical
data from each user. The web user interface in the cloud
server can synchronously display real-time ECG and heart
sound signals. Doctors can remotely guide patients in
conducting regular auscultation through video chat APPs,
such as FaceTime and Skype, and doctors can observe
patients’ ECG and heart sound signals on the web user
interface. The platform can provide important benefits to
patients who suffer from chronic diseases. Patients can
reduce considerable waiting time in hospitals and can be
diagnosed in remote places.
The AI-based analyzing system for big data is still under
construction. In the future, a recurrent neural network model
will be applied in this system to identify heart murmurs and
to aid doctors during their diagnosis.
III. E
XPERIMENTAL
R
ESULT
The proposed stethoscope is completely developed, and
is permitted by the Taiwan Food and Drug Administration
(TFDA) of our country to conduct human trials in the
National Cheng Kung University Hospital.
A. Hardware implementation
Fig. 8(a) shows the PCBs of the heart sound measuring
device, which include a hear sound analog front-end circuit
and an MCU. The heart sound size is 21.04 mm*18.63 mm,
and the MCU is circular with a diameter of 25.6 mm. Fig.
(a) (b)
Fig. 8. (a) MCU and heart sound PCBs of heart sound measuring
device (b) The proposed heart sound measuring device
(a) (b)
Fig. 9. (a) MCU and ECG PCBs of ECG measuring device (b) The
proposed ECG measuring device
Fig. 10. Actual intelligent stethoscope measurement
8(b) shows the assembled heart sound measuring device with
a size of 81.69 mm*45.73 mm*23 mm. Fig. 9(a) shows the
PCBs of the ECG measuring device, which include an ECG
analog front-end circuit and an MCU. The ECG measuring
device is circular with a diameter of 25.6 mm. Fig. 9(b)
shows the assembled ECG-measuring device with a size of
85.05 mm*30.36 mm*18.55 mm.
Fig. 10 demonstrates the practical measurement on the
human body. ECG measurement is a single-lead
measurement, and is set on the left-upper part of the chest.
Meanwhile, heart sound measurement can freely move
among the four auscultatory sites. The measured signals are
displayed in the APP, as shown in Fig. 11. The signals can
be easily observed, and S1 and S2 can be identified by using
R peak and T wave as reference. Fig. 12 shows the frequency
analysis of heart sound signals, and the main frequency of
heart sounds is located at 20–200 Hz, which meets the
practical situation.
B. Software development
Fig. 11 shows the APP user interface, which includes all
the above mentioned functions, such as recording signals,
replaying historical data, and uploading data. The proposed
web user interface in the cloud server can display ECG and
heart sound signals, as shown in Fig. 13.
Tables I and II presents the comparison between this
work and other digital stethoscopes. Compared with other
studies, this study synchronously completes the measurement
of ECG and heart sound signals and utilizes the cloud server
and IoT in this system. Compare with traditional digital
stethoscopes that only measure heart sound signals, this
study adds ECG measurement and provides considerable
diagnosis references to doctors.
IV. C
ONCLUSION
This study provides three contributions as follows. First,
the proposed stethoscope can enhance medical usage
efficiency. Doctors in different divisions, such as family
medicine, pediatrics or otolaryngology, can perform precise
diagnosis by using this intelligent stethoscope. Therefore,
this intelligent stethoscope can save medical resources and
can detect cardiovascular diseases in advance. Second,
doctors and trained medical students can enhance their heart
sound auscultation skills by using this intelligent
stethoscope. Third, this intelligent stethoscope can be applied
in telemedicine, which is a convenient home-care device for
elderly people and those who suffer from chronic diseases.
A
CKNOWLEDGMENTS
The authors greatly appreciate the support from the Taiwan Semiconductor
Research Institute, the Ministry of Science and Technology, and Southern
Taiwan Science Park, Taiwan (Grant MOST 107-2218-E-006-034 and AZ-
13-05-28-107).
R
EFERENCES
[1] V. T. Nkomo, J. M. Gardin, T. N. Skelton, J. S. Gottdiener, C. G.
Scott, M. Enriquez-Sarano, “Burden of valvular heart diseases: a
population-based study,” The Lancet, vol. 368, pp. 1005–1011,
September 2006.
[2] LaHaye S1, Lincoln J, Garg V, “Genetics of valvular heart disease,”
Curr Cardiol Rep, 16(6):487, June 2014, doi:10.1007/s11886-014-
0487-2.
[3] C. Aguilera-Astudillo, M. Chavez-Campos, A. Gonzalez-Suarez, J. L.
Garcia-Cordero, “A low-cost 3-D printed stethoscope connected to a
smartphone,” 38
th
Annual International Conference of the IEEE
Engineering in Medicine and Biology Society, August 2016, pp. 4365-
4368.
[4] W. Y. Shi, J. Mays, J.-C. Chiao, “A wireless stethoscope,” IEEE
MTT-S 2015 International Microwave Workshop Series on RF and
Wireless Technologies for Biomedical and Healthcare Applications
(IMWS-BIO), September 2015, pp.197-198.
[5] J. E. Suseno, M. Burhanudin, “The signal processing of heart sound
from digital stethoscope for identification of heart condition using
wavelet transform and neural network,” International Conference on
Informatics and Computational Sciences (ICICoS), November 2017,
pp. 153-157.
[6] Ekohealth, ”DUO Stethoscope + EKG,” Ekohealth.com [Online]
Available: https://ekohealth.com/duo/ [Accessed 2 Feb. 2019].
S1 S2
Fig 13. Proposed web interface screenshot
TABLE
I.
P
EFORMANCE
C
OMPARSION
This work 2016[3] 2015[4]
Measuring Heart sound and
ECG
Heart
sound
Heart
sound
Wireless Bluetooth NO 2.4 G RF
User
Interface
Smart Phone APP
WEB
Smart
Phone APP
PC
Database Cloud N/A N/A
Human trials Yes N/A N/A
TABLE
II.
E
XISTING
P
RODUCTS
C
OMPARSION
This work 3M[1] EKO[2]
Measuring
Heart sound
and
ECG
Heart
sound
Heart sound
and
ECG
Wireless Bluetooth Bluetooth Bluetooth
User
Interface
Smart Phone APP
WEB
PC Smart Phone
APP
Database Cloud Computer Smart Phone
FDA No (TFDA) Yes Yes
S1 S2
Fig. 11. Actual intelligent stethoscope measurement screenshot
Fig. 12. Frequency analysis of heart sound signals
