Conference PaperPDF Available

A Panoramic Study of Obstructive Sleep Apnea Detection Technologies

Authors:

Abstract and Figures

This study offers a literature research reference value for bioengineers and practitioner medical doctors. It could reduce research time and improve medical service efficiency regarding Obstructive Sleep Apnea (OSA) detection systems. Much of the past and the current apnea research, the vital signals features and parameters of the SA automatic detection are introduced.The applications for the earlier proposed systems and the related work on real-time and continuous monitoring of OSA and the analysis is given. The study concludes with an assessment of the current technologies highlighting their weaknesses and strengths which can set a roadmap for researchers and clinicians in this rapidly developing field of study.
Content may be subject to copyright.
A Panoramic Study of Obstructive Sleep Apnea
Detection Technologies
Laiali Almazaydeh, Khaled Elleithy and Miad Faezipour
Department of Computer Science and Engineering
University of Bridgeport
Bridgeport, CT 06604, USA
{lalmazay, Elleithy, mfaezipo} @bridgeport.edu
AbstractThis study offers a literature research reference value
for bioengineers and practitioner medical doctors. It could
reduce research time and improve medical service efficiency
regarding Obstructive Sleep Apnea (OSA) detection systems.
Much of the past and the current apnea research, the vital signals
features and parameters of the SA automatic detection are
introduced.The applications for the earlier proposed systems and
the related work on real-time and continuous monitoring of OSA
and the analysis is given. The study concludes with an assessment
of the current technologies highlighting their weaknesses and
strengths which can set a roadmap for researchers and clinicians
in this rapidly developing field of study.
Keywords: sleep apnea, PSG, ECG, EEG, SpO
2
.
I. INTRODUCTION
Sleep apnea (SA) in the form of Obstructive sleep apnea
(OSA) is becoming the most common respiratory disorder
during sleep, which is characterized by cessations of airflow to
the lungs. These cessations in breathing must last more than 10
seconds to be considered an apnea event. Apnea events may
occur 5 to 30 times an hour and may occur up to four hundred
times per night in those with severe SA [1].
The most frequent night symptoms of SA can include
snoring, nocturnal arousals, sweating and restless sleep.
Moreover, like all sleeping disorders, symptoms of sleep apnea
do not occur just during the night. Daytime symptoms can
range from morning headaches, depression, impaired
concentration and excessive sleepiness which cause mortality
from traffic and industrial accidents. However, these symptoms
are not definitive to detect SA syndrome [2] [3].
In fact, SA is not a problem to be taken lightly, since it is
associated with a major risk factor of health implications and
increased cardiovascular disease and sudden death. It has been
linked to irritability, depression, sexual dysfunction, high blood
pressure (hypertension), learning and memory difficulties, in
addition to stroke and heart attack [2][3].Several treatments
options for OSA patients include weight loss, positional
therapy, oral appliances, surgical procedures and continuous
positive airway pressure (CPAP). CPAP is a common and
effective treatment especially for patients with moderate to
severe OSA. CPAP devices are masks worn during sleep that
improves oxygen saturation and reduces sleep fragmentation
[4].
However, statistics show that around 100 million people
worldwide, where in the US from 18 to 50 million people, are
suspected to have OSA, of which more than 80% remain
undiagnosed [5]. The trouble of having examinations
discourages patients prone to OSA undergo at the overnight
clinical research through polysomnographic data.
Polysomnography (PSG) is a complicated procedure and
certain way of assessing the OSA problem. Complete PSG
includes the monitoring of the breath airflow, respiratory
movement, oxygen saturation (SpO
2
), body position,
electroencephalography (EEG), electromyography (EMG),
electrooculography (EOG), and electrocardiography (ECG)
[6].However, PSG has received many criticisms from some
researchers. This is due to several reasons, including first, the
inconvenience since it requires the patient to be connected to
numerous sensors and to stay in hospital for one night. Second,
it is expensive. The average cost for a PSG is $2,625 due to the
need for the study to take place in a specially equipped medical
facility, in addition to the requirement of having a sleep lab
staff overnight, trained in „scoring‟ the resultant measurements
manually. Third, along wait list of up to 6 months is caused by
limited availability of PSG [7].
According to the American Academy of Sleep Medicine
(AASM), the Apnea-Hypopnea Index (AHI) is used to describe
the number of complete and partial apnea events per hour of
sleep and it is calculated to assess OSA syndrome severity.
OSA severity is usually determined as follows: AHI 5-15
indicates mild, 15-30 indicates moderate and over 30 indicates
severe OSA syndrome. Therefore, patients are diagnosed with
OSA if they have five or more apnea events per hour of sleep
during a full night sleep period [8].
However, new simplified diagnostic methods and
continuous screening of OSA is needed, in order to have a
major benefit of the treatment on OSA outcomes. In this study,
we investigate the different current apnea researches which
provide an alternative to the expensive PSG visual scoring
method which is commonly used today to assess a patient‟s
sleep quality.
Much of the current apnea research is ranging from off-line
computer-based systems for automatic evaluation of apneas by
analyzing different signals stored in PSG records to
comprehensive portable real-time devices that enable the
patient to be diagnosed and receive feedback at home in order
to alert of the apnea event and help the patient to recover.
In the following sections, we glance at a variety of sleep
apnea detection and monitoring methods. Then, we conclude
our paper in section V, and highlight some directions for future
research.
II. COMPUTER AIDED SLEEP APNEA DETECTION
Over the past few years most of the related research has
focused on presenting methods for the automatic processing of
different statistical features of different signals such as thorax
and abdomen effort signals, nasal air flow, oxygen saturation,
electrical activity of the heart (ECG), and electrical activity of
the brain (EEG) for the detection of SA.
The validated database used to assess the detection
algorithms in the related research studies is supplied online
from the PhysioNet web site [9]. The Apnea Database available
in PhysioNet has been created to support such studies. All
apneas in the recordings are either obstructive or mixed and it
does not contain episodes of pure central apnea or of Cheyne-
Stokes respiration. The full overnight polysomnogram
recordings were divided into a set of one-minute segments.
Each segment was annotated based on visual scoring of
disordered breathing during sleep and if at any time during that
minute there was evidence of sleep apnea the segment was
classified as “apnea”; otherwise it was classified as “normal”.
Segments containing hypopneas were also classed as apnea
[10].
Several studies show that the brain waves signal,
Electroencephalogram (EEG), which indicates states of mental
activity ranging from concentrated cognitive efforts to
sleepiness [11], is able to diagnose SA. Wavelet transforms and
an artificial neural network (ANN) algorithm were applied to
the EEG signal in [12] to find a solution to the problem of
identifying SA episodes. The EEG signals can be classified into
four frequency bands of basis waves, namely as delta (δ), theta
(θ), alpha (α) and beta (β). When an episode of SA occurs, the
EEG signal shifts above the delta frequency band. Then, sleep
EEG activity shifts from a delta wave to theta and alpha waves
frequency bands in the range of 4~14 Hz once episode of SA
ends. The system‟s identification results achieved a sensitivity
of approximately 69.64% and a specificity of approximately
44.44%. However, even though this study yielded promising
initial results, it still requires improvement since the EEG
signal characteristic of a SA is easily contaminated by artifacts,
therefore, a preprocessor circuit is needed to eliminate EEG
signal artifacts and enable the system to recognize SA episodes.
Many studies show that detection of OSA can be performed
through the Electrocardiogram (ECG) signal due to cyclic
variations in the duration of a heartbeat. This consists of
bradycardia during apnea followed by tachycardia upon its
cessation [13]. In our previous published research, we
developed a modelbased on a linear kernel Support Vector
Machines (SVMs)using a selective set of RR-interval features
from short duration epochs of the ECG signal. The results show
that our automated classification system can recognize epochs
of SA with a high degree of accuracy, approximately 96.5%
[14].
Arterial oxygen saturation (SpO
2
) measured by pulse
oximetry can be useful in OSA diagnosis as clinical experience
indicates that an apneic event is frequently accompanied by a
fall in the SpO
2
signal (oxygen desaturation) [15]. The study
carried out by Marcos et al. [16] provided an automated means
for interpretation of SpO
2
signals, based on (LD) classifier and
the linear combination of spectral and nonlinear features from
SpO
2
recordings using principal component analysis (PCA). In
addition, in [17] we further developed a Neural Network(NN)
as a predictive tool for OSA using the SpO
2
signal features and
evaluated its effectiveness.
It has been reported that, the snoring is a common finding
in people with OSA. OSA is generally caused by a blocked of
the airflow airway. Therefore, the snoring must become due to
the vibration of soft tissues when the airflow stimulates the ill
structure in the upper airway during sleep [18]. Of all methods
for diagnosing OSA, the formants estimation method is most
widely used. The formants information contains the essential
acoustic properties of the upper airway. It has been discovered
by studies that there is a correlation between the state of the
upper airway and the first formant frequency. A narrower upper
airway is usually led to a higher first formant frequency.
Therefore, Andrew et al. [19] and [20] proposed fixed formant
frequency thresholds to detect the hypopneic snores which
must be higher than that of the typical ones.
Recently, based on the tracheal breathing sounds recording
analysis during sleep, which can be used for respiratory flow
estimation and distinguish the changes in breathing pattern
recognition of the patient, the study in [21] reports a new fully
automatic technology for OSA detection. Different parameters
were investigated to distinguish the breathing level during each
individual apnea event, including the total energy of the breath
sound segments. After collection of data, each parameter was
then fuzzified with a sigmoid function and the fuzzy output of
the fuzzy functions are added together to classify the sound
signals. The results show high sensitivity and specificity values
of more than 90% in differentiating normal respiration from
disordered breathing patients.
In the present studies, the researchers provide
complementary information with combined different
physiological signals, in order to obtain additional information
to that provided by classical methods to evaluate sleep quality
and detect apnea. In some studies, ECG and SpO
2
data have
been bridged to analyze sleep data. As the blood oxygen
saturation falls during apnea, the resultant increase in heart rate
and blood pressure causes stress and potential injury to the
parts of the cardiovascular system [22]. In [23], the authors
analyze various feature sets and a combination of classifiers
based on the arterial oxygen saturation signal measured by
pulse oximetry (SpO
2
) and the ECG. In this work, the Bagging
with REP Tree classifier achieved sensitivity of 79.75%,
specificity of 85.89% and overall accuracy of 84.40%.
Because of the desaturation event that activates the
sympathetic nervous system, the relationship between periodic
changes in the SpO
2
profile and in the EEG pattern due to
apnea events during the night was investigated in [24]. The
combined spectral analysis of these two signals achieved 91%
sensitivity, 83.3% specificity and 88.5% accuracy in OSA
diagnosis.
The first successful preliminary attempts to directly assess
the interactions of power spectral of sleep EEG and ECG
signals in detecting OSA events is presented in [25].
Consistency between these two signals over different frequency
bands (0-128 Hz) were evaluated before, during and after an
OSA terminations event (with/without arousals) in non-REM
as well as REM sleep.
III. HOME-RECORDING FOR SA DETECTION
Nowadays, much of the current apnea research is being
done on providing portable devices that monitor those
experiencing apnea during the day. The device could act as an
inexpensive and convenient way for doctors to diagnose SA
patients and as a means for collecting data on apnea sufferers to
determine the severity of the condition once diagnosed. More
specifically, this may help in the initial assessment of patients
with suspected OSA in order to prioritize patients. Patients with
utmost need of treatment will go through complete PSG
recordings within a sensible time frame; meanwhile those who
are free of apnea symptoms will avoid the cumbersome
procedure [26].
Various portable monitor devices already exist in the
market. ApneaLink
TM
Plus Home Sleep Apnea Test Device is
one of the carriage able in home sleep test diagnostic devices
that records up to four channels of information: respiratory
effort, pulse, oxygen saturation and nasal flow. The patient can
sleep normally while ApneaLink
TM
Plus monitors his/her sleep,
checking breathing patterns and the amount of oxygen in
his/her blood and recording possible apneas or other breathing
abnormalities [27]. Also, SleepStrip
TM
may be a simple and
effective tool for OSA diagnostic strategy. This device has to
be worn for a minimum of five hours of sleep, and the actual
device is placed on the individuals face where the two flow
sensors (oral and nasal thermistors) are placed in just below the
nose and above the upper lip to capture the breath of individual
patient. For all samples combined, sensitivity and specificity
values ranged from 80-86% and 57-86% respectively [26].
Also, WM ARES is a home sleep test device that records
heart rate, airflow, respiratory effort and oxygen saturation
[28]. When the patient wakes in the morning, after removing
the tube from the nose and the tape and sensor from the finger,
he/she returns the device to the clinician for analysis. The
device contains a detailed record of the patient‟s personal sleep
patterns, which can be downloaded, analyzed and processed in
the clinician‟s computer. The clinician will then identify if the
person is suffering from sleep apnea.
In [29], a new screening test for OSA was implemented on
a Personal Digital Assistant (PDA) platform to perform the test
at home during the patient‟s nightly rest. The Bluetooth ECG
sensor, made by Corscience [30] is integrated into this
platform, and the algorithm running on the PDA calculates an
index that quantifies the magnitude of the heart beats rate
variability power spectrum alterations. After the patient‟s first
night using the device at home, the collection of test results are
transmitted directly from the PDA to the hospital via the
internet either by a WiFi connection, or by GPRS/UMTS
connection. Once the healthcare staffs have evaluated the
results, they will notify the patient whether the collected test
results are conclusive or not. If the results are conclusive the
patient should return the device. If needed; however, the patient
may be asked to repeat the test again to collect additional data
the following night. However, there is a loss of efficiency in
the use of the wireless network because normal ECGs are also
sent, which implies a high cost.
The portable device hardware design of an FPGA for home
preliminary screening of SA syndromes in [31] stores a
combination of three signals data of three sensors, namely the
nasal air flow and the thorax and abdomen effort signals of
overnight sleep on a Secure Digital card. Later, the sleep
specialist at the clinic uses an algorithm for the evaluation and
detection of SA. The device is relatively inexpensive and
simple to use to diagnose more cases of SA.
Habul et al. [32] developed a diagnostic device for initial
test at home that measures three vital signals, namely “the
respiratory rate measurement, the oxygen concentration in
blood and chest oscillations. The system architecture is divided
into 5 parts, the microcontroller, the external communications,
data storage, power management, and signal conditioning part.
The data will be transmitted wirelessly and stored on the
storage device. After the patient has finished sleeping, the next
morning he or she can bring the data received on the storage
device to a clinic‟s office, where the physician can interpret the
data and determine what the patient‟s condition is. However,
the device will reduce the cost for the patient because the
patient does not have to pay for an overnight stay at the sleep
center” [32].
In using vision based analysis to diagnose OSA in [33]
there has been effective use of two SONY infrared camcorders
(DCR-HC-30E) that work together in order to capture 10 video
clips from three different angles. General body movement is
continuously monitored and updated in a 2D breathing activity
template. After collection of video data, offline analysis is used
to detect abnormal breathing and to facilitate diagnosis of OSA.
Furthermore, after a careful meta-analysis of literature for
twenty-five various tools and devices used to screen and detect
SA by Ross et al. [34], it was discovered that only two of these
were done at home, all others were performed under
supervision in the sleep laboratory. The studies results gave
sensitivity values ranging from 78-100% and specificity values
ranging from 62-100%. However, the related issues such as
reliability, compliance, prices and safety, equipment failure
rates were largely ignored.
IV. REAL-TIME SYSTEMS FOR CONTINUOUS DETECTION
AND SCREENING OF SLEEP APNEA
Although the systems in [26-34] provide home based OSA
diagnosis, but all of them record the physiological sleep data of
patients to memory card. Then, the patient must load these data
into medical center computers where physicians use specialized
software to analyze the data. Meanwhile, some patient can
experience life threatening episodes by not receiving proper
feedback notification from a medical center. Hence, to reduce
the waiting times, as an alternative proposal to that scenario,
the real time monitoring systems that promotes not only a
transmission of physiological data but also a real time analysis
of these data in order to alert of the apnea event and help
patients to recover is performed in researches that appear in
[35-40].
Sechang et al. [35] propose a wireless OSA monitoring
system, which enables the patient to be diagnosed and receive
feedback at home. The system supports monitoring five
different biomedical signals continuously, namely,
electrocardiogram with dry electrodes, body position, nasal
airflow, abdomen/chest efforts and oxygen saturation. A
wireless transmitter unit in the system sends the measured
signals from sensors to a receiver unit with Zigbee
communication. The receiver unit, which has two wireless
modules, Zigbee and Wi-Fi, receives signals from the
transmitter unit and retransmits signals to the remote
monitoring system with Zigbee and Wi-Fi communication,
respectively.
In [36] the implementation of HealthGear, a real-time
monitoring wearable system with a blood oximeter to monitor
the pateint‟s blood oxygen level and pulse is presented. The
three main hardware components of HealthGear‟s include: an
oximetry sensor, a data transmission module and a smart
phone. The sensor is connected wirelessly via Bluetooth to a
smart phone which collects, analyzes and transmits the stored
physiological data, and presenting it to the patient in an
understandable way. The after mentioned study addresses with
20 participants how HealthGear manages to acquire, process,
store, and display the medical information.
The study in [37] develops a wearable biomedical system
embedded in a comfortable glove. The work is based on the
photoplethysmographic (PPG) signal coming from a standard
SpO
2
wrapped sensor placed in one of the fingers for the
continuous monitoring of SA patient at home. The real-time
monitoring is performed through the glove communications
with an internet gateway connected with a remote station.
When the number of SA events crosses a guard level, the alarm
is released.
More recent works [38, 39] implement real-time monitoring
systems that detect apneic events while the patient is sleeping.
This monitoring system constitutes of SpO
2
values analysis
from the Medical, Inc., 4100 Digital Pulse Oximeter. The
oximeter uses the Bluetooth serial line profile to send SpO
2
current values every second to the PDA. The classifier of apnea
episodes has been built using the Bagging method that uses the
decision tree ADTree as base classifier. However, it must be
taken into consideration that the classifier has only been trained
and tested with the eight records of Apnea-ECG Database from
the Physionet website that contain SpO
2
records, and when
validated against Apnea-ECG Database, provides an accuracy
of 93%. Moreover, the system is limited to the detection of SA,
but proposed system can be used for the detection of a variety
of sleep disorders.
Another recent work, Apnea MedAssist [40], was
implemented on Android operating system (OS) based
smartphones. The smartphone provides initialization,
configuration, and synchronization of Bluetooth connectivity to
an off-the-shelf one Lead ECG sensor used for recording heart
activity on a per 1-min epoch basis. The fully automated
processing platform on a smartphone, which was implemented
to process ECG and generate input features for the SVM
classifier to recognize OSA events, shows a high degree of
accuracy for both home and clinical care applications.
However, to increase the accuracy to the ones considered here,
more extracted input features from ECG or other biomedical
sensors such as SpO
2
can be added.
V. CONCLUSIONS
Table 1 provides a summary of all related work with
respect to performance measures, signals employed, technique
used, test set and size, decision method chosen to signify if
there is an apnea on a specific signal interval or not, whether it
is run in real time, offline or portable, hardware used in the
implementation, and more extra features such as the cost.
From such results, we can see that most of the approaches
make use of the whole signal in order to perform the analysis,
and the validation provided by the tests have been performed
on the Apnea-ECG Database. Furthermore, we can categorize
the various systems that deal with OSA monitoring, that is,
those portable commercially available as well as research
proposals into off-line and real-time systems. The main
categories include (A) Off-line systems that help to diagnose
SA using the automatic computer analysis on downloaded
recorded data. For example, Zhao et al. [18], ApneaLink
TM
Plus [27], and other portable commercial systems that only
record the signals (SpO
2
and airflow) to perform off-line
analysis. These systems still have various limitations resulting
from the fact that the classification is not performed in the
place where the signal is acquired. (B) Systems that perform
local real-time OSA monitoring. For example, MedAssist [40]
lies within this category.
In general, the diversity of the studies designs and
objectives were very high and the methodological rigor of these
studies as assessments of diagnostics and monitoring tests was
low.Thus, to enhance the utility of this literature, we are
workingon developing a portable monitoring system to
facilitate the self-administered sleep tests in familiar
surroundings environment closer to the patients' normal sleep
habits.Therefore, the patient does not need hospitalization and
can be diagnosed and receive feedback at home, as it eases
follow-up and retesting after treatment. In this system, the sleep
data will be sent wirelessly via Bluetooth to a nearby smart-
phone for processing and storage. Moreover, the data can be
uploaded to cloud and transmitted to a hospital to keep the
individual's health records. This not only will assist physicians
and patients in planning for sleep apnea treatment, but will also
offer access to a large pool of sleep data for investigations in
this challenging field through providing benchmark data that
can be used by researchers to enhance their used mechanisms
and tools.
After the successful design and implementation of the OSA
system, it is planned to be experimentally tested in order to
evaluate its accuracy and practicality. The tests will take place
in a local hospital for a set of patients who have symptoms of
OSA. In addition, a different set of other subjects without SA
symptoms will test the system to verify its false positive
accuracy.
REFERENCES
[1] Sleep Disorders Guide. www.sleepdisorderguide.com.
[2] N. Israel, A. Tarasiuk, and Y. Zigel,Nocturnal Sound Analysis for the
Diagnosis of Obstructive Sleep Apnea,” in Proceedings of the 32
nd
IEEE
International Conference on Engineering in Medicine and Biology
Society (EMBS 2010), pp. 6146-6149, Sep. 2010.
TABLE 1 Comparison Current OSA Detection Techniques
Research
work
Performance [%]
Signal
analyzed
Techniques
employed
Decision
method
Test set & size
Realtime/Of
fline/
Portable
Hardware used
Se
Acc
Lin et al.
[12]
2006
69.64
NA
EEG
Wavelet
Transforms
& ANN
Threshold
(delta wave
frequency)
slp59 EEG
recording from the
MIT-BIH sleep
database
Offline
NA
Almazaydeh
et al. [14]
2012
100
96.5
ECG
Feature
Extractions
SVM
classifier
70 ECG recording
from physionet
Offline
NA
Marcos et al.
[16]
2010
97
93
SpO
2
Spectral and
nonlinear
features
LDA classifier
214 SpO
2
signals
Offline
NA
Almazaydeh
et al. [17]
2012
87.5
93.3
SpO
2
Feature
Extractions
NN
8 SpO
2
recording
from physionet
Offline
NA
Zhao et al.
[18]
2011
90
NA
Snoring
Formant
frequency
Personalized
threshold
12 simple snores
(7 males, 5
females) and 30
OSA patients (27
males, 3 females)
Online
a non-contact
unidirectional
microphone
placed about 0.3m
above the patient's
mouth
simultaneously
with their full PSG
study
Yadollahi et
al. [21]
2009
90
NA
Tracheal
breathing
Total energy
of the breath
sounds
segment
fuzzy
functions
40 patients
Online
Sony (ECM-77B)
Microphone
Xie et al.
[23]
2012
79.75
84.4
Spo
2
and
ECG
Features set
Bagging with
REP Tree
classifier
25 subjects PSG
recording
Offline
NA
Alvarez et al.
[24] 2009
91
88.5
Spo
2
and
EEG
Spectral
analysis
Forward
stepwise
logistic
regression
(LR)
148 subjects PSG
recording
Offline
NA
Shochat et
al. [26]
2002
80-86
NA
Airflow
NA
NA
402 patients
Online,
Portable
oral and nasal
thermistors
Wang et al.
[33]
2006
NA
NA
Body
movement
Video
information
A
continuously
updated 2D
breathing
activity
template
2 subjects
Online
Two SONY infrared
camcorders (DCR-
HC-30E)
Angius et al.
[37]
2008
NA
NA
PPG
Frequency
analysis
Guard level
(fixed
threshold)
20 volunteers
Remote
realtime
SpO2wrapped
sensor, remote station
Burgos et al.
[38,39]
2010
NA
93
SpO
2
Features set
decision tree
ADTree
classifier
8 records from
Physionet
Realtime,
portable
Medical, Inc., 4100
Digital Pulse
Oximeter, PDA
Bsoul et al.
[40] 2011
96
NA
ECG
Features
measure
SV classifier
ECG recording
from physionet
Realtime,
portable
ECG sensor,
smartphone, server
[3] A. Yilmaz, and T. Dundar, Home Recording for Pre-Phase Sleep Apnea
Diagnosis by Holter Recorder Using MMC Memory,” in Proceedings of
the 2010 IEEE International Conference on Virtual Environments
Human-Computer Interfaces and Measurement Systems (VECIMS), pp.
126-129, Sep. 2010.
[4] “Choosing a CPAP,” American Sleep Apnea Association,
www.sleepapnea.org/resources/pubs/cpap/htm.
[5] SleepMedInc. www.sleepmed.md.
[6] D. Baraglia et al.,Automated Sleep Scoring and Sleep Apnea Detection
in Children,” in Proceedings of SPIE 6039, 2005.
[7] “Sleep Study Cost”, Shop & Compare Healthcare Facilities and Cost
with NewChoiceHealth.com. Web. 03 Dec. 2010.
www.newchoicehealth.com/SleepStudyCost.
[8] P. Chazal, C. Heneghan, and W. Mcnicholas, Multimodal Detection of
Sleep Apnoea Using Electrocardiogram and Oximetry Signals,”
Philosophical Transactions of The Royal Society A: Mathematical,
Physical and Engineering Sciences, vol. 367, no. 1887, pp. 369-389,
2009.
[9] PhysioNet, The Research Resource for Complex Physiologic Signals,
http://www.physionet.org/.
[10] P. de Chazal, R. Reilly, and C. Heneghan, “Automatic Sleep Apnoea
Detection Using Measures of Amplitude and Heart Rate Variability
From The Electrocardiogram,” in Proceedings of the 16th IEEE
International Conference in Pattern Recognition., vol. 1, pp. 775778,
2002.
[11] D. Liu, Z. Pang, and S. Lloyd, “A Neural Network Method for Detection
of Obstructive Sleep Apnea and Narcolepsy Based on Pupil Size and
EEG,” in IEEE Transactions on Neural Networks., vol. 19, no. 2,
pp.308318, 2008.
[12] R. Lin, R. Lee, C. Tseng, H. Zhou, C. Chao, and J. Jiang, “A New
Approach for Identifying Sleep Apnea Syndrome Using Wavelet
Transform and Neural Networks,” Biomedical Engineering:
Applications, Basis & Communications., vol. 18, no. 3, pp. 138
143,2006.
[13] P. Chazal, T. Penzel, and C. Heneghan, “Automated Detection of
Obstructive Sleep Apnoea at Different Time Scales Using the
Electrocardiogram, Institute of Physics Publishing., vol. 25, no. 4, pp.
967983, Aug. 2004.
[14] L. Almazaydeh, K. Elleithy, and M. Faezipour, “Detection of obstructive
sleep apnea through ECG signal features,” in Proceedings of the IEEE
International Conference on Electro Information Technology (IEEE
eit2012), pp. 1-6, May. 2012.
[15] M. Canosa, E. Hernandez, and V. Moret, “Intelligent Diagnosis of Sleep
Apnea Syndrome,” In IEEE Engineering in Medicine and Biology
Magazine., vol. 23, no. 2, pp. 7281, 2004.
[16] J. Marcos, R. Hornero, D. Álvarez, F. del Campo, and M. Aboy,
“Automated Detection of Obstructive Sleep Apnoea Syndrome from
Oxygen Saturation Recordings Using Linear Discriminant Analysis,”
Medical & Biological Engineering & Computing, vol. 48, pp. 895902,
2010.
[17] L. Almazaydeh, M. Faezipour, and K. Elleithy,A Neural Network
System For Detection Of Obstructive Sleep Apnea Through SpO
2
Signal
Features,” (IJACSA) International Journal of Advanced Computer
Science and Applications, vol. 3, no. 5, pp. 7-11, Jun. 2012.
[18] Y. Zhao, H. Zhang, W. Liu, and S. Ding, “A Snoring Detector for
OSAHS Based on Patient‟s Individual Personality,” In 3rd International
Conference in Awareness Science and Technology (iCAST)., pp. 24-27,
2011.
[19] K. Andrew, S. Tong, et al. Could Formant Frequencies of Snore Signals
Be an Alternative Means for the Diagnosis of Obstructive Sleep
Apnea?, Sleep Medicine, vol. 9, pp. 894-898, Dec. 2008.
[20] K. Andrew, T. Koh, E. Baey, et al., Speech-Like Analysis of Snore
Signals for the Detection of Obstructive Sleep Apnea, In International
Conference on Biomedical and Pharmaceutical Engineering 2006
(ICBPE 2006), pp. 99-103, Dec. 2006.
[21] A. Yadollahi, and Z. Moussavi, “Acoustic Obstructive Sleep Apnea
Detection”, In Proceedings of IEEE Conference on Engineering in
Medicine and Biology Society (EMBC 2009), pp. 7110-7113, Sep. 2009.
[22] T. Micsik, S. Benedek, and Z. Benyo, “A Novel Method for the
Detection of Apnea and Hypopnea Events in Respiration Signals”, IEEE
Transactions on Biomedical Engineering, vol. 49, no. 9, pp. 936-942,
Sep. 2002.
[23] B. Xie, and H. Minn, “Real Time Sleep Apnea Detection by Classifier
Combination,” in IEEE Transactions on Information Technology in
Biomedicine, vol. 16, no. 3, pp. 469-477, May. 2012.
[24] D. Alvarez, R. Hornero, J. Marcos, F. Campo, and M. Lopez, “Spectral
Analysis of Electroencephalogram and Oximetric Signals in Obstructive
Sleep Apnea Diagnosis,” in Proceedings of the 31st IEEE International
Conference on Engineering in Medicine and Biology Society (EMBS
2009), pp. 400-403, Sep. 2009.
[25] H. Khandoker,K. Karmaker, and M. Palaniswami, “Analysis of
coherence between Sleep EEG and ECG Signals during and after
Obstructive Sleep Apnea Events,” In Proceeedings of 30th IEEE
International Conference on Engineering in Medicine and Biology
Society (EMBS 2008), pp. 3876-3879, 2008.
[26] T. Shochat, N. Hadas, M. Kerkhofs, et al., “The SleepStrip
TM
: An
Apnoea Screener for the Early Detection of Sleep Apnoea Syndrome,”
European Respiratory Journal, vol. 19, pp. 121-126, 2002.
[27] ResMed, www.resmed.com.
[28] M. Stuart,Sleep Apnea Devices: The Changing of the Guard,” START-
UP journal, vol. 15, no. 10, pp. 1-8, Oct. 2010.
[29] A. Otero, S. Dapena, P. Felix, J. Presedo, and M. Tarasco, “A Low Cost
Screening Test for Obstructive Sleep Apnea that can be Performed at the
Patient‟s Home,” In Proceedings of the 6th IEEE International
Symposium on Intelligent Signal Processing, pp. 199-204, Aug 2009.
[30] Corscience GmbH & Co. http://www.corscience.es.
[31] K. Al-Ashmount, H. Hamed, and A. Morsy, “FPGA-based Sleep Apnea
Screening Device for Home Monitoring,” In Proceedings of 28th IEEE
International Conference on Engineering in Medicine and Biology
Society (EMBS 2006), pp. 5948-5951, 2006.
[32] M. Habul, R. Hallmark, A. Kotlyanskiy, P. Parejk, and R. Kamoua,
“Sleep Apnea Diagnosis Device,” In Proceedings of 2011 IEEE Long
Island conference in Systems, Applications and Technology (LISAT), pp.
1-5, 2011.
[33] C. Wang, A. Ahmed, and A. Hunter, “Vision Analysis in Detecting
Abnormal Breathing Activity in Application to Diagnosis of Obstructive
Sleep Apnoea,” In Proceedings of the 28th IEEE International
Conference on Engineering in Medicine and Biology Society (EMBS
2006), pp. 4469-4473, 2006.
[34] S. Ross, I. Sheinhait, H. Katherine, et al., “Systematic Review and Meta-
Analysis of the Literature Regarding the Diagnosis of Sleep Apnea,”
Sleep Journal, vo. 23, no. 4, pp. 519-532, 2000.
[35] O. Sechang, K. Hyeokjun, and V. Vijay, “Ubiquitous Health Monitoring
System for Diagnosis of Sleep Apnea With Zigbee Network and
Wireless LAN,” Journal of Nanotechnology in Engineering and
Medicine, vol. 2, no. 2, 2011.
[36] N. Oliver, and F. Mangas, “HealthGear: Automatic Sleep Apnea
Detection and Monitoring with a Mobile Phone,” Journal of
Communications, vol. 2, no. 2, pp. 1-9, March 2007.
[37] G. Angius, and L. Raffo, “A Sleep Apnoea Keeper in a Wearable Device
for Continuous Detection and Screening During Daily Life,” Computers
in Cardiology, pp. 433-436, 2008.
[38] A. Burgos, A. Goni, A. Illarramendi, and J. Bermudez, “Real-Time
Detection of Apneas on a PDA,” IEEE Transactions on Information
Technology in Biomedicine, vol. 14, no. 4, pp. 995-1002, July 2010.
[39] A. Goni, A. Burgos, L. Dranca, et al., “Architecture, Cost-Model and
Customization of Real-Time Monitoring Systems Based On Mobile
Biological Sensor Data-Streams,” Computer Methods and Programs in
Biomedicine, vol. 96, no. 2, pp. 141-157, May 2009.
[40] M. Bsoul, H. Minn, and L. Tamil, Apnea MedAssist: Real-time Sleep
Apnea Monitor Using Single-Lead ECG,” in IEEE Transactions on
Information Technology in Biomedicine, vol. 15, no. 3, pp. 416-427,
May 2011.
... The increasing number of patients who have not been diagnosed with SA cause the necessity of improving medical systems and methods of diagnosis of SA with the differentiation of obstructive and central SA out of hospital [4,5]. One way to solve this problem is to develop systems and devices for monitoring SA at home. ...
Article
Full-text available
The work presents a set of significant indicators and a method of detection of sleep apnea episodes based on the cardiovascular and respiratory systems’ state analysis by ECG signal, pulse wave signals and respiratory movements signals. The algorithms of biological signal processing based on wavelet analysis for studying the dynamics of indicators important for diagnosis in time and spectral range, an algorithm for detecting apnea episodes, which allows to differentiate sleep apnea of central and obstructive type, are given. The proposed algorithms are based on the method of smart monitoring of sleep apnea, which allows to identify episodes of sleep apnea during the patient’s stay outside the hospital.
Conference Paper
Full-text available
Obstructive sleep apnea (OSA) is a common disorder in which individuals stop breathing during their sleep. Most of sleep apnea cases are currently undiagnosed because of expenses and practicality limitations of overnight polysomnography (PSG) at sleep labs, where an expert human observer is needed to work over night. New techniques for sleep apnea classification are being developed by bioengineers for most comfortable and timely detection. In this paper, an automated classification algorithm is presented which processes short duration epochs of the electrocardiogram (ECG) data. The automated classification algorithm is based on support vector machines (SVM) and has been trained and tested on sleep apnea recordings from subjects with and without OSA. The results show that our automated classification system can recognize epochs of sleep disorders with a high degree of accuracy, approximately 96.5%. Moreover, the system we developed can be used as a basis for future development of a tool for OSA screening.
Article
Full-text available
Obstructive sleep apnea (OSA) is a common disorder in which individuals stop breathing during their sleep. Most of sleep apnea cases are currently undiagnosed because of expenses and practicality limitations of overnight polysomnography (PSG) at sleep labs, where an expert human observer is needed to work over night. New techniques for sleep apnea classification are being developed by bioengineers for most comfortable and timely detection. In this paper, an automated classification algorithm is presented which processes short duration epochs of the electrocardiogram (ECG) data. The automated classification algorithm is based on support vector machines (SVM) and has been trained and tested on sleep apnea recordings from subjects with and without OSA. The results show that our automated classification system can recognize epochs of sleep disorders with a high degree of accuracy, approximately 96.5%. Moreover, the system we developed can be used as a basis for future development of a tool for OSA screening.
Article
Full-text available
Obstructive sleep apnea (OSA) is a common disorder in which individuals stop breathing during their sleep. These episodes last 10 seconds or more and cause oxygen levels in the blood to drop. Most of sleep apnea cases are currently undiagnosed because of expenses and practicality limitations of overnight polysomnography (PSG) at sleep labs, where an expert human observer is required. New techniques for sleep apnea classification are being developed by bioengineers for most comfortable and timely detection. In this study, we develop and validate a neural network (NN) using SpO2 measurements obtained from pulse oximetry to predict OSA. The results show that the NN is useful as a predictive tool for OSA with a high performance and improved accuracy, approximately 93.3%, which is better than reported techniques in the literature.
Article
Full-text available
A method for the automatic processing of the electrocardiogram (ECG) for the detection of disordered breathing associated with obstructive sleep apnoea is presented. The method provides a minute-by-minute analysis of night-time single lead ECG recordings. An independently validated database of 35 ECG recordings acquired from normal subjects and subjects with obstructive and mixed sleep apnoea, each of approximately eight hours in duration, was used throughout the study. A wide variety of features based on heart beat intervals and an electrocardiogram derived respiratory signal were considered. Classifiers based on linear and quadratic discriminants were used. Results show that a 90% success rate in correctly identifying one-minute segments containing disordered breathing is achievable.
Article
The Greek word "apnea" literally means "without breath." The three types of apnea are obstructive, central, and mixed. Obstructive is the most common type. Despite the difference in the root cause of the three types of apnea, people with untreated sleep apnea stop breathing repeatedly during their sleep, sometimes hundreds of times during the night and often for 1 min or longer. Sleep apnea can cause high blood pressure and other cardiovascular diseases, memory problems, weight gain, impotency, and headaches. Sleep apnea may be responsible for job impairment and motor vehicle accidents. Apart from the physical health risks, sleep apnea can lead to social problems when left undiagnosed or untreated such as a high amount of unnecessary health care cost. Sleep disorder is diagnosed with polysomnography, an overnight sleep study, by monitoring electrical activity of brain, heart, eye movement, muscle activity, breathing pattern, and other physiological signals. Because polysomnography requires overnight monitoring by a sleep technologist with full sleep staging, polysomnography is expensive, inconvenient, time consuming, and labor intensive. Although some systems provide home based diagnosis, most systems record the sleep data in a memory card. The patient must send the memory card to a medical center through the mail or internet. The real-time monitoring is not performed and a patient can experience life threatening episodes by not receiving proper feedback from a medical center or a physician. We propose a wireless health monitoring system to diagnose sleep apnea, which enables the global monitoring of biomedical signals. A patient does not need hospitalization and can be diagnosed and receive feedback at home. The system supports monitoring five different biomedical signals to diagnose sleep apnea: electrocardiogram with dry electrodes (no gel), body position, nasal airflow, abdomen/chest efforts and oxygen saturation. The system consists of three units: a wireless transmitter, a wireless receiver, and a monitoring unit. A wireless transmitter unit sends the measured signals from sensors to a receiver unit with Zigbee communication. The receiver unit, which has two wireless modules, Zigbee and Wi-Fi, receives signals from the transmitter unit and retransmits signals to the remote monitoring system with Zigbee and Wi-Fi communication, respectively. By using both the Zigbee network and wireless LAN, the system can achieve low power consumption in a local monitoring area with the feature of the Zigbee standard. The system also provides wide data coverage and easily extends its sensor network to the Internet with the Wi-Fi standard. The features of the system and the results of the continuous monitoring of vital signals are presented.
Conference Paper
A conventional diagnostic tool for assessing Obstructive Sleep Apnea Hypopnea Syndrome (OSAHS) is polysomnography (PSG), which is expensive and uncomfortable for patients. It is an important and urgent topic to find a non-invasive and low-cost diagnostic approach for OSAHS detection. Recently, the snore signal analysis receives much attention due to its potential capability for OSAHS detection. In this paper, we propose a novel method for diagnosing OSAHS based on patient's individual personality. First, the first formant frequencies of each snorer are classified into two clusters by K-means clustering. And then, using the first cluster center of each snorer, we set a personalized threshold to distinguish the hypopneic snores from the normal ones. Since the proposed threshold varies with each individual, the patient's individual personality can be overcome effectively. Experimental results show the validity of the proposed detector. In the experiments, the sensitivity of our method can achieve 90% and the specificity can achieve 91.67%.
Article
Sleep apnea is a disorder in which people are unable to breathe while they are sleeping. The goal of this project is to design a medical device that can be used as a diagnostic tool for preliminary diagnosis of sleep apnea. The signals measured will be the respiratory rate measurement, the oxygen concentration in blood and chest oscillations. The system architecture will be divided into 5 parts, the microcontroller, the external communications, data storage, power management, and signal conditioning part. The data stored will be transmitted wirelessly to a clinic's office. The physician can interpret the data received from the medical device the next morning and conclude what the patient's condition is. There will be sensors designed to measure airflow, temperature and oxygen level. The design for the airflow and temperature measurements will be made through a Wheatstone Bridge circuit using thermistors. To measure the saturation of oxygen in the blood, a pulse oximetry technique will be used. Also, to measure chest oscillations, an accelerometer will be used. The sleep apnea diagnostic device will be available at a low cost and can be administered at home, so patients can afford to purchase it. The device will also reduce costs for the patient because the patient does not have to pay for an overnight stay at the clinic to determine if he or she has sleep apnea. High costs may occur if the doctor decides the patient needs further tests, but for patients who do not have sleep apnea, however think they are inflicted with the disorder, will not have to pay for an expensive one night stay at the clinic.
Article
This paper investigates the automated detection of a patient's breathing rate and heart rate from their skin conductivity as well as sleep stage scoring and breathing event detection from their EEG. The software developed for these tasks is tested on data sets obtained from the sleep disorders unit at the Adelaide Women's and Children's Hospital. The sleep scoring and breathing event detection tasks used neural networks to achieve signal classification. The Fourier transform and the Higuchi fractal dimension were used to extract features for input to the neural network. The filtered skin conductivity appeared visually to bear a similarity to the breathing and heart rate signal, but a more detailed evaluation showed the relation was not consistent. Sleep stage classification was achieved with and accuracy of around 65% with some stages being accurately scored and others poorly scored. The two breathing events hypopnea and apnea were scored with varying degrees of accuracy with the highest scores being around 75% and 30%.
Article
Since 1999, PhysioNet (http://physionet.org/) has offered free access via the web to large collections of recorded physiologic signals and related open-source soft-ware. This presentation offers an overview of PhysioNet and discusses its benefits to the global community of re-searchers, clinicians, educators, and students, with exam-ples that include support of exploratory data analysis, de-velopment and validation of experimental methods, collab-oration among geographically distant investigators, and stimulating rapid progress on specific questions via open research and engineering challenges.