Conference PaperPDF Available

Sleep Apnea Detection by Real-time Heart Rate Variability Analysis

Authors:

Abstract

Sleep Apnea Detection by Real Heart Rate Variability Analysis; The obstructive sleep apnea (OSA) is a chronic, lifelong medical condition that effects on the sleep, health, and quality of life. Breakdown of the patients on the basis of the acute situations can be useful in providing appropriate therapy. In this study we present a classifier to predict the obstructive sleep apnea (OSA) based on heart rate variability of patients. We used the recorded ECG signals from PhysioNet Database. At first, in the preprocessing, stage the noise from the electrocardiogram (ECG) signal removed and R spikes were detected to generate heart rate variability (HRV). The HRV signal is measured by calculating the time between R spikes on an ECG period [2]. Then, the HRV signal segmented to multiple windows with same durations. The choice of the optimum window duration was determined based on the correct detection of the apnea event period. The next stage was related to linear and non -linear feature extraction. We used paired sample t-test that is a statistical technique to compare two periods (apnea and non-apnea). A significant result was indicated with p- value less than 0.05. These features were used as the inputs of two different classifier include Multi-Layer Perceptron (MLP) and Radial Basic Function (RBF) to find the best method to distinguish patients with higher apnea risk. The results showed that the MLP classifier is more capable to separate the apnea periods from control. The MSE and R index for detecting the control, OSA and CSA event were 0.051-0.94, 0.057-0.93 and 0.053-0.94, respectively. In conclusion, based on proposed algorithm the HRV signal and feature presented in this paper had the lowest mean square error for the detection the apnea event of the non-apnea.
¶Ì¸ve ʬ̬u ½Z» ½Z] w¿ ÉË~a Ì̤e \¸« dÆm
ÌzeµÔfyY ÉZÅ [YÂy ÄÀaM
Ê¿Y|Ë» ʼË¯ |¼v»
˺
*,
,|Ìu ZË|Æ»
2
1
Ê°a Ê|ÀÆ» Á dY|Æ] Ã|°¿Y{, Ê°a ¹Â¸ |uYÁ ,½YÆeÊ»ÔY {YM ÃZ´¿Y{, ½YÆe.½YËY , iautmu.ac.irkarimi.m@
2
Ê°a Ê|ÀÆ» Á dY|Æ] Ã|°¿Y{, Ê°a ¹Â¸ |uYÁ ,½YÆeÊ»ÔY {YM ÃZ´¿Y{, ½YÆe.½YËY , mahdyarheydar@gmail.com
ÄÔy
Ê°Ë Y cÓÔfyY ËZ ÉY] Ê¨Àe cÔ°» į [YÂy { ÉZ¼Ì] |À¯ Ê» {ZnËY ½YZ¼Ì] ÄÀaM dY į hZ] » ÉY] É{ZË cÔ°
.{Â ½YZ¼Ì] ±» Ä] nÀ» {Zu ËY { |¿YÂe Ê» ½M «Â¼] ½Z»{ Á tÌv Ìze ¹| .{Â Ê» {Y§Y ʧY³ÂÀ¼̸a ÃÁ»Y
Ìze ÂÀ» Ä] µÂ^« ¶]Z« Á ®Ë ÄÀaM [YÂy ¶«Y|u Âu Á ËZ»M ¾ËY ÉÓZ] ÄÀËÅ ¶Ì·|] Z»Y .dY24 dÆm Z¼Ì] ÄfZ
ÉZÅf»YZa d^i ºÆ» ÉZÅÂf¯Z§ ®Ë ÉÁ ] Ä·Z» µZu { ½Z¬¬v» ,ÄÂ]» ®Ë·ÂË̧, ÊËZZÀ dÆm |ÀfÅ ÊËZÅf»YZa
½ÂÀ¯Ze .|ËZ¼¿ ÊËY] ®¼¯ ÄÀaM ĸ¼u {Y|y ÊËZZÀ { Ê·Â^« ¶]Z« |u Ze |¿YÂf] d^i ʳ{Z ] ÃÁÔ į ÊeZ·Z» ÂÀ» Ä]
ÊËZZÀ Ĩ«Á ÉZÅ Ê¨Àe ³Â·Z¨¿YÁf°·Y Âe Éf¼¯ ¿Y» Z] Z¼Ì] Y Ê^¸« µZÀ´Ì d^i Ä°ÀËY ¶Ì·|] Z»Y ,dY Ã| ¹Zn¿Y ¹Y
¹Z´ÀÅ {§ Z¿y dÌÁ ÊËZZÀ dÆm ¹Y³ÂË{Z¯Áf°·Y µZÀ´Ì Y Ã{Z¨fY ÉÁ ] ʸY ¯¼e Ä·Z¬» ¾ËY { ,dY ÃY¼Å
.dY Ã{Â] [YÂy ¦¸fz» ¶uY» { Ä·Z¬» ¾ËY Y Ã{Y{ ÃZ´ËZa{Y|¿ZfY MIT/BIH .dY Ã| Ã{Z¨fY ¾ ZÌ» ¥Yv¿Y Á Âf»
½YZ¼Ì]ÄÀaM yZ Á ½|] ¹m yZ ,- ]Y] \Ìef] ÄÀaMÂbËZÅ10ά50 Ã{Á|v»)28 Ze68 ,(µZ6ά4/31 Ã{Á|v»)1/25 Ze
5/42 Á (¹³Â¸Ì¯ 3/20ά20
/24 Ã{Á|v»)7/1 Ze9/90( .] Ä] ÊËZZÀ ÂÀ»[YÂy ¾Ìu { Z¼Ì] Z¿y dÌ«Â» ,ËÁ ʳ
ÉZÅ ÊyÌ£ Á ÊyY Z¼Ì] \¸« ½Z] w¿ ÉË~aÌ̤e µZÀ´Ì Yf a Á kYzʳËÁ [Zzf¿Y Y \ZÀ» ÉZÅ Y Ã{Z¨fY Z]
ÉZ»M Ì·Z¿M Ìze Ä] ÊZ ÄËZa ]Ze Á ÄËÓ |Àq ½Áfba Ê^ ÉZÅ Ä°^ Âe , lËZf¿ .| ÄfyY{a ½YZ¼Ì] { ÄÀaM ÄÌ^
ÄËÓ |Àq ½fba Ê^ Ä°^ Y Ã{Z¨fY ,{Y{ ½Z¿ ÉZ ¦¸fz» ZfyZ Z] , d¿YÂe¾Ì´¿ZÌ» cYÁ~n» Zy Á\Ë ÊË°¿
Y] \Ìef] Y051/0 Á94/0 ÉY]µfÀ¯ d·Zu Á057/0 Á93/0 ÉY] d·ZuÉ{Y|¿Y ÄÀaM Á dËZÆ¿ {053/0 Á94/0 ÉY]
[YÂy É¯» ÄÀaM d·Zu .|À¯ {ZnËY ¾ËY]ZÀ]É{ZÆÀÌa ºfË´·Y Y Ã{Z¨fY Z] Ä·Z¬» ¾ËY {, Z] ½| {Á|v» ,cZ^Zv» ºnuÅZ¯
ʳËÁ {Y|e Z]ÉZy ½{¯ º¯ Ĭ^ É|À] Z¼Ì] Zy ËY¹Z´ÀÅ
[YÂy Á Ê].| ÊËZZÀ
:É|̸¯ cZ¼¸¯ÄÀaM ,
ʳËÁ ,Ê^¸« µZÀ´Ì ÉZÅÉ|À] Ĭ^ ,ÉZ»M Ì·Z¿M ,ÊyÌ£ Á Êy
* Corresponding author: ½YÆe Ê°a ¹Â¸ |uYÁ Ê»ÔY {YM ÃZ´¿Y{ ,Ê°a Ê|ÀÆ» Á dY|Æ] Ã|°¿Y{ ʼ¸ cPÌÅ Â,ZË{ZfY
Email: karimi.m@iautmu.ac.ir
2017 24th national and 2nd International Iranian Conference on Biomedical Engineering (ICBME), Amirkabir University of
Technology, Tehran, Iran, 30 November - 1 December 2017
978-1-5386-3609-1/17/$31.00 ©2017 IEEE 227
1- Ä»|¬»
µZ»¿ {Y§Y { ¹Z´ÀÅ { \¸« ½Z] Á ½ÂyZ§Ä¿Z^ duYfY
.|À¯ Ê» Y|Ìa ÅZ¯ ,Ä¿YÁ Àe a ÉZÅ dÌ·Z§ ¹| ¶Ì·|]
{§ Ä°ÌeÂ { ¾ËY]ZÀ] [YÂy Á ʧZ¯\ZÀ» hZ] ,|Z] ÄfY{
ÅZ¯Z§ c|dÌ· ĸ \¸« d»Ô Y dËZÆ¿ { Á Á ʼm
ÉfÌ] ÊuÁ
Y{Ây] {Â] |ÅYÂy . ËZ cÓÔfyY Y Ê°Ë ZÌ] Á
Z¿y[YÂy É{Y|¿Y ÄÀaM Ä»Zm {Y§Y {
1
|Ë{e ½Á|] į dY
.{Y{ ÊËY] ÌiZe {Y§Y d»Ô Á Ã»Á ÉZÅ dÌ·Z§ ÉÁ ] Z]
Äf§³ cÂ cZ¬Ì¬ve Ä] ÄmÂe2 Ze4 { ÊÀ ÃZ] { ZÌ¿{ {Y§Y |
|ÀfÅ [YÂy ÄÀaM µÔfyY » { µZ¿ZÌ»[1] .µÔfyY ¾ËY į
fÌ] {Á|u ¨Àe « Z]Ã{ Ã{¿Za Ze [YÂy { ÄÌ¿Zi Á ÉZŠĨ«Á
¨Àe { Ê·YÂf»{Â Ê» z», \^ ½YÌ» ½| º¯ ½Ì¯Y
½Ây {Ã| hZ] Á{§ ½|Ëa {¹Z´ÀÅ {§ [YÂy .{Â Ê» ÄÀaM
[YÂy { {Â Ê» ºÌ¬e ¸z» Á É¯» ,É{Y|¿Y Äf{ Ä Ä] .
{Ma Ê¿Z«Â§ ÉZÅ d¼« ,¨Àe ÉY] Ôe º£ Ä] É{Y|¿Y ÄÀ
ÅZ¯ ÊÀÌ] Y YÂÅ ½ZËm Á |À]YÂy Ê» ºÅ ÉÁ ¨Àe ºfÌ
(ÄÀaM ÂbÌÅ) Äf§ZË .(ÄÀaM) {Â Ê» ¦«Âf» M d¸ Ä] É¯» ÄÀa
¤» { ¨Àe ¯» {°¸¼ { µÔfyY µÔfyY ¾ËY .{Â Ê» {ZnËY
{Â Ê» Ã|ÅZ» ¤» Ä«Z \ÌM Á É¤» Äf° Ä] Ôf {Y§Y { .
M¸z» ÄÀa Ì¿ Á dY ¿ Á{ Å Y Ê^̯efÌ] ½Y|À¼·Z {
{Â Ê» Ã|ÅZ». Ê¿ZƳZ¿ ÅZ¯½YÌ» ½Ây ½Ì¯YhZ] ËY§Y
½Ây Z§« ÄnÌf¿ { Á Ä] Â^n» Zb¼a fÌ] c| Z] ½Ây
|ÅYÂydY{ į\^ {ZnËY cÔ°»É|m {Â Ê»[2] . ¾ÌÀr¼Å
ʧZ¯ [YÂy ¾fY{ ¹| cÂ { ËY§Y {§ ÊÀÌ·Â¿Y d»ÁZ¬»
¿ d]ZË{ Âz] d]ZË{ ÉZ¼Ì] Ä] Ôf]Y y Á Äf§ZË2 ËY§Y Y
¾ËY]ZÀ] .|Å{ Ê»{ M Ä°ÌeÂ ÄÀaYÂy É{Y|¿Y[ «Â¼] ½Z»{
¿ hZ] |¿YÂe Ê»¼Ì] Á] É|m ÉZÅ ÉZ ¶»ZÊ¿ZƳZ¿ ±»
Ê^¸«
2
ʼfËM {ZnËY ¶Ì·|], ÉZ¼Ì]Ê^¸« Ê¿Z¬fuY ÊËZZ¿
3
,ÉZ¼Ì]
Ê^¸«-Êf]ZË{
4
Äf° ,ÊfyZÀ Á ʬ¸y cÓÔfyY , {Â ÃÌ£ Á-3[
6] .É{Y§Y ©Yf§Y Á Ìze d̼ÅY ¾ËY]ZÀ]į ºWÔ Á Ä¿Z¿ ÉZÅ
Z°M ÄÀaM É{Y|¿Y [YÂy Âz] y Á ¦a |À¸] Á ]ÉY|Ì °»
{ \Y Y{¿,| .| Ê» ¿ Ä] ÉÁ ÉZÌ] cZ·Z» ½ÂÀ¯Ze
ÉZ¼Ì] Ìze ÄÀÌ» {OSA Ä] Ä»Y{Y { į dY Ã| ¹Zn¿Y
.{Â Ê» ÄfyY{a cZ¬Ì¬ve ¾ËY Y É{Y|e Ê]
Âe į ʬ̬ve {Nobili L µZ { ½YZ°¼Å Á2011 [7] ]
Ê ½YZ¼Ì] ÉÁ» ,d§³ cÂ Zq{ į É{Y§Y į | z
[YÂy µÔfyY |ÀfÅ  ÉZ¼Ì] \ZÀ»Z¿ {°¸¼ Y ÊZ¿ į
. Obstructive sleep apnea (OSA)
. Sudden cardiac death (SCD)
. Congestive heart failure (CHF)
. diabetic heart disease
dY Ê¨Àe ºfÌ [YÂy { Ê¿ZƳZ¿ ±» Ä] nÀ» |¿YÂe Ê»
{Â. Âe į É´Ë{ Ä·Z» ¾ÌÀr¼ÅKatagal P ½YZ°¼Å Á
µZ {2008 ÉÁ ] į {Y{ ½Z¿ | ¹Zn¿Y Ê Z¼Ì] É{Y|e
¿Z¯§ ½YÌ» ÉZ¼Ì] {ÂmÁ Z] cÔ¼uOSA ÊÆmÂe ¶]Z« ËY§Y
cÔ¼u ½YÌ» ÅZ¯ Á µfÀ¯ Ä] nÀ» ÉZ¼Ì] ¾ËY ½Z»{ Á ÄfY{
dY Ã| Ê]8[. ¹Y³Â·Z¨¿YÁf°·Y ÉZÅ µZÀ´Ì Y Ã{Z¨fY
5
Á
½Ây ½Ì¯Y Z^Y t
6
cZ¬Ì¬ve ʼ° ÉZÅ µZÀ´Ì Ì¿ Á
f§³ cÂ ÄÀaM Ìze ÃÂu { į dY É´Ë{ lËZf¿ į dY Ä
Z»Y µÂ^« ¶]Z« É{Á|u Ze ½MdY Ã{Â^¿ z] dËZ[9-11] .
2- Á Á {Y»
2-1- Ã{Y{ ÃZ´ËZa ʧ»
Ã| Ã{Z¨fY ÉZÅ Ã{Y{ ¾ËY {Ä·Z» Y
d¿ÂË̧ Ã{Y{ ÃZ´ËZa [12]
Âe į dY Ã| d^i ¾Ì¸]Á{ ÃZ´¿Y{ ½ZfZ¼Ì] [13] Ê» ,
.|Z]14 Z] Ã{Y{Ã{Z¨fY ¸a ÃZ´f{ YÊʧY³ÂÀ»Â Y25 {§
¶»Z4 Á ½21 Y ZÆÀe Ä·Z¬» ¾ËY { .dY Ã|Ì d^i Ä] {»
µZÀ´Ì¹Y³ÂË{Z¯Áf°·Y
7
Ì¿ Á d^i d·ÂÆ ¶Ì·|] {§ÌiZe
ËY ºÌ¬f»Ã|»M Ìa Z¼Ì] ÉY] É{Y|¿Y ÄÀaM ĸ¼u d·Zu {
[YÂy ,µZÀ´Ì ¾ËY ÉÁ ] .dY Ã| Ã{Z¨fYZ¯§ Ŀ¼¿ ¿
µZÀ´Ì ÉY{] ECG 128 |Ì·) eÅV2( |Z] Ê» . Á Âe
½YZ¼Ì] ¾ ZÌ» ¥Yv¿YÄÀaM yZ Á ½|] ¹m yZ ,
-
]Y] \Ìef] ÄÀaMÂbËZÅ10ά50 Ã{Á|v»)28 Ze68 ,(µZ6ά4/31
Ã{Á|v»)1/25 Ze5/42 Á (¹³Â¸Ì¯ 3/20ά20/24 Ã{Á|v»)7/1 Ze
9/90( .{Â]
2-2- É{ZÆÀÌa Á
cZ¬Ì¬ve ½Z¿ dY Ã{Y{µZÀ´Ì ECG ®ËZf§ |¿ Êy ÄfY
·R» ÉYY{ Ä°¸]ĨķZ¬» ¾ËY { .dY ÊyÌ£ ÉZÅ ÂÀ» Ä]
[YÂy É{Y|¿Y ÄÀaM ÊËZZÀ µZÀ´Ì Y Ã{Z¨fY Z]ECG, Y|f]Y
Æ ©] Ë¿ )50-60 eÅ( ÄÀÌ» y Á Âe µZÀ´Ì ºfË´·Y
d·ÂËÁ [14] | ¥~u. Ã{Z¨fY Z] b ÀÌ°b»Ze Á ¾a ºfË´·Y Y
[15] k» ÊËZZÀ Ä]
R µZÀ´Ì ¶Ì°e ÁHRV | ÄfyY{a
[16] . Y ÂÀ»HRV {ZnËY ®Ë Ã| {ZnËY ¶Y§ Y Ê¿Z» É
Á{ Å ¾Ì]R µZÀ´Ì { Ê·YÂf»ECG dY .{ Ä»Y{Y Y ÉZÅ Á
ÃÂu ,½Z»½Z»- ÊyÌ£ ¾ÌÀr¼Å Á ¿Z¯§ Ä] ÂÀ» kYzfY
cYÌ̤e µZÀ´Ì Y ʳËÁ ½Z] | Ã{Z¨fY \¸« . ZÌ» ¥Yv¿Y
¶Y§½Z] Ä] ½Z]
8
¹Á{ ÄË ,¾Ì´¿ZÌ» ]» ¥ÔfyY ¶Y§
. Electroencephalogram (EEG)
. Blood oxygen saturation level (SPO2)
. Electrocardiogram (ECG)
. Standard Deviation of NN Intervals (SDNN)
2017 24th national and 2nd International Iranian Conference on Biomedical Engineering (ICBME), Amirkabir University of
Technology, Tehran, Iran, 30 November - 1 December 2017
978-1-5386-3609-1/17/$31.00 ©2017 IEEE 228
½Z] ½Z] Ä]
1
¾ÌÀr¼Å Á |{§ {Y|y µZ¼fuY Ä] ½Z] ¶YÂ
½Z] Ê·YÂf» Y fÌ]50 ÄÌ¿Zi ʸ̻
2
{ į |¿{Â] ÊËZŠʳËÁ
.| Ã{Z¨fY ½M Y ½Z» ÃÂu Ä·{Z» ÉZÅ 1 Ze 3 , ]YÁ ÊZË
ʳËÁ½Z» ÃÂu Y Ã| kYzfY É Ê» ½Z¿ Y |Å{[17].
¦
N
i
i
RRRR
N
SDNN
1
_
)(
1
(1)
)2(
¦
1
1
2
1
)(
1
1
N
i
ii
RRRR
N
RMSSD
)3 (
100
50
50
}50|{|50
1
1
u
!
¦
N
NN
pNN
msRRRRNN
N
i
ii
Ä^Zv»½Z» ÃÂu ÉZŠʳËÁ- ¦¸fz» ÉZÅ Äf{ Ä] ¿Z¯§
¶»Z 5] ËÂe±
3
¶ËÁÀ´ËÁ ËÂe ,
4
ÃZe¯ ½Z» ÄË§ ¶Ë|^e ,
5
,
¶Ë|^ed·ÂËÁ
6
, Éf»YZa µ|»Êf³Z]{Ây
7
ÃÌ£ Áe Ê» ºÌ¬
|¿Â. Ä·Z¬» ¾ËY { Y Ã{Z¨fY Z]ZÀ´Ì ½YÂe ¦Ìµ
8
Á Á
Ì£®Ëf»YZa d·ÂËÁ ½YÂe |¿Z] { µZÀ´Ì ÉZÅ Ê¿Z¯§ ¦¸fz»
¶»Z ¾ÌËZa ʸÌy ¿Z¯§
9
)0 Ze04/0 (eÅ ¿Z¯§ ,
¾ÌËZa
10
)04/0 Ze15/0 ,(eÅ ÓZ] ¿Z¯§
11
)15/0 Ze4/0 (eÅ Á
ÓZ] Ä] ¾ÌËZa ¿Z¯§ d^¿
12
kYzfY .|HF ½YÂe ¦Ì {
µZÀ´Ì HRV´¿ZÌ] dÌ·Z§d¼« Á Ê^ ºfÌ ®ÌeZb¼YZa
LF ´¿ZÌ] dÌ·Z§d¼« Z¯{Ây Ê^ ºfÌ ®ÌeZb¼
µfÀ¯ .dY \¸« ½Z] w¿ Ã|ÀÀ¯¾ËY]ZÀ] d^¿ ,½YÂe
HF
LF
Ê»
Ä] |¿YÂeʳËÁ ®Ë ½YÂÀµ{Ze į {Â Ã{Z¨fY ÉY Á ®ÌeZb¼
Ê» Ê]ZËY Y ®ÌeZb¼YZa|À¯ [18-19].
Ä»Y{Y {]Z ½Z] w¿ ÉË~aÌ̤e µZÀ´Ì ÊyÌ£ cY} Ä] ÄmÂe
Ã| |¿e É{ cZ¿Z¿ Ì·Z¿M Á Y ,\¸«
13
]20[ kYzfY ÉY]
.| Äf§³ ®¼¯ ÊyÌ£ ÃÂu { ÊËZŠʳËÁ ÉY] Á ¾ËY
Ä^Zv»\Ë ½YÌ» Ê´f^¼Å
14
ZÀ´Ì { ®Ì»ZÀË{ ÉYY{ ÉZÅ µ
. Root Mean Square of the Successive Differences
(RMSSD)
. Number of pairs of successive NN intervals that
differ by more than 50 ms (pNN50)
. Burg
. Wigner-Ville distribution
. Short-time Fourier transform (STFT)
. Wavelet transform
. Autoregressive (AR)
. Power spectral density function (PSD)
. Very low frequency (VLF)
 . Low Frequency (LF)
 . High Frequency (HF)
 . Low Frequency to High Frequency (LF/HF)
 . Detrended Fluctuation Analysis (DFA)
 . Correlation Coefficient
f»YZa ®Ë Á ¾ËY { .{̳ Ê» Y« Ã{Z¨fY {»Į Ê» ʧ»
{Â Á ¾ËY Ée] .dY Ê¿Z» É Ê´f^¼Å ½YÌ» ´¿ZÌ] į
{ ºfË´·Y ¾ËY ÊËYZ¯ Ä] ÃÂu ¾ËY { Á ËZ Ä] d^¿
.dY ZfËY Ì£ Ê¿Z» ÉZÅ É { {Âm» Ê´f^¼Å Ä^Zv»
5/0<Ƚ<
0 Ê´f^¼Å Ã|ÀÅ{ ½Z¿,ʨÀ» 1<Ƚ<5/0 ´¿ZÌ]
Ê´f^¼Åd^j» Ä°ÌeÂ{ Á5/0=Ƚ=1 ,Ƚ Á1>Ƚ |Z] \Ìef]
Ê¿Z» É Ê´f^¼ÅZ¿ Ã|ÀÅ{ ½Z¿ Ë¿ ,
f
1
Ê¿ÁY] Ë¿ Á .dY
Y ¾ËY Ä] ,|Z] d^j» ´Ë|°Ë Z] f»YZa Á{ Ê´f^¼Å \Ë ³
Ä·Z» į ÊËZ§ { į dZÀ» ,Ã| ¹Zn¿Y Ê] Á ®Ë ËY§Y
´Ë{ f»YZa ËY§Y Z] f»YZa Ì¿ Á f»YZa ½M ÅZ¯Z] ÅZ¯
´Ë{ f»YZa dY ÃY¼Å . ´Ë|°Ë Z] f»YZa Á{ Ê´f^¼Å \Ë ³Y
¹Zn¿Y Ê] Á Ä·Z» į ÊËZ§ { į dZÀ» ¾ËY Ä] ,|Z] ʨÀ»
,Ã| f»YZa ®Ë ËY§Y ´Ë{ f»YZa ÅZ¯ Z]Á ½M ÅZ¯
f»YZa ´Ë{ f»YZa ËY§Y Z]dY ÃY¼Å.
{ |À¿YÂe ʼ¿ ÊmYzfY ÉZŠʳËÁ Ê»Z¼e Ä°ÀËY Ä] ÄmÂe Z]
ÊËZZÀ ÄÀaM ĸ¼u {Y|y ÃZ] cZ^ Ä] Á Z] Ã|ÀÀ¯ ®¼¯
\ZÀ» ʳËÁ [Zzf¿Y ¹| cÂ { Ìze ºfÌ ÊËYZ¯ ´Ë{
Ä·Z¬» ¾ËY { ,d§ZË |ÅYÂy ÅZ¯ Âe Y§Y ¹¿SPSS
15
Äz¿
24 Á½Â»M ÉZ»M t ÉZŠĿ¼¿ÊmÁ
16
į ÊËZŠʳËÁ [Zzf¿Y Ä]
|ÀfY{ µZ»¿ Á [YÂy ÄÀaM ÃZ] Á{ ¾Ì] ÉY{ ÊÀ» ÉË~aËZ¼e
.| ÄfyY{alËZf¿ d|] Ã|»M Y ¿ ÉZ»M Ê¿Z»Y{ZÀ» į {Â]
Y|¬»p-value Y f¼¯05/0 |Z] . ÉZÅ Ä°^ Y Ä»Y{Y { Ê^
ÊÂÀ»
17
ze dÆm.dY Ã| Ã{Z¨fY ÄÀaM ÉZ¼Ì] ÌANN
{ Ì¿ Á ÊËZZÀ ¦¸fz» ÉZÅ ÄÀÌ»ÉZÅ{]Z¯ cÁZ¨f» |À¿Z»
¾Ì¼ze ZÅf»YZa {» Ã{Z¨fY Y« .|¿̳ Ê» Ä°^ ÉZÅ Ê^
ÉYY{ ½Âr¼Å ÊËZËY» d ÓZ] { cZ^Zv» ,ÉY» d̸]Z«
¥~u ,Ë¿ d»ÁZ¬» { ]Y] Ì̤e ZaZÅf»Y Á d̸]Z« »M Á
|ÀfÅ ÉË~a ºÌ¼e[21] .Ä] ʸ¯Â ZfyZ Ä°^ ÉZÅ Ê^ Y
Ê» ½YÂe Ä] Á{ Äf{ ʸ¯ Ä°^ ÉZÅ ÂzÌa Á Ä°^ ÉZÅ
Êf³Z] .{¯ É|À] Ĭ^ Ä°^ ÉZÅ ÂzÌa ÉYY{ ZfyZ Ã{Z
ÉY |ÀfÅ Á ¶°» ÉY|ËZa Ä°^ ÉZÅ Êf³Z] Y |¿Y|¿. Ä°^
ÉZÅ ÂzÌa ÉYY{ Y¿Y ʨ¸fz» |ÀfÅ.
Ĭ^ Á{ Y Ä·Z¬» ¾ËY {
ÄËÓ |Àq ½Áfba Ã|ÀÀ¯ É|À]
18
ÊZ ÄËZa ]Ze Á
19
dÆm
Y|¿Y ÄÀaM ĸ¼u {Y|y ÃZ] ÊËZZÀ{.| Ã{Z¨fY [YÂy É ¸]
 . Statistical Package for the Social Sciences
(SPSS)
 . paired sample t-test
 . Artificial Neural Network (ANN)
 . Multi-layer perceptron (MLP)
 . Radial basis function (RBF)
2017 24th national and 2nd International Iranian Conference on Biomedical Engineering (ICBME), Amirkabir University of
Technology, Tehran, Iran, 30 November - 1 December 2017
978-1-5386-3609-1/17/$31.00 ©2017 IEEE 229
¹Y³ZË{ ¶° { Ã| ÊuY1 Ì£ Y ÄÀaM {Y|y ÊËZZÀ ÃÂv¿ ,
Z¿ Y ¦¸fz» Ê¿Z» ÉZÅ ÃZ] { ÄÀaM.|Å{ Ê» ½
¶°1: [YÂy ÄÀaM ÊËZZÀ ÂÀ» Ä] ¹Y³ZË{ ¸] ÊuY
Ä°^ ÉZÅ Ê^ ½Áfba Y ¾Ë|Àq ÄËÓ Y Ã³ ¶»Z ®Ë
ÄËÓ ,É{ÁÁ ®Ë ÄËÓ ÊmÁy Á ½ZÆÀa ÄËÓ |Àq ®Ë ¶Ì°e
Ã| .dY.
Ä°^ Ê^ Y Ä·Z¬» ¾ËY { Ã{Z¨fY {» ®Ë ÄËÓ ¶Ì°e ½ZÆÀa
Ã|
dY į ¾ËY ÄËÓ ½ZÆÀa ¶»Z Ã³ ¶]Z«Ì£ ÉZÅ Ã|ÅZ» ÉY
.dY Å Ã³ ½ZÆÀa Ê]Ze ¼m Y Y{ ½Á É{ÁÁ .dY ]Ze
{» ,Ã{Z¨fY ]Ze Z µZ§ ¹Z¿ {Y{ Á Y ¿ { ÉZÅ ½Á
Ã³ Å ÉY] Ã| Äf§³Ä] ĸÌÁ ºfË´·Y ¾Ì¼ze z» Ê»
{Â.
Ä·Z¬» ¾ËY { ½ZÆÀa ÄËÓ ÉZÅ ½Á¿ ÉY] Z µZ§ ]Ze Y
®Ì·Â]bÌÅ d¿Y¿Ze
1
Ê]Ze į Ë~a ªf» Á Y|¿Y¯ ,ʬ̬u Ã{Â] Á
] Y ®Ë Ze ¨ ¾Ì] ÊmÁy Ë{Z¬» ,É{ÁÁ ¦¸fz» Ë{Z¬» ÉYY Ä]
,|¿Y{³ Ê» Ã{Z¨fY .| µZ§ ]Ze Y Ì¿ ÊmÁy ÄËÓ ½Á¿ ÉY]
d»Ô Z
2
Y|¬» Ä ÉYY{ ]Ze ¾ËY ÊmÁy į1 ,1- Á0 ,dY
) ÉZÅ Ä·{Z» .| Ã{Z¨fY4) Á (5( ÊZË ]YÁ ½Z¿ Y ]YÂe ¾ËY
.|Å{ Ê»
)4 (
xx
xx
ee
ee
xf
)(
)5 (
01
00
01
)(
!
x
x
x
xf
{» ÉÂv¿ Ä] |ËZ] Ä°^ Å ,[¸» xZa Ä] ½|Ì ÂÀ» Ä]
.{̳ Y« »MÁ ÉZÅ »M Ì]´¿Z ÃÂv¿ aY{ Ä°^
ÉÁ Ã{Y{ ÉZÅ É{ÁÁ dY Ä°^ »M dÆm É{ZË ÉZÅ Á .
ÂmÁÊË´·Y »M Äf{ Á{ Ä] ʸ¯ Â Ä] į {Y{ {
3
ÉY Äf{ Á
4
.|¿Â Ê» ºÌ¬e Ä] (É{ÁÁ Y{]) ´·Y ®Ë Y|f]Y ÊË´·Y Á {
. Hyperbolic tangent
. Sign
. Pattern mode
. Batch mode
Ê» Y« Ä^Zv» {» ½M Y ÊZ¿ ÊmÁy Á {Â Ê» µZ¼Y Ä°^
»M Ä] ÄmÂe Z] b .{Â Ê» {ZnËY ÉY Äv· ÉZy Á {̳
Ê¿ZÁ] ½Á Ã| ¯}
5
Z] |Ë|m É{ÁÁ Y{] b Á Ã|
dY ½M Á ¾ËY ¶°» .{Â Ê» µZ¼Y Ä°^ Ä] |Ë|m ÉZÅ ½Á
ÉZÅ ½Á {ZnËY hZ] {Â {YÁ tÌvZ¿ É{ÁÁ É´·Y ®Ë ³Y į
Y Y{a |¿YÂe Ê» ½Á ¾ËY Y Ã{Z¨fY ¾ËY]ZÀ] Á Ã| Â]»Z¿
Y ¶Ì·|] Á ¾ËY { Z»Y .¯ ¶°» Zq{ É|] ¶uY» { Ä°ÀË
É{ÁÁ É´·Y ®Ë ÉYY Ä] ½Z» Å { Ä°^ ÊmÁy ÉZy
ZÌ» ]Ze Ê¿Z ¶«Y|u ,{Â Ê» tÌve
6
ʸv» cÂ Ä]
7
Á
¶Ì·|] ÄÀÌÆ] ÉZÅ ½Á Ä] ½|Ì ÄnÌf¿ { Á Ã| ¹Zn¿Y ÉY Äv·
.{Â] |ÅYÂy e Ã{Z Äv· Å { ´·Y ®Ë d̼ÅY »M Á {
n°Ë cÂ Ä] ZÅ´·Y ¹Z¼e ,ÉY Äf{ a Á {Â Ê» Ã{Y{ Ä°^ Ä] Z
Y ½Á Ê¿ZÁ] ÉY] Zy ½M Y ,ÊmÁy ÉZy ¾Ì´¿ZÌ» Ä^Z
Ä»Y{Y Zy ZÌ» ]Ze ½{¯ ºÌÀÌ» Ze Z¯ ¾ËY Á {Â Ê» Ã{Z¨fY
{ ¶°» ¹| Ä] ½YÂe Ê» Á ¾ËY ÉZËY» Y .{¯ |ÅYÂy Y|Ìa
{ µÂ¼» Ì£ ½Á |Àq ®Ë {ÂmÁ Z] Êfu ½Á ÊËY´¼Å
fz» ¶uY»»M ¦¸ {¯ ÃZY, ¾Ì´¿ZÌ» Y Á ¾ËY { YË
.{Â Ê» Ã{Z¨fY ½Á Ê¿ZÁ] ÂÀ» Ä] ZÅZy ¼n»
Ê^ Ä°^RBF Ó ®Ë Z] ÃY¼Å Â¸m ÉZÅ Ä°^ ¿ Y Ì¿ ÄË
ÊmÁy Á ½ZÆÀa ÄËÓ ÉZÅ ½Á¿ ÉY] Z µZ§ ]YÂe .dY ½ZÆÀa
Ä°^ »M .|¿Â Ê» [Zzf¿Y Êy Á ¾Ì³ \Ìef]RBF
{ ,
É{ÁÁ ÉZÅ Ã{Y{ Y Ã{Z¨fY Z] µÁY ĸu» da ½Á|] cÂ Ä]
8
Á ¹Zn¿Y É|À] ÄÂy ÉZÅ ºfË´·Y ®¼¯ Ä]Á Ã| ]Ze ÉZÅf»YZa
Á ¯Y») ÊZ ÄËZa (ÄÌuZ¿  ĸu» { b .|ËM Ê» d|]
Z] Á Y ÊmÁy Á ½ZÆÀa ÄËÓ ¾Ì] ÉZÅ ½Á Ä^Zv» ÉY] ¹Á{
da
9
\Ì ÅZ¯ ÉZÅ ºfË´·Y Á
10
Êy ½ÂÌ³ Á
.{Â Ê» Ã{Z¨fY
½ÂÀ¯Ze ºfË´·Y ÉZŠʨ¸fz» ÂÀ» Ä]»M Ä°^ ÄWYY Ã|
.dY dË» Å Á d^¿ Ä] Á ËZ Á d«{ ½YÌ» {
.{Â Ê» ÄÔy É̳{ZË d ÄÀÌÆ] ºfË´·Y Y Ä·Z¬» ¾ËY {
cY¯Z» ±^¿Â· ÉZ
11
dÆm Ê^ZÀ» ÊËY´¼Å d ÉYY{ į
Ã| ½ZÌ] ÉZÅ Ä°^ »M ÂÀ» Ä] ,dY [¸» ÊmÁy {ZnËY
.| Ã{Z¨fY ¶^« ÉZÅ z] {ÉY] ¾ÌÀr¼Å »M fÆ] Ä°^ Á
Z¨f» Z̬» ¶Ì·|] ,ZÅ É{ÁÁ Y ) Ä·{Z»6( ½{¯ µZ»¿ ÉY]
É{ÁÁ Ä°^ Ä] µZ¼Y dÆm Ã{Z¨fY |. ,Ä·{Z» ¾ËY {
i
x
. Update
. Performance
. Local
. Unsupervised
. Supervised
 . Gradient descent
 . Levenberg Marquardt
2017 24th national and 2nd International Iranian Conference on Biomedical Engineering (ICBME), Amirkabir University of
Technology, Tehran, Iran, 30 November - 1 December 2017
978-1-5386-3609-1/17/$31.00 ©2017 IEEE 230
,ÄÌ·ÁY É{ÁÁ Y|¬»
mean
x
,ÄÌ·ÁY É{ÁÁ ¾Ì´¿ZÌ» Y|¬»
max
x
Y|¬»
,ÄÌ·ÁY É{ÁÁ º¼Ë¯Z»
min
x
Á ÄÌ·ÁY É{ÁÁ º¼ÌÀÌ» Y|¬»
Ni
x
.dY ÄÌ·ÁY É{ÁÁ Ã| µZ»¿ Y|¬»
)6 (
5.05.0
minmax
u xx
xx
x
meani
Ni
{ ZÆf¿Y ÂÀ» Ä] ZËY]Ê lËZf¿ Ä°^ ÉZÅ Ê^ ,ÊÂÀ» Y
ÉZÅZÌ» ¾Ì´¿ZÌ» cYÁ~n» Zy
1
Á\Ë ÊË°¿ Y]
2
ª]Z»
) ÉZÅ Ä·{Z»7 Á ()8 (.| Ã{Z¨fY { Ä·{Z» ¾ËYY ÂÀ»
N
,ZÅ É{ÁÁ {Y|e
actual
y
Y|¬»ÊmÁy Ê«YÁ ,
predicted
y
Á Ã| ÊÀÌ] Ìa Y|¬»
y
ÊmÁy ¾Ì´¿ZÌ» Y|¬» .dY
)7 (
¦
N
i
predictedactual
iyiy
N
MSE
1
2
))()((
1
)8(
¦¦
¦
22
)()(
))((
predictedpredictedactualactual
predictedpredictedactualactual
yyyy
yyyy
R
3- ÉZ ÄÌ^ lËZf¿
Z¼Ì] cÁZ¨f» ËY Y ¦¸fz» ÉZŠʳËÁ kYzfY Y a ,
cÂ Ä] ÊnËZf¿ ÉZÅ ¶°2 Ze4 |»M d|]¶° .2 ÉZÅf»YZa
Y Ã| kYzfYµZÀ´Ì ½Z» ÃÂu Y \¸« ½Z] w¿ ÉË~aÌ̤e
.|Å{ Ê» ½Z¿ Ã|»M d|] Ë{Z¬» ÄËZ¬» Z¼Ì] į |Å{ Ê» ½Z¿
É{Y|¿Y ÄÀaM d·Zu Ä] d^¿ f¼¯ \¸« ½Z] µfÀ¯ d·Zu {
.{Y{ É¯» ÄÀaM d·Zu Y f¼¯ Ì¿ [YÂy É{Y|¿Y d·Zu Á [YÂy
ʳËÁ Z] Z^eY { Â» ¾ËY °mean RRI .{Zf§Y ©Z¨eY
ʳËÁ Y|¬»SDNN µfÀ¯ Zu { Ì¿ Y fÌ]OSA .dY
Ë{Z¬» ÄËZ¬» {» { ʳËÁ ¾ËY OSA ÁCSA |Å{ Ê» ½Z¿
{ ʳËÁ ¾ËY Y|¬»CSA Y fÌ]OSA į dY Äf§ZË ÅZ¯
{ fÌ] cYÌ̤e Ã|ÀÅ{ ½Z¿CSA ´Ë{ d·Zu Á{ Ä] d^¿
.dY į Ê¿Z» { ʳ|ÌrÌa į d§³ ÄnÌf¿ ½YÂe Ê» ¾ËY]ZÀ]
b Á É{Y|¿Y ÄÀaM d¼ Ä] µfÀ¯ µZu Y Z¼Ì] ËY ÄÀaM
d·ÂËÁ ½YÂe ¦Ì Á Y fÆ] Ê] ÉY] .{Â Ê» fÌ] É¯»
.| Ã{Z¨fY Ê¿Z¯§ ÉZÅf»YZa Ä^Zv» ÉY]¶°3 , ÄËZ¬»
½Z» ÉZŠʳËÁ- ½YÂe ¦Ì Ê·Z´q Y Ã| kYzfY ¿Z¯§
[YÂy ¦¸fz» ËY { d·ÂËÁ .|Å{ Ê» ½Z¿ Y Y Ã{Z¨fY lËZf¿
Z¿ ¿Z¯§ ½Z» ÉZŠʳËÁ µfÀ¯ ËY Y {§ Ä°Ì»Z´ÀÅ ,{Y{ ½
ÉZŠʳËÁ { ¶]Z« ËY§Y {Á Ê» [YÂy ÄÀaM d¼ Ä]VLF ,
LF ÁHF .|f§Y Ê» ©Z¨eY ÉZŠʳËÁ VLF ÁLF Z¼Ì] Ä°Ì¿Z»
dÌÁ {OSA {Y{ Y« µfÀ¯ d·Zu Ä] d^¿ ÁCSA fÌ]
ʳËÁ {» { .dYHF
d·Zu Á dY Y«] d·Zu ¾Ì¼Å Ì¿
OSA Äf°¿ .dY ÉfÌ] Y|¬» ÉYY{ ´Ë{ dÌÁ Á{ Ä] d^¿
. Mean squared error (MSE)
. Goodness of fit index
Y|¬» d§Y d̼ÅY WZuHF d·Zu {CSA Ä] d^¿ į dY
dÌÁOSA .dY Ã{Y{ d^¿ {Ây Ä] Y Éf¼¯ Y|¬» µfÀ¯ Á
ʳËÁ
HF
LF
{ d·ZuCSA ZfyY {Ây Ä] Y Y|¬» ¾ËfÌ]
dÌÁ Á Ã{Y{OSA Y µfÀ¯ Á É|] ÉZÅ Ä^e { Y|¬» ¿
ʳËÁ Ä°ÀËY Ä] ÄmÂe Z] .|¿Y{ Y«LF ÁHF Z] \ZÀf» \Ìef]
{ Z¼Ì] ¾f§³ Y« ,|ÀfÅ ®ÌeZb¼ Á ®ÌeZb¼YZa [ZY
Á Ã| ®Ì¿Zb¼YZa Á ®ÌeZb¼ [ZY |e hZ] ÄÀaM dÌÁ
) µZ³YÁ ÂeZb¼ µ{Ze Âf¯Z§ ¾f§ÓZ] hZ] ÄnÌf¿ {
HF
LF
Ê» (
¾ÌÀr¼Å .{ÂY|¬» Ã| |¿e É{ cZ¿Z¿ yZ)
1
D
( d·Zu {
CSA ËY ËZ Ä] d^¿®Ë Ä] ®Ë{¿ dY ¿ įZ Ã|ÀÅ{ ½
ºfÌ ʳ|ÌrÌa Äm{ ËY§Y d·Zu Ä] d^¿ ÄÀaM ËY {
µfÀ¯ ʳËÁ ¾ÌÀr¼Å .dY
2
D
{ Ì¿ d·Zu CSA Ä] ®Ë{¿
5/0 ½Z¿ Y Ê¿Z» É Ê´^¼ÅZ¿ ½YÌ» Ê¿ Ä] į dY Ã|
.|Å{ Ê» Y|¬»yZ Á Y Ã{Z¨fY Z] Ã|»M d|] ÉZÅDFA
¶° {4 .dY Ã| ÄÔy
¶°2: ÉZŠʳËÁ ÄËZ¬»Ê¿Z» ¹Z´ÀÅ ¦¸fz» ËY {
[YÂy
¶°3: ÉZŠʳËÁ ÄËZ¬»½Z»- ¿Z¯§ Y Ê·Z´q Y Ã| kYzf
d·ÂËÁ ½YÂe ¦Ì [YÂy ¦¸fz» ËY {
2017 24th national and 2nd International Iranian Conference on Biomedical Engineering (ICBME), Amirkabir University of
Technology, Tehran, Iran, 30 November - 1 December 2017
978-1-5386-3609-1/17/$31.00 ©2017 IEEE 231
¶°4: ÉZŠʳËÁ ÄËZ¬» Ã| kYzfY µZf¯§ ËY {
[YÂy ¦¸fz»
ÊÀ» ÉË~aËZ¼e į ÊËZŠʳËÁ Ä] ÄmÂe Z] Ä»Y{Y { ¾Ì] ÉY{
|ÀfY{ µZ»¿ d·Zu Á ÄÀaM ½Z» µÁ|m ª]Z»)1( ʳËÁ \̯e Ä]
Á | ÄfyY{a Y ÊËZZÀ dÆm ÊÂÀ» Ê^ ÉZÅ Ä°^
.| ÄfyY{a [YÂy ¹Z´ÀÅ ¦¸fz» ËY { {§ dÌÁ ÊËZn¿M Y
{ dY{ |ÅYÂz¿ ÉfÆ] xZa ÃY¼Še] ÉZŠʳËÁ \̯e į
Ä°^ »M dÆm ¦¸fz» ÉZŠʳËÁ \̯e Y Ä·Z¬» ¾ËY
Ë{ cZ^ Ä] .| [Zzf¿Y \̯e ¾ËfÆ] dËZÆ¿ { Á | Ã{Z¨fY ´
Ä̸¯ ZfyZ ¦¸fz» ÉZŠʳËÁÉ{ÁÁ Âe Ä°^ Ê^
½Áfba q ÄËÓ Á Ä°^ ]Ze ÄËZa ÊZ {» »M Á ½Â»M
|Àf§³ Y« į { ÄnÌf¿ z» ,| lËZf¿ Ä] d{ Ã|»M Y Ä°^
Ê^ ½Áfba |Àq ÄËÓ Y ÉfÆ] {°¸¼ d^¿ Ä] Ä°^ ]Ze
ÄËZa ÊZ Y{Ây] .{Â] Zf¿ Y Ã{Z¨fY Ä°^ ÉZÅ Ê^ {
d·Zu ½Â»M µÁ|m {2 ½Z¿ .dY Ã| Ã{Y{
µÁ|m1: ÉZŠʳËÁ ÉË~aËZ¼e¦¸fz» ËY { Ã| ʧ»
µÁ|m2: lËZf¿Ä°^ ¦¸fz» ZfyZ Z] ʳËÁ \̯e ¾ËfÆ]
4- É̳ ÄnÌf¿
¨Àe cÓÔfyY [YÂy Ê d̨̯ |¿YÂe Ê» ʳ|¿ ÌiZe dve Y {Y§Y
Á Ã{Y{ Y« ʨÀ»±» y [YÂy { ½YZ¼Ì] { Y Zq{ ÊËZZ¿
^¸«Ë|e Ê| .|ËZ¼¿ ¾ËYÄ·Z¬» ¥|Å Z]ÊËZZÀ É{Y|¿Y ÄÀaM Á
É¯» [YÂy ½YZ¼Ì] Y Ã| d^i Ê^¸« ÉZÅ µZÀ´Ì Y Ã{Z¨fY Z]
[YÂy ¦¸fz» ¶uY» ¾Ìu { .| ¹Zn¿Y lËZf¿ ½Z¿ Ã|»M d|]
{ |¿YÂe Ê» É{ZÆÀÌa ºfË´·Y {Y ÉË~aÌ̤e µZÀ´Ì Y Ã{Z¨fY Z]
ÔfyY Ê·Â^« ¶]Z« cÂ Ä] \¸« ½Z] w¿ ÊËZZÀ Y [YÂy
.|À¯ Y ºÆ» Á Z] ÉZŠʳËÁÁ Ã| ÄWYY { Ä·Z¬» ¾ËY ʳ{Z
d^i ħ Ä] ½Á¬» Á ½{Â] ÂÀ» Ä] [Zzf¿Y ½YZ¼Ì] Z]y Á ÓZ]
½YZ¼Ì] ªÌ«{ É´·Z]£ ÉY] dÆm ®a Ä] ÄmY» Ìze
f¬Ì«{Y [YÂy Ê¨Àe cÓÔfyY ½Z»{ Ád. ÉZÅ dË{Á|v» Y
ÅÁa Zu, |e{Y fÌ] ʸÌy Ã|ÀÀ¯ d¯ ½Y{» Ä] d^¿
½Z¿ {Y|e ËY§Y Z] ½YÂe Ê» į {Â] ½Z¿ d^i Á Ã|ÀÀ¯ d¯
,ZÅ ½M Y µZÀ´Ì ÊfÌÀm µ{Ze.{¯ ºÅY§ Y ÉfÆ]
5. mY»
1. Hornero, R., et al., (2007´ 8WLOLW\ of approximate entropy from
overnight pulse oximetry data in the diagnosis of the obst ructive sleep
apnea syndrome´ Biomedical Engineering, IEEE Transactions on. 54(1),
pp 107-113.
2. Khandoker, A.H. and M. Palaniswami, (2011), ³modelin g respiratory
movement signals during central and obstructive sleep apnea events using
electrocardiRJUDP´ Annals of biomedical engineering, 39(2), pp 801 -811.
3. Phillipson EA. (2013 ³Sleep apnea´ In: Longo D, Fauci A, Kasper D,
Hauser S, Jameson J, Jameson J, editors. Harrison's principles of internal
medicine. 18th ed. New York, NY: McGraw-Hill. pp 2186-2189.
4. Phillips B, Kryger MH. (2005), ³Management of obstructive sleep
apnea- hypopnea syndrome: overview´ In: Kryger MH, Roth T, Dement
WC, editors . Principles and practice of sleep medicine. 4th ed.
Philadelphia, PA: Saunders.
5. Manni R, Terzaghi M. (2 010´ &RPRUELGLW\ between epilepsy and
sleep disorders´ Epilepsy Res, 90(3), pp 171-177.
6. Kryger MH, Basner R. (2011 ), ³Management of obstructive sleep apnea
in adult [Online]. Available from:
URL:http://www.uptodate.com/contents/management -of-obstructive-
sleep-apnea-in-adults.
7. Nobili L, Proserpio P, Rubboli G, Montano N, Didato G, Tassinari CA.
(2011³Sudden unexpected death in epilepsy (SUDEP) and sleep´ Sleep
Med Rev; 15(4), pp 237-46.
8. Katagal P, Yardi N. (2008 ´ The relationship between sleep and
epilepsy´ Seminars in Pediatric Neurology, 15(2), pp 42-48.
9. Shen, K.Q., et al., A feature selection method for multilevel mental
fatigue EEG classification. IEEE transactions on bio-medical engineering,
2007. 54(7): p. 1 231-1237.
10. Koley, B. and D. Dey, On-line Detection of Apnea/Hypopnea Events
Using SpO2 Signal: A Rule-based Approach Employing Binary Classifier
Models. 2014.
11. Várady, P., et al., A novel method for the detection of apnea and
hypopnea events in respiration signals. Biomedical Engineering, IEEE
Transactions on, 20 02. 49(9): p. 936 -942.
12. Goldberger, A.L., et al., Physiobank, physiotoolkit, and ph ysionet
components of a new research resource for complex physiologic signals.
Circulation, 2000. 101 (23): p. e215-e220 .
13 6W 9LQFHQW¶V 8QLYHUVLW\ +RVSLWDO8QLYHUVLW\ &ROOHJH 'XEOLQ
SleepApneaDatabase.(2008).[Online].Available:
http://www.physionet.org/pn3/ucddb/.
14. V.Viknesh & P. Ram Prashanth, Matlab implementation of ECG
signal processing, IOSR Journal of VLSI and Signal Processing (IOSR-
JVSP), Volume 3, Issue 1 (Sep. ± Oct. 2013), PP 40-47.
15. Pan, J., and Tompkins, W.J., "A real time QRS detection algorithm,".
IEEE Transaction on Biomedical Engineering,Vol. 32, No.3,pp. 230-36,
1985.
16. Thuraisingham RA, Preprocessing RR interval time series for heart
rate variability analysis and estimates of standard deviation of RR
intervals, Comput Methods Programs Biomed. 2006 Jul;83(1):78-82.
17. European Society of Cardilogy. Heart rate variability: Standarts of
measurements, physiological interpretation, and clinical use. European
Heart Journal, 17:354±381, March 1996.
18. I. Gritti1, S. Defendi, C. Mauri1, G. Banfi1, P. Duca, G. S. Roi, ''Heart
rate variability standards of measurement, physiological interpretation,
and clinical use in Mountain Marathon Runners during Sleep and
after Acclimatization at 3480 m'', Journal of Behavioral and Brain
Science, Vol. 3 No. 1, 2013.
19. T. Force, ''Heart rate variability: standards of measurement,
physiological interpretation and clinical use. Task Force of the European
Society of Car diology and the North American Societ y of Pacing and
Electrophysiology'', Circu lation, Vol. 93, No. 5, March 1996.
20. Peng, C.K.; et al. (1994). "Mosaic organization of DNA
nucleotides". Phys. Rev. E. 49: 1685±
1689. doi:10.1103/physreve.49.1685.
21. 7' %DW]HO .< /HH ³$ GLDJRQDOO\ UHFXUUHQW neural network
approach to sensorless operation of the permanent magnet synchronous
PRWRU´,((( Power Engineering Society Summer Meeting, vol.4,2000.
2017 24th national and 2nd International Iranian Conference on Biomedical Engineering (ICBME), Amirkabir University of
Technology, Tehran, Iran, 30 November - 1 December 2017
978-1-5386-3609-1/17/$31.00 ©2017 IEEE 232
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
Obstructive sleep apnea (OSA) causes a pause in airflow with continuing breathing effort. In contrast, central sleep apnea (CSA) event is not accompanied with breathing effort. CSA is recognized when respiratory effort falls below 15% of pre-event peak-to-peak amplitude of the respiratory effort. The aim of this study is to investigate whether a combination of respiratory sinus arrhythmia (RSA), ECG-derived respiration (EDR) from R-wave amplitudes and wavelet-based features of ECG signals during OSA and CSA can act as surrogate of changes in thoracic movement signal measured by respiratory inductance plethysmography (RIP). Therefore, RIP and ECG signals during 250 pre-scored OSA and 150 pre-scored CSA events, and 10 s preceding the events were collected from 17 patients. RSA, EDR, and wavelet decomposition of ECG signals at level 9 (0.15-0.32 Hz) were used as input to the support vector regression (SVR) model to recognize the RIP signals and classify OSA from CSA. Using cross-validation test, an optimal SVR (radial basis function kernel; C = 2(8) and ε = 2(-2) where C is the coefficient for trade-off between empirical and structural risk and ε is the width of ε-insensitive region) showed that it correctly recognized 243/250 OSA and 139/150 CSA events (95.5% detection accuracy). Independent test was performed on 80 OSA and 80 CSA events from 12 patients. The independent test accuracies of OSA and CSA detections were found to be 92.5 and 95.0%, respectively. Results suggest superior performance of SVR using ECG as the surrogate in recognizing the reduction of respiratory movement during OSA and CSA. Results also indicate that ECG-based SVR model could act as a potential surrogate signal of respiratory movement during sleep-disordered breathing.
Article
Full-text available
The monitoring of breathing dynamics is an essential diagnostic tool in various clinical environments, such as sleep diagnostics, intensive care and neonatal monitoring. This paper introduces an innovative signal classification method that is capable of on-line detection of the presence or absence of normal breathing. Four different artificial neural networks are presented for the recognition of three different patterns in the respiration signals (normal breathing, hypopnea, and apnea). Two networks process the normalized respiration signals directly, while another two use sophisticatedly preprocessed signals. The development of the networks was based on training sets from the polysomnographic records of nine different patients. The detection performance of the networks was tested and compared by using up to 8000 untrained breathing patterns from 16 different patients. The networks which classified the preprocessed respiration signals produced an average detection performance of over 90%. In the light of the moderate computational power used, the presented method is not only viable in clinical polysomnographs and respiration monitors, but also in portable devices.
Chapter
Sleep-disordered breathing (SDB) exists along a continuum of severity, and the boundaries between pathologic and normal breathing are still not precisely defined. At one end of the spectrum is snoring of varying degrees of intensity; at the other is severe obstructive sleep apnea with profound hypoxemia and sleep disturbance. In between those extremes are varying degrees of inspiratory airflow resistance, including respiratory-effort related arousals (RERAs), hypopneas, and frank apneas. Other respiratory abnormalities that may be caused or exacerbated by sleep include hypoxemia, hypoventilation, and central apneas. The reported prevalence of obstructive sleep apneaa-hypopnea (OSAH) is rising, driven by the increasing recognition of persons who have the disorder and by the increasing ubiquity of risk factors, notably obesity and aging, that enhance the risk of SDB.
Article
This paper presents an online method for automatic detection of apnea/hypopnea events, with the help of oxygen saturation (SpO 2) signal, measured at fingertip by Bluetooth nocturnal pulse oximeter. Event detection is performed by identifying abnormal data segments from the recorded SpO 2 signal, employing a binary classifier model based on a support vector machine (SVM). Thereafter the abnormal segment is further analyzed to detect different states within the segment, i.e., steady, desaturation, and resaturation, with the help of another SVM-based binary ensemble classifier model. Finally, a heuristically obtained rule-based system is used to identify the apnea/hypopnea events from the time-sequenced decisions of these classifier models. In the developmental phase, a set of 34 time domain-based features was extracted from the segmented SpO 2 signal using an overlapped windowing technique. Later, an optimal set of features was selected on the basis of recursive feature elimination technique. A total of 34 subjects were included in the study. The results show average event detection accuracies of 96.7% and 93.8% for the offline and the online tests, respectively. The proposed system provides direct estimation of the apnea/hypopnea index with the help of a relatively inexpensive and widely available pulse oximeter. Moreover, the system can be monitored and accessed by physicians through LAN/WAN/Internet and can be extended to deploy in Bluetooth-enabled mobile phones.
Article
The risk of sudden unexpected death is considered to be notably higher in patients with epilepsy with respect to the general population. Sudden unexpected death in epilepsy (SUDEP) is probably caused by the peri-ictal concurrence of a number of different predisposing and precipitating factors. Among these, the presence of a seizure before the fatal event is the only feature that seems to be constantly present. Different mechanisms, namely cardiac arrhythmias, respiratory dysfunctions, dysregulation of systemic or cerebral circulation have been suggested as potential physiopathological mechanisms. Moreover, clinical data seem to suggest that SUDEP could occur preferentially during sleep. In order to assess a possible relationship between sleep and SUDEP, we have analyzed studies in which sufficient information about the circumstances of deaths was available. Our analysis confirms that the relationship between sleep and SUDEP is not given by chance as the percentage of possible sleep-related SUDEP is higher than 40% in the majority of studies. We will discuss the possible longstanding and precipitating mechanisms involved in the interaction between sleep and epilepsy likely to favour SUDEP occurrence. In this perspective, possible preventive measures will be hypothesized.
Article
Despite being relatively common and potentially able to have clinical and pathophysiological consequences, the comorbidity between epilepsy and sleep disorders is poorly investigated in the literature and rarely taken into consideration by clinicians in general practice. There is increasing evidence that obstructive sleep apnoea (OSA) coexists in epilepsy (in 10% of unselected adult epilepsy patients, 20% of children with epilepsy and up to 30% of drug-resistant epilepsy patients). A few lines of evidence suggest that continuous positive airway pressure treatment of OSA in epilepsy patients improves seizure control, cognitive performance and quality of life. Parasomnias and epileptic seizures can coexist in the same subject making the differential diagnosis of these conditions particularly challenging. In childhood, a frequent association between epilepsy and NREM arousal parasomnias, enuresis and rhythmic movement disorder has been documented. A particular pattern of association has been found between nocturnal frontal lobe epilepsy (NFLE) and NREM arousal parasomnias, the latter being found in the personal or family history of up to one third of NFLE patients. As far as REM parasomnias are concerned, REM sleep behaviour disorder, unrecognised or misdiagnosed, has been found to co-occur in 12% of elderly epilepsy patients. Patients with epilepsy often complain of poor, non-restorative sleep; however, insomnia in epilepsy is poorly investigated, with the literature giving conflicting prevalence data and no information on the impact of this disorder on seizure control, or on the best therapeutic approach to insomnia in this particular group of patients. A greater awareness, among clinicians, of the comorbidities between sleep disorders and epilepsy may help to prevent misdiagnosis and mistreatment. Sleep hygiene measures in epilepsy need to be more comprehensive, taking into account the various pathologies that may underlie disordered sleep in epilepsy patients.
Article
Although the indications for treatment of obstructive sleep apnea are controversial, there is general agreement that treatment is warranted in patients with a combined apnea and hypopnea frequency of 20 events per hour of sleep. General therapeutic measures include abstinence from alcohol and weight loss in obese patients. A minority of patients with specific abnormalities of the nose, pharynx, or jaw may require surgical management. In the absence of specific upper airway abnormalities, the most effective treatment is nasal continuous positive airway pressure or, if necessary, bilevel positive airway pressure during sleep. The role of uvulopalatopharyngoplasty remains controversial, due in part to the reported variability in long-term efficacy of this procedure. Drug therapy may be effective in selected patients.