International Journal of Engineering Trends and Technology (IJETT) – Volume 28 Number 8 - October 2015
Feature Extraction of Foetal Heart Rate Variability Using MATLAB
Mrs. Sanhita Manna
Assistant Professor, Department of Electronics & Instrumentation,
GITAM University, GITAM School of Technology, Nagadenahalli, Doddaballapur Taluk, Bengaluru, Karnataka, India
Abstract— For the foetal health monitoring it is very
important to analyse the foetal heart rate variability. There is
ongoing effort to develop computer based advanced method to
assists physician. Here a method is presented to extract the
feature of foetal heart rate variability. The extracted feature
shows good agreement with the traditional feature extraction
method.
Keywords — Foetal Heart Rate, Standard Deviation,
Poincare Plot, MECG
I. INTRODUCTION
Heart defect is one of the common birth defect and it is
leading the cause of the birth related death [1]. The continuous
foetal heart signal monitoring and analysis are the key aspects
to reduce this unforseen risk. The foetal electrocardiogram
(FECG) provides significant clinical information about the
health condition of the foetus. However due to the overlap of
the maternal electrocardiogram (MECG) in time domain and
frequency domain, the FECG have relatively low signal to
noise ratio, compared to MECG [2]. The normal foetal heart
rate is varied in between 110 bpm and 160 bpm.
In the last few years, many studies tried to demonstrate the
nonlinear nature of the foetal heart rate variability signal [3].
In this paper one feature extraction method has described.
II. ANALYSIS
The interaction between sympathetic nervous system (SNS)
and parasympathetic nervous system (PNS) is reflecting the
foetal heart rate variability [4]. During stressful condition the
sympathetic nervous system provides a compensatory
mechanism to improve the foetal heart rate [4].
The foetal heart rate variability can be analysed in three
different methods i.e. the time domain, frequency domain and
nonlinear method [5].
Fig. 1: RR interval time series
mean of FHR (min(FHR (i))). Next the standard deviation of
RR intervals can be define as [6]
....1
Where N is the total number of successive interval series.
The SDNN reflects the overall variation within the RR
interval series, where as the standard deviation of successive
RR differences (SDSD) given by
(a)
(b)
Fig.2: (a) RR(s) distribution, (b) HR (beats/min) distribution
....2
Where sFHR(i) is the value of the signal FHR(i) taken
every 2.5 second.
TABLE I
Variable
Mean RR
SDNN
Mean HR
STD HR
RMSSD
Units
ms
ms
1/min
1/min
ms
Value
623.8
62.5
97.29
11.52
40.8
B. Frequency Domain Method
In frequency domain methods a power spectrum density
(PSD) is calculated for the RR series. In the FHR variability
analysis the PSD estimation is performed using FFT based and
parametric AR modelling based. The advantage of FFT based
methods is the simplicity of implementation, while the AR
spectrum yields improved resolution especially for short
samples [7].
A. Time Domain Method
One of the easiest method is the time domain methods as it
is calculated from the series of successive RR interval or the
ISSN: 2231-5381
http://www.ijettjournal.org
Page 423
International Journal of Engineering Trends and Technology (IJETT) – Volume 28 Number 8 - October 2015
Fig. 3: FFT spectrum
TABLE II: FFT SPECTRUM RESULT
Frequency Band
Peak
Power
Power
(Hz)
(ms2)
(%)
VLF(0-0.04 Hz)
0.0039
1327
78.9
LF(0.04-0.15 Hz)
0.0625
125
7.4
HF(0.15-0.4 Hz)
0.3242
231
13.7
Fig. 6: Detrended Fluctuation plot (DFA)
TABLE IV: NONLINEAR RESULT
Variable
Units
Poincare plot
SD1
SD2
Approximate entropy(ApEn)
Sample entropy
Detrended Fluctuation
α1
α2
Fig. 4: AR spectrum
TABLE III: AR SPECTRUM RESULT
Frequency Band
Peak
Power
Power
(Hz)
(ms2)
(%)
VLF(0-0.04 Hz)
0.0039
2491
86.6
LF(0.04-0.15 Hz)
0.0430
196
6.8
HF(0.15-0.4 Hz)
0.3555
189
6.6
C. Nonlinear Method:
The nonlinear properties of FHRV
using measures such as Poincare plot
sample entropy [10],[11], detrended
[12],[13], correlation dimension [14],
plots [16]-[18].
have been analysed
[8], [9] approximate
fluctuation analysis
[15], and recurrence
(ms)
(ms)
Value
28.9
83.6
0.732
0.645
0.811
1.138
III CONCLUSION
The foetal heart rate variability is most important signal to
measure to understand foetus hypoxia. The signal analysis is
done by using MATLAB and result shows good agreement to
the theoretical value.
REFERENCES
Fig. 5 Poincare plot
One of the common nonlinear method plots is the Poincare
plot. It is a graphical representation of the correlation between
successive RR intervals.
[1]
Reza Sameni and Gari D. Cliford, A Review of Fetal ECG Signal
Processing Issues and Promising Directions, The Open Pacing,
Electrophysiology & Theory Journal, 2010.
[2]
Ungureanu M, Bergmans J, Oei S, Strungaru R. Fetal ECG extraction
during labor using an adaptive maternal beat subtraction technique.
Biomedizinische Technik 2007; 52(1):56–60.
[3]
Cerutti S, Baselli G, Civardi S, Ferrazzi E, Marconi A, Pagani M,
Pardi G. Variability analysis of fetal heart rate signals as obtained
from abdominal electrocardiographic recordings. Journal of Perinat
Med Official J of the WAPM 1986; 14(6):445–52.
[4]
Shahad N. Al-Yousif and M. A. Mohd. Ali, Cardiotocography Trace
Pattern Evaluation Using MATLAB, International conference on
Biomedical Engineering and Technology: IPCBB, 2011, vol. 11.
Widrow B, Glover J, McCool J, Kaunitz J, Williams C, Hearn R,
Zeidler J, Dong E, Goodlin R. Adaptive noise cancelling: Principles
and applications. Proceedings of the IEEE 1975;63(12):1692–716.
[5]
Detrended fluctuation analysis (DFA) measured the
correlation within the signal. The correlation is extended for
different time scales.
ISSN: 2231-5381
[6]
Dimitris G., Ioannis G. Tsoulos, G. Georgoulas and E. Glavas,
Classification of Foetal Heart Rate using Grammatical Evaluation,
IEEE SIPS, pp. 7803–9333, Mar. 2005.
http://www.ijettjournal.org
Page 424
International Journal of Engineering Trends and Technology (IJETT) – Volume 28 Number 8 - October 2015
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
J. Mateo and P. Laguna. Improved heart rate variability signal analysis
from the beat occurrence times according to the IPFM model. IEEE
Trans Biomed Eng, 47(8):985{996, August 2000.
M. Brennan, M. Palaniswami, and P. Kamen. Do existing measures of
Poincare plot geometry reflect nonlinear features of heart rate
variability, IEEE Trans Biomed Eng, 48(11):1342{1347, November
2001.
S. Carrasco, M.J. Caitan, R. Gonzalez, and O. Yanez., Correlation
among Poincare plot indexes and time and frequency domain measures
of heart rate variability. J Med Eng Technol, 25(6):240{248,
November/December 2001
O. Rompelman. Rhythms and analysis techniques. In J. Strackee and
N. Westerhof, editors, The Physics of Heart and Circulation, pages
101-120. Institute of Physics Publishing, Bristol, 1993.
P. Grassberger and I. Procaccia. Characterization of strange attractors.
Phys Rev Lett, 50:346-349, 1983.
T. Penzel, J.W. Kantelhardt, L. Grote, J.-H. Peter, and A. Bunde.
Comparison of detrended uctu-ation analysis and spectral analysis for
heart rate variability in sleep and sleep apnea. IEEE Trans Biomed
Eng, 50(10):1143{1151, October 2003.
G.D. Pinna, R. Maestri, A. Di Cesare, R. Colombo, and G. Minuco.
The accuracy of power-spectrum analysis of heart-rate variability from
annotated RR lists generated by Holter systems. Physiol Meas,
15:163{179, 1994
[15]
H.V. Huikuri, T.H. M•akikallio, P. Raatikainen, J. Perki•om•aki, A.
Castellanos, and R.J. Myerburg. Prediction of sudden cardiac death:
appraisal of the studies and methods assessing the risk of sudden
arrhythmic death. Circulation, 108(1):110-115, July 2003
[16]
C.L. Webber Jr. and J.P. Zbilut. Dynamical assessment of
physiological systems and states using recurrence plot strategies. J
Appl Physiol, 76:965-973, 1994.
[17]
E.J.M. Weber, C.M. Molenaar, and M.W. van der Molen. A nonstationary test for the spectral analysis of physiological time series
with an application to respiratory sinus arrhythmia. Psychophysiol,
29(1):55-65, January 1992.
[18]
J.P. Zbilut, N. Thomasson, and C.L. Webber. Recurrence quantication
analysis as a tool for the nonlinear exploration of nonstationary
cardiac signals. Med Eng Phys, 24:53-60, 2002.
[19]
Winnie Rachel Cherian, D.J.Jagannath and A.Immanuel Selvakumar,
Comparison of Algorithms for Fetal ECG Extraction, IJETT Vol. 9,
No-11, Mar 2014.
[20]
Cerutti S, Baselli G, Civardi S, Ferrazzi E, Marconi A, Pagani M, Pardi
G. Variability analysis of fetal heart rate signals as obtained from
abdominal electrocardiographic recordings. J of Perinat Med Official
J of the WAPM 1986; 14(6):445–52.
[21]
Martens S, Rabotti C, Mischi M, Sluijter R. A robust fetal ECG
detection method for abdominal recordings. Physiol Meas 2007;28:
373–88.
P.S. Hamilton and W.J. Tompkins. Quantitative investigation of QRS
detection rules using the MIT/BIH arrhythmia database. IEEE Trans
Biomed Eng, 33(12):1157{1165, December 1986.
ISSN: 2231-5381
http://www.ijettjournal.org
Page 425