Title: Biomedical Signal Processing Forum, August 30th 2005
1Biomedical Signal Processing Forum, August 30th
2005
2BioSens (2/2) wavelet based morphological ECG
analysis
- Joel Karel, Ralf Peeters, Ronald L. Westra
- Maastricht University
- Department of Mathematics
Biomedical Signal Processing Forum, August 30th
2005
3Biosens prehistory from 1994 onwards
TU Delft micro electronics
UM mathematics
Richard Houben Medtronic Bakken Research
Maastricht
Other companies
UM Fysiology Prof. M. Allessie
4STW project Biosens 2004-2008 Organization
research consortium
5Research topics studied so far
- 1. Efficient analog implementation of wavelets
- Analog implementation of wavelets allows
low-power consuming wavelet transforms for e.g.
implantable devices - Wavelets cannot be implemented in analog circuits
directly but need to be approximated A good
approximation approach will allow reliable
wavelet transforms - 2. Epoch detection and segmentation
- Application of the Wavelet Transform Modulus
Maxima method to T-wave detection in cardiac
signals. - 3. Optimal discrete wavelet design for cardiac
signal processing - What is the best wavelet relative to the data and
pupose?
6-
- 1. Efficient implementation of analog wavelets
- Biosens team J.M.H. Karel (PhD-student),
- dr. R.L.M. Peeters, dr. R.L. Westra
- Accepted papers BMSC 2005 (Houffalize, Belgium),
IFAC 2005 (Prague, Czech Republic), and CDC/EDC
2005 (Sevilla, Spain)
7Implementation of wavelets
- Analog implementation of wavelets allows
low-power consuming wavelet transforms for e.g.
implantable devices - Wavelets cannot be implemented in analog circuits
directly but need to be approximated
8Wavelet approximation considerations
- A good approximation approach will allow reliable
wavelet transforms - will allow low-order implementation (low-power
consuming) of wavelet transforms - allows approximation of various types of wavelets
- is relatively easy applicable
9Wavelet approximation methodology
10Initial high-order discrete time MA-system
- Sampled wavelet function
- Required i.r.
- State-space realization in controllable companion
form
11Model reduction with balance and truncate
- Balance Lyapunov-equations
- Note that P is identity matrix
- Reduce system based on Hankel singular values
12L2-approach
- Wavelet is approximated by impulse response of
system - The model class is determined by the system whos
i.r. is used as a starting point
13State-space realization
14Morlet approximation
15Daubechies 3 wavelet
16Wavelet approximation methodology
- Allows approximation of wavelet function that
previously could not be approximated - Allows approximation of more generic functions
- Publications accepted for IFAC 2005 (Prague,
Czech Republic), CDC/ECC 2005 (Sevilla, Spain),
and BMSC 2005 (Houffalize, Belgium),
17-
- 2. ECG morphological analysis using designer
wavelets - Biosens team J.M.H. Karel (PhD-student), dr.
R.L.M. Peeters, dr. R.L. Westra - Graduation students Pieter Jouck, Kurt
Moermans, Maarten Vaessen
18Wavelet based signal analysis
19Signal morphology and optimal wavelet design (1)
- Design wavelet such that they detect morphology
in signal - Wavelet transform can be seen as convolution
- Maximum values if wavelet resembles signal in an
L2 sense - Fit a wavelet to signal or create optimal wavelet
in wavelet domain
20Signal morphology and optimal wavelet design (2)
- Can be used to detect epochs
- Detecting morphologies related to pathologies
- Computational efficient
21Example of optimized wavelet for R-peak detection
22Resulting transform
23- 2.1 Application of the wavelet
- Transform Modulus Maxima
- method to T-wave detection
- in cardiac signals
- J.M.H. Karel, P. Jouck, R.L. Westra
24T-wave detection in electrocardiograms
- Pilot study based on state-of-the-art approach
(e.g. Li 1995, Butelli 2002) - WTMM-based algorithm
- Approximation of singular value (Lipschitz
coefficient) did not show to be particularly
discriminating - Approach successful on a variety of RT-wave
morphologies - Classification strategy rather ad hoc
25Epoch detection and segmentation
26Example Application of the Wavelet Transform
Modulus Maxima method to T-wave detection in
cardiac signals
27Objectives
- ECG segmentation epoch detection
- Characterization of epoch
28Testcase T-wave detection
- Complications with T-wave detection
- low amplitude
- Wide variety of T-waves types
- fuzzy positioning
29Conventional Methodes
- 1st step Filter ? filtering of fluctuations and
artifacts - Different types of filters
- Differential filters
- Digital filters
- 2d step Signal comparsion using threshold
30Conventional Methods
- Advantages
- Simple and straightforward methodology
- Ease of implementation
- Disadvantages
- Sensitive for stochastic fluctuations
- Bad detection of complexes with low amplitudes
31Wavelet Transform
- Wavelet transform of signal f using wavelet ?
- Frequency and time domain
- Spectral analysis by scaling with a (dilation)
- Temporal analysis by translation with b
32Example WTMM
33WTMM-based QRS detection (1)
- Dyadic transform scales 21 22 23 24
34WTMM-based QRS detection (2)
- Identification of Modulus Maxima
- QRS-complex ? 2 modulus maxima (MM)
- Find all MM on all scales
- Delete redundant MMs
- 2 positive or 2 negative recurring MMs
- Proximate MM multiples (too close for comfort)
35WTMM-based QRS detection (3)
- Positioning of R-peak
- Zero-crossing between positive and negative MMs
36Adjustments for T-wave detection
37Adjustments for T-wave detection
- Search for Modulus Maxima
- Only MMs above a given threshold
- Position of T-wave peak
- Normal T-wave ? MM pare ? zero-crossing
- T-wave with single increase/decrease ? one MM
near peak
38Results
- Sensitivity of WTMM-based methode to
- Low T-wave amplitudes
- Noise and stochastic variation
- Baseline-drift
- complex T-wave morphology
39Results
40Results
- Testcases Signals from i) MIT-BIH database, and
ii) Cardiology department of Maastricht
University. - Performance
- Number of True Positive detections (TruePos)
- Number of False Positive detections (FalsePos)
- Number of True Negative detections (TrueNeg)
- Number of False Negative detections (FalseNeg)
- Total number of peaks (TotalPeak)
- Percentage of detected T-waves (Sensitivity)
- Percentage of correct detections (PercCorr)
41Testcase 1
42Testcase 2
43Testcase 3
44Testcase 4
45Discussion
- Problems
- Low amplitude high noise levels
- Extremely short ST-intervals
- Not all types of T-wave are detected
- Improvements
- Automatic scale adjuster
- More decision rules
- Learning algorithm
46Conclusion
- Reliable method
- Robust and noise-resistent
- Good performance in sense of sensitivity and
percentage correct (typically gt 85)
47- 2.2 Optimal discrete wavelet
- design for cardiac signal
- processing
- J.M.H. Karel, R.L.M. Peeters and R.L. Westra
EMBC 2005 ( 27th Annual International Conference
of the IEEE Engineering in Medicine and Biology
Society), 1-4 September 2005, Shanghai,
Peoples Republic of China
48Optimal Wavelet Design
- What is a good wavelet for a given signal and a
given purpose? - Freedom in choice for analyzing wavelet ?(t)
- Best output ( wavelet coefficients ck) are well
positioned in frequency-temporal space, i.e.
sparse representation - Essentials perfect reconstruction, orthogonal
wavelet-multi resolution structure, vanishing
moments of wavelets, flatness of filter, smooth
wavelets - Measure the performance of a given signal x(t)
and a trial wavelet ?(t) with a criterion
function V?
49Optimal Wavelet Design
- Orthogonal wavelets and filter banks
50Wavelet analysis and synthesis
- Low pass filter with transfer function C(z)
- High pass filter with transfer function D(z)
- Combination with down-sampling
- ? has compact support ? C, D are FIR
51Wavelets analysis and synthesis
52Filter bank
53Filter bank
54Polyphase decomposition
55Polyphase decomposition
the alternating flip construction relates
coefficients c and d
the remaining orthogal freedom in the 2n ck can
be expressed by a reparametrization in n new
parameters ?i the polyphase matrices R and ?
56Polyphase decomposition
The polyphase matrices R and ? in terms of
parameters ?i are defined as
Then define the 2x2 matrix H as
57Polyphase decomposition
Matrix H can be partitioned as
with
58Polyphase decomposition
Now the matrices C and D can be written as
59Design criteria on the wavelet put constraints on
C and D
- The polyphase decomposition handles the
orthogonality of the filterbank - Another desirable property are the vanishing
moments - This puts constraints on C and D, e.g. C(z) has
zeros at z 1 0, ditto corresponding flatness
of C and D at high/low frequencies - The condition of vanishing moments translates to
a linear set of constraints on the filter
coefficients c and d
60Design criteria on the wavelet put constraints on
C and D
- For instance one vanishing moment states
- this translates to a constraint on c and d
- And so also to a constraint on the ?s
61Computing the wavelet and scaling function
- The scaling function ? relates to the wavelet ?
via the dilation equation. - Using the alternating flip construction this
states
62Computing the wavelet and scaling function
- The resulting functions ? and ? may be
discontinuous and fractal - The dilation relation allows an iteration scheme
for the coefficients
63Tree structure for dyadic scales
64Criteria to measure the quality of a wavelet for
given data
- The energy of the original signal x(t) is
- Because of the orthogonality this energy is
preserved in the wavelet representation
- Therefore, the L2-norm is not a suitable
criterion
65Wavelet design criteria
- Philosophy and algorithm
- Guiding principle proposed here is to aim for
maximization of the variance. - This is achieved by either
- maximization of the variance of the absolute
values of the wavelet coefficients ? minimization
of L1-norm - maximization of the variance of the squared
wavelet coefficients ? maximization of L4-norm
66Algorithm
67Wavelet Design Criteria
- Remarks
- The criteria min L1 and max L4 have been
investigated to design wavelets for various given
signals. - When all wavelet coefficients are taken into
account and no weighting is applied, both
criteria tend to produce similar results. - However, when only a few scales are taken into
account (e.g. by weighting) the results may
become different in case of minimization of the
L1-norm energy tends to be placed in scales not
taken into account, whereas in case of
maximization of the L4-norm this does not happen.
68Experimentation
- I. Reconstruction of artificially generated noisy
signals - Make an artificial sparse signal x(t) by setting
only a few wavelet-coefficients ck to non-zero
values - Note that this signal probably has a small
L1-norm (sparse) and a large L4-norm (large
variance) - Add some white noise v(t) and apply the
wavelet-design algorithm to this signal x(t)
v(t) - The reconstructed signal x(t) fits perfectly
with x(t) up to a signal-to-noise ratio (snr) 1
69Experimentation
- II. Reconstruction of a reference signal
- Reference signal x(t) is obtained by averaging
comparable episodes from ECG signals from the MIT
Physionet normal sinus rhythm database - Resulting smooth signal is upsampled
70Input signal averaged ECG
71Experimentation
- Local optima in the ?-parameter space
- Consider the situation with n3, i.e. three ?s
- Because ?1 ?2 ?3 ?/4 this situation has two
degrees of freedom - Now we can plot the L1-criterion in the (?2,
?3)-plane and study local optima
72Local Optima
73Experimentation
- Local optima in the ?-parameter space
- The coefficients of the local optima closely
resemble the Daubechies 2 filter coefficients - This observation gives a rationale for the use of
the Daubechies 2 wavelet for cardiac signals - When the sum of ?2 and ?3 already is close to ?/4
only one degree of freedom is effectively used - some of the minor local optima resemble the
Daubechies 3 wavelet, however to lesser extent
74Experimentation
- The optimal number ?-parameters
- The filter size of the wavelet filter is
determined by the number n of ?s used - A large number of ?s gives freedom to fit the
wavelet to the signal but also increases the
complexity - For n 1 to 25 the L1-criterion averaged over
1000 random starting points is computed - The graph is rather flat between n 5 and n
20, and n 8 is a reasonable choice
75Criterion versus number of ?s
n 8
76Experimentation
- Practical evaluation
- Next we design the best fitting wavelet for the
test set episode 103 of the MIT-BIH arrythmia
database. This is a 360 Hz annotated ECG signal. - Two wavelets were designed using 8 ?s by
minimizing a criterion function V, with - for ?1 V L1-norm of the wavelet transform
- for ?2 V L4-norm of the wavelet transform
77L1-L4 wavelet
78Experimentation
- Quality of the wavelet transform of the
reference averaged ECG-signal with the
L4-criterion maximized wavelet - The L4-wavelet has a fractal structure. The
available degrees of freedom are not used to
place poles at z -1 as with the Daubechies
wavelets - The fractal nature of the L4-wavelet is not
relevant in processing, as only the coefficients
are used in computations - Moreover, the L4-wavelet is very effective in the
wavelet decomposition of the reference signal.
One single strong wavelet coefficient marks the
location of largest correlation
79Wavelet transform of reference ECG signal with
the designed L4-wavelet
80Experimentation
- Comparison of the L1- and L4-wavelets with the
Daubechies 2 wavelet - The wavelet transform of the MIT-testset with was
compared for the three wavelets (L1, L4, and
Daubechies), on only a single level (scale) - This level was selected for each wavelet
individually to maximize performance - A binary vector was constructed of all the
wavelet coefficients of which the absolute value
exceeded a certain threshold. These locations
were related to the locations of the original
signal - The beat annotations in the testset were used as
a reference to locate the QRS-complex.
81Experimentation
- Comparison of the L1- and L4-wavelets with the
Daubechies 2 wavelet - There are 1688 normal QRScomplexes in the
MIT-testset. If the binary vector corresponding
to the wavelet transform has detected a peak
within 20 samples (56ms) of the marker, it is
assumed that the QRS-complex is detected. If a
peak is detected but no marker is within 20
samples, a false detection is assumed - This comparison yielded the following results
82Results
Comparison of the L1- and L4-wavelets with the
Daubechies 2 wavelet
83Conclusions
- Comparison of the L1- and L4-wavelets with the
Daubechies 2 wavelet - The table shows that the performance of the
Daubechies 2 wavelet is quite good - The L4 optimized wavelet however shows superior
performance - Furthermore the L4-wavelet is more robust with
respect to the choice of threshold value, which
may be a large advantage in practical
applications
84References
1 Gilbert Strang,Truong Nguyen, Wavelets and
Filter Banks, Wellesley-Cambridge Press,
1996. 2 N. Neretti, N. Intrator, An adaptive
approach to wavelet filter design, Proc. IEEE
int. workshop on neural networks for signal
processing, 2002. 3 A.L. Goldberger et al..
Physiobank, physiotoolkit, and physionet Complex
physiologic signals. Circulation,
101(23)e215e220, June 2000. 4 Cuiwei Li et
al., Detection of ECG characteristic points using
wavelet transforms. IEEE Transactions on
Biomedical Engineering, 42(1)2128, January
1995. 5 Stephane Mallat. A Wavelet Tour of
Signal Processing. Ac. Press, 1999. 6 Ivo
Provaznýk, Ji Kozumplýk. Wavelet transform in
electrocardiography - data compression.
International Journal of Medical Informatics,
45(1-2)111128, June 1997. 7 M.P. Wachowiak et
al., Waveletbased noise removal for biomechanical
signals a comparative study. IEEE Transactions
on Biomedical Engineering, 47(3)360368, March
2000.
85Focus for further research
- Analysis of mathematical morphology of cardiac
signals - Characterization of cardiac signals (clustering)
- Expert input
- Design of optimal multiwavelets
- Development of online algorithm based on optimal
wavelets and hardware implementation
86Discussion
Dr. Ronald L. Westra Department of
Mathematics Maastricht University westra_at_math.unim
aas.nl
Biomedical Signal Processing Forum, August 30th
2005