scispace - formally typeset
Search or ask a question
Book ChapterDOI

Arrhythmia classification using local hölder exponents and support vector machine

TL;DR: A novel hybrid Holder-SVM detection algorithm for arrhythmia classification that classifies 160 of Normal sinus rhythm, 25 of Ventricular bigeminy, 155 of Atrial fibrillation and 146 of Nodal (A-V junctional) rhythm with 96.94% accuracy is proposed.
Abstract: We propose a novel hybrid Holder-SVM detection algorithm for arrhythmia classification. The Holder exponents are computed efficiently using the wavelet transform modulus maxima (WTMM) method. The hybrid system performance is evaluated using the benchmark MIT-BIH arrhythmia database. The implemented model classifies 160 of Normal sinus rhythm, 25 of Ventricular bigeminy, 155 of Atrial fibrillation and 146 of Nodal (A-V junctional) rhythm with 96.94% accuracy. The distinct scaling properties of different types of heart rhythms may be of clinical importance.

Content maybe subject to copyright    Report

Citations
More filters
Journal ArticleDOI
TL;DR: This review describes the main features of predictive clinical data mining and focuses on two specific aspects of particular interest: the methods able to deal with temporal data and the efforts performed to translate molecular medicine results into clinically useful data mining models.
Abstract: Predictive data mining in clinical medicine deals with learning models to predict patients' health. The models can be devoted to support clinicians in diagnostic, therapeutic, or monitoring tasks. Data mining methods are usually applied in clinical contexts to analyze retrospective data, thus giving healthcare professionals the opportunity to exploit large amounts of data routinely collected during their day-by-day activity. Moreover, clinicians can nowadays take advantage of data mining techniques to deal with the huge amount of research results obtained by molecular medicine, such as genetic or genomic signatures, which may allow transition from population-based to personalized medicine. The current challenge is to exploit data mining to build models able to take into account the dynamic and temporal nature of clinical care and to exploit the variety of information available at the bedside. This review describes the main features of predictive clinical data mining and focus on two specific aspects of particular interest: the methods able to deal with temporal data and the efforts performed to translate molecular medicine results into clinically useful data mining models. © 2011 John Wiley & Sons, Inc. WIREs Data Mining Knowl Discov 2011 1 416-430 DOI: 10.1002/widm.23

93 citations

Journal ArticleDOI
TL;DR: In this paper, sufficient and necessary conditions for L 2(ℝ2) functions with uniform and pointwise Holder exponent α > 0 were derived for the special case with 1-dimensional singularity line.
Abstract: Using Hart Smith’s and curvelet transforms, new necessary and new sufficient conditions for an L 2(ℝ2) function to possess Holder regularity, uniform and pointwise, with exponent α>0 are given. Similar to the characterization of Holder regularity by the continuous wavelet transform, the conditions here are in terms of bounds of the transforms across fine scales. However, due to the parabolic scaling, the sufficient and necessary conditions differ in both the uniform and pointwise cases. We also investigate square-integrable functions with sufficiently smooth background. Specifically, sufficient and necessary conditions, which include the special case with 1-dimensional singularity line, are derived for pointwise Holder exponent. Inside their “cones” of influence, these conditions are practically the same, giving near-characterization of direction of singularity.

19 citations


Cites methods from "Arrhythmia classification using loc..."

  • ...This regularity information can be used for example in denoising, classification, or re-construction of signals [ 16 , 19, 21]....

    [...]

Proceedings ArticleDOI
01 Nov 2009
TL;DR: This paper proposes a Holder-SVM detection algorithm using a novel hybrid arrangement of binary and multiclass SVMs designed to take care of class imbalance rampant in biomedical signals, and significantly reduces the number of false negatives.
Abstract: Automatically classifying ECG recordings for Malignant Ventricular Arrhythmia is fraught with several difficulties. Even normal ECG signals exhibit only quasi-periodic nature, and contain various irregularities. The key to more accurate detection is the use of position, and amount of local singularities in the signals.In this paper, we propose a Holder-SVM detection algorithm using a novel hybrid arrangement of binary and multiclass SVMs designed to take care of class imbalance rampant in biomedical signals. As a result, we significantly reduce the number of false negatives – patients falsely classified as normal. We used the MIT-BIH Arrhythmia database for even different arrhythmias. We compare our hybrid SVM with a suitable conventional SVM, and show better results.We also use the new arrangement for features proposed earlier, and demonstrate the gain in accuracy. Our concept of hybrid SVM is applicable to a wide variety of multiclass classification problems.

17 citations


Cites methods from "Arrhythmia classification using loc..."

  • ...Testing Predicted classification Data Classes N P AFIB NOD VF BI B 369 N 355(356)355 5(5)5 4(3)4 1(1)1 1(1)1 1(1)1 2(2)2 26 P 15(13)13 10(12)12 1(1)1 0(0)0 0(0)0 0(0)0 0(0)0 101 AFIF 7(4)3 0(1)1 90(92)93 1(1)1 0(0)0 3(3)3 0(0)0 33 NOD 0(0)0 0(0)0 2(2)2 24(24)24 5(5)5 2(2)2 0(0)0 31 VF 0(4)0 0(0)0 0(0)0 5(2)3 22(21)24 4(4)4 0(0)0 81 BI 1(1)0 0(0)0 1(1)1 11(11)11 2(2)2 64(64)65 2(2)2 11 B 0(0)0 0(0)0 0(0)0 2(2)2 0(0)0 4(4)4 5(5)5 Sens 93....

    [...]

  • ...The extraction of features from each rhythm was done as explained in (3)....

    [...]

  • ...369 N 354(354) 1(1) 2(2) 5(5) 2(2) 4(4) 1(1) 26 P 6(5) 20(21) 0(0) 0(0) 0(0) 0(0) 0(0) 101 AFIF 3(2) 0(0) 98(98) 0(0) 0(1) 0(0) 0(0) 33 NOD 4(3) 0(1) 0(0) 29(29) 0(0) 0(0) 0(0) 31 VF 6(5) 0(0) 0(0) 0(0) 25(26) 0(0) 0(0) 81 BI 1(1) 0(0) 0(0) 0(0) 0(0) 80(80) 0(0)...

    [...]

  • ...1In (3), the term “hybrid” was used to indicate the dual concepts of Hölder and SVM whereas this paper describes the larger notion of making any multiclass SVM truly hybrid as described in § II....

    [...]

  • ...Prior work in computation of LHE for different applications such as (3), (2), (1) have used only the lowest scale (which is scale 1)....

    [...]

Journal ArticleDOI
TL;DR: In this paper, the mixed Holder spectra of finitely many functions can be recovered from a Legendre transform of a concave scaling function based on simultaneous wavelet leaders, and the typical optimality of this upper bound in a product of continuous Besov or oscillation spaces is studied.
Abstract: In this paper, we show that the mixed Holder spectra of finitely many functions can be recovered from a Legendre transform of a concave scaling function based on simultaneous wavelet leaders. We first prove that this Legendre transform yields an upper bound valid for uniform Holder functions. We then study the typical optimality (in the sense of Baire’s category) of this upper bound in a product of continuous Besov or oscillation spaces.

10 citations

Iyad Batal1
01 Jan 2015

6 citations


Additional excerpts

  • ...In [22], Joshi et al....

    [...]

References
More filters
Journal ArticleDOI
TL;DR: Issues such as solving SVM optimization problems theoretical convergence multiclass classification probability estimates and parameter selection are discussed in detail.
Abstract: LIBSVM is a library for Support Vector Machines (SVMs). We have been actively developing this package since the year 2000. The goal is to help users to easily apply SVM to their applications. LIBSVM has gained wide popularity in machine learning and many other areas. In this article, we present all implementation details of LIBSVM. Issues such as solving SVM optimization problems theoretical convergence multiclass classification probability estimates and parameter selection are discussed in detail.

40,826 citations

Book
Vladimir Vapnik1
01 Jan 1995
TL;DR: Setting of the learning problem consistency of learning processes bounds on the rate of convergence ofLearning processes controlling the generalization ability of learning process constructing learning algorithms what is important in learning theory?
Abstract: Setting of the learning problem consistency of learning processes bounds on the rate of convergence of learning processes controlling the generalization ability of learning processes constructing learning algorithms what is important in learning theory?.

40,147 citations


"Arrhythmia classification using loc..." refers methods in this paper

  • ...In this work,we have developed a novel hybrid Hölder-SVM detection algorithmfor arrhythmia classification....

    [...]

  • ...Ḧolder-SVM methodology was found to provide superior performance....

    [...]

  • ...For each of the extracted rhythms, we computed the features to be used by SVM for classification in the following manner....

    [...]

  • ...SVM is being extensively used for several classification and regression applications....

    [...]

  • ...In Section 2, we briefly describe the methodology for computation of the local singularities and provide a short introduction to SVM. Section 3 highlights our approach for classification along with results achieved....

    [...]

Journal ArticleDOI
TL;DR: Decomposition implementations for two "all-together" multiclass SVM methods are given and it is shown that for large problems methods by considering all data at once in general need fewer support vectors.
Abstract: Support vector machines (SVMs) were originally designed for binary classification. How to effectively extend it for multiclass classification is still an ongoing research issue. Several methods have been proposed where typically we construct a multiclass classifier by combining several binary classifiers. Some authors also proposed methods that consider all classes at once. As it is computationally more expensive to solve multiclass problems, comparisons of these methods using large-scale problems have not been seriously conducted. Especially for methods solving multiclass SVM in one step, a much larger optimization problem is required so up to now experiments are limited to small data sets. In this paper we give decomposition implementations for two such "all-together" methods. We then compare their performance with three methods based on binary classifications: "one-against-all," "one-against-one," and directed acyclic graph SVM (DAGSVM). Our experiments indicate that the "one-against-one" and DAG methods are more suitable for practical use than the other methods. Results also show that for large problems methods by considering all data at once in general need fewer support vectors.

6,562 citations


"Arrhythmia classification using loc..." refers methods in this paper

  • ...For multi-class classification, we used popular One-Against-One method [14]....

    [...]

Journal ArticleDOI
TL;DR: It is shown that this method provides a natural generalization of the classical box-counting techniques to fractal functions (the wavelets actually play the role of “generalized boxes”).
Abstract: The multifractal formalism originally introduced for singular measures is revisited using the wavelet transform. This new approach is based on the definition of partition functions from the wavelet transform modulus maxima. We demonstrate that the f(α) singularity spectrum can be readily determined from the scaling behavior of these partition functions. We show that this method provides a natural generalization of the classical box-counting techniques to fractal functions (the wavelets actually play the role of “generalized boxes”). We report on a systematic comparison between this alternative method and the structure function approach which is commonly used in the context of fully developed turbulence. We comment on the intrinsic limitations of the structure functions which possess fundamental drawbacks and do not provide a full characterization of the singularities of a signal in many cases. We show that our method based on the wavelet transform modulus maxima decomposition works in most situations and ...

562 citations


"Arrhythmia classification using loc..." refers background or methods in this paper

  • ...This is formalized by its relation [6] with the H ölder exponent: Ws,x0( f ) ∝ |s|h(x0), s→ 0 (2)...

    [...]

  • ...The WTMM based formalism developed by Muzy [6] as described above provides global estimates of scaling properties of time series....

    [...]

Journal ArticleDOI
TL;DR: In this paper, the wavelet transform modulus maxima is used to describe the scaling properties of singular measures of fractal objects, and it is shown that the generalized fractal dimensions D q and the f (α) singularity spectrum can be determined from the scaling behavior of these partition functions.
Abstract: The multifractal formalism originally introduced to describe statistically the scaling properties of singular measures is revisited using the wavelet transform. This new approach is based on the definition of partition functions from the wavelet transform modulus maxima. We demonstrate that very much like thermodynamic functions, the generalized fractal dimensions D q and the f ( α ) singularity spectrum can be readily determined from the scaling behavior of these partition functions. We show that this method provides a natural generalization of the classical box-counting techniques to fractal signals, the wavelets playing the role of “generalized boxes”. We illustrate our theoretical considerations on pedagogical examples, e.g., devil's staircases and fractional Brownian motions. We also report the results of some recent application of the wavelet transform modulus maxima method to fully developed turbulence data. That we emphasize the wavelet transform as a mathematical microscope that can be further used to extract microscopic informations about the scaling properties of fractal objects. In particular, we show that a dynamical system which leaves invariant such an object can be uncovered form the space-scale arrangement of its wavelet transform modulus maxima. We elaborate on a wavelet based tree matching algorithm that provides a very promising tool for solving the inverse fractal problem. This step towards a statistical mechanics of fractals is illustrated on discrete period-doubling dynamical systems where the wavelet transform is shown to reveal the renormalization operation which is essential to the understanding of the universal properties of this transition to chaos. Finally, we apply our technique to analyze the fractal hierarchy of DLA azimuthal Cantor sets defined by intersecting the inner frozen region of large mass off-lattice diffusion-limited aggregates (DLA) wit a circle. This study clearly lets out the existence of an underlying multiplicative process that is likely to account for the Fibonacci structural ordering recently discovered in the apparently disordered arborescent DLA morphology.

468 citations


"Arrhythmia classification using loc..." refers methods in this paper

  • ...The computation of singularity strength and the transformation from τ(q) to spectrum of singularities D(h) is given by the Legendre transformation[8]: h(q) = dτ(q)/dq; D[h(q)] = qh(q)− τ(q) (4)...

    [...]