Improving medical diagnostic accuracy of ultrasound Doppler signals by combining neural network models

There are a number of different quantitative models that can be used in a medical diagnostic decision support system including parametric methods (linear discriminant analysis or logistic regression), nonparametric models (k nearest neighbor or kernel density) and several neural network models. The complexity of the diagnostic task is thought to be one of the prime determinants of model selection. Unfortunately, there is no theory available to guide model selection. This paper illustrates the use of combined neural network models to guide model selection for diagnosis of ophthalmic and internal carotid arterial disorders. The ophthalmic and internal carotid arterial Doppler signals were decomposed into time-frequency representations using discrete wavelet transform and statistical features were calculated to depict their distribution. The first-level networks were implemented for the diagnosis of ophthalmic and internal carotid arterial disorders using the statistical features as inputs. To improve diagnostic accuracy, the second-level networks were trained using the outputs of the first-level networks as input data. The combined neural network models achieved accuracy rates which were higher than that of the stand-alone neural network models.

[1]  Elif Derya.  Comparison of eigenvector methods with classical and model-based methods in analysis of internal carotid arterial Doppler signals , 2003 .

[2]  Simon Haykin,et al.  Neural Networks: A Comprehensive Foundation , 1998 .

[3]  David H. Wolpert,et al.  Stacked generalization , 1992, Neural Networks.

[4]  Yoichi Hayashi,et al.  Combining neural network predictions for medical diagnosis , 2002, Comput. Biol. Medicine.

[5]  E. Ritenour.  Doppler Ultrasound: Physics, Instrumentation and Clinical Applications , 1990 .

[6]  Elif Derya Übeyli,et al.  Detection of ophthalmic artery stenosis by least-mean squares backpropagation neural network , 2003, Comput. Biol. Medicine.

[7]  Bidyut Baran Chaudhuri,et al.  Efficient training and improved performance of multilayer perceptron in pattern classification , 2000, Neurocomputing.

[8]  Elif Derya Übeyli,et al.  Determination of Behcet disease with the application of FFT and AR methods , 2002, Comput. Biol. Medicine.

[9]  Metin Akay,et al.  Wavelet applications in medicine , 1997 .

[10]  Ingrid Daubechies,et al.  The wavelet transform, time-frequency localization and signal analysis , 1990, IEEE Trans. Inf. Theory.

[11]  B. H. Blott,et al.  Review of neural network applications in medical imaging and signal processing , 1992, Medical and Biological Engineering and Computing.

[12]  Yu Zhang,et al.  Doppler ultrasound signal denoising based on wavelet frames. , 2001, IEEE transactions on ultrasonics, ferroelectrics, and frequency control.

[13]  Roberto Battiti,et al.  First- and Second-Order Methods for Learning: Between Steepest Descent and Newton's Method , 1992, Neural Computation.

[14]  Eric B. Bartlett,et al.  An information theoretic approach for combining neural network process models , 1999, Neural Networks.

[15]  James V. Hansen,et al.  Data mining of time series using stacked generalizers , 2002, Neurocomputing.

[16]  Vivian West,et al.  Model selection for a medical diagnostic decision support system: a breast cancer detection case , 2000, Artif. Intell. Medicine.

[17]  B. Yegnanarayana,et al.  Artificial Neural Networks , 2004 .

[18]  I A Basheer,et al.  Artificial neural networks: fundamentals, computing, design, and application. , 2000, Journal of microbiological methods.

[19]  Elif Derya Übeyli,et al.  Spectral broadening of ophthalmic arterial Doppler signals using STFT and wavelet transform , 2004, Comput. Biol. Medicine.

[20]  Elif Derya Übeyli,et al.  Spectral analysis of internal carotid arterial Doppler signals using FFT, AR, MA, and ARMA methods , 2004, Comput. Biol. Medicine.

[21]  Elif Derya Übeyli,et al.  Application of classical and model-based spectral methods to ophthalmic arterial Doppler signals with uveitis disease , 2003, Comput. Biol. Medicine.

[22]  Kai Liu,et al.  A novel large-memory neural network as an aid in medical diagnosis applications , 2001, IEEE Transactions on Information Technology in Biomedicine.

[23]  Volker Tresp,et al.  Combining Regularized Neural Networks , 1997, ICANN.

[24]  Vivian West,et al.  Improving diagnostic accuracy using a hierarchical neural network to model decision subtasks , 2000, Int. J. Medical Informatics.

[25]  Selwyn Piramuthu,et al.  Using Feature Construction to Improve the Performance of Neural Networks , 1998 .

[26]  Elif Derya.  CLASSIFICATION OF INTERNAL CAROTID ARTERIAL DOPPLER SIGNALS USING WAVELET-BASED NEURAL NETWORKS , 2003 .

[27]  Mohammad Bagher Menhaj,et al.  Training feedforward networks with the Marquardt algorithm , 1994, IEEE Trans. Neural Networks.

[28]  Elif Derya Übeyli,et al.  Neural network analysis of internal carotid arterial Doppler signals: predictions of stenosis and occlusion , 2003, Expert Syst. Appl..

[29]  B. Sigel,et al.  A brief history of Doppler ultrasound in the diagnosis of peripheral vascular disease. , 1998, Ultrasound in medicine & biology.

[30]  William G. Baxt,et al.  Use of an Artificial Neural Network for Data Analysis in Clinical Decision-Making: The Diagnosis of Acute Coronary Occlusion , 1990, Neural Computation.

[31]  Ian H. Witten,et al.  Issues in Stacked Generalization , 2011, J. Artif. Intell. Res..

[32]  J H Smith,et al.  The application of an artificial neural network to Doppler ultrasound waveforms for the classification of arterial disease , 1996, International journal of clinical monitoring and computing.

[33]  D. Evans.  Doppler Ultrasound: Physics Instrumentation and Clinical Applications , 1989 .

[34]  N A Gough,et al.  Neural network analysis of Doppler ultrasound blood flow signals: a pilot study. , 1997, Ultrasound in medicine & biology.