CLASSIFYING MYOPATHY AND NEUROPATHY NEUROMUSCULAR DISEASES USING ARTIFICIAL NEURAL NETWORKS

Kocer S.

INTERNATIONAL JOURNAL OF PATTERN RECOGNITION AND ARTIFICIAL INTELLIGENCE, cilt.24, sa.5, ss.791-807, 2010 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 24 Sayı: 5
  • Basım Tarihi: 2010
  • Doi Numarası: 10.1142/s0218001410008184
  • Dergi Adı: INTERNATIONAL JOURNAL OF PATTERN RECOGNITION AND ARTIFICIAL INTELLIGENCE
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Sayfa Sayıları: ss.791-807
  • Anahtar Kelimeler: Electromyogram (EMG), Multilayer Perceptron (MLP), Radial Basis Function (RBF), Support Vector Machine (SVM), SUPPORT VECTOR MACHINES, PATTERN-RECOGNITION, POWER SPECTRUM, EMG, CLASSIFICATION, SIGNALS, DISORDERS, ALGORITHM, DIAGNOSIS, SYSTEMS
  • Gazi Üniversitesi Adresli: Hayır

Özet

The aim of this study is to classify myopathy and neuropathy neuromuscular diseases using artificial neural networks. Coefficients were obtained from these EMG signals by applying Fast Fourier Transform (FFT), Autoregressive (AR), and Cepstral spectral analysis methods. Each of these coefficients was used as input data for Multilayer Perceptron (MLP), Radial Basis Function (RBF), and Support Vector Machine (SVM). After these inputs were individually trained in MLP, RBF and SVM classification systems, their classification and test performances were examined. The results revealed that the highest prediction was in SVM classification system, whereas the best analysis method was found to be FFT. The results show that the combination of FFT with SVM topology has provided the area under the ROC curve of 0.953, which is considered within the acceptable range.