A New Automatic Method of Parkinson Disease Identification Using Complex-Valued Neural Network
Rashidah Funke Olanrewaju 1, Nur Syarafina Zahari 1,
and
Aibinu Abiodun Musa 2
1. Department of Electrical and Computer Engineering, Faculty of Engineering, International Islamic University Malaysia, 50728 Kuala Lumpur
2. Federal University of Technolog, Minna, Nigeria
2. Federal University of Technolog, Minna, Nigeria
Abstract—A new automatic method of parkinson detection and classification using Complex Valued Neural Network (CVNN) is proposed in this paper. The proposed methodology used one of recently introduced dysphonia measure as part of its input data. The selected measures are those that are robust to many uncontrollable variations in individual and environments. The three selected dysphonia measures are converted from time domain to frequency domain by application of Discrete Fourier Transform (DFT) on the data. The frequency domain converted measures are fed to CVNN and the output of CVNN serves as input to the parkinson disease classifier for classification purpose. Result obtained by application of this technique on parkinson data resulted in classification performance of 96% accuracy.
Index Terms—complex backpropagation algorithm, Complex-Valued Data (CVD), Complex Valued Neural Network (CVNN), Fast Fourier Transform (FFT), parkinson disease (PD)
Cite: Rashidah Funke Olanrewaju, Nur Syarafina Zahari, and Aibinu Abiodun Musa, "A New Automatic Method of Parkinson Disease Identification Using Complex-Valued Neural Network," Journal of Medical and Bioengineering, Vol. 6, No. 1, pp. 25-28, June 2017. doi: 10.18178/jomb.6.1.25-28
Cite: Rashidah Funke Olanrewaju, Nur Syarafina Zahari, and Aibinu Abiodun Musa, "A New Automatic Method of Parkinson Disease Identification Using Complex-Valued Neural Network," Journal of Medical and Bioengineering, Vol. 6, No. 1, pp. 25-28, June 2017. doi: 10.18178/jomb.6.1.25-28
Array