Comparison of Artificial Neural Networks for Low-Power ECG-Classification System
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Electrocardiogram (ECG) classification has become an essential task of modern day wearable devices, and can be used to detect cardiovascular diseases. State-of-the-art Artificial Intelligence (AI)-based ECG classifiers have been designed using various artificial neural networks (ANNs). Despite their high accuracy, ANNs require significant computational resources and power. Herein, three different ANNs have been compared: multilayer perceptron (MLP), convolutional neural network (CNN), and spiking neural network (SNN) only for the ECG classification. The ANN model has been developed in Python and Theano, trained on a central processing unit (CPU) platform, and deployed on a PYNQ-Z2 FPGA board to validate the model using a Jupyter notebook. Meanwhile, the hardware accelerator is designed with Overlay, which is a hardware library on PYNQ. For classification, the MIT-BIH dataset obtained from the Physionet library is used. The resulting ANN system can accurately classify four ECG types: normal, atrial premature contraction, left bundle branch block, and premature ventricular contraction. The performance of the ECG classifier models is evaluated based on accuracy and power. Among the three AI algorithms, the SNN requires the lowest power consumption of 0.226 W on-chip, followed by MLP (1.677W), and CNN (2.266 W). However, the highest accuracy is achieved by the CNN (95%), followed by MLP (76%) and SNN (90%).
Abstract
Electrocardiogram (ECG) classification has become an essential task of modern day wearable devices, and can be used to detect cardiovascular diseases. State-of-the-art Artificial Intelligence (AI)-based ECG classifiers have been designed using various artificial neural networks (ANNs). Despite their high accuracy, ANNs require significant computational resources and power. Herein, three different ANNs have been compared: multilayer perceptron (MLP), convolutional neural network (CNN), and spiking neural network (SNN) only for the ECG classification. The ANN model has been developed in Python and Theano, trained on a central processing unit (CPU) platform, and deployed on a PYNQ-Z2 FPGA board to validate the model using a Jupyter notebook. Meanwhile, the hardware accelerator is designed with Overlay, which is a hardware library on PYNQ. For classification, the MIT-BIH dataset obtained from the Physionet library is used. The resulting ANN system can accurately classify four ECG types: normal, atrial premature contraction, left bundle branch block, and premature ventricular contraction. The performance of the ECG classifier models is evaluated based on accuracy and power. Among the three AI algorithms, the SNN requires the lowest power consumption of 0.226 W on-chip, followed by MLP (1.677W), and CNN (2.266 W). However, the highest accuracy is achieved by the CNN (95%), followed by MLP (76%) and SNN (90%).
참고문헌 (19)
- R. Lozano, "Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010 : a systematic analysis for the global burden of disease study 2010", The Lancet, 380(9859), pp.2095~2128 (2013)
- M. Mitra, "Cardiac Arrhythmia Classification Using Neural Networks with Selected Features", first international conference on Computational Intelligence:Modeling Techniques and Applications (CIMTA), pp.76~84 (2013)
- T. Habte, T, "Ultra Low Power ECG Processing System for IoT Devices: Analog Circuits and Signal Processing", pp.1~14 (2019)
- T. E. Taha, "A survey on Classification of ECG Signal Study", 6(5), pp.11~16 (2016)
- C. V. Maggio, "Quantification of Ventricular Repolarization Dispersion Using Digital Processing of the Surface ECG" (2012)
- M. Aqil, "ECG signal denoising by Discrete Wavelet Transform", 13(09), pp.51~68 (2017)
- Bultheel Bull. Belg, "Chapter 4. Data Processing and Analysis, Math Wavelet Transform", pp.35~42 (1995), IEEE Pub.
- M. Sanaullah, "A review of higher order statistics and spectra in communication systems", 13(4), pp.31~50 (2013)
- E. D. Ubeyli, "Combining recurrent neural networks with eigenvector methods for classification of ECG beats", Digital Signal Processing, 19(2), pp.320~329 (2009)
- J. Park, "Arrhythmia detection from heartbeat using the k-nearest neighbor classifier", IEEE International Conference on Bioinformatics and Biomedicine, pp.15~22 (2013)
- S. Kiranyaz, "Real-time patient-specific ECG classification by 1-D convolutional neural networks", IEEE Transactions on Biomedical Engineering, 63(3), pp.664~675 (2016)
- T. J. Jun, "ECG arrhythmia classification using a 2-D convolutional neural network", ~ (2018)
- S. Savalia, "Cardiac Classification by Multi-Layer Perceptron and Convolutional Neural Networks", 5(35), pp.1~12 (2018)
- A. Das, "Unsupervised heart-rate estimation in wearable with liquid states and a probabilistic readout", Neural Networks, 99, pp.134~147 (2018)
- A. Amirshahi, "Ultra Low-Power and Real-time ECG classification Based on STDP and R-STDP Neural networks for Wearable Devices", IEEE Transactions on Biomedical Circuits and Systems (TBioCAS), 13(6), pp.1483~1493 (2019)
- K. Guo, "A survey of FPGA Based Neural Network Accelerator", ACM Transactions on Reconfigurable Technology and Systems, 9(4), pp.11~ (2017)
- C. V. Nguyen, "A non-contact infection screening system using medical radar and Linux-embedded FPGA : Implementation and preliminary validation", Informatics in Medicine Unlocked, 16, pp.1~9 (2019)