Synthesis of Embedded SystemC Design: A Case Study of Digital Neural Networks
Proceedings of the conference on Design, automation and test in Europe - Volume 3
Automatic Classification of Heartbeats Using Wavelet Neural Network
Journal of Medical Systems
| Hi-index | 0.00 |
At the present scenario, one of the main causes ofdeath in developed and in emerging countries is thecardiovascular related diseases. Most of these deathscould be avoided if there was a pre-monitoring and a pre-diagnosticof these cardiac arrhythmia and myocardialisquemy by using an electrocardiogram (ECG) tool.In this scenario, this work proposes a system to helpthe doctor to detect cardiac arrhythmia. As reference, ituses the Normal, Fusion and PVC signals of the MITdatabase. Then, we extract the principal characteristics ofthe signal by means of the Principal Component Analysis(PCA) technique. One key-point in this work is the inputsignals extraction, which are captured in the sameamount. So, the number of segments for each signal is thesame. After signal preprocessing, they are applied to anArtificial Neural Network Multilayer Perceptron (ANNMLP). The MLP with 5 neurons was verified to have thebest accuracy. Based on this idea (the use of the sameinformation amount for all input signal types), weachieved better results in comparison with other works inthe field. This consideration is very important due to thefact that the ANN could be more sensible to the signal typewith major predominance.Keywords: Electrocardiogram (ECG); Artificial NeuralNetwork (ANN); Pattern Recognition; PrincipalComponent Analysis (PCA).