AN INTEGRATED FRAMEWORK FOR PARALLEL TRAINING AND CLASSIFICATION OF ECG SIGNAL

Apply

AN INTEGRATED FRAMEWORK FOR PARALLEL TRAINING AND CLASSIFICATION OF ECG SIGNAL

Journal Title: Elysium Journal of Engineering Research and Management - Year 2015, Vol 2, Issue 5

Abstract

Recently, various techniques are developed to analyze and classify the Electrocardiograph (ECG) signal to know the threaten results. Many real time development tools such as VLSI, DSP, etc., are available to carry out this analysis process. During the implementation process, the problems addressed are circuit complexity and time computation due to the separate operation of training and classification. In existing works, the training of features of an ECG signal performed and stored in the database, the arrival of new data perform the comparison process to show the status of signal. But, in proposed work, parallel computation of training and classification optimize the circuit complexity and time consumption. Moreover, the integration of MATLAB with XILINX platform assures the suitability of proposed system in real time medical diagnosis applications. In this paper, a Discrete Fourier Transform (DFT) is used to extract the features of an ECG signal. Then, an approximate multiplier architecture in XILINX modifies the exponent processing element architecture with Gaussian Kernel (GK) function in QRS detection strategy optimizes the complexity level. Finally, the Support Vector Machine (SVM) classifier is used to classify the extracted features with the maximum accuracy. The proposed algorithms reduces the components used leads to power and time reduction. The comparative analysis between proposed integrated framework and existing methods confirms the effectiveness.

Authors and Affiliations

Ramya M, Karthikeyan P.

Keywords

Butter-Worth Filter (BW) CORDIC Discrete Fourier Transform (DFT) Electrocardiograph (ECG) Gaussian Kernel (GK) Support Vector Machine (SVM).

EP ID EP400055
DOI -
Views 129
Downloads 0