Adaptive Filtering and Compression of Bio-Medical Signals using Neural Netwo
Kalyan Chatterjee1, Mandavi2, Prasannjit3, Nilotpal Mrinal4, S.Dasgupta5
1Kalyan Chatterjee, Computer Science Engineering, Bengal college of Engineering & Technology, Durgapur, India.
2Mandavi, information Technology, Bengal College of Engineering & Technology, Durgapur, India.
3Prasannjit, Information Technology, Bengal College of Engineering & Technology, Durgapur, India.
4Nilotpal Mrinal, Information Technology, Bengal College of Engineering & Technology, Durgapur, India.
5S.Dasgupta, Computer Science Engineering, Bengal College of Engineering & Technology, Durgapur, India.
Manuscript received on January 30, 2013. | Revised Manuscript received on February 16, 2013. | Manuscript published on February 28, 2013. | PP: 323-327 | Volume-2 Issue-3, February 2013. | Retrieval Number: C1120022313/2013©BEIESP
Open Access | Ethics and Policies | Cite
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Biomedical signals are often contaminated by noise. Thus, noise removal and subsequently their lossless compression is also very necessary. This paper presents an adaptive filtering technique for removing noise from ECG signal using the Recursive Least Square (RLC) method. Twelve significant features are extracted from an echocardiogram (ECG) dataset. After carrying out noise cancellation followed by Recursive Least Square method filtered, ECG signal is obtained. Moreover we have also compressed the ECG signals. The filtered signals are used as input to the artificial neural network. Finally these samples which are used in the database are trained and tested using the Back Propagation Algorithm. The compression ratio is observed to be 0.9745583.It is further observed that input signals are same as the supervised signals used in the network. This paper presents experimental results which demonstrates the usefulness of adaptive filtering and data compression in several bio-medical applications.
Keywords: Adaptive filtering, Data compression, back propagation, Recursive least square method.