Adaptive Noise Cancellation Techniques for Impedance Cardiography Signal Analysis

• Published in: International journal of innovative technology and exploring engineering 2019-07, Vol.8 (9), p.122-130
• Main Authors: Rahman, Md Zia Ur, Mirza, ShafiShahsavar, Krishna, K. Murali
• Format: Article
• Language: English
• ISSN: 2278-3075; 2278-3075
• DOI: 10.35940/ijitee.I7531.078919

Impedance Cardiography (ICG) evaluation facilitates the volume of heart stroke in the sudden cardiac arrest. It is a noninvasive method for measurement of stroke volume, cardiac output monitoring and observing the hemodynamic parameters by changes in the body blood volume. Bloodvolume changes caused due to various physiological processes is extracted in the form of the variations in the impedance of the body segment. In the real time clinical environment during the extraction the ICG signals are influenced with several artifacts.As these artifacts are not stationary in nature, we can’t predict their characteristics. Hence,we developed several hybrid adaptive filtering mechanisms to improve the ICG signals resolution. Least mean square (LMS) algorithm is the basic enhancement technique in the adaptive filtering. However, in the non-stationery situation the LMS algorithm suffers with low rate of convergence and weight drift problems. In this paper we developed some hybrid variantsof LMS algorithm those are Leaky LMS (LLMS) for ICG signal enhancement. More over to progress the convergence rate, filtering capability and to reduce the computational complexity we also developed various sign versions of LLMS algorithms. The sign variants of LLMS algorithms are sign regressor LLMS (SRLLMS), Sign LLMS (SLLMS), and Sign Sign LLMS (SSLLMS). Severaladaptive signal enhancement units (ASEUs) are developed based on adaptive algorithms and performance is evaluated on the real ICG signal taken from MIT-BIT database. To ensure the efficiency of these algorithms, four experiments were performed to eliminate the various artifacts such as sinusoidal artifacts (SA), respiration artifacts (RA), muscle artifacts (MA) and electrode artifacts (EA). Among these techniques, the ASEU associated with SRLLMS performs better in the artifacts filtering process. The signal to noise ratio improvement (SNRI) for this algorithm is calculated as 9.3388 dBs, 5.7514 dBs, 8.4449 dBs and 8.7358 dBs respectively for SA, RA, MA and EA. Hence, the SRLLMS based ASEUs are more suitable in ICG signal filtering in real time health care sensing systems.