Adaptive neuro-fuzzy interference cancellation for ubiquitous wearable ECG sensor node

An efficient method to extract noiseless Electrocardiogram (ECG) signal which is utilized for diagnostics purpose is presented. An adaptive neuro-fuzzy filtering which is basically a nonlinear system structure presented here for the noise cancellation of biomedical signals (like ECG, PPG etc) measured by ubiquitous wearable sensor node (USN node). This paper presents non-linear adaptive filter which uses fuzzy neural network (FNN) to treat with the unknown noise and artifacts present in biomedical signals. The presented work based on ANFF (Adaptive Neuro Fuzzy Filter), where adaptation process includes neural network learning ability and fuzzy if-then rules with the optimal weight setting ability. ANFF is basically a fuzzy filtering implemented in the framework of adaptive neural networks environment. ANFF setting parameters such as the training epochs, number of MFs for each input and output, type of MFs for each input and output, learning algorithm etc. Finally simulated experimental results are presented for proper validation.