Local adaptive de-noising techniques in transform domain for EMCG de-noising

There are various de-noising algorithms and optimization methods for different signal and noise characteristics. However, the signals used in real application may deviate from the model. For example: signal and/or noise may not be stationary or a proper model for them may nor be available. MCG (magnetocardiography) is an example signal, where conventional de-noising methods are not giving satisfactory results. Local adaptive processing allows to modify filtering parameters according to the specific properties of different "portions" of a signal. In this paper a methodology for adopting the transform domain local adaptive processing to the specific task of MCG signal de-noising is introduced.