Time–Frequency Methods and Brain Rhythm Signal Processing

摘要: Physiological signal analysis could encompass any of the methods being brought to bear upon a problem in physiology. In practice, physiological is narrower field than this because signals share common features which limit applicability different methods. An important property that they typically possess determined by their historical contribution organism fitness. This gives rise identifying and classifying present signal. The primary statistical feature are nonstationary, meaning characteristics change with time, variety techniques exist analyze such data. suitability data dependent, particularly degree manner changes time frequency. well-known specialized have been developed for particular sets (e.g., Achermann Borbely, 1998). Developing data-specific methodologies presents its own difficulties terms hypothesis verifiability separation into training testing methodology). article, I focus on introducing standard well-defined relevant processing brain rhythms as well organization time–frequency activities hippocampal rhythm. general, analyses consist three stages: (1) Transformation domain; (2) Quantification properties (3) Grouping together these properties. For example, wavelet transform (time–frequency transform) coefficients (quantification) might be extracted clustered (grouping) (Quiroga et al., 2002). Filtering neurophysiological performed (an implicit step), threshold chosen quantification), remaining averaged (grouping). Below, suggest generally useful way incorporate steps minimum additional assumptions, i.e., transformation, followed characterization domain’s local activity, then mixture distribution analysis.