Electrophysiological signals such as the EEG, MEG, or LFPs have been extensively studied over the last decades, and elaborate signal\r\nprocessing algorithms have been developed for their analysis. Many of these methods are based on time-frequency decomposition\r\nto account for the signals� spectral properties while maintaining their temporal dynamics. However, the data typically exhibit\r\nintra- and interindividual variability. Existing algorithms o?en do not take into account this variability, for instance by using\r\n??xed frequency bands. ?is shortcoming has inspired us to develop a new robust and ??exible method for time-frequency analysis\r\nand signal feature extraction using the novel smooth natural Gaussian extension (snaGe) model. ?e model is nonlinear, and its\r\nparameters are interpretable. We propose an algorithm to derive initial parameters based on dynamic programming for nonlinear\r\n??tting and describe an iterative re??nement scheme to robustly ??t high-order models. We further present distance functions to be\r\nable to compare different instances of our model. ?e method�s functionality and robustness are demonstrated using simulated as\r\nwell as real data. ?e snaGe model is a general tool allowing for a wide range of applications in biomedical data analysis.
Loading....