Decomposition of Biomedical Signals in Spatial and Time-frequency Modes

Summary

Objectives: The purpose of this paper is to introduce a new method for spatial-time-frequency analysis of multichannel biomedical data. We exemplify the method for data recorded with a 31-channel Philips biomagnetometer.

Methods: The method creates approximations and decompositions of spatiotemporal signal distributions using elements (atoms) chosen from a very large and redundant set (dictionary). The method is based on the Matching Pursuit algorithm, but it uses atoms that are distributed both in time and space (instead of only time-distributed atoms in standard Matching Pursuit). The time-frequency distribution of signal components is described by Gabor atoms and their spatial distribution is modeled by spatial modes. The spatial modes are created with the help of Bessel functions. Two versions of the method, differing in the definition of spatial properties of the atoms, are presented.

Results: The technique was validated on simulated data and real magnetic field data. It was used for removal of powerline noise from multichannel magnetoencephalography data, extraction of high-frequency somatosensory evoked fields and for separation of partially overlapping T- and U-waves in magnetocardiography.

Conclusions: The method allows for parameterization of multichannel data in the time-frequency as well as in the spatial domains. It thus can be used for signal preserving filtering simultaneously in time, frequency, and space. Applications are e.g. the elimination of artifact components, extraction of components with biological meaning, and data exploration.

Keywords

Magnetoencephalography - signal analysis - Matching Pursuit - filtering

• References

• 1 Della Penna S, Cianflone F, Del Gratta C, Erné SN, Granata C, Pizzella V, Russo M, Romani GL. A 500 channel MEG system, Proceedings of the 14th International Conference on Biomagnetism. August 8-12, 2004. Boston, Massachusetts, USA: pp 619-620.
• 2 Mallat S, Zhang Z. Matching pursuit with timefrequency dictionaries. IEEE Trans Signal Proc 1993; 41 (12) 3397-3415.
• 3 Blinowska KJ, Durka PJ, Zygierewicz J. Timefrequency analysis of brain electrical activity – Adaptive approximations. Methods Inf Med 2004; 43 (01) 70-73.
• 4 Gribonval R. Piecewise linear source separation. In: Proc SPIE 03. Vol. 5207. Wavelets: Applications in Signal and Image Processing. San Diego: California; 2003
• 5 Durka PJ, Matysiak A, Montes EM, Sosa PV, Blinowska KJ. Multichannel matching pursuit and EEG inverse solutions. J Neurosci Methods 2005; 15 148 (01) 49-59.
• 6 Studer D, Hoffmann U, Koenig T. From EEG dependency multichannel matching pursuit to sparse topographic EEG decomposition. Journal of Neuroscience Methods 2006; 153: 261-275.
• 7 Davis G, Mallat S, Zhang Z. Adaptive time-frequency approximations with matching pursuits. Wavelets: Theory, Algorithms and Applications. 1994 pp 271-293.
• 8 Gratkowski M, Haueisen J, Arendt-Nielsen L, CN Chen A, Zanow F. Time-Frequency Filtering of MEG Signals with Matching Pursuit. J Physiol (Paris) 2006; 99 (01) 47-57.
• 9 Durka PJ. Time-frequency analyses of EEG PhD dissertation.. Warsaw University, 1999
• 10 Dössel O, David B, Fuchs M, Krüger J, Lüdeke KM, Wischmann HA. A 31-channel SQUID sys- tem for biomagnetic imaging. Applied Superconductivity 1993; 1 Nos. 10-12 1813-1825.
• 11 Kreyszig. Advanced Engineering Mathematics. John Wiley & Sons; 1993
• 12 Nelder JA, Mead R. A simple method for function minimization. Computer Journal 1965; 7: 308-313.
• 13 Ben-Israel A, Greville Thomas NE. Generalized Inverses. Springer; 2003
• 14 Nuwer MR, Aminoff M, Desmedt J, Eisen AA, Goodin D, Matsuoka S, Maugiere F, Shibasaki H, Sutherling W, Vibert JF. IFCN recommended standards for short latency somatosensory evoked potentials. Electroencephalography and clinical Neurophysiology 1994; 91: 6-11.
• 15 Haueisen J, Schack B, Meier T, Curio G, Okada Y. Multiplicity in the high-frequency signals during the short-latency somatosensory evoked cortical acivity in humans. Clinical Neurophysiology 2001; 112: 1316-1325.
• 16 Curio G. Linking 600-Hz “spikelike” EEC/MEG wavelets (“sigma-bursts”) to cellular substrates – Concepts and caveats. Journal of Clinical Neurophysiology 2000; 17: 377-396.
• 17 Hashimoto I. High-frequency oscillations of somatosensory evoked potentials and fields. Journal of Clinical Neurophysiology 2000; 17: 309-320.
• 18 Ikeda H, Leyba L, Bartolo A, Wang YZ, Okada YC. Synchronized spikes of thalamocortical axonal terminals and cortical neurons are detectable outside the pig brain with MEG. Journal of Neurophysiology 2002; 87: 626-630.
• 19 Goernig M, Tute C, Liehr M, Lau S, Haueisen J, Figulla H-R, Leder U. Spatiotemporal correlation analyses: a new procedure for standardisation of DC magnetocardiograms. Biomed Tech (Berl) 2006; 51: 198-200.
• 20 Comani S, Mantini D, Merlino B. et al. Integrated software suite for magnetocardiographic data analysis – A proposal based on an interactive programming environment. Methods Inf Med 2005; 44 (01) 114-123.
• 21 Sahu P, Lim PO, Rana BS, Struthers AD. QT dispersion in medicine: electrophysiological Holy Grail or fool’s gold?. Q J Med 2000; 93: 425-431.
• 22 Christov I, Simova I. Q-onset and T-end delineation: assessment of the performance of an automated method with the use of a reference database. Physiol Meas 2007; 28: 213-221.
• 23 Hilgenfeld B, Haueisen J. Simultaneous suppression of disturbing fields and localization of magnetic markers by means of multipole expansion. Biomagnetic Research and Technology 2004; 2: 6.