METHOD AND SYSTEM FOR DETERMINING A THRESHOLD FOR SPIKE DETECTION OF ELECTROPHYSIOLOGICAL SIGNALS

• US 20110071766A1
• Filed: 05/25/2009
• Published: 03/24/2011
• Est. Priority Date: 05/28/2008
• Status: Active Grant

• Current Assignee: Medtronic Bakken Research Center B.V. (Medtronic Plc)
• Original Assignee: Koninklijke Philips Electronics N.V. (Koninklijke Philips N.V.)
• Inventors: Dolan, Kevin T., Martens, Hubert Cecile Francois
• Granted Patent: US 9,202,114 B2
• Time in Patent Office: Days
• US Class Current: 702/19
• CPC Class Codes: G06F 18/00 Pattern recognition
• CPC Class Codes: G06F 2218/10 by analysing the shape of a waveform, e.g. extracting parameters relating to peaks

Abstract

A method is provided for determining a threshold (81, 82) for spike (12) detection in an electrophysiological signal (11). The method comprises a step of determining an estimated envelope (31) of the electrophysiological signal (11), a step of, based on the estimated envelope (31), determining an estimated Gaussian noise, a step of determining a distribution (51) of instantaneous amplitudes of the estimated Gaussian noise, a step of determining a mode (61) of the distribution (51) of instantaneous amplitudes, and determining the threshold (81, 82) based on the mode (61) of the distribution (51) of instantaneous amplitudes.

Claims

1. A method for determining a threshold (81, 82) for spike (12) detection in an electrophysiological signal (11), the method comprising the steps of:

(21) determining an estimated envelope (31) of the electrophysiological signal (11),

(22) based on the estimated envelope (31), determining an estimated Gaussian noise,

(24) determining a mode (61) of the distribution (51) of the estimated signal envelope (31), and

(25) determining the threshold (81, 82) based on the mode (61) of the distribution (51) of the estimated envelope (31).

2. A method for determining a threshold (81, 82) as claimed in claim 1, further comprising a step (23) of determining a distribution (51) of instantaneous amplitudes of the estimated Gaussian noise, and wherein the distribution (51) of the estimated envelope (31) is represented by the distribution (51) of the instantaneous amplitudes.

3. A method for determining a threshold (81, 82) as claimed in claim 1,

wherein the determining of the estimated envelope (31) comprises computing an analytic signal, the analytic signal comprising a real part and an imaginary part, the real part comprising the electrophysiological signal (11), the imaginary part comprising a Hilbert transform of the electrophysiological signal (11), and

wherein the distribution (51) of instantaneous amplitudes is derived from the analytic signal.

4. A method for determining a threshold (81, 82) as claimed in claim 1,

wherein the determining of the estimated envelope (31) comprises determining local extrema (42) of the electrophysiological signal (11), and

wherein the distribution (51) of instantaneous amplitudes is derived from a distribution of the local extrema (42).

5. A method for determining a threshold (81, 82) as claimed in claim 1, wherein the threshold (81, 82) is determined to be a fixed multiple of the mode of the distribution of instantaneous amplitudes.

6. A method for determining a threshold (81, 82) as claimed in claim 1, wherein the determining of the threshold (81, 82) comprises fitting a smooth curve (71) to a shape of the distribution (51) of instantaneous amplitudes over an interval around the mode (61).

7. A method for determining a threshold (81, 82) as claimed in claim 1, wherein the smooth curve (71) is a parabola.

8. A method for determining a threshold (81, 82) as claimed in claim 1, wherein the smooth curve (71) is a Rayleigh distribution.

9. A method for presenting a statistical analysis of electrophysiological signals, the method comprising

measuring an electrophysiological signal at one or more positions,

for each one of the measured electrophysiological signals, performing the steps of the method for determining a threshold (81, 82) as claimed in claim 1 to determine, for each one of the measured electrophysiological signals, an estimated Gaussian noise and/or the threshold (81, 82),

performing a statistical analysis on the measured electrophysiological signals based on the estimated Gaussian noise and/or on the threshold of the measured electrophysiological signals,

presenting to the user, a result of the statistical analysis.

10. A method for presenting a statistical analysis of electrophysiological signals as claimed in claim 9, wherein the statistical analysis further comprises determining an average rate of spikes at each one of the positions.

11. A method for presenting a statistical analysis of electrophysiological signals as claimed in claim 9, wherein the presenting comprises presenting the estimated Gaussian noise.

12. A method for presenting a statistical analysis of electrophysiological signals as claimed in claim 9, wherein the statistical analysis further comprises sorting spikes based on an analysis of firing rates present in the electrophysiological signals.

13. A computer program product, which program is operative to cause a processor to perform a method as claimed in claim 1.

14. A system (130) for determining a threshold (81, 82) for spike (12) detection in an electrophysiological signal (11), the system comprising:

an input (131) for receiving the electrophysiological signal (11),

a processing unit (132), configured to;

determine an estimated envelope (31) of the electrophysiological signal (11),

based on the estimated envelope (31), determine an estimated Gaussian noise,

determine a mode (61) of the distribution (51) of the estimated envelope (31), and

determine the threshold (81, 82) based on the mode (61) of the distribution (51) of the estimated envelope (31).

15. A system (130) as claimed in claim 14, further comprising:

a component analysis subsystem (133) for performing a statistical analysis on the measured electrophysiological signals based on the estimated Gaussian noise and/or on the threshold of one or more electrophysiological signals measured at different positions, and

an output (134) for presenting to the user, a result of the statistical analysis as a function of the one or more different positions.