Methods and systems for denoising a neural recording signal

• US 20060276719A1
• Filed: 06/01/2005
• Published: 12/07/2006
• Est. Priority Date: 06/01/2005
• Status: Active Grant

• Current Assignee: Advanced Bionics AG (Sonova Holding AG)
• Original Assignee: Advanced Bionics LLC (Sonova Holding AG)
• Inventors: Litvak, Leonid M., Kulkarni, Abhijit
• Granted Patent: US 7,447,549 B2
• Time in Patent Office: Days
• US Class Current: 600/544
• CPC Class Codes: A61B 5/24 Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
• CPC Class Codes: A61B 5/6849 in combination with a needle set
• CPC Class Codes: A61B 5/7217 of noise originating from a therapeutic or surgical apparatus, e.g. from a pacemaker

Abstract

Methods of denoising a neural recording signal include correlating a neural recording signal with a number of basis functions to obtain a number of weights and then multiplying the weights with the basis functions to obtain a denoised neural recording signal. The basis functions are derived using principal component analysis. Systems for denoising a neural recording signal include one or more devices configured to correlate a neural recording signal with the basis functions to obtain the weights and then multiply the weights with the basis functions to obtain the denoised neural recording signal.

Claims

1. A method of denoising a neural recording signal, said method comprising:

correlating a neural recording signal with a number of basis functions derived using principal component analysis to produce a number of weights; and

multiplying said weights with said number of basis functions to obtain a denoised neural recording signal.

2. The method of claim 1, wherein said number of basis functions is at least one and less than or equal to seven.

3. The method of claim 1, further comprising:

collecting a number of neural recording signals m₁ through m_n into a measurement matrix M; and

using said principal component analysis to derive a matrix V_full that comprises a complete set of basis functions that account entirely for said measurement matrix M.

4. The method of claim 3, further comprising selecting said number of basis functions that are correlated with said neural recording signal from said complete set of basis functions within said matrix V_full by analyzing at least one or more of a percent of unaccounted variance of said neural recording signals in said measurement matrix M and a difference of standard deviations of an error in a beginning portion versus an end portion of said neural recording signals in said measurement matrix M.

5. The method of claim 1, wherein said step of multiplying said weights with said number of basis functions comprises use of the following equation:

6. The method of claim 5, further comprising describing said noise signal with a covariance matrix CD_n=E[n n′

]=T E[n n′

]T′

=T C_nT′

.

7. The method of claim 1, further comprising:

designating said neural recording signal as including an evoked neural response if an amplitude of said denoised neural recording signal exceeds a confidence interval.

8. The method of claim 1, further comprising stimulating a neural tissue with a stimulus pulse to evoke said neural recording signal.

9. The method of claim 8, further comprising determining a minimum amplitude of said stimulus pulse that evokes a neural response signal.

10. The method of claim 1, wherein said neural recording signal corresponds to an auditory nerve.

14. The system of claim 11, further comprising a neural stimulator configured to evoke said neural recording signal by stimulating a neural tissue with a stimulus pulse.

15. The system of claim 14, wherein said neural stimulator comprises at least one or more of a cochlear implant system, a spinal cord stimulator, a deep brain stimulator, and a microstimulator.

11. A system for denoising a neural recording signal, said system comprising one or more devices configured to:

correlate a neural recording signal with a number of basis functions derived using principal component analysis to produce a number of weights; and

multiply said weights with said number of basis functions to obtain a denoised neural recording signal.

12. The system of claim 11, wherein said number of basis functions is at least one and less than or equal to seven.

13. The system of claim 11, wherein said one or more devices are further configured to:

designate said neural recording signal as including an evoked neural response if an amplitude of said denoised neural recording signal exceeds a confidence interval.

16. A system for denoising a neural response signal, said system comprising:

means for correlating a neural recording signal with a number of basis functions derived using principal component analysis to produce a number of weights; and

means for multiplying said weights with said number of basis functions to obtain a denoised neural recording signal.

17. The system of claim 16, wherein said number of basis functions is at least one and less than or equal to seven.

18. The system of claim 16, further comprising:

means for designating said neural recording signal as including an evoked neural response if an amplitude of said denoised neural recording signal exceeds a confidence interval.

19. The system of claim 16, further comprising means for stimulating a neural tissue with a stimulus pulse to evoke said neural recording signal.

20. The system of claim 19, further comprising means for determining a minimum amplitude of said stimulus pulse that evokes a neural response signal.