Arrangement for analyzing local bioelectric currents in biological tissue complexes

• US 4,974,602 A
• Filed: 08/15/1989
• Issued: 12/04/1990
• Est. Priority Date: 08/16/1988
• Status: Expired due to Fees

• Current Assignee: Siemens AG
• Original Assignee: Siemens AG
• Inventors: Abraham-Fuchs, Klaus, Schneider, Siegfried, Roehrlein, Gerhard
• Primary Examiner(s): Jaworski, Francis
• Assistant Examiner(s): Manuel, George
• Application Number:
  US07/393,895
• Time in Patent Office: 476 Days
• Field of Search: 128/731, 128/732, 128/734, 128/696, 364/413.06, 364/413.03
• US Class Current: 600/544
• CPC Class Codes: A61B 5/245 specially adapted for magnetoencephalographic [MEG] signals
• CPC Class Codes: A61B 5/35 by template matching
• CPC Class Codes: A61B 5/369 Electroencephalography [EEG] devices for psychotechnics A61B5/16
• CPC Class Codes: A61B 5/372 Analysis of electroencephalograms

Abstract

In an arrangement for analyzing local bioelectric currents in biological tissue complexes, electrical signals measured by EEG and/or MEG sensors are supplied to a pattern recognition unit in which defined signal patterns are acquired and are provided with time markings to produce "templates." The templates are forwarded to a correlation unit which compares the continuous measured values to the templates, and calculates a correlation coefficient based on the comparison. A threshold definition unit identifies signal patterns from the continuously measured signal which transgress the threshold, and supplies the threshold-transgressing signal patterns to a comparison unit which compares the correlated signal to the threshold-transgressing signal. Signal patterns are selected based on this comparison which are temporally and spatially identical, and are supplied to an averaging unit, which forms temporal averages of the selected signal patterns. A localization unit identifies the geometrical location of any source in the pattern which is pathologically electrically active. These geometrical locations are supplied to a monitor together with a tomographic portrayal of the same region of the subject, and are superimposed on the tomograph display.

Claims

1. An arrangement for analyzing local bioelectric currents in biological tissue complexes in a patient comprising:

a plurality of sensors adapted to be spatially distributed over a region of a patient in which a biological tissue complex to be analyzed is disposed, said sensors generating electrical signals based on measured filed quantities generated by local bioelectrical currents;

means for recognizing individual signal patterns within said electrical signals and for temporally limiting said signal patterns to form a template for each signal pattern;

means for storing said templates;

means for spatially and chronologically correlating said electrical signals, after formation of said templates, with said stored templates to form a space-time function and a correlation signal derived from said space-time function;

means for deriving a significant threshold for said correlation signal from said electrical signals;

means for comparing said correlation signal to said threshold and for enabling entry, on the basis of the comparison, of selected ones of the identified signal patterns into a means for chronologically averaging said identified signal patterns;

means for identifying a localized spatial source within said biological tissue complex from the averaged identified signal patterns; and

means for superimposing an image identifying the location of said source on a separately generated tomogram of said biological tissue complex.

2. An arrangement as claimed in claim 1, wherein said sensors are sensor which generate said electrical signals corresponding to measured electrical field quantities generated by said biological electrical currents.

3. An arrangement as claimed in claim 2, wherein said sensors are sensors which generate said electrical signals corresponding to measured magnetic field quantities generated by said biological electrical currents.

4. An arrangement as claimed in claim 1, wherein said electrical signals are analog electrical signals, and further comprising:

means for converting said analog electrical signals into corresponding digital electrical signals; and

means for digitally filtering selected noise frequencies out of said digital electrical signals.

5. An arrangement as claimed in claim 1, wherein said means for recognizing individual signal patterns is a means for automatically recognizing individual signal patterns.

6. An arrangement as claimed in claim 1, wherein said means for automatic recognition of individual signal patterns is programmable, and further comprising means for entering selected programmable criteria for recognition of individual signal patterns into said means for automatic recognition of individual signal patterns.

7. An arrangement as claimed in claim 1, wherein said means for recognition of individual signal patterns is a means for visual manual recognition of individual signal patterns.

8. An arrangement as claimed in claim 1, wherein said means for spatially and temporally correlating comprises:

a first calculating means for forming a spatial correlation function K_R (t_i) and for averaging said spatial correlation function over a time interval defined by the length of said templates according to the formula ##EQU4## second calculating means for forming a temporal correlation function averaged over a space defined by the spatial distribution of said signal sensors constituting said templates according to the formula ##EQU5## and third calculating means for multiplying K_R (t_i) and K_T (t_i), wherein C₁ . . . C_N denote N sensor measuring channels in an arbitrary topical distribution relative to said biological tissue complex, S_C.sbsb.i (t_i) denotes a signal in the channel C_i at a point in time t_i and τ

_o . . . τ

_i . . . τ

_M denote said time interval of a template, beginning at τ

_o and ending at τ

_M with samples at a time τ

_i in said time interval, with τ

_o <

τ

_i <

τ

_M.

9. An arrangement as claimed in claim 1, wherein said means for spatially and temporally correlating comprises means for spatially and temporally correlating said electrical signals with said templates to form said spacetime correlation function K_RT (t_i) according to the formula ##EQU6## wherein C₁ . . . C_N denote N sensor measuring channels in an arbitrary topical distribution relative to said biological tissue complex S_C.sbsb.i (t_i) denotes a signal in the channel C_i at a point in time t_i, and τ

_o . . . τ

_i . . . τ

_M denote said time interval of a template, beginning at τ

_o and ending at τ

_M with samples at a time τ

_i in said time interval, with τ

_o <

τ

_i <

τ

_M.

10. An arrangement as claimed in claim 1, further comprising means for connecting said means for averaging to said templates memory for supplying the output of said means for averaging to said template memory for continual updating of said template.

11. An arrangement as claimed in claim 1, further comprising means for connecting said means for averaging to said means for defining said threshold for continually updating said threshold.

12. An arrangement as claimed in claim 1, further comprising means for summing said electrical signals and for supplying the sum of said electrical signals to said means for recognizing said individual signal patterns.

13. An arrangement as claimed in claim 1, further comprising:

means for intermediately storing said correlation signal;

means for calculating a frequency histogram of the stored correlation signal;

wherein said means for forming a threshold is a means for setting a similarity threshold characteristic for the signal pattern which was the basis for the correlation signal stored in the means for intermediately storing based on a defined deviation of said frequency histogram from a Gaussian normal distribution curve; and

wherein said means for comparing is a means for comparing said intermediately stored correlation signal to said similarity threshold and for enabling entry of the signal section which was the basis for the correlation signal stored in the means for intermediately storing to said means for averaging upon transgression of said similarity threshold.

14. An arrangement as claimed in claim 1, wherein said means for spatially and temporally correlating is an array processor computer means for forming said correlation signal using a fast convolution algorithm.