DD299509A7 - METHOD AND CIRCUIT ARRANGEMENT FOR EVENT-RELATED NONLINEAR TOPOLOGICAL FUNCTIONAL ANALYSIS - Google Patents

Abstract

The invention relates to a method and a circuit arrangement for event-related nonlinear topological functional analysis. The aim of the invention is to determine dynamically non-linear function-relevant topological parameters. Signals that have been measured at several channels are processed into signals that have been processed using stimulus-related signals, while the processed signals and the linear signals are converted into nonlinear signal components by temporal analysis. The latter are then converted into nonlinear parameters by spatial approximation to form nonlinear topological parameters and presented the nonlinear topological parameters as image values. For the first time, event-related nonlinear topological parameters are determined with comparatively low computational complexity.

Description

will be charged.

10. Circuit arrangement for event-related nonlinear topological function analysis, characterized in that the multi-channel measured signals (SM) an input of a signal analyzer (2) and

the stimulus-related signals (SR) are fed to a first input of a memory module (3),

an output of the signal analyzer (2) having a first input of a subtractor (4) and a second input of the memory module (3),

an output of the memory module (3) with a second input of the difference former (4), an output of the difference former (4) with an input of a parameter analyzer (5), an output of the parameter analyzer (5) with an input of a topological approximator (6) and for an output of the topological approximator (6) the nonlinear topological parameters (PTL) are present.

For this 1 page drawing

Application blot of the invention

The invention relates to a method and a circuit arrangement for event-related nonlinear topological functional analysis. »

The invention is used for functional and performance diagnostics for evoked potentials, EMG reactions and other multichannel signals in medical diagnostics, in machine diagnostics, in image sequence analysis and in the de-topological nonlinear analysis of other event-related data faults.

Characteristic of the known state of the art Methods are known for determining dynamically non-linear transmission properties due to reactions to Double Pulse Sequences (Sclabassi, R.I. Krieger, D.N. Berger, T.W .: A systems theoretic approach to the study of CNC function. Annals of Biomedical Engineering 16, New York, 1988, 17-34). Also known are methods for frequency-selective analysis

EEG changes that can be triggered by inner intentions or external stimuli (P. F. Prior, D. E. Maynard:

Monitoring Cerebral Function. Amsterdam, Elsevier 1986, D.GIannitrapi: The Electrophysiology of Intellectual Functions. Basel,

Karger, 1985).

An event-related non-linear analysis of interactions based on function-specific determinations Signal characteristics are not realistic. Also known are methods for determining dynamically non-linear transmission properties due to reactions

on double pulse sequences (Sclabassi, R.I., Kriegor, D.N., Berger, T.W .: A systems theoretic approach to the study of CNC function.

Annals of Biomedical Engineering 16, New York, 1988, 17-34). An extraction of topologically relevant parameters based on this and their topographic processing is not realized. Methods are known for triggering impulse responses and their rich presentation, for example during mapping

evoked hermetic potentials (Duffy, F.H .: Topographic Mapping of Brain Electric Activity, Boston ..., Butterworth 1986,

7-28). Furthermore, methods for topological interpolation of signal parameters are known.

The approximation of dynamically nonlinear temporal-spatial transfer processes is not yet in these methods

realized.

Object of the invention The object of the invention is to provide a method and a circuit arrangement for event-related nonlinear topological To provide functional analysis that is capable of dynamically nonlinear functionally relevant topological parameters

determine.

Furthermore, methods and circuitry with minimal technical and computational effort and

minimum computing time can be realized.

Explanation of the essence of the invention The invention is based on the object, a method and a circuit arrangement for event-related non-linear

specify topological function analysis, with which temporal-spatial dynamically non-linear transmission or

Processing systems efficiently characterized and which according to a topological fi nalysis efficient dynamically

nonlinear topological parameters are determined.

With the help of the invention, a more differentiated systromo-theoretically substantiated characterization of topological Transmission systems are realized. The object is achieved by the method according to the invention characterized in that multi-channel measured signals Signal analysis to processed signals, these using stimulus-related signals to linear signals, the processed Signals and the linear signals by subtraction to njchtllnearen signal components, these by temporal analysis to

Nonlinear parameters, these are processed by spatial approximation to nonlinear topological parameters and the nonlinear topological parameters are presented as image values.

The object is achieved by the circuit arrangement according to the invention characterized in that the multi-channel measured signals

an input of a signal analyzer and the stimulus-related signals are supplied to a first input of a memory module, an output of the signal analyzer with a first input of a subtractor and a second input of the

Memory modules, an output of the Speicliei module with a second input of the subtractor, an output of the Difference generator with one input of a parameter analyzer, one output of the parameter analyzer with one input

are connected to a topological approximator and the nonlinear topological parameters are present at an output of the topological approximator.

Structure and operation of the invention will be described below. The multi-channel measured signals are the input of a signal analyzer and the stimulus-related signals are the

supplied to the first input of the memory module.

The multi-channel measured signals are processed by signal analysis to processed signals. The output of the signal analyzer is connected to the first input of the differentiator and the second input of the Memory modules connected. In the memory module, the processed signals become the linear signals using the stimulus-related signals

processed.

The output of the memory module is connected to the second input of the subtractor. There, the processed signals and the linear signals by subtraction to the non-linear signal components

processed.

The output of the subtractor is connected to the input of the parameter analyzer. There, the non-linear signal components are processed by temporal analysis to non-linear parameters. The output of the parameter analyzer is connected to the input of the topological approximator. There the nonlinear parameters become the nonlinear topological parameters through spatial approximation

processed.

At the output of the topological approximator are the nonlinear topological parameters, which serve as image values

to get presented.

The invention enables an objective analysis of dynamically nonlinear interactions of event-related Signal changes and an objective analysis of dynamically nonlinear transmission properties as topologic

Parameter.

The signal analyzer is preferably constructed in such a case that the multi-channel measured signals according to the relationship SV (M) = F (SM (M)) -F (SM (Mr))

be processed to processed signals.

This will highlight a more efficient potential profiles, in particular for the further processing according to the invention

the signals is essential, sintered.

In another case, the signal analyzer is constructed in such a way that the signals measured in several channels have the relationship SV (M) = ISM (M) I

be processed to processed signals.

By this variant, another group of significant signal characteristics triggered signal changes

characterized, which has not been used.

The different variants take into account different aspects of information processing. They are therefore good

individually and in combination advantageous, for example in medical diagnostics applicable.

The Spoichereinheit is constructed so that for single pulses in the stimulus-related signal, the processed signals as linear Signals are stored. The subtractor is constructed so that the linear signals depend on the relationship SNL (M) »SV (M) - SS (M) - SS (M - dt)

be charged to don non-linear signal components. In this way, system-theoretically justified non-linear transmission properties were determined with minimal computational effort.

In one case, the parameter analyzer is constructed in such a way that the nonlinear parameters can be selected from the nonlinear signal components by the selection at analysis times according to the relationship '.

PNL (k) = SNL (k, ni)

be determined.

In a second case, the parameter analyzer is constructed so that the non-linear parameters at analysis times from the

non-linear signal components according to the relationship

n2

PNL (k) = 1 / (1 + n2-n1) £ SNL (M) I = n1

be determined.

In a third case, the parameter analyzer is constructed so that the non-linear parameters at analysis times from the

non-linear signal components according to the relationship

n2

Σ i ♦ SNL (M) i = n1 PNL (k) «

n2

Σ SNL (M) i = n1

be determined.

These variants take into account different signal characteristics and are adapted to the task

selected.

The determined nonlinear parameters contain important characteristics of the nonlinear transmission behavior, the

So far wedar have been analyzed in technical even in medical systems.

The topological approximator is constructed so that the nonlinear parameters are nonlinear topological parameters

after the relationship

PTL (b) = ΣW (b) * PNL (k) k = 1

will be charged.

Linking nonlinear parameters that dynamically contain nonlinear transmission properties to nonlinear ones

Topological parry has the advantage that hitherto unknown spatio-temporal relationships are instructed and displayed instructively and with minimal effort.

In this review. ang, the topological interpolation allows an instructive representation not previously visible Contexts. In this way, dynamically non-linear function-relevant topological parameters are effectively determined. embodiment

The invention is presented using the example of the evaluation of the event-related electrical brain activity of a subject. Fig. 1 shows the circuit diagram of the circuit arrangement for event-related nonlinear topological functional analysis. The multi-channel measured signals SM are the input of a signal analyzer 2 and the stimulus-related signals SR are supplied to the first input of the memory module 3.

The multi-channel measured signals SM are processed by Slgnalanalyee 2u processed signals SV. The output of the signal analyzer 2 is connected to the first input of the difference former 4 and the second input of the Memory modules 3 connected. In the memory module 3, the processed signals SV become linear using the stimulus-related signals SR Signals SS processed. The output of the memory module 3 is connected to the second input of the difference former 4. There, the processed signals SV and the linear signals SS by subtraction to the non-linear Signal components SNL processed. The output of the difference former 4 is connected to the input of the parameter analyzer 5. There, the nonlinear signal components SNL are processed by temporal analysis to nonlinear parameters PNL. The output of the parameter analyzer 5 is connected to the input of the topological approximator 6. There, the nonlinear parameters PNL become the nonlinear topological by spatial approximation Parameters PTL processed. At the output of the topological approximator 6 are the nonlinear topological parameters PTL, which are referred to as Image values are presented. At the entrance of the examination configuration 7, the stimulus-related signals SR are present. A pulse generator 1 supplies as stimulus-related signal SR single pulses and double pulses of variable distance in a row. The

stimulus-related signal SR controls a Fojostimulator belonging to the examination configuration 7, which in the patient or

Subjects visually evoked potentials triggers. These will continue in the context of the examination configuration 7 Surface electrodes on the head of the patient with a conventional EEG device 16 channels measured and a Computer configuration, consisting of analysis computer and personal computer, fed, A-D converted and an average Procedure undercut. As a result of this procedure, the multi-channel measured signals SM are at the output of the Inspection, Configuration 7. The further processing is advantageously realized in software in this embodiment. In a first embodiment variant, the measured data is band-pass filtered and zero-corrected according to the

relationship

SV (M) = F (SM (M)) - F (SM (Mr)).

The frequency range was 5-100Hz. The reference time tr is in the pre-stimulus interval. In a second embodiment, the measured data SM becomes an envelope according to the relationship

SV (M) = | M (M) |

processed data SV, which correspond to the continuous energy content. The processed data SV can advantageously be further smoothed.

For individual pulses in the relzbezoge ien signal SR, the processed signals SV are stored as linear signals SS. The Distances of the double pulses were stepwise dt = 20 ms-300 ms. The linear signals SS become the relationship SNL (M) = SV (M) - SS (M) - SS (M - dt)

to non-linear signal components SNL charged.

The determination of the nonlinear parameter PNL is carried out in the first variant by nonline & re parameter PNL Selection at analysis times ni from the nonlinear signal components SNL according to the relationship PNL (k) = SNL (k, ni)

be determined.

In the second variant, the determination of the nonlinear parameter PNL is performed by adding nonlinear parameters PNL Analysis time points ni from the nonlinear signal components SNL according to the relationship

n2

PNL (k) = 1 / (1 + n2 - n1) £ SNL (k, i) i = n1

be determined.

In the second variant, the determination of the nonlinear parameter PNL is performed by adding nonlinear parameters PNL Analysis time points ni from the nonlinear signal components SNL according to the relationship

n2

Σ i * SNL (k, i) i = n1 PNL (k) =

n2

Σ SNL (k, i) = n1

be determined.

According to the intersections with the zero line, the following was chosen:

SNL (t-1) <0 SNL (t + 1)> 0 SNLIt -1)> 0

^{n1 = tfor} '.SNUt

^{n2 = t for} I SNUt + 1X0

in a range of 80mm to 120ms under interactive control.

All variants were evaluated. The approximation of the nonlinear topological parameters PTL to be represented as pixels b is achieved by the fact that the

nonlinear parameter PNL to nonlinear topological parameters PTL according to the relationship

PTL (b) = ΣW (b) * PNL (k) k = 1

will be charged.

The nonlinear topological parameters PTL are displayed as an image on a display. These images are used for the subjective evaluation of the topological distribution of the non-linear information transmission devices of the nerversystem or can be subjected to an objective image analysis.

Documents considered:

Prior, P.F .; Maynard, D.E .: Monitoring Cerebral Function. Amsterdam, Elsevier, 1986

- Giannitrapi D .: The Electrophysiology of Intellectual Functions. Basel, Karger, 1985

- Duffy, F. Α .: Topography Mapping of Brain Electric Activity. Boston, Butterworth, 1986, pp. 27-28

Sclabassi, R. J .; Krieger, D.N., Berger, T.W .: A systems theoretic approach to the study of CNC function. Annals of Biomedinal Engineering 16, New York, 1988, p. 17-34

Claims (9)

1. A method for event-related nonlinear topological function analysis using signal analysis method and the topographic approximation, characterized in that multi-channel measured signals (SM) by Signalanalyae to processed signals (SV), the processed signals (SV) using stimulus-related signals (SR) to linear signals (SS), the processed signals (SV) and the linear signals (SS) by differentiation to non-linear signal components (SNL), non-linear signal components (SNL) are temporally analyzed into nonlinear parameters (PNL), the non-linear P '-qmeters (PNL) are spatially approximated to nonlinear topological parameters (PTL), and the non-linear topological parameters (PTL) are presented as image values. 2. The method according to claim 1, characterized in that the multi-channel measured signals (SM) according to the relationship ' SV (k) = F (SM (M)) -F (SM (Mr)) be processed to processed signals (SV). 3. The method according to claim 1, characterized in that the multi-channel measured signals (SM) according to the relationship SV (M) = | SM (M) | be processed to processed signals (SV). 4. The method according to claim 1, characterized in that for single pulses in the stimulus-related signal (SR), the processed signals (SV) are stored as linear signals (SS). 5. Process according to claims 1 and 2 or 3 or 4, characterized in that the linear signals (SS) according to the relationship SNL (M) = SV (M) -SS (M) -SS (M-dt) be calculated to the non-linear signal components (SNL). 6. The method according to claims 1 and 5, characterized in that the non-linear parameters (PNL) by selection at analysis time points (ni) from the non-linear signal components (SNL) according to the relationship PNL (k) = SNL (k, ni) be determined. 7. The method according to claims 1 and 5, characterized in that the non-linear parameters (PNL) at analysis time points (ni) from the non-linear signal components (SNL) according to the relationship n2 PNL (k) = 1 / (1 + n2 - n1) £ SNL (M) i = n1 be determined. 8. The method according to claims 1 and 5, characterized in that the nonlinear parameters (PNL) at analysis time points (ni) from the non-linear signal components (SNL) according to the relationship PNL (k) be determined. 9. The method according to claims 1 and 6 or 7 or 8, characterized in that the nonlinear parameter (PNL) to nonlinear topological parameters (PTL) according to the relationship PTL (b) = W (b) * PNL (k) k = 1