Patent US4812138A successfully refreshed from Google Patents.
Patent US5417211A successfully refreshed from Google Patents.
US5417211A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Method for the classification of field patterns generated by electrophysiological activities
Assignee: Siemens AG
•
Filing: 1993-08-03
•
Granted: 1995-05-23
•
Expires: 2013-08-03
Abstract
Abstract A method for classifying field patterns generated by electrophysiological activities occurring inside the body of a living subject includes the steps of measuring field patterns arising as a result of the electrophysiological activities outside the body of the subject using a multi-channel measuring apparatus, generating feature vectors corresponding to the measured field patterns, supplying the feature vectors to an adaptive classifier, and training the adaptive classifier with training field patterns which have been generated by a localizable surrogate model of the electrophysiological activity. The method includes the further step of generating a probability value for each field pattern at an output of the adaptive classifier which indicates the probability with which each field pattern can be generated by a selected localizable surrogate model. This classification method greatly reduces the number of possible patterns which must be reviewed by a physician in order to evaluate the electrophysiological activity of the subject.
Inventors
- 👤 Klaus Abraham-Fuchs
- 👤 Martin Schlang
- 👤 Johann Uebler
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-03 11:42
CPC Classifications
G; G16; G16H; G16H50/00; G16H50/20; G06; G06F; G06F2218/00; G06F2218/22; Y; Y10; Y10S; Y10S128/00; Y10S128/92; Y10S128/925; Y10S706/00; Y10S706/902; Y10S706/924
Claims Analysis
Total Claims: 15
Claim 1
Independent Claim
159 words
A method for classifying spatial field patterns generated by electrophysiological activities occurring in the body of a living subject, comprising the steps of: generating a plurality of training field patterns respectively from a plurality of localizable surrogate models and at least one non-localizable surrogate model of electrophysiological activity; training an adaptive classifier with said training field patterns; measuring a plurality of spatial field patterns outside of said subject, generated by electrophysiological activities in said subject, with a multi-channel measuring apparatus to obtain a plurality of respective measured field patterns; generating a plurality of feature vectors respectively representative of each measured field pattern; supplying said feature vectors to a plurality of inputs of said adaptive classifier; generating a probability value for each measured field pattern at an output of said classifier identifying a probability with which said that measured field pattern can be generated by a selected localizable surrogate model; and classifying said spatial field patterns dependent on said probability.
Claim 2
Independent Claim
159 words
A method for classifying spatial field patterns generated by electrophysiological activities occurring in the body of a living subject, comprising the steps of: generating a plurality of training field patterns respectively from a plurality of localizable surrogate models and at least one non-localizable surrogate model of electrophysiological activity; training an adaptive classifier with said training field patterns; measuring a plurality of spatial field patterns outside of said subject, generated by electrophysiological activities in said subject, with a multi-channel measuring apparatus to obtain a plurality of respective measured field patterns; generating a plurality of feature vectors respectively representative of each measured field pattern; supplying said feature vectors to a plurality of inputs of said adaptive classifier; generating a probability value for each measured field pattern at an output of said classifier identifying a probability with which said that measured field pattern can be generated by a selected localizable surrogate model; and classifying said spatial field patterns dependent on said probability.
Claim 3
Dependent Claim
34 words
A method as claimed in claim 1 wherein the step of generating said feature vectors comprises generating said feature vectors in a pre-processing unit connected between said multi-channel measuring apparatus and said adaptive classifier.
Claim 4
Dependent Claim
24 words
A method as claimed in claim 2 comprising the additional step of conducting an amplitude normalization of measured values in each of said channels.
Claim 5
Dependent Claim
28 words
A method as claimed in claim 1 comprising the additional step of superimposing noise on the respective training field pattern from said at least one non-localizable surrogate model.
Claim 6
Dependent Claim
65 words
A method as claimed in claim 1 comprising the step of employing an at least two-layer neural network as said adaptive classifier, said neural network having a plurality of input neurons to which signals from said channels of said multi-channel measuring apparatus are supplied, and an output, forming said output of said adaptive classifier, at which said probability values for said field patterns are present.
Claim 7
Dependent Claim
44 words
A method as claimed in claim 5 wherein the step of employing an at least two-layer neural network as said adaptive classifier comprises employing a neural network having a plurality of input neurons equal to the number of channels of said multi-channel measuring apparatus.
Claim 8
Dependent Claim
52 words
A method as claimed in claim 5 wherein said multi-channel measuring apparatus has at least 30 channels, and wherein the step of employing said neural network comprises employing a neural network having a plurality of hidden neurons approximately equal to one and one-half the plurality of channels of said multi-channel measuring apparatus.
Claim 9
Dependent Claim
61 words
A method as claimed in claim 1 comprising the additional steps of: connecting said multi-channel measuring apparatus and said output of said adaptive classifier to an image display unit; and displaying a chronological curve of said field patterns together with a probability value for which the respective field patterns can be generated by said surrogate model on said image display unit.
Claim 10
Dependent Claim
55 words
A method as claimed in claim 1 comprising the additional steps of: comparing said output of said adaptive classifier to a threshold; forwarding said probability value from said output as a forwarded value if said probability value exceeds said threshold; and forwarding a value of zero from said output as said forwarded value when said
Claim 11
Dependent Claim
32 words
A method as claimed in claim 9 comprising the additional step of: setting said probability value to zero if said probability value fails to exceed said threshold for a minimum chronological duration.
Claim 12
Dependent Claim
62 words
A method as claimed in claim 9 further comprising the steps of: weighting each signal in the respective channels of said multi-channel measuring apparatus with a weighting factor dependent on a plurality of positive field values and negative field values in a field pattern; and multiplying said forwarded value from said output by said weighting factor to obtain a modified probability value.
Claim 13
Dependent Claim
20 words
A method as claimed in claim 1 wherein the step of measuring said electrophysiological activities comprises measuring biomagnetic field patterns.
Claim 14
Dependent Claim
30 words
A method as claimed in claim 1 comprising the additional step of: defining said surrogate model as an electrical current dipole disposed in a space having a uniform electrical conductivity.
Claim 15
Dependent Claim
19 words
A method as claimed in claim 13 comprising the additional step of defining said space as a conductive sphere.
References Cited (Backward Citations)
US4974602A
Cited
US5152288A
Cited
US5092353A
Cited
US5255362A
Cited
US5218529A
Cited
EP0477434A1
Cited
US5280792A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1998-05-22 | 1998-11-21 | 1999-05-23 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2002-05-22 | 2002-11-21 | 2003-05-23 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2006-05-22 | 2006-11-21 | 2007-05-23 | $7700.00 | Window Closed / Lapsed |