Surgery – Diagnostic testing – Cardiovascular
Reexamination Certificate
2001-02-19
2003-06-10
Wolfe, Willis R. (Department: 3747)
Surgery
Diagnostic testing
Cardiovascular
C600S300000, C600S523000
Reexamination Certificate
active
06577892
ABSTRACT:
The invention concerns an apparatus for processing body signals, which includes a sensor for picking up in particular electrical signals from a living body, and means for preparing picked-up signals for further processing, and at least one first memory for a picked-up measurement signal or signal portions.
BACKGROUND OF THE ART
It is frequently a matter of concern to identify given features in for example electrically recorded signals. For example, it may be a matter of interest to identify T-waves or QRS-complexes in an electrocardiogram and to determine as accurately as possible the time at which they occur. To analyse such signals, it is desirable to locate certain events or signal features, from the point of view of time.
With that background in mind, the present invention is concerned in particular with picking up intracardially recorded signals, in particular in an implanted device. For reasons relating to power requirement and the limited amount of space involved, only restricted resources for signal processing and analysis are available in an implanted device.
Various apparatuses and methods of feature identification of heart signals and for signal analysis are already known. European patent No 0 487 429 for example discloses an apparatus in which is stored a sequence of values of such parameters which correspond to an active cardiac cycle. The apparatus includes comparison devices to compare the stored parameter values of the active cardiac cycle to previously stored items of information about those values and, in the event of a positive comparison result, to trigger a signal. German patent applications Nos 199 38 376 and 199 63 246 which are not prior publications also concern apparatuses in which a measured cardiac signal is compared to previously formed and stored comparison signals in order in the case of DE 199 38 376 to identify fusion events in the electrostimulation of the heart and in the case of DE 199 63 246 to detect the circulatory effect of extrasystoles.
U.S. Pat. No. 5,439,483 also discloses the use of a wavelet transform for the classification of tachycardias. In U.S. Pat. No. 5,778,881 a wavelet transform is additionally combined with a hidden Markov modelling in order to be able to detect P-R-waves in each case as Markov states with a reduced number of wavelet coefficients. It is further proposed therein that a set of wavelet coefficients, which is typical for the respective result, can be automatically updated in the event of rapid changes in the signal morphology, in order to make the analysis independent of short-term fluctuations in the physiological signals for example due to physical stress.
The aim of the present invention, in physiological signals such as intracardial electrocardiograms (ECG) and intracardial dynamic impedance patterns (IDZ), is to determine the occurrence of certain features and the moments in time thereof, hereinafter also referred to as time location, in an efficient manner. That should be effected as reliably as possible even in the presence of noise and interference signals. The degree of accuracy of time location of a feature is to correspond to the sampling rate, that is to say the time raster with which the physiological signals are recorded. Feature recognition is also to function in the event of a variation in signal morphology.
Further aims of the invention concern improved feature analysis functions.
Therefore the object of the present invention is to permit better or more efficient signal analysis in comparison with the state of the art at least in individual areas and thus very substantially to attain the above-indicated aims.
SUMMARY OF THE INVENTION
In accordance with the invention, that object is attained by an apparatus of the kind set forth in the opening part of this specification, which includes at least one second memory which contains a predeterminable comparison signal which is finite in respect of time, and signal comparison means which are connected to the second and the first memories and which are adapted for sliding or continuous comparison of signal portions, which overlap in respect of time, of the measurement signal in the first memory to the comparison signal stored in the second memory and for output of a correlation coefficient representing the similarity of each compared signal portion of the measurement signal to the comparison signal.
The integral of the comparison signal in relation to time or its sum of time-discrete signal values is preferably zero. In that way, the correlation signal formed by the correlation coefficients is a signal which is of a mean value of about zero. That greatly simplifies subsequent signal analysis.
In this case, the measurement signal is preferably an intracardial electrogram (ECG) or an intracardially recorded dynamic impedance pattern (IDZ). If those signals are not recorded continuously but in time-discrete manner with a sampling rate, the first memory has a sequence of time-discrete measurement values representing the measurement signal. The comparison signal is also stored in the second memory as a finite sequence of time-discrete values. In that case, instead of the integral of the comparison signal in respect of time preferably its sum of all discrete signal values is zero. That complies with the requirements made in relation to wavelets. Therefore the comparison signal can also be referred to as a comparison wavelet.
A crucial difference in relation to the wavelet transform is that the apparatus does not provide a two-dimensional result, like the wavelet transform, as the comparison pattern is not subjected to time scaling for the investigation of each measurement signal portion over a given frequency range. That decisively reduces the computing power required.
The level of resolution in respect of time of the measurement signal in the first memory is preferably the same as that of the comparison signal in the second memory. Each measurement signal portion to be compared to the comparison signal then corresponds to the comparison signal, in respect of the time duration and the number of discrete measurement values. The correlation coefficients formed by the comparison of the signals also form a sequence of time-discrete values which respectively represent the similarity of precisely one signal portion in the first memory to the comparison signal. The corresponding signal portions of the measurement signal in the first memory preferably overlap in that case in such a way that they are displaced relative to each other only by a discrete time step corresponding to the time resolution of the measurement signal. Due to this sliding comparison, the procedure produces a time-discrete correlation signal which is formed by the correlation coefficients which occur in succession in respect of time, as the result of comparison of the signals. The level of time resolution of the correlation signal is then the same as that of the measurement signal and the comparison signal. The correlation signal however can also be formed with a lower level of time resolution if the comparison operation is implemented only for each second, third or n-th time step. The measurement signal portions which are used for comparison with the comparison signal are then displaced relative to each other from one comparison operation to another comparison operation in each case by two, three or n time steps. In a practical context however it will frequently be advantageous to record the measurement and the comparison signals with a correspondingly lower sampling rate which defines the time raster, if the lower level of time resolution of the signal still permits secure feature identification and reliable signal comparison.
An essential feature of the apparatus is the second memory which contains the comparison signal which can be predetermined in accordance with the respective signal feature to be detected and in particular is variable for adaptation to a varying signal morphology by the apparatus itself.
A preferred apparatus is one in which the signal comparison means are connected to a
Biotonik Mess -und Therapiegeraete GmbH & Co. Ingenieurbuero Ber
Grant Stephen L.
Hahn Loeser + Parks LLP
Wolfe Willis R.
LandOfFree
Apparatus for processing body signals does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Apparatus for processing body signals, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus for processing body signals will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3108049