ECG ANALYSIS FOR DIAGNOSIS OF HEART FAILURE AND CARDIOVASCULAR DISEASE USING SIGNALS OBTAINED FROM AN IMPLANTABLE MONITOR

Abstract

An implantable device for monitoring of a patient's electrocardiogram, has a detection element for determining electrocardiogram signals of a heart of the patient. The signals are indicative of the electrocardiogram of the patient. The implantable device includes a processor that is configured to determine from the signals at least one parameter of the electrocardiogram. The at least one parameter is an R:S ratio defined as the ratio between the absolute value of the R complex and the absolute value of the S complex.

Claims

1. An implantable device for monitoring a patient's electrocardiogram, the device comprising: a detection element for acquiring electrocardiogram signals of a heart of the patient, the signals being indicative of the electrocardiogram of the patient; a processor connected to said detection element and configured to determine from the signals at least one parameter of the electrocardiogram, the at least one parameter being an R:S ratio defined as a ratio between an absolute value of an R complex and an absolute value of an S complex.

2. The device according to claim 1, wherein said processor is configured to record a plurality of QRS amplitudes during a first sampling period and a plurality of QRS amplitudes during a later second sampling period, and to determine a net change of the QRS amplitude between the two sampling periods using a relationship
QRS=QRS.sub.(i)QRS.sub.(i-t), wherein QRS.sub.(i) is an average QRS amplitude of the later second sampling period and QRS.sub.(i-t) is an average QRS amplitude of the first sampling period.

3. The device according to claim 2, wherein said processor is configured to determine and record a plurality of net changes of the QRS amplitude as a function of time.

4. The device according to claim 3, wherein said processor is configured to determine and record the plurality of net changes of the QRS amplitude from a time at which the implantable device is first implanted and started until a time at which the net change of the QRS amplitude has been determined most recently.

5. The device according to claim 1, wherein said processor is configured to determine and record a plurality of R-R durations.

6. The device according to claim 5, wherein said processor is configured to determine a measure of the plurality of R-R durations and said processor is further configured to determine a variance S.sub.(Rd).sup.2 of the measure using a relationship $S_{\left(\mathrm{Rd}\right)}^2=\frac{1}{n-1}.Math.\underset{i=1}{\overset{n}{.Math.}}.Math.{\left({\mathrm{Rd}}_i-\stackrel{\_}{\mathrm{Rd}}\right)}^2,$ wherein n is a number of recorded R-R durations, Rd.sub.i is an i.sup.th recorded R-R duration, and Rd is the measure.

7. The device according to claim 6, wherein the measure is a mean of the plurality of R-R durations.

8. The device according to claim 6, wherein said processor is configured to determine and record a plurality of variances of measures of R-R durations as a function of time.

9. The device according to claim 8, wherein said processor is configured to determine and record at least one or a plurality of F values according to $F=\frac{S_{\mathrm{Rd}\left(\right)}^2}{S_{\mathrm{Rd}\left(\right)}^2},$ wherein S.sub.Rd().sup.2 is a variance of the plurality of variances that has been determined more recently than a variance S.sub.Rd().sup.2, which is also a variance of the plurality of variances.

10. The device according to claim 9, wherein said processor is configured to determine and record a difference F between two of the recorded F values Fx and Fy according to F=FyFx, wherein Fx has been determined and recorded before Fy.

11. The device according to claim 10, wherein said processor is configured to determine and record a plurality of differences of recorded F values.

12. The device according to clam 1, wherein said processor is configured to determine and record a plurality of Q-T durations.

13. The device according to claim 12, wherein said processor is configured to determine a mean of the plurality of Q-T durations according to ${\stackrel{\_}{\mathrm{QT}}}_o=\frac{1}{n}.Math.\left(\underset{i=1}{\overset{n}{.Math.}}.Math.{\mathrm{QT}}_i\right),$ wherein n is a number of recorded Q-T durations, and QT.sub.i is an i.sup.th recorded Q-T duration.

14. The device according to claim 12, wherein said processor is configured to determine and record a plurality of means of Q-T durations as a function of time.

15. The device according to claim 14, wherein said processor is configured to determine and record a difference QT between two of the recorded means QT.sub.x and QT.sub.y according to QT=QT.sub.yQT.sub.x, wherein QT.sub.x has been determined and recorded before QT.sub.y.

16. The device according to claim 15, wherein said processor is configured to determine and record a plurality of differences QT between recorded means.

17. The device according to claim 1, wherein said processor is configured to determine and record a plurality of R:S ratios.

18. The device according to claim 17, wherein said processor is configured to determine a mean of said plurality of R:S ratios according to ${\stackrel{\_}{\left(R.Math.\text{:}.Math.S\right)}}_0=\frac{1}{n}.Math.\left(\underset{i=1}{\overset{n}{.Math.}}.Math.{\left(R.Math.\text{:}.Math.S\right)}_i\right),$ wherein n is a number of recorded R:S ratios, and (R:S).sub.i is an i.sup.th recorded R:S ratio.

Description

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWING

[0068] FIG. 1 shows a schematic representation of a device according to the invention and an electrocardiogram detected by the device;

[0069] FIG. 2 shows an algorithm for determining a difference (net change) between average QRS amplitudes carried out by a processing unit of the device;

[0070] FIG. 3 shows an algorithm for determining a variance of a measure for an R-R duration carried out by a processing unit of the device;

[0071] FIG. 4 shows an algorithm for conducting an F test and for determining a difference between F values;

[0072] FIG. 5 shows an algorithm for determining a difference (net change) between average R-R durations amplitudes carried out by a processing unit of the device;

[0073] FIG. 6 shows an algorithm for determining a difference (net change) between average Q-T durations carried out by a processing unit of the device; and

[0074] FIG. 7 shows an algorithm for determining a difference (net change) between average R:S ratios carried out by a processing unit of the device.

DETAILED DESCRIPTION OF THE INVENTION

[0075] Referring now to the figures of the drawing in detail and first, particularly, to FIG. 1 thereof, there is shown an embodiment of an implantable device 1 for monitoring of a patient's electrocardiogram (ECG).

[0076] The device 1 comprises a detection element (e.g. a lead or a comparable suitable means) 2 for determining electrocardiogram signals S of a heart H of a human or animal patient when the device is implanted in the vicinity of the heart H. The device 1 thus forms a subcutaneous implantable device 1. The signals S are indicative of the electrocardiogram ECG of the patient, which ECG is indicated on the right hand side of FIG. 1. The ECG shows a P wave, QRS complex, a T wave, and a further succeeding QRS complex (the P wave is omitted here) in order to indicate an R-R interval. The U wave following the T wave is not shown. Also indicated is a P-R interval, a QRS amplitude (the sum of the R.sub.peak and S.sub.peak as indicated), QRS duration, and a Q-T inverval.

[0077] Furthermore, the cardiac monitoring device 1 includes a processor 3, or processing unit 3, that is configured to determine from the signals S at least one parameter of the electrocardiogram ECG or a quantity derived from such a parameter, the ECG, particularly for prognosis or diagnosing of CVD or HF or for determining a progress in HF/CVD.

[0078] Particularly, the processor 3 is configured to conduct algorithms for determining those parameters or quantities, particularly for monitoring CVD/HF as will be described in more detail below.

[0079] Further, the device 1 comprises a telemetry unit 4, via which the parameter and/or quantities generated by the processor 3 can be transmitted to a remote external device located outside the human or animal body (e.g. a smart phone, a computer, or a remote server), via which the parameters/quantities can be graphically displayed and/or further analyzed in order to prognose/diagnose and evaluate progression of CVD, HF, or other pathologies referenced previously, of the patient.

[0080] FIG. 2 illustrates an exemplary method for establishing the algorithm to determine a difference (net change) between average QRS amplitudes carried out by a processing unit 3. Here, the electrocardiogram signal is detected by the detection element 2 (step S1). The processing unit 3 detects the QRS complexes (QRS Amplitude Detection Algorithm, step S2) in the received ECG signals (S1) and forms corresponding output variables (QRS output variable, step S3) which are recorded (buffered outputs, step S4). The processing unit 3 further allows an operator to program/select (step S5) the sampling period (programmable sampling period (buffer size)), i.e., the number n of samples, or the period of time over which samples are averaged to determine (average calculation, step S6) an average QRS amplitude QRS.sub.(i) (step S7). The determined averages are recorded (buffered outputs, step S8), wherein in turn the number of recorded averages can be programmed/selected (programmable buffer size, step S9). From these recorded averages an earlier average is selected (prior output value QRS.sub.(i-t.sub.n.sub.), step S10) and a net change (difference) of the QRS amplitude QRS is determined (final output, step S11) according to

QRS=QRS.sub.(i)QRS.sub.(i-t) Eq (1):

wherein (i) indicates the more recent sampling period over which QRS amplitudes have been collected and averaged, while (i-t) indicates the prior sampling period which may be fixed or programmed. The determined net changes QRS in QRS amplitude over the time between the considered sampling periods are recorded/stored (stored history, step S12) to have a history of these values for later analysis, wherein the number of stored net changes QRS can be programmed (programmable buffer size, step S13).

[0081] Equation (1) provides the means by which QRS amplitudes may be evaluated as an indicator of CVD/HF. Particularly, in instances where QRS is negative, the negative value may suggest that factors influencing QRS amplitude like LVEDD, LVESD, LVH, LLE, or Amylosis may developing or worsening and therefore the patient has a poor or worsening prognostic outlook. This makes an educated assumption that QRS amplitude will be negatively correlated with the severity of the diseased state.

[0082] Additionally, by plotting or tracking the QRS or QRS measurements over time (from implant to the most recent sample) one may be able to provide diagnostic information about the efficacy of treatment and patient prognosis.

[0083] Furthermore, Equations (2-1), (2-2), and (2-3) below provide the means by which R-R variability in the detected ECG (cf. FIG. 1) may be evaluated as an indicator of CVD/HF.

[0084] This is used in an embodiment according to FIG. 3, wherein the processing unit 3 is configured to conduct an algorithm for calculating a variance S.sub.(Rd).sup.2 of a measure Rd for the R-R interval (duration) as indicated in FIG. 1.

[0085] For this, according to FIG. 3, the electrocardiogram signal S is detected by the detection element 2 (step S1), and the processing unit 3 detects the QRS complexes (QRS detection algorithm, step S2) in the received ECG signals S and uses time stamps for marking two successive R waves (R-wave marker time stamp, step S3), which are recorded (buffered outputs (2-sample), step S4) and used to determine the corresponding R-R interval (delta of outputs calculation (R-R interval), step S5), which is recorded (buffered delta outputs, step S6), wherein the number of recorded R-R intervals can be selected/programmed (programmable sampling period (buffer size), step S7).

[0086] The recorded R-R intervals (S6) are averaged (average calculation of deltas, step S9) to form a measure (e.g. average) Rd of the R-R intervals (output Rd, step S10) which is used in conjunction with the individual R-R intervals (most recent delta Rd.sub.i, step S8) that have been determined in step S5, respectively, in order to determine a variance of the measure Rd according to

[00007]$\begin{array}{cc}{S}_{\left(\mathrm{Rd}\right)}^2=\frac{1}{n-1}.Math.\underset{i=1}{\overset{1}{.Math.}}.Math.{\left({\mathrm{Rd}}_i-\stackrel{\_}{\mathrm{Rd}}\right)}^2& \mathrm{Eq}.Math..Math.\left(2.Math.\text{-}.Math.1\right)\end{array}$

wherein n denotes the number of samples.

[0087] The resultant variance S.sub.(Rd).sup.2 from Eq (2-1) is then recorded (stored history, step S13) for further analysis and tracking, wherein the number of stored variances S.sub.(Rd).sup.2 can be selected/programmed (programmable buffer size, step S12).

[0088] Further, according to an embodiment shown in FIG. 4, stored values (stored history S.sub.(Rd).sup.2, S3) for the variance may then be compared to one another using an F-Test which may indicate the relative change in R-R variability across time. Here, more recent variance values S.sub.Rd().sup.2 are used (S4) together with an earlier (or the earliest measured) value S.sub.Rd().sup.2 (S5) in order to determine corresponding F values (F-test calculation, step S6) according to

[00008]$\begin{array}{cc}F=\frac{S_{\mathrm{Rd}\left(\right)}^2}{S_{\mathrm{Rd}\left(\right)}^2}& \mathrm{Eq}.Math..Math.\left(2.Math.\text{-}.Math.2\right)\end{array}$

F values greater than one indicate an increase in R-R variability which may indicate an improvement in patient prognosis, whereas values less than one indicate a negative prognosis. Values of 1 are diagnostic of no observed change.

[0089] The time between each compared measure in Eq. (2-2) is particularly programmable/selectable (programmable sample period; S1, S2). Multiple comparisons may also be made and displayed for additional diagnostic and statistical data.

[0090] Additionally, as indicated in FIG. 4, the products of Eq. (2-2) may be stored (stored history, S7) and used for further analysis and tracking. Specifically, Equation (2-3) may provide an index on the progression of autonomic dysfunction over time by comparing two programmable/selectable results (programmable sample period; S8, S9, programmed recent value Fy, S11, and programmed historic value, S12) from the stored products (S7) from Eq (2-2), i.e., by calculating (change calculation F, step S10)

F=FyFx Eq (2-3):

[0091] The time between two programmable/selectable results is, according to embodiments, a period of time that has elapsed between the Fx and Fy measures. It could also be a time interval in that the device is programmed to assess the difference between x and y. For example, a programmed interval of three months means that the device is programmed to compare x to y at three month intervals. Alternatively, the device may be programmed such that it continuously compares each progressive measure to a value that occurred three months earlier.

[0092] Here, particularly, negative values for F may be used as an index of poor prognosis and are diagnostic of a progression of CDV/HF. The time between each compared measure in Eq (2-3) may be programmable/selectable. Multiple comparisons may also be made and displayed for additional diagnostic and statistical data. The values may be stored (stored history, S13).

[0093] Furthermore, according to an embodiment shown in FIG. 5, tachycardia may also be tracked utilizing the variable Rd as well. However, instead of calculating a variance to the population of samples a mean Rd.sub.o will be generated, as shown in Equation (3-1).

[0094] Also here, as indicated in FIG. 5, the electrocardiogram signal S is detected by the detection element 2 (S1), and the processing unit 3 detects the QRS complexes (QRS detection algorithm, S2) in the received ECG signals S and uses time stamps for marking two successive R waves (QRS marker time stamp, S3), which are recorded (buffered outputs (2-sample), S4) and used to determine the corresponding R-R interval (delta of outputs, S5), which is recorded (buffered delta outputs, S6), wherein the number of recorded R-R intervals can be selected/programmed (programmable sample period, S7).

[0095] The recorded R-R intervals Rd.sub.i (step S6) are averaged (Rd.sub.o calculation, S8) to form a mean Rd.sub.o of the R-R intervals according to

[00009]$\begin{array}{cc}{\stackrel{\_}{\mathrm{Rd}}}_o=\frac{1}{n}.Math.\left(\underset{i=1}{\overset{n}{.Math.}}.Math.{\mathrm{Rd}}_i\right)& \mathrm{Eq}.Math..Math.\left(3.Math.\text{-}.Math.1\right)\end{array}$

wherein the number n of samples or the sampling duration may be selectable/programmable (S7).

[0096] The determined means are recorded (Rd.sub.0 stored history, S11) to have a history of means and differences Rd are determined (Rd calculation, S14) according to

Rd=Rd.sub.yRd.sub.x Eq (3-2):

using programmed values (S12, S13), namely a more recent value Rd.sub.y and an earlier value Rd.sub.x.

[0097] Particularly, the individual differences Rd are recorded (stored history, S15) and may be displayed graphically or as calculated by Eq (3-2).

[0098] A value smaller than zero for Eq (3-2) or a negative correlation between time and the values obtained from Eq (3-1) may be diagnostic of worsening CVD/HF.

[0099] Furthermore, according to yet another embodiment shown in FIG. 6, Q-T elongation may be tracked utilizing formulae identical to Eqs. (3-1) and (3-2) by replacing the variable Rd with the measured duration between the Q to the T waves (Q-T) (cf. FIG. 1).

[0100] For this, according to FIG. 6, the electrocardiogram signal S is detected by the detection element 2 (S1), and the processing unit 3 detects the Q and T complexes (QT detection algorithm, S2) in the received ECG signals S and uses time stamps for marking a Q wave and successive T wave (Q marker time stamp, S3, and T marker time stamp, S5), which are recorded (buffered outputs, S4) and used to determine the corresponding Q-T interval QT.sub.i (QT duration, S6), which is recorded (buffered QT duration outputs, S9). From these Q-T durations, a mean is determined (QT.sub.o calculation, S8) according to

[00010]$\begin{array}{cc}{\stackrel{\_}{\mathrm{QT}}}_o=\frac{1}{n}.Math.\left(\underset{i=1}{\overset{n}{.Math.}}.Math.{\mathrm{QT}}_i\right)& \mathrm{Eq}.Math..Math.\left(4.Math.\text{-}.Math.1\right)\end{array}$

wherein the number n of recorded Q-T intervals can be selected/programmed (programmable sample period, S7).

[0101] Particularly, determination of the individual Q-T interval may be most easily implemented by measuring from the peak (absolute maximum for each complex) or preferably through other methods such as a start or end detection method. In the instance the Q wave is not present or has poor resolution, the P-wave or R-wave may be utilized as a non-superior alternative.

[0102] Furthermore, the determined means are recorded (QT.sub.o stored history, S12) so as to have a history of the means and differences QT are determined (QT calculation, S15) according to

QT=QT.sub.yQT.sub.x Eq (4-2):

using programmed values (S13, S14), namely a more recent value QT.sub.y and an earlier value QT.sub.x.

[0103] Particularly, the individual differences QT are recorded (stored history, S16) and may be displayed graphically or as calculated by Eq (4-2).

[0104] A value greater than zero for QT as shown in Eq (4-2) or a positive correlation between time and the equivalent values QT.sub.o obtained from Eq (4-1) may be diagnostic of a worsening of a HF/CDV (e.g. indicator of pump failure, electrical conduction changes, ischaemia, risk for arrhythmia, rhythm disturbances, or ventricular aneurysm).

[0105] Finally, according to the embodiment shown in FIG. 7, the ratio between the absolute value of the R and S complexes, as calculated in Equation (5-1), may provide information diagnostic to RVH and LVH. In particular, devices implanted with the ECG axis oriented parallel to the midline or in a positive deviation toward the left of the midline but less than perpendicular to midline, will exhibit an increase in the R:S ratio over time in the event RVH is worsening, while the value will decrease from the time of implant if LVH is worsening. The same formulae and methodology associated with comparing values, as performed for Q-T elongation and R-R duration, may be utilized to provide this diagnostic data.

[0106] Particularly, according to FIG. 7, the electrocardiogram signal S is detected by the detection element 2 (S1), and the processing unit 3 detects the R and S complexes (RS detection algorithm, S2) in the received ECG signals S and determines their amplitudes (R amplitude, S3, and S amplitude, S4). From these amplitudes the respective R:S ratio ((R:S) calculation, S5) is determined according to

[00011]$\begin{array}{cc}{\left(R.Math.\text{:}.Math.S\right)}_0=\frac{.Math.R_{\mathrm{peak}}.Math.}{.Math.S_{\mathrm{peak}}.Math.}& \mathrm{Eq}.Math..Math.\left(5.Math.\text{-}.Math.1\right)\end{array}$

and recorded (buffered (R:S) outputs, S6). From these R:S ratios, a mean is determined ((R:S).sub.o calculation, S7) according to

[00012]$\begin{array}{cc}{\stackrel{\_}{\left(R.Math.\text{:}.Math.S\right)}}_0=\frac{1}{2}.Math.\left(\underset{i=1}{\overset{n}{.Math.}}.Math.{\left(R.Math.\text{:}.Math.S\right)}_i\right)& \mathrm{Eq}.Math..Math.\left(5.Math.\text{-}.Math.2\right)\end{array}$

wherein the number n of recorded ratios can be selected/programmed (programmable sample period, S8).

[0107] Furthermore, the determined means (R:S).sub.o are recorded (stored history, S11) so as to have a history of the means and differences (R:S) are determined ((R:S) calculation, S14) according to

(R:S)=(R:S).sub.y(R:S).sub.x Eq (5-3):

using programmed values (S12, S13), namely a more recent value (R:S).sub.y and an earlier value (R:S).sub.x.

[0108] Particularly, the individual differences (R:S) are recorded (stored history, S15) and may be displayed graphically or as calculated by Eq (5-3).

[0109] Further, particularly, a value greater than zero for the difference (R:S) or a positive correlation between time and the means (R:S).sub.0 points to a worsening of RVH, and a value smaller than zero for the difference (R:S) or a negative correlation between time and the means (R:S).sub.0 points to a worsening of LVH.

[0110] In FIGS. 5 and 7, S9 and S10 refer to a programmable sample period, respectively. In FIG. 6, S10 and S11 refer to a programmable sample period, respectively.