Method and device for monitoring and optimizing time trigger stability by P and / or T waves

Abstract

The invention relates to a device for monitoring and optimizing the time-triggered stability of extracorporeal circulation support, as well as to an extracorporeal circulation support control unit comprising such a device and to a corresponding method. Accordingly, the present invention proposes a method for monitoring and optimizing time-triggered stability of extracorporeal circulation support, comprising the steps of: receiving an electrocardiogram signal of a supported patient within a predetermined duration; determining a P wave, an R wave and a T wave from the received electrocardiogram signal; a predetermined trigger signal is determined from the electrocardiogram signal of the current cardiac cycle, taking into account the at least one detected P-wave and / or the at least one detected T-wave.

Description

Technical Field

[0001] This invention relates to a method and apparatus for monitoring and optimizing time-triggered stability of extracorporeal circulation support. Background Technology

[0002] To stabilize patients in cases of cardiac pump failure or cardiac pumping dysfunction, extracorporeal circulation support systems providing mechanical support have been developed. These systems can be rapidly connected to the circulatory system, improving blood flow and perfusion to organs, including the coronary arteries, and preventing hypoxia. For example, a blood pump can be connected to a venous access via a venous cannula and to an arterial access via an arterial cannula, for aspirating and pumping blood, respectively. Blood flow can be supplied from the low-pressure side (e.g., via an oxygenator) to the high-pressure side, thereby supporting the patient's circulation.

[0003] It can record electrocardiogram (ECG) measurements and use them to control external support, thereby determining the corresponding characteristic amplitudes of different phases of the cardiac cycle. For example, the R wave, characteristic of the systolic phase of the cardiac cycle, can usually be easily distinguished from other phases of the cardiac cycle. The R wave can be R-triggered and, in conjunction with a predetermined delay time, control the blood pump during the subsequent diastolic phase.

[0004] Cardiopulmonary bypass support can be used for both spontaneous and stimulating rhythms. Stimulating rhythms can include excitatory electrical stimulation of the heart via pacemaker (PM), implantable cardioverter defibrillator (ICD), or cardiac resynchronization therapy (CRT), or non-excitatory electrical stimulation of the heart via cardiac contractility modulation (CCM), as well as combinations of excitatory and non-excitatory electrical stimulation. In all cases, arrhythmias can occur during support, affecting the electrocardiogram (ECG) signal. For example, atrioventricular conduction disorders, ventricular bundle branch block, or even the bradycardia or tachycardia phases of the rhythm can distort the ECG signal, leading to electrophysiological or pathophysiological changes. This can significantly alter the waveform of the ECG signal, potentially affecting or increasing the difficulty of R-wave detection required for cardiopulmonary bypass support.

[0005] Therefore, there may be uncertainties regarding accurate detection, especially when the R wave is unclear in the ECG signal and / or cannot be clearly detected due to fixed parameter settings. This can result in the R wave not being detected, or the timing of R wave detection not coinciding with the actual physiological time of ventricular depolarization. This can severely impact cardiopulmonary bypass support, as the support pulse may not be applied at the expected point in diastole.

[0006] Furthermore, if cardiac stimulation is applied to the patient in addition to cardiopulmonary bypass support, stimulation-related ECG signal interference may occur. For example, the stimulation pulse may be applied within the QRS complex of the corresponding cardiac cycle, and its amplitude, duration, and form may differ (e.g., it may also contain multiple positive and negative stimulation pulses). In such cases, the corresponding R wave may not be detected, or amplitude changes in the ECG signal may be mistakenly detected as R waves. In these situations, cardiopulmonary bypass support cannot be provided with the ideal stability required by the patient.

[0007] Therefore, it is necessary to improve the validation of time-triggered stability of cardiopulmonary bypass support (preferably under different physiological and / or clinical conditions) and optimize the time-triggered stability during cardiopulmonary bypass support. Summary of the Invention

[0008] Based on existing technologies, the purpose of this invention is to improve, verify, and optimize the time-triggered stability of extracorporeal circulation support.

[0009] The above objectives are achieved through the independent claims. Preferred embodiments are described in the dependent claims, the specification, and the drawings.

[0010] Accordingly, this invention proposes a method for monitoring and optimizing the time-triggered stability of extracorporeal circulation support, comprising the following steps: - Receive ECG signals from supported patients within a predetermined time period; - Identify the P wave, R wave, and T wave from the received electrocardiogram signal; - Determine a predetermined trigger signal from the ECG signal of the current cardiac cycle, taking into account at least one detected P wave and / or at least one detected T wave.

[0011] Typically, for cardiopulmonary bypass systems, only one R wave is detected from the corresponding cardiac cycle; ideally, an additional signal of the QRS complex containing the R wave is also detected. As mentioned earlier, this leads to uncertainty in the correct detection of the R wave, especially in the presence of arrhythmias that affect the ECG signal. For example, an R wave may be detected at the wrong time (when the detected amplitude change does not physiologically correspond to an R wave); or an R wave may not be detected in the ECG signal at all; furthermore, two R waves may be detected in the case of a single ventricular depolarization, for example, when a prominent delta wave appears at the beginning of the QRS complex in Wolf-Parkinson-White syndrome, and in the case of intermittent left and right bundle branch block. Therefore, it is impossible to ensure that cardiopulmonary bypass control has ideal reliability and accuracy.

[0012] According to the present invention, by additionally considering the P wave and / or T wave, another influencing factor is incorporated into the determination of the trigger signal, which significantly improves the reliability and accuracy of correctly detecting the trigger signal. For example, the continuous sequence of P wave, R wave, and T wave within the cardiac cycle itself indicates that the R wave is detected between the P wave and the T wave. In this case, the possibility of misinterpreting artifacts, i.e., waves outside the PT time interval, as R waves can be eliminated.

[0013] Furthermore, the detection of P waves can indicate, for example, that atrial depolarization has occurred. For instance, an additional inverted P wave following an R wave may indicate retrograde conduction between the ventricle and atrium during ventricular stimulation; while an inverted P wave preceding an R wave may indicate the presence of a left atrial rhythm, possibly accompanied by sinus bradycardia. Atrial fibrillation without P wave detection can usually be ruled out: if a P wave is detected, the detected R wave can be considered to be triggered by atrioventricular node conduction, and therefore cannot be considered to be triggered by an alternative rhythm.

[0014] T-wave detection can also be used to confirm prior excitation. The detection of a T-wave and two R-waves in a single cardiac activity may indicate bundle branch block due to QRS complex widening (e.g., caused by a prominent delta wave), resulting in a so-called "double count." If no T-wave is detected but an R-wave is detected, there may be invalid right ventricular and / or left ventricular stimulation pulses; in this case, the detected R-wave does not correspond to ventricular depolarization and should not be used as a trigger signal.

[0015] On the other hand, the T wave determined for the current cardiac cycle can, for example, confirm the validity of a trigger signal determined in the form of a P wave (especially an R wave). For example, an R wave may initially be identified as a trigger signal, but it can only ultimately be used as a trigger signal if a subsequent (normal) T wave is identified in the current cardiac cycle. In this case, a delay time can be defined or set based on the R wave for the drive signal supporting cardiopulmonary bypass (e.g., for activating the blood pump in subsequent diastole), but the drive signal is only output if a T wave is detected in the current cardiac cycle. Specifically, the drive signal can be output directly after the T wave, or alternatively, when a tilt at the beginning of the T wave is detected. Furthermore, the T wave itself can also be identified as a trigger signal, for example, if a P wave and / or R wave have been pre-detected in the current cardiac cycle, and the corresponding delay time of the drive signal after the P wave or R wave coincides with the T wave. Thus, the drive of the blood pump supported by cardiopulmonary bypass can, for example, coincide directly with the beginning of diastole or the end of systole.

[0016] In this way, the detected P waves and / or T waves can be used to automatically optimize trigger stability and make adjustments as necessary, such as by adjusting the control of the cycle support accordingly. Using the detected P waves and / or T waves, any potential variability in the ECG signal can be detected and accounted for. For example, if it is determined that a trigger signal cannot be identified, or if an error is suspected in a trigger signal determined from the detected P waves and / or T waves, a corresponding signal or warning signal can be output.

[0017] However, if a trigger signal can be determined based on the detected P wave and / or T wave, or if a correct trigger signal has been determined, then a drive signal and / or control signal for extracorporeal circulation support can be output based on that trigger signal (preferably in conjunction with a predetermined delay time). Alternatively, it can be configured to request input or confirmation from healthcare personnel.

[0018] The predetermined duration may be specifically defined by the treatment duration or a (periodic) portion of the treatment duration, such that the reception of ECG signals preferably continues continuously during patient treatment. This duration may also include at least the current cardiac cycle and at least one previous cardiac cycle, preferably at least two to five previous cardiac cycles. The received ECG signals may be specifically stored in memory or temporarily saved for a predetermined number of cardiac cycles to facilitate the determination of trigger signals and / or necessary evaluation or processing of the ECG signals. A rolling buffer of consecutive cardiac cycles may be configured to continuously acquire the current cardiac cycle and the predetermined number of previous cardiac cycles.

[0019] P waves and / or T waves can be detected for the current cardiac cycle and / or at least one previous cardiac cycle, preferably at least or only for the immediately preceding previous cardiac cycle.

[0020] Preferably, the P wave is detected at least for the current cardiac cycle, which also allows for immediate confirmation or adjustment of the trigger signal to be determined for the current cardiac cycle. In other words, high-resolution monitoring of the trigger signal to be determined can be achieved, and adjustments can be made as necessary. As mentioned earlier, the T wave can also provide feedback for the determined or output trigger signal for the current cardiac cycle. For example, the end of the T wave coinciding with the end of systole or the beginning of diastole can indicate the potential application time of the support pulse, thereby determining whether the trigger signal was detected at the correct time or whether the delay time of the support pulse was set correctly.

[0021] Alternatively, P waves and / or T waves can be detected for one or more previous cardiac cycles. This allows for the detection of any changes in the ECG signal waveform prior to the current cardiac cycle, thereby further improving the accuracy of P wave and T wave detection and trigger signal determination for the current cardiac cycle, while also taking into account changes in the T wave during the current cardiac cycle. For example, if there were no T waves (and / or P waves) in the previous cardiac cycle, a warning signal can be issued.

[0022] Accordingly, it can be configured to determine the P wave at least for the current cardiac cycle and the T wave at least for the immediately preceding cardiac cycle. This allows determination of whether anticipated repolarization or the regular sequence of the previous cardiac cycle occurred from the previous cardiac cycle, while ensuring no overlap between the P wave of the current cardiac cycle and the T wave of the previous cardiac cycle. In this way, the validity of the determined P wave and / or subsequently determined R wave can be enhanced or confirmed, thereby allowing the P wave and R wave of the current cardiac cycle to be identified as trigger signals.

[0023] Furthermore, by averaging the T waves of multiple cardiac activities to detect microvolt-level T-wave alternation, ventricular repolarization can be identified and monitored, and this can be used to optimize support pulses. This allows for the detection and elimination of triggering issues during ST-segment elevation or depression.

[0024] In addition to detecting P, R, and T waves in the time domain, learning-based biosignal processing and analysis methods can be applied using machine learning models (such as linear regression, support vector regression, neural networks, support vector machines, nearest neighbor algorithms, feedforward neural networks (FFNN), recurrent neural networks (RNN), and long short-term memory networks (LSTM)). This can further improve the time-triggered stability during extracorporeal circulation support. Using learning-based methods, statistical parameters such as mean, median, standard deviation, percentiles, skewness, and kurtosis can be determined; furthermore, the parameters of P, T, and R waves in the frequency domain can also be determined.

[0025] Statistical evaluation can be performed on ECG signals corresponding to each P wave and / or T wave, for example, by averaging or normalizing the corresponding cardiac cycle, thereby reducing the weight of small changes in the ECG signal waveform. For example, P waves and / or T waves can also be detected (only) for the immediately preceding cardiac cycle, thus taking into account any variability in the ECG signal in real time, similar to detection for the current cardiac cycle.

[0026] The trigger signal is preferably determined based on the PR interval between the P wave and R wave in the corresponding cardiac cycle. When a predetermined threshold for the PR interval is exceeded, the determination of the predetermined trigger signal and / or the predetermined trigger signal to be determined can be adjusted. In other words, when the threshold is exceeded, the detection type can be adjusted (e.g., by adjusting the parameters used for detection, such as the minimum, maximum, slope, and / or time interval of the predetermined trigger signal to be determined (e.g., the R wave); or the predetermined trigger signal itself can be adjusted so that, for the current cardiac cycle or a subsequent cardiac cycle, the trigger signal to be determined is a P wave or a T wave, rather than the previously determined R wave. Thus, the type of amplitude change specific to the cardiac cycle phase to be determined can be adjusted.

[0027] As mentioned earlier, the detection of the P wave itself has advantages. For example, it can detect or routinely rule out atrial fibrillation, and determine whether atrial fibrillation occurs intermittently, thus allowing consideration of possible alternative rhythms and associated lower heart rates. Furthermore, analyzing the P wave morphology helps differentiate between sinus rhythm and left atrial rhythm, as well as retrograde conduction between the ventricles and atria.

[0028] Furthermore, the PR interval can be used to more reliably detect other cardiac conditions affecting excitation, such as atrioventricular conduction disorders, ventricular bundle branch block, and / or additional conduction pathways between the atria and ventricles (the so-called Kent pathway). For example, if the corresponding threshold is exceeded, atrioventricular block may be present; if the interval is within or below the corresponding threshold, overt Wolf-Parkinson-White syndrome (WPW syndrome) may be present, in which anterograde additional conduction between the atria and ventricles occurs via Kent fibers, leading to QRS complex changes and prolongation. In addition, the combination of maximum pre-excitation of the QRS complex via Kent fibers and the possible presence of Mahaim fibers between the atrioventricular node (AV node) and right ventricular myocardium may alter the ECG interval. Prolonged PR intervals may, on the one hand, cause the P wave to merge with the preceding T wave, increasing the difficulty of P or T wave detection if the detection method is not adjusted accordingly; on the other hand, prolonged PR intervals may also be accompanied by QRS complex widening or electrophysiological changes, making the R wave undetectable. However, if the R wave is to be identified as the trigger signal, a prolonged PR interval may reduce the stability of the trigger signal. Considering the PR interval also has advantages, as it describes the onset of the systolic phase of the cardiac cycle. If a prolonged PR interval or one exceeding a certain threshold is detected, another trigger signal can be assigned to the current cardiac cycle or at least subsequent cardiac cycles; for example, replacing the previously identified R wave with a P wave, to ensure ideal trigger stability.

[0029] The PR interval is used as the criterion for judgment in the above and below texts; alternatively, if a Q wave needs to be (clearly) detected in the electrocardiogram signal in addition to the R wave, the PQ interval can also be considered.

[0030] Preferably, when a predetermined threshold for the PR interval is exceeded for at least one previous cardiac cycle, the detection of the P wave, R wave, and / or T wave in the current cardiac cycle is adjusted. As mentioned earlier, if the PR interval is prolonged, the P wave may merge with the T wave, making P wave and / or T wave detection more difficult, and QRS complex widening or electrophysiological changes in the QRS complex may also occur (e.g., reduced R wave amplitude, or even disappearance of the R wave). By adjusting the detection method, the detection of the current cardiac cycle can be adapted to the cardiac disease pattern.

[0031] For example, the detection method can be adjusted to provide different tilt values ​​or detection times to be detected; the corresponding bandpass filter can also be changed; alternative ECG leads can be selected for the corresponding detection (which can be spatially and / or anatomically separated from the previously used ECG leads, for example); in addition, multiple ECG leads can be considered to optimize the signal (e.g., the corresponding ECG signal can be processed by "signal averaging", and averaging is preferred).

[0032] When there is no anterograde conduction via the Kent pathway, the threshold for the PR interval is preferably at least 120 ms. This is because, with anterograde conduction via the Kent pathway, the threshold for the PR interval shortens to several milliseconds, and the end of the P wave may overlap with the start of the QRS complex, resulting in a 0 ms RQ interval. Depending on the delta wave characteristics and the selected ECG lead, the threshold for the PR interval with anterograde conduction via the Kent pathway is preferably between 0 and 120 ms. This allows for a higher probability of determining whether the PR interval is non-physiological (e.g., whether it corresponds to first-degree atrioventricular block). The threshold for the PR interval is preferably between 150 ms and 400 ms, specifically between 200 ms and 300 ms. This threshold, for example, may correspond to first- or second-degree atrioventricular block; exceeding this threshold may result in QRS complex variations and impaired R wave detection.

[0033] By analyzing the PR interval and P wave morphology, first-degree atrioventricular block (AV block I), second-degree type I AV block (“Wenckebach”), second-degree type II AV block (“Mobitz”), third-degree AV block (AV block III), and sinoatrial block can be assessed. This analysis of atrioventricular biosignals can further support the temporal stability of R triggering.

[0034] In the absence of atrioventricular conduction obstruction and anterograde conduction between the atria and ventricles via an additional Kent pathway, the PQ interval can range from 120 ms to 200 ms, with a threshold selected accordingly. This threshold can also depend on the cardiac cycle duration and / or heart rate, as a higher heart rate in physiological conditions generally results in a shorter expected PQ interval. Accordingly, the PQ interval threshold can correspond to at least 20%, preferably at least 30% or at least 35% of the cardiac cycle duration. For example, in the absence of atrioventricular conduction obstruction, the PQ interval threshold can also be less than 200 ms, but should be higher than a patient-specific (patho)physiological value, specifically at least 10% higher than the patient's (patho)physiological value (e.g., a threshold of 180 ms when the patient's (patho)physiological value is 160 ms). Therefore, variations in the PR interval due to bradycardia or tachycardia can be considered.

[0035] In ventricular stimulation using a pacemaker, retrograde conduction from the ventricle to the atrium can occur, producing additional P waves with altered morphology (e.g., suppressed P waves). This retrograde P wave can particularly affect the pacing behavior and R triggering of dual-chamber pacemakers. For example, an RP interval with a retrograde conduction time of 200 to 300 ms may lead to premature ventricular stimulation, resulting in tachycardia, affecting hemodynamics, and causing RR interval fluctuations. Therefore, detecting P waves and determining the PR and RP intervals is beneficial for assessing intermittent or permanent retrograde conduction between the ventricle and atrium, thereby optimizing the temporal stability of R triggering.

[0036] For example, to reduce the weight of individual, rare, and / or inconspicuous "outliers," the PR interval and / or RP interval of at least two cardiac cycles can be averaged. Accordingly, the time interval between the corresponding P wave and R wave of the corresponding cardiac cycle is averaged, i.e., at least two PR intervals (RP intervals) are averaged. Preferably, the PR interval and / or RP interval are averaged over consecutive cardiac cycles. For example, a mean with standard deviation or a median based on a histogram can be determined for multiple cardiac cycles. Thus, the detection and / or the trigger signal to be determined is adjusted, for example, only when the mean, standard deviation, or median exceeds the corresponding statistical threshold. The advantage of this approach is that small and / or temporary changes do not immediately trigger corresponding parameter adjustments, thereby offsetting or smoothing the impact of adjustments on circulatory support.

[0037] However, the number of PR and / or RP intervals to be averaged can be kept small, for example, limited to 2 to 5 cardiac cycles. Only when the threshold is exceeded can the presence of pathological changes and / or retrograde conduction from the ventricle to the atrium be verified; at the same time, rapid adjustment of the detection and / or trigger signal to be determined can be achieved.

[0038] Preferably, the PR interval and / or RP interval of at least two consecutive cardiac cycles are compared. When the PR interval and / or RP interval are continuously prolonged, the determination of the predetermined trigger signal and / or the predetermined trigger signal to be determined is adjusted.

[0039] For example, a prolonged PR interval may be caused by Wenckebach block; a prolonged RP interval may be caused by retrograde Wenckebach phenomenon. For instance, the PR interval may be prolonged over two to three consecutive cardiac cycles, in which case conduction fails, meaning no R wave or ventricular excitation occurs. For example, in the case of Wenckebach block, only one in every two or three P waves may conduct. This pathophysiological condition may mean that R waves are not detected, and the P wave is used as a trigger signal instead.

[0040] This continuous prolongation is particularly likely to occur when a corresponding threshold is exceeded. For example, the degree of prolongation can be determined by the absolute difference or relative to previous cardiac cycles. For example, the prolongation may be 50 ms to 100 ms, and / or at least 20%.

[0041] Alternatively, the trigger signal can be determined based on the RT interval between the R wave (or multiple P waves in the case of retrograde conduction between the ventricle and atrium) and the T wave in the corresponding cardiac cycle. Preferably, the determination of the predetermined trigger signal and / or the predetermined trigger signal to be determined is adjusted when a predetermined threshold for the RT interval is exceeded.

[0042] Considering the RT interval (or similarly, the QT interval) can also more reliably detect cardiac diseases affecting excitation, such as ventricular excitation reversal disorders with or without bundle branch block. For example, RT interval prolongation may result from the fusion of the T wave with the subsequent P wave and / or QRS complex widening. These QRS complex changes may be caused by an additional conduction pathway between the atrium and ventricle (i.e., the so-called anterograde Kent pathway). The Kent pathway can be located at the level of the right and / or left ventricular valves and can conduct continuously or intermittently anterograde or retrogradely. For example, anterograde conduction through the Kent pathway via the left lateral wall produces a so-called delta wave, resulting in a corresponding widening of the QRS complex and a shortening of the PR interval. In Wolf-Parkinson-White syndrome (WPW syndrome), the left ventricle is prematurely excited due to the additional anterograde conduction through the Kent pathway, leading to QRS complex widening. A shortened PR interval can indicate overt WPW syndrome and is accompanied by QRS complex widening.

[0043] The threshold of the RT interval can be exceeded as an indicator of pathological conditions, such as enabling early adjustment of the acquired and / or determined trigger signal. Therefore, trigger stability can be significantly improved even in the presence of arrhythmias. As previously mentioned, detecting and considering the T wave not only improves trigger stability itself (e.g., by verifying whether excitation has occurred), but also, with the help of the RT interval, allows the trigger signal to be more precisely adapted to specific pathophysiological states.

[0044] T-wave shortening may occur in certain heart conditions or arrhythmias. While the threshold preferably represents an upper time limit, it can also include a lower time limit. For example, the threshold can be defined as a range that should neither be lower than nor higher than under normal circumstances.

[0045] The preferred setting is: if a predetermined threshold for the RT interval is exceeded for at least one previous cardiac cycle, then the detection of the P wave, R wave and / or T wave in the current cardiac cycle is adjusted.

[0046] The T wave appears at the end of systole, following the P and R waves, which helps optimize the detection of subsequent cardiac cycles. For example, if the RT interval is prolonged, the P wave of the current cardiac cycle may merge with the T wave of the immediate preceding cardiac cycle. Accordingly, more precise detection of the P wave may be required (e.g., by changing parameters, such as a predetermined frequency range for the P wave).

[0047] The preferred threshold for the RT interval is at least 350 ms. Similar to the PR interval described above, this allows for the identification and consideration of non-physiological RT intervals (which may require adjustments to the detection and / or the trigger signal to be determined).

[0048] The threshold for the RT interval can also be set to at least 450 ms. This allows for the detection of certain types of arrhythmias (e.g., atrioventricular block) and their consideration when determining trigger signals. As mentioned above regarding the PR interval, this threshold can also depend on the duration of the cardiac cycle and / or heart rate. It should be noted that under physiological conditions, a higher heart rate results in a shorter expected QT interval and consequently a shorter RT interval. Therefore, the RT interval threshold can correspond to at least 35% of the cardiac cycle duration, preferably at least 40%, or even at least 45%.

[0049] As described above regarding the PR interval, the RT interval can be averaged over at least two cardiac cycles.

[0050] Therefore, T-wave detection alone can enable monitoring of the current cardiac cycle; the RT interval can also identify potential pathophysiological changes; furthermore, T-wave detection can be used, for example, to detect arrhythmias. Accordingly, if the T-wave amplitude of at least one previous cardiac cycle contains negative values ​​or the total value is negative, it is preferable to adjust the determination of the predetermined trigger signal, the predetermined trigger signal to be determined, and / or the detection of the P wave, R wave, and / or T wave of the current cardiac cycle.

[0051] Specifically, this method can be used to adjust subsequent cardiac cycles. If the amplitude is negative or significantly flattened, the result may indicate bundle branch block in the cardiac conduction system.

[0052] The trigger signal is preferably determined based on the QRS interval between the Q wave and S wave in the corresponding cardiac cycle. When a predetermined threshold for the QRS interval is exceeded, the determination of the predetermined trigger signal and / or the predetermined trigger signal to be determined is adjusted. Alternatively, if a predetermined threshold for the QRS interval is exceeded for at least one previous cardiac cycle, the detection of the P wave, R wave, and / or T wave in the current cardiac cycle is adjusted.

[0053] Due to arrhythmias (especially in the case of atrioventricular bundle branch block), the QRS complex may be widened or prolonged, or the ECG waveform within the QRS complex may change significantly (e.g., the R wave is absent, or its amplitude is much lower than normal when it appears). If the QRS interval threshold is exceeded, this situation (e.g., due to bundle branch block) may prevent the identification of a trigger signal (e.g., the R wave) within the QRS complex. In this case, another amplitude change (preferably the P wave) can be identified as the trigger signal; or any parameter values ​​required for detection can be adjusted to optimize the detection of subsequent cardiac cycles (e.g., adjusting the frequency range set for detecting the QRS complex, and / or setting a lower threshold for detecting the R wave).

[0054] The threshold for the QRS interval is preferably at least 120 ms. This allows for the exclusion of physiological values ​​(especially between 90 ms and 110 ms) and the detection or confirmation of arrhythmias.

[0055] Each of the aforementioned PR, RT, or QRS intervals can improve trigger stability, thereby enhancing cardiopulmonary bypass support; however, it is preferable to consider at least two of these intervals when determining the trigger signal. This allows for more accurate detection of any arrhythmia, and even early prediction during cardiopulmonary bypass control. A combination of PR, RT, and QRS intervals is preferred, as these intervals can ideally collectively constitute the electrocardiographic features.

[0056] The detection of P waves, R waves, and T waves, as well as the determination of each interval, can be performed automatically (e.g., by executing the method on a computer). This process can be performed, for example, by the processor of the electrocardiogram device or the extracorporeal circulation support control unit. Consequently, the detection method can also be automatically adjusted, or adjustments can be provided / suggested, thereby continuously monitoring and improving trigger stability.

[0057] Thresholds and / or predetermined trigger signals for one or more intervals within a corresponding cardiac cycle can be based on pathological electrocardiogram (ECG) characteristics. For example, a threshold for the PR interval can be set for arrhythmias such as first-degree atrioventricular block; in this case, the P wave (instead of the R wave) can also be defined as the trigger signal. Thus, pathological features characterizing a particular cardiac disease pattern can, where necessary, specify the ECG signal waveform or the duration of one or more specific intervals.

[0058] Pathological ECG features are determined and / or predetermined for the current cardiac cycle based on intervals detected in previous cardiac cycles. For example, pathological ECG features can be predetermined for a patient's cardiopulmonary bypass support to optimize the triggering stability of circulatory support from the initial stage. For example, the ECG signal of the supported patient can be compared with a corresponding threshold for a specific interval (such as the PR interval) in a set or built-in pathological ECG feature.

[0059] However, pathological ECG features can also be determined using ECG signals from previous cardiac cycles. Therefore, appropriate thresholds can be determined by performing “online” analysis of ECG signals during the current cycle support (e.g., through statistical evaluation and learning-based methods in the time and frequency domains). For example, intervals in consecutive cardiac cycles (such as the PR and / or RT intervals) can be detected and determined, defined as the median of a histogram, mean, or with standard deviation or percentiles, as well as skewness and kurtosis (curvature).

[0060] For example, ECG signals can be evaluated "online" while cardiopulmonary bypass support is being performed, and optionally used for learning-based methods. ECG signals can be received immediately (preferably continuously) after acquisition or detection. During ECG signal detection and cardiopulmonary bypass support, the initially acquired data can be validated and labeled, subsequently used to determine one or more thresholds for subsequent cardiac cycles. Data can also be evaluated or analyzed with a delay during cardiopulmonary bypass support to achieve "near-line" analysis. Learning-based methods can include linear regression, support vector machines (SVM), artificial neural networks, feedforward neural networks (FFNN), recurrent neural networks (RNN), and long short-term memory networks (LSTM).

[0061] The electrocardiogram (ECG) signal and corresponding data can be at least partially or completely buffered (preferably for a predetermined duration). Thus, for example, when determining the trigger signal and / or various thresholds, the evaluation unit can take into account the waveform of the patient's own ECG signal, thereby improving or enhancing monitoring accuracy and, where necessary, improving the accuracy of trigger signal localization.

[0062] In addition, the trigger signal to be determined can be predetermined based on previous cardiac cycles. For example, if the PP interval or RR interval detected in a minimum number (optionally consecutive) of cardiac cycles is consistent and has small deviations, the P wave or R wave can be designated as the trigger signal.

[0063] In addition to predetermined ECG features, pathological ECG features can be determined based on previous cardiac cycles, allowing for immediate monitoring once trigger stability is established, while also being immediately adapted to the patient's specific condition.

[0064] Finally, intervals detected from previous cardiac cycles can be compared with a dataset of “offline” assessments containing ECG signals from supported patients with common pathological features to be treated. The ECG features that best match the ECG signals in the dataset can be predefined as parameters, with learning-based methods particularly applicable to determine the degree of match.

[0065] For example, P waves, R waves, and T waves in “offline” stored datasets can be labeled to improve the accuracy of detected intervals. The labeled data can be used to specify statistically optimized thresholds for the intervals used (corresponding to their respective pathologies); additionally or alternatively, appropriate frequency ranges and / or appropriate amplitude values ​​for P waves, R waves, and T waves can be predetermined to improve the matching degree by comparing the patient’s ECG signals with the dataset.

[0066] For each cardiac pathology, a range of intervals, frequency ranges, and / or amplitude ranges can be specified or predetermined, and / or a trigger signal to be determined within a defined interval (such as the RT interval). The interval used can be specifically selected from the PR interval, RT interval, and / or QRS interval.

[0067] To improve the detection accuracy of amplitude changes in electrocardiogram (ECG) signals, P waves, R waves, and / or T waves can be detected after passing through a bandpass filter. The preferred frequency range for P waves is 5 Hz to 20 Hz, for R waves it is 10 Hz to 25 Hz, and / or for T waves it is 1 Hz to 5 Hz. According to the present invention, it has been recognized that processing ECG signals through the aforementioned frequency ranges can significantly improve the detection accuracy of each amplitude change. If the detection method needs to be adjusted during cyclic support (e.g., due to exceeding a threshold for a specific interval within the cardiac cycle), one or more frequency ranges can be adjusted accordingly for the detection settings.

[0068] Typically, to control cardiopulmonary bypass support, the R wave or R wave spike is identified from the corresponding cardiac cycle and used as a suitable trigger signal. As mentioned earlier, under certain pathophysiological conditions, R wave detection may be timed incorrectly or affected (e.g., the detected amplitude change is not an R wave). For example, if the patient receives cardiac stimulation, the stimulation may produce artifacts in the electrocardiogram signal, also leading to the above situation. Therefore, in some cardiac conditions, the R wave may not provide the ideal or required trigger stability for cardiopulmonary bypass support.

[0069] In this case, the predetermined trigger signal may be an R wave; taking into account at least one detected P wave and / or at least one detected T wave, alternative or additional trigger signals (such as P wave or T wave) may be selected.

[0070] Therefore, the R wave can be predetermined as the trigger signal; however, if the threshold of the PR interval is exceeded, for example, the trigger signal can be adjusted, such as using the P wave as the trigger signal for at least one subsequent cardiac cycle. In this way, even in the event of ECG signal distortion or variation, the trigger signal can be determined and used with sufficient stability.

[0071] Using the P wave also offers the following advantages: for example, the PR interval of the current cardiac cycle can be utilized, and even the trigger signal can be adjusted for the current cardiac cycle. The P wave and PR interval do not conflict with or overlap with the subsequent diastolic phase. A trigger signal with an appropriate predetermined delay time can be provided or output by an algorithm to control cardiopulmonary bypass support during the subsequent diastolic phase, even if the R wave cannot be detected, or cannot be detected with sufficient reliability.

[0072] The preferred configuration is as follows: based on a determined trigger signal, a signal supporting extracorporeal circulation is output with a delay time; the predetermined delay time takes into account at least one detected P wave and / or at least one detected T wave.

[0073] The circulatory support signal can be, for example, a control signal for cardiopulmonary bypass (such as a control signal for a blood pump). For instance, when the PR interval and / or QRS interval are prolonged, the delay time can be adjusted to ensure that a signal is not mistakenly delivered during an unexpected phase of the cardiac cycle. Adjusting the delay time prevents cardiopulmonary bypass from applying a support pulse during the systolic phase of the cardiac cycle.

[0074] Furthermore, it can be configured to automatically adjust or suggest delay times based on the trigger signal to be determined. This process is advantageous, for example, if the trigger signal changes (e.g., a P wave replaces the R wave as the trigger signal). A predetermined delay time or delay may be insufficient to prevent overlap with systole, thus requiring adjustment of the delay time. For example, an R-wave-based delay time can prolong the PR interval, or alternatively, prolong the PQ interval.

[0075] If neither a Q wave, P wave, nor R wave is detected, the delay time can be adjusted (e.g., by incorporating a hemodynamic sensor) in the event of excitation formation and / or conduction disturbances and / or cardiac stimulation. Adjusting the delay time can involve complex cardiac electrophysiological and hemodynamic conditions, particularly those affected by electromechanical coupling. For example, cardiac electromechanical decoupling can result in cardiac electrical excitation without cardiac mechanical contraction. However, a hemodynamic sensor can be used to provide an alternative or redundant system to enable optimized triggering even in these situations.

[0076] The delay time of the current cardiac cycle can also be determined from the detected QT time of at least one previous cardiac cycle; the QT time can be normalized using the heart rate determined from the electrocardiogram signal.

[0077] The QT duration corresponds to the time interval in the cardiac cycle from the onset of the Q wave in the QRS complex to the end of the subsequent T wave. Physiologically, the end of the T wave coincides with the closure of the aortic valve, marking the end of systole or the beginning of diastole. A longer QT duration is advantageous because, empirically, it roughly corresponds to the duration of the subsequent diastolic phase. The delay time and support pulse duration can be adjusted to provide the ideal timing for the trigger signal at the corresponding stage of the cardiac cycle.

[0078] Normalization and / or consideration of heart rate has the following advantages: QT time is typically dependent on the patient's heart rate, and variations in QT time are expected to accompany variations in heart rate. Conversely, determining the heart rate-optimized delay time in this way is particularly advantageous. Therefore, normalized QT time already accounts for the variability of heart rate. Thus, the delay time can be selected using the Framingham formula to represent the heart rate-based QT time. In this way, the (theoretical) delay time of the current cardiac cycle can be calculated using the normalized value and the current heart rate. The normalized value can be based on empirical values, such as those predetermined using datasets previously evaluated and validated "offline." For example, the delay time can be calculated using the following variant of the Framingham formula: Delay time = 380 - 154 × (1 - 60 / heart rate) In the case of right ventricular pacing (RVP), the QT time can be advantageously determined by stimulating the T time. In rate-adaptive pacemakers, the stimulation frequency during exercise can be controlled by measuring the RVP T time.

[0079] Heart rate is preferably determined by the time interval between two R waves in a continuous cardiac cycle (the so-called RR interval) or the time interval between two P waves in a continuous cardiac cycle (the so-called PP interval). Therefore, the heart rate of the current cardiac cycle can also be considered; the delay time can be determined and optimized for each cardiac cycle, and if the heart rate changes, the delay time will be almost automatically corrected. For example, if an increase in heart rate leads to a shortening of the systolic phase, the delay time can be adjusted accordingly. Thus, when determining the delay time, the corresponding shortening of the QT time can be considered by taking into account the increase in heart rate.

[0080] Therefore, the trigger signal is highly likely to be output at a non-systolic time point, thereby minimizing or even avoiding afterload. If the P wave is identified as the trigger signal, in addition to the heart rate-optimized delay time, the PQ interval or PR interval can be extrapolated to the delay time. For example, the delay time can be automatically corrected after a change in the trigger signal.

[0081] Alternatively, impedance cardiography can be used to determine the delay time using the so-called "left ventricular ejection time (LVET)," in which the end of the LVET interval marks the closure of the aortic valve.

[0082] Furthermore, the present invention also proposes an apparatus for monitoring and optimizing the time-triggered stability of cardiopulmonary bypass support, the apparatus comprising: (i) an interface for receiving electrocardiogram (ECG) signals from a supported patient over a predetermined duration; and (ii) an evaluation unit configured to determine P waves, R waves, and T waves from the received ECG signals. The evaluation unit is also advantageously configured to determine a predetermined trigger signal from the ECG signal of the current cardiac cycle, taking into account at least one detected P wave and / or at least one detected T wave. Specifically, the apparatus can be configured to perform the method of the present invention described above.

[0083] The device can receive electrocardiogram (ECG) signals from supported patients via its interface; these ECG signals can be processed by working memory and / or stored in a storage medium. For example, the ECG signals can be communicatively connected to an assessment unit via the interface; the assessment unit may include both working memory and a storage medium. Preferably, ECG signals are received from multiple spatially separated ECG leads via the interface; multiple ECG signals can also be received. Therefore, the assessment unit is advantageously configured to assess ECG signals to determine P waves, R waves, T waves, and trigger signals.

[0084] Preferably, the device is part of an electrocardiogram (ECG) device; however, the device may also be communicatively connected to at least one ECG device via the aforementioned interface (one or more); preferably, the device is integrated into an extracorporeal circulation support control unit. The control unit may be configured to output an extracorporeal circulation support control signal with a predetermined delay time, taking into account at least one P wave and / or at least one T wave, after a predetermined time following a specific trigger signal.

[0085] In the control unit, the device can preferably be provided as an integrated electrocardiogram (ECG) device, or connected to the control unit as an ECG device. This allows the control unit to be used independently of other components and results in a compact design. Preferably, the ECG device is integrated into a single housing of the extracorporeal circulation support system (e.g., integrated into a sensor box as an ECG card or ECG module).

[0086] The control unit may also be located in a console with a user interface for inputting and reading system settings (especially parameters of the extracorporeal circulation support system's blood pump and / or electrocardiogram equipment). For example, the console may include a touchscreen and / or a display with a keyboard that the user can operate. The control unit preferably operates, drives, controls, adjusts, and monitors the blood pump; it enables the blood pump to be synchronized with the cardiac cycle, particularly with trigger signals provided by the patient.

[0087] Other advantages, possible implementations and further improvements of this method have been described in detail above in conjunction with the control unit. To avoid repetition, the relevant content will not be repeated here, but the relevant content of this disclosure still applies to this topic. Attached Figure Description

[0088] The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings: Figure 1 The electrocardiogram signal curves of continuous cardiac cycles under physiological conditions are shown. Figure 2 It shows Figure 1 The curves of the electrocardiogram signals shown are labeled with the characteristics of the electrocardiogram. Figure 3 The changes in electrocardiogram curves during additional anterograde conduction via the Kent pathway in WPW syndrome are shown. Figure 4 The changes in electrocardiogram curves in atrioventricular reentrant tachycardia (AVRT) with retrograde Kent pathway are shown; Figure 5 It shows the relationship with Figure 4 Similar changes in the electrocardiogram curve, accompanied by intermittent right bundle branch block; Figure 6 An advantageous embodiment of the device of the present invention is shown. Detailed Implementation

[0089] The preferred embodiments are described below with reference to the accompanying drawings. In these drawings, the same, similar, or functionally similar elements are labeled with the same reference numerals; to avoid repetition, some repeated descriptions of these elements are omitted.

[0090] Figure 1 The electrocardiogram (ECG) signal curves for two consecutive cardiac cycles within a predetermined duration are shown; these curves are physiological curves. Each cardiac cycle phase is color-coded: each cardiac cycle begins with atrial depolarization, i.e., the P wave, characterized by amplitude changes preceding the characteristic R wave, with a current value I of approximately 0.1 mV.

[0091] Each cardiac cycle is followed by a characteristic and easily identifiable R wave or R wave spike, with a current value I of approximately 1 mV. The R wave, which is usually easily detectable under physiological conditions, is caused by ventricular depolarization and marks the beginning of systole. The R wave is preceded by a Q wave and followed by an S wave, both of which have low negative current values; these amplitude variations together constitute the so-called QRS complex.

[0092] After the R wave induces excitation, ventricular repolarization or excitation recovery occurs at the end of ventricular contraction, manifested as a T wave. In this case, the current value I of each amplitude is approximately 0.1mV, which is green and represents an overall positive current value.

[0093] In this example, the electrocardiogram signal comes from the lead I electrocardiogram of two cardiac activities under sinus rhythm, and normal atrioventricular conduction (AV conduction) can be observed.

[0094] Figure 2 right Figure 1Two key features of cardiac activity were automatically detected or determined. Based on these features, specific intervals and amplitude variations in each cardiac cycle could be determined; by comparing them with predetermined thresholds, trigger signals could be determined or adjusted as needed.

[0095] In this example, the onset and end of the P wave, the position of the R wave, the onset and end of the QRS complex, and the onset and end of the T wave were automatically detected. The detection process was particularly facilitated by employing appropriate bandpass filters to process the ECG signal: a 10Hz to 25Hz bandpass filter for the QRS complex, a 5Hz to 20Hz bandpass filter for the P wave, and a 1Hz to 5Hz bandpass filter for the T wave. Based on the expected frequency range and (if applicable) the expected maximum heart rate and / or the expected duration of each ECG feature or amplitude variation, corresponding features in the ECG signal can be advantageously and automatically detected through appropriately configured feature extraction.

[0096] Intervals, such as the PR interval, QRS interval, and RT interval or QT interval, can be determined using a labeled cardiac cycle. In this example, the PR interval is 170 ms to 185 ms, the QRS interval is 95 ms to 110 ms, and the QT interval or RT interval is approximately 385 ms to 435 ms. The intervals of each cardiac cycle can vary from person to person and may change with heart rate.

[0097] Based on a defined interval and comparison with an appropriate predetermined threshold, the detection of the trigger signal and / or P wave, R wave, and / or T wave can be adjusted when the corresponding threshold is exceeded. In any case, the trigger signal can be determined with higher reliability, thereby significantly improving the triggering stability of cardiopulmonary bypass support.

[0098] If the predetermined threshold is exceeded, there may be signs of arrhythmia (such as atrioventricular block or bundle branch block). As mentioned earlier, this can lead to conduction failure: for a given cardiac cycle, ventricular depolarization may not occur or may not be adequately detected.

[0099] For example, low-amplitude R waves (represented by lowercase "r") may occur, or no R waves may be formed at all; double R waves (low amplitude) and / or low-amplitude Q waves and / or low-amplitude S waves may also occur.

[0100] Examples of bundle branch block include (incomplete) right bundle branch block or left bundle branch block, which may be intermittent or continuous. Atrioventricular block is particularly characterized by double R waves in the QRS complex (e.g., an M-shaped Rs-R' curve). This waveform may depend on the corresponding ECG lead; for example, in the case of left bundle branch block, it may form in lead I, V5, and / or V6. In leads I, aVI, V5, and / or V6, the R wave may have a positive amplitude; while in leads III, aVR, aVF, V1, V2, and / or V3, the R wave may have a negative amplitude. In cases of right bundle branch block, such as in leads I, aVL, V5, and / or V6, the S wave may be deeper and / or wider; in leads III, aVL, V1, and / or V2, the R wave may be higher and wider, and splitting may be particularly likely.

[0101] Based on the waveform of the electrocardiogram signal and / or the amplitude variations within the QRS complex, the type of arrhythmia can be determined and taken into account when identifying the trigger signal. This process is preferably performed in conjunction with consideration of at least one P wave and / or T wave.

[0102] In addition to bundle branch block or conduction system block, left anterior or left posterior fascicular block may also occur; bifascicular block may also be present (e.g., right bundle branch block with corresponding fascicular block). As previously mentioned, these pathophysiological conditions can be diagnosed by P waves, T waves, and / or QRS complexes, and should be considered when identifying trigger signals. Furthermore, ventricular and / or supraventricular conduction disturbances (bradycardia or tachycardia) may also be present.

[0103] Figure 3 Multiple electrocardiogram (ECG) leads are shown, namely the ECG curves of lead III, lead V1, and lead V6 of the surface ECG, and the right ventricular ECG RV34 is also shown below; the markings above define the time intervals of 1000ms.

[0104] This ECG curve corresponds to the pathological curve of Wolf-Parkinson-White syndrome (WPW syndrome), exhibiting additional anterograde conduction via the Kent pathway. It is evident that in this case, the additional anterograde conduction occurs intermittently (i.e., during the first and third cardiac events). Therefore, the additional premature conduction between the atria and ventricles leads to a shortened PQ interval; premature ventricular contractions can still be observed during the second cardiac event. To accurately detect the R wave, the corresponding PQ or PR interval can be shortened, or the relevant thresholds can be adjusted, so that stable triggering based on the detected R wave can be achieved even under this pathological curve.

[0105] Figure 4The ECG curve for another cardiac pathological condition (i.e., atrioventricular reentrant tachycardia with retrograde Kent pathway (AVRT)) is shown. Figure 3 Similarly, narrow QRS complexes are also present in this example. The ECG curve was recorded in the absence of right and left bundle branch block, targeting leads I, II, III, V1, and V6 of the surface ECG. A right atrial ECG HRA34 is also shown below. Figure 3 Similarly, the marker above also defines a 1000ms time interval.

[0106] In cases of intermittent right bundle branch block, the electrocardiogram (ECG) curve will change accordingly, such as... Figure 5 As shown in the figure, this diagram illustrates the widening of the QRS complex during right bundle branch block, with the most pronounced effect in leads I and V1. Therefore, identifying this QRS widening can not only confirm or validate the pathological curve but also adjust the prognostic period to support or facilitate the detection of the future R wave.

[0107] for Figure 1 and Figure 2 The right bundle branch block shown can be used to analyze the QRS interval or frequency range in the electrocardiogram.

[0108] Figure 6 The diagram schematically illustrates a device 10 for monitoring time-triggered stability according to the present invention. In this embodiment, device 10 is designed as an electrocardiogram (ECG) module. ECG signals can be received via interface 12, and signals can be transmitted to other devices or equipment without special coupling. Therefore, this interface can achieve communication coupling with an extracorporeal circulation support system or extracorporeal circulation support device; it can also be designed together with the interface of a control unit.

[0109] Device 10 may optionally be designed as an electrocardiogram (ECG) device, which may be integrated into or connected to a control unit to provide a control unit for the extracorporeal circulation support system. This control unit can be used independently of the provision of other components and has a compact design. Preferably, device 10 may be integrated into a single housing of the extracorporeal circulation support system (e.g., integrated into a sensor housing).

[0110] The ECG signal 14 of the patient can be received via interface 12; the corresponding data can be temporarily stored in the working memory and optionally stored in the storage medium 18 of the device 10. The ECG signal 14 is forwarded from interface 12 or via the working memory to the evaluation unit 16; the evaluation unit 16 is also communicatively connected to the storage medium 18 and configured to receive or read at least one pathological ECG feature 20 (which includes the evaluated ECG signal or corresponding threshold or parameter values ​​for each pathology) and compare the ECG signal 14 with at least one ECG feature 20.

[0111] ECG feature 20 can be selected based on optimal superposition (e.g., the statistical maximum superposition of the average PR interval, QRS interval, and / or RT interval of multiple ECG signals 14). This can serve as the basis for determining trigger signal 22 and / or detecting P wave, R wave, and / or T wave in subsequent cardiac cycles. In other words, appropriate thresholds and / or frequency ranges of the intervals from pathological ECG feature 20 can be used as the basis for further processing and use of ECG signal 14. Specifically, the selected pathological ECG feature can be used to determine or select trigger signal 22 from ECG signal 14 in subsequent cardiac cycles, which is used to control and synchronize the blood pump of the extracorporeal circulation support system. For this purpose, trigger signal 22 can be forwarded to interface 12 or provided at interface 12; for example, the control unit can retrieve or receive trigger signal 22 at interface 12.

[0112] If subsequent cardiac cycles exceed one or more thresholds, or fail to (routinely) adhere to the threshold range set based on pathological ECG feature 20, the evaluation unit 16 may again compare the data with the dataset stored "offline" in the storage medium 18 (i.e., with other pathological ECG features 20) to further optimize the process and improve trigger stability. This process may be performed periodically, but is preferably performed continuously to identify and consider any changes in the ECG signal 14 as quickly as possible.

[0113] The various features shown in the exemplary embodiments may be combined and / or replaced without departing from the scope of the invention.

[0114] List of reference numerals 10 devices 12 interfaces 14 ECG signals 16 assessment units 18 storage media 20 Pathological Electrocardiographic Features 22 trigger signal

Claims

1. A method for monitoring and optimizing the temporal stability of extracorporeal circulation support, comprising the following steps: - receiving an electrocardiogram signal of a supported patient over a predetermined time length; - determining P-waves, R-waves and T-waves from the received electrocardiogram signal; and - determining a predetermined trigger signal from the electrocardiogram signal of the current cardiac cycle in consideration of at least one detected P-wave and / or at least one detected T-wave.

2. The method of claim 1, wherein, The P-wave and / or the T-wave is / are detected for the current cardiac cycle and / or for at least one preceding cardiac cycle, preferably for at least or only the immediately preceding cardiac cycle.

3. The method of claim 2, wherein, The P-wave is detected for at least the current cardiac cycle and the T-wave is detected for at least the immediately preceding cardiac cycle.

4. The method of claim 1, 2, or 3, wherein, The trigger signal is determined based on the P-R interval between the P-wave and the R-wave in the respective cardiac cycle.

5. The method of claim 4, wherein, The determination of the predetermined trigger signal and / or the predetermined trigger signal to be determined is adjusted when a threshold value predetermined for the P-R interval is exceeded.

6. The method of claim 5, wherein, The predetermined trigger signal to be determined is adjusted from the R-wave to the P-wave or from the P-wave to the R-wave as the trigger signal.

7. The method of any one of claims 4 to 6, wherein, The detection of the P-wave, the R-wave and / or the T-wave of the current cardiac cycle is adjusted if for at least one preceding respective cardiac cycle a threshold value predetermined for the P-R interval is exceeded.

8. The method of any one of claims 5-7, wherein, The threshold value for the P-R interval is at least 150 ms.

9. The method of claim 8, wherein, The threshold value for the P-R interval is between 150 ms and 400 ms, preferably between 200 ms and 300 ms.

10. The method of any one of claims 5 to 9, wherein, The threshold value for the P-R interval corresponds to at least 20%, preferably at least 30% or at least 35% of the duration of the cardiac cycle.

11. The method of any one of claims 4 to 10, wherein, The P-R intervals of at least two cardiac cycles are averaged.

12. The method of any one of claims 4 to 11, wherein, The P-R intervals of at least two consecutive cardiac cycles are compared, and the determination of the predetermined trigger signal and / or the predetermined trigger signal to be determined is adjusted when the P-R intervals are consecutively prolonged.

13. The method of any of the preceding claims, wherein, The trigger signal is determined based on the R-T interval between the R-wave and the T-wave in the respective cardiac cycle.

14. The method of claim 13, wherein, The determination of the predetermined trigger signal and / or the predetermined trigger signal to be determined is adjusted when a threshold value predetermined for the R-T interval is exceeded.

15. The method of claim 13 or 14, wherein, The detection of the P-wave, the R-wave and / or the T-wave of the current cardiac cycle is adjusted if for at least one preceding respective cardiac cycle a threshold value predetermined for the R-T interval is exceeded.

16. The method of claim 14 or 15, wherein, The threshold value for the R-T interval is at least 350 ms.

17. The method of claim 16, wherein, The threshold value for the R-T interval is at least 450 ms.

18. The method of any one of claims 14 to 17, wherein, The threshold value for the R-T interval corresponds to at least 35%, preferably at least 40% or at least 45% of the duration of the cardiac cycle.

19. The method of any one of claims 13 to 18, wherein, The R-T intervals of at least two cardiac cycles are averaged.

20. The method of any of the preceding claims, wherein, The determination of the predetermined trigger signal, the predetermined trigger signal to be determined and / or the detection of the P-wave, the R-wave and / or the T-wave of the current cardiac cycle is adjusted if the T-wave amplitude of at least one preceding cardiac cycle comprises a negative value or is negative in total.

21. The method of any of the preceding claims, wherein, The trigger signal is determined based on the QRS interval between the Q-wave and the S-wave in the respective cardiac cycle, wherein the determination of the predetermined trigger signal and / or the predetermined trigger signal to be determined is adjusted when a threshold value predetermined for the QRS interval is exceeded; and / or wherein the detection of the P-wave, the R-wave and / or the T-wave of the current cardiac cycle is adjusted if for at least one preceding respective cardiac cycle a threshold value predetermined for the QRS interval is exceeded.

22. The method of claim 21, wherein, The threshold value for the QRS interval is at least 120 ms.

23. The method of any of the preceding claims, wherein, The predetermined trigger signal and / or the threshold value for one or more intervals within the respective cardiac cycle is based on a pathologic electrocardiogram feature.

24. The method of claim 23, wherein, The pathologic electrocardiogram feature is determined and / or predetermined for the current cardiac cycle based on intervals recorded in a previous cardiac cycle.

25. The method of claim 24, wherein, The intervals detected from the previous cardiac cycle are compared to a data set evaluated "offline", which contains electrocardiogram signals of supported patients with a corresponding pathology, and wherein the electrocardiogram feature is predetermined based on the best match to the electrocardiogram signals in the data set.

26. The method of claim 25, wherein, For each known pathology and the respective interval, a range of intervals, a range of frequencies and / or a range of amplitudes and / or a trigger signal to be determined is predetermined.

27. The method of any one of the preceding claims, wherein, The P-wave, the R-wave and / or the T-wave are detected by means of a bandpass filter, wherein the frequency range of the P-wave is preferably 5 Hz to 20 Hz, the frequency range of the R-wave is preferably 10 Hz to 25 Hz, and / or the frequency range of the T-wave is preferably 1 Hz to 5 Hz.

28. The method of any of the preceding claims, wherein, The predetermined trigger signal is the R-wave, and by taking into account the at least one detected P-wave and / or the at least one detected T-wave, a P-wave or a T-wave is selected as the trigger signal to be determined.

29. The method of any one of the preceding claims, wherein, Based on the determined trigger signal, a signal for extracorporeal circulation support is outputted with a delay time, which is predetermined by taking into account the at least one detected P-wave and / or the at least one detected T-wave.

30. The method of claim 29, wherein, The delay time for the current cardiac cycle is determined based on a QT time detected from at least one previous cardiac cycle, which is normalized and / or taken into account based on a heart rate determined from the electrocardiogram signal.

31. A device (10) for monitoring and optimizing the temporal trigger stability of extracorporeal circulation support, comprising: an interface (12) for receiving an electrocardiogram signal (14) of a supported patient over a predetermined length of time; and an evaluation unit (16) configured to determine P-waves, R-waves and T-waves from the received electrocardiogram signal (14), wherein the evaluation unit (16) is further configured to determine a predetermined trigger signal (22) from the electrocardiogram signal (14) of the current cardiac cycle taking into account the at least one detected P-wave and / or the at least one detected T-wave.

32. The apparatus (10) of claim 31, wherein The device is configured to perform the method according to any one of claims 2 to 30.

Images