LAP signal processing for automatic calculation of A / V ratio.

Abstract

Systems and methods are described herein for cardiac monitoring and therapy using left atrial pressure. In many embodiments, a pressure sensor implantable in the left atrium of a patient's heart records a pressure signal. The left atrial pressure can be used to calculate an A / V ratio that has clinical significance, which can then be used to guide therapy and / or as part of an open-loop or closed-loop automated therapy system.

Description

[Technical field]

[0001] (CROSS REFERENCE TO RELATED APPLICATIONS) This application is filed under 35 U.S.C. §119(e) with U.S. Provisional Patent Application No. 63 / 287,003, filed December 7, 2021, entitled “LAP SIGNAL PROCESSING TO AUTOMATICALLY CALCULATE A / V RATIO”, U.S. Provisional Patent Application No. 63 / 287,009, filed December 7, 2021, entitled “EXERTION RESPONSE DETECTION BY AN IMPLANTABLE SENSOR”, U.S. Provisional Patent Application No. 63 / 287,016, filed December 7, 2021, entitled “IMPROVED SIGNAL PROCESSING OF BLOOD PRESSURE FROM AN IMPLANTABLE SENSOR BY PRE-CLASSIFICATION OF PATIENTS”, and U.S. Provisional Patent Application No. 63 / 287,016, filed December 17, 2021, entitled “SYSTEMS AND METHODS FOR HEART This application claims the benefit of and priority to U.S. Provisional Patent Application No. 63 / 291,253, entitled "SYSTEMS AND METHODS FOR PUMONARY CONGESTION MONITORING AND TREATMENT," filed on December 17, 2021, and U.S. Provisional Patent Application No. 63 / 291,270, entitled "SYSTEMS AND METHODS FOR PUMONARY CONGESTION MONITORING AND TREATMENT," filed on December 17, 2021. The disclosures of U.S. Provisional Patent Application Nos. 63 / 287,003, 63 / 287,009, 63 / 287,016, 63 / 291,253, and 63 / 291,270 are incorporated by reference in their entireties herein for all purposes.

[0002] The present invention relates generally to the use of cardiac signals to monitor the health of a patient, and more particularly to the use of left atrial pressure signals to predict and treat cardiac conditions. [Background technology]

[0003] The human heart has four chambers: the left and right atria and the left and right ventricles. The atria receive blood from the veins and the ventricles eject blood into the arteries. The right-sided structures receive deoxygenated blood from the body and pump it to the lungs where it is oxygenated. The left-sided structures receive freshly oxygenated blood from the lungs and pump it to the rest of the body. The heart also contains four major valves: the mitral (bicuspid), tricuspid, aortic, and pulmonary. Prosthetic heart valves are life-saving medical devices that can be implanted in the heart to replicate the function of the natural heart valves when cardiac function is lost. Summary of the Invention

[0004] Systems and methods are illustrated for LAP signal processing according to aspects of the invention. One aspect includes a method of treating a cardiac condition in a patient, comprising: acquiring a left atrial pressure (LAP) signal using an atrial cardiac sensor implanted in the patient's left atrium, identifying a set of A waves and a set of V waves in the LAP signal, calculating an A / V ratio based on an average pressure amplitude of the set of A waves and an average pressure amplitude of the set of V waves, determining an appropriate therapy for the patient based on the A / V ratio, and providing the appropriate therapy to the patient. [Brief description of the drawings]

[0005] [Figure 1] FIG. 1 is an atrial implant cardiac monitoring system according to one embodiment of the present invention. [Diagram 2] FIG. 2 is a block diagram of a cardiac signal processor according to an embodiment of the present invention. [Figure 3A] FIG. 3A illustrates exemplary A and V waves, according to an embodiment of the present invention. [Figure 3B] FIG. 3B is a diagram illustrating key points on exemplary A and V waves, according to an embodiment of the present invention. [Figure 4] FIG. 4 is a flow chart illustrating a process for calculating A / V according to an embodiment of the present invention. [Diagram 5]FIG. 5 is an exemplary left atrial pressure (LAP) signal, in accordance with an embodiment of the present invention. [Figure 6] FIG. 6 is an exemplary spectral analysis of a typical LAP signal, in accordance with an embodiment of the present invention. [Figure 7A] FIG. 7A shows LAP signals reflecting large, slow breathing and shallow, fast breathing, respectively, in accordance with an embodiment of the present invention. [Figure 7B] FIG. 7B shows LAP signals reflecting large, slow breathing and shallow, fast breathing, respectively, in accordance with an embodiment of the present invention. [Figure 8] FIG. 8 is a denoised LAP signal according to one embodiment of the present invention. [Figure 9] FIG. 9 is an analyzed LAP signal according to an embodiment of the present invention. [Figure 10] FIG. 10 illustrates the relationship between the LAP signal and the electrocardiogram (EKG) signal, according to an embodiment of the present invention. [Figure 11] FIG. 11 shows a labeled and analyzed LAP signal according to an embodiment of the present invention. [Figure 12A] FIG. 12A illustrates a baseline V wave and an elevated V wave, respectively, according to an embodiment of the present invention. [Figure 12B] FIG. 12B illustrates a baseline V wave and an elevated V wave, respectively, according to an embodiment of the present invention. [Figure 13A] FIG. 13A shows a single AV cycle and a longer time scale LAP signal, respectively, of a patient in atrial fibrillation, in accordance with an embodiment of the present invention. [Figure 13B] FIG. 13B shows a single AV cycle and a longer time scale LAP signal, respectively, of a patient in atrial fibrillation, in accordance with an embodiment of the present invention. [Figure 14] FIG. 14 is a flow chart illustrating a process for determining cardiac variability across different exercise states, according to an embodiment of the present invention. [Figure 15] FIG. 15 illustrates a diagram showing change in LAP as a function of change in heart rate, in accordance with an embodiment of the present invention. [Figure 16]FIG. 16 is a diagram illustrating change in respiration rate as a function of change in heart rate, in accordance with an embodiment of the present invention. [Figure 17] FIG. 17 illustrates the change in A / V ratio as a function of change in heart rate, in accordance with an embodiment of the present invention. [Figure 18] FIG. 18 is a flow chart illustrating a process for dynamically presenting clinical data based on relevance to a particular patient, according to an embodiment of the present invention. [Figure 19] FIG. 19 illustrates the two main types of structural changes during heart failure, according to an embodiment of the present invention. [Figure 20] FIG. 20 illustrates a LAP signal showing enlarged V-wave pressure resulting from mitral regurgitation, in accordance with an embodiment of the present invention. [Figure 21] FIG. 21 illustrates mitral stenosis, according to an embodiment of the present invention. [Figure 22] FIG. 22 illustrates a LAP signal showing enlarged A-wave pressure resulting from mitral stenosis, in accordance with an embodiment of the present invention. [Diagram 23] FIG. 23 is a flow chart illustrating a process for calculating mitral valve index according to an embodiment of the present invention. [Figure 24] FIG. 24 is a diagram illustrating key points on the LAP signal for calculating the mitral valve index, according to an embodiment of the present invention. [Diagram 25] FIG. 25 is a diagram illustrating a minimum delta in a LAP signal, in accordance with an embodiment of the present invention. [Figure 26] FIG. 26 is a diagram illustrating a large minimum delta in a LAP signal in accordance with an embodiment of the present invention. [Figure 27] FIG. 27 is a diagram illustrating a negative minimum delta in a LAP signal, in accordance with an embodiment of the present invention. [Figure 28] FIG. 28 is a first process for calculating a pulmonary congestion risk index according to an embodiment of the present invention. [Figure 29A] FIG. 29A is a chart that diagrammatically illustrates a first process on an arbitrary sample data set, according to an embodiment of the present invention. [Figure 29B]FIG. 29B is a chart that diagrammatically illustrates the first process on an arbitrary sample data set, according to an embodiment of the present invention. [Diagram 30] FIG. 30 is a second process for calculating a Pulmonary Congestion Risk Index according to an embodiment of the present invention. [Figure 31A] FIG. 31A is a chart that diagrammatically illustrates a second process on an arbitrary sample data set, according to an embodiment of the present invention. [Figure 31B] FIG. 31B is a chart that diagrammatically illustrates a second process on an arbitrary sample data set, according to an embodiment of the present invention. [Diagram 32] FIG. 32 is a third process for calculating a pulmonary congestion risk index according to an embodiment of the present invention. [Figure 33A] FIG. 33A is a chart that diagrammatically illustrates a third process on an arbitrary sample data set, according to an embodiment of the present invention. [Figure 33B] FIG. 33B is a chart that diagrammatically illustrates a third process on an arbitrary sample data set, according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0006] Despite significant advances in medical care, cardiovascular disease is the leading cause of death worldwide. If underlying problems are detected early, patient outcomes can be significantly more positive. Cardiac indices such as heart rate and respiratory rate are common measurements used for diagnosis and monitoring. Often, a patient's condition can be better understood in the context of the difference in cardiac function between resting and exercising states. However, cardiac function is most often measured when the patient is at rest, e.g., sitting in an examination room, lying on an examination table, etc. Measuring a patient during an exercising state (e.g., moving around, walking fast, climbing stairs, etc.) can be difficult and / or cumbersome, as the monitoring equipment must be able to move with the patient or a static, exercising environment with measurement capabilities must be provided.

[0007] Continuous (or even voluntary) monitoring of these indicators at any time throughout the day can be difficult without the immediate assistance of a medical professional. Although some modern wearable devices (e.g., smart watches) have biometric measurement capabilities, they may be unreliable in situations where a high degree of accuracy is required. Furthermore, conventional wearable devices cannot record certain pressure measurements, such as (but not limited to) left atrial pressure (LAP), which may be used for cardiac monitoring and treatment, as discussed herein.

[0008] To address these issues, implantable medical devices with recording capabilities may enable consistent and frequent measurements of biological functions. An implantable atrial implant cardiac monitoring system according to various aspects of the present invention allows new and more accurate measurements (including LAP measurements) to be obtained. To take advantage of these new data, the systems and methods described herein can record and calculate cardiac indices that have clinical relevance for both early detection of cardiac disease, as well as other disease states and treatment directions. In many aspects, derived cardiac indices such as (but not limited to) heart rate, respiratory rate, mean blood pressure, A / V ratio, etc., are highly accurate and can be calculated over many different exercise states and / or over long periods of time to obtain an accurate picture of the difference between exercise and resting states. Changes in the ratios between various cardiac indices in different states can be used to direct diagnosis and treatment.

[0009] In many embodiments, the atrial implanted cardiac monitoring system can measure LAP. LAP is a measurement that is traditionally estimated to be difficult to measure directly (e.g., via transseptal left heart catheterization) and has significant associated risks. LAP can be estimated in many ways, including during Swan-Ganz catheter insertion, where balloon flotation technology is used to measure pulmonary capillary wedge pressure. LAP can also be estimated by measuring left ventricular end-diastolic pressure. However, these estimates may be inaccurate and may be insufficient for precise, personalized medicine.

[0010] In some embodiments, the implanted cardiac monitoring system allows for the optional direct recording of LAP as a continuous signal, at a high sampling rate. However, only measuring LAP is insufficient for high-quality cardiac monitoring. As discussed herein, the LAP signal can be processed to extract clinically significant information about the patient instantaneously, and can be used to determine a treatment protocol for medical staff. Of particular note, the A / V ratio described herein has particular clinical relevance. For example, the LAP signal itself may not change significantly during atrial fibrillation (AF), but the derived A / V ratio may indicate changes in disease state. In particular, the A / V ratio can be a useful indicator when dynamics are informative. In contrast to (or as well as) absolute cardiac pressure values, e.g., when measuring changes in cardiac function between resting and exercise states, in many embodiments, changes in A / V ratios can be used to identify and subsequently treat any number of conditions, including (but not limited to) heart failure (HF), AF, hypertension, pulmonary hypertension, mitral regurgitation, mitral stenosis, aortic stenosis, cardiac rhythm disorders (e.g., hyperventricular tachycardia, ventricular tachycardia, etc.), etc. As can be readily appreciated, this list of conditions is not exhaustive, and any number of different conditions in which A / V ratios are clinically relevant can be analyzed, depending on the requirements of a particular application of an embodiment of the invention.

[0011] Not only are there many different types of cardiovascular disease, but there are also many different types of patients. Since each person is unique to at least some degree, personalized treatment plans can significantly improve patient outcomes. To rapidly create personalized treatment plans, patients can be pre-classified into at least one unique class that attempts to group highly similar patients, and information can be rapidly presented to medical professionals based on that classification. Furthermore, depending on the particular classification of a given patient, various cardiac indicators may be analyzed differently. For example, normal ranges may be different for different classifications. As a further example, certain metrics may be calculated in different ways, and A / V ratios may be calculated and / or presented differently when a particular patient class is unlikely to exhibit defined A and / or V waves.

[0012] The LAP can also be used to calculate heart rate and respiratory rate with high accuracy using the systems and methods described herein. The process of calculating the A / V ratio is disclosed in detail below. In conjunction with patient information, the LAP, as well as other cardiac signals and / or indices, can be used by the systems and methods described herein to monitor patient health and develop personalized treatment plans. Cardiac monitoring systems are described below.

[0013] Atrial Heart Monitoring System At a high level, a cardiac monitoring system includes an implanted sensor capable of generating a LAP signal, and a processing device capable of extracting clinically relevant information from the LAP signal. Systems including implanted sensors are primarily discussed below, but as can be readily appreciated, the processes described herein may also be implemented using a LAP signal, or an estimate thereof, recorded using an external or temporarily implanted measurement device.

[0014] Referring now to FIG. 1, an atrial cardiac monitoring system is illustrated according to one embodiment of the present invention. The system 100 includes an atrial cardiac sensor 110 implanted in the left atrium of a patient. In many embodiments, the atrial cardiac sensor is a sensor that has the ability to directly measure cardiac function and transmit the recorded data outside the body. In various embodiments, the atrial cardiac sensor can record a LAP signal inside the left atrium. In many embodiments, the atrial cardiac sensor can also measure other different cardiac functions, including (but not limited to) blood pressure, blood flow velocity, heart rate, flow direction, turbulence, and / or any other cardiac metrics appropriate to the requirements of a particular application of an embodiment of the present invention. The recording of these metrics over time is referred to herein as a "cardiac signal."

[0015] In many embodiments, the atrial cardiac sensor is capable of recording at least multiple cardiac cycles (typically at least 10-20 seconds) at a high sampling rate sufficient to capture the pressure waveform (typically about 100 Hz). In many embodiments, the atrial cardiac sensor does not record continuously, but only when powered via a wireless power source. Thus, data recorded in a single recording session may typically record only 3-25 cardiac cycles. In many embodiments, the recording session lasts for a predetermined period of time, e.g., 10-60 seconds. Although many more cardiac cycles may be recorded via continuous power supply, clinically relevant data may be extracted from data describing only this small time frame.

[0016] The atrial cardiac sensor 110 can transmit the recorded cardiac signal to the transceiver 120. In many aspects, the transceiver obtains the cardiac signal from the atrial cardiac sensor using any of a number of different wireless communication methods. In various aspects, the transceiver may control and / or power the operation of the atrial cardiac sensor. For example, the transceiver can remotely power the atrial cardiac sensor using wireless power transmission (e.g., radio frequency induction). In many aspects, powering the atrial cardiac sensor triggers the recording and transmission process. In some aspects, the atrial cardiac sensor has on-board memory that stores recorded data that can later be transmitted to the transceiver. In many aspects, the transceiver can instruct the atrial cardiac sensor to make measurements over a predetermined period of time. In various aspects, the transceiver can instruct the atrial cardiac sensor to automatically record a predetermined number of times per day at a predetermined interval. In various aspects, the transceiver can obtain external heart rate measurements via any of many different types of wearable heart rate monitoring devices and trigger the atrial cardiac sensor to take a recording in response to an increase in heart rate. The transceiver may include a memory used to store cardiac signal data received from the atrial cardiac sensor for later transmission.

[0017] The transceiver may further transmit data acquired from the atrial cardiac sensor to one or more cardiac signal processors 130. In many embodiments, the data is transmitted over a network, such as, but not limited to, the Internet. However, any number of different networks, wireless connections, wired connections, or any combination thereof may be used to transmit the data. In various embodiments, the transceiver is implemented as a smartphone. In many embodiments, the transceiver is a wearable device, such as, but not limited to, a smart watch, a chest worn device, and / or any other wearable suitable for the requirements of a particular application of an embodiment of the present invention. As can be readily appreciated, many different devices can be used as the transceiver, so long as they have the appropriate communication and processing capabilities.

[0018] The cardiac signal processor is a computing device that can process measurements obtained by the atrial cardiac sensor and extract clinically relevant indices. In many aspects, the cardiac signal processor can provide an individualized treatment plan based on the clinically relevant indices and patient information. The cardiac signal processor can be implemented on any number of different computing platforms, such as, but not limited to, a personal computer (132), a server system (134), a smartphone (136), and / or any other computing platform having sufficient computing power appropriate to the requirements of a particular application of aspects of the present invention. Furthermore, in many aspects, the transceiver and the cardiac signal processor may be implemented as part of the same hardware platform.

[0019] In many aspects, the transceiver and / or cardiac signal processor may further identify when the atrial cardiac sensor requires replacement. In many aspects, unexpected changes in the received signal may be used to determine the need for replacement. For example, degradation of signal quality, shifts in pressure readings over time not evidenced by other measurements, and / or any other changes in the signal may be used as appropriate to the requirements of a particular application of aspects of the present invention. In various aspects, the expected life of the atrial cardiac sensor may be used as part of the determination. In various aspects, the atrial cardiac sensor may transmit a signal indicating that replacement is required. In many aspects, the atrial cardiac sensor may be incorporated into a prosthetic heart valve and may then indicate when the prosthetic heart valve requires replacement.

[0020] As can be readily appreciated, a particular architecture for an atrial heart monitoring system is shown in Figure 1, however, any number of different architectures (e.g., those utilizing a transceiver and / or a cardiac signal processor) may be utilized without departing from the scope or spirit of the present invention. The cardiac signal processor is described in more detail with respect to Figure 2 below.

[0021] 2, a cardiac signal processor according to an embodiment of the present invention is illustrated. Cardiac signal processor 200 includes a processor 210. In many embodiments, the processor is any logical processing circuit capable of executing the cardiac monitoring processes described herein. The processor may be implemented using any (or combination and / or plurality) of a central processing unit, a graphic processing unit, a field programmable gate array, an application specific integrated circuit, and / or any other logical processing circuit as appropriate to the requirements of a particular application of an embodiment of the present invention. Cardiac signal processor 200 further includes an input / output (I / O) interface 220. The I / O interface may be any circuit or group of circuits capable of enabling communication with a transceiver and / or other computing devices. In various embodiments, a number of different communication methods may be implemented using the I / O interface as appropriate to the requirements of a particular application of an embodiment of the present invention.

[0022] The cardiac signal processor 200 further includes a memory 230. The memory 230 may be implemented using volatile memory, non-volatile memory, and / or any combination thereof. The memory 230 includes a cardiac monitoring application 232 that may direct the processor to perform various cardiac monitoring processes, such as those described herein. In many aspects, the memory 230 may include cardiac signal data describing cardiac signals acquired from an atrial cardiac sensor over time, and patient data describing the patient from whom the atrial cardiac sensor data was acquired. As can be readily appreciated, a particular architecture for the cardiac signal processor is shown in FIG. 2, although any number of different architectures may be utilized without departing from the scope or spirit of the present invention. The cardiac monitoring processes are described in further detail below.

[0023] Atrial Cardiac Monitoring and Therapy Process Atrial cardiac monitoring and therapy processes according to many embodiments of the present invention involve the utilization of cardiac signals that generate clinically actionable data. A cardiac signal of particular importance for many applications is the LAP signal, which can then be used to generate an A / V ratio with significant clinical relevance. The A / V ratio can be used alone or in conjunction with other cardiac signals and / or cardiac indices to determine an individualized treatment plan for a given patient. For example, in many embodiments, if the A / V ratio falls below 0.6, it can indicate that the diuretic should be increased by 50%. As a further example, if the mean pressure increases above 40 mmHg without a change in the A / V ratio, the diuretic should not be increased, but instead the hypertension medication should be increased. However, any number of different treatment protocols can be used depending on the patient's condition as appropriate to the requirements of a particular application of an embodiment of the present invention. Additionally, the A / V ratio can be used as part of an early warning system for the health of the patient or for closed / open loop control of drug delivery. For example, changes in the cardiac indices described herein can be used to automatically trigger appropriate drug delivery, for example, via an infusion pump or other automated delivery method. Although the LAP and derived A / V ratio are not the only cardiac signals that are imported, the process of calculating the A / V ratio is discussed below as an initial foundation.

[0024] Calculate the AA / V ratio The left atrial pressure over the duration of one cardiac cycle can be decomposed into two main features: the A wave and the V wave. The A wave reflects the pressure rise due to the contraction of the atria (sometimes called presystole or systole). The V wave reflects the pressure rise due to the surge of blood into the atrium after it is pushed out of the right ventricle during systole (sometimes called systole or ventricular systole). Referring now to FIG. 3A, an exemplary LAP signal is illustrated over a single cardiac cycle, labeled with A-wave and V-wave components, according to an embodiment of the present invention. Additionally, five important stages can be mapped onto the LAP signal, which are shown in FIG. 3B according to an embodiment of the present invention. As mentioned above, the A wave indicates right atrial contraction at the end of diastole. The C wave indicates the expansion of the mitral valve leaflets into the left atrium during early systole. The X descent indicates left atrial relaxation in mid systole. The V wave again indicates the rapid filling of the right atrium in late systole. Finally, Y descent indicates the rapid emptying of the left atrium into the left ventricle during early diastole into left ventricular filling.

[0025] The A / V ratio is the ratio between the A wave mean amplitude and the V wave mean amplitude. It is a very sensitive parameter because changes in the A / V ratio are faster and more pronounced than changes in the overall mean pressure in response to the onset of congestion or edema. At first glance, the A / V ratio may appear simple to calculate, but due to certain perturbations introduced by the living human body, extraction is in fact a non-trivial task. Referring now to FIG. 4, a process for calculating the A / V ratio is illustrated, according to an embodiment of the present invention.

[0026] The process 400 includes acquiring (410) a LAP signal. In many embodiments, the LAP signal is acquired from an atrial cardiac sensor as described above. Although the LAP signal for one cardiac cycle is discussed with respect to FIGS. 3A and B, a LAP signal over many cardiac cycles is significantly more complex. In real-world scenarios, the LAP signal is not always easy to interpret because there are variations at different LAP stages. There may also be variations due to pressure generated by breathing, which typically contributes to the LAP in the lower frequency band. An example of a LAP signal according to an embodiment of the present invention is illustrated in FIG. 5.

[0027] The LAP signal is decomposed (420) as part of the spectral analysis to obtain cardiac indices. In many embodiments, the LAP signal may be decomposed into frequency components using a Fast Fourier Transform (FFT). In various embodiments, the LAP signal may be decomposed using wavelet analysis. The decomposed LAP signal provides the respiration frequency (typically 0.1-0.5 Hz), the heart rate frequency (typically 0.8-1.7 Hz at rest), and the dominant frequency of the A / V complex (near 2 Hz, roughly twice the heart rate). Small peaks at higher frequencies are typically harmonics of the basal respiration and cardiac pressure frequencies. An exemplary spectral analysis of a LAP signal according to an embodiment of the invention is shown in FIG. 6. As can be readily appreciated, other spectral analysis methodologies capable of separating respiration rate from heart rate can be used depending on the requirements of a particular application of an embodiment of the invention. In various embodiments, if the respiration rate cannot be separated from the heart rate, it can be an indication that the patient may be holding their respiration rate.

[0028] After detecting the respiration and heart rate frequencies, the respiration contribution can be subtracted 430 from the cardiac pressure component of the LAP signal by applying a low pass filter with a frequency cutoff between the respiration and heart rate frequencies. Note that respiration rate and magnitude may be derived from the signal and may be important to cardiac patients. Shallower and faster breathing may indicate shortness of breath, an important symptom for various cardiac conditions, such as pulmonary congestion and pulmonary edema. Exemplary LAP signals with respiratory components indicative of large slow breathing and shallow and fast breathing, respectively, according to an embodiment of the present invention, are shown in FIG. 7A and 7B. An exemplary noise-removed LAP signal with the respiratory contribution subtracted according to an embodiment of the present invention is shown in FIG. 8. Typically, heart rate starts at about 40-90 beats per minute (BPM) and may rise during exercise up to 220 BPM. A typical respiration rate starts at about 5-20 breaths per minute (RPM) and may rise during exercise up to 100 RPM. However, as can be readily appreciated, these numbers are merely representative of a typical population and may vary due to individual idiosyncrasies.

[0029] Once the noise has been removed, the LAP signal may be parsed (440) into even and odd waves. In many embodiments, the waveform is parsed into discrete waves at a frequency approximately twice the heart rate. In many embodiments, a local minimum in a short time window centered at twice the heart rate is selected as the division point. However, any number of different methodologies can be used to divide the waveform. For example, machine learning models can be used depending on the requirements of a particular application of an embodiment of the invention. An analyzed LAP signal according to an embodiment of the invention is shown in FIG. 9.

[0030] The even and odd waves can then be labeled 450 as A or V waves. A labeled version of the analyzed LAP signal according to an embodiment of the present invention is shown in FIG. 10. Since A and V waves alternate, it can be assumed that all odd waves belong to one group and all even waves belong to another group. A waves can often be identified as having a double peak (A and C waves), while V waves tend to have a single peak. The X descent is more typically flatter at the bottom than the Y descent, which tends to be more pointed. In many embodiments, a machine learning model may be trained and used to identify A and V waves in the analyzed LAP signal. For example, the machine learning model may be trained using a training data set made of LAP signals whose A and V waves have been annotated. However, other methodologies, such as (but not limited to) searching for a double peak in a single wave, can be utilized depending on the requirements of a particular application of an embodiment of the present invention. For example, in many embodiments, for each analyzed wave, if the slope graph of the wave has two zero crossings apart from the edges, an A wave can be assigned. If this is true for every other wave, a signature can be confirmed. It is important to note that for some patients, the A wave may not always be present. In various embodiments, failure to automatically detect the A wave may alert a medical professional for further investigation and / or treatment recommendations.

[0031] In various aspects, only a portion of the analyzed signal is processed to increase computational efficiency. In many aspects, the size of the portion can be increased until a confidence threshold is reached. The wave assignments may then be extrapolated across the entire analyzed LAP signal, even and odd waves. In various aspects, different portions across the analyzed signal can be processed separately to increase confidence in the extrapolation.

[0032] In various aspects, a machine learning model for identifying A and V waves in an analyzed LAP signal may be trained using a training data set that includes data collected from actual cardiac patients along with EKG data. The EKG data can be used to determine the V wave, which occurs later since the A wave typically occurs before the systole identified as the QRS complex on the EKG. This relationship is shown in FIG. 11 in accordance with an aspect of the present invention. The EKG data can be used to rapidly generate sufficiently reliable labeled analyzed LAP signal training data without requiring manual labeling of the training data.

[0033] Once the A and V waves have been identified, the average pressure amplitudes of the A and V waves are calculated 460. In many embodiments, to accomplish this, the average pressure and peak-to-peak pressure amplitudes of all A waves and all V waves are calculated. The A / V ratio may then be calculated 470 as the ratio between the A wave average amplitude and the V wave average amplitude.

[0034] As mentioned above, the A / V wave is a very sensitive parameter and therefore proper calculation is often important. To illustrate this, an analyzed LAP signal with a baseline V wave and an analyzed LAP signal with an elevated V wave are shown in Figs. 12A and B, respectively, according to an embodiment of the present invention. In the illustrated example, the overall mean pressure rises by only 20% from about 12 mmHg to about 14 mmHg, but the A / V ratio drops from about 1 to 0.4, a 60% change. A 2 mmHg change in LAP is generally considered not clinically significant, but in this case the A / V ratio may identify a pathological change not covered by the V wave. Although a particular process is described above with respect to the discussion of Figs. 3-12B, any of a variety of other processes for calculating the A / V ratio may be utilized as appropriate to the requirements of a particular application according to various embodiments of the present invention.

[0035] As a further example, an elevated V wave may be an indication of mitral regurgitation (MR). When the mitral valve leaks during systole, blood from the left ventricle returns to the left atrium at high pressure. The left ventricle may then generate a systolic pressure of 100-200 mmHg, which is much greater than the LAP of 10-20 mmHg. MR may then be considered as the primary cause of elevated LAP. Changes in MR may be an important monitoring factor in the treatment of cardiac conditions. For example, worsening of MR over time is an important indicator in the treatment of heart failure. MR may be caused by many factors, including fluid retention and increased left ventricular preload. The A / V ratio may act as a more sensitive and specific parameter for monitoring MR than the LAP mean pressure, and is therefore useful in the care, monitoring, and treatment of patients with cardiac conditions.

[0036] By way of yet further example, some pathological changes to the A-wave can also be detected as changes in the A / V ratio. The most significant changes in the A-wave, especially in heart failure patients, are typically caused by arrhythmias common to heart failure, such as atrial fibrillation or atrial flutter. In these conditions, the left atrium is fibrillating or fluttering instead of contracting. Thus, the left atrium does not generate pressure and the A-wave is corrupted. A corrupted A-wave in a cardiac cycle and an LAP signal over many cycles showing a corrupted A-wave according to an embodiment of the present invention are shown in Figs. 12A and B, respectively. Again, as can be seen, the overall effect on the average LAP is minimal, often masked by the elevated V-wave, but the A / V ratio is close to zero. In particular, heart failure patients may switch from normal sinus rhythm to atrial fibrillation to compensate. A system that monitors only the average LAP misses this important clinical indicator, which is revealed via the A / V ratio.

[0037] In contrast to collapsed A-waves, high A-waves (both alone and in A / V ratio) may indicate poor tissue compliance / tamponade, constrictive pericarditis, and / or cardiomyopathy. In many embodiments, a sudden increase in pressure (sudden upward and downward deflection) reflected in LAP and / or A / V ratio may be an indicator of pulmonary embolism and therefore may be an important indicator for patients with a highly hypercoagulable state. As can be easily understood from the above examples, the A / V ratio is a clinically significant indicator. In many embodiments, the A / V ratio may be calculated over time to monitor the health of a patient. Changes in the A / V ratio may be used to reveal specific pathologies that have traditionally gone unnoticed for much longer and direct treatment to those specific problems. As such, the A / V ratio is an important indicator to calculate in many scenarios. However, it is not necessary to calculate the A / V ratio in all embodiments of the present invention, and many other indicators may be calculated and used by the cardiac monitoring system described herein. For example, in many embodiments, the atrial cardiac monitoring system may monitor a patient in multiple sessions over time to create a dynamic health profile. In particular, the patient may be instructed to record data during different usage conditions to measure cardiovascular performance. The exercise monitoring process is described further below.

[0038] B. Exercise monitoring An atrial cardiac monitoring system can record measurements while the patient is mobile. Typically, blood pressure monitoring via an implantable sensor is performed while the patient is at rest, meaning sitting or lying down. Similar tests from non-implantable devices, such as (but not limited to) a pulmonary artery catheter, are also performed with the patient lying on an examination table. As a result, the obtained pressure signal is typically limited to resting conditions. Even if the patient is previously active, recordings are usually made during the resting state. Unfortunately, some changes to the pressure only occur when the patient is exercising on his or her own. Therefore, measuring changes during exercise states is important in cardiac patient care. As mentioned above, a cardiac monitoring system can include an atrial cardiac sensor that can record measurements during exercise states, and the data acquired during different exercise states can be processed to extract clinically relevant information.

[0039] 14, a process for motion monitoring is illustrated according to an embodiment of the present invention. Process 1400 includes acquiring 1410 a first cardiac signal recorded while the patient is at rest. In many embodiments, the first cardiac signal is a blood pressure signal. In various embodiments, the first cardiac signal is a LAP signal. A first set of cardiac indices is calculated 1420 from the first cardiac signal. In many embodiments, the first set of cardiac indices includes heart rate, mean blood pressure, respiratory rate, and A / V ratio (if the cardiac signal is a LAP signal). However, depending on the placement (and number) of implanted cardiac sensors, different cardiac signals and cardiac indices may be acquired and calculated, respectively.

[0040] The process 1400 further includes obtaining 1430 a second cardiac signal recorded while the patient is in an exercise state and calculating 1440 a second set of cardiac indices. In many embodiments, the changes between the first set of cardiac indices and the second set of cardiac indices may be plotted 1450 for review by a medical professional. Additionally, a ratio between the first and second sets of cardiac indices may be calculated 1460 to derive clinically relevant information.

[0041] For example, in many embodiments, the ratio between the average pressure change and the heart rate change (ΔP / ΔHR) may be calculated. An example plot of average LAP as a function of heart rate according to an embodiment of the invention is shown in FIG. 15. Additionally, in many embodiments, the ratio between the respiration rate change and the heart rate change (ΔRR / ΔHR) is calculated. An example plot of respiration rate as a function of heart rate according to an embodiment of the invention is shown in FIG. 16. In various embodiments, the ratio between the average pressure change and the heart rate change (A / V ratio exercise response) is calculated. An example plot of LAP A / V ratio as a function of heart rate according to an embodiment of the invention is shown in FIG. 17. Using these indicators (and others), it can be easily determined if the patient's response to exercise is worsening, which is a clear indicator of a worsening cardiac condition. This knowledge can be used to immediately initiate treatment (or otherwise medically intervene). In many embodiments, this process may be performed on a regular schedule (or as desired) to continually check the health of the heart.

[0042] LAP and / or A / V ratio can be used as an indicator of heart failure (or as a component thereof), tracking mean pressure variations over time. Tracking cardiac conditions can provide insight into multiple conditions, including (but not limited to) heart failure, coronary artery disease, ischemia, and many others. In various aspects, exercise monitoring can complement or replace stress testing, which is typically performed using a treadmill or pharmacological induction. Because implanted cardiac sensors can collect data at any time of the patient's day with minimal impact, tracking can occur at a more granular level than in the clinic. In the clinic, tests are typically performed during a single resting state and a single exercise state, which are then compared. Performing these tests can be a significant time commitment for the patient and may not accurately reflect normal conditions for several reasons, including (but not limited to) stress conditions due to the medical environment, discrepancies between pharmacologically induced and exercise induced exercise, and many other reasons.

[0043] In many embodiments, these data are collected and plotted on a curve. Heart rate can be used as an indicator of exercise state, and other cardiac indicators, such as LAP, A / V ratio, respiration rate, etc., can be plotted over time. Portions of the graph may be colored or otherwise annotated by exercise state for easy reading. In various embodiments, the curve for each indicator may be labeled with warnings for indicators that are outside of a healthy range, that constitute an unhealthy delta between resting and exercise states, or any other annotation that may be clinically useful. In various embodiments, the curves can be used to characterize cardiac conditions and / or quantify the severity of cardiac disease.

[0044] The above discussion focuses on changes in cardiac condition across different exercise states, as can be easily understood, but a similar process involving measurements over time can yield valuable information regardless of exercise state. For example, diurnal variations (night and day) can provide clinical insight in tracking a patient's condition. Furthermore, measuring heart rate over time can enable detection of arrhythmias, atrial fibrillation, supraventricular tachycardia, ventricular tachycardia, long QT syndrome, and / or any other heart rate affecting condition appropriate to the requirements of a particular application of an embodiment of the present invention. In various embodiments, heart rate patterns may be established using one or more methods, including (but not limited to) measuring A / V timing and based on the pressure morphology of the LAP signal. If other measurement tools are available, additional data can be used to confirm the accuracy of the measurements. For example, it is expected that the V wave will usually coincide with the echocardiogram (ECG) T wave.

[0045] With detailed and continuous access to medical information provided by the systems and methods described herein, personalized medicine becomes increasingly feasible. Indeed, the systems and methods described herein can utilize the data to present different information to medical professionals based on a patient's profile to maximize quality of care. Patient classification and personalized treatment processes are discussed further below.

[0046] C. Personalized Cardiac Medicine Cardiac patients are highly heterogeneous, which can mean that similar cardiac signals will be different in different patients. As discussed throughout, the various systems and methods described herein can optionally acquire a wide variety of cardiac signals and generate clinically relevant cardiac indices with increased sensitivity for each patient. The systems and methods described herein can be used to classify patients into any of a number of subcategories, and the cardiac signals and indices can be presented to medical personnel in a context that best fits the patient's overall classification.

[0047] 18, a process for personalized presentation of cardiac information according to an embodiment of the present invention is illustrated. Process 1800 includes acquiring 1810 patient data. In many embodiments, the patient data includes background patient data. The background patient data may be entered manually by a medical professional and / or may be acquired via an electronic medical record. The patient may then be parameterized 1820 based on the patient's background patient data to generate a patient profile. Parameters of the patient profile may include (but are not limited to): gender, type of cardiac condition (heart disease with preserved (diastolic) ejection fraction - HFpEF, heart disease with reduced (systolic) ejection fraction - HFrEF), hypertension (none, mild, severe), pulmonary hypertension group (group I - primary PH (idiopathic), group II - secondary PH (driven by elevated LAP), group III - active pulmonary disease, other), arrhythmia (none, paroxysmal atrial fibrillation; persistent atrial fibrillation, other), mitral regurgitation (mild, moderate, severe), other valvular pathologies (regurgitation, stenosis), ischemia and / or post-ischemic scar tissue, pacemaker (yes, no), and / or any other relevant parameters appropriate to the requirements of a particular application of an embodiment of the present invention.

[0048] The patient is classified (1830) based on the parameterized patient profile. In many embodiments, each unique combination of parameter values ​​may be treated as a separate class. In many embodiments, the patient profile is provided to a machine learning model, which generates a classification of the input profile. Classification using machine learning is described in more detail below. It may be assumed that each class of patients is more homogeneous, and that the LAP waveforms of each patient in the class may be interpreted in the same or similar manner. One or more cardiac signals are then acquired (1840) and cardiac indices are calculated (1850). In many embodiments, depending on the classification of the patient from which the cardiac signal was recorded, the method of calculating the cardiac indices may be different to be more appropriate for the type of cardiac disease the patient presents. The cardiac indices may then be displayed (1860) to the medical professional in the context of the classification via a user interface. In many embodiments, the classification is presented with an index. In various embodiments, different normal / abnormal ranges may be presented for one or more cardiac indices based on the classification. In various embodiments, different cardiac indices may be highlighted as being more likely to be clinically relevant based on the classification. In many aspects, different treatment methods may be suggested based on the cardiac index and the classification. As can be appreciated, any number of different personalized medicine applications may be implemented without departing from the scope or spirit of the present invention. Turning now to classification using machine learning, several possible alternatives are presented. In many aspects, a training dataset is generated that includes patient profiles annotated with classifications and used to train a supervised machine learning model. In various aspects, a classification is assigned to each normal cardiac index range. Providing the patient profile to the trained machine learning model can generate a classification for the patient profile, which may then be used to interpret the cardiac index.

[0049] In many embodiments, the supervised machine learning model trained is a random forest model or a support vector machine. In the case of a random forest model, multiple decision trees can be generated to generate a final model prediction. In contrast, a support vector machine can maximize the boundaries between different classes that maximize the margin between them. In various embodiments, two or more models are used. When two or more models are used to classify the patient profile, both classifications can be provided. In some embodiments, a definitive classification is provided only if the model outputs agree. Any discrepancies can be flagged to the medical professional. Additionally, a classification confidence value may be provided to represent the classification probability generated by the model. If the confidence value falls below a predetermined threshold (e.g., between 75-99%), a positive classification may not be provided, although this number can be adjusted to the user's risk tolerance.

[0050] As will be appreciated, other supervised machine learning models can be used to generate classifications without departing from the scope or spirit of the present invention. In various aspects, unsupervised machine learning models are used to eliminate the burden of generating training data. Although in some cases the accuracy may be lower, unsupervised learning allows for an unbiased classification process that may reveal new insights. For example, a clustering approach may be used on aggregated patient data into which an immediate patient profile is introduced.

[0051] In many aspects, where machine learning is utilized for classification, the patient's LAP signal and / or derived cardiac indices may be incorporated into a patient profile and provided to the machine learning model to further aid in the classification process. In various aspects, dimensionality reduction may be applied based on the waveform features prior to configuring the machine learning model.

[0052] The use of machine learning for classification can be particularly useful when integrated into an electronic health record (EHR) system. The EHR may already be composed of patient profiles annotated with patient outcomes and diagnoses. In this way, the EHR itself can be used as a training dataset or, in the case of unsupervised learning, as an aggregate set of patient profiles. In a sufficiently populated EHR, classification can be elevated to the level of disease clusters or individual disease level classification across healthy cardiac index ranges. In some aspects, the healthy cardiac index range can be further derived in the context of the EHR by observing the range of discharge cardiac indexes across the EHR population for that particular class.

[0053] D. Implant Monitoring Overall cardiac health monitoring remains an important function. For many patients fitted with an atrial cardiac sensor, additional implants may be implanted (or already implanted) to maintain cardiac health. However, implanted cardiovascular medical devices may degrade over time or be damaged or otherwise impaired by accident. LAP and / or A / V ratios can be used to measure the functionality of implantable medical devices, such as shunts and / or heart valves. In many aspects, the pressure difference between different sides of the shunt or valve can be used to build a pressure profile, which in turn is indicative of blood flow. Measurements can be made using sensors implanted on both sides of the shunt / valve, or via a sensor on one side of the shunt / valve and an external measurement modality. In some aspects, the valves and / or shunts may include cardiac sensors capable of pressure measurements as described herein on one or both sides where the required measurements can be made.

[0054] For example, mitral valve disease is a common form of cardiovascular disease. Two mitral valve abnormalities are mitral regurgitation and mitral stenosis. Mitral valve disease is often a contributing factor to heart failure. In many embodiments, mitral valve disease status can be monitored through the use of a mitral valve index (MVI), which reflects the deterioration of the mitral valve. The MVI may be calculated based on LAP measurements over time and updated as new LAP measurements are obtained. When the MVI exceeds a pre-specified level, the systems and methods described herein can be used to warn that mitral valve intervention is required (e.g., surgical percutaneous valve repair). In many embodiments, prosthetic heart valves are monitored using the MVI. If a prosthetic heart valve begins to deteriorate, the MVI is similarly affected and can be used to trigger an alarm and subsequent repair or replacement.

[0055] As can be readily appreciated, although sensors configured for implantation primarily in the atrium are discussed above, similar sensors can be made to measure pressure in other areas of the circulatory system, such that such pressure differentials can be obtained outside of the left atrium. In many embodiments, a mitral valve index is calculated to monitor the condition of the implant and / or overall cardiac health.

[0056] Mitral valve disease can lead to changes in pressure and flow in the left atrium. As mentioned above, there are two main mitral valve abnormalities: mitral regurgitation and mitral stenosis. In mitral regurgitation, the valve does not close properly, and blood leaks back from the left atrium into the left ventricle. Because the ventricle generates much higher pressure than the atrium, the backflow causes a pressure spike in the atrium at a time corresponding to ventricular contraction. In many aspects, this appears as an elevated V wave on the LAP signal. Mitral regurgitation is common in heart failure patients.

[0057] There are two main types of heart failure: heart failure with reduced ejection fraction (HFrEF) and heart failure with preserved ejection fraction (HFpEF), which are associated with two different structural changes in the heart. Approximately 65% ​​of HF patients are HFrEF patients. They have an enlarged heart and a reduced ejection fraction. When the heart is severely enlarged, the leaflets of the mitral valve are pulled away from each other and cannot close. The remaining 35% of HF patients are HFpEF patients. The muscular walls of the heart ventricles are thicker, fibrotic, and stiffer than normal. The mitral valve typically does not leak in HFpEF, but the stiff ventricles create diastolic dysfunction, where ventricular filling by the atrium is longer and creates higher pressure in the atrium. A graphical representation of the different structural changes in HFrEF and HFpEF hearts compared to normal is shown in FIG. 19. The systems and methods described herein can analyze LAP signal waveforms to diagnose both HFrEF and HFpEF and describe the differences between them. In HFrEF, mitral regurgitation tends to be more common and manifests as high V waves (e.g., peak V wave LAP above 25 mmHg) even with moderate mean LAP values ​​(5-25 mmHg). In HFpEF, both A and V waves tend to be elevated, thus resulting in more elevated mean LAP (e.g., above 20 mmHg), while the V wave is also moderately elevated (e.g., also elevated by about 20 mmHg). As an example, a LAP signal showing enlarged V wave pressure resulting from mitral regurgitation is shown in FIG. 20, according to an embodiment of the present invention. It is important to note that caval valve stenosis can also cause elevated A wave pressure.

[0058] Mitral stenosis occurs when the valve is diseased, calcified, or otherwise thickened and enlarged, causing the leaflets to not open completely. During atrial contraction, the mitral valve is normally wide open, allowing blood to flow into the ventricle. A stenosed valve opens only slightly, thus increasing pressure within the atrium. An illustration of a normal mitral valve compared to mitral stenosis is shown in FIG. 21. In the LAP signal, this is reflected by an elevated A-wave. An exemplary LAP signal with an elevated A-wave due to mitral stenosis in accordance with an embodiment of the present invention is shown in FIG. 22.

[0059] To detect and monitor mitral regurgitation and / or mitral stenosis, cardiac sensors described herein can be used to record LAP signals. Multiple LAP signals can be acquired over time and analyzed to determine the health of a patient's mitral valve. In many embodiments, cardiac indices can be used to quantify the severity of a mitral valve condition. In many embodiments, the A / V ratio and the minimum pressure point between the A and V waves are used to calculate a mitral valve index (MVI). When the MVI exceeds a threshold, an indication that intervention is required can be provided. In many embodiments, the intervention is a surgical percutaneous valve repair. In some embodiments, the intervention is a surgical percutaneous valve replacement. In various embodiments, the intervention is a replacement of a prosthetic heart valve.

[0060] 23, a process for calculating MVI according to an embodiment of the present invention is illustrated. Process 2300 includes acquiring (2310) a LAP signal. In many embodiments, multiple LAP signals are obtained. From the LAP signal, an A / V ratio is calculated (2320). During the calculation of the A / V ratio, A and V waves are identified. For each A and V wave, local maxima and minima are identified (2330). In many embodiments, the minima occur at the AV and VA transitions. A minimum at the AV wave transition may be referred to as an "X depression." Similarly, a minimum at the VA wave transition may be referred to as a Y depression. In various embodiments, a maximum occurs at the peak of each wave. An example of a normal LAP signal marked with minimum and maximum values ​​according to one embodiment of the present invention is shown in FIG. 24. Typically, the minimum at the AV transition is slightly higher than the minimum at the VA transition.

[0061] The mean maximum and minimum pressures are calculated (2340). The difference between the mean A wave maximum and the mean V wave maximum is calculated (2350), ΔMax. The difference between the mean AV minimum and the VA minimum is calculated (2360), ΔMin. In many embodiments, an increase in ΔMin may indicate diastolic dysfunction. An exemplary normal ΔMin according to an embodiment of the invention is shown in FIG. 25. In contrast, an exemplary increased ΔMin according to an embodiment of the invention is shown in FIG. 26. In various embodiments, an inverted ΔMin (e.g., a negative ΔMin resulting from a greater mean VA pressure compared to the mean AV pressure) may indicate delayed ventricular contraction in conditions such as heart block or long QT syndrome. An exemplary negative ΔMin according to an embodiment of the invention is shown in FIG. 27.

[0062] The A / V ratio, ΔMax, and ΔMin are then factored together to calculate the MVI (2370). In many embodiments, the MVI can be calculated as follows:

number

[0063] E. Cardiopulmonary monitoring Heart failure (HF) is a condition in which a patient's heart becomes compromised and is unable to generate sufficient blood output to supply the patient's organs' demand for oxygen. When the condition becomes chronic, permanent changes to cardiac structure and hemodynamics occur, often resulting in chronic symptoms that impair quality of life. The most common symptom is pulmonary congestion, and HF is often referred to as congestive heart failure (CHF).

[0064] Pulmonary congestion is the result of pulmonary hypertension and poor blood flow through the lungs. Heart failure increases blood pressure in the left atrium of the heart, which pumps blood out of the lungs. As left atrial pressure (LAP) increases, blood pressure in the lungs also increases, causing fluid to escape from the capillaries in the lungs into the air cavities (alveoli) in the lungs, causing congestion. With severe pulmonary congestion, the patient is effectively drowning and cannot get enough air into the lungs. This often requires hospitalization for treatment.

[0065] Treatment includes medications to minimize pulmonary congestion. Medications include diuretics to reduce the patient's general fluid volume, and antihypertensives to lower blood pressure. High doses tend to negatively impact quality of life by causing frequent urination and fatigue associated with low blood pressure. Therefore, optimization of medical management is important in maintaining quality of life for HF patients.

[0066] Traditionally, patients monitor heart failure by monitoring blood pressure, heart rate, pulse, weight, and mood (e.g., energy levels, tightness of clothing, swelling, shortness of breath, coughing, etc.). These are often subjective and can be difficult for patients to describe and difficult for physicians to understand. Furthermore, as discussed above, traditional wearable devices cannot record certain pressure measurements, such as (but not limited to) left atrial pressure (LAP), which can be used for cardiopulmonary monitoring and treatment, as discussed herein. Patient outcomes can be significantly more positive if worsening pulmonary congestion can be predicted and treated more aggressively, for example, by increasing drug dosage before congestion becomes too severe and requires hospitalization.

[0067] In many embodiments, the pulmonary congestion risk index is calculated as the area under the mean LAP signal curve above a hypertensive level (also called the "LAP threshold") to monitor the patient. In many aspects, at least one LAP signal is recorded daily. Each day, a new LAP signal is analyzed in conjunction with the previously recorded daily LAP signals for the patient to trend the data and determine the risk of developing pulmonary congestion. The risk may be represented by a pulmonary congestion risk index. Furthermore, due to noise in the LAP signals and / or temporal variability in their acquisition, filtering and / or averaging techniques may be applied. When the pulmonary congestion risk index exceeds a threshold, an alarm may be provided that can trigger the patient's treatment to prevent severe congestion.

[0068] Congestion is caused by the diffusion of fluid from the blood into the pulmonary cavities. The rate of diffusion is controlled by the pressure gradient from the blood to the air cavity. As LAP increases, pulmonary blood pressure also increases, leading to faster accumulation of fluid in the lungs. Excess fluid is normally slowly expelled from the lungs by reabsorption and exhalation. If the rate of fluid accumulation is higher than the rate of fluid expulsion, fluid may accumulate in the lungs, leading to severe congestion. Thus, integration of blood pressure rise over time may be a predictor of pulmonary congestion.

[0069] The pulmonary congestion risk index is a measure of the severity and length of time that a patient's average LAP exceeds the hypertension level. In many embodiments, the hypertension level is 20 mmHg, which can be considered as high pressure in the left atrium. However, chronic HF patients on medication may sustain much higher pressures, and a physician may select a higher threshold. Similarly, some patients may benefit from a lower threshold that reflects high pressure on the heart, depending on their particular condition. The time window of the index can be pre-determined for analysis, typically 3 to 30 days. The pulmonary congestion risk index can be calculated by finding all events within the time window where the daily average LAP exceeds the threshold, and summing the difference between the average LAP and the threshold for all events within the time window.

[0070] 28, a first process for calculating a pulmonary congestion risk index according to an embodiment of the present invention is shown. Process 2800 includes setting 2810 a high blood pressure level and a pulmonary congestion alarm level. In many embodiments, if the high blood pressure level is 20 mmHg, the alarm level is set to 30 mmHg. The alarm level can be set to any value depending on the particular patient. In many embodiments, the high blood pressure level is 20 mmHg, but this number may vary depending on the particular patient as described above. A LAP signal is received daily 2820, and an average LAP per day is calculated each day 2830. In many embodiments, the LAP signal is recorded by an atrial cardiac sensor. The LAP signal may be received at a cardiac signal processor via a transceiver as discussed above. In many embodiments, multiple LAP signals are recorded each day and averaged together. In many embodiments, the LAP signal is recorded each day during a similar exercise state, e.g., while at rest.

[0071] If the daily average LAP for a particular day is greater than the hypertension level, the difference between the daily average LAP and the hypertension level is calculated (2840). A pulmonary congestion risk index is then calculated (2850) by summing the daily differences over a moving index time window. If the pulmonary congestion risk index exceeds the pulmonary congestion alarm level, an alarm is triggered and treatment is recommended (2860), and in many embodiments, implemented. Now referring to Figures 29A and 29B, process 2900 is illustrated graphically on an arbitrary set of sample data according to an embodiment of the present invention. Figure 29A shows the average LAP over time, and the newly shaded area in Figure 29B shows the value of the pulmonary congestion risk index.

[0072] Although a process such as process 2800 can be used to calculate a pulmonary congestion risk index, the daily measurements may jump up and down and have a level of noise associated with them. To reduce false alarms, the daily LAP measurements may be smoothed by averaging the measurements over consecutive days. Referring now to FIG. 30, a second process for calculating a pulmonary congestion risk index using smoothing according to an embodiment of the present invention is shown.

[0073] The process 3000 includes setting (3010) a high blood pressure level and a pulmonary congestion alarm level. In many embodiments, the high blood pressure level is 20 mmHg, although this number may vary depending on the particular patient as discussed above. In various embodiments, the alarm level is between 0 and 200 mmHg. A LAP signal is received daily (3020), and a daily average LAP is calculated for each day (3030). In many embodiments, the LAP signal is recorded by an atrial cardiac sensor. The LAP signal may be received at a cardiac signal processor via a transceiver, as discussed above. In many embodiments, multiple LAP signals are recorded each day and averaged together. In many embodiments, LAP signals are recorded each day during similar exercise conditions, e.g., while at rest. A running average LAP is further calculated (3040). In many embodiments, the running average LAP is taken over a three day window, although a two to ten day window may also be used.

[0074] If the moving average LAP is greater than the hypertension level, the difference between the moving average LAP and the hypertension level is calculated (3050). A pulmonary congestion risk index is then calculated by summing the differences over the time window of the index (3060). If the pulmonary congestion risk index exceeds the pulmonary congestion alarm level, an alarm is triggered and treatment is recommended (3070), and in many embodiments, implemented. Now referring to Figures 31A and 31B, the process 3000 is illustrated graphically on an arbitrary set of sample data according to an embodiment of the present invention. Figure 31A shows the average LAP over time, and the new shaded area in Figure 31B shows the value of the pulmonary congestion risk index.

[0075] Although the process 3000 introduces smoothing and can help reduce noise, day-to-day variability in the LAP signal is often the cause of more than noise. Higher variability may indicate that the patient is less stable, and less stable patients are at higher risk of developing pulmonary congestion, but with little warning. Often, patients may be stable at times and not stable at times when the risk is elevated. In many embodiments, this risk can be quantified by calculating the standard deviation of the LAP against the smoothed LAP signal and taking into account the variability of the pulmonary congestion risk index. Now referring to FIG. 3200, a third process for calculating a pulmonary congestion risk index is shown, accounting for variability according to an embodiment of the present invention.

[0076] The process 3200 includes setting (3210) a high blood pressure level and a pulmonary congestion alarm level. In various embodiments, the alarm level is 30 mmHg, but daily variations tend to add 5-10 mmHg each day. This variability can be accounted for by setting the high blood pressure level approximately 5 mmHg higher. In various embodiments, the high blood pressure level is set at 20 mmHg and the alarm level is set between 50-60 mmHg. In many embodiments, the high blood pressure level is 20 mmHg, but this number can vary depending on the particular patient as discussed above. A LAP signal is received each day (3220) and for each day, an average LAP per day is calculated (3230). In many embodiments, the LAP signal is recorded by an atrial cardiac sensor. The LAP signal may be received at a cardiac signal processor via a transceiver, as discussed above. In many embodiments, multiple LAP signals are recorded each day and averaged together. In many embodiments, LAP signals are recorded daily during similar exercise conditions, for example while at rest.

[0077] A moving N-day average LAP is also calculated (3240). In many embodiments, the moving average LAP is taken over a 3-day window N=3, although windows of 2-10 days may also be used. An M-day LAP standard deviation is calculated (3250). In many embodiments, M>N. In various embodiments, N=3, and M=5. In many embodiments, M=N+2.

[0078] If the N-day moving average LAP+2M-day standard deviation is greater than the hypertension level, the difference between the moving average LAP+2M-day standard deviation and the hypertension level is calculated (3260). A pulmonary congestion risk index is then calculated by summing the differences over the index time frame (3270). If the pulmonary congestion risk index exceeds the pulmonary congestion alarm level, an alarm is triggered and treatment is recommended (3280), and in many embodiments, implemented. Now referring to Figures 11A and 11B, the process 3200 is graphed on an arbitrary set of sample data, where N=3 and M=5, in accordance with one embodiment of the present invention. Figure 33A shows the average LAP over time, and the new shaded area in Figure 33B shows the value of the pulmonary congestion risk index.

[0079] Although three different specific processes for calculating the Pulmonary Congestion Risk Index are discussed above, as can be readily appreciated, the processes can be modified with different moving windows, thresholds, and other modified parameters for a particular patient as appropriate to the requirements of a particular application of an embodiment of the present invention.

[0080] Further description of the embodiment Provided below is a list of examples, each of which may include aspects of any of the other examples disclosed herein. Additionally, aspects of any of the examples described above may be implemented in any of the numbered examples provided below.

[0081] Example 1. A method of treating a cardiac condition in a patient, comprising: acquiring a left atrial pressure (LAP) signal using an atrial cardiac sensor implanted in the patient's left atrium; identifying a set of A waves and a set of V waves within the LAP signal; calculating an A / V ratio based on the average pressure amplitude of the set of A waves and the average pressure amplitude of the set of V waves; Determining appropriate treatment for a patient based on the A / V ratio; and providing an appropriate treatment to the patient.

[0082] The method of treating a cardiac condition in a patient according to any of the embodiments described herein, particularly embodiment 1, wherein identifying the set of A waves and the set of V waves in the LAP signal comprises decomposing the LAP signal.

[0083] Example 3. A method of treating a cardiac condition in a patient according to any embodiment described herein, particularly embodiment 2, wherein decomposing the LAP signal comprises applying a Fourier transform.

[0084] Example 4. A method of treating a cardiac condition in a patient as described in any embodiment herein, particularly embodiment 2, wherein decomposing the LAP signal comprises applying wavelet analysis.

[0085] Example 5. A method of treating a cardiac condition in a patient according to any embodiment described herein, particularly Example 2, further comprising determining the patient's respiratory rate by locating a dominant frequency between 0.1 and 0.5 Hz in the resolution.

[0086] Example 6. A method of treating a cardiac condition in a patient according to any embodiment described herein, particularly Example 2, further comprising determining a heart rate of the patient by locating a dominant frequency between 0.8 Hz and 1.7 Hz in the decomposition.

[0087] Example 7. A method of treating a cardiac condition in a patient according to any embodiment described herein, particularly embodiment 2, further comprising subtracting frequency components associated with respiratory rate from the LAP signal to produce a noise-removed LAP signal.

[0088] Example 8. A method of treating a cardiac condition in a patient according to any embodiment described herein, particularly Example 7, further comprising parsing the denoised LAP signal into an even wave set and an odd wave set.

[0089] Example 9. A method of treating a cardiac condition as described in any embodiment herein, particularly Example 8, wherein the even wave sets and odd wave sets are exclusively labeled as A wave sets or V wave sets.

[0090] Example 10. A method of treating a cardiac condition as described in any of the examples described herein, particularly Example 9, wherein the exclusionary labeling is performed using a machine learning model trained using a training dataset comprising a denoised LAP signal annotated with A and V waves.

[0091] Example 11. A method of treating a cardiac condition as described in any of the examples described herein, particularly Example 9, wherein the exclusionary labeling is performed by waveform analysis, including identification of a double peak in the A wave.

[0092] Example 12. A method of treating a cardiac condition according to any of the examples described herein, particularly Example 1, wherein the appropriate treatment comprises increasing the dosage of diuretic administered to the patient by between 40% and 60% in response to the A / V ratio decreasing to less than 0.6.

[0093] Example 13. A method of treating a cardiac condition as described in any of the examples described herein, particularly Example 1, wherein the appropriate treatment comprises increasing the dosage of a hypertension medication while maintaining the patient on a diuretic dosage in response to an increase in the patient's mean blood pressure of more than 40 mmHg without a change in A / V ratio.

[0094] Example 14. A method of treating a cardiac condition according to any embodiment described herein, particularly Example 1, wherein the appropriate treatment is a medication that is automatically delivered to the patient via an infusion pump.

[0095] Example 15. A system for treating a cardiac condition in a patient, comprising: A processor; A memory, the memory comprising: Obtain a recorded left atrial pressure (LAP) signal using an atrial cardiac sensor; Identifying the set of A waves and the set of V waves within the LAP signal; Calculate the A / V ratio based on the average pressure amplitude of the set of A waves and the average pressure amplitude of the set of V waves; and a memory including a cardiac monitoring application configured to determine an appropriate therapy for a patient based on the A / V ratio.

[0096] Example 16. A system for treating a cardiac condition of a patient as described in any of the embodiments described herein, particularly Example 15, wherein the cardiac monitoring application is further configured to cause the processor to decompose the LAP signal to identify a set of A waves and a set of V waves in the LAP signal.

[0097] Example 17. A system for treating a cardiac condition of a patient as described in any embodiment described herein, particularly example 16, wherein the decomposition comprises application of a Fourier transform.

[0098] Example 18. A system for treating a cardiac condition in a patient as described in any embodiment described herein, particularly example 16, wherein the decomposition comprises application of wavelet analysis.

[0099] Example 19. A system for treating a cardiac condition of a patient as described in any of the embodiments described herein, particularly Example 16, wherein the cardiac monitoring application is further configured to cause the processor to determine the patient's respiratory rate by locating a dominant frequency between 0.1 and 0.5 Hz in resolution.

[0100] Example 20. A system for treating a cardiac condition of a patient as described in any of the embodiments described herein, particularly Example 16, wherein the cardiac monitoring application is further configured to cause the processor to determine the patient's heart rate by locating a dominant frequency between 0.8 Hz and 1.7 Hz in the resolution.

[0101] Example 21. A system for treating a cardiac condition of a patient as described in any of the embodiments described herein, particularly Example 16, wherein the cardiac monitoring application is further configured to cause the processor to subtract frequency components associated with respiratory rate from the LAP signal to generate a noise-removed LAP signal.

[0102] Example 22. A system for treating a cardiac condition of a patient as described in any of the embodiments described herein, particularly Example 21, wherein the cardiac monitoring application is further configured to cause the processor to analyze the noise-removed LAP signal into a set of even waves and a set of odd waves.

[0103] Example 23. A system for treating a cardiac condition as described in any embodiment herein, particularly Example 22, wherein the even wave sets and odd wave sets are exclusively labeled as A wave sets or V wave sets.

[0104] Example 24. A system for treating a cardiac condition as described in any of the embodiments described herein, particularly Example 23, wherein the exclusionary labeling is performed using a machine learning model trained using a training dataset comprising a denoised LAP signal annotated with A and V waves.

[0105] Example 25. A system for treating a cardiac condition as described in any of the embodiments described herein, particularly as described in Example 23, wherein the exclusionary labeling is performed by waveform analysis, including identification of a double peak in the A wave.

[0106] Example 26. A system for treating a cardiac condition as described in any embodiment herein, particularly Example 15, wherein the appropriate treatment includes increasing a dose of diuretic administered to the patient by between 40% and 60% in response to the A / V ratio falling below 0.6, and the cardiac monitoring application is further configured to cause the processor to deliver the dose of diuretic via an infusion pump communicatively coupled to the processor.

[0107] Example 27. A system for treating a cardiac condition as described in any of the embodiments described herein, particularly Example 15, wherein the appropriate treatment includes increasing a dosage of a hypertension medication while maintaining a dosage of a diuretic to the patient in response to an increase in the patient's mean blood pressure of more than 40 mmHg without a change in A / V ratio, and the cardiac monitoring application is further configured to cause the processor to deliver the dosage of the diuretic via an infusion pump communicatively coupled to the processor.

[0108] Example 28. A system for treating a cardiac condition as described in any of the embodiments described herein, particularly Example 15, wherein the appropriate therapy is a medication that is automatically delivered to the patient via an infusion pump communicatively coupled to the processor.

[0109] Example 29. A cardiac monitoring system for calculating an A / V ratio, comprising: an atrial cardiac sensor configured to be implanted in the patient's left atrium to record a left atrial pressure (LAP) signal; A transceiver comprising: wirelessly providing power to the atrial cardiac sensor; a transceiver configured to receive a LAP signal from an atrial cardiac sensor; A processor; A memory containing a cardiac monitoring application, the cardiac monitoring application causing a processor to: Identifying the set of A waves and the set of V waves within the LAP signal; and a memory configured to cause an A / V ratio to be calculated based on an average pressure amplitude of the set of A waves and an average pressure amplitude of the set of V waves.

[0110] Example 30. A cardiac monitoring system according to any embodiment described herein, particularly Example 29, wherein the cardiac monitoring application is further configured to cause the processor to decompose the LAP signal to identify a set of A waves and a set of V waves in the LAP signal.

[0111] Example 31 A cardiac monitoring system as described herein, in any embodiment, particularly example 30, wherein the decomposition comprises application of a Fourier transform.

[0112] Example 32 A cardiac monitoring system as described herein in any embodiment, particularly example 30, wherein the decomposition comprises application of wavelet analysis.

[0113] Example 33. A cardiac monitoring system as described in any of the embodiments herein, particularly example 30, wherein the cardiac monitoring application is further configured to cause the processor to determine the patient's respiratory rate by locating a dominant frequency between 0.1 and 0.5 Hz in resolution.

[0114] Example 34. Any of the embodiments described herein, particularly example 30, wherein the cardiac monitoring application is further configured to cause the processor to determine the patient's heart rate by locating a dominant frequency between 0.8 Hz and 1.7 Hz in the resolution.

[0115] Example 35. A cardiac monitoring system as described in any of the embodiments described herein, particularly example 30, wherein the cardiac monitoring application is further configured to cause the processor to subtract frequency components associated with respiratory rate from the LAP signal to generate a noise-removed LAP signal.

[0116] Example 36. A cardiac monitoring system according to any of the embodiments described herein, particularly example 35, wherein the cardiac monitoring application is further configured to cause the processor to analyze the noise-removed LAP signal into a set of even waves and a set of odd waves.

[0117] Example 37. A cardiac monitoring system as described in any embodiment herein, particularly example 36, wherein the even wave sets and odd wave sets are exclusively labeled as A wave sets or V wave sets.

[0118] Example 38. A cardiac monitoring system as described in any of the embodiments described herein, particularly Example 37, wherein the exclusionary labeling is performed using a machine learning model trained using a training dataset comprising a denoised LAP signal annotated with A and V waves.

[0119] Example 39. A cardiac monitoring system as described in any of the embodiments described herein, particularly example 37, wherein the exclusion labeling is performed by waveform analysis, including identification of a double peak in the A wave.

[0120] Example 40. A method for calculating an A / V ratio, comprising: acquiring a left atrial pressure (LAP) signal using an atrial cardiac sensor configured to be implanted in the patient's left atrium; identifying a set of A waves and a set of V waves within the LAP signal; and calculating an A / V ratio based on the average pressure amplitude of the set of A waves and the average pressure amplitude of the set of V waves.

[0121] Example 41. The method of calculating the A / V ratio as described in any of the embodiments described herein, particularly Example 40, wherein the cardiac monitoring application is further configured to cause the processor to decompose the LAP signal to identify a set of A waves and a set of V waves in the LAP signal.

[0122] Example 42. A method of calculating the A / V ratio as described in any embodiment described herein, particularly Example 41, wherein the decomposition comprises application of a Fourier transform.

[0123] Example 43. A method of calculating the A / V ratio as described in any embodiment herein, particularly Example 41, wherein the decomposition comprises application of wavelet analysis.

[0124] Example 44. The method of calculating the A / V ratio of any of the embodiments herein, particularly Example 41, wherein the cardiac monitoring application is further configured to cause the processor to determine the patient's respiratory rate by locating a dominant frequency, in a resolution, between 0.1 and 0.5 Hz.

[0125] Example 45. The method of calculating the A / V ratio of any embodiment described herein, particularly Example 41, wherein the cardiac monitoring application is further configured to cause the processor to determine the patient's heart rate by locating a dominant frequency between 0.8 Hz and 1.7 Hz in the resolution.

[0126] Example 46. The method of calculating the A / V ratio as described in any of the embodiments described herein, particularly Example 41, wherein the cardiac monitoring application further configures the processor to subtract frequency components associated with the respiratory rate from the LAP signal to generate a noise-removed LAP signal.

[0127] Example 47. The method of calculating the A / V ratio as described in any of the embodiments described herein, particularly Example 46, wherein the cardiac monitoring application is further configured to cause the processor to analyze the noise-removed LAP signal into a set of even waves and a set of odd waves.

[0128] Example 48. A method of calculating the A / V ratio as described in any of the examples described herein, particularly Example 47, wherein the set of even waves and the set of odd waves are exclusively labeled as the set of A waves or the set of V waves.

[0129] Example 49. A method of calculating the A / V ratio as described in any of the examples described herein, particularly Example 48, wherein the exclusionary labeling is performed using a machine learning model trained using a training dataset comprising a denoised LAP signal annotated with A and V waves.

[0130] Example 50. A method of calculating the A / V ratio as described in any of the examples herein, particularly as described in Example 48, wherein the exclusion labeling is performed by waveform analysis, including identification of a double peak in the A wave.

[0131] Example 51. A method for monitoring a patient's response to exercise, comprising: recording a first blood pressure signal while the patient is in a resting state using an atrial cardiac sensor implanted in the patient's left atrium; Detecting a change from a resting state to a moving state; recording a second blood pressure signal during the exercise state in response to detecting a change in the patient's condition using the atrial cardiac sensor; receiving, at a cardiac signal processor, the first and second blood pressure signals; calculating a first plurality of cardiac indices based on the first blood pressure signal using a cardiac signal processor; calculating a second plurality of cardiac indices based on the second blood pressure signal using a cardiac signal processor; determining, using the cardiac signal processor, a change in the patient's health status based on a change between the first and second plurality of cardiac indices; and treating the patient in response to the change in health status.

[0132] Example 52. A method of monitoring a patient's response to exercise as described in any embodiment herein, particularly embodiment 51, wherein the transceiver instructs the atrial cardiac sensor to perform recording of first and second blood pressure signals.

[0133] Example 53. A method of monitoring a patient's response to exercise as described in any embodiment described herein, particularly example 51, wherein the first blood pressure signal is obtained while the patient is in a resting state.

[0134] Example 54. A method of monitoring a patient's response to exercise as described in any embodiment herein, particularly as described in Example 51, wherein the second blood pressure signal is obtained while the patient is in an active state.

[0135] Example 55. The method of monitoring a patient's response to exercise as described in any embodiment described herein, particularly example 51, wherein the first and second blood pressure signals are left atrial pressure signals.

[0136] Example 56. A method of monitoring a patient's response to exercise as described in any embodiment herein, particularly Example 51, wherein the first plurality of cardiac indices and the second plurality of cardiac indices comprise A / V ratio.

[0137] Example 57. A method of monitoring a patient's response to exercise as described in any embodiment described herein, particularly example 51, wherein the first plurality of cardiac indicators and the second plurality of cardiac indicators include respiratory rate.

[0138] Example 58. A method of monitoring a patient's response to exercise as described in any embodiment described herein, particularly example 51, wherein the first plurality of cardiac indicators and the second plurality of cardiac indicators include heart rate.

[0139] Example 59. A method of monitoring a patient's response to exercise as described in any embodiment herein, particularly example 51, wherein the treatment is automated delivery of a medication via an infusion pump.

[0140] Example 60. A cardiac patient motion monitoring system, comprising: It is implanted in the patient's left atrium, a first blood pressure signal during a patient resting state; a first blood pressure signal during the patient's exercise state; and a second blood pressure signal during the patient's exercise state. 1. A cardiac signal processor comprising: receiving first and second blood pressure signals; Calculating a first plurality of cardiac indices based on the first blood pressure signal; Calculating a second plurality of cardiac indices based on the second blood pressure signal; and a cardiac signal processor configured to provide an alert indicative of a patient's health condition based on a change between the first and second plurality of cardiac indices.

[0141] Example 61. A cardiac patient motion monitoring system as described in any of the examples described herein, particularly example 60, comprising: obtaining first and second blood pressure signals from an atrial cardiac sensor; The cardiac patient motion monitoring system further comprises a transceiver configured to transmit the first and second blood pressure signals to the cardiac signal processor.

[0142] Example 62. A cardiac patient motion monitoring system as described in any embodiment described herein, particularly example 61, wherein the transceiver is further configured to power the atrial cardiac sensor.

[0143] Example 63. A cardiac patient motion monitoring system as described in any embodiment herein, particularly example 60, wherein powering the atrial cardiac sensor triggers the atrial cardiac sensor to record the first cardiac signal.

[0144] Example 64. A cardiac patient motion monitoring system as described in any embodiment described herein, particularly example 60, further comprising an infusion pump, and wherein the cardiac signal processor is further configured to automatically deliver medication to the patient via the infusion pump based on the patient's health status.

[0145] Example 65. A cardiac patient motion monitoring system as described in any embodiment described herein, particularly example 60, wherein the first and second blood pressure signals are left atrial pressure signals.

[0146] Example 66. A cardiac patient motion monitoring system according to any embodiment described herein, particularly example 60, wherein the first plurality of cardiac indices and the second plurality of cardiac indices include an A / V ratio.

[0147] Example 67. A cardiac patient motion monitoring system as described in any embodiment herein, particularly example 60, wherein the first plurality of cardiac indicators and the second plurality of cardiac indicators include respiratory rate.

[0148] Example 68. A cardiac patient motion monitoring system as described in any embodiment herein, particularly example 60, wherein the first plurality of cardiac indicators and the second plurality of cardiac indicators include heart rate.

[0149] Example 69. A cardiac patient motion monitoring system, comprising: It is implanted in the patient's left atrium, a first blood pressure signal during a patient resting state; a first blood pressure signal during the patient's exercise state; and a second blood pressure signal during the patient's exercise state. 1. A cardiac signal processor comprising: receiving first and second blood pressure signals; calculating a first heart rate (HR), a first mean blood pressure (P), a first respiratory rate (RR), and a first A / V ratio based on the first blood pressure signal; calculating a second HR, a second P, a second RR, and a second A / V ratio based on the second blood pressure signal; calculating a P / HR ratio between the change between the first P and the second P and the change between the first HR and the second HR; calculating an RR / HR ratio between the change between the first RR and the second RR and the change between the first HR and the second HR; calculating an (A / V) / HR ratio between the change between the first A / V ratio and the second A / V ratio and the change between the first HR and the second HR; and a cardiac signal processor configured to provide an alert indicating a negative health condition when at least one of the P / HR ratio, the RR / HR ratio, and the (A / V) / HR ratio exceed a respective threshold value.

[0150] Example 70. A cardiac patient motion monitoring system as described in any embodiment herein, particularly example 69, further comprising an infusion pump, and wherein the cardiac signal processor is further configured to automatically infuse medication to the patient to treat the cardiac condition in response to a negative health state.

[0151] Example 71. A cardiac patient motion monitoring system, comprising: an atrial cardiac sensor configured to be implanted in the patient's left atrium to record a plurality of pressure signals over the course of at least one day; 1. A cardiac signal processor comprising: Receives multiple pressure signals; Calculating a heart rate (HR) and an A / V ratio for each pressure signal of the plurality of pressure signals; Plot a curve with A / V ratio on the y-axis and recording time on the x-axis. Identifying times when the patient was in an exercise state and times when the patient was in a resting state based on the HR; Annotating the curve with the identified motion and rest states; and a cardiac signal processor configured to display the curve.

[0152] Example 72. A cardiac signal processor comprising: Calculating a mean blood pressure (P) for each pressure signal of the plurality of pressure signals; Calculating the P / HR for each pair of consecutive pressure signals in the plurality of pressure signals; Plot additional curves with P / HR ratio on the y-axis and recording time on the x-axis. annotating the additional curves with the identified motion and rest states; A cardiac patient motion monitoring system as described in any embodiment described herein, particularly embodiment 71, further configured to display additional curves.

[0153] Example 73. A cardiac signal processor comprising: For each pair of consecutive pressure signals in the plurality of pressure signals, calculate (A / V) / HR; Plot additional curves where the (A / V) / HR ratio is on the y-axis and recording time is on the x-axis. annotating the additional curves with the identified motion and rest states; A cardiac patient motion monitoring system as described in any embodiment described herein, particularly embodiment 71, further configured to display additional curves.

[0154] Example 74. A cardiac patient use monitoring system as described in any embodiment described herein, particularly example 71, wherein the cardiac signal processor is further configured to provide a warning indicating a negative health condition based on the curve.

[0155] Example 75. A method for monitoring cardiac patient motion, comprising: receiving a plurality of pressure signals from a cardiac sensor implanted in a left atrium of a heart of a cardiac patient; Calculating a heart rate (HR) and an A / V ratio for each pressure signal of the plurality of pressure signals; Plotting a curve with A / V ratio on the y-axis and recording time on the x-axis; Identifying times when the patient was in an exercise state and times when the patient was in a resting state based on the HR; annotating the curve with the identified motion and rest states; and displaying the curve.

[0156] Example 76. A method for monitoring cardiac patient motion as described in any embodiment herein, particularly as described in example 75, comprising: calculating a mean blood pressure (P) for each pressure signal of the plurality of pressure signals; calculating a P / HR for each pair of consecutive pressure signals in the plurality of pressure signals; Plotting an additional curve with the P / HR ratio on the y-axis and the recording time on the x-axis; annotating the additional curves with the identified motion and rest states; and displaying the additional curve.

[0157] Example 77. A method for monitoring exercise in a cardiac patient as described in any of the embodiments described herein, particularly as described in Example 75, comprising: calculating (A / V) / HR for each pair of consecutive pressure signals in the plurality of pressure signals; Plotting an additional curve with the (A / V) / HR ratio on the y-axis and recording time on the x-axis; annotating the additional curves with the identified motion and rest states; and displaying the additional curve.

[0158] Example 78. A method of monitoring cardiac patient motion as described in any embodiment described herein, particularly example 75, wherein the cardiac signal processor is further configured to provide a warning indicating a negative health condition based on the curve.

[0159] Example 79. A method for differentially presenting individualized patient health indicators to a medical professional, comprising: receiving a left atrial blood pressure waveform from an atrial cardiac sensor implanted in the patient's left atrium; acquiring background medical information of a patient, the background medical information of the patient including values ​​of a plurality of categorical parameters relating to the patient; generating a parameterized patient profile based on the patient's background information; Classifying the patient based on a parameterized patient profile; Calculating cardiac metrics including mean left atrial pressure (LAP), peak LAP, heart rate, respiratory rate, A / V ratio, and exercise response based on the left atrial blood pressure waveform; and displaying the cardiac indices based on the classification such that an indication is provided as to whether each cardiac index is within a normal range for the classification.

[0160] Example 80. A parameterized patient profile is sex, Heart condition, Pulmonary hypertension group, arrhythmia, Mitral regurgitation, Ischemic scar tissue, and The method according to any embodiment described herein, particularly embodiment 79, including parameters, for pacemaker implantation.

[0161] Example 81. The method of any of the embodiments described herein, particularly example 79, wherein each unique combination of parameters in a parameterized patient profile is associated with a classification.

[0162] Example 82. The method described in any example described herein, particularly example 79, wherein to classify the patient, the patient profile is provided to a machine learning model configured to provide a classification.

[0163] Example 83 The method of any of the embodiments described herein, particularly example 82, wherein the machine learning model is a supervised machine learning model.

[0164] Example 84 The method of any of the embodiments described herein, particularly Example 83, wherein the machine learning model is a random forest model.

[0165] Example 85 The method of any of the embodiments described herein, particularly Example 83, wherein the machine learning model is a support vector machine.

[0166] Example 86. The method described in any of the examples described herein, particularly Example 83, wherein the supervised machine learning model is trained using a training dataset comprising patient profiles from an electronic medical record system in which patient outcomes are annotated.

[0167] Example 87. The method of any of the examples described herein, particularly example 82, wherein the machine learning model is an unsupervised clustering model, and the unsupervised clustering model is provided with patient profiles from an electronic medical record system, and the patient profiles are for classifying patients as belonging to a particular cluster.

[0168] Example 88. The method described in any of the examples described herein, particularly example 87, wherein the normal range of cardiac index for a particular cluster is derived from cardiac index values ​​from the electronic medical records of patients in the particular cluster at the time of discharge.

[0169] Example 89. An electronic medical record system for differentially presenting personalized health indicators, comprising: A processor; A memory, the memory comprising: receiving a left atrial blood pressure waveform from an atrial cardiac sensor implanted in the left atrium of the patient; acquiring background medical information of the patient, the background medical information of the patient including values ​​of a plurality of categorical parameters relating to the patient; generating a parameterized patient profile based on the patient's background information; Classifying patients based on a parameterized patient profile; Calculating cardiac metrics including mean left atrial pressure (LAP), peak LAP, heart rate, respiratory rate, A / V ratio, and exercise response based on the left atrial pressure waveform; and a memory including an application configured to cause the cardiac indices to be displayed based on the classification such that an indication is provided as to whether each cardiac index is within a normal range for the classification.

[0170] Example 90. A parameterized patient profile is sex, Heart condition, Pulmonary hypertension group, arrhythmia, Mitral regurgitation, Ischemic scar tissue, A system as described in any embodiment herein, particularly embodiment 89, including parameters for pacemaker implantation.

[0171] Example 91. A system according to any of the embodiments described herein, particularly example 89, wherein each unique combination of parameters in a parameterized patient profile is associated with a classification.

[0172] Example 92. Any of the embodiments described herein, particularly the system described in Example 89, wherein, to classify the patient, the application further instructs the processor to provide the patient profile to a machine learning model configured to provide a classification.

[0173] Example 93. Any of the embodiments described herein, particularly the system described in Example 92, wherein the machine learning model is a supervised machine learning model.

[0174] Example 94. The system of any of the embodiments described herein, particularly example 93, wherein the machine learning model is a random forest model.

[0175] Example 95. Any of the embodiments described herein, particularly the system described in Example 93, wherein the machine learning model is a support vector machine.

[0176] Example 96. Any of the examples described herein, particularly the system described in Example 93, wherein the supervised machine learning model is trained using a training dataset including patient profiles from an electronic medical record system in which patient outcomes are annotated.

[0177] Example 97. Any of the examples described herein, particularly the system described in Example 92, wherein the machine learning model is an unsupervised clustering model, and the unsupervised clustering model is provided with patient profiles from an electronic medical record system, and the patient profiles are for classifying patients as belonging to a particular cluster.

[0178] Example 98. A system as described in any of the embodiments described herein, particularly example 97, wherein the normal range of a cardiac index for a particular cluster is derived from cardiac index values ​​from the patient's electronic medical record in the particular cluster at the time of discharge.

[0179] Example 99. A method of treating a cardiac condition in a patient, comprising: acquiring a left atrial pressure (LAP) signal using an atrial cardiac sensor implanted in the patient's left atrium; identifying a set of A waves and a set of V waves within the LAP signal; calculating an A / V ratio based on the average pressure amplitude of the set of A waves and the average pressure amplitude of the set of V waves; Calculating the average maximum amplitude of the set of A waves; Calculating the mean maximum amplitude of the set of waves V; Calculating the average minimum amplitude of the inflection point in the transition of wave A to wave V in the LAP signal; Calculating the average minimum amplitude of the inflection point at the transition of wave V to wave A in the LAP signal; Calculating the difference ΔMax between the average maximum amplitude of the set of A waves and the average maximum amplitude of the set of V waves; Calculating the difference ΔMin between the average minimum amplitude of the inflection point in the transition of wave A to wave V and the average minimum amplitude of the inflection point in the transition of wave V to wave A; Calculating a mitral valve index (MVI) based on the A / V ratio, ΔMax, and ΔMin; and determining a treatment for the patient based on the MVI exceeding a threshold.

[0180] Example 100 The method of any of the embodiments described herein, particularly example 99, wherein the treatment is implantation of a prosthetic heart valve.

[0181] Example 101 The method of any of the embodiments described herein, particularly example 99, wherein the treatment is replacement of a prosthetic heart valve.

[0182] Example 102 The method of any of the embodiments described herein, particularly example 99, wherein the appropriate treatment is surgical percutaneous valve repair.

[0183] Example 103. MVI is

number

[0184] Example 104. The method of any of the embodiments described herein, especially Example 103, wherein C1 and C2 are between 1 and 5.

[0185] Example 105. The method according to any of the examples described herein, particularly example 104, wherein the threshold is 1 and an MVI>1 indicates a healthy patient not requiring treatment.

[0186] Example 106 The method of any of the examples described herein, particularly example 99, further comprising providing the determined treatment.

[0187] Example 107. A mitral valve monitoring and therapy system, comprising: an atrial cardiac sensor configured to be implanted in the patient's left atrium to record a left atrial pressure (LAP) signal; A transceiver comprising: wirelessly providing power to the atrial cardiac sensor; a transceiver configured to receive a LAP signal from an atrial cardiac sensor; A processor; A memory containing a cardiac monitoring application, the cardiac monitoring application causing a processor to: Identifying the set of A waves and the set of V waves within the LAP signal; Calculate the A / V ratio based on the average pressure amplitude of the set of A waves and the average pressure amplitude of the set of V waves; Calculate the mean maximum amplitude of the set of A waves, Calculate the mean maximum amplitude of the set of V waves, Calculate the average minimum amplitude of the inflection point in the transition of wave A to wave V in the LAP signal; Calculate the average minimum amplitude of the inflection point at the transition of wave V to wave A in the LAP signal; Calculate the difference, ΔMax, between the mean maximum amplitude of the set of A waves and the mean maximum amplitude of the set of V waves; Calculate the difference, ΔMin, between the average minimum amplitude of the inflection points in the transition of wave A to wave V and the average minimum amplitude of the inflection points in the transition of wave V to wave A; Calculate the mitral valve index (MVI) based on the A / V ratio, ΔMax, and ΔMin. and a memory configured to cause a patient therapy decision to be made based on the MVI crossing a threshold.

[0188] Example 108. The system of any of the embodiments described herein, particularly example 107, wherein the treatment is implantation of a prosthetic heart valve.

[0189] Example 109. The system of any of the embodiments described herein, particularly example 107, wherein the treatment is replacement of a prosthetic heart valve.

[0190] Example 110. The system of any of the embodiments described herein, particularly example 107, wherein the treatment is surgical percutaneous valve repair.

[0191] Example 111. MVI is

number

[0192] Example 112. Any of the embodiments described herein, particularly the system described in example 111, wherein C1 and C2 are between 1 and 5.

[0193] Example 113. A system according to any of the embodiments described herein, particularly example 107, wherein the threshold is 1 and an MVI>1 indicates a healthy patient not requiring treatment.

[0194] Example 114. An artificial heart valve, comprising: a sensor configured to record a left atrial pressure (LAP) signal; a transmitter configured to wirelessly transmit the LAP signal to an external monitoring device; The prosthetic heart valve, wherein the monitoring device is configured to determine, based on the LAP signal, whether the prosthetic heart valve requires replacement.

[0195] Example 115. To determine whether a prosthetic heart valve requires replacement, a monitoring device is provided. Identifying the set of A waves and the set of V waves within the LAP signal; Calculate the A / V ratio based on the average pressure amplitude of the set of A waves and the average pressure amplitude of the set of V waves; Calculate the mean maximum amplitude of the set of A waves, Calculate the mean maximum amplitude of the set of V waves, Calculate the average minimum amplitude of the inflection point in the transition of wave A to wave V in the LAP signal; Calculate the average minimum amplitude of the inflection point at the transition of wave V to wave A in the LAP signal; Calculate the difference, ΔMax, between the mean maximum amplitude of the set of A waves and the mean maximum amplitude of the set of V waves; Calculate the difference, ΔMin, between the average minimum amplitude of the inflection points in the transition of wave A to wave V and the average minimum amplitude of the inflection points in the transition of wave V to wave A; Calculate the mitral valve index (MVI) based on the A / V ratio, ΔMax, and ΔMin. The prosthetic heart valve of any embodiment described herein, particularly embodiment 114, further configured to determine that the prosthetic heart valve requires replacement when MVI<1.

[0196] Example 116. MVI is

number

[0197] Example 117. The prosthetic heart valve of any embodiment described herein, particularly example 115, wherein C1 and C2 are between 1 and 5.

[0198] Example 118. A method for monitoring pulmonary congestion in a patient, comprising: acquiring a plurality of left atrial pressure (LAP) signals using an atrial cardiac sensor implanted in the patient's left atrium, each LAP signal being acquired on a different day among a plurality of days; calculating an average daily LAP for each given day of the plurality of days based on the LAP signals acquired on the given day; If the daily average LAP is greater than the hypertension level value for a given day, calculating the difference between the daily average LAP and the hypertension level; and calculating a pulmonary congestion risk index based on the calculated difference.

[0199] Example 119 The method of any of the embodiments described herein, particularly example 118, wherein the number of days is consecutive.

[0200] Example 120 The method of any embodiment described herein, particularly example 118, further comprising recommending treatment if the pulmonary congestion risk index is greater than the alarm level.

[0201] Example 121. The method of any of the embodiments described herein, particularly example 120, wherein the hypertension level is 20 mmHg and the alarm level is 30 mmHg.

[0202] Example 122 The method of any embodiment described herein, particularly example 120, wherein the recommended treatment is mitral valve replacement.

[0203] Example 123 The method of any embodiment described herein, particularly example 120, wherein the recommended treatment is mitral valve repair.

[0204] Example 124. A method for monitoring pulmonary congestion in a patient, comprising: acquiring a plurality of left atrial pressure (LAP) signals using an atrial cardiac sensor implanted in the patient's left atrium, each LAP signal being acquired on a different day among a plurality of days; calculating an average daily LAP for each given day of the plurality of days based on the LAP signals acquired on the given day; calculating a moving average LAP based on a number of consecutive most recent daily average LAP values; If the moving average LAP is greater than the hypertension level value for a given day, calculating the difference between the moving average LAP and the hypertension level; and calculating a pulmonary congestion risk index based on the calculated difference.

[0205] Example 125 The method of any of the embodiments described herein, particularly example 124, wherein the number of days is consecutive.

[0206] Example 126 The method of any embodiment described herein, particularly example 118, further comprising recommending treatment if the pulmonary congestion risk index is greater than the alarm level.

[0207] Example 127. The method of any embodiment described herein, particularly example 120, wherein the hypertension level is 20 mmHg and the alarm level is between 0 and 200 mmHg.

[0208] Example 128 The method of any embodiment described herein, particularly example 126, wherein the recommended treatment is mitral valve replacement.

[0209] Example 129. The method of any of the embodiments described herein, particularly example 126, wherein the recommended treatment is mitral valve repair.

[0210] Example 130. The method of any of the embodiments described herein, particularly Example 124, wherein the number of consecutive most recent average daily LAP values ​​is between 2 and 10.

[0211] Example 131. A method for monitoring pulmonary congestion in a patient, comprising: acquiring a plurality of left atrial pressure (LAP) signals using an atrial cardiac sensor implanted in the patient's left atrium, each LAP signal being acquired on a different day among a plurality of days; calculating an average daily LAP for each given day of the plurality of days based on the LAP signals acquired on the given day; calculating a moving average LAP based on a first number of consecutive most recent daily average LAP values; calculating a LAP standard deviation based on a second number of consecutive most recent daily average LAP values; calculating the difference between the moving average LAP plus two LAP standard deviations and the hypertension level if the moving average LAP plus two LAP standard deviations is greater than the hypertension level value for a given day; and calculating a pulmonary congestion risk index based on the calculated difference.

[0212] Example 132 The method of any of the embodiments described herein, particularly example 131, wherein the number of days is consecutive.

[0213] Example 133 The method of any embodiment described herein, particularly example 131, further comprising recommending treatment if the pulmonary congestion risk index is greater than the alarm level.

[0214] Example 134. The method of any of the embodiments described herein, particularly example 133, wherein the hypertension level is 20 mmHg and the alarm level is between 50-60 mmHg.

[0215] Example 135 The method of any of the embodiments described herein, particularly example 133, wherein the recommended treatment is mitral valve replacement.

[0216] Example 136 The method of any of the embodiments described herein, particularly example 133, wherein the recommended treatment is mitral valve repair.

[0217] Example 137. The method described in any of the embodiments described herein, particularly example 131, wherein the first and number of consecutive most recent average daily LAP values ​​are between 2 and 10, and the first number is less than the second number.

[0218] Example 138. A system for monitoring pulmonary congestion in a patient, comprising: an atrial cardiac sensor configured to be implanted in a patient's left atrium to record a plurality of left atrial pressure (LAP) signals, each LAP signal of the plurality of LAP signals being acquired on a different day among a plurality of days; A transceiver comprising: wirelessly providing power to the atrial cardiac sensor; a transceiver configured to receive a LAP signal from an atrial cardiac sensor; A processor; A memory containing a cardiopulmonary monitoring application, the cardiopulmonary monitoring application causing a processor to: Acquire multiple LAP signals, calculating an average daily LAP for each given day of the plurality of days based on the LAP signals acquired on the given days; If the daily average LAP is greater than the hypertension level value for a given day, calculate the difference between the daily average LAP and the hypertension level; and a memory for instructing the system to calculate a pulmonary congestion risk index based on the calculated difference.

[0219] Example 139. Any of the embodiments described herein, particularly the system described in example 138, wherein the number of days is consecutive.

[0220] Example 140. Any of the embodiments described herein, particularly the system described in Example 138, wherein the cardiopulmonary monitoring application further instructs the processor to recommend therapy if the pulmonary congestion risk index is greater than the alarm level.

[0221] Example 141. A system according to any of the embodiments described herein, particularly example 140, wherein the hypertension level is 20 mmHg and the alarm level is 30 mmHg.

[0222] Example 142. The system of any of the embodiments described herein, particularly example 140, wherein the recommended treatment is mitral valve replacement.

[0223] Example 143. The system of any of the embodiments described herein, particularly example 140, wherein the recommended treatment is mitral valve repair.

[0224] Example 144. A system for monitoring pulmonary congestion in a patient, comprising: an atrial cardiac sensor configured to be implanted in a patient's left atrium to record a plurality of left atrial pressure (LAP) signals, each LAP signal of the plurality of LAP signals being acquired on a different day among a plurality of days; A transceiver comprising: wirelessly providing power to the atrial cardiac sensor; a transceiver configured to receive a LAP signal from an atrial cardiac sensor; A processor; A memory containing a cardiopulmonary monitoring application, the cardiopulmonary monitoring application causing a processor to: Acquire multiple LAP signals, calculating an average daily LAP for each given day of the plurality of days based on the LAP signals acquired on the given days; Calculate a moving average LAP based on the number of consecutive most recent daily average LAP values; If the moving average LAP is greater than the hypertension level value for a given day, calculate the difference between the moving average LAP and the hypertension level; and a memory for instructing the system to calculate a pulmonary congestion risk index based on the calculated difference.

[0225] Example 145. Any of the embodiments described herein, particularly the system described in example 144, wherein the number of days is consecutive.

[0226] Example 146. A system as described in any of the embodiments described herein, particularly example 144, wherein the cardiopulmonary monitoring application further instructs the processor to recommend therapy if the pulmonary congestion risk index is greater than the alarm level.

[0227] Example 147. Any of the embodiments described herein, particularly the system described in example 146, wherein the hypertension level is 20 mmHg and the alarm level is between 0 and 200 mmHg.

[0228] Example 148. The system of any of the embodiments described herein, particularly example 146, wherein the recommended treatment is mitral valve replacement.

[0229] Example 149. The system of any of the embodiments described herein, particularly example 146, wherein the recommended treatment is mitral valve repair.

[0230] Example 150. Any of the embodiments described herein, particularly the system described in example 124, wherein the number of consecutive most recent daily average LAP values ​​is between 2 and 10.

[0231] Example 151. A system for monitoring pulmonary congestion in a patient, comprising: an atrial cardiac sensor configured to be implanted in a patient's left atrium to record a plurality of left atrial pressure (LAP) signals, each LAP signal of the plurality of LAP signals being acquired on a different day among a plurality of days; A transceiver comprising: wirelessly providing power to the atrial cardiac sensor; a transceiver configured to receive a LAP signal from an atrial cardiac sensor; A processor; A memory containing a cardiopulmonary monitoring application, the cardiopulmonary monitoring application causing a processor to: Acquire multiple LAP signals, calculating an average daily LAP for each given day of the plurality of days based on the LAP signals acquired on the given days; Calculating a moving average LAP based on a first number of consecutive most recent daily average LAP values; Calculating the LAP standard deviation based on a second number of consecutive most recent daily average LAP values; calculating the difference between the moving average LAP plus two LAP standard deviations and the hypertension level if the moving average LAP plus two LAP standard deviations is greater than the hypertension level value for a given day; and a memory for instructing the system to calculate a pulmonary congestion risk index based on the calculated difference.

[0232] Example 152. Any of the embodiments described herein, particularly the system described in example 151, wherein the number of days in the plurality of days is consecutive.

[0233] Example 153. A system as described in any embodiment described herein, particularly example 151, wherein the cardiopulmonary monitoring application further instructs the processor to recommend therapy if the pulmonary congestion risk index is greater than the alarm level.

[0234] Example 154. Any of the embodiments described herein, particularly the system described in example 153, wherein the high blood pressure level is 20 mmHg and the alarm level is between 50 and 60 mmHg.

[0235] Example 155. The system of any of the embodiments described herein, particularly example 153, wherein the recommended treatment is mitral valve replacement.

[0236] Example 156. The system of any of the embodiments described herein, particularly example 153, wherein the recommended treatment is mitral valve repair.

[0237] Example 157. Any of the embodiments described herein, particularly the system described in Example 151, wherein the first and a number of consecutive most recent average daily LAP values ​​are between 2 and 10, and the first number is less than the second number.

[0238] Although specific systems and methods for cardiac health monitoring and treatment have been discussed above, many different systems and methods can be implemented in accordance with many different aspects of the invention. It is therefore to be understood that the invention may be practiced otherwise than as specifically described without departing from the scope and spirit of the invention. The present aspects are therefore to be considered in all respects as illustrative and not restrictive. The scope of the invention should therefore be determined not by the illustrated embodiments, but by the appended claims and their equivalents.

Claims

1. 1. A method of treating a cardiac condition in a patient, comprising: acquiring a left atrial pressure (LAP) signal using an atrial cardiac sensor implanted in the patient's left atrium; identifying a set of A waves and a set of V waves within the LAP signal; calculating an A / V ratio based on the average pressure amplitude of the set of A waves and the average pressure amplitude of the set of V waves; determining an appropriate treatment for the patient based on the A / V ratio; and providing said appropriate treatment to said patient.

2. 2. The method of treating a cardiac condition of a patient of claim 1, wherein identifying the set of A waves and the set of V waves in the LAP signal comprises decomposing the LAP signal.

3. 3. The method of treating a cardiac condition of a patient of claim 2, wherein decomposing the LAP signal comprises applying a Fourier transform.

4. 3. The method of treating a cardiac condition of a patient of claim 2, wherein decomposing the LAP signal comprises applying wavelet analysis.

5. 3. The method of treating a cardiac condition of a patient of claim 2, further comprising determining a respiratory rate of the patient by locating a dominant frequency in the decomposition between 0.1 and 0.5 Hz.

6. 3. The method of treating a cardiac condition of a patient of claim 2, further comprising determining a heart rate of the patient by locating a dominant frequency in the decomposition between 0.8 Hz and 1.7 Hz.

7. 3. The method of treating a cardiac condition of a patient of claim 2, further comprising subtracting frequency components associated with respiration rate from the LAP signal to produce a de-noised LAP signal.

8. 8. The method of treating a cardiac condition of a patient of claim 7, further comprising parsing the denoised LAP signal into a set of even waves and a set of odd waves.

9. 9. The method of treating a cardiac condition according to claim 8, wherein the even wave sets and the odd wave sets are exclusively labeled as A wave sets or V wave sets.

10. 10. The method of treating a cardiac condition of claim 9, wherein the exclusionary labeling is performed using a machine learning model trained using a training dataset that includes a denoised LAP signal annotated with A and V waves.

11. 10. The method of treating a cardiac condition according to claim 9, wherein the exclusionary marking is performed by waveform analysis, including identification of a double peak in the A wave.

12. 10. The method of treating a cardiac condition of claim 1, wherein the appropriate therapy comprises increasing the dosage of diuretic administered to the patient by between 40% and 60% in response to the A / V ratio decreasing to less than 0.

6.

13. 2. The method of treating a cardiac condition of claim 1, wherein the appropriate treatment comprises increasing the dosage of a hypertensive medication while maintaining the dosage of a diuretic to the patient in response to an increase in the patient's mean blood pressure of more than 40 mmHg without a change in A / V ratio.

14. 10. The method of treating a cardiac condition according to claim 1, wherein the appropriate therapy is a medication that is automatically delivered to the patient via an infusion pump.

15. 1. A system for treating a cardiac condition of a patient, comprising: A processor; A memory containing a cardiac monitoring application, the cardiac monitoring application causing the processor to: Obtaining a recorded left atrial pressure (LAP) signal using an atrial cardiac sensor; identifying a set of A waves and a set of V waves within the LAP signal; calculating an A / V ratio based on the average pressure amplitude of the set of A waves and the average pressure amplitude of the set of V waves; and a memory including an application configured to determine an appropriate therapy for the patient based on the A / V ratio.

16. 16. The system for treating a cardiac condition of a patient of claim 15, wherein the cardiac monitoring application further configures the processor to decompose the LAP signal to identify the set of A waves and the set of V waves in the LAP signal.

17. 20. The system for treating a cardiac condition of a patient of claim 16, wherein the decomposition comprises application of a Fourier transform.

18. 20. The system for treating a cardiac condition of a patient of claim 16, wherein the decomposition comprises application of wavelet analysis.

19. 17. The system for treating a cardiac condition of a patient of claim 16, wherein the cardiac monitoring application is further configured to cause the processor to determine a respiratory rate of the patient by locating a dominant frequency in the decomposition between 0.1 and 0.5 Hz.

20. 17. The system for treating a cardiac condition of a patient of claim 16, wherein the cardiac monitoring application is further configured to cause the processor to determine a heart rate of the patient by locating a dominant frequency in the decomposition between 0.8 Hz and 1.7 Hz.

21. 17. The system for treating a cardiac condition of a patient of claim 16, wherein the cardiac monitoring application further configures the processor to subtract frequency components associated with respiration rate from the LAP signal to generate a de-noised LAP signal.

22. 22. The system for treating a cardiac condition of a patient of claim 21, wherein the cardiac monitoring application further configures the processor to analyze the noise-removed LAP signal into an even wave set and an odd wave set.

23. 23. The system for treating a cardiac condition of claim 22, wherein the even wave sets and the odd wave sets are exclusively labeled as A wave sets or V wave sets.

24. 24. The system for treating a cardiac condition of claim 23, wherein the exclusionary marking is performed using a machine learning model trained using a training data set comprising a denoised LAP signal annotated with A and V waves.

25. 24. The system for treating a cardiac condition according to claim 23, wherein the exclusionary marking is performed by waveform analysis, including identification of a double peak in the A wave.

26. 16. The system for treating a cardiac condition of claim 15, wherein the appropriate therapy includes increasing a dose of diuretic administered to the patient by between 40% and 60% in response to the A / V ratio dropping below 0.6, and the cardiac monitoring application further configured to cause the processor to deliver the dose of diuretic via an infusion pump communicatively coupled to the processor.

27. 16. The system for treating a cardiac condition of claim 15, wherein the appropriate therapy includes increasing a dosage of a hypertension medication while maintaining a dosage of a diuretic to the patient in response to an increase in the patient's mean blood pressure of more than 40 mmHg without a change in A / V ratio, and the cardiac monitoring application is further configured to cause the processor to deliver the dosage of diuretic via an infusion pump communicatively coupled to the processor.

28. 16. The system for treating a cardiac condition of claim 15, wherein the appropriate therapy is a medication that is automatically delivered to the patient via an infusion pump communicatively coupled to the processor.

Images