A method and system for estimating the remaining life of an implantable medical device battery

Abstract

The application discloses a kind of implantable medical equipment battery remaining life estimation method and system, comprising: collecting the electric quantity information in equipment operation, including the running time of each hardware module in equipment, heart pacing frequency and data transmission length, based on electric quantity information and single pacing pulse energy, from three aspects of cumulative power consumption, event-triggered power consumption and data transmission type power consumption;Based on the historical pacing data consisting of heart pacing times, calculate the pacing dependency index and the pacing fluctuation index and construct the time evolution model of pacing dependency, combine the adjustment coefficient and the pacing fluctuation index and the pacing dependency index determined based on the time evolution model, predict the pacing frequency and power consumption under different working conditions in the future;Based on the remaining capacity of battery determined by cumulative power consumption and predicted power consumption, the remaining service life of equipment battery is obtained by discrete time step decrement calculation.This can improve the accuracy of battery remaining life estimation.

Description

Technical Field

[0001] This invention belongs to the field of implantable medical device technology, specifically relating to a method and system for estimating the remaining battery life of implantable medical devices. Background Technology

[0002] With the continuous development of medical electronics and miniaturization technologies, implantable medical devices (IMDs) are increasingly widely used in clinical practice, becoming an important technical means for the long-term management of chronic diseases. These devices, implanted in the patient's body, can continuously monitor physiological signals and execute corresponding interventions when abnormalities are detected, thereby achieving long-term control and treatment of diseases. Compared to traditional methods relying on hospital monitoring and manual intervention, IMDs can significantly reduce the frequency of patient visits, improve convenience, and provide doctors with more continuous and accurate physiological data. Among the many applications of IMDs, the cardiac pacemaker is a typical example, mainly used to treat arrhythmias such as bradycardia. It senses the heart's electrical activity and delivers electrical pulses when necessary to maintain a stable heart rhythm, which is crucial for ensuring patient safety.

[0003] Because pacemakers need to operate stably within the body for extended periods, their power system relies entirely on an internal battery. As the device's sole energy source, the battery's performance directly impacts its lifespan and reliability. When the battery runs out of power, surgical replacement is typically required, increasing the patient's physical burden and surgical risks, as well as incurring additional medical costs. Therefore, device design must prioritize maximizing battery energy utilization efficiency within limited space constraints, while simultaneously accurately assessing the battery's remaining lifespan. This allows doctors and patients to plan replacements in advance, mitigating the risks of sudden device failure.

[0004] In practical applications, the factors affecting battery life are diverse and complex. On the one hand, the battery's material system and capacity parameters determine its initial energy reserve. Currently commonly used battery types include lithium-iodine batteries (Li-I) and lithium-carbon fluoride batteries (Li-CFx), which feature high energy density, low self-discharge rate, and long service life. On the other hand, the power consumption characteristics of the device during operation play a decisive role in battery energy consumption. Specifically, the device consists of multiple functional modules, including application-specific integrated circuits (ASICs), microcontrollers, storage units, and communication modules. These modules exhibit significant differences in power consumption under different operating modes. For example, in low-power standby or monitoring states, the system current is small, while in states such as pacing pulse delivery, data processing, or wireless communication, the current increases significantly in a short period of time. Furthermore, the energy consumption of the pacing pulse itself is closely related to parameters such as pacing amplitude, pulse width, and electrode impedance, resulting in a non-linear variation in the energy consumption of a single pacing event.

[0005] Furthermore, due to individual patient differences, the degree of pacemaker dependence varies significantly among patients. Some patients may only require pacing support under specific circumstances, while others are highly dependent on pacemakers to maintain a basic heart rate. This directly leads to large differences in the number of pacing cycles between individuals. Simultaneously, a patient's physiological state is not constant; factors such as activity level, disease progression, and changes in cardiac function can all cause pacing needs to exhibit dynamic changes over time. That is, pacing frequency and mode may fluctuate at different times. This dual characteristic of "individual differences + time variations" results in a complex pattern of multi-source superposition and dynamic changes in device power consumption.

[0006] Therefore, under the combined effects of the aforementioned multiple factors, the energy consumption process of the battery not only exhibits significant nonlinear characteristics but also demonstrates discreteness and uncertainty in the time dimension, making it extremely difficult to accurately predict the remaining battery life. How to comprehensively consider the multi-source power consumption characteristics of the device and the dynamic changes in patient usage behavior under limited computing resources to make a more accurate and reliable estimate of the remaining battery life has become a key technical problem that urgently needs to be solved in the field of implantable medical devices.

[0007] In the prior art, as described in patent application CN111601639A, a method for estimating the remaining lifespan of a power supply for implantable medical devices is proposed. This method obtains power supply parameters (such as battery voltage) and device operating parameters, and calculates the estimated duration for the power supply to reach a pre-recommended replacement time (pre-RRT) threshold and a recommended replacement time (RRT) threshold, respectively. Specifically, a first estimated lifespan value is obtained by calculating the time to the pre-RRT threshold based on the power supply parameters and adding a preset timer duration; a second estimated lifespan value is obtained by calculating the time to the RRT threshold based on the power supply parameters. The smaller of the two values ​​is then used as the estimated remaining lifespan of the power supply. Simultaneously, this method incorporates a timer mechanism, starting timing after the power supply reaches the pre-RRT threshold, and triggering RRT when either the timer expires or the power supply parameters reach the corresponding threshold are met.

[0008] Prior art, such as patent application US20120130439A1, proposes a hybrid battery life monitoring system and method for implantable medical devices. Its core lies in simultaneously employing two battery life estimation techniques: first, an energy counter accumulates the energy consumed during device operation to generate an energy-based remaining life estimate; second, a voltage monitor monitors the battery voltage in real time to generate a voltage-based remaining life estimate. A calculator dynamically selects which estimation result to use based on the battery's lifespan stage—in the early stages of battery life (when the voltage curve is flat and the voltage method is difficult to accurately determine), the energy-based estimate is used; in the later stages of battery life (when the voltage begins to drop sharply and the voltage method becomes more accurate), the voltage-based estimate is switched to.

[0009] Existing technology, such as patent application CN105960263A, proposes a method for dynamically monitoring the battery life of implantable medical devices. The core of this method is that when the battery voltage first drops to the Pre-RRT threshold, the device queries a pre-calculated lookup table based on the current actual usage conditions (including parameters such as pacing pulse amplitude, pulse width, and pacing percentage) to determine an initial "remaining lifespan duration" value, and triggers subsequent lifespan indicator states (such as RRT, ERI, and EOS) based on this value. During the period from the triggering of Pre-RRT to the triggering of ERI, the device monitors changes in usage conditions daily. If parameters such as the pacing percentage change significantly, the lookup table is queried again, and the remaining lifespan duration is updated to the smaller of the current lookup value and the previous remaining value (i.e., only decreasing, not increasing), thereby dynamically adjusting the remaining lifespan estimate.

[0010] While existing technologies can estimate battery life to some extent, they still have significant shortcomings: First, they primarily model based on state parameters such as battery voltage, internal resistance, or capacity, failing to fully consider the impact of actual device usage on power consumption, leading to discrepancies between predicted results and actual usage. Second, power consumption processing typically employs a holistic or averaging approach, failing to differentiate between different types of power consumption such as pacing pulses, hardware operation, and data communication, making it difficult to accurately reflect the actual energy consumption process. Third, they do not consider the discreteness and dynamic changes in power consumption over time, making it difficult to characterize the impact of instantaneous high-power events on lifespan. Fourth, they lack modeling for individual patient differences and changes in pacing needs, making personalized predictions impossible. Fifth, they typically only output a single lifespan estimate, lacking a description of the range of lifespan variations, thus limiting their application value in clinical decision-making to some extent. Summary of the Invention

[0011] In view of the above, in order to improve the accuracy of estimating the remaining life of implantable pacemaker batteries, the present invention provides a method and system for estimating the remaining life of implantable medical device batteries. Based on the power statistics during device operation and the patient's pacing behavior characteristics, the battery energy consumption is cumulatively calculated and the future power consumption is predicted, thereby realizing the dynamic estimation of the remaining life of the battery.

[0012] To achieve the above-mentioned objectives, an embodiment provides a method for estimating the remaining battery life of an implantable medical device, comprising the following steps: Collect power information during device operation, including the running time of each hardware module in the device, the number of heart pacings, and the data transmission length. Based on the power information and the energy of a single pacing pulse, calculate the cumulative power consumption from three aspects: continuous power consumption, event-triggered power consumption, and data transmission power consumption. Based on historical pacing data consisting of cardiac pacing counts, a pacing dependence index and a pacing variability index are calculated, and a time evolution model of pacing dependence is constructed. By combining the adjustment coefficient, the pacing variability index, and the pacing dependence index determined based on the time evolution model, the number of pacing counts and power consumption under different working conditions in the future can be predicted. Based on the remaining battery capacity determined by the cumulative power consumption and the predicted power consumption, the remaining lifespan of the device battery is calculated by decreasing the discrete time step.

[0013] Preferably, the energy of a single pacing pulse is calculated based on the pacing amplitude, pulse width, electrode impedance, and battery voltage; Alternatively, a lookup table combined with an interpolation algorithm can be used to pre-store pulse energy values ​​for typical pacing amplitude, pulse width, electrode impedance, and battery voltage combinations, and the corresponding pulse energy can be quickly obtained during operation through interpolation.

[0014] Preferably, based on power information and the energy of a single pacing pulse, the cumulative power consumption is calculated from three aspects: continuous power consumption, event-triggered power consumption, and data transmission power consumption, including: The cumulative power consumption is the sum of continuous power consumption, event-triggered power consumption, and data transmission power consumption; Among them, continuous power consumption is the product of the power consumption per unit time of each hardware module and the running time; Event-triggered power consumption is the sum of the product of the number of atrial pacings and the energy of the corresponding atrial pacing pulse, and the product of the number of ventricular pacings and the energy of the corresponding ventricular pacing pulse. The power consumption for data transmission is the product of the data transmission length and the power consumption per unit data length.

[0015] Preferably, the pacing dependence index and pacing variability index are calculated based on historical pacing data consisting of the number of cardiac pacings, including: in, The pacing dependence index. The pacing volatility index, and The values ​​are divided into the average number of atrial and ventricular pacings within a preset time window. This represents the theoretical maximum number of pacing cycles per unit of time. and Let be the weight coefficient, and satisfy... , and It is divided into the standard deviation of the number of atrial and ventricular pacing.

[0016] Preferably, a time evolution model of pacing dependence is constructed to describe the dynamic changes of pacing dependence over time; The time evolution model adopts a linear model, an exponential model, a constant model, a polynomial model, or a piecewise function model to accommodate individual differences among different patients; Specifically, based on historical pacing data, the changing trend of pacing dependence is analyzed. By calculating the rate of change, slope, or fluctuation characteristics of pacing dependence over time, the corresponding time evolution model parameters are selected or adjusted so that the model can adaptively reflect the patient's current and future physiological changes.

[0017] Preferably, by combining the adjustment coefficient, the pacing variability index, and the pacing dependence index determined based on a time evolution model, the number of pacing cycles under different operating conditions in the future is predicted, including: First, calculate the future... Baseline pacing counts per time step : in, This represents the average number of pacing cycles. Represents the theoretical maximum number of pacing cycles, the difference term This indicates the potential growth space for pacing demand. The pacing dependence index is determined based on a time evolution model; Then, short-term fluctuation factors are superimposed on the baseline pacing count to obtain a unified pacing count prediction model to predict the pacing count under different working conditions in the future: in, The pacing volatility index, For the predicted number of pacing sessions, This is an adjustment coefficient used to characterize different power consumption conditions. When, it means that the impact of short-term fluctuations is not considered; when When this occurs, it indicates that the pacing demand is higher than average, corresponding to a higher power consumption condition; when When the pacing demand is below average, it indicates a lower power consumption condition.

[0018] Preferably, calculating the predicted power consumption based on the predicted number of pacing cycles includes: in, For the future i Predicted power consumption at time step and Each of the future i The continuous power consumption and data transmission power consumption of the time step. and These represent the predicted number of atrial and ventricular pacing cycles, respectively. and These represent the energy of a single atrial and ventricular pacing, respectively.

[0019] Preferably, based on the remaining battery capacity determined by cumulative power consumption and the predicted power consumption, the remaining battery life is calculated by decreasing the discrete time step, including: in, For the first The remaining battery capacity at the time step. To predict power consumption, For the first The remaining battery capacity at the time step. The adjustment coefficient is used to characterize different power consumption conditions; When the following conditions are met: At that time, corresponding time step That is, the remaining service life under power consumption conditions. .

[0020] Preferably, by selecting different adjustment coefficients Within a unified computing framework, the maximum lifespan, average lifespan, and minimum lifespan can be obtained, thus generating interval prediction results for battery lifespan. And satisfy ,in For the remaining battery life, and These are the minimum lifespan and the maximum lifespan, respectively. Average lifespan; When the adjustment coefficient Continuous values ​​are used to represent lifetime prediction results at different confidence levels.

[0021] To achieve the above-mentioned objectives, embodiments of the present invention provide a system for estimating the remaining battery life of an implantable medical device, comprising: The power information statistics module is used to collect power information during device operation, including the running time of each hardware module in the device, the number of heartbeats, and the data transmission length. The pulse energy calculation module is used to calculate the energy of a single pacing pulse. The cumulative power consumption calculation module is used to calculate the cumulative power consumption from three aspects: continuous power consumption, event-triggered power consumption, and data transmission power consumption, based on power information and the energy of a single pacing pulse. The future power consumption prediction module is used to calculate the pacing dependence index and pacing fluctuation index based on historical pacing data consisting of the number of heart pacings and to construct a time evolution model of pacing dependence. Combining the adjustment coefficient, pacing fluctuation index and pacing dependence index determined based on the time evolution model, it predicts the number of pacings and power consumption under different working conditions in the future. The remaining life estimation module is used to calculate the remaining lifespan of the device battery by using the battery's remaining capacity determined by the cumulative power consumption and the predicted power consumption, through discrete time step decrement.

[0022] Compared with the prior art, the beneficial effects of the present invention include at least the following: 1. This invention estimates battery life based on power information statistics and actual power consumption modeling, without relying on battery voltage or discharge characteristic curves. This effectively avoids errors caused by the unstable voltage-capacity relationship in the early stages of battery use, thereby improving the accuracy and stability of remaining life estimation.

[0023] 2. This invention divides device power consumption into hardware operation power consumption, event-triggered power consumption, and data transmission power consumption, and uses different methods for modeling and calculation of each, thereby achieving a refined description of multi-source heterogeneous power consumption, which can more realistically reflect the energy consumption of the device during actual operation.

[0024] 3. This invention introduces a pacing dependence index to quantitatively model the degree of patient dependence on pacemakers, enabling life prediction to be dynamically adjusted according to changes in the patient's physiological state, thereby improving the individual adaptability of the prediction results.

[0025] 4. This invention introduces a pacing fluctuation index to characterize the short-term fluctuation characteristics of pacing demand and incorporates it into power consumption prediction, thereby reflecting the range of power consumption fluctuation during actual use and improving the predictive model's adaptability to complex usage scenarios.

[0026] 5. By setting adjustment coefficients, this invention constructs multiple power consumption conditions under a unified model framework, enabling unified calculation of lifespan under different usage scenarios. This yields the maximum, average, and minimum battery lifespan, providing interval-based prediction results and improving the reference value and reliability of the prediction results.

[0027] 6. This invention uses a discrete-time recursive method to gradually update the remaining battery capacity, which can intuitively reflect the dynamic process of battery energy consumption over time. At the same time, it avoids relying on complex battery models and has the advantages of simple implementation and high computational efficiency, making it suitable for resource-constrained implantable devices.

[0028] 7. The method of the present invention can be adaptively adjusted based on historical pacing data, and has good scalability and versatility, and can be applied to different patients, different pacing modes and different equipment configurations. Attached Figure Description

[0029] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0030] Figure 1 This is a system structure block diagram according to an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the relationship between battery voltage and battery capacity. Figure 3 This is a schematic diagram of the current change curve during actual operation of an embodiment of the present invention; Figure 4This is a flowchart illustrating the implementation of the remaining lifetime estimation algorithm in this embodiment of the invention. Figure 5 This is a schematic diagram of the overall structure of the battery remaining life estimation system in an embodiment of the present invention; Figure 6 This is a flowchart illustrating the overall process of estimating battery remaining life in an embodiment of the present invention. Detailed Implementation

[0031] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not limit the scope of protection of this invention.

[0032] Existing technologies that rely on battery voltage or discharge curves to estimate remaining capacity suffer from inaccurate estimates because there is no clear correlation between voltage and capacity in the early stages of battery use (such as the PSP stage). Existing technologies also fail to provide detailed modeling of the multi-source power consumption of implantable medical devices under different operating modes, making it difficult to accurately reflect the impact of various energy consumption factors, such as hardware operation, pacing pulses, and data communication, on total power consumption. Furthermore, existing technologies typically rely solely on historical average power consumption for lifespan prediction, neglecting the dynamic changes in patient pacing needs, resulting in predictions that fail to reflect actual usage scenarios. Finally, the lack of modeling methods for individual patient differences prevents the quantification of patient dependence on the pacemaker, further impacting the accuracy of lifespan predictions.

[0033] To address the aforementioned technical problems, this invention provides a scheme for estimating the remaining battery life of implantable medical devices, which can be applied to applications such as... Figure 1 The implantable medical device system shown includes an in-vivo device 101 and external devices 108 and 109. The in-vivo device 101 includes a dedicated integrated chip 102, a microcontroller 103, an off-chip memory 104, a low-power Bluetooth chip 105, an accelerometer 106, and a battery 107; the external devices 108 and 109 can be a programmable controller and a mobile terminal, respectively.

[0034] The dedicated integrated chip 102 is used to sense and process cardiac electrical signals, as well as generate and output pacing pulses. Specifically, it amplifies, filters, and digitizes the ECG signals, and outputs pacing pulses with specific amplitude, pulse width, and frequency according to set parameters. It can also detect parameters such as battery voltage, battery internal resistance, and electrode impedance. The microcontroller 103 is used to implement system control and data processing functions, including acquiring and storing cardiac activity data, managing the working status of each hardware module, performing data interaction, and performing power consumption statistics and battery life estimation. In other words, the battery remaining life estimation scheme of this invention is centrally executed in the microcontroller 103. The off-chip memory 104 is a non-volatile memory used to store patient ECG data, event records, and device configuration parameters, maintaining data integrity even in the event of a power outage.

[0035] An accelerometer 106 is used to sense the patient's motion state and positional changes, thereby assisting in adjusting pacing parameters to improve pacing fit and patient comfort. A low-power Bluetooth chip 105 is used to enable wireless communication between the device and an external terminal, supporting data transmission and remote monitoring functions. A battery 107 provides the energy required for the operation of the various hardware modules of the in-body device. In the external device, a programmer 108 is used to configure parameters, read data, and monitor the status of the implanted device, supporting pacing parameter adjustment and abnormal status alerts; a mobile terminal 109 communicates with the in-body device via Bluetooth to achieve remote data acquisition and alarm information alerts.

[0036] like Figure 2 As shown, a battery performance curve applicable to implantable medical devices (IMDs) is presented, which describes the relationship between battery voltage and consumed capacity under a given load condition. Figure 2 The system marks several key status points, including Beginning of Service (BOS), Elective Replacement Indicator (ERI), and End of Service (EOS). BOS indicates the start of clinical use of the device; ERI indicates that the battery is about to run out, during which the device can continue to operate normally for a preset period to allow for replacement scheduling; EOS indicates that the battery is completely depleted and the device will no longer be able to operate normally. The period between BOS and ERI is the Prolonged Service Period (PSP), and the period between ERI and EOS is the Projected Service Period (PSL). Figure 2 It is known that during the PSP stage, there is no obvious linear relationship between battery voltage and remaining capacity, so it is difficult to accurately estimate the remaining battery capacity based solely on battery voltage.

[0037] Therefore, this embodiment provides a method for estimating the remaining battery life of an implantable medical device, such as... Figure 6 As shown, it includes the following steps: S1 collects power information during device operation, including the running time of each hardware module in the device, the number of heart pacings, and the data transmission length. Based on the power information and the energy of a single pacing pulse, the cumulative power consumption is calculated from three aspects: continuous power consumption, event-triggered power consumption, and data transmission power consumption.

[0038] like Figure 3 As shown, the equipment exhibits different current variation characteristics during actual operation, corresponding to different working states and operating modes. Based on the time characteristics and triggering mechanisms of current changes, energy consumption can be divided into three categories: continuous power consumption, event-triggered power consumption, and data transmission power consumption, and each category is calculated using different power consumption statistics methods.

[0039] First, obtain the power consumption information of the device during operation, including the operating time of each hardware module in different working modes, the number of atrial and ventricular pacing pulses delivered (collectively referred to as cardiac pacing counts), and the amount of data transmission between the hardware modules. Based on the above power consumption information, calculate the cumulative power consumption of the device since power-on.

[0040] Figure 3 The current change at point 301 is gradual and stable, indicating that the device is in a low-power operating state. At this time, the dedicated integrated chip 102, microcontroller 103, off-chip memory 104, and accelerometer 106 are all in low-power mode, and the low-power Bluetooth module 105 is off. The overall current remains at a low level. This addresses the issue of continuous power consumption. The calculation method is based on time accumulation, specifically the product of the power consumption per unit time and the running time of each hardware module, expressed as: in, Power consumption per unit time This refers to the runtime of the corresponding operating state. In practice, during the equipment manufacturing phase, the static current of each hardware module under different operating modes is calibrated and tested, and a lookup table is created and stored in the microcontroller 103. During equipment operation, by recording the duration of each hardware module under different operating states and combining this data with the lookup table data, accurate cumulative calculation of continuous power consumption is achieved.

[0041] The currents marked 302 and 303 in the diagram exhibit a characteristic of instantaneous peak followed by a rapid decline, corresponding to the pacing pulse delivery process. 302 corresponds to lower amplitude pacing, and 303 to higher amplitude pacing. This relates to the power consumption of this type of event-triggered pulse. The calculation method is based on frequency statistics, specifically calculated by multiplying the power consumption per unit of pacing by the number of pacing cycles. More specifically, it is the sum of the product of the number of atrial pacing cycles and the corresponding energy of a single atrial pacing pulse, and the product of the number of ventricular pacing cycles and the corresponding energy of a single ventricular pacing pulse, expressed as: in, , These represent the number of atrial and ventricular pacing cycles, respectively. , This refers to the energy of a single pacing pulse in the atrium and ventricle.

[0042] The energy of a single pulse is related to the pacing amplitude, pulse width, electrode impedance, and battery voltage, and can be calculated based on these parameters. Considering the limited computing resources of embedded systems, this invention preferably uses a lookup table combined with an interpolation algorithm to pre-store the pulse energy values ​​for typical parameter combinations. During runtime, the corresponding pulse energy is quickly obtained through interpolation, thereby reducing computational complexity.

[0043] The current at point 304 in the diagram represents a short-term peak, corresponding to device communication processes, including RF communication, wired interface communication, and Bluetooth communication. This addresses the power consumption associated with this type of data transmission. The calculation method is based on data volume statistics, specifically the product of data transmission length and power consumption per unit data length, expressed as: in, Power consumption per unit data length This represents the data transmission length. Due to the differences in power consumption characteristics among different communication methods, the unit data energy of each communication interface was calibrated during the experimental phase, and a lookup table was created and stored in the device. In actual operation, the cumulative calculation of communication energy was achieved by statistically analyzing the data transmission volume of each communication module and combining it with the lookup table.

[0044] In summary, the cumulative power consumption of the device over any given time period It can be represented as: By using the above classification modeling method, the complex current change process is decomposed into three quantifiable energy statistical models. This not only accurately reflects the power consumption characteristics under different operating conditions, but also effectively reduces computational complexity, providing a reliable data foundation for subsequent future power consumption prediction based on pacing behavior.

[0045] When combined with the initial battery capacity This will give you the current remaining battery capacity. : S2 calculates the pacing dependence index and pacing fluctuation index based on historical pacing data consisting of the number of heart pacings, and constructs a time evolution model of pacing dependence. Combining the adjustment coefficient, pacing fluctuation index, and pacing dependence index determined based on the time evolution model, it predicts the number of pacings and power consumption under different operating conditions in the future. Based on the remaining battery capacity determined by the cumulative power consumption and the predicted power consumption, it calculates the remaining lifespan of the device battery by decreasing the discrete time step.

[0046] In this embodiment of the invention, based on the current remaining battery capacity of the device, in order to accurately predict future power consumption, a pacing behavior model is constructed based on historical pacing data, and a discrete-time recursive method is combined to estimate the remaining battery life. Figure 4 As shown, the specific process for estimating the remaining battery life includes: Step 401: Initialize system parameters, including remaining battery capacity. Time step i and the adjustment coefficients corresponding to different power consumption conditions. k ; Step 402: Calculate the pacing volatility index (PVI) based on historical pacing data within a preset time window to characterize the short-term volatility of pacing demand. Steps 403-406: During the discrete-time recursion process, for each time step, the pacing dependence index at the current moment is calculated sequentially. basal pacing count and the revised predicted number of pacemakers And based on this, calculate the energy consumption at that time step. ; Step 407, check the remaining battery capacity The battery capacity is gradually reduced as time progresses. Step 408: Determine the remaining capacity Should the value be reduced to below the preset threshold? Step 409, when determining the remaining capacity When the value drops below a preset threshold, the current time step is recorded as the remaining service life under the corresponding operating condition. .

[0047] Specifically, the number of atrial and ventricular pacing events per day for each patient is counted within a preset time window and recorded as follows: and ,in And calculate its average value. and The preset time window is preferably 15 to 30 days.

[0048] Furthermore, the standard deviation of the number of atrial and ventricular pacing events was calculated. and : Based on the above statistical results, the Pacing Dependency Index (PDI) is defined as follows: in, This represents the theoretical maximum number of pacing cycles per unit of time. and Let be the weight coefficient, and satisfy... The physical meaning of PDI is the ratio of a patient's current pacing demand to the theoretical maximum pacing capacity. It is used to reflect the patient's long-term dependence on the pacemaker, thereby establishing a quantitative mapping relationship between the patient's physiological state and the device's energy consumption.

[0049] Furthermore, the Pacing Variability Index (PVI) is defined as follows: The physical meaning of PVI is the normalized fluctuation of the number of pacing cycles relative to its average value. It describes the magnitude of change in pacing demand over a short timescale, thus improving power consumption prediction from static estimation to dynamic range estimation. A large PVI indicates strong fluctuations in pacing demand; a small PVI indicates relatively stable pacing demand.

[0050] Given the dynamic changes in patients' physiological state during long-term follow-up after pacemaker implantation, their dependence on the pacemaker is not constant but may exhibit various trends over time. On the one hand, as the function of the cardiac conduction system gradually deteriorates, some patients experience a decline in their spontaneous rhythm-regulating ability, leading to a gradual increase in pacemaker dependence. On the other hand, with medication, lifestyle adjustments, or improvements in cardiac function, patients' spontaneous pacing ability may also recover to some extent, thus stabilizing or even temporarily reducing pacemaker dependence. Therefore, a single fixed parameter is insufficient to accurately characterize the changing patterns of pacemaker dependence.

[0051] Therefore, considering the differences in physiological states among different patients, the changing trends of pacing dependence over time are not entirely the same. A time evolution model of pacing dependence is introduced. The model describes the dynamic changes in pacing dependence over time. The pacing dependence time evolution model is not limited to linear, exponential, or constant models; multinomial and piecewise function models can also be used to accommodate individual differences among patients. Furthermore, in practical implementation, the changing trend of pacing dependence can be analyzed based on historical multi-period data. For example, by calculating its rate of change, slope, or fluctuation characteristics over time, the corresponding evolution model parameters can be selected or adjusted, enabling the model to adaptively reflect the patient's current and future physiological changes and improve model adaptability.

[0052] Based on this, first calculate the future number... Baseline pacing counts per time step : in, This represents the average number of pacing cycles. Represents the theoretical maximum number of pacing cycles, the difference term This indicates the potential growth space for pacing demand.

[0053] Furthermore, short-term fluctuation factors are superimposed on the baseline pacing frequency, and an adjustment coefficient is introduced. A unified pacing frequency prediction model was obtained to predict pacing frequency. : in, This is an adjustment coefficient used to characterize different power consumption conditions. Specifically, when... When the time is not considered, it means that the impact of short-term fluctuations is not taken into account. At this time, the number of pacing times is equal to the baseline number of pacing times, reflecting the average power consumption condition; when When this occurs, it indicates that the pacing demand is higher than average, corresponding to a higher power consumption condition; when This indicates that the pacing demand is below average, corresponding to a lower power consumption condition. In a preferred embodiment, the adjustment coefficient takes a discrete set of values. These correspond to three typical operating scenarios: low power consumption, average power consumption, and high power consumption.

[0054] Based on the above prediction results of pacing frequency, and considering the device power consumption structure, the future number of pacing cycles is predicted. Energy consumption at each time step is calculated. This energy consumption is still decomposed according to the three models used in the cumulative power consumption calculation, including hardware operating power consumption, pacing pulse power consumption, and data transmission power consumption. The pacing pulse power consumption is calculated from the predicted number of pacing cycles and the energy of a single pulse. Hardware operating power consumption and data transmission power consumption can be averaged based on historical statistical results or dynamically adjusted according to changes in the device's operating mode. Therefore, in the future... Total power consumption of time step It can be represented as: in, This indicates the hardware power consumption during that time step. Indicates data transmission power consumption. , These represent the energy of a single atrial and ventricular pacing, respectively.

[0055] Furthermore, by performing recursive calculations over multiple discrete time steps, the power consumption sequence over a future period can be obtained. This sequence serves as input for subsequent remaining lifetime estimation. Let the initial capacity be... Then we have: As the time step increases, the battery capacity gradually decreases, until the following condition is met: At that time, corresponding time step This refers to the remaining service life under this operating condition: By selecting different adjustment coefficients Maximum lifetime can be achieved within a unified computing framework. Average lifespan and minimum lifespan This results in an interval prediction of battery life: And satisfy Furthermore, the adjustment coefficient Continuous values ​​can also be used to represent lifetime prediction results at different confidence levels, thereby improving the flexibility and applicability of the prediction results.

[0056] like Figure 5As shown in the embodiment, an implantable medical device battery remaining life estimation system includes: a control and scheduling module 501, used for periodically scheduling each functional module and triggering data updates and life estimation processes; a power information statistics module 502, used for collecting power information such as the working time of each hardware module, the number of atrial and ventricular pacings, and the amount of data transmission during device operation; a pulse energy calculation module 503, used for calculating the energy of a single pacing pulse based on pacing amplitude, pulse width, electrode impedance, and battery voltage, preferably using a lookup table combined with an interpolation algorithm; and a future power consumption prediction module 504, used for calculating based on historical pacing data. The system includes a pacing dependence index (PDI) and a pacing volatility index (PVI), and constructs a time evolution model of pacing dependence. Combining the adjustment coefficient, the pacing volatility index, and the pacing dependence index determined based on the time evolution model, it predicts the number of pacing cycles and power consumption under different operating conditions in the future. A cumulative power consumption calculation module 505 is used to calculate the cumulative power consumption from three aspects: continuous power consumption, event-triggered power consumption, and data transmission power consumption, based on power information and the energy of a single pacing pulse. A remaining lifespan estimation module 506 is used to calculate the remaining lifespan of the device battery based on the remaining battery capacity determined by the cumulative power consumption and the predicted power consumption, by decreasing the discrete time step.

[0057] The solution provided in this invention constructs a battery life estimation model based on actual usage behavior, transforming battery energy consumption from traditional battery state-driven to usage behavior-driven, so as to more realistically reflect the energy changes during device operation; it performs structured modeling of device power consumption, dividing power consumption into hardware operation power consumption, pacing pulse power consumption, and data transmission power consumption, and uses different calculation methods to accurately count and accumulate them.

[0058] The solution provided in this invention constructs a pacing dependence index based on historical pacing data to characterize the patient's long-term dependence on the pacemaker and to serve as a trend driver for future power consumption changes; and constructs a pacing fluctuation index based on the statistical characteristics of the number of pacings within a preset time window to characterize the fluctuation characteristics of pacing behavior on a short time scale and to reflect the uncertainty of power consumption. The solution provided in this invention, under a unified model framework, introduces the pacing dependence index and pacing fluctuation index as trend and fluctuation terms into the power consumption prediction process, respectively, to achieve collaborative modeling of long-term power consumption trends and short-term fluctuation characteristics. Based on the pacing behavior model, it predicts the future number of pacing cycles and combines multi-source power consumption calculations to realize a power consumption prediction mechanism driven by behavior prediction. It also uses a discrete-time recursive method to dynamically update the remaining battery capacity, simulating the gradual decay of battery energy over time. By introducing adjustment coefficients to construct different power consumption conditions, the maximum lifespan, average lifespan, and minimum lifespan are obtained under a unified calculation framework, achieving interval-based estimation of battery lifespan.

[0059] The specific embodiments described above illustrate the technical solution and beneficial effects of the present invention in detail. It should be understood that the above description is only the most preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, additions, and equivalent substitutions made within the scope of the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for estimating the remaining battery life of an implantable medical device, characterized in that, Includes the following steps: Collect power information during device operation, including the running time of each hardware module in the device, the number of heart pacings, and the data transmission length. Based on the power information and the energy of a single pacing pulse, calculate the cumulative power consumption from three aspects: continuous power consumption, event-triggered power consumption, and data transmission power consumption. Based on historical pacing data consisting of cardiac pacing counts, a pacing dependence index and a pacing variability index are calculated, and a time evolution model of pacing dependence is constructed. By combining the adjustment coefficient, the pacing variability index, and the pacing dependence index determined based on the time evolution model, the number of pacing counts and power consumption under different working conditions in the future can be predicted. Based on the remaining battery capacity determined by the cumulative power consumption and the predicted power consumption, the remaining lifespan of the device battery is calculated by decreasing the discrete time step.

2. The method for estimating the remaining battery life of an implantable medical device according to claim 1, characterized in that, The energy of a single pacing pulse is calculated based on the pacing amplitude, pulse width, electrode impedance, and battery voltage. Alternatively, a lookup table combined with an interpolation algorithm can be used to pre-store pulse energy values ​​for typical pacing amplitude, pulse width, electrode impedance, and battery voltage combinations, and the corresponding pulse energy can be quickly obtained during operation through interpolation.

3. The method for estimating the remaining battery life of an implantable medical device according to claim 1, characterized in that, Based on power information and the energy of a single pacing pulse, the cumulative power consumption is calculated from three aspects: continuous power consumption, event-triggered power consumption, and data transmission power consumption, including: The cumulative power consumption is the sum of continuous power consumption, event-triggered power consumption, and data transmission power consumption; Among them, continuous power consumption is the product of the power consumption per unit time of each hardware module and the running time; Event-triggered power consumption is the sum of the product of the number of atrial pacings and the energy of the corresponding atrial pacing pulse, and the product of the number of ventricular pacings and the energy of the corresponding ventricular pacing pulse. The power consumption for data transmission is the product of the data transmission length and the power consumption per unit data length.

4. The method for estimating the remaining battery life of an implantable medical device according to claim 1, characterized in that, The pacing dependence index and pacing variability index are calculated based on historical pacing data consisting of the number of cardiac pacings, including: in, The pacing dependence index. The pacing volatility index, and The values ​​are divided into the average number of atrial and ventricular pacings within a preset time window. This represents the theoretical maximum number of pacing cycles per unit of time. and Let be the weight coefficient, and satisfy... , and It is divided into the standard deviation of the number of atrial and ventricular pacing.

5. The method for estimating the remaining battery life of an implantable medical device according to claim 1, characterized in that, A time evolution model of pacing dependence is constructed to describe the dynamic changes of pacing dependence over time; The time evolution model adopts a linear model, an exponential model, a constant model, a polynomial model, or a piecewise function model to accommodate individual differences among different patients; Specifically, based on historical pacing data, the changing trend of pacing dependence is analyzed. By calculating the rate of change, slope, or fluctuation characteristics of pacing dependence over time, the corresponding time evolution model parameters are selected or adjusted so that the model can adaptively reflect the patient's current and future physiological changes.

6. The method for estimating the remaining battery life of an implantable medical device according to claim 1, characterized in that, By combining the adjustment coefficient, the pacing variability index, and the pacing dependence index determined based on a time evolution model, the number of pacing cycles under different operating conditions in the future is predicted, including: First, calculate the future... Baseline pacing counts per time step : in, This represents the average number of pacing cycles. Represents the theoretical maximum number of pacing cycles, the difference term This indicates the potential growth space for pacing demand. The pacing dependence index is determined based on a time evolution model; Then, short-term fluctuation factors are superimposed on the baseline pacing count to obtain a unified pacing count prediction model to predict the pacing count under different working conditions in the future: in, The pacing volatility index, For the predicted number of pacing sessions, This is an adjustment coefficient used to characterize different power consumption conditions. When, it means that the impact of short-term fluctuations is not considered; when When this occurs, it indicates that the pacing demand is higher than average, corresponding to a higher power consumption condition; when When the pacing demand is below average, it indicates a lower power consumption condition.

7. The method for estimating the remaining battery life of an implantable medical device according to claim 1, characterized in that, Predicted power consumption is calculated based on the predicted number of pacemakers, including: in, For the future i Predicted power consumption at time step and Each of the future i The continuous power consumption and data transmission power consumption of the time step. and These represent the predicted number of atrial and ventricular pacing cycles, respectively. and These represent the energy of a single atrial and ventricular pacing, respectively.

8. The method for estimating the remaining battery life of an implantable medical device according to claim 1, characterized in that, Based on the remaining battery capacity determined by cumulative power consumption and the predicted power consumption, the remaining battery life is calculated by decreasing the discrete time step, including: in, For the first The remaining battery capacity at the time step. To predict power consumption, For the first The remaining battery capacity at the time step. The adjustment coefficient is used to characterize different power consumption conditions; When the following conditions are met: At that time, corresponding time step That is, the remaining service life under power consumption conditions. .

9. The method for estimating the remaining battery life of an implantable medical device according to claim 1, characterized in that, By selecting different adjustment coefficients Within a unified computing framework, the maximum lifespan, average lifespan, and minimum lifespan can be obtained, thus generating interval prediction results for battery lifespan. And satisfy ,in For the remaining battery life, and These are the minimum lifespan and the maximum lifespan, respectively. Average lifespan; When the adjustment coefficient Continuous values ​​are used to represent lifetime prediction results at different confidence levels.

10. A system for estimating the remaining battery life of an implantable medical device, characterized in that, include: The power information statistics module is used to collect power information during device operation, including the running time of each hardware module in the device, the number of heartbeats, and the data transmission length. The pulse energy calculation module is used to calculate the energy of a single pacing pulse. The cumulative power consumption calculation module is used to calculate the cumulative power consumption from three aspects: continuous power consumption, event-triggered power consumption, and data transmission power consumption, based on power information and the energy of a single pacing pulse. The future power consumption prediction module is used to calculate the pacing dependence index and pacing fluctuation index based on historical pacing data consisting of the number of heart pacings and to construct a time evolution model of pacing dependence. Combining the adjustment coefficient, pacing fluctuation index and pacing dependence index determined based on the time evolution model, it predicts the number of pacings and power consumption under different working conditions in the future. The remaining life estimation module is used to calculate the remaining lifespan of the device battery by using the battery's remaining capacity determined by the cumulative power consumption and the predicted power consumption, through discrete time step decrement.

Images