A medical system configured to determine a person's health status based on blood pressure changes detected by an implanted optical sensor.

Abstract

An exemplary implantable medical device comprises a housing configured for subcutaneous implantation at a location within a patient; an optical sensor located on or inside the housing and configured to generate an optical signal indicating blood movement within the patient's blood vessels adjacent to a location over a period of time; and a processing circuit configured to determine a plurality of blood pressure levels over a period of time based on the optical signal, determine blood pressure changes over a period of time based on the determined plurality of blood pressure levels, determine the patient's health status based on the determined blood pressure changes over a period of time, and output an instruction for the determined health status.

Description

Technical Field

[0001]

[0001] This application claims the benefit of U.S. Provisional Patent Application No. 63 / 498,903, filed Apr. 28, 2023, the entire contents of which are incorporated herein by reference.

[0002]

[0002] (Field of the Invention) The present disclosure generally relates to assessing health status based on blood pressure.

Background Art

[0003]

[0003] Hypertension represents an important factor in the development of cardiovascular diseases. One in three adults in the United States is diagnosed with hypertension every year. Uncontrolled hypertension dramatically increases the risk of diseases such as coronary artery disease, congestive heart failure, kidney failure, vision impairment, and stroke. Therefore, blood pressure measurement and monitoring are important indicators essential for the treatment and management of a patient's overall health.

Summary of the Invention

[0004]

[0004] The present disclosure generally aims at techniques for detecting changes in blood pressure based on one or more signals generated by an optical sensor of an implantable medical device to facilitate the determination of the health status of a patient, such as hypertension. More specifically, the present disclosure aims at techniques for evaluating an optical signal to determine changes in blood pressure over a period, such as one day or more, to determine the health status of a patient. A processing circuit extracts specific features from the received optical signal and applies the extracted features to an artificial intelligence model, such as a machine learning model or other suitable model, to determine changes in blood pressure over a period and / or to determine the health status of a patient.

[0005]

[0005] For many patients, blood pressure is measured only during visits to the clinic and / or hospital. In such patients, changes in blood pressure and health status between visits may not be detected. Some patients can use conventional mechanical plethysmography blood pressure measurement devices at home, but the quality of these measurements still depends on the compliance and ability of the patient or caregiver, and the measurements are taken relatively infrequently.

[0006]

[0006] In the case of some medical devices, such as watches, patches, bands, fitness trackers, or other external devices configured to detect blood pressure using, for example, photoplethysmograpic techniques, the position and orientation of the sensors used to detect blood pressure values ​​may vary between patients and within a given patient over time. External medical devices also require the user to adhere to wearing the device.

[0007]

[0007] The techniques and systems of the present disclosure may be implemented in implantable medical devices (IMDs) that can detect optical signals indicating blood pressure values ​​continuously, for example autonomously, and periodically and / or on a trigger basis, without human intervention, while implanted subcutaneously in a patient for several months or years. Continuously monitored optical signals and / or features continuously derived from continuously monitored optical signals, used by processing circuits to determine the patient's health status based on determined blood pressure changes over a period of time, can improve the sensitivity and / or specificity of blood pressure changes. For example, by continuously monitoring blood pressure and determining changes in blood pressure in the continuously monitored blood pressure, these determined blood pressure changes may not be detectable by clinic visits or by the patient manually measuring blood pressure, thus leading to improved sensitivity and / or specificity of blood pressure changes. In some examples, improving the sensitivity and / or specificity of blood pressure changes over a period of time can facilitate a more accurate determination of hypertension, cardiac health status, and the risk of sudden cardiac death, which may lead to clinical interventions to control hypertension, such as medication and ablation.

[0008]

[0008] Unlike conventional blood pressure detection systems, the techniques and systems of the present disclosure may use machine learning models to more accurately determine a patient's health status. In some examples, the machine learning model is trained on a set of training instances, one or more of which contain data showing the relationship between various features of the received optical signal (including features of a particular blood pressure level) and the classification of changes in blood pressure levels over a period of time. Since the machine learning model is potentially trained on thousands or millions of training instances, it may reduce the amount of classification error when classifying health status based on blood pressure changes compared to conventional blood pressure detection systems. In addition, the techniques and systems of the present disclosure may be implemented in implantable medical devices (IMDs) that can detect blood pressure values ​​continuously, for example autonomously and periodically and / or on a trigger basis, without human intervention, while implanted subcutaneously in a patient for months or years. Processing circuits in the IMD or other devices in a medical system including an IMD may perform millions of operations per second on the received optical signal data to determine blood pressure changes and determine the patient's health status, for example, using a machine learning model. Using the techniques of this disclosure in conjunction with IMD may be advantageous when a physician cannot be continuously present with a patient over several weeks or months to assess blood pressure levels, and / or when performing millions of actions on blood pressure levels over several weeks or months is not practically feasible in the physician's mind using the techniques of this disclosure, including, for example, the use of machine learning models.

[0009]

[0009] Reducing errors in classifying health conditions using machine learning models that implement the techniques of the present disclosure can yield one or more technical and clinical benefits. In some examples, a medical system that determines a patient's health condition based on the determined blood pressure changes over a period of time, by determining multiple blood pressure values ​​over a period of time based on received optical signals, and determining blood pressure changes over that period based on blood pressure levels, and by applying a trained machine learning model to features extracted from the received optical signals, for example, can help the determination of health conditions such as the severity of hypertension have higher specificity and sensitivity. For example, if a system or computing device that determines a patient's health condition such as the severity of hypertension has higher specificity and sensitivity, the number of false positives can be reduced. In some examples, using a machine learning model as described in the present disclosure can yield higher specificity and sensitivity when determining a patient's health condition such as the severity of hypertension. This higher specificity and sensitivity can increase the confidence of other devices, users, and / or clinicians in the accuracy of determining a patient's health condition such as the severity of hypertension. In some cases, this improved reliability in determining a patient's health status, such as the severity of hypertension, can improve the usefulness of a system or computing device, as clinicians, users, or other computing devices may not use and / or rely on determinations that do not exceed specificity and sensitivity thresholds. The systems and techniques of the Disclosure that use machine learning models can also more flexibly classify or predict a patient's health status, such as the severity of hypertension, from specific portions of an optical signal, by eliminating the need to construct an explicit set of rules within the IMD, which could otherwise be too diverse in size to actually implement and process for each new portion of the optical signal detected for a patient. With the more accurate health status classifications, such as the severity of hypertension, provided by the machine learning models used in conjunction with the techniques of the Disclosure, physicians and caregivers can also provide better tailored care, treatment, and interventions for patients experiencing health events such as hypertension.

[0010]

[0010] In one example, the present disclosure describes an implantable medical device comprising: a housing configured for subcutaneous implantation at a location within a patient; an optical sensor located on or inside the housing and configured to generate an optical signal indicating blood movement in the patient's blood vessels adjacent to a location over a period of time; and a processing circuit configured to determine a plurality of blood pressure levels over a period of time based on the optical signal, determine a change in blood pressure over a period of time based on the determined plurality of blood pressure levels, determine the patient's health status based on the determined change in blood pressure over a period of time, and output an instruction for the determined health status.

[0011]

[0011] In another example, the present disclosure describes a system comprising: an implantable medical device (IMD) comprising: a housing configured for subcutaneous implantation at a location in a patient; and an optical sensor located on or within the housing and configured to generate an optical signal indicating blood movement in the patient's blood vessels adjacent to a location over a period of time; and one or more computing devices communicably connected to the IMD, the one or more computing devices comprising: a memory; and a processing circuit connected to the memory, the processing circuit configured to determine a plurality of blood pressure levels over a period of time based on the optical signal; determine a change in blood pressure over a period of time based on the determined plurality of blood pressure levels; determine the patient's health status based on the determined change in blood pressure over a period of time; and output an instruction for the determined health status.

[0012]

[0012] In another example, the present disclosure describes a method for operating a processing circuit in a medical system, the method comprising: the processing circuit receiving optical signals indicating blood movement within the patient's blood vessels via an implantable medical device including an optical sensor implanted subcutaneously in the patient; the processing circuit determining a plurality of blood pressure levels of the patient over a period of time based on the optical signals; the processing circuit determining a change in blood pressure over a period of time based on the determined plurality of blood pressure levels; the processing circuit determining the patient's health status based on the determined change in blood pressure over a period of time; and the processing circuit outputting an indication of the determined health status.

[0013]

[0013] In another example, the present disclosure describes a non-temporary computer-readable storage medium that stores instructions causing one or more processors, when executed, to receive optical signals indicating blood movement within a patient's blood vessels via an implantable medical device, which includes at least an optical sensor implanted subcutaneously within the patient; determine a plurality of blood pressure levels of the patient over a period of time based on the optical signals; determine a change in blood pressure over a period of time based on the determined plurality of blood pressure levels; determine a health status of the patient based on the determined change in blood pressure over a period of time; and output an instruction for the determined health status.

[0014]

[0014] This summary is intended to provide an overview of the subject matter described herein. It is not intended to provide an exclusive or comprehensive description of the apparatus and methods described in detail in the accompanying drawings and the following description. Further details of one or more embodiments are provided in the accompanying drawings and the following description. [Brief explanation of the drawing]

[0015] [Figure 1A] This is a conceptual diagram illustrating an exemplary system for determining a health status, using several examples from the present disclosure. [Figure 1B]This block diagram shows an exemplary system comprising an access point, a network, an external computing device such as a server, and one or more other computing devices, which can be connected to the IMD and / or computing device shown in Figure 1A. [Figure 2A] This block diagram shows an exemplary configuration of the IMD in Figures 1A and 1B, according to some examples of the present disclosure. [Figure 2B] Figure 1A is a block diagram illustrating an exemplary configuration of a computing device, as shown in some examples of the present disclosure. [Figure 3A] These are conceptual diagrams of detected pulsating waveforms, illustrated by some examples of the disclosure. [Figure 3B] These are conceptual diagrams of detected pulsating waveforms, illustrated by some examples of the disclosure. [Figure 3C] These are conceptual diagrams of detected pulsating waveforms, illustrated by some examples of the disclosure. [Figure 4A] Figures 1A to 2A are conceptual perspective views illustrating the exemplary configuration of the IMD. [Figure 4B] Figures 1A to 2A are schematic diagrams showing exemplary configurations of the IMD. [Figure 4C] Figures 1A to 2A are conceptual perspective views illustrating the exemplary configuration of the IMD. [Figure 5] This flowchart illustrates an exemplary technique for operating a system to determine a person's health status. [Figure 6] This is a conceptual diagram illustrating an exemplary machine learning model configured to determine blood pressure levels, changes in blood pressure levels, and / or health status. [Figure 7] This is a conceptual diagram illustrating an exemplary training process for an artificial intelligence model, as shown in the examples of this disclosure. [Figure 8A] This flowchart illustrates some examples of determining a health status based on optical signals in this disclosure. [Figure 8B] This flowchart illustrates some examples of determining a health status based on optical signals in this disclosure. [Figure 8C]A flowchart for determining a health condition based on an optical signal according to some examples of the present disclosure. [Figure 9A] A conceptual diagram for determining a health condition based on applying an optical signal to each machine learning model. [Figure 9B] A conceptual diagram for determining a health condition based on applying an optical signal to each machine learning model.

Mode for Carrying Out the Invention

[0016]

[0028] Various types of medical devices detect a patient's blood pressure. An implantable medical device (IMD) can include an optical sensor for detecting blood pressure. The optical sensor used by the IMD to detect blood pressure may be integrated with the housing of the IMD and / or connected to the IMD via one or more elongated leads. Examples of IMDs that can be configured to monitor blood pressure include pacemakers and implantable defibrillators that can be connected to intravascular or extravascular leads, and pacemakers having a housing configured to be implanted within the heart that can be leadless. An example of a pacemaker configured for intracardiac implantation is the Micra (trademark) Transcatheter Pacing System available from Medtronic, Inc. Some IMDs that do not provide therapy, such as implantable patient monitors, can be configured to detect blood pressure. An example of such an IMD is the Reveal LINQ (trademark) and LINQ II (trademark) Insertable Cardiac Monitor (ICM) available from Medtronic, Inc., which can be inserted subcutaneously. Such IMDs can facilitate relatively long-term monitoring of patients during normal daily activities and can transmit the collected data to a network service such as the Medtronic Carelink (trademark) network periodically.

[0017]

[0029] Any medical device configured to detect blood pressure values ​​via an implantable optical sensor, including examples identified herein, may implement the techniques of this disclosure for evaluating the optical signal to determine a patient's health status based on changes in blood pressure over time. For example, specific features may be extracted from the optical signal, and the extracted features may be applied to a machine learning model to determine the health status, thereby providing accurate determination of health status, such as the severity of hypertension, with high specificity and sensitivity. Techniques of this disclosure for determining a patient's health status, such as the severity of hypertension, can facilitate the determination of the severity of hypertension, cardiac health status, and the risk of sudden cardiac death, which may lead to clinical interventions to control hypertension, such as medication and ablation.

[0018]

[0030] Figure 1A is a conceptual diagram illustrating an exemplary system 2 for determining a health status using a machine learning model based on the reception of signals from an optical sensor. As shown in Figure 1A, system 2 includes a computing device 12. The computing device 12 may be a computing device used in a home, outpatient, clinic, or hospital environment. The computing device 12 may include, for example, a clinician programmer, a desktop computer, a laptop computer, a workstation, a server, a mainframe, a cloud computing system, a smartphone, a smartwatch, or a combination thereof. The computing device 12 may be configured to receive input from a user, such as a clinician or patient 4, output information to the user, or both, via a user interface device 13 ("UI 13"). In some examples, the UI 13 may include a display such as a touch-sensitive display (e.g., a liquid crystal display (LCD) or a light-emitting diode (LED) display), one or more buttons, one or more keys (e.g., a keyboard), a mouse, one or more dials, one or more switches, a speaker, one or more lights, or a combination thereof.

[0019]

[0031] The computing device 12 may be communicatively connected to the implantable medical device (IMD) 10. In some examples, the IMD 10 may be configured to be implanted subcutaneously in the patient 4 and may include (multiple) optical sensors 62 (as shown in Figure 2A). In some examples, the IMD 10 may include (multiple) additional sensors 61 (as shown in Figure 2A) in addition to the (multiple) optical sensors 62. Some examples of physiological signals that can be detected by such a device may include blood pressure, electrocardiogram (ECG) signals, heart rate, cardiac output, heart sounds, impedance, cardiac motion, respiratory signals, perfusion signals, activity and / or posture signals, pressure signals, blood oxygen saturation signals, body composition, fluid impedance signals, and blood glucose or other blood component signals. In some examples, in addition to the (multiple) optical sensors 62, the IMD 10 may include electrodes and other sensors 61 for detecting physiological signals of the patient 4, collecting and storing physiological data, and detecting the onset of symptoms based on such signals. In some examples, the (multiple) optical sensors 62 of the IMD10 are photoplethysmography (PPG) sensors. In some examples, the IMD10 takes the form of another ICM similar to a Reveal LINQ® or LINQ II ICM®, for example, a modified or modified version of the LINQ® ICM.

[0020]

[0032] The IMD10 may be configured to collect detected physiological signals and / or data based on detected physiological signals and / or transmit them to a computing device 12. For example, the IMD10 may detect the blood pressure value of patient 4, collect the detected blood pressure value and / or communicate with the computing device 12. In some examples, a subcutaneous optical sensor, such as the PPG sensor of the IMD10, may be placed at a desired location. The subcutaneous optical sensor signal may reflect the movement of blood in the blood vessels from the heart toward the measurement location as a wave-like motion. The output of the subcutaneous optical sensor 62 over a period of time may be used as a surrogate for the blood pressure level and blood pressure changes of patient 4.

[0021]

[0033] With each contraction of the left ventricle of the heart, the left ventricle releases blood, generating a pulse pressure that travels throughout the patient's arteries. This pulse can be detected at various locations in the patient, including various subcutaneous implantable sites. The IMD10 may be configured to be implanted subcutaneously. Multiple optical sensors 62 may be located within the IMD10.

[0022]

[0034] According to the techniques of this disclosure, the IMD 10 and / or computing device 12 can determine the patient's health status based on changes in blood pressure over time, and the changes in blood pressure may be changes in various features from the PPG waveform. In some examples, the computing device 12 may receive (multiple) hemodynamic parameters of the patient, such as blood pressure, from the IMD 10. In some examples, the hemodynamic parameter may be blood pressure.

[0023]

[0035] Figure 1B is a block diagram illustrating an exemplary system comprising an access point 20, a network 22, an external computing device such as a server 24, and one or more other computing devices 30A-30N (collectively, "Computing Devices 30"), which can be connected to the IMD 10 and external devices 23 via a network 22 according to one or more techniques described herein. In this example, the IMD 10 may communicate with the external devices 23 via a first wireless connection and with the access point 20 via a second wireless connection using a communication circuit 54. In the example of Figure 1B, the access point 20, the external devices 23, the server 24, and the computing devices 30 are interconnected and can communicate with each other via the network 22. In some examples, the external device 23 in Figure 1B may be a computing device 12 as shown in Figure 1A.

[0024]

[0036] Access point 20 may include a device that connects to network 22 via one of various connections, such as telephone dial-up, digital subscriber line (DSL), or cable modem connection. In other examples, access point 20 may be connected to network 22 via various forms of connections, including wired or wireless connections. In some examples, access point 20 may be a user device, such as a tablet or smartphone, which may be located in the same place as the patient. IMD 10 may be configured to transmit data, such as optical signals, to access point 20. Access point 20 may then transmit the retrieved data to server 24 via network 22.

[0025]

[0037] In some cases, server 24 may be configured to provide a secure storage site for data collected from IMD 10 and / or external device 23. In some cases, server 24 may collect data from web pages or other documents via computing device 30 for viewing by trained professionals such as clinicians. One or more embodiments of the system illustrated in Figure 1B may implement common network technologies and functions similar to those provided by the Medtronic CareLink® network. In some examples, server 24 may communicate with computing device 30 via network 22. For example, server 24 may transmit data analysis, such as blood pressure changes, to computing device 30, external device 23, or any other computing device via network 22. For example, server 24 may transmit health status or blood pressure changes to computing device 30, external device 23, or any other computing device via network 22.

[0026]

[0038] In some examples, one or more of the computing devices 30 may be a tablet or other smart device located with the clinician, allowing the clinician to program the IMD 10, receive alerts from the IMD 10, and / or query the IMD 10. For example, the clinician can access data collected by the IMD 10 via the computing device 30 to check the patient's condition, such as when patient 4 is between clinician visits. In some examples, the clinician can input instructions for medical interventions for patient 4 into an application running on the computing device 30, based on the patient's health status determined by the IMD 10, an external device 23, a server 24, or any combination thereof, or based on other patient data known to the clinician. The device 30 can then transmit the instructions for medical interventions to another computing device 30 located with patient 4 or patient 4's caregiver. For example, such instructions for medical interventions may include instructions to change the dosage, timing, or choice of medication, instructions to schedule a clinician's visit, or instructions to request medical treatment. In a further example, the computing device 30 can generate alerts to patient 4 based on the patient's medical condition, which may enable patient 4 to proactively seek medical treatment before being instructed to do so. In this way, patient 4 is empowered to take action to address their medical condition as needed, which may help improve patient 4's clinical outcome.

[0027]

[0039] In the example shown in Figure 1B, the server 24 includes, for example, a storage device 26 for storing data retrieved from IMD 10, and a processing circuit 28. Although not shown in Figure 1B, the computing device 30 may similarly include a storage device and a processing circuit. The processing circuit 28 may include one or more processors configured to implement functions and / or process instructions for execution within the server 24. For example, the processing circuit 28 may be capable of processing instructions stored in memory 26. The processing circuit 28 may include, or be connected to, a communication circuit, which may include any suitable hardware, firmware, software, or any combination thereof for communicating with another device. In some examples, a description of a processing circuit 28 that outputs a signal, such as a classification signal, may include the processing circuit 28 causing the communication circuit of the server 4 to output a signal. The processing circuit 28 may include, for example, a microprocessor, DSP, ASIC, FPGA, or equivalent discrete or integrated logic circuit, or any combination of the aforementioned devices or circuits. Therefore, the processing circuit 28 may include any suitable structure for performing the functions attributed to the processing circuit 28 herein, whether hardware, software, firmware, or any combination thereof. The processing circuit of computing device 12, the processing circuit 28 of server 24, and / or the processing circuit of computing device 30 may implement any of the techniques described herein to analyze the optical signals received from IMD 10, for example, to determine the health status of patient 4. For example, the processing circuit 28 may determine the health status of patient 4 based on changes in blood pressure over time, and changes in blood pressure may include changes in various feature segments from the PPG waveform.

[0028]

[0040] The storage device 26 may include a computer-readable storage medium or a computer-readable storage device. In some examples, the memory 26 includes one or more short-term memory or long-term memory. The storage device 26 may include, for example, RAM, DRAM, SRAM, magnetic disk, optical disk, flash memory, or EPROM or EEPROM. In some examples, the storage device 26 is used to store data indicating instructions for execution by the processing circuit 28.

[0029]

[0041] Figure 2A is a block diagram showing an example configuration of the IMD10 shown in Figures 1A and 1B. As shown in Figure 2A, the IMD10 may include a processing circuit 50, a memory 56, one or more sensors 61 which may include one or more optical sensors 62, a detection circuit 52 connected to one or more sensors 61 and electrodes 16A and 16B (collectively referred to as "electrodes 16"), and a communication circuit 54. As used herein, "sensor" may refer to any sensor described herein, including electrodes 16 and optical sensors 62.

[0030]

[0042] One or more sensors 61 of the IMD10 may detect physiological parameters or signals of patient 4. The (multiple) sensors 61 may include one or more accelerometers (e.g., a triaxial accelerometer), a temperature sensor, a pressure sensor, a heart sound sensor (e.g., a microphone or accelerometer), or other sensors. The electrodes 16 may be configured to detect the electrocardiogram or other electrographic signals of patient 4, and / or the impedance of tissue fluid adjacent to the electrodes. One or more optical sensors 62 may include one or more photodetectors 64 configured to receive and / or detect optical signals, such as reflected light signals from (multiple) optical emitters 63. The optical signals received and / or detected by the (multiple) optical sensors 62 may be called optical signals. In some examples, one or more of the (multiple) optical sensors 62 may be configured as PPG sensors, for example, by being configured to receive light reflected by blood in one or more blood vessels. In some examples, the optical sensors 62 may be contained in the same sensor package and / or implemented using the same (multiple) transducers. The IMD10 may, in some cases, include one or more optical sensors 62, each including two or more light emitters 63 and one or more photodetectors 64.

[0031]

[0043] The IMD10, for example, the optical sensor 62, can generate a signal proportional to instantaneous blood pressure fluctuations. In some examples, the IMD10 and / or the optical sensor 62 may be calibrated so that the IMD10 can determine blood pressure values ​​based on the signal. In some examples where the IMD10 is not calibrated in this way, the optical signal may nevertheless be useful for tracking fluctuations in the amplitude / morphology of blood pressure, such as systolic and / or diastolic blood pressure and other features described herein, over a period of time. In some examples, this period may be one week, two weeks, one month, two months, and / or other periods. The days used for the period may be adjacent days (e.g., Monday to Friday) or non-adjacent days (e.g., every Monday).

[0032]

[0044] Although the techniques for determining the health status are described herein as being performed by the processing circuit 50 of the IMD 10, such techniques may be performed in whole or in part by the processing circuits of any one or more devices of System 2, such as the processing circuit 230 of the computing device 12, the processing circuit 28 of the server 24, or the processing circuits of other computing devices or computing systems, such as various implantable medical devices, servers, cloud computing systems, or any other computing devices or combinations of computing devices.

[0033]

[0045] The processing circuit 50 may be configured to determine multiple blood pressure levels over a period of time based on an optical signal. In some examples, the blood pressure levels may include one or more of the patient's systolic or diastolic blood pressure. In some examples, the blood pressure levels may be values ​​derived from the optical signal at different times. In some examples, the optical signal and / or blood pressure levels may represent various features, such as cardiac features, which will be further described below. Based on the determined multiple blood pressure levels, the processing circuit 50 may determine the blood pressure change over a period of time, determine the patient's health status based on the determined blood pressure change over a period of time, and output an indication of the determined health status. In some examples, the health status is the severity of hypertension.

[0034]

[0046] In some examples, the processing circuit 50 may use multiple blood pressure levels over a period such as three months to determine various changes in blood pressure levels during that period, including short-term changes (e.g., changes in the daily mean, mode, and median) and / or long-term changes (e.g., changes in the weekly mean, mode, median, and monthly mean, mode, and median). Based on these various determined changes, the processing circuit 50 may determine the patient's health status, such as an increase / decrease in hypertension, an increase / decrease in the hypertension risk score, or other changes. The processing circuit 50 may output indicators of the determined health status to the computing device 12, which may facilitate the determination of cardiac health status and the risk of sudden cardiac death, and may lead to clinical interventions to control hypertension, such as medication and renal denervation.

[0035]

[0047] The (multiple) optical sensors 62 may be configured as PPG sensors. The (multiple) optical emitters 63 may be configured to emit optical signals belonging to a specific wavelength spectrum, and the (multiple) photodetectors 64 may be configured to receive reflected optical signals corresponding to the specific wavelength spectrum emitted by the (multiple) optical emitters 63. In some examples, the (multiple) optical emitters 63 may be configured to emit optical signals belonging to a specific wavelength spectrum. Some examples of specific wavelength spectra may be amber wavelength spectra, green wavelength spectra, yellow wavelength spectra, blue wavelength spectra, or other wavelength spectra. The (multiple) photodetectors 64 may be configured to receive optical signals of the corresponding specific wavelength spectrum, such as amber wavelength spectra, green wavelength spectra, yellow wavelength spectra, blue wavelength spectra, or other wavelength spectra.

[0036]

[0048] The processing circuit 50 may be configured to extract one or more feature segments, such as cardiac feature segments, from an optical signal. In some examples, the feature segments that can be extracted from the optical signal may include one or more of the following: interpeak interval, systolic peak, time to systolic peak, systolic area, systolic gradient trajectory, diastolic peak, diastolic area, diastolic gradient trajectory, diastolic peak / systolic peak ratio, Boolean dichroic notch, dichroic notch height, mean, standard deviation, kurtosis, and / or skewness. In some examples, the processing circuit 50 may be configured to apply the extracted cardiac feature segments to a machine learning model to determine the patient's health status.

[0037]

[0049] In some examples, the processing circuit 50 may be configured to extract one or more cardiac features from an optical signal, determine changes in the extracted cardiac features over a certain period of time, and apply one or more of the extracted cardiac features or the determined changes in the extracted cardiac features to a machine learning model to determine the patient's health status.

[0038]

[0050] In some examples, the processing circuit 50 may apply the optical signals and / or levels or features described herein to a machine learning model to determine the patient's health status. In some examples, the processing circuit 50 may be configured to identify the wavelength spectrum of the optical signal. In some examples, the processing circuit 50 may select a machine learning model trained on training data having a wavelength spectrum corresponding to the wavelength spectrum of the optical signal, and apply the selected machine learning model corresponding to the wavelength spectrum of the optical signal to the optical signal.

[0039]

[0051] In some examples, the IMD10 may further include electrodes 16 and (multiple) sensors 61 other than (multiple) optical sensors 62 to monitor one or more physiological parameters of the patient, where one or more physiological parameters are different from blood pressure. In some examples, the (multiple) sensors 61 may include accelerometers. In some examples, the physiological parameters may include one or more of heart rate, cardiac output, heart sounds, or impedance (e.g., fluid state).

[0040]

[0052] The processing circuit 50 may be further configured to determine the health status of patient 4 based on the determined blood pressure changes and one or more physiological parameters of the patient. For example, the processing circuit 50 may apply one or more physiological parameters along with the optical signal to a machine learning model to determine the patient's health status. In some examples, the processing circuit 50 may determine and / or update the health status based on one or more monitored physiological parameters, and the health status may be determined based on changes in blood pressure.

[0041]

[0053] In some examples, the processing circuit 50 can determine the patient's health status by applying the optical signal detected via the electrode 16 and impedance measurements including one or more of the actual impedance and reactive impedance to a machine learning model. In some examples, the processing circuit 50 can extract features from the optical signal and the detected impedance measurements including one or more of the actual impedance and reactive impedance, and apply the extracted features to a machine learning model to determine the patient's health status.

[0042]

[0054] Figure 2B is a block diagram showing an example configuration of computing device 12. In some examples, computing device 12 may take the form of a smartphone, laptop, tablet computer, personal digital assistant (PDA), smartwatch, or other wearable computing device. In some examples, computing device 30 and / or server 24 may be configured similarly to computing device 12 shown in Figure 2B.

[0043]

[0055] As shown in the example in Figure 2B, the computing device 12 may be logically divided into user space 202, kernel space 204, and hardware 206. Hardware 206 may include one or more hardware components that provide an operating environment for components running in user space 202 and kernel space 204. User space 202 and kernel space 204 may represent different sections or segmentations of memory, but kernel space 204 provides processes and threads with higher privileges than user space 202. For example, kernel space 204 may include an operating system 220 that operates with higher privileges than components running in user space 202.

[0044]

[0056] As shown in Figure 2B, the hardware 206 includes a processing circuit 230, a memory 232, one or more input devices 234, one or more output devices 236, one or more sensors 238, and a communication circuit 240. Although shown in Figure 2B as a standalone device for example, the computing device 12 can be any component or system including a processing circuit or other suitable computing environment for executing software instructions, and does not necessarily have to include one or more elements shown in Figure 2B.

[0045]

[0057] The processing circuit 230 is configured to perform a function within the computing device 12 and / or to process instructions for execution. For example, the processing circuit 230 may be configured to receive and process instructions stored in memory 232 that provide the functionality of components included in kernel space 204 and user space 202 to perform one or more operations in accordance with the techniques of this disclosure. Examples of the processing circuit 230 include any one or more microprocessors, controllers, GPUs, TPUs, DSPs, ASICs, FPGAs, or equivalent discrete or integrated logic circuits.

[0046]

[0058] Memory 232 may be configured to store information within the computing device 12 for processing during the operation of the computing device 12. In some examples, memory 232 is described as a computer-readable storage medium. In some examples, memory 232 includes temporary memory or volatile memory. Examples of volatile memory include RAM, DRAM, SRAM, and other forms of volatile memory well known in the art. In some examples, memory 232 also includes one or more memories configured for long-term storage of information, for example, including non-volatile storage elements. Examples of such non-volatile storage elements include magnetic hard disks, optical disks, floppy disks, flash memory, or electrically programmable memories (EPROM) or electrically erasable and programmable memories (EEPROM) memory. In some examples, memory 232 includes cloud-related storage.

[0047]

[0059] One or more input devices 234 of the computing device 12 may receive input from, for example, a patient 4, a clinician, or another user. Examples of input include tactile input, voice input, dynamic input, and visual input. Examples of input devices 234 may include a mouse, keyboard, voice response system, camera, buttons, control pad, microphone, presence detection component or contact detection component (e.g., a screen), or any other device for detecting input from a user or machine.

[0048]

[0060] One or more output devices 236 of the computing device 12 may generate output to, for example, patient 4 or another user. Examples of output include tactile output, sensory output, audio output, and visual output. The output devices 236 of the computing device 12 may include presence detection screens, sound cards, video graphics adapter cards, speakers, cathode ray tube monitors, liquid crystal displays (LCDs), light-emitting diodes (LEDs), or any type of device for generating tactile output, audio output, and / or visual output.

[0049]

[0061] One or more sensors 238 of the computing device 12 may detect physiological parameters or signals of patient 4. The (multiple) sensors 238 may include electrodes, accelerometers (e.g., triaxial accelerometers), optical sensors, impedance sensors, temperature sensors, pressure sensors, heart sound sensors (e.g., microphones or accelerometers), and other sensors, as well as detection circuits (e.g., including ADCs) similar to those described above with respect to IMD 10 and Figure 2A.

[0050]

[0062] The communication circuit 240 of the computing device 12 can communicate with other devices by sending and receiving data. The communication circuit 240 can receive patient parameter data, such as optical signals, from the communication circuit in the IMD 10. The communication circuit 240 may include a network interface card, such as an Ethernet card, an optical transceiver, a radio frequency transceiver, or any other type of device capable of sending and receiving information. For example, the communication circuit 240 may include a wireless transceiver configured to communicate according to standards or protocols such as 3G, 4G, 5G, WiFi (e.g., 802.11 or 802.15 ZigBee), Bluetooth®, or Bluetooth® Low Energy (BLE).

[0051]

[0063] As shown in Figure 2B, the health monitoring application 250 runs in the user space 202 of the computing device 12. The health monitoring application 250 may be logically divided into a presentation layer 252, an application layer 254, and a data layer 256. The presentation layer 252 may include a user interface (UI) component 260 that generates and renders the user interface of the health monitoring application 250.

[0052]

[0064] The data layer 256 may include parameter data 290 and optical signal data 292 received from the IMD 10 via the communication circuit 240 and stored in memory 232 by the processing circuit 230. The application layer 254 may include, but is not limited to, a situation analyzer 270 and a model configuration service 272. The situation analyzer 270 may determine a health status based on the optical signal data 292 and, optionally, other parameter data 290 generated by the IMD 10, as described herein. The situation analyzer 270 may determine a health status based on the application of data as input to one or more models 294, which may include one or more machine learning models, algorithms, decision trees, and / or thresholds. In an example where model 294 includes one or more machine learning models, the situation analyzer 270 may apply feature vectors derived from the data to the (multiple) models.

[0053]

[0065] The model configuration component 272 may be configured to develop a model 294 based on machine learning. Exemplary machine learning techniques that may be employed to generate (multiple) models 294 may include various learning styles such as supervised learning, unsupervised learning, and semi-supervised learning. Representative types of algorithms include Bayesian algorithms, Markov models, Hawkes processes, clustering algorithms, decision tree algorithms, regularization algorithms, regression algorithms, example-based algorithms, artificial neural network algorithms, deep learning algorithms, and dimensionality reduction algorithms. Various examples of specific algorithms include Bayesian linear regression, boosted decision tree regression, neural network regression, backpropagation neural networks, convolutional neural networks (CNNs), long short-term networks (LSTMs), a priori algorithms, K-means clustering, k-nearest neighbor method (kNN), learning vector quantization (LVQ), self-organizing maps (SOMs), locally weighted learning (LWL), ridge regression, least absolute shrinkage and selection operator (LASSO), elastic networks, least-angle regression (LARS), principal component analysis (PCA), and principal component regression (PCR).

[0054]

[0066] Figures 3A to 3C show examples of optical signals 365A to 365C (collectively referred to as "optical signals 365") received by (multiple) optical sensors 62. Processing circuits of System 2, for example, processing circuits 50, 23, and / or 230, may perform feature extraction to extract features from the optical signals. In some examples, features that can be extracted from the optical signals include one or more of the following: inter-peak spacing 358, systolic peak 367, time to systolic peak 368, systolic area 369A, systolic gradient trajectory, diastolic peak 357, diastolic area 369B, diastolic gradient trajectory, diastolic peak / systolic peak ratio, Boolean dichroic notch 355, dichroic notch height 355, mean, standard deviation, kurtosis, and / or skewness. As shown in Figure 3B as an example, one or more of the following features are extracted from the optical signal 365: time to systolic peak 368, inter-peak interval 358, dichroic notch 355, diastolic peak 357, and / or systolic peak 367. As shown in Figure 3C as an example, the feature areas systolic area 369A and diastolic area 369B are extracted from the optical signal 365. The processing circuits of System 2, for example, processing circuits 50, 23, and / or 230, can apply the extracted features to a machine learning model to determine the health status. Figures 3A to 3C show examples of the dichroic notch 355, diastolic peak 357, and systolic peak 367, where the dichroic notch is the lowest point of a waveform, such as a pressure waveform, indicating the end of systole. The dichroic notch may occur after aortic valve closure and precede a secondary overlap wave. The systolic wave is the result of the direct pressure wave traveling from the left ventricle to the periphery of the body, and the systolic peak is the point of maximum pressure in the systolic wave. The diastolic wave is the result of the pressure wave being reflected by the arteries of the lower body, and the diastolic peak is the point of maximum pressure in the diastolic wave.

[0055]

[0067] In some examples, by extracting specific features from the received optical signal that are determined to correspond to blood pressure and / or specific health conditions related to blood pressure, as described above, according to the inventive techniques of this disclosure, the machine learning model can be more focused and spend a significant portion of its nodes determining blood pressure levels, changes in blood pressure levels, and / or health conditions based on changes in blood pressure over a period of time.

[0056]

[0068] The processing circuit of System 2 may be configured to run an artificial intelligence (AI) engine that operates according to one or more models, such as machine learning models. The machine learning models may include any number of different types of machine learning models, such as neural networks, deep neural networks, convolutional neural networks, recurrent neural networks such as long-term short-term memory networks, and high-density neural networks. In some examples, the inputs of various feature units to the AI ​​engine may be supplied as direct inputs to different layers within the network, and not necessarily before the convolutional layers. While machine learning models are described, the techniques described in this disclosure are also applicable to other types of AI models, including rule-based models and finite state machines.

[0057]

[0069] Machine learning, in general, can enable computing devices to analyze input data and identify actions that should be taken in response to that input data. Each machine learning model can be trained using training data that reflects likely input data. The training data may or may not be labeled (meaning that the correct actions to be taken based on the sample of training data are either explicitly stated or not explicitly stated).

[0058]

[0070] Training a machine learning model may be guided (in the sense that a designer, such as a computer programmer, can instruct the training to guide the machine learning model to identify the correct action considering the input data) or unguided (in the sense that the machine learning model is not guided by the designer to identify the correct action considering the input data). In some cases, machine learning models are trained through combinations of labeled and unlabeled training data, guided and unguided training, or possibly combinations thereof. Examples of machine learning include nearest neighbor, naive Bayes, decision trees, linear regression, support vector machines, neural networks, k-means clustering, Q-learning, time-lag, deep adversarial networks, evolutionary algorithms, or other supervised, unsupervised, semi-supervised, or reinforcement learning algorithms for training one or more models.

[0059]

[0071] The System 2 processing circuit may utilize machine learning, such as a deep learning algorithm or model (e.g., a neural network or a deep belief network), to generate a score indicating blood pressure levels and / or changes in blood pressure levels in order to determine the health status. The System 2 processing circuit may train a deep learning model to represent the relationship between the feature portions extracted from the received optical signals described above, the blood pressure levels, changes in blood pressure levels, and / or the health status. For example, the System 2 processing circuit may train a deep learning model using optical signals from other patients. In some examples, the System 2 processing circuit may train a deep learning model by adjusting the weights of the hidden layers of a neural network model to balance the contribution of each input (e.g., the characteristics of the image input) in accordance with determining the blood pressure levels, changes in blood pressure levels, and / or the health status. In some examples, the System 2 processing circuit may train a deep learning model by using optical signals received from patient 4 in addition to using optical signals from other patients.

[0060]

[0072] Once the deep learning model is trained, the system's processing circuitry can acquire data such as features extracted from the received optical signal and apply it to the trained deep learning model.

[0061]

[0073] The output of the deep learning model may include blood pressure levels, changes in blood pressure levels, and / or a score indicating health status. For example, the score may indicate the severity of hypertension in patient 4.

[0062]

[0074] Figure 4A is a conceptual diagram showing IMD10A, which may be an exemplary configuration of IMD10 shown in Figures 1A-2A as an implantable cardiac monitor (ICM). In the example shown in Figure 4A, IMD10A may be embodied as a monitoring device having a housing 412, a proximal electrode 16A, a distal electrode 16B, and (multiple) optical sensors 62. The (multiple) optical sensors 62 may be positioned at various locations on IMD10A. The housing 412 may further comprise a first main surface 414, a second main surface 418, a proximal end 420, and a distal end 422. The housing 412 surrounds the electronic circuitry located inside IMD10A and protects the circuitry housed inside from bodily fluids. An electrical feedthrough provides an electrical connection between electrode 16A and electrode 16B.

[0063]

[0075] In the example shown in Figure 4A, the IMD10A is defined by length L, width W, and thickness or depth D, and is in the form of an elongated rectangular prism, where length L is much greater than width W, and width W is greater than depth D. In one example, the geometric shape of the IMD10A, particularly the width W being greater than depth D, is chosen to allow the IMD10A to be inserted under the patient's skin using a minimally invasive procedure and to remain in the desired orientation during insertion. For example, the device shown in Figure 4A includes radial asymmetry along the longitudinal axis (particularly the rectangular shape) that maintains the device in the appropriate orientation after insertion. For example, the distance between the proximal electrode 16A and the distal electrode 16B may be in the range of 30 mm to 55 mm, 35 mm to 55 mm, and 40 mm to 55 mm, or any range or individual interval of 25 mm to 60 mm. In addition, the IMD10A may have a length L in the range of 30 mm to about 70 mm. In other examples, the length L may be in the range of 5mm to 60mm, 15mm to 50mm, 40mm to 60mm, 45mm to 60mm, or any length or range between approximately 5mm and approximately 80mm. In addition, the width W of the main surface 414 may be in the range of 5mm to 15mm, 3mm to 10mm, or any single width or range between 3mm and 15mm. The thickness of the depth D of the IMD10A may be in the range of 2mm to 9mm. In other examples, the depth D of the IMD10A may be in the range of 2mm to 5mm, in the range of 5mm to 15mm, or any single depth or range between 2mm and 15mm. In addition, the IMD10A according to an example of this disclosure has a geometric shape and size designed for ease of implantation and patient comfort. The example of the IMD10A described in this disclosure may have a volume of 3 cubic centimeters (cm) or less, 1.5 cubic centimeters or less, or any volume between 3 and 1.5 cubic centimeters.

[0064]

[0076] In the example shown in Figure 4A, when inserted into the patient, the first principal surface 414 faces outward toward the patient's skin, and the second principal surface 418 is located opposite the first principal surface 414. In addition, in the example shown in Figure 4A, the proximal end 420 and distal end 422 are rounded to reduce discomfort and irritation to surrounding tissues when inserted under the patient's skin. The IMD10A, including the instrument and method for inserting the IMD10A, is described, for example, in U.S. Patent Application Publication 2014 / 0276928, which is incorporated herein by reference in its entirety.

[0065]

[0077] Using the proximal electrode 16A and distal electrode 16B, cardiac signals, such as EGM signals, are detected intrathoracic or extrathoracic, which may be submuscular or subcutaneous. The EGM signals may be stored in the memory of the IMD 10A, and the data may be transmitted to another medical device via the integrated antenna 426A, which may be another implantable device or an external device such as a computing device 12. In some examples, electrodes 16A and 16B may be used in addition to, or instead of, detecting any biopotential signals of an object from any implanted location, which may be EGM, electroencephalogram (EEG), electromyogram (EMG), or nerve signals.

[0066]

[0078] In the example shown in Figure 4A, the proximal electrode 16A is very close to the proximal end 420, and the distal electrode 16B is very close to the distal end 422. In this example, the distal electrode 16B is not limited to a flat outward surface, but may extend from the first principal surface 414 onto the second principal surface 418 around a rounded edge 424 and / or end face 425, so that the electrode 16B has a three-dimensional curved shape. In some examples, the electrode 16B is an uninsulated portion of the metal (e.g., titanium) part of the housing 412.

[0067]

[0079] In the example shown in Figure 4A, the proximal electrode 16A is located on the first principal surface 414 and is substantially flat and outward-facing. However, in other examples, the proximal electrode 16A may utilize the three-dimensional curved shape of the distal electrode 16B to provide a three-dimensional proximal electrode (not shown in this example). Similarly, in other examples, the distal electrode 16B may utilize a substantially flat, outward-facing electrode located on the first principal surface 414, similar to the electrode shown with respect to the proximal electrode 16A.

[0068]

[0080] These various electrode configurations allow for configurations in which the proximal electrode 16A and the distal electrode 16B are located on both the first principal surface 414 and the second principal surface 418. In other configurations, such as that shown in Figure 4A, only one of the proximal electrode 16A or the distal electrode 16B is located on both principal surfaces 414 and 418, and in yet another configuration, both the proximal electrode 16A and the distal electrode 16B are located on either the first principal surface 414 or the second principal surface 418 (for example, the proximal electrode 16A is located on the first principal surface 14 and the distal electrode 16B is located on the second principal surface 418). In another example, the IMD 10A may include electrodes on both principal surfaces 414 and 418 at or near the proximal and distal ends of the device, so that a total of four electrodes are included in the IMD 10A. Electrodes 16A and 16B may be formed from several different types of biocompatible conductive materials, such as stainless steel, titanium, platinum, iridium, or alloys thereof, and may utilize one or more coatings such as titanium nitride or fractal titanium nitride.

[0069]

[0081] In the example shown in Figure 4A, the proximal end 420 includes a header assembly 428 which includes one or more of the proximal electrode 16A, an integrated antenna 426A, a movement-preventing projection 432, and / or suture holes 434. The integrated antenna 426A is located on the same principal surface as the proximal electrode 16A (i.e., the first principal surface 414) and is also included as part of the header assembly 428. The integrated antenna 426A enables the IMD 10A to transmit and / or receive data. In other examples, the integrated antenna 426A may be formed on the principal surface opposite to the proximal electrode 16A, or it may be incorporated into the housing 412 of the IMD 10A. In the example shown in Figure 4A, the movement-preventing projection 432 is located adjacent to the integrated antenna 426A and protrudes away from the first principal surface 414 to prevent longitudinal movement of the device. In the example shown in Figure 4A, the anti-movement projection 432 includes a plurality (e.g., nine) of small bumps or protrusions extending away from the first principal surface 414. As described above, in other examples, the anti-movement projection 432 may be located on the principal surface opposite to the proximal electrode 16A and / or integrated antenna 426A. In addition, in the example shown in Figure 4A, the header assembly 428 includes a suture hole 434 that provides another means of securing the IMD 10A to the patient to prevent movement after insertion. In the illustrated example, the suture hole 434 is located adjacent to the proximal electrode 16A. In one example, the header assembly 428 is a molded header assembly made of polymer or plastic material, which may be integrated with or separable from the principal part of the IMD 10A.

[0070]

[0082] Figure 4B is a functional schematic diagram of an IMD10A as shown in Figure 4A, according to an embodiment of the present disclosure. The IMD10A may include a proximal electrode 16B located at the proximal end 422, a distal electrode 16A located at the distal end 420, (multiple) optical sensors 62, an integrated antenna 426A, an electrical circuit 400, and a power supply 402. In particular, the electrical circuit 400 is connected to the proximal electrode 16B and the distal electrode 16A to detect cardiac signals and monitor events. The electrical circuit 400 may also be connected to transmit and receive communications via the integrated antenna 426A. The power supply 402 provides power to the electrical circuit 400 and any other components that require power. The power supply 402 may include one or more energy storage devices, such as one or more rechargeable or non-rechargeable batteries. In some examples, the electrical circuit 400 includes a processing circuit 50 and a storage device 56, such as memory, as shown in Figure 2, where the memory 56 is operably connected to the processing circuit 50 and configured to store machine learning models.

[0071]

[0083] In the example shown in Figure 4B, the electrical circuit 400 can receive raw EGM signals monitored by the proximal electrode 16B and the distal electrode 16A, and raw optical signals monitored by the (multiple) optical sensors 62. The electrical circuit 400 may include components / modules for converting the raw EGM signals into processed EGM signals that can be analyzed to detect detected events, and for converting the raw optical signals into processed optical signals that can be analyzed to detect detected events. Although not shown, the electrical circuit 400 may include any discrete and / or integrated electronic circuit components that implement analog and / or digital circuits that can analyze the optical signals to generate the functions described above for determining the patient's health status. For example, the electrical circuit 400 may include analog circuits, such as preamplifiers, filtering circuits, and / or other analog signal conditioning circuits. The module may also include digital circuits, such as digital filters, combinational logic circuits or sequential logic circuits, state machines, integrated circuits, processors (shared, dedicated, or grouped) that run one or more software programs or firmware programs, memory devices, or any other suitable components or combinations thereof that provide the described functionality.

[0072]

[0084] In one example, the electrical circuit 400 includes a detection unit for monitoring the EGM signals detected by the proximal electrode 16A and distal electrode 16B, respectively, and the optical signals received by the (multiple) optical sensors 62. In one example, the electrical circuit 400 includes a processing circuit 50 used to receive information about the detected events and implementing one or more algorithms for determining the patient's health status. Furthermore, the analog voltage signals received from electrodes 16A and 16B are passed to an analog-to-digital (A / D) converter included in the electrical circuit 400 and can be stored in a memory unit (not shown) included as part of the electrical circuit 400 for subsequent analysis using firmware executed by a processor included as part of the electrical circuit 400.

[0073]

[0085] In some examples, the housing 412 (Figure 4A) may be a sealed housing configured for subcutaneous implantation in a patient, and at least the power supply 402, memory, and processing circuit 50 are housed in a sealed case.

[0074]

[0086] Figure 4C is a perspective view showing another IMD10B, which may be another exemplary configuration of IMD10 derived from Figures 1A to 2A. IMD10B in Figure 4C may be configured substantially similarly to IMD10A in Figure 4A, and the differences between them are described herein.

[0075]

[0087] The IMD10B may include a leadless subcutaneous implantable monitoring device, such as an ICM. The IMD10B includes a housing having a base 440 and an insulating cover 442. The IMD10B includes (multiple) optical sensors 62. Proximal electrodes 16C and distal electrodes 16D may be formed or disposed on the outer surface of the cover 442. For example, the various circuits and components of the IMD10B described below with respect to Figure 2A may be formed or disposed on the inner surface of the cover 442 or within the base 440. In some examples, the battery or other power source of the IMD10B may be included within the base 440. In the illustrated example, the antenna 426B may be formed or disposed on the outer surface of the cover 442, but in some examples, it may be formed or disposed on the inner surface. In some examples, the insulating cover 442 may be positioned above the open base 440 such that the base 440 and the cover 442 surround the circuits and other components, protecting them from fluids such as bodily fluids.

[0076]

[0088] Circuits and components may be formed inside the insulating cover 442, for example, by using flip-chip technology. The insulating cover 442 may be inverted on the base 440. When inverted and placed on the base 440, the components of the IMD 10B formed inside the insulating cover 442 may be positioned within the gap 444 defined by the base 440. Electrodes 16C and 16D and the antenna 426B may be electrically connected to the circuits formed inside the insulating cover 442 through one or more vias (not shown) formed through the insulating cover 442. The insulating cover 442 may be formed of sapphire (i.e., corundum), glass, parylene, and / or any other suitable insulating material. The base 440 may be formed of titanium or any other suitable material (e.g., biocompatible material). Electrodes 16C and 16D may be formed from stainless steel, titanium, platinum, iridium, or an alloy thereof. In addition, electrodes 16C and 16D may be coated with materials such as titanium nitride or fractal titanium nitride, but other suitable materials and coatings may be used for such electrodes.

[0077]

[0089] In the example shown in Figure 4C, the housing of IMD10B defines a length L, a width W, and a thickness or depth D, and, similar to IMD10A in Figure 4C, it is in the form of an elongated rectangular prism where the length L is much greater than the width W and the width W is greater than the depth D. For example, the distance between the proximal electrode 16C and the distal electrode 16D may be in the range of 30 mm to 50 mm, 35 mm to 45 mm, or about 40 mm. In addition, IMD10B may have a length L in the range of 30 mm to about 70 mm. In other examples, the length L may be in the range of 5 mm to 60 mm, 40 mm to 60 mm, 45 mm to 55 mm, or about 45 mm. In addition, the width W may be in the range of 3 mm to 15 mm, such as about 8 mm. The thickness D of IMD10B may be in the range of 2 mm to 15 mm, or 3 mm to 5 mm, or about 4 mm. IMD10B may have a volume of 3 cubic centimeters (cm) or less, or 1.5 cubic centimeters or less, for example, about 1.4 cubic centimeters.

[0078]

[0090] In the example shown in Figure 4C, when inserted subcutaneously within the patient, the outer surface of the cover 442 faces outward toward the patient's skin. In addition, as shown in Figure 4C, the proximal end 446 and distal end 448 are rounded to reduce discomfort and irritation to surrounding tissues when inserted.

[0079]

[0091] Figure 5 is a flowchart illustrating exemplary techniques of operating system 2. As shown in Figure 5, the processing circuit can determine multiple blood pressure levels over a period of time based on optical signals (e.g., received reflected light signals) detected by a subcutaneously implanted medical device (e.g., an implantable cardiac monitor) (500). Based on the determined multiple blood pressure levels, the processing circuit 50 can determine blood pressure changes over a period of time (502). Based on the determined blood pressure changes, the processing circuit 50 can determine the health status of patient 4 (504). The processing circuit 50 can output instructions for the determined health status to the computing device 12 (506), which may facilitate the determination of cardiac health status and the risk of sudden cardiac death, and may lead to clinical interventions to control hypertension, such as medication and renal denervation.

[0080]

[0092] Figure 6 is a conceptual diagram showing an exemplary machine learning model 600 configured to determine blood pressure levels, changes in blood pressure levels, and / or health status. Machine learning model 600 is an example of the machine learning model described above. Machine learning model 600 is an example of a deep learning model or deep learning algorithm trained to determine blood pressure levels, changes in blood pressure levels, and / or health status. IMD10, computing device 12, server 24, and / or one or more other computing devices can train, store, and / or utilize machine learning model 600, but in other examples, other devices can apply inputs associated with a particular patient to machine learning model 600. As described above, in other examples, other types of machine learning models and deep learning models or machine learning algorithms and deep learning algorithms may be used. For example, the ResNet-18 convolutional neural network model can be used. Some non-exclusive examples of models that can be used for transfer learning include AlexNet, VGGNet, GoogleNet, ResNet50, or DenseNet. Some non-restrictive examples of machine learning techniques include Support Vector Machines, the K-Nearest Neighbor algorithm, and Multi-layer Perceptron.

[0081]

[0093] As shown in the example in Figure 6, the machine learning model 600 may include three layers. These three layers include an input layer 602, a hidden layer 604, and an output layer 606. The output layer 606 includes the output from the transfer function 605 of the output layer 606. The input layer 602 represents each of the input values ​​X1 to X4 provided to the machine learning model 600. In some examples, the input values ​​may be any of the values ​​input to the machine learning model as described above. For example, the input values ​​may be one or more feature parts extracted from the received optical signal as described above. In some examples, the input values ​​may be one or more of the following: inter-peak interval, systolic peak, time to systolic peak, systolic area, systolic gradient trajectory, diastolic peak, diastolic area, diastolic gradient trajectory, diastolic peak / systolic peak ratio, Boolean dichroic notch, dichroic notch height, mean, standard deviation, kurtosis, and / or skewness. In addition, in some examples, the input values ​​for the machine learning model 1500 may include additional data, such as data related to one or more additional parameters of patient 4.

[0082]

[0094] Each input value for each node in the input layer 602 is provided to each node in the hidden layer 604. In the example in Figure 6, the hidden layer 604 consists of two layers, one with four nodes and the other with three nodes, but in other examples, fewer or more nodes may be used. Each input from the input layer 602 is multiplied by a weight and then summed at each node in the hidden layer 604. During the training of the machine learning model 600, the weights of each input are adjusted to establish a relationship between the extracted cardiac features and determining blood pressure levels, changes in blood pressure levels, and / or health status. In some examples, one hidden layer may be incorporated into the machine learning model 600, or three or more hidden layers may be incorporated into the machine learning model 600, each layer containing the same or different number of nodes.

[0083]

[0095] The results of each node in the hidden layer 604 are applied to the transfer function of the output layer 606. The transfer function may be linear or nonlinear, depending on the number of layers in the machine learning model 600. Exemplary nonlinear transfer functions may be sigmoid functions or normalization functions. The output 607 of the transfer function may be a classification of scores indicating the patient's health status, such as blood pressure level, changes in blood pressure level, and / or severity of hypertension, generated by a computing device such as the processing circuit 50 or by a computing device, in response to applying cardiac features extracted from the received optical signal to the machine learning model 600.

[0084]

[0096] As shown in the example above, by extracting one or more specific cardiac features from the received optical signal and applying the extracted cardiac features to a machine learning model such as machine learning model 600 to determine a score indicating the patient's blood pressure level, changes in blood pressure level, and / or the patient's health status, the processing circuit 50 can determine the severity of the patient's hypertension with high accuracy, specificity, and sensitivity. This can facilitate the determination of the patient's cardiac health status and the risk of sudden cardiac death, which may lead to clinical interventions to control hypertension, such as medication and renal denervation.

[0085]

[0097] Figure 7 shows an example of a machine learning model 702 trained using supervised learning and / or reinforcement learning techniques. The machine learning model 702 can be implemented using any number of models for supervised learning and / or reinforcement learning, including, but not limited to, artificial neural networks, decision trees, naive Bayes networks, support vector machines, or k-nearest neighbor models. In some examples, one or more of the IMD 10, computing device 12, server 24, and / or other computing devices first train a machine learning model 702 corresponding to an received optical signal based on a training set of metrics. The training set 700 may include a set of feature vectors, where each feature in the feature vector represents a value of a particular metric. One or more of the IMD 10, computing device 12, server 24, and / or other computing devices may select a training set containing a set of training instances, where each training instance includes associations between one or more respective optical signal feature parts of their respective optical signal and their respective optical signal. The prediction or classification by the machine learning model 702 can be compared 704 with the target output 703. Based on the error signal representing the comparison, the processing circuit implementing the learning / training function 705 may send or apply modifications to the weights of the machine learning model 702, or otherwise modify / update the machine learning model 702. For example, one or more of the IMD 10, computing device 12, server 24, and / or (multiple) other computing devices may, for each training instance in the training set, modify the machine learning model 702 to change the score generated by the machine learning model 702 in response to subsequent optical signals applied to the machine learning model 702, based on the respective heart features and respective optical signals of the training instance.

[0086]

[0098] Figures 8A to 8C show different examples of flowcharts illustrating how the processing circuit 50 determines a health status based on optical signals received by (multiple) optical sensors 62. For example, the processing circuit 50 may preprocess, normalize, and / or filter (810) the optical signals 805. The processing circuit 50 may further extract features as described above (815). The processing circuit 50 may select the extracted features as described above (820). As shown in Figures 8A to 8B, the processing circuit 50 may input the selected features into a machine learning model 825. As shown as an example in Figure 8A, the processing circuit 50 may generate outputs of blood pressure event flags 830, such as classifying blood pressure indicating a health status. Based on the classification and / or health status, the processing circuit 50 may determine whether or not to generate a health alert. As shown in Figure 8B as an example, the processing circuit 50 may generate outputs of blood pressure waveform values ​​such as diastolic blood pressure, systolic blood pressure, and / or mean blood pressure (835). The processing circuit 50 may determine the health status based on the determined blood pressure values. The processing circuit 50 may determine whether to generate a health alert based on the determined health status and / or the determined blood pressure values. As shown in Figure 8C, the processing circuit 50 may determine the health status based on selected features and determine whether to generate a health alert based on the determined health status.

[0087]

[0099] Figures 9A and 9B show different examples of processing circuits of System 2, e.g., processing circuits 50, 23, and / or 230, which determine a health status based on optical signals received by (multiple) optical sensors 62 and / or features selected from the optical signals, as described above. For example, a machine learning model 900A, as shown in Figure 9A, may be a blood pressure event classification-based convolutional neural network. As shown in Figure 9A, the input 902 may be optical signals received by (multiple) optical sensors 62. The processing circuit may perform one or more convolutional actions 904A and 904B (collectively referred to as "convolution 904") and one or more subsampling actions 906A and 906B (collectively referred to as "subsampling 906") on the input data 902 to determine a map of features 908 that can be input for the classifier 912, for example, using additional features 910 derived from the input data 902. For example, as shown in the example in Figure 9A, the health status may be an indicator of a classified blood pressure event 914. For example, an optical signal classified as Class 1 may indicate a health status that would generate a health alert 916. An optical signal classified as Class 2 may indicate a health status that would not require the generation of a health alert.

[0088]

[0100] In some examples, the machine learning model may be a blood pressure regression-based CNN900B, as shown in Figure 9B. As shown in Figure 9B, the input 902 may be an optical signal received by (multiple) optical sensors 62. The processing circuit of System 2 may perform a series of operations on the input optical signal sample 902. In the example of Figure 9B, the processing circuit of System 2 may generate an input sequence 920 based on the optical signal sample and apply the input sequence to CNN922. The output of CNN922 may be applied to LSTM924. The processing circuit of System 2 applies the output of LSTM924 to one or more fully connected layers 926 that generate the regression output of model 900B. For example, as shown in the example of Figure 9B, the health status may be an index of blood pressure values, such as systolic blood pressure or diastolic blood pressure. The processing circuit may determine the health status based on the determined blood pressure value. The processing circuit may determine whether to generate a health alert 930 based on the determined health status and / or the determined blood pressure value.

[0089]

[0101] The technologies described herein may be implemented, at least in part, in hardware, software, firmware, or a combination thereof. For example, various aspects of the technologies described may be implemented in one or more processors or processing circuits, including one or more microprocessors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or any other equivalent integrated or discrete logic circuits, and any combination of such components. The terms “processor” or “processing circuit” may generally refer to any of the aforementioned logic circuits, either alone or in combination with other logic circuits or any other equivalent circuits. A control unit including hardware may also perform one or more of the technologies described herein.

[0090]

[0102] Such hardware, software, and firmware may be implemented within the same device or in separate devices to support the various operations and functions described herein. In addition, any of the described units, circuits, or components may be implemented together or separately as discrete but interoperable logic devices. Descriptions of different features of a circuit or unit are intended to highlight different functional aspects and do not necessarily imply that such a circuit or unit must be implemented by separate hardware or software components. Rather, the functionality associated with one or more circuits or units may be performed by separate hardware or software components, or integrated within a common or separate hardware or software component.

[0091]

[0103] The technologies described herein may also be embodied or encoded in a computer-readable medium, such as a computer-readable storage medium, which may contain instructions that may be described as non-temporary media. Instructions embedded in or encoded in a computer-readable storage medium may, for example, cause a programmable processor or other processor to execute a method when the instructions are executed. Examples of computer-readable storage media include random access memory (RAM), read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electronically erasable programmable read-only memory (EEPROM), flash memory, hard disks, CD-ROMs, floppy disks, cassettes, magnetic media, optical media, or any other computer-readable storage media.

[0092]

[0104] The following examples illustrate the techniques described herein.

[0093]

[0105] Example 1: The implantable medical device comprises a housing configured for subcutaneous implantation at a certain location within a patient; an optical sensor located on or inside the housing and configured to generate an optical signal indicating blood movement in the patient's blood vessels adjacent to a certain location over a certain period of time; and a processing circuit configured to determine a plurality of blood pressure levels over a certain period of time based on the optical signal, determine blood pressure changes over a certain period of time based on the determined plurality of blood pressure levels, determine the patient's health status based on the determined blood pressure changes over a certain period of time, and output an instruction for the determined health status.

[0094]

[0106] Example 2: The implantable device described in Example 1, wherein the optical sensor is a photoplethysmography sensor.

[0095]

[0107] Example 3: An embedded device according to any one of Examples 1 to 2, wherein the optical sensor is configured to emit an optical signal belonging to a specific wavelength spectrum, and the optical sensor is configured to receive a reflected optical signal of a specific wavelength spectrum.

[0096]

[0108] Example 4: The implantable device according to Example 3, wherein the specific wavelength spectrum is an amber wavelength spectrum or a green wavelength spectrum.

[0097]

[0109] Example 5: An implantable device according to any one of Examples 1 to 4, wherein the processing circuit is further configured to determine the patient's health status based on determined blood pressure changes over a certain period of time, by extracting one or more cardiac features from the optical signal and applying the extracted cardiac features to a machine learning model to determine the patient's health status.

[0098]

[0110] Example 6: An implantable device according to any one of Examples 1 to 4, wherein the processing circuit is further configured to determine the patient's health status based on determined blood pressure changes over a certain period of time, by extracting one or more cardiac features from an optical signal, determining changes in the extracted cardiac features over a certain period of time, and applying one or more of the extracted cardiac features or the determined changes in the extracted cardiac features to a machine learning model to determine the patient's health status.

[0099]

[0111] Example 7: An implantable device according to any one of Examples 1 to 4, wherein the processing circuit is further configured to apply optical signals to a machine learning model to determine the patient's health status.

[0100]

[0112] Example 8: An implantable device according to any of Examples 5 to 7, wherein the machine learning model is trained with training data having wavelength spectra corresponding to the wavelength spectra of optical signals.

[0101]

[0113] Example 9: An embedded device according to any of Examples 1 to 8, wherein the processing circuit is further configured to identify the wavelength spectrum of the received optical signal.

[0102]

[0114] Example 10: An implantable device according to any one of Examples 1 to 9, further comprising one or more second sensors for monitoring one or more physiological parameters of a patient, the one or more physiological parameters being different from blood pressure, and the processing circuit further configured to determine the patient's health status based on the determined blood pressure changes and one or more physiological parameters of the patient.

[0103]

[0115] Example 11: The implantable device according to Example 10, wherein one or more second sensors comprise at least one of an accelerometer or an electrode.

[0104]

[0116] Example 12: An implantable device according to any one of Examples 10 to 11, wherein one or more physiological parameters include at least one of heart rate, cardiac output, heart sounds, or impedance.

[0105]

[0117] Example 13: The implantable device is an implantable cardiac monitor, the implantable cardiac monitor comprising: a power supply operably connected to a processing circuit; a memory operably connected to the processing circuit and configured to store a machine learning model; a distal electrode operably connected to the processing circuit; a proximal electrode operably connected to the processing circuit; and a sealed housing configured for subcutaneous implantation in a patient, wherein at least the power supply, memory, and processing circuit are located within the sealed housing, and the housing has length, width, and depth, the length being greater than the width, the width being greater than the depth, the length being in the range of 5 mm to 60 mm, the width being in the range of 5 mm to 15 mm, and the depth being in the range of 5 mm to 15 mm, as described in any of Examples 1 to 12.

[0106]

[0118] Example 14: The health status is the severity of hypertension, and the implantable device is as described in any of Examples 1 to 13.

[0107]

[0119] Example 15: The system comprises an implantable medical device (IMD) comprising a housing configured for subcutaneous implantation at a location within a patient, and an optical sensor located on or inside the housing and configured to generate an optical signal indicating blood movement in the patient's blood vessels adjacent to a location over a period of time, and one or more computing devices communicably connected to the IMD, the one or more computing devices comprising a memory, and a processing circuit connected to the memory, the processing circuit configured to determine a plurality of blood pressure levels over a period of time based on the optical signal, determine a change in blood pressure over a period of time based on the determined plurality of blood pressure levels, determine the patient's health status based on the determined change in blood pressure over a period of time, and output an instruction for the determined health status.

[0108]

[0120] Example 16: The system described in Example 15, wherein the optical sensor is a photoplethysmography sensor.

[0109]

[0121] Example 17: The system according to any one of Examples 15 to 16, wherein the optical sensor is configured to emit an optical signal belonging to a specific wavelength spectrum, and the optical sensor is configured to receive a reflected optical signal of a specific wavelength spectrum.

[0110]

[0122] Example 18: The system according to Example 17, wherein the specific wavelength spectrum is the amber wavelength spectrum or the green wavelength spectrum.

[0111]

[0123] Example 19: The system according to any one of Examples 15 to 18, wherein the processing circuit is further configured to determine the patient's health status based on determined blood pressure changes over a certain period of time, by extracting one or more cardiac features from the optical signal and applying the extracted cardiac features to a machine learning model to determine the patient's health status.

[0112]

[0124] Example 20: The system according to any one of Examples 15 to 18, wherein the processing circuit is further configured to determine the patient's health status based on determined blood pressure changes over a certain period of time, by extracting one or more cardiac features from an optical signal, determining changes in the extracted cardiac features over a certain period of time, and applying one or more of the extracted cardiac features or the determined changes in the extracted cardiac features to a machine learning model to determine the patient's health status.

[0113]

[0125] Example 21: The system according to any one of Examples 15 to 18, wherein the processing circuit is further configured to apply the optical signal to a machine learning model to determine the patient's health status.

[0114]

[0126] Example 22: The system according to any one of Examples 19 to 21, wherein the machine learning model is trained with training data having wavelength spectra corresponding to the wavelength spectra of optical signals.

[0115]

[0127] Example 23: The system according to any one of Examples 15 to 22, wherein the processing circuit is further configured to identify the wavelength spectrum of the optical signal.

[0116]

[0128] Example 24: The system according to any one of Examples 15 to 23, wherein the implantable medical device further includes one or more second sensors for monitoring one or more physiological parameters of the patient, the one or more physiological parameters being different from blood pressure, and the processing circuit is further configured to determine the patient's health status based on the determined blood pressure changes and one or more physiological parameters of the patient.

[0117]

[0129] Example 25: The system according to Example 24, wherein one or more second sensors include at least one of an accelerometer or an electrode.

[0118]

[0130] Example 26: The system according to any one of Examples 23-24, wherein one or more physiological parameters are at least one of heart rate, cardiac output, heart sounds, or impedance.

[0119]

[0131] Example 27: The implantable device is an implantable cardiac monitor, the implantable cardiac monitor comprising: a power supply operably connected to an implantable cardiac monitor processing circuit; a memory operably connected to the implantable cardiac monitor processing circuit and configured to store a machine learning model; a distal electrode operably connected to the implantable cardiac monitor processing circuit; a proximal electrode operably connected to the implantable cardiac monitor processing circuit; and a sealed housing configured for subcutaneous implantation in a patient, wherein at least the power supply, memory, and implantable cardiac monitor processing circuit are located within the sealed case, and the housing has length, width, and depth, with the length being greater than the width and the width being greater than the depth, the length being in the range of 5 mm to 60 mm, the width being in the range of 5 mm to 15 mm, and the depth being in the range of 5 mm to 15 mm, as described in any of Examples 15 to 26.

[0120]

[0132] Example 28: The health status is the severity of hypertension, using the system described in any of Examples 15-27.

[0121]

[0133] Example 29: A method for operating a processing circuit of a medical system includes: the processing circuit receiving an optical signal indicating blood movement within the patient's blood vessels via an implantable medical device including an optical sensor implanted subcutaneously in the patient; the processing circuit determining multiple blood pressure levels of the patient over a period of time based on the optical signal; the processing circuit determining blood pressure changes over a period of time based on the determined multiple blood pressure levels; the processing circuit determining the patient's health status based on the determined blood pressure changes over a period of time; and the processing circuit outputting an indication of the determined health status.

[0122]

[0134] Example 30: The optical sensor is a photoplethysmography sensor, as described in Example 29.

[0123]

[0135] Example 31: The optical signal belongs to a specific wavelength spectrum, as described in any of Examples 29 to 30.

[0124]

[0136] Example 32: The method according to Example 31, wherein the specific wavelength spectrum is an amber wavelength spectrum or a green wavelength spectrum.

[0125]

[0137] Example 33: A method according to any one of Examples 29 to 32, for determining a patient's health status based on determined blood pressure changes over a period of time, further comprising: a processing circuit extracting one or more cardiac features from an optical signal; and a processing circuit applying the extracted cardiac features to a machine learning model to determine the patient's health status.

[0126]

[0138] Example 34: A method according to any one of Examples 29 to 32 for determining a patient's health status based on determined blood pressure changes over a period of time, comprising: a processing circuit extracting one or more cardiac features from an optical signal; the processing circuit determining changes in the extracted cardiac features over a period of time; and the processing circuit applying one or more of the extracted cardiac features or the determined changes in the extracted cardiac features to a machine learning model to determine the patient's health status.

[0127]

[0139] Example 35: The method according to any one of Examples 29 to 32, further comprising a processing circuit applying an optical signal to a machine learning model to determine the patient's health status.

[0128]

[0140] Example 36: The method according to any one of Examples 29 to 35, wherein the machine learning model is trained with training data having wavelength spectra corresponding to the wavelength spectra of optical signals.

[0129]

[0141] Example 37: The method according to any one of Examples 29 to 36, further comprising identifying the wavelength spectrum of an optical signal.

[0130]

[0142] Example 38: The method according to any one of Examples 29 to 36, further comprising: a processing circuit receiving one or more physiological parameters of a patient via one or more second sensors, wherein the one or more physiological parameters are different from blood pressure; and the processing circuit determining the patient's health status based on the determined blood pressure changes and the one or more physiological parameters of the patient.

[0131]

[0143] Example 39: The method according to Example 38, wherein one or more second sensors include at least one of an accelerometer or an electrode.

[0132]

[0144] Example 40: The method according to any one of Examples 38 to 39, wherein one or more physiological parameters are at least one of heart rate, cardiac output, heart sounds, or impedance.

[0133]

[0145] Example 41: The health status is the severity of hypertension, according to any of the methods in Examples 29 to 40.

[0134]

[0146] Example 42: A non-temporary computer-readable storage medium that stores instructions causing one or more processors, when executed, to receive optical signals indicating blood movement within a patient's blood vessels via an implantable medical device, including at least an optical sensor implanted subcutaneously within the patient; determine multiple blood pressure levels of the patient over a period of time based on the optical signals; determine blood pressure changes over a period of time based on the determined multiple blood pressure levels; determine the patient's health status based on the determined blood pressure changes over a period of time; and output instructions for the determined health status.

[0135]

[0147] Various embodiments are described. These embodiments and other embodiments are within the scope of the following claims.

Claims

1. It is an implantable medical device, A housing configured for subcutaneous implantation at a specific location within the patient, An optical sensor, wherein the optical sensor is located on or inside the housing and is configured to generate an optical signal indicating blood movement within the blood vessels of the patient near a certain location over a certain period of time. A processing circuit, Based on the optical signal, multiple blood pressure levels over a certain period are determined, Based on the multiple blood pressure levels determined above, the blood pressure changes over a certain period are determined, Based on the blood pressure changes determined over the aforementioned period, the patient's health status is determined, An implantable medical device comprising a processing circuit configured to output instructions for the determined health status.

2. The optical sensor is configured to emit an optical signal belonging to a specific wavelength spectrum, The embedded device according to claim 1, wherein the optical sensor is configured to receive a reflected light signal of the specific wavelength spectrum.

3. Based on the blood pressure changes determined over the aforementioned period, the processing circuit determines the patient's health status, One or more cardiac features are extracted from the optical signal, The implantable device according to any one of claims 1 to 2, further configured to apply the extracted cardiac features to a machine learning model to determine the patient's health status.

4. Based on the blood pressure changes determined over the aforementioned period, the processing circuit determines the patient's health status, One or more cardiac features are extracted from the optical signal, The changes in the extracted cardiac features over the aforementioned period are determined, The implantable device according to any one of claims 1 to 2, further configured to apply one or more of the extracted cardiac feature portions or the determined changes in the extracted cardiac feature portions to a machine learning model to determine the health status of the patient.

5. The aforementioned processing circuit is The implantable device according to any one of claims 1 to 3, further configured to apply the optical signal to a machine learning model to determine the health status of the patient.

6. The embedded device according to any one of claims 3 to 5, wherein the machine learning model is trained with training data having a wavelength spectrum corresponding to the wavelength spectrum of the optical signal.

7. The embedded device according to any one of claims 1 to 6, wherein the processing circuit is further configured to identify the wavelength spectrum of the received optical signal.

8. One or more second sensors for monitoring one or more physiological parameters of the patient, wherein the one or more physiological parameters are different from blood pressure, and further comprising one or more second sensors, The implantable device according to any one of claims 1 to 7, wherein the processing circuit is further configured to determine the patient's health status based on the determined blood pressure change and one or more physiological parameters of the patient.

9. The implantable device is an insertable cardiac monitor, and the insertable cardiac monitor is A power supply operably connected to the aforementioned processing circuit, A memory configured to be operably connected to the processing circuit and to store the machine learning model, A distal electrode operably connected to the processing circuit, A proximal electrode operably connected to the processing circuit, A sealed housing configured for subcutaneous implantation within the patient, comprising a sealed housing in which at least the power supply, memory, and processing circuit are located within the sealed case, The housing has a length, width, and depth, The length is greater than the width, and the width is greater than the depth. The aforementioned length is within the range of 5 millimeters (mm) to 60 mm. The aforementioned width is within the range of 5 mm to 15 mm. The implantable device according to any one of claims 1 to 8, wherein the depth is in the range of 5 mm to 15 mm.

10. It is a system, An implantable medical device (IMD), A housing configured for subcutaneous implantation at a specific location within the patient, An implantable medical device (IMD) includes an optical sensor, the optical sensor located on or inside the housing, and configured to generate an optical signal indicating blood movement within the patient's blood vessels near a certain location over a period of time; One or more computing devices communicably connected to the IMD, wherein the one or more computing devices are Memory and A processing circuit connected to the memory, wherein the processing circuit is Based on the optical signal, multiple blood pressure levels over a certain period are determined, Based on the multiple blood pressure levels determined above, the blood pressure changes over a certain period are determined, Based on the blood pressure changes determined over the aforementioned period, the patient's health status is determined, A system comprising one or more computing devices, including a processing circuit configured to output instructions for the determined health status.

11. The optical sensor is configured to emit an optical signal belonging to a specific wavelength spectrum, The system according to claim 10, wherein the optical sensor is configured to receive a reflected light signal of the specific wavelength spectrum.

12. Based on the blood pressure changes determined over the aforementioned period, the processing circuit determines the patient's health status, One or more cardiac features are extracted from the optical signal, The system according to any one of claims 10 to 11, further configured to apply the extracted cardiac features to a machine learning model to determine the patient's health status.

13. Based on the blood pressure changes determined over the aforementioned period, the processing circuit determines the patient's health status, One or more cardiac features are extracted from the optical signal, The changes in the extracted cardiac features over the aforementioned period are determined, The system according to any one of claims 10 to 11, further configured to apply one or more of the extracted cardiac feature portions or the determined changes in the extracted cardiac feature portions to a machine learning model to determine the health status of the patient.

14. The aforementioned processing circuit is The system according to any one of claims 10 to 11, further configured to apply the optical signal to a machine learning model to determine the health status of the patient.

15. The system according to any one of claims 12 to 14, wherein the machine learning model is trained with training data having a wavelength spectrum corresponding to the wavelength spectrum of the optical signal.

Images