Smart Machines and Machine Learning for Hemodynamic Support Devices

Abstract

Methods and systems are provided for using a smart hemodynamic support device positioned in one region of the body in combination with machine learning to infer and / or detect a condition in another region of the body operatively connected by blood flow, comprising: a trained machine learning (ML) algorithm, a first probability of a condition being present in a second region of the subject's body, different from the first region, based on received data;

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to U.S. Provisional Patent Application No. 63 / 285,302, filed December 2, 2021, which is incorporated by reference in its entirety.

[0002] Technical Field The present disclosure relates to hemodynamic support devices, and in particular to devices that infer problems in one region of a subject's body based on conditions detected in other regions of the body. [Background technology]

[0003] background Conventional medical devices designed to be inserted into or onto specific locations on a patient's body routinely incorporate sensors. Such sensors reflect the conditions at wherever the medical device is positioned. Although information received from the sensors can be displayed and ranges for normal operation determined, such approaches have a number of disadvantages. For example, when information from a sensor is determined to fall outside of normal ranges, such approaches may not be able to determine whether there is an actual problem and / or the most likely cause of the problem. Summary of the Invention [Means for solving the problem]

[0004] overview In some embodiments, a system for detecting and / or inferring a condition may be provided. The system may include a hemodynamic support device configured to be positioned in a first region of a subject's body and at least one processor operably coupled to the hemodynamic support device. The at least one processor may be configured to receive data from the hemodynamic support device. The at least one processor may also be configured to determine, using at least one trained machine learning (ML) algorithm, a first probability of the condition being present in a second region of the subject's body based on the received data, the second region being different from the first region.

[0005] In some embodiments, the at least one processor may include a first processor and a second processor. The first processor may be configured to receive data from the hemodynamic support device and send the data to the second processor over the network. The second processor may be configured to receive data from the first processor and use a trained machine learning algorithm to determine a first probability of a condition existing in a second region of the subject's body based on the received data. In some embodiments, the second processor may be further configured to send the first probability to the first processor.

[0006] In some embodiments, the system may include a remote device. In some embodiments, the second processor may be further configured to send the first probability to the remote device. The remote device may be configured to display the first probability, or text or an image representing the first probability. In some embodiments, the remote device may be configured to send the second data and the third data to the at least one processor, and to receive the first probability, the second probability, and the third probability from the at least one processor. In some embodiments, non-user identifiable data is sent to or from the remote device. In some embodiments, the remote device is a mobile phone, tablet, or laptop. A user, such as a nurse, clinician, medical personnel, etc., may be a user of the remote device.

[0007] In some embodiments, the at least one trained ML algorithm may include a first ML algorithm trained on historical data collected from multiple medical device models (e.g., some or all of Abiomed's Impella® blood pump models). In some embodiments, the at least one trained ML algorithm may include a second ML algorithm trained on data collected from a single medical device model (e.g., only Abiomed's Impella® 5.5 with Smart Assist).

[0008] In some embodiments, after making a first probability determination, the system may be configured to continue to refine the probability estimates based on new information transmitted to the one or more processors.

[0009] In some embodiments, the at least one processor may be further configured to receive second data (e.g., entered by a user at a remote device or sent from a medical device to the one or more processors) after receiving the data from the hemodynamic support device. The second data may relate to a third region of the subject's body different from the first region and the second region. For example, if the first region is the patient's left heart and the second region is the patient's right heart, the third region may be the patient's vena cava. In some embodiments, the second data may include a value related to central venous pressure (CVP). In some embodiments, the at least one trained ML algorithm may be further configured to determine a second probability of the condition based on the data from the hemodynamic support device and the second data.

[0010] In some embodiments, the at least one processor may be configured to receive third data after receiving the second data, the third data relating to a fourth region of the subject's body different from the first region, the second region, and the third region. For example, if the first, second, and third regions are the left heart, the right heart, and the vena cava, the fourth region may be, for example, the pulmonary artery. In some embodiments, the third data may include a value relating to pulmonary artery pulsatility (PAP). In some embodiments, the at least one trained ML algorithm may be further configured to determine a third probability of the condition based on the data from the hemodynamic support device, the second data, and the third data.

[0011] At least one trained ML algorithm may also be trained to take into account the derived data. In some embodiments, the at least one processor may be further configured to derive at least one parameter, and the at least one trained ML algorithm may be further configured to determine a probability (e.g., a first, second, or third probability as disclosed herein) based on the at least one parameter, where the at least one parameter may be central venous pressure (CVP), right atrial pressure (RAP), minimum, maximum, and / or mean value of RAP, right ventricular end-diastolic pressure (RVEDP), pulmonary artery pressure (PAP), mean, systolic, and / or diastolic PAP, pulmonary artery pressure index (PAPI), and / or echo-based parameters of right heart function. In some embodiments, PAPI may be calculated by subtracting diastolic pulmonary artery value (PAdia) from systolic pulmonary artery value (PAsys) and then dividing the difference by RAP or CVP. In some embodiments, the echo-based parameters of right heart function may be right ventricular (RV) diameter, RV volume, RV stroke volume index (RVSVI) value, RV stroke work index (RVSWI) value, and / or tricuspid annular systolic excursion (TAPSE) value.

[0012] In some embodiments, the at least one processor may be further configured to derive at least one parameter, and the at least one trained ML algorithm may be further configured to determine a probability (e.g., a first, second, or third probability as disclosed herein) based on the at least one parameter, where the at least one parameter may be LV end-diastolic pressure (LVEDP), pump aspiration, pump alarm rate and / or type, cardiac output, LV contractility, LV relaxation, pulse wave velocity, ejection fraction, a statistical metric of a parameter included in the data from the hemodynamic support device, and / or a systolic value, diastolic value, mean, median, minimum, maximum, delta, or pulse of a parameter included in the data from the hemodynamic support device.

[0013] In some embodiments, the system may include other data collection devices. For example, in one embodiment, the system may include an additional device (or devices) operably coupled to at least one processor. The one or more additional devices may include a sensor, where the sensor may be positioned in or on a third region of the patient's body. For example, a watch with a sensor may be placed around the subject's wrist, or a patch with a sensor may be placed around the subject's chest.

[0014] In some embodiments, the at least one trained ML algorithm may be further configured to determine a probability (e.g., a first probability, a second probability, or a third probability as disclosed herein) based on data received from a sensor of the additional device.

[0015] In some embodiments, the data received from the additional device's sensors may include heart rate, values ​​related to blood oxygen, values ​​related to an electrocardiogram (ECG), skin temperature, or acceleration.

[0016] In some embodiments, the data from the hemodynamic support device may include first information regarding left heart contractile function and second information regarding suction or pump flow in the left heart. In some embodiments, the first information may include a left ventricular (LV) contractility value, an aortic (AO) pulse pressure and / or a pulsatility value, or a combination thereof. In some embodiments, the data from the hemodynamic support device may include an aortic (AO) pressure, a left ventricular (LV) pressure, a pump motor speed, a pump motor current, an LV-AO pressure gradient, a pump flow, a cardiac output, a native cardiac output, an LV rate, an AO pulse rate, or a combination thereof. In some embodiments, the data from the hemodynamic support device may also include an LV volume (e.g., by conductance), a heart rate, a heart rhythm, an arterial pressure, blood oxygenation, or a combination thereof.

[0017] In some embodiments, the at least one processor may be further configured to analyze the determined probability in various ways. In some embodiments, the one or more processors may be configured to determine whether the first probability is above a first threshold and / or below a second threshold. In some embodiments, the one or more processors may be configured to determine a trend in the probability of the subject's condition over time. In some embodiments, the one or more processors may be configured to determine whether the rate of change described by the trend is above a threshold rate and / or if the trend continues, the probability will be above the first threshold or below the second threshold within a predetermined period of time. In some embodiments, the one or more processors may be configured to identify one or more key factors that cause the probability to be above the first predetermined threshold and / or below the second predetermined threshold. In some embodiments, the one or more processors may be configured to determine a trend in the identified key factors over time. In some embodiments, various combinations of these are used.

[0018] In some embodiments, the at least one processor may be further configured to alert the user when the first probability is determined to be above a first predetermined threshold or below a second predetermined threshold, when the rate of change explained by the trend exceeds a threshold rate, and / or when it is determined that if the trend continues, the probability will be above the first threshold or below the second threshold within a predetermined time period.

[0019] In some embodiments, the at least one processor may be further configured to receive input from a user responsive to the alert. For example, the input responsive to the alert may include data indicating that a particular treatment has been administered or that an existing treatment has been discontinued. In some embodiments, the alert may include options for the user to select. In some embodiments, the options may include testing and / or treatment for the condition.

[0020] In some embodiments, the at least one processor may be further configured to track the probability of the condition over time to determine whether a treatment administered is effective in reducing the risk of the condition.

[0021] In some embodiments, a method for detecting and / or inferring a condition may be provided. The method may include receiving data from a hemodynamic support device positioned in a first region of the subject's body. In some embodiments, the data may include, for example, aortic (AO) pressure, left ventricular (LV) pressure, pump motor speed, pump motor current, LV-AO pressure gradient, pump flow, cardiac output, intrinsic cardiac output, LV rate, AO pulse rate, or combinations thereof. In some embodiments, the data may also include LV volume by conductance, heart rate, heart rhythm, arterial pressure, blood oxygenation, or combinations thereof.

[0022] The method may include determining, using a trained machine learning algorithm, a first probability of the condition being present in a second region of the subject's body based on the received data, the second region being different from the first region.

[0023] In some embodiments, the data may be received from the hemodynamic support device by a first device (e.g., a local controller with a first processor) and the first probability may be determined by a second device (e.g., a remote server with a second processor). In some embodiments, the method may include sending the data over a network to the second device. In some embodiments, the method may include sending the first probability to the first device. In some embodiments, the method may include sending the first probability to a third device, the third device configured to display the first probability or text or an image representing the first probability.

[0024] In some embodiments, the method may include providing a first ML algorithm trained on historical data collected from multiple medical device models. In some embodiments, the method may include providing a second ML algorithm trained on data collected from a single medical device model.

[0025] In some embodiments, the method may include receiving second data after receiving the data from the hemodynamic support device (e.g., from a user-controlled device, such as a mobile phone, or from a medical device controlled by the user), the second data relating to a third region of the subject's body, different from the first region and the second region. In some embodiments, the method may include determining a second probability of the condition based on the data from the hemodynamic support device and the second data, using at least one trained ML algorithm. In some embodiments, the method may include receiving third data after receiving the second data from the user (e.g., from a user-controlled device, etc.), the third data relating to a fourth region of the subject's body, different from the first region, the second region, and the third region. In some embodiments, the method may include determining a third probability of the condition based on the data from the hemodynamic support device, the second data, and the third data, using at least one trained ML algorithm.

[0026] As disclosed herein, the method may also include deriving at least one parameter that may be used to determine one or more probabilities. Such parameters may include central venous pressure (CVP), right atrial pressure (RAP), minimum, maximum, and / or mean RAP, right ventricular end-diastolic pressure (RVEDP), pulmonary artery pressure (PAP), mean, systolic, and / or diastolic PAP, pulmonary artery pressure index (PAPI), echo-based parameters of right heart function, LV end-diastolic pressure (LVEDP), pump aspiration, pump alarm rate and / or type, cardiac output, LV contractility, LV relaxation, pulse wave velocity, ejection fraction, statistical metrics of parameters included in data from a hemodynamic support device, and / or systolic, diastolic, mean, median, minimum, maximum, delta, or pulse of parameters included in data from a hemodynamic support device, or combinations thereof.

[0027] In some embodiments, the method may include transmitting the first probability, the second probability, and the third probability to a user controlled device (e.g., a cell phone, a tablet, a laptop, etc.). The method may include preventing user identifiable data from being sent to or from the user controlled device.

[0028] In some embodiments, the method may include receiving data from an additional device (e.g., a watch or patch) including a sensor, the sensor being positioned in or on a third region of the patient's body. The data received from the additional device may include, for example, a heart rate, a value related to blood oxygen, a value related to an electrocardiogram (ECG), skin temperature, or acceleration. In some embodiments, the at least one trained ML algorithm may further determine the first probability based on the data received from the sensor of the additional device.

[0029] In some embodiments, the method may include determining whether the first probability is above a first threshold and / or below a second threshold. In some embodiments, the method may include determining a trend over time of the probability of the subject's condition, and optionally determining whether the rate of change described by the trend is above a threshold rate, and / or if the trend continues, whether the probability will be above the first threshold or below the second threshold within a predetermined period of time. In some embodiments, the method may include identifying one or more key factors that cause the probability to be above the first predetermined threshold and / or below the second predetermined threshold. In some embodiments, the method may include determining a trend over time of the identified key factors.

[0030] In some embodiments, the method may include alerting the user of a time when a certain condition is met. In some embodiments, the user may be alerted when a first probability is determined to be above a first predetermined threshold or below a second predetermined threshold, when a rate of change explained by the trend exceeds a threshold rate, and / or when it is determined that if the trend continues, the probability will be above the first threshold or below the second threshold within a predetermined time period.

[0031] In some embodiments, the method may include receiving input from a user in response to the alert. In some embodiments, the input in response to the alert may include data indicating that a particular treatment has been administered or that an existing treatment has been discontinued. In some embodiments, the alert may include options for the user to select, the options including testing and / or treatment for the condition. In some embodiments, the method may include tracking the probability of the condition over time. In some embodiments, the method may include determining whether the administered treatment is (or was) effective in reducing the risk of the condition.

[0032] BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and, together with the summary of the invention provided above and the detailed description of the embodiments provided below, serve to explain the principles of the invention. [Brief description of the drawings]

[0033] [Figure 1] FIG. 1 is a schematic diagram of an embodiment of a system according to an embodiment of the present disclosure. [Diagram 2] 1 is a representation of one embodiment of a user interface. [Figure 3A] FIG. 1 is a block diagram illustrating an embodiment of a machine learning model configuration. [Figure 3B] FIG. 1 is a block diagram illustrating an embodiment of a machine learning model configuration. [Figure 3C] FIG. 1 is a block diagram illustrating an embodiment of a machine learning model configuration. [Figure 4] 1 is a flow chart illustrating one embodiment of a validation and training process. [Diagram 5] 1 is a flow chart illustrating one embodiment of a cycle of collecting data and updating probabilities. [Figure 6] 1 is a flow chart illustrating one embodiment for detecting and / or inferring a condition. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0034] Detailed Description The following description and drawings explain the principles of the present disclosure. Therefore, it is understood that a person skilled in the art can devise various arrangements that are not explicitly described or illustrated herein but can embody the principles of the present invention and fall within its scope. Moreover, it is expressly intended that all examples listed herein are merely for illustration to assist the reader in understanding the principles and concepts of the present disclosure by the inventors to advance the art, and are not limited to the examples and conditions specifically listed as such. Moreover, the term "or" as used herein refers to non-exclusive, i.e., unless otherwise indicated (e.g., "or else" or "or in the alternative"). Also, the various embodiments described herein are not necessarily mutually exclusive, since some embodiments can be combined with one or more other embodiments to form new embodiments.

[0035] Many of the innovative teachings of the present application are described with particular reference to presently preferred exemplary embodiments. However, it should be understood that this class of embodiments provides only a few of the many advantageous uses of the teachings disclosed herein. In general, statements made in the specification of the present application do not necessarily limit any of the various claimed disclosures. Moreover, some statements may apply to some disclosed features, but not to others.

[0036] Disclosed herein are methods and systems that use a smart hemodynamic support device positioned in one region of a subject's body and machine learning to infer and / or detect the condition of other regions of the subject's body, such as those operatively connected via the blood flow. For example, the inventors have realized a benefit of the disclosed methods and systems that may enable detection of right heart failure (RHF) from signals received from a pump positioned in the left heart.

[0037] In some embodiments, a system for detecting and / or inferring a condition may be provided. Referring to Figure 1, the system 100 may include a hemodynamic support device 110 configured to be positioned in a first region 103 of a body of a subject 101. The hemodynamic support device may include a blood pump 112, which may be coupled, for example, to a distal end of a catheter 113. In Figure 1, the blood pump is shown as being located in the heart 102, specifically the left heart.

[0038] The system may include at least one processor, such as processor 120. The processor may be operably coupled to the hemodynamic support device by either a wired or wireless connection. In some embodiments, the hemodynamic support device may be removably coupled to the controller 121 by one or more wires 126. The controller may include the processor 120 operably coupled to a memory 122, a non-transitory computer-readable storage medium 123, a display 124, and / or one or more controls 125, such as mechanical controls (e.g., buttons or knobs) or non-mechanical controls (e.g., a touch screen). As will be appreciated, the non-transitory computer-readable storage medium may include instructions that, when executed by the processor, configure the processor to perform certain steps.

[0039] The hemodynamic support device may include one or more sensors 111. Such sensors may be any suitable sensors for collecting the intended data, as described herein, and may include, for example, electrodes, optical sensors, and / or pressure sensors.

[0040] In some embodiments, the at least one processor can be configured to receive data from a hemodynamic support device.

[0041] In some embodiments, data from the hemodynamic support device may include, for example, aortic (AO) pressure, left ventricular (LV) pressure, pump motor speed, pump motor current, LV-AO pressure gradient, pump flow, cardiac output, intrinsic cardiac output, LV rate, AO pulse rate, or combinations thereof. In some embodiments, the data may also include LV volume (e.g., by conductance), heart rate, heart rhythm, arterial pressure, blood oxygenation, or combinations thereof.

[0042] In some embodiments, the one or more processors may be configured to determine or derive a parameter based on the data received from the hemodynamic support device. In some embodiments, the at least one parameter may be: LV end-diastolic pressure (LVEDP); pump aspiration; pump alarm rate and / or type; cardiac output (CPO); LV contractility; LV relaxation; pulse wave velocity; ejection fraction; a statistical metric of a parameter included in the data from the hemodynamic support device; and / or a systolic value, a diastolic value, a mean, a median, a minimum, a maximum, a delta, or a pulse of a parameter included in the data from the hemodynamic support device.

[0043] In some embodiments, cardiac output (CPO) may be calculated by Mean Arterial Pressure (MAP) x CO(Cardiac Output) / 451, where Mean Arterial Pressure (MAP) = [(Systolic Blood Pressure - Diastolic Blood Pressure) / 3] + Diastolic Blood Pressure.

[0044] The one or more processors may be configured to use at least one trained machine learning (ML) algorithm to determine, based on the received data, a first probability of a condition being present in a second region of the subject's body, the second region being different from the first region.

[0045] In some embodiments, the condition may be right heart failure (RHF), although additional or alternative conditions may also be incorporated based on where the hemodynamic support device is placed.

[0046] In some embodiments, the data from the hemodynamic support device may include first information regarding the function of the region in which the hemodynamic support device is located and second information regarding the function of the hemodynamic support device, in some embodiments, the first information may include information regarding left heart systolic function (e.g., left ventricular (LV) contractility, aortic (AO) pulse pressure and / or pulsatility, or a combination thereof) and the second information is related to suction or pump flow within the left heart.

[0047] In some embodiments, a processor in the controller may include at least one trained ML algorithm and thus may be configured to make such determinations locally. In some embodiments, the processor 120 may be configured to display some or all of the data received from the hemodynamic support device on the display 124. In some embodiments, the processor 120 may be configured to display some or all of the determinations (e.g., one or more probabilities) from one or more ML algorithms. These values ​​may be displayed in any suitable manner; for example, in some embodiments, a number or other text may be displayed, a graph or trend may be displayed, or both.

[0048] In some embodiments, the local controller may not be configured to make the decision. In some embodiments, the at least one processor may include a first processor (e.g., processor 120) and a second processor (e.g., processor 130) that may be located remotely, such as on a cloud-based server. As will be appreciated, the second processor, like the first processor, may be coupled to memory 132 and non-transitory computer-readable storage medium 133.

[0049] In some embodiments, the first processor can be configured to receive data from the hemodynamic support device and send the data over the network to the second processor.

[0050] In some embodiments, the second processor may be configured to receive data from the first processor and use a trained machine learning algorithm to determine a first probability of a condition being present in a different region of the subject's body based on the received data, i.e., the second processor may incorporate one or more trained ML algorithms rather than, for example, a local controller.

[0051] In some embodiments, the second processor (e.g., processor 130) may be configured to send the determined probabilities to the first processor (e.g., processor 120). The first processor may then display the probabilities, or a representation of the probabilities, on a display.

[0052] In some embodiments, the system may include a remote device 140 that may be associated with a user 145. For example, a user may include a health care provider, a researcher, a subject, or an assistant.

[0053] In some embodiments, the remote device may be a mobile phone, laptop, tablet, or other computer or computing device. The remote device may include a display 141, which may be a touch-sensitive display. The remote device may be configured to provide a user interface for a user, e.g., to allow the user to interact with the remote device. The user interface may include one or more screens to be displayed to the user. The remote device may include input devices (not shown), which may include a touch-sensitive display, a mouse, a keyboard, etc., to allow the user to make selections, input data, and / or interact with the user interface.

[0054] In some embodiments, the one or more processors that determine the probability (e.g., processor 120 or processor 130) may be further configured to send the first probability to a remote device, which may be sent over one or more networks, such as the Internet. In some embodiments, the remote device may be configured to display the first probability, or text or an image representing the first probability.

[0055] 2, which may be shown, for example, on a remote device and / or a controller. The user interface may include one or more display screens 200, each of which may be divided into one or more sections, such as a first section 210, a second section 220, and a third section 230. Each section may be configured to display different information.

[0056] In some embodiments, the first section may display the most recently determined probability in the first section 210. While shown here as a specific value of "62%", it will be understood that in some embodiments, a range of values ​​(e.g., ">95%" or "70-80%") may be displayed. In some embodiments, this display may be provided as text or an image representing the first probability. For example, the interface may display text that is a description of the probability (e.g., a "low", "medium", or "high" risk or probability status or rating, such as a rating on a scale of 1-5, or terms that define a range of probabilities, such as "unlikely", "probable", "likely", "highly likely", etc.). These values ​​or textual representations may be color coded, e.g., green for a probability less than 50%, yellow for a probability greater than 50% and less than 75%, and red for a probability greater than 75%. Alternatively, or in addition, the interface may display an image which may be a rating, e.g., a star rating from 1 to 5 stars, a graph showing probability data over time, an icon, e.g., up and down arrows, or a color-coded representation (e.g., red, yellow, or green tiles to indicate a determined probability).

[0057] In some embodiments, the second section 220 may display additional data, which may include one or more graphs 221 and one or more text fields 222, which may include, for example, words or numbers describing some or all of the data received, derived, or determined by the one or more processors. In some embodiments, this may include, for example, one or more graphs showing trends in the data, or bar graphs showing data categorized by pump speed setting. In some embodiments, this may include one or more text fields displaying descriptions and / or values ​​of parameters that may be of interest to the user (here, only "37.8" is displayed).

[0058] In some embodiments, the third section 230 may include one or more text fields, which may include, for example, a text field 231 describing the treatment option, a data entry field 232 that allows the user to enter required values ​​(e.g., CVP values), and / or a field 233 for additional data that allows the user to enter information such as the particular alternative treatment being undertaken. The third section 230 may also include one or more selectable buttons 235 that may indicate to the user that they can read a particular text field, enter data into the field, submit the data entered into the field to another processor, and / or seek assistance.

[0059] In some embodiments, when more information is provided than can appear on a single screen, the entire display may be scrolled up or down for proper viewing, hi some embodiments, the information within each section may be scrolled independently.

[0060] In some embodiments, the system may include a graphical interface for displaying signals received from the hemodynamic support device and / or information determined from the signals. As will be appreciated, the graphical interface may be on one or more devices, including, for example, a patient console, a computer, and / or a mobile device. The graphical user interface may be the same or different on each display.

[0061] As will be appreciated, in some embodiments, the screens may be configurable by the user. For example, in some embodiments, the user may determine which information (e.g., which sections) are visible on the screen (e.g., to avoid having to scroll up and down to see desired information).

[0062] In some embodiments, the one or more trained ML algorithms may be based on historical data, which may include data collected from devices other than hemodynamic support devices, and / or may include data collected from one or more hemodynamic support device models. For example, in some embodiments, they may be trained using data collected from some or all of the Impella® blood pumps offered by Abiomed Inc. In some embodiments, they may be trained using data from only a single hemodynamic support model; for example, data collected only from a particular blood pump, such as the Impella® CP blood pump offered by Abiomed Inc.

[0063] 3A, in some embodiments, the system can include a database 310 that contains all data used for training and validation purposes. Such data can be collected from multiple databases and can include, for example, data such as results from Abiomed Inc.'s Global cVAD study, and data logs from blood pump controllers. The training data set may be fed to an ML algorithm (running on a processor) along with various parameters 321 (e.g., measured or derived parameters including, but not limited to, CVP, RAP, minimum, maximum and / or mean RAP, RVEDP, PAP, mean, systolic and / or diastolic PAP, PAPI, echo-based parameters of right heart function, LVEDP, pump aspiration, pump alarm rate and / or type, cardiac output, LV contractility, LV relaxation, pulse wave velocity, ejection fraction, statistical metrics of parameters included in the data from the hemodynamic support device, and / or systolic, diastolic, mean, median, minimum, maximum, delta, or pulse of parameters included in the data from the hemodynamic support device, or combinations thereof) to produce a trained model 330, which may then be validated and used to generate probabilities and the like of the present disclosure, as understood in the art.

[0064] In some embodiments, rather than a single model being generated from a large database, two or more models may be generated. For example, in one embodiment, as shown in FIG. 3B, a first database 311 is used to train a first model 330, and a second database 312 is used to train a second model 350. In such an embodiment, the first database may include historical data from multiple sources, while the second database may include only data related to the particular model that has been installed in the subject (e.g., if an Impella® CP blood pump is positioned on the subject's heart, the second model is trained using data collected only from the Impella® CP blood pump). Here, the training data set from the first database may be fed into an ML algorithm (running on a processor) along with various parameters 321, resulting in a first trained model 330, which may then be validated and used to generate probabilities, etc. of the present disclosure, as understood in the art. In some embodiments, data from the second dataset may be provided to a second ML algorithm (running on a processor) along with various parameters 341, resulting in a second trained model 350. In some embodiments, the first trained model may be provided to the second ML algorithm along with data from the second dataset and various parameters.

[0065] In some embodiments, these ML models may be configured to output a probability of a state occurring. In some cases, each model may independently determine the probability of a state occurring, and these independent determinations may be combined to determine a final probability. In some embodiments, a third ML algorithm may be trained by providing the trained models 330, 350 to a third ML algorithm (running on a processor), along with various parameters 361, resulting in a third trained model 370.

[0066] In some embodiments, the data may be fed into a system architecture, such as the Cygnus-X system architecture, to train a machine learning algorithm to identify a defined condition, such as right heart failure (RHF). In some embodiments, as shown in FIG. 4, a process 400 may include providing 410 a database, such as data from the National Cardiogenic Shock Initiative (National CSI). The process may include matching 420 data from the database to pump data. The process may also include labeling 430 the data to indicate whether such data is associated with a target condition (e.g., RHF). The process may include identifying 440 key signals associated with the condition, then designing and training 450 a model as understood in the art, then validating 460 the model. In some embodiments, the process may include iterating the process and enriching the process by validating 470 using secondary data sets, such as from a different database or a different study.

[0067] The model may be, for example, a neural network. In some embodiments, the neural network may be a feedforward neural network, a radial basis function (RBF) neural network, a multi-layer perceptron, a convolutional neural network (CNN), or a recurrent neural network (RNN).

[0068] In some embodiments, the system may be configured to receive cycles of gathering data, updating probabilities, and gathering more data. In some cases, this may involve sending and receiving information from a user-controlled device and having the user perform one or more tasks to gather information.

[0069] This flow chart is shown in Figure 5. In some embodiments of the process 500, the hemodynamic support device may first be introduced 510 into a first region of the subject.

[0070] Thereafter, a first cycle 520 is shown in which data may be received from a hemodynamic support device 521 and the one or more processors may determine the probability of a condition existing and / or occurring 522. The one or more processors may then inform or display the determined probability to a user 523, and optionally display other data as disclosed herein.

[0071] As a result of this notification or indication, the user (or an automated device) may then perform a task 550, such as measuring a parameter of the subject, such as measuring CVP.

[0072] Thereafter, a second cycle 540 is shown in which data, e.g., CVP values, and optionally more data from the hemodynamic support device, may be received 541. The one or more processors may then determine updated probabilities of the condition existing and / or occurring based on the available data, e.g., CVP values ​​542, and may then inform or display the updated probabilities, and optionally other data as disclosed herein, to the user 543.

[0073] As a result of this second notification or indication, the user (or the automated device) may then perform another task, such as measuring a parameter of the subject, such as measuring PAP 570.

[0074] Next, a third cycle 560 is shown in which data, e.g., a PAP value, and optionally more data from the hemodynamic support device, may be received 561. The one or more processors may, for example, derive a PAPI value and determine a further updated probability of the condition existing and / or occurring based on the available data, e.g., the PAP and / or PAPI value 562, and may then inform or display to the user the further updated probability, and optionally other data as disclosed herein 563.

[0075] As a result of this third notification, the user may, for example, engage in a treatment pathway 570. This may be a treatment pathway proposed by one or more processors and displayed to the user.

[0076] It will be appreciated that different arrangements may be used. For example, in some embodiments, a user may decide to engage 570 in a processing method after receiving a first or second notification. In some embodiments, multiple cycles may occur in which no user action is required.

[0077] Thus, in some embodiments, the at least one processor may be configured to receive second data after receiving data from the hemodynamic support device, where the second data relates to a third region of the subject's body that is different from the first region and the second region. For example, if the first region (where the hemodynamic support device is located) is the left heart and the second region (where the condition may exist) is the right heart, the third region may be, for example, the vena cava or the pulmonary artery. Thus, the second data may include, for example, a CVP or PAP value. In some embodiments, the medical instrument or device used to collect the second data also sends the data to the at least one processor of the system. In other embodiments, the user may input data into the user's remote device (e.g., the user's phone, tablet, laptop, etc.) to send the data to the at least one processor.

[0078] In some embodiments, the at least one trained ML algorithm can be further configured to determine a second probability of the condition based on the data from the hemodynamic support device and the second data.

[0079] Similarly, in some embodiments, the at least one processor may be configured to receive third data after receiving the second data from the user, the third data relating to a fourth region of the subject's body distinct from the first region, the second region, and the third region. In some embodiments, the at least one trained ML algorithm may be further configured to determine a third probability of the condition based on the data from the hemodynamic support device, the second data, and the third data.

[0080] In some embodiments, the at least one processor may be further configured to derive at least one parameter from at least a portion of the received data, and the at least one trained ML algorithm may be further configured to determine a probability (e.g., a first, second, and / or third probability) based on the at least one parameter. Non-limiting examples of such derived parameters include central venous pressure (CVP), right atrial pressure (RAP), minimum, maximum, and / or mean RAP, right ventricular end-diastolic pressure (RVEDP), pulmonary artery pressure (PAP), mean, systolic, and / or diastolic PAP, pulmonary artery pressure index (PAPI), and / or echo-based parameters of right heart function. In some embodiments, PAPI may be calculated by subtracting diastolic pulmonary artery value (PAdia) from systolic pulmonary artery value (PAsys), and then dividing the difference by RAP or CVP. In some embodiments, the echo-based parameters of right heart function are right ventricular (RV) diameter, RV volume, RV stroke volume index (RVSVI) value, RV stroke work index (RVSWI) value, and / or tricuspid annular systolic excursion (TAPSE) value.

[0081] In some embodiments, the system may include a remote device that may be configured to transmit the second data and the third data to the at least one processor and to receive the first probability, the second probability, and the third probability from the at least one processor. In some embodiments, non-user identifiable data is sent to the remote device.

[0082] 1, in some embodiments, the system may include an additional device (or devices) including a sensor. For example, in some embodiments, the system may include a first device 150 (here, a patch) that includes a sensor 155. In some embodiments, each additional device is used in or on the subject's body in a different region of the body than the other additional devices, and different than the first, second, third, and fourth regions of the body (here, a patch was used on the subject's chest, while the first through fourth regions were in and around the heart).

[0083] In some embodiments, the system may include a single additional device. In some embodiments, the system may include two or more additional devices. As shown in FIG. 1, a second additional device 151 (here a watch or bracelet) is shown that includes a sensor 156.

[0084] In some embodiments, the at least one trained ML algorithm may be further configured to determine a probability of a state (e.g., a first, second, or third probability) based on data received from a sensor of the additional device.

[0085] The additional device and associated sensors may provide data regarding any suitable parameters For example, in some embodiments, data received from the sensors of the additional device may include heart rate, blood oxygenation values, electrocardiogram (ECG) values, skin temperature, acceleration, or combinations thereof.

[0086] In some embodiments, the at least one processor may be further configured to make decisions in an effort to assist in providing a human interpretation to the data. For example, in some embodiments, the at least one processor may be configured to determine whether the probabilities (e.g., the first, second, and / or third probabilities) are above a first threshold and / or below a second threshold.

[0087] In some embodiments, the one or more processors may determine a trend over time of the probability of the subject's condition, and may optionally determine whether the rate of change described by the trend exceeds a threshold rate, and / or if the trend continues, whether the probability will exceed a first threshold or fall below a second threshold within a predetermined period of time. For example, if the first threshold rate is 80%, the second threshold rate is 20%, and the latest probability was 70%, the latest probability determination will not exceed the first threshold or fall below the second threshold, so in this example, no alarm will be triggered. However, if the probability has been trending steadily upward from 45% to 70% for the past 5 minutes (e.g., the trend defines a rate of change of +5% per minute), the probability may be estimated to exceed the 80% threshold in just over 2 minutes. This determination may be useful if the predetermined period is 5 minutes (e.g., if the system determines that the high threshold may be exceeded within 5 minutes, which is not ideal).

[0088] In some embodiments, the one or more processors may be configured to identify one or more major factors that cause the probability to be above a first predetermined threshold and / or below a second predetermined threshold. For example, looking at the trained ML model, it may be clear what weightings may be given to what features. For example, each such feature may be assigned to a different biological component, biological function, and / or parameter (e.g., CVP, PAPI, etc.). With such factors, it may be relatively straightforward to determine what one or more components, functions, and / or parameters have the highest overall impact on the determined probability. In some embodiments, a single component, function, and / or parameter may be identified. In some embodiments, multiple components, functions, and / or parameters may be identified.

[0089] In some embodiments, the one or more processors may determine trends over time in the identified major factors. For example, the one or more processors may be configured to identify at least one factor each time the probability is determined to be above a threshold value. After five minutes, there may be several identified factors, which may be listed, for example, shown in a Pareto chart on the display. As will be appreciated, other displays may be used in other embodiments. In some embodiments, the impact of multiple factors may be tracked over time, even if they are not considered to be "major" factors.

[0090] In some embodiments, some combinations of these may be included.

[0091] In some embodiments, the at least one processor may be configured to generate an alert when the first probability is determined to be above a first predetermined threshold or below a second predetermined threshold, when the rate of change described by the trend exceeds a threshold rate, and / or when it is determined that if the trend continues, the probability will be above the first threshold or below the second threshold within a predetermined time period. In some embodiments, the alert includes a visual signal to the user. In some embodiments, the alert may include an audio or tactile signal to the user. In some embodiments, a text message or email is sent to the user. In some embodiments, the alert may appear on a display of the controller. In some embodiments, the alert may appear on a remote device of the user (e.g., laptop, cell phone, etc.). In some embodiments, the alert may include one or more options for the user to select. In some embodiments, the alert may include text indicating testing and / or treatment for the condition.

[0092] In some embodiments, the at least one processor may be further configured to receive input from a user in response to the alert. In some embodiments, this may include receiving data indicating that a particular treatment has been administered or that an existing treatment has been discontinued. In some embodiments, this may include receiving data indicating an option from which the alert was selected.

[0093] In some embodiments, the at least one processor may be further configured to track the probability of the condition over time to determine whether a treatment administered is effective in reducing the risk of the condition. For example, if input from a user in response to an alert indicates that the user has begun administering a particular treatment, the system may continue to monitor the probability over time and provide feedback to the user indicating whether the treatment is or has been effective in reducing or eliminating the condition.

[0094] In some embodiments, a method for detecting and / or inferring a condition may be provided. Referring to FIG. 6, a method 600 may include receiving 610 data from a hemodynamic support device positioned in a first region of a subject's body. The method may include determining 620, using a trained machine learning algorithm, based on the received data, a first probability of the condition being present in a second region of the subject's body, the second region being different from the first region.

[0095] As disclosed herein, in some embodiments, the method may include sending 615, over a network, data from a first processor (that received the data from the hemodynamic support device) to a second processor configured to determine the probability using a trained machine learning algorithm.

[0096] As disclosed herein, in some embodiments, the method may include sending 628 the first probability over a network to at least one processor (e.g., the first processor or a processor in a different remote device).

[0097] In some embodiments, the method may then include displaying 629 the first probability, or text or an image representing the first probability. Other data, graphs, etc. may also be animated as disclosed herein. For example, this may be displayed on a console controlling the hemodynamic support device, on a mobile device, etc.

[0098] In some embodiments, the method may include providing at least one trained ML algorithm 605. In some embodiments, this may include providing a first ML algorithm trained on historical data collected from multiple medical device product lines. In some embodiments, this may include providing a second ML algorithm trained on data collected from a single medical device product line.

[0099] In some embodiments, the method may include receiving 630 second data, after receiving the data from the hemodynamic support device, e.g., data from a user-controlled device as disclosed herein, the second data relating to a third region of the subject's body different from the first region and the second region. In some embodiments, a user may have entered the data into the device. In some embodiments, the device may have measured and transmitted the data.

[0100] In some embodiments, the method may include determining 640 a second probability of the condition based on the data from the hemodynamic support device and the second data using the at least one trained ML algorithm.

[0101] As disclosed herein, in some embodiments, the method may include sending 648 the second probability over a network to at least one processor (e.g., the first processor or a processor in a different remote device).

[0102] In some embodiments, the method may then include displaying 649 the second probability, or text or an image representing the second probability. Other data, graphs, etc. may also be animated as disclosed herein. This may be displayed, for example, on a console controlling the hemodynamic support device, on a mobile device, etc.

[0103] In some embodiments, the method may include receiving 650 third data, for example from a user-controlled device, after receiving the second data from the user, as disclosed herein, the third data relating to a fourth region of the subject's body different from the first region, the second region, and the third region.

[0104] In some embodiments, the method may include determining 660 a third probability of the condition based on the data from the hemodynamic support device, the second data, and the third data, using the at least one trained ML algorithm.

[0105] As disclosed herein, in some embodiments, the method may include sending 668 the third probability over a network to at least one processor (e.g., the first processor or a processor in a different remote device).

[0106] In some embodiments, the method may then include displaying 669 the third probability, or text or an image representing the third probability. Other data, graphs, etc. may also be animated as disclosed herein. This may be displayed, for example, on a console controlling the hemodynamic support device, on a mobile device, etc.

[0107] In some embodiments, prior to determining the at least one probability, the method may include deriving 625, 645, 665 at least one parameter. As disclosed herein, the at least one parameter may be, for example, central venous pressure (CVP), right atrial pressure (RAP), minimum, maximum, and / or mean RAP, right ventricular end-diastolic pressure (RVEDP), pulmonary artery pressure (PAP), mean, systolic, and / or diastolic PAP, pulmonary artery pressure index (PAPI), and / or an echo-based parameter of right heart function.

[0108] In some embodiments, the method may include receiving data from an additional device (e.g., a wearable device or patch) including a sensor, as disclosed herein, positioned in or on a third region of the patient's body. As disclosed herein, in some embodiments, the at least one trained ML algorithm may further determine the first probability based on data received from a sensor of the additional device. As disclosed herein, the additional device may provide any of a wide range of relevant data. For example, in some embodiments, one or more sensors in or on the additional device may collect data including heart rate, values ​​related to blood oxygen, values ​​related to an electrocardiogram (ECG), skin temperature, acceleration, or a combination thereof.

[0109] In some embodiments, the method may include making additional determinations 626, 646, 666 regarding the determined probability. For example, in some embodiments, this may include determining whether the probability is above a first threshold and / or below a second threshold as disclosed herein. In some embodiments, this may include determining a trend in the probability of the subject's condition over time, and optionally determining whether the rate of change described by the trend is above a threshold rate, and / or if the trend continues, whether the probability will be above the first threshold or below the second threshold within a predetermined time period. In some embodiments, this may include identifying one or more key factors that cause the probability to be above the first predetermined threshold and / or below the second predetermined threshold. In some embodiments, this may include determining a trend over time in the identified key factors. In some embodiments, this may include combinations of these.

[0110] In some embodiments, the method may include alerting 670 the user when the first probability is determined to be above a first predetermined threshold or below a second predetermined threshold, when the rate of change explained by the trend exceeds a threshold rate, and / or when it is determined that if the trend continues, the probability will be above the first threshold or below the second threshold within a predetermined time period. In some embodiments, the alert may include options for the user to select, for example, including testing and / or treatment for the condition.

[0111] In some embodiments, the method may include receiving input from a user responsive to the alert 675. In some embodiments, the input responsive to the alert may include data indicating that a particular treatment has been administered or that an existing treatment has been discontinued.

[0112] In some embodiments, the method may include tracking 680 the probability of the condition over time. In some embodiments, the method may include determining 685 whether a treatment administered is effective in reducing the risk of the condition (e.g., based on the tracked probability).

[0113] Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the disclosure described herein which equivalents are intended to be encompassed by the following claims.

Claims

1. 1. A system for detecting and / or inferring a condition, comprising: a hemodynamic support device configured to be positioned in a first region of a subject's body; at least one processor operably coupled to the hemodynamic support device, receiving data from the hemodynamic support device; and Using at least one trained machine learning (ML) algorithm, determine a first probability of a condition being present in a second region of the subject's body that is different from the first region based on the received data. at least one processor configured to Including, the system.

2. The at least one processor a first processor, receiving data from the hemodynamic support device; and Sending the data to a second processor over a network a first processor configured to: a second processor, receiving data from the first processor; and Using the trained machine learning algorithm, determine the first probability of the condition being present in the second region of the subject's body based on the received data. a second processor configured to The system of claim 1 , comprising:

3. The system of claim 2 , wherein the second processor is further configured to send the first probability to the first processor.

4. further comprising a remote device; the second processor is further configured to send the first probability to the remote device; and The system of claim 2 , wherein the remote device is configured to display the first probability or text or an image representing the first probability.

5. 10. The system of claim 1, wherein the at least one trained ML algorithm comprises a first ML algorithm trained with historical data collected from multiple medical device product lines.

6. 6. The system of claim 5, wherein the at least one trained ML algorithm further comprises a second ML algorithm trained with data collected from a single medical device product line.

7. the at least one processor is configured to receive second data after receiving the data from the hemodynamic support device, the second data relating to a third region of the subject's body, different from the first region and the second region; and 7. The system of claim 6, wherein the at least one trained ML algorithm is further configured to determine a second probability of the condition based on the data from the hemodynamic support device and the second data.

8. The system of claim 7 , wherein the second data includes a value related to central venous pressure (CVP).

9. the at least one processor is configured to receive third data after receiving the second data from a user, the third data relating to a fourth region of the subject's body, different from the first region, the second region, and the third region; and 8. The system of claim 7, wherein the at least one trained ML algorithm is further configured to determine a third probability of the condition based on the data from a hemodynamic support device, second data, and third data.

10. the third data includes a value related to pulmonary artery pulsatility (PAP); the at least one processor is further configured to derive at least one parameter, and the at least one trained ML algorithm is further configured to determine the second probability and / or the third probability based on the at least one parameter, wherein the at least one parameter is central venous pressure (CVP), right atrial pressure (RAP), minimum, maximum, and / or mean RAP, right ventricular end-diastolic pressure (RVEDP), pulmonary artery pressure (PAP), mean, systolic, and / or diastolic PAP, pulmonary artery pressure index (PAPI), and / or an echo-based parameter of right heart function; PAPI is calculated by subtracting the diastolic pulmonary artery value (PAdia) from the systolic pulmonary artery value (PAsys), then dividing the difference by the RAP or CVP, 10. The system of claim 9, wherein the echo-based parameters of right heart function are right ventricular (RV) diameter, RV volume, RV stroke volume index (RVSVI) value, RV stroke work index (RVSWI) value, and / or tricuspid annular systolic excursion (TAPSE) value.

11. 10. The system of claim 9, further comprising a remote device configured to transmit the second data and the third data to the at least one processor and to receive the first probability, the second probability, and the third probability from the at least one processor.

12. The system of claim 11 , wherein non-user identifiable data is sent to or from the remote device.

13. 10. The system of claim 1, further comprising an additional device operably coupled to the at least one processor, the additional device including a sensor, the sensor positioned in or on a third region of the patient's body.

14. 14. The system of claim 13, wherein the at least one trained ML algorithm is further configured to determine the first probability based on data received from the sensor of the additional device.

15. The system of claim 14 , wherein the data received from the sensor of the additional device includes a heart rate, a blood oxygen value, an electrocardiogram (ECG) value, skin temperature, or acceleration.

16. 10. The system of claim 1, wherein the first region of the subject's body is the left heart and the second region of the subject's body is the right heart.

17. 17. The system of claim 16, wherein the data from the hemodynamic support device includes first information regarding left heart systolic function and second information regarding suction or pumping flow within the left heart.

18. 18. The system of claim 17, wherein the first information comprises left ventricular (LV) contractility, aortic (AO) pulse pressure and / or pulsatility, or a combination thereof.

19. 2. The system of claim 1, wherein the data from the hemodynamic support device includes aortic (AO) pressure, left ventricular (LV) pressure, pump motor speed, pump motor current, LV-AO pressure gradient, pump flow, cardiac output, intrinsic cardiac output, LV rate, AO pulse rate, or a combination thereof.

20. 20. The system of claim 19, wherein the data from the hemodynamic support device also includes LV volume by conductance, heart rate, heart rhythm, arterial pressure, blood oxygenation, or a combination thereof.

21. 21. The system of claim 20, wherein the at least one processor is further configured to derive at least one parameter, and the at least one trained ML algorithm is configured to determine the first probability based on the data from the hemodynamic support device and the at least one parameter, wherein the at least one parameter is LV end-diastolic pressure (LVEDP), pump aspiration, pump alarm rate and / or type, cardiac output, LV contractility, LV relaxation, pulse wave velocity, ejection fraction, a statistical metric of a parameter included in the data from the hemodynamic support device, and / or a systolic value, a diastolic value, a mean, a median, a minimum, a maximum, a delta, or a pulse of a parameter included in the data from the hemodynamic support device.

22. The at least one processor further comprises: determining whether the first probability is above a first threshold and / or below a second threshold; determining a trend over time in the subject's probability of the condition, and optionally whether the rate of change explained by the trend exceeds a threshold rate, and / or if the trend continues, whether the probability will exceed the first threshold or fall below the second threshold within a predetermined time period; identifying one or more primary factors that cause the probability to be above the first threshold and / or below the second threshold; Determine trends over time for identified key drivers; or Do a combination of them The system of claim 1 configured to:

23. 23. The system of claim 22, wherein the at least one processor is further configured to alert a user when the first probability is determined to be above the first threshold or below the second threshold, when the rate of change explained by the trend exceeds the threshold rate, and / or when it is determined that the probability will be above the first threshold or below the second threshold within the predetermined time period if the trend continues.

24. 24. The system of claim 23, wherein the at least one processor is further configured to receive input from the user in response to the alert.

25. 10. The system of claim 1, wherein the at least one processor is further configured to track the probability of the condition over time to determine whether a treatment administered is effective in reducing the risk of the condition.

26. 1. A method for detecting and / or inferring a condition, comprising: receiving data from a hemodynamic support device positioned in a first region of the subject's body; using a trained machine learning algorithm to determine, based on the received data, a first probability of a condition being present in a second region of the subject's body that is different from the first region; A method comprising:

Images