Patent Application
20220287579
The present technology is generally related to a system and method for continuous non-invasive blood pressure (CNIBP) measurement, for example using deep learning artificial intelligence (AI) utilizing differential pulse transit time (DPTT).
In the field of medicine, doctors often desire to monitor certain physiological characteristics of their patients. Accordingly, a wide variety of devices have been developed for monitoring many such physiological characteristics. Such devices provide doctors and other healthcare personnel with the information they need to provide the best possible healthcare for their patients. As a result, such monitoring devices have become an indispensable part of modern medicine.
One technique for monitoring certain physiological characteristics of a patient uses attenuation of light to determine physiological characteristics of a patient. This is used in pulse oximetry, and the devices built are based upon pulse oximetry techniques. Light attenuation is also used for regional or cerebral oximetry. Oximetry may be used to measure various blood characteristics, such as the oxygen saturation of hemoglobin in blood or tissue, the volume of individual blood pulsations supplying the tissue, and/or the rate of blood pulsations corresponding to each heartbeat of a patient. The signals can lead to further physiological measurements, such as respiration rate, glucose levels or blood pressure.
Many conventional medical monitors require attachment of a sensor to a patient in order to detect physiologic signals from the patient and to transmit detected signals through a cable to the monitor. These monitors process the received signals and determine vital signs such as the patient's pulse rate, respiration rate, and arterial oxygen saturation. For example, a pulse oximeter is a finger sensor that can include two light emitters and a photodetector. The sensor emits light into the patient's finger and transmits the detected light signal to a monitor. The monitor includes a processor that processes the signal, determines vital signs (e.g., pulse rate, respiration rate, arterial oxygen saturation), and displays the vital signs on a display.
Other monitoring systems include other types of monitors and sensors, such as electroencephalogram (EEG) sensors, blood pressure cuffs, temperature probes, air flow measurement devices (e.g., spirometer), and others. Some wireless, wearable sensors have been developed, such as wireless EEG patches and wireless pulse oximetry sensors.
Determination of blood pressure non-invasively and continuously presents a significant technical challenge in the medical device industry. For that reason, blood pressure is typically measured intermittently via a separate blood pressure cuff or continuously using invasive techniques, for example using of an invasive arterial line, with the various monitoring devices being connected to one or more patient monitors to present patient measurements.
What is needed in the art are systems and methods allowing for continuous, non-invasive blood pressure measurement.
The techniques of this disclosure generally relate to systems and methods for continuous non-invasive blood pressure (CNIBP) measurement. In exemplary aspects described herein, CNIBP is measured using deep learning artificial intelligence (AI) utilizing differential pulse transit time (DPTT).
In one aspect, a patient monitoring system includes a first PPG sensor in contact with a patient at a first location, the first sensor providing first data over a first time period related to the patient to determine one or more patient parameters; a second PPG sensor in contact with a patient at a second patient location different from the first location, the second sensor providing second data over said first time period related to the patient to determine one or more patient parameters; and a processor configured to: compare at least a portion of the first data and at least a portion of the second data in order to calculate a differential pulse transit time (DPTT) between the first and second locations; and determine continuous non-invasive blood pressure (CNIBP) using the first data, the second data and the DPTT.
In another aspect, a method for patient monitoring includes configuring a first PPG sensor to contact a patient at a first location, the first sensor configured to provide first data over a first time period related to the patient to determine one or more patient parameters; configuring a second PPG sensor to contact a patient at a second patient location different from the first location, the second sensor providing second data over said first time period related to the patient to determine one or more patient parameters; and with a processor: comparing at least a portion of the first data and at least a portion of the second data in order to calculate a differential pulse transit time (DPTT) between the first and second locations; and determining continuous non-invasive blood pressure (CNIBP) using the first data, the second data and the DPTT.
In an exemplary aspect, the processor compares at least one fiducial point in the first data and in the second data in order to calculate DPTT. Exemplary fiducial points include: a peak of the pulse, the trough of the pulse; or the location of maximum upslope gradient.
In another exemplary aspect, at least a portion of the first sensor data, at least a portion of the second sensor data, and the calculated DPTT are input into a deep learning AI model to determine CNIBP. Exemplary deep learning AI models include an LSTM model, a CNN model, and a hybrid CNN-LSTM model.
The details of one or more aspects of the disclosure are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the techniques described in this disclosure will be apparent from the description and drawings, and from the claims.
Many aspects of the present disclosure can be better understood with reference to the following drawings. The components in the drawings are not necessarily to scale. Instead, emphasis is placed on illustrating clearly the principles of the present disclosure. The drawings should not be taken to limit the disclosure to the specific embodiments depicted, but are for explanation and understanding only.
The following disclosure describes systems and methods for continuous non-invasive blood pressure (CNIBP) measurement. In exemplary aspects described herein, CNIBP is measured using deep learning artificial intelligence (AI) utilizing differential pulse transit time (DPTT).
In exemplary aspects, devices, systems, and/or methods configured in accordance with embodiments of the present technology can include one or more sensors or probes associated with (e.g., contacting) a patient that can be configured to capture data (e.g., temperature, blood pressure, heart rate, arterial oxygen saturation, etc.) related to a patient. The devices, systems, and/or methods can transmit the captured data to a monitoring device, hub, mobile patient management system (MPM), or the like. In some embodiments, the devices, systems, and/or methods can analyze the captured data to determine and/or monitor one or more patient parameters. In these and still other embodiments, the devices, systems, and/or methods can trigger alerts and/or alarms when the devices, systems, and/or methods detect one or more patient parameter abnormalities.
In some embodiments, one or more sensors or probes associated with (e.g., contacting) a patient can be configured to capture data related to a patient, e.g. as a PPG signal. A signal representing light intensity versus time or a mathematical manipulation of this signal (e.g., a scaled version thereof, a log taken thereof, a scaled version of a log taken thereof, etc.) may be referred to as the photoplethysmograph (PPG) signal. In addition, the term “PPG signal,” as used herein, may also refer to an absorption signal (i.e., representing the amount of light absorbed by the tissue) or any suitable mathematical manipulation thereof. The amount of light detected or absorbed may then be used to calculate any of a number of physiological parameters, including oxygen saturation (the saturation of oxygen in pulsatile blood, SpO2), an amount of a blood constituent (e.g., oxyhemoglobin), as well as a physiological rate (e.g., pulse rate or respiration rate) and when each individual pulse or breath occurs. For SpO2, red and infrared (IR) wavelengths may be used because it has been observed that highly oxygenated blood will absorb relatively less Red light and more IR light than blood with a lower oxygen saturation. By comparing the intensities of two wavelengths at different points in the pulse cycle, it is possible to estimate the blood oxygen saturation of hemoglobin in arterial blood, such as from empirical data that may be indexed by values of a ratio, a lookup table, and/or from curve fitting and/or other interpolative techniques.
In exemplary aspects described herein, CNIBP is measured using deep learning artificial intelligence (AI) utilizing differential pulse transit time (DPTT) utilizing plural PPG signals obtained via separate locations of a patient. The DPTT may be defined as the difference in time that a pulse wave takes to arrive at two distinct arterial locations. The DPTT derived from the PPG signals acquired from pulse oximeter probes placed at such distinct locations may be used as an input in an AI model to determine CNIBP.
Specific details of several embodiments of the present technology are described herein with reference to
The sensor 14 also includes a sensor body 46 to house or carry the components of the sensor 14. The body 46 includes a backing, or liner, provided around the emitter 16 and the detector 18, as well as an adhesive layer (not shown) on the patient side. The sensor 14 may be reusable (such as a durable plastic clip sensor), disposable (such as an adhesive sensor including a bandage/liner), or partially reusable and partially disposable.
In the embodiments shown, the sensors 14 is communicatively coupled to the patient monitor 12. In certain embodiments, the sensors may include a wireless module configured to establish a wireless communication 15 with the patient monitor 12 using any suitable wireless standard. For example, the sensors may include a transceiver that enables wireless signals to be transmitted to and received from an external device (e.g., the patient monitor 12, a charging device, etc.). The transceiver may establish wireless communication 15 with a transceiver of the patient monitor 12 using any suitable protocol. For example, the transceiver may be configured to transmit signals using one or more of the ZigBee standard, 802.15.4x standards WirelessHART standard, Bluetooth standard, IEEE 802.11x standards, or MiWi standard. Additionally, the transceiver may transmit a raw digitized detector signal, a processed digitized detector signal, and/or a calculated physiological parameter, as well as any data that may be stored in the sensor, such as data relating to wavelengths of the emitters 16, or data relating to input specification for the emitters 16. Additionally, or alternatively, the emitters 16 and detectors 18 of the sensor 14 may be coupled to the patient monitor 12 via a cable 24 through a plug 26 (e.g., a connector having one or more conductors) coupled to a sensor port 29 of the monitor. In certain embodiments, the sensor 14 is configured to operate in both a wireless mode and a wired mode. Accordingly, in certain embodiments, the cable 24 is removably attached to the sensor 14 such that the sensor 14 can be detached from the cable to increase the patient's range of motion while wearing the sensor 14. It should be recognized that wired or wireless configurations, as with sensor 14, are also contemplated with regard to sensor 13, as well as optional blood pressure cuff 11, which are shown in
The patient monitor 12 is configured to calculate physiological parameters of the patient relating to the physiological signal received from the sensors 13, 14. For example, the patient monitor 12 may include a processor configured to calculate the patient's arterial blood oxygen saturation, tissue oxygen saturation, pulse rate, respiration rate, blood pressure, blood pressure characteristic measure, autoregulation status, brain activity, and/or any other suitable physiological characteristics. Additionally, the patient monitor 12 may include a monitor display 30 configured to display information regarding the physiological parameters, information about the system (e.g., instructions for disinfecting and/or charging the sensor 14), and/or alarm indications. The patient monitor 12 may include various input components 32, such as knobs, switches, keys and keypads, buttons, etc., to provide for operation and configuration of the patient monitor 12. The patient monitor 12 may also display information related to alarms, monitor settings, and/or signal quality via one or more indicator lights and/or one or more speakers or audible indicators. The patient monitor 12 may also include an upgrade slot 28, in which additional modules can be inserted so that the patient monitor 12 can measure and display additional physiological parameters.
Because the sensors 13, 14 may be configured to operate in a wireless mode and, in certain embodiments, may not receive power from the patient monitor 12 while operating in the wireless mode, the sensors 13, 14 may include a battery to provide power to the components of the sensor (e.g., the emitter(s) 16 and the detector(s) 18). In certain embodiments, the battery may be a rechargeable battery such as, for example, a lithium ion, lithium polymer, nickel-metal hydride, or nickel-cadmium battery. However, any suitable power source may be utilized, such as, one or more capacitors and/or an energy harvesting power supply (e.g., a motion generated energy harvesting device, thermoelectric generated energy harvesting device, or similar devices).
As noted above, in an embodiment, the patient monitor 12 is a pulse oximetry monitor and the sensor 14 is a pulse oximetry sensor. The sensor 14 may be placed at a site on a patient with pulsatile arterial flow, typically a fingertip, toe, forehead or earlobe, or in the case of a neonate, across a foot. Additional suitable sensor locations include, without limitation, the neck to monitor carotid artery pulsatile flow, the wrist to monitor radial artery pulsatile flow, the inside of a patient's thigh to monitor femoral artery pulsatile flow, the ankle to monitor tibial artery pulsatile flow, and around or in front of the ear. As shown in
As we have noted, exemplary systems and methods described herein determine continuous non-invasive blood pressure (CNIBP) using an AI model, where the inputs include, among other possible inputs, differential pulse transit time (DPTT). Referring to
However,
In exemplary embodiments, and with further reference to
In further exemplary embodiments, systems and methods described herein make use of both wavelengths from two or more PPG signals in a deep learning model. For example,
While
In other exemplary embodiments, inputs may be provided corresponding to features derived from the raw PPG signals. For example, these may include characteristic features from each PPG waveform including: pulse duration, relative position of maximum upslope of the systolic rise, peak location and amplitude, perfusion index, baseline trend, respiratory cycle information, area of upstroke, downstroke, max gradient of upslope, baseline value, etc. Additionally, for each feature a sequence of values over time may be used. These may be once per period of time (i.e. once per second, or once per pulse, or a single value from a time window of longer period (e.g. 15 seconds, 30 seconds, etc.) In addition to these features, the DPTT calculated between each signal may be provided as an input. Thus, a matrix of feature values from each signal and the corresponding DPTT may be constructed, as shown generally at 500 in
Table 1, below outlines further exemplary features, for example relative to a finger PPG and the DPTT between the finger and a forehead:
In exemplary embodiments described herein, the AI model is trained to calculate a blood pressure signal from the provided inputs. This may be any characteristic blood pressure including, for example, the systolic pressure (SP), diastolic pressure (DP), mean arterial pressure (MAP) or pulse pressure (PP).
In further exemplary embodiments, feature matrices are input into the training cycle of a deep learning model with a target BP value associated with it. The model is then trained to associate the PPG morphological feature sequences with the blood pressure values. The model may then be tested using a test set of feature sequences previously unseen by the model to estimate the associated blood pressure. In this way a model may be generated with a given performance in terms of associating the PPG-based input to a BP.
In exemplary embodiments described herein, the deep learning model is a long short-term memory (LSTM) machine learning model, with an exemplary architecture illustrated generally at 600 in
In further exemplary embodiments, the deep learning model is a convolutional neural network (CNN) model, an exemplary architecture of which is shown generally at 800 in
In additional exemplary embodiments, the deep learning model is a hybrid CNN-LSTM model, as is shown generally at 1000 in the flowchart of
In additional exemplary embodiments, the CNIBP system and method may be at least periodically calibrated, e.g., to account for possible loss of accuracy over time, e.g., due to confounders affecting the model being used. In this case, the CNIBP system may be intermittently calibrated using a blood pressure cuff, such as cuff 11 in
The above detailed descriptions of embodiments of the technology are not intended to be exhaustive or to limit the technology to the precise form disclosed above. Although specific embodiments of, and examples for, the technology are described above for illustrative purposes, various equivalent modifications are possible within the scope of the technology, as those skilled in the relevant art will recognize. For example, while steps are presented in a given order, alternative embodiments can perform steps in a different order. Furthermore, the various embodiments described herein can also be combined to provide further embodiments.
The systems and methods described herein can be provided in the form of tangible and non-transitory machine-readable medium or media (such as a hard disk drive, hardware memory, etc.) having instructions recorded thereon for execution by a processor or computer. The set of instructions can include various commands that instruct the computer or processor to perform specific operations such as the methods and processes of the various embodiments described here. The set of instructions can be in the form of a software program or application. The computer storage media can include volatile and non-volatile media, and removable and non-removable media, for storage of information such as computer-readable instructions, data structures, program modules or other data. The computer storage media can include, but are not limited to, RAM, ROM, EPROM, EEPROM, flash memory or other solid-state memory technology, CD-ROM, DVD, or other optical storage, magnetic disk storage, or any other hardware medium which can be used to store desired information and that can be accessed by components of the system. Components of the system can communicate with each other via wired or wireless communication. The components can be separate from each other, or various combinations of components can be integrated together into a monitor or processor or contained within a workstation with standard computer hardware (for example, processors, circuitry, logic circuits, memory, and the like). The system can include processing devices such as microprocessors, microcontrollers, integrated circuits, control units, storage media, and other hardware.
From the foregoing, it will be appreciated that specific embodiments of the technology have been described herein for purposes of illustration, but well-known structures and functions have not been shown or described in detail to avoid unnecessarily obscuring the description of the embodiments of the technology. To the extent any materials incorporated herein by reference conflict with the present disclosure, the present disclosure controls. Where the context permits, singular or plural terms can also include the plural or singular term, respectively. Moreover, unless the word “or” is expressly limited to mean only a single item exclusive from the other items in reference to a list of two or more items, then the use of “or” in such a list is to be interpreted as including (a) any single item in the list, (b) all of the items in the list, or (c) any combination of the items in the list. Where the context permits, singular or plural terms can also include the plural or singular term, respectively. Additionally, the terms “comprising,” “including,” “having” and “with” are used throughout to mean including at least the recited feature(s) such that any greater number of the same feature and/or additional types of other features are not precluded.
From the foregoing, it will also be appreciated that various modifications can be made without deviating from the technology. For example, various components of the technology can be further divided into subcomponents, or various components and functions of the technology can be combined and/or integrated. Furthermore, although advantages associated with certain embodiments of the technology have been described in the context of those embodiments, other embodiments can also exhibit such advantages, and not all embodiments need necessarily exhibit such advantages to fall within the scope of the technology. Accordingly, the disclosure and associated technology can encompass other embodiments not expressly shown or described herein.
