WO2025128655A1

WO2025128655A1 - System and method for detection of pulselessness using a wearable computing device

Info

Publication number: WO2025128655A1
Application number: PCT/US2024/059500
Authority: WO
Inventors: Jacob Eugene SUNSHINE; Anran WANG; Kamal Girish SHAH; Tajinder Singh Gadh; Yiwen Chen; Enxun WEI; Pramod RUDRAPATNA; Edward SHI
Original assignee: Google LLC
Current assignee: Google LLC
Priority date: 2023-12-12
Filing date: 2024-12-11
Publication date: 2025-06-19
Anticipated expiration: 2026-06-12

Abstract

A computing system, a computer implemented method, and a wearable computing device to detect pulselessness in a wearer of the wearable computing device while minimizing the risk of false positives is provided. A series of passive, semi-active, and/or active algorithms utilizing time domain analysis, gate-based analysis, and/or a machine-learned model can be implemented to analyze motion and photoplethysmogram sensor data to verify if a pulselessness event associated with the wearer has occurred and to alert a third party of such an event.

Description

SYSTEM AND METHOD FOR DETECTION OF PULSELESSNESS USING A WEARABLE COMPUTING DEVICE

RELATED APPLICATIONS

[0001] The present disclosure claims priority to U.S. Provisional Application No.: 63/609,146 filed on December 12, 2023, which is incorporated herein by reference in its entirety.

FIELD OF THE INVENTION

[0002] The present disclosure relates generally to wearable computing devices, and more particularly, to systems and methods for detection of pulselessness of a wearer using a wearable computing device.

BACKGROUND

[0003] Recent advances in technology, including those available through consumer devices, have provided for corresponding advances in health detection and monitoring. For instance, certain wearable devices include a variety of sensors for measuring multiple biological parameters and collecting data that can be beneficial to a user of the device, such as a heart rate sensor in the form of a photoplethysmogram (PPG) sensor, motion sensors (e.g., a gyroscope, an altimeter, an accelerometer, etc.), multi-purpose electrical sensors compatible with electrocardiogram (ECG) and electrodermal activity (EDA) applications, infrared sensors, a temperature sensor, an ambient light sensor, Wi-Fi, GPS, a vibration sensor, a speaker, and a microphone, among others. For example, wearable computing devices such as fitness trackers and smart watches are able to determine information relating to the pulse or motion of a person wearing the device. Such information can be used to determine if the wearer is experiencing a serious cardiac event, and this information can then be used to notify an emergency contact, first responders, and the like. Due to capabilities of conventional devices, it may be difficult to determine if data is associated with pulselessness or the device has been loosened or taken off (doffed) by the wearer. This can increase the risk of false positives, which can, in turn, expend emergency resources at times when such resources are not needed.

[0004] Accordingly, improved systems and methods for the accurate detection and notification of pulselessness of a wearer by one or more sensors in a wearable computing device would be welcomed in the art, with an emphasis on systems and methods that reduce the occurrence of false positives.

SUMMARY OF THE INVENTION

[0005] Aspects and advantages of embodiments of the present disclosure will be set forth in part in the following description, or can be learned from the description, or can be learned through practice of the embodiments.

[0006] In one aspect, a computing system to detect pulselessness in a wearer of a wearable computing device is provided. The computing system includes one or more processors; and one or more non-transitory computer-readable media that collectively store instructions that, when executed by the one or more processors, cause the computing system to perform operations, the operations comprising, as part of a first algorithm: (i) receiving, via the one or more processors, signal data from a motion sensor and a photoplethysmogram sensor associated with the wearable computing device; (ii) performing, via the one or more processors, time domain analysis on a green light signal received from the photoplethysmogram sensor to quantify a decrease in an alternating current component of the green light signal, wherein if the decrease in the alternating current component of the green light signal is above a predetermined percentage over a predetermined time period, pulselessness is suspected, and wherein if the decrease in the alternating current component of the green light signal is less than or equal to the predetermined percentage over the predetermined time period, pulselessness is not suspected; (iii) if pulselessness is suspected in step (ii), performing, via the one or more processors, gate-based analysis on an acceleration signal received from the motion sensor to quantify an acceleration change over a predetermined time period, wherein if the acceleration change is less than a predetermined value, pulselessness is suspected, and wherein if the acceleration change over the predetermined time period is greater than or equal to the predetermined value, then pulselessness is not suspected; and (iv) if pulselessness is suspected in step (ii) and/or step (iii), implementing, via the one or more processors, a machine-learned model trained using a dataset of signal data from a prior predetermined amount of time to determine if the signal data received from the motion sensor, the photoplethysmogram sensor, or both is indicative of pulselessness.

[0007] In one aspect, the operations can further include, as part of the first algorithm: (iii)(a) if pulselessness is suspected in step (iii) and prior to performing step (iv), performing, via the one or more processors, gate-based analysis on an ambient light signal received from the photoplethysmogram sensor to quantify a change in a peak to peak amplitude of the ambient light signal, wherein if the change in the peak to peak amplitude of the ambient light signal is less than a predetermined level over a predetermined time period, pulselessness is suspected, and wherein if the change in the peak to peak amplitude of the ambient light signal is greater than or equal to a predetermined level over a predetermined time period, pulselessness is not suspected.

[0008] In another aspect, the operations can further include, as part of the first algorithm: (v) evaluating, via the one or more processors, the time domain analysis and the machine-learned model determinations from the signal data to predict whether or not pulselessness is suspected, and if pulselessness is suspected, to initiate further investigation by the computing system.

[0009] In yet another aspect, the first algorithm can be passive.

[0010] In still another aspect, if pulselessness is suspected, the operations can further include, as part of a second algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; and estimating, via the one or more processors, a signal-to-noise ratio of a green light signal received from the photoplethysmogram sensor at one or more predetermined frequencies, wherein if the signal- to-noise ratio is less than or equal to a predetermined value at one or more of the predetermined frequencies, pulselessness is suspected, and wherein if the signal-to-noise ratio is greater than the predetermined value at one or more of the predetermined frequencies, pulselessness is not suspected.

[0011] In one more aspect, if pulselessness is suspected, the operations can further include, as part of a second algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; and estimating, via the one or more processors, a signal-to-noise ratio of an infrared light signal received from the photoplethysmogram sensor at one or more predetermined frequencies, wherein if the signal- to-noise ratio is less than or equal to a predetermined value at one or more of the predetermined frequencies, pulselessness is suspected, and wherein if the signal-to-noise ratio is greater than the predetermined value at one or more of the predetermined frequencies, pulselessness is not suspected.

[0012] In an additional aspect, if pulselessness is suspected, the operations can further include, as part of a second algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; and performing, via the one or more processors, gate-based analysis on an ambient light signal received from the photoplethysmogram sensor to quantify a change in a peak to peak amplitude of the ambient light signal, wherein if the change in the peak to peak amplitude of the ambient light signal is less than a predetermined level over a predetermined time period, pulselessness is suspected, and wherein if the change in the peak to peak amplitude of the ambient light signal is greater than or equal to a predetermined level over a predetermined time period, pulselessness is not suspected.

[0013] In one more aspect, if pulselessness is suspected, the operations can further include, as part of a second algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; and performing, via the one or more processors, gate-based analysis on an acceleration signal received from the motion sensor to quantify an acceleration change over a predetermined time period, wherein if the acceleration change is less than a predetermined value, pulselessness is suspected, and wherein if the acceleration change over the predetermined time period is greater than or equal to the predetermined value, then pulselessness is not suspected.

[0014] In another aspect, if pulselessness is suspected, the operations can further include, as part of a third algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; generating, via the one or more processors, a first visual notification to provide an alert the wearer of the wearable computing device by controlling, via the one or more processors, a display screen of the wearable computing device to display the first visual notification; delivering, via the one or more processors, a haptic stimulus through the wearable computing device; and analyzing, via the one or more processors, the signal data received from the motion sensor and the photoplethysmogram sensor to determine if the signal data received during delivery of the haptic stimulus is indicative of motion and/or pulsatility, wherein if the signal data is not indicative of motion and/or pulsatility, then pulselessness is suspected, and wherein if the signal data is indicative of motion and/or pulsatility, then pulselessness is not suspected.

[0015] In still another aspect, if pulselessness is suspected, the operations can further include, as part of a fourth algorithm: generating, via the one or more processors, a second visual notification for a predetermined period of time to alert the wearer of the wearable computing device by controlling, via the one or more processors, the display screen of the wearable computing device to display the second visual notification; and if the wearer has not responded to the second visual notification after the predetermined period of time has passed, initiating an emergency alert communication, via the one or more processors, with a third party. [0016] In one more aspect, a computer-implemented method for detection of pulselessness of a wearer of a wearable computing device is provided that is capable of carrying out the functionality described above with respect to the computing system.

[0017] In still another aspect, a wearable computing device is provided that is capable of carrying out the functionality described above with respect to the computing system.

[0018] These and other features, aspects, and advantages of various embodiments of the present disclosure will become better understood with reference to the following description and appended claims. The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate example embodiments of the present disclosure and, together with the description, serve to explain the related principles.

BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Detailed discussion of embodiments directed to one of ordinary skill in the art is set forth in the specification, which makes reference to the appended figures, in which: [0020] FIG. 1 provides a perspective view of a wearable computing device on a wrist of a user according to one embodiment of the present disclosure;

[0021] FIG. 2 provides a front perspective view of a wearable computing device according to one embodiment of the present disclosure;

[0022] FIG. 3 provides a rear perspective view of the wearable computing device of FIG. 2;

[0023] FIG. 4 illustrates various components of an example system that can be utilized according to one embodiment of the present disclosure;

[0024] FIG. 5 provides a schematic diagram of an example set of devices that are able to communicate according to one embodiment of the present disclosure;

[0025] FIG. 6A depicts a flow chart diagram of a first algorithm contemplated by the present disclosure;

[0026] FIG. 6B depicts a block diagram of an example computing system that performs a machine-learned portion of the first algorithm of FIG. 6 A according to example embodiments of the present disclosure;

[0027] FIG. 6C depicts a block diagram of an example computing device that performs a machine-learned portion of the first algorithm of FIG. 6 A according to example embodiments of the present disclosure;

[0028] FIG. 6D depicts a block diagram of an example computing device that performs a machine-learned portion of the first algorithm of FIG. 6 A according to example embodiments of the present disclosure; [0029] FIG. 7 depicts a flow chart diagram of a second algorithm contemplated by the present disclosure;

[0030] FIG. 8 depicts a flow chart diagram of a third algorithm contemplated by the present disclosure; and

[0031] FIG. 9 depicts a flow chart diagram of a fourth algorithm contemplated by the present disclosure.

[0032] Reference numerals that are repeated across plural figures are intended to identify the same features in various implementations.

DETAILED DESCRIPTION

[0033] Reference now will be made in detail to embodiments of the invention, one or more examples of which are illustrated in the drawings. Each example is provided by way of explanation of the invention, not limitation of the invention. In fact, it will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the scope or spirit of the invention. For instance, features illustrated or described as part of one embodiment can be used with another embodiment to yield a still further embodiment. Thus, it is intended that the present invention covers such modifications and variations as come within the scope of the appended claims and their equivalents.

Overview of Wearable Computing Device for Detecting Pulselessness [0034] In general, when a period of pulselessness of a wearer is detected by sensors contained on or within a wearable computing device, it is critical to alert an emergency contact, first responders, and the like that the wearer is in need of immediate medical attention. However, there are situations when data collected by the wearable computing device may inaccurately detect pulselessness when, for instance, the wearer has doffed the wearable device or the wearable device is being improperly worn so that the sensors that typically detect a wearer’s heartbeat do not have sufficient contact with the wearer to detect the wearer’s heartbeat. For instance, when a wearable computing device is removed or worn loosely, there is no skin contact or inconsistent skin contact or shifting area of skin contact, which may cause inaccurate data and a false positive pulselessness reading. To avoid such false positives and to accurately detect pulselessness when it occurs, the present disclosure utilizes one or more algorithms, such as a series of passive, semi-passive, and active algorithms in a sequential process to accurately detect pulselessness using the sensors present on the wearable computing device. It should be understood that passive algorithms consume data as-is from one or more sensors of the wearable computing device, without changing the sensor configuration, and typically analyze sensor data retrospectively or in real time (as the data are received); semi-passive algorithms consume data as-is from one or more sensors, but may analyze data prospectively (thereby introducing a delay on the order of seconds or minutes before a semi-passive algorithm's output is reported); and active algorithms force one or more sensors into specific configurations before processing the sensors' data. If the one or more algorithms (e.g., one or more passive, semi-passive, and/or active algorithms) detects pulselessness, the wearable computing device and methods contemplated herein can initiate a check-in with the wearer, and if the wearer does not respond within a first predetermined time period, a third party can be alerted, such as an emergency contact, first responders, emergency services, and/or the like, after the wearer does not respond within a second predetermined time period.

[0035] In particular, the present disclosure is related to a wearable computing device, particularly a wearable computing device having a photoplethysmogram (PPG) sensor and a motion sensor, as well as a computer-implemented method, for accurately detecting pulselessness of the wearer via one or more algorithms. The present inventors have found that the particular algorithm(s) described herein reduce the likelihood of false positives of pulselessness and improve accuracy of pulselessness detection in wearable computing devices.

[0036] In general, the PPG sensor(s) of the present disclosure use one or more emitters, such as light-emitting diodes (LEDs) to emit controlled pulses of light and use one or more detectors, such as photodiodes, to capture the returned light. The emitted photons reflect off the skin, tissue, bones, blood, etc. of a wearer, and a processer controls the emitter(s) and converts the analog current received by the detector(s) into a digital PPG signal. Signal changes associated with peripheral perfusion, due to contract of the heart, enable the wearable computing device to measure the wearer’s pulsatility and heart rate in the PPG signal. The PPG sensors can be configured for use at various light wavelengths such as green (centered at 528 nanometers (nm)), red (centered at 660 nm), and infrared (centered at 940 nm), where the amplitude of the reflected light for the green, red, and infrared wavelengths changes with every heartbeat. Typically, the peak to peak amplitude for green PPG signals is greater than the peak to peak amplitude for red PPG signals and infrared PPG signals when there is a clear pulsatile signal such as that associated with a heartbeat.

[0037] Meanwhile, the motion sensor(s) of the present disclosure are configured for sensing and outputting movement data indicative of the motion of the wearer of the wearable monitoring device. For example, the motion sensors can include one or more accelerometers for sensing movement data. In some embodiments, the wearable monitoring device can include one or more accelerometers for sensing acceleration or other movement data in each of, for example, three directions (x, y, and z), which may be orthogonal. For instance, the accelerometer can be a triaxial accelerometer. The motion sensor(s) additionally can include one or more gyroscopes for sensing rotation data. In some embodiments, the wearable monitoring device can include one or more gyroscopes for sensing rotation about each of, for example, three axes, which may be orthogonal. The motion sensor(s) additionally can include one or more altimeters, such as a pressure or barometric altimeter. Data collected by the motion sensor(s) can be used to identify periods of relative stillness, which can suggest that a loss of pulse event may have occurred since purposeful movement as would be detected by the motion sensor(s) is not consistent with central pulselessness events.

[0038] As will be discussed in more detail below, the devices, methods, and systems of the present disclosure analyze signals collected from the motion sensor(s) and PPG sensor(s) to identify incipient loss of pulse with high confidence while minimizing false positives. This can be accomplished via a one or more algorithms running on the wearable computing device that can detect potential loss of pulse, verify such loss of pulse, and connect the wearer to an emergency contact, emergency services, first responders, and/or the like.

[0039] For instance, pulselessness can be detected via a first, passive, always-on algorithm, a second, semi-passive algorithm, a third, active algorithm, and a fourth, active algorithm. Once pulselessness is detected via the one or more of the algorithms referenced above, the devices, methods, and systems of the present disclosure can initiate communication with an emergency contact, first responders, emergency services, and/or the like, while if no pulselessness is detected via the one or more algorithms, no communication is initiated. Accordingly, the disclosed devices, systems, and methods allow for pulselessness of the wearer of the wearable computing device to be more accurately determined, which in turn results in a reduction in the number of false positives, which reduces the strain on first responders and other emergency personnel when responding to pulselessness alerts from wearable computing devices.

[0040] With reference now to the figures, example embodiments of the present disclosure will be discussed in further detail.

[0041] Referring now to the drawings, FIGS. 1-3 illustrate perspective views of a wearable computing device 102 according to the present disclosure. In particular, as shown in FIG. 1, the wearable computing device 102 may be worn on a user’s forearm 101 like a wristwatch. Thus, as shown, the wearable computing device 102 may include a wristband 103 for securing the wearable computing device 102 to the user’s forearm 101. However, it should be appreciated that the wearable computing device 102 may be worn at any other suitable location by a user, such as, for example, on an ankle. It should be further appreciated that the wearable computing device 102 can include a ring, band, earring, necklace, or any other wearable device known by one of skill in the art. In addition, as shown in FIGS. 1-3, the wearable computing device 102 has an outer covering 105 and a housing 104 that contains the electronics associated with the wearable computing device 102. For example, in an embodiment, the outer covering 105 may be constructed of glass, polycarbonate, acrylic, or similar. Further, as shown in FIGS. 1 and 2, the wearable computing device 102 includes an electronic display screen 106 arranged within the housing 104 and viewable through the outer covering 105. Moreover, as shown, the wearable computing device 102 may also include one or more buttons 108 that may be implemented to provide a mechanism to activate various sensors of the wearable computing device 102 to collect certain health data of the user. Moreover, in an embodiment, the electronic display 106 may cover an electronics package (not shown), which may also be housed within the housing 104.

[0042] Referring particularly to FIG. 3, one or more motion sensors 124 as described in detail above may be contained within the housing 104 of the wearable computing device 102. Further, the housing 104 of the wearable computing device 102 further includes a dorsal wrist-side face 110 configured to sit against a dorsal wrist of a user when being worn by the user, and one or more PPG sensors 126 as described above can be disposed on the dorsal wrist-side face 110 of the housing 104 of the wearable computing device 102 so as to maintain skin contact with the user when being worn on the wrist by the user.

[0043] Additionally, a plurality of sensor electrodes 125 can be positioned on the dorsal wrist-side face 110 of the housing 104 so as to maintain skin contact with the user when being worn on the wrist by the user. Thus, in such embodiments, each of the sensor electrodes 125 may be configurable to measure, at least, electrical impedance of the user at a location of the skin contact on the dorsal wrist. In some embodiments, the wearable computing device 102 may also include at least one additional biometric sensor electrode in addition to the PPG sensors 126 and the impedance sensor electrodes. In such embodiments, the additional biometric sensor electrode 125 may include one or more temperature sensors (such as an ambient temperature sensor or a skin temperature sensor), a humidity sensor, a light sensor, a pressure sensor, a microphone, or an optical sensor. In some instances, the sensor electrodes 125 may generally be arranged around the PPG sensors 126. [0044] Referring now to FIG. 4, components of an example computing system 100 of the wearable computing device 102 that can be utilized in accordance with various embodiments are illustrated. In particular, as shown, the computing system 100 may also include at least one processor 112 communicatively coupled to the motion sensor(s) 124 the PPG sensor(s) 126, and any other sensors present, such as the sensor electrodes 125. Moreover, in an embodiment, the processor(s) 112 may be a central processing unit (CPU) or graphics processing unit (GPU) for executing instructions that can be stored in a memory 114, such as flash memory or DRAM, among other such options.

[0045] For example, in an embodiment, the memory 114 may include RAM, ROM, FLASH memory, or other non-transitory digital data storage, and may include a control program comprising sequences of instructions 118 which, when loaded from the memory 114 and executed using the processor(s) 112, cause the processor(s) 112 to perform the functions that are described herein. As would be apparent to one of ordinary skill in the art, the computing system 100 can include many types of memory, data storage, or computer-readable media, such as data storage for program instructions for execution by any suitable processor. The same or separate storage can be used for images or data, a removable memory can be available for sharing information with other devices, and any number of communication approaches can be available for sharing with other devices.

[0046] In addition, as shown, the computing system 100 includes any suitable display 106, such as a touch screen, organic light emitting diode (OLED), or liquid crystal display (LCD), although devices might convey information via other means, such as through audio speakers, projectors, or casting the display or streaming data to another device, such as a mobile phone, wherein an application on the mobile phone displays the data.

[0047] The computing system 100 may also include one or more wireless components 212 operable to communicate with one or more electronic devices within a communication range of the particular wireless channel. The wireless channel can be any appropriate channel used to enable devices to communicate wirelessly, such as Bluetooth, cellular, NFC, Ultra- Wideband (UWB), or Wi-Fi channels. It should be understood that the computing system 100 can have one or more conventional wired communications connections as known in the art.

[0048] The computing system 100 also includes one or more power components 208, such as a battery operable to be recharged through conventional plug-in approaches, or through other approaches such as capacitive charging through proximity with a power mat or other such device. In further embodiments, the computing system 100 can also include at least one additional input-output (I/O) component 122 able to receive conventional input from a user. This conventional input can include, for example, a push button, touch pad, touch screen, wheel joystick, keyboard, mouse, keypad, or any other such device or element whereby a user can input a command to the computing system 100. In another embodiment, the I/O component(s) 122 may be connected by a wireless infrared or Bluetooth or other link as well in some embodiments. In some embodiments, the computing system 100 may also include a microphone or other audio capture element that accepts voice or other audio commands. For example, in particular embodiments, the computing system 100 may not include any buttons at all, but might be controlled only through a combination of visual and audio commands, such that a user can control the wearable computing device 102 without having to be in contact therewith. In certain embodiments, the I/O components 122 may also include one or more of the sensor electrodes 125 described herein, optical sensors, barometric sensors (e.g., altimeter, etc.), and the like.

[0049] Still referring to FIG. 4, the computing system 100 may also include a motion sensor 124, as well as a driver 214 and a photoplethy smogram (PPG) sensor 126 that includes at least some combination of one or more emitters 127 and one or more detectors 128 for measuring data for one or more metrics of a human body via optical signals, such as for a person wearing the wearable computing device 102. In such embodiments, as described above with reference to FIG. 3, for example, the PPG sensor 126 may be arranged within the housing 104 and at least partially exposed through the dorsal wrist-side face 110 of the housing 104. Thus, the sensor electrodes 125 may be positioned around the PPG sensor 126 on the wrist-side face 110 of the housing 104. In alternative embodiments, the various components of the PPG sensor 126 may be positioned around the sensor electrodes 125 and/or in another other suitable configuration such as adjacent to, interspersed with, surrounded by, or on top of the PPG sensor 126.

[0050] The emitters 127 and detectors 128 of FIG. 4 are capable of being used, in one example, for obtaining optical PPG measurements as part of the PPG sensor 126. Some PPG technologies rely on detecting light at a single spatial location, adding signals taken from two or more spatial locations, or an algorithmic combination thereof. Both of these approaches result in a single spatial measurement from which the heart rate (HR) estimate (or other physiological metrics) can be determined. In some embodiments, a PPG device employs a single light source coupled to a single detector (i.e., a single light path). Alternatively, a PPG device may employ multiple light sources coupled to a single detector or multiple detectors (i.e., two or more light paths). In other embodiments, a PPG device employs multiple detectors coupled to a single light source or multiple light sources (i.e., two or more light paths). In some cases, the light source(s) may be configured to emit one or more of green, red, infrared (IR) light, as well as any other suitable wavelengths in the spectrum. For example, a PPG device may employ a single light source and two or more light detectors each configured to detect a specific wavelength or wavelength range. In some cases, each detector is configured to detect a different wavelength or wavelength range from one another. In other cases, two or more detectors are configured to detect the same wavelength or wavelength range. In yet another case, one or more detectors configured to detect a specific wavelength or wavelength range different from one or more other detectors). In embodiments employing multiple light paths, the PPG device may determine an average of the signals resulting from the multiple light paths before determining an HR estimate or other physiological metrics.

[0051] Moreover, in an embodiment, the emitters 127 and detectors 128 may be coupled to the processor 112 directly or indirectly using driver circuitry by which the processor 112 may drive the emitters 127 and obtain signals from the detectors 128. Further, a server computing system 130 can communicate with the wireless networking components 212 via the one or more networks 180, which may include one or more local area networks, wide area networks, UWB, and/or internetworks using any of terrestrial or satellite links. In some embodiments, the server computing system 130 executes control programs and/or application programs that are configured to perform some of the functions described herein.

[0052] Referring now to FIG. 5, a schematic diagram of an environment 300 in which aspects of various embodiments can be implemented is illustrated. In particular, as shown, a user might have a number of different devices that are able to communicate using at least one wireless communication protocol. For example, as shown, the user might have a smartwatch 302 or fitness tracker (such as wearable computing device 102), which the user would like to be able to communicate with a smartphone 304 and a tablet computer 306. The ability to communicate with multiple devices can enable a user to obtain information from the smartwatch 302, e.g., data captured using a sensor on the smartwatch 302, using an application installed on either the smartphone 304 or the tablet computer 306. The user may also want the smartwatch 302 to be able to communicate with a service provider 308, or other such entity, which is able to obtain and process data from the smartwatch and provide functionality that may not otherwise be available on the smartwatch or the applications installed on the individual devices. In addition, as shown, the smartwatch 302 may be able to communicate with the service provider 308 through at least one network 310, such as the Internet or a cellular network, or may communicate over a wireless connection such as Bluetooth® to one of the individual devices, which can then communicate over the at least one network. There may be a number of other types of, or reasons for, communications in various embodiments.

[0053] In addition to being able to communicate, a user or wearer may also want the devices to be able to communicate in a number of ways or with certain aspects. For example, the user or wearer may want communications between the devices to be secure, particularly where the data may include personal health data or other such communications. The device or application providers may also be required to secure this information in at least some situations. The user may want the devices to be able to communicate with each other concurrently, rather than sequentially. This may be particularly true where pairing may be required, as the user may prefer that each device be paired at most once, such that no manual pairing is required. The user may also desire the communications to be as standards-based as possible, not only so that little manual intervention is required on the part of the user but also so that the devices can communicate with as many other types of devices as possible, which is often not the case for various proprietary formats. A user may thus desire to be able to walk in a room with one device and have such device automatically communicate with another target device with little to no effort on the part of the user. In various conventional approaches, a device will utilize a communication technology such as Wi-Fi to communicate with other devices using wireless local area networking (WLAN). Smaller or lower capacity devices, such as many Internet of Things (loT) devices, instead utilize a communication technology such as Bluetooth®, and in particular Bluetooth Low Energy (BLE) which has very low power consumption.

[0054] In further embodiments, the environment 300 illustrated in FIG. 5 enables data to be captured, processed, and displayed in a number of different ways. For example, data may be captured using sensors on the smartwatch 302, but due to limited resources on the smartwatch 302, the data may be transferred to the smartphone 304 or the service provider 308 (or a cloud resource) for processing, and results of that processing may then be presented back to that user on the smartwatch 302, smartphone 304, and/or another such device associated with that user, such as the tablet computer 306. In at least some embodiments, a user may also be able to provide input such as health data using an interface on any of these devices, which can then be considered when making that determination. Devices and Systems for Pulselessness Detection Algorithms

[0055] As mentioned above, the data collected from the motion sensor(s) 124 and the PPG sensor(s) 126 can be utilized in one or more digital signal processing and/or machine-learned algorithms in order to detect pulselessness of the wearer of the wearable computing device 102. Referring to FIG. 6 A, the first algorithm 400, which is passive, obtains sensor inputs 402 from the motion sensor(s) 124 and the PPG sensor(s) 126, including when the PPG sensor(s) 126 has its emitters 127 turned on and off (ambient), is utilized in a passive manner. This means that the first algorithm 400 is always on and can run in the low-power microcontroller unit (MCU) of the wearable computing device 102. Thus, the first algorithm 400 is low power and low memory and can be used in the first stage of identifying incipient loss of pulse by the wearer of the wearable computing device 102. Generally, the first algorithm 400 consumes data from the PPG sensor 126 and the motion sensor 124 and uses a combination of digital signal processing and machine-learned models in steps 404, 406, 408, 410, and 412, which are described in more detail below, to detect possible loss of pulse, where a final decision 414 as to pulselessness is made after receiving the output 412 of the first algorithm 400. In some embodiments, the PPG sensor 126 and the motion sensor 124 are sampled at a frequency of about 10 Hertz to about 150 Hertz, such as about 20 Hertz to about 120 Hertz, such as about 25 Hertz to about 100 Hertz, while the output 412 is reported at a frequency of about 0.3 Hertz to about 1 Hertz, such as about 0.4 Hertz to about 0.8 Hertz, such as about 0.5 Hertz.

[0056] Turning now to the specific components of the first algorithm 400, at step 404, which includes time domain data analysis, the first algorithm 400 analyzes the green wavelength PPG signal data to identify large drops in the alternating current (AC) component of the measured PPG signal in the time domain. This is because the received PPG signal includes five main parts, of which only the light reflected by the pulsatile blood contributes to the AC component during periods of stillness, which can be indicative of pulselessness of the wearer. The five main parts of the signal are (1) light reflected from blood, which is the signal of interest, and includes a pulsatile, AC component and a non-pulsatile direct current (DC) component, (2) light reflected from tissue such as skin and bone, which includes a constant, DC component, (3) ambident light, which typically includes a constant, DC component over short durations and is typically compensated for by an analog front end controller that drives the emitter and digitizes the measured detector current, (4) noise, which typically includes Johnson-Nyquist noise (thermal noise) and shot noise, among others, which have Gaussian and Poisson distributions, respectively, and which can be approximated as constant, DC components over short durations, and (5) motion artifacts, which is when motion by the wearer may introduce transient artifacts in the received PPG signal, and where it is noted that periods with substantial motion are identified in step 406 and are not identified as pulseless. The PPG signal may be intermittent in wearers who are moving and/or who wear the wearable computing device 102 loosely. In this scenario, the PPG signal may have a periodic component corresponding to when the watch is in contact with (or not in contact with) the skin. This is identified in either step 406 or step 408 of the first algorithm 400 as discussed in more detail below. It is noted that the pulsatile component of PPG signal may drop substantially if the wearer takes off (doffs) the wearable computing device 102, and this type of motion artifact can also be identified in step 406 or step 408 to avoid false positives of pulselessness.

[0057] To identify possible signs of loss of pulse, the step 404 time domain analysis of the passive first algorithm 400 quantifies the AC component of the PPG signal and detects when the AC component drops substantially. This includes: (1) normalizing the measured PPG signal by the emitted current from the LED or emitter 127 (also referred to as the current transfer ratio because the number of reflected photons is directly proportional to the number of emitted photons and normalizing by the LED current directly accounts for changes to the number of emitted photons per unit time, (2) bandpass filtering to about 0.5 Hertz to about 4 Hertz (also referred to as the bandpass-filtered current ratio), where this frequency range is selected because typical pulse rates are between 30 beats per minute and 200 beats per minute (0.5 Hertz to 3.3 Hertz), and filtering to 0.5 Hertz to 4 Hertz ensures that the signal-to-noise ratio at these frequencies is maximized, (3) quantifying the approximate AC component because since the AC component is the time-varying component of a signal, the peak-to-peak intensity of the bandpass filtered signal is computed over short durations (e.g., 2 seconds to 4 seconds, such as 3 seconds) to identify when pulsatility drops with minimal latency, and (4) identifying if the AC component drops substantially, where a drop in the amplitude of 50% or greater may suggest loss of pulse, warranting further investigation. If a large drop in the AC component of the normalized, filtered PPG signal is detected, then additional sensor data is investigated further to confirm whether loss of pulse may have occurred.

[0058] If loss of pulse is suspected after completing step 404 of the first algorithm 400, then, in step 406, then motion sensor 124 data that is input into the first algorithm 400 is subjected to gate-based analysis (e.g., analysis in which data from one or more sensors such as a motion sensor or accelerometer is quantitatively analyzed to determine if the purported loss of pulse event is likely to correspond to a true loss of pulse event) by looking at a change in signal over a specified time period to identify stillness of the wearer. Since the AC component of the PPG signal may be affected by transient user motion, periods of relative stillness are identified using a motion sensor 124 such as a triaxial accelerometer sensor as an orthogonal sensing modality to PPG. Such a motion sensor 124 can detect motion in three orthogonal spatial axes (x, y, z) and reports the acceleration components across these axes (typically at 100 Hertz). Since the motion sensor 124 components may be relative to the wearable computing device 102’s current orientation (i.e., are not necessarily based on an absolute coordinate system such as the Earth’s axes), the relative change across all accelerometer axes over rolling 10 second windows is computed to identify periods of stillness via the equation below:

A[Motion SenSOr] (Xmax ^— Xmin) + (ymax ^— ymin) + (Zmax ^— Zmin)

[0059] The epoch of duration of 10 seconds is sufficiently short to identify loss of pulse events that may occur in users who were moving just prior (e.g., walking, running, exercising, or otherwise engaged in everyday activities) and is sufficiently long to minimize the risk of false positives from users doffing the wearable computing device 102.

[0060] During this step 406 of motion sensor data analysis, if the relative motion sensor 124 change is less than about 0.15 g (about 1.47 m/s²), then the epoch is labeled as relatively still and the purported loss of pulse event warrants additional investigation and the next step 408 of the first algorithm 400 can be initiated.

[0061] Optionally, step 408 can include ambient PPG data analysis to reduce the risk of false positives of pulselessness from a wearer doffing the wearable computing device 102. In particular, the AC component of the PPG signal may also show artifacts from the user wearing the wearable computing device 102 loosely (with no or intermittent skin contact) or from the user doffing the wearable computing device 102. Loose wear may cause the wearable computing device 102 to intermittently decouple from the user’s skin. Similarly, doffing the wearable computing device 102 completely decouples the wearable computing device 102 from the user’s skin. In both cases, the change in optical coupling will manifest as a change in the measured PPG signal when the emitter 127 (e.g., LED) of the PPG sensor 126 is off (i.e., the ambient PPG channel).

[0062] Possible loose wear or doffing events can be identified in step 408 by computing the relative change of the signal of the ambient PPG channel over rolling 60 second windows/epochs via gate-based analysis by looking at a change in signal over a specified time period according to the equation below, where larger values indicate inconsistent ambient PPG measurements:

AfAmbient PPG channel signal] = [Ambient PPG signal]_max - [Ambient PPG signal]_min

[0063] The ambient channel associated with the infrared “low latency off body” (LLOB) PPG channel is utilized for this algorithm component. If the relative change in the ambient PPG channel is about 1000 units or greater in amplitude, then the epoch can be labeled as inconsistent and the purported loss of pulse event is thus determined to be unreliable. Meanwhile, if the relative change in the ambient PPG channel is below about 1000 units in amplitude, then the epoch can be labeled as consistent and the purported loss of pulse event warrants additional investigation and the next step 410 of the first algorithm 400 can be initiated.

[0064] Next, in step 410, preprocessing and feature generation for the machine learning component of the first algorithm 400 is conducted, where it is to be understood that this is another always-on component of the first algorithm 400. The feature generation or extraction is pulled from the PPG sensor 126 signal, the motion sensor 124 signal, and the PPG sensor 126 ambient signal. This feature extraction portion of the first algorithm 400 is designed to keep low-power and low-memory constraints in mind while also capturing important representations upon which a deep neural network (DNN) model could make a decision. Specifically, a streaming approach is used where, as real-time signals become available, a set of features are generated about every 2 seconds, where the features are kept within a 60 second sliding segment in the system 100’s memory 114. The preprocessing and feature calculation are linear in time and space complexity, which makes them suitable for always-on processing on wearable computing devices. The features in step 410 are calculated as follows, referring to Table 1 :

Table 1 : List of possible features derived from sensors for machine learning

[0065] If the digital signal processing-based steps 404, 406, and 408 of the first algorithm 400 identify a possible loss of pulse event, then the machine learning-based step 412 of the first algorithm 400 can be initiated to add an additional probabilistic check. The machine learning portion of the first algorithm 400 does not consume raw sensor data (which is sampled at greater than or equal to about 25 Hertz). Rather, the machine-learned model uses features from the past 60 second segments derived from the raw sensors that are computed in step 410.

[0066] Each feature set is generated once every 2 seconds using the last 5 seconds of signals. Hence, there are a total of 28 feature sets within a 60 second segment. To improve generalization, a few middle feature sets can be removed, which may capture artifacts of the tourniquet inflation-the modality to train the pulseless algorithm) and use only 22 feature sets where each set contains 24 features. Additionally, these are folded from the middle to have 11 folded feature sets where each set contains 48 combined features to improve locality. [0067] In one embodiment, the machine-learned model can generally include a few convolutional neural network (CNN) layers and its architecture is shown in Table 2 below, although other models are also contemplated by the present disclosure and generally described in detail below:

Table 2: CNN Machine-Learned Model Example

[0068] It should be noted that the dense layer has one output unite, and the output of the machine-learned model is a number within zero to one, which indicates the likelihood that the first 25 seconds of the segment being analyzed is pulsatile, and the last 25 seconds of the segment is pulseless.

[0069] To further enhance the utility of the machine-learned model, two more machine learning inferences can be conducted over the last 64 second and 68 second segments, with removal of an extra two or four feature sets from the middle, respectively, so an input of 22 feature sets are taken. Since the machine learning inferences of step 412 are only conducted a few times per hour, battery life is not significantly affected by the additional runs.

[0070] After completing steps 404, 406, 408 (optional), and 410 of the first algorithm 400 as described above, a final decision can be made based on the results in step 414. To avoid false positives of pulselessness from the first algorithm 400, a pulseless event decision is reported if and only if: (1) the past 10 seconds acceleration peak to peak amplitude is less than about 0.15 g, (2) at least 50% of the last 5 seconds of PPG signals are less than 10% in amplitude of the 5 second PPG signals that occurred 30 seconds ago, (3) (optional) the peak to peak amplitude of the ambient PPG signal over the past 60 seconds is less than 500 units, and (4) if conditions (1), (2), and (3)(optional) are met, nine machine learning inferences are run across three 2-second strides, and 60 second, 64 second, and 68 second segment durations, respectively, where at least 8 out of the 9 inference results should be more than 0.94 on a scale of 0 to 1.

[0071] Specifically, if at time t, conditions (1), (2), and (3)(optional) are met, then machine learning inferences are run using the features from the following 9 segments: [0072] (l) t-58 to t+2

(2) t-62 to t+2

(3) t-66 to t+2

(4) t-56 to t+4

(5) t-60 to t+4

(6) t-64 to t+4

(7) t-54 to t+6

(8) t-58 to t+6

(9) t-62 to t+6

[0073] FIG. 6B depicts a block diagram of an example computing system 500 that can performs the machine learning portion of the first algorithm 400 in step 412 according to example embodiments of the present disclosure. The system 500 includes the wearable computing device 102 described above, a server computing system 130, and a training computing system 150 that are communicatively coupled over a network 180.

[0074] The wearable computing device 102 includes one or more processors 112 and a memory 114. The one or more processors 112 can be any suitable processing device (e.g., a processor core, a microprocessor, an ASIC, an FPGA, a controller, a microcontroller, etc.) and can be one processor or a plurality of processors that are operatively connected. The memory 114 can include one or more non -transitory computer-readable storage media, such as RAM, ROM, EEPROM, EPROM, flash memory devices, magnetic disks, etc., and combinations thereof. The memory 114 can store data 116 and instructions 118 which are executed by the processor 112 to cause the wearable computing device 102 to perform operations.

[0075] In some implementations, the wearable computing device 102 can store or include one or more models pulselessness detection models 120 that are used in step 412 of the first algorithm. For example, the one or more models pulselessness detection models 120 can be or can otherwise include various machine-learned models such as neural networks (e.g., deep neural networks) or other types of machine-learned models, including non-linear models and/or linear models. Neural networks can include feed-forward neural networks, recurrent neural networks (e.g., long short-term memory recurrent neural networks), convolutional neural networks or other forms of neural networks. Some example machine-learned models can leverage an attention mechanism such as self-attention. For example, some example machine-learned models can include multi-headed self-attention models (e.g., transformer models).

[0076] In some implementations, the one or more models pulselessness detection models 120 can be received from the server computing system 130 over network 180, stored in the wearable computing device memory 114, and then used or otherwise implemented by the one or more processors 112. In some implementations, the wearable computing device 102 can implement multiple parallel instances of a single pulselessness detection model 120.

[0077] Additionally or alternatively, one or more pulselessness detection models 140 can be included in or otherwise stored and implemented by the server computing system 130 that communicates with the wearable computing device 102 according to a client-server relationship. For example, the pulselessness detection models 140 can be implemented by the server computing system 140 as a portion of a web service. Thus, one or more models 120 can be stored and implemented at the wearable computing device 102 and/or one or more models 140 can be stored and implemented at the server computing system 130.

[0078] The wearable computing device 102 can also include one or more user input components 122 that receives user input. For example, the user input component 122 can be a touch-sensitive component (e.g., a touch-sensitive display screen or a touch pad) that is sensitive to the touch of a user input object (e.g., a finger or a stylus). The touch-sensitive component can serve to implement a virtual keyboard. Other example user input components include a microphone, a traditional keyboard, or other means by which a user can provide user input.

[0079] The server computing system 130 includes one or more processors 132 and a memory 134. The one or more processors 132 can be any suitable processing device (e.g., a processor core, a microprocessor, an ASIC, an FPGA, a controller, a microcontroller, etc.) and can be one processor or a plurality of processors that are operatively connected. The memory 134 can include one or more non-transitory computer-readable storage media, such as RAM, ROM, EEPROM, EPROM, flash memory devices, magnetic disks, etc., and combinations thereof. The memory 134 can store data 136 and instructions 138 which are executed by the processor 132 to cause the server computing system 130 to perform operations.

[0080] In some implementations, the server computing system 130 includes or is otherwise implemented by one or more server computing devices. In instances in which the server computing system 130 includes plural server computing devices, such server computing devices can operate according to sequential computing architectures, parallel computing architectures, or some combination thereof.

[0081] As described above, the server computing system 130 can store or otherwise include one or more pulselessness detection models 140. For example, the models 140 can be or can otherwise include various machine-learned models. Example machine-learned models include neural networks or other multi-layer non-linear models. Example neural networks include feed forward neural networks, deep neural networks, recurrent neural networks, and convolutional neural networks. Some example machine-learned models can leverage an attention mechanism such as self-attention. For example, some example machine-learned models can include multi-headed self-attention models (e.g., transformer models).

[0082] The wearable computing device 102 and/or the server computing system 130 can train the models 120 and/or 140 via interaction with the training computing system 150 that is communicatively coupled over the network 180. The training computing system 150 can be separate from the server computing system 130 or can be a portion of the server computing system 130.

[0083] The training computing system 150 can include one or more processors 152 and a memory 154. The one or more processors 152 can be any suitable processing device (e.g., a processor core, a microprocessor, an ASIC, an FPGA, a controller, a microcontroller, etc.) and can be one processor or a plurality of processors that are operatively connected. The memory 154 can include one or more non-transitory computer-readable storage media, such as RAM, ROM, EEPROM, EPROM, flash memory devices, magnetic disks, etc., and combinations thereof. The memory 154 can store data 156 and instructions 158 which are executed by the processor 152 to cause the training computing system 150 to perform operations. In some implementations, the training computing system 150 includes or is otherwise implemented by one or more server computing devices.

[0084] The training computing system 150 can include a model trainer 160 that trains the machine-learned models 120 and/or 140 stored at the wearable computing device 102 and/or the server computing system 130 using various training or learning techniques, such as, for example, backwards propagation of errors. For example, a loss function can be backpropagated through the model(s) to update one or more parameters of the model(s) (e.g., based on a gradient of the loss function). Various loss functions can be used such as mean squared error, likelihood loss, cross entropy loss, hinge loss, and/or various other loss functions. Gradient descent techniques can be used to iteratively update the parameters over a number of training iterations.

[0085] In some implementations, performing backwards propagation of errors can include performing truncated backpropagation through time. The model trainer 160 can perform a number of generalization techniques (e.g., weight decays, dropouts, etc.) to improve the generalization capability of the models being trained.

[0086] In particular, the model trainer 160 can train the pulselessness detection models 120 and/or 140 based on a set of training data 162. The training data 162 can include, for example, data associated with past instances of pulselessness for the wearer or others based on acquired sensor data from a motion sensor 124 and/or a PPG sensor 126 using the processes described above.

[0087] In some implementations, if the user has provided consent, the training examples can be provided by the wearable computing device 102. Thus, in such implementations, the model 120 provided to the wearable computing device 102 can be trained by the training computing system 150 on user-specific data received from the wearable computing device 102. In some instances, this process can be referred to as personalizing the model.

[0088] The model trainer 160 includes computer logic utilized to provide desired functionality. The model trainer 160 can be implemented in hardware, firmware, and/or software controlling a general purpose processor. For example, in some implementations, the model trainer 160 includes program files stored on a storage device, loaded into a memory and executed by one or more processors. In other implementations, the model trainer 160 includes one or more sets of computer-executable instructions that are stored in a tangible computer-readable storage medium such as RAM, hard disk, or optical or magnetic media. [0089] The network 180 can be any type of communications network, such as a local area network (e.g., intranet), wide area network (e.g., Internet), or some combination thereof and can include any number of wired or wireless links. In general, communication over the network 180 can be carried via any type of wired and/or wireless connection, using a wide variety of communication protocols (e.g., TCP/IP, HTTP, SMTP, FTP), encodings or formats (e.g., HTML, XML), and/or protection schemes (e.g., VPN, secure HTTP, SSL).

[0090] The machine-learned models described in this specification may be used in a variety of tasks, applications, and/or use cases. In some implementations, the input to the machine- learned model(s) of the present disclosure can be sensor data. The machine-learned model(s) can process the sensor data to generate an output. As an example, the machine-learned model(s) can process the sensor data to generate a recognition output. As another example, the machine-learned model(s) can process the sensor data to generate a prediction output. As another example, the machine-learned model(s) can process the sensor data to generate a classification output. As another example, the machine-learned model(s) can process the sensor data to generate a segmentation output. As another example, the machine-learned model(s) can process the sensor data to generate a visualization output. As another example, the machine-learned model(s) can process the sensor data to generate a diagnostic output. As another example, the machine-learned model(s) can process the sensor data to generate a detection output.

[0091] In some implementations, the input to the machine-learned model(s) of the present disclosure can be statistical data. Statistical data can be, represent, or otherwise include data computed and/or calculated from some other data source. The machine-learned model(s) can process the statistical data to generate an output. As an example, the machine-learned model(s) can process the statistical data to generate a recognition output. As another example, the machine-learned model(s) can process the statistical data to generate a prediction output. As another example, the machine-learned model(s) can process the statistical data to generate a classification output. As another example, the machine-learned model(s) can process the statistical data to generate a segmentation output. As another example, the machine-learned model(s) can process the statistical data to generate a visualization output. As another example, the machine-learned model(s) can process the statistical data to generate a diagnostic output.

[0092] FIG. 6B illustrates one example computing system 500 that can be used to implement the machine learning aspects of the present disclosure. Other computing systems can be used as well. For example, in some implementations, the wearable computing device 102 can include the model trainer 160 and the training dataset 162. In such implementations, the models 120 can be both trained and used locally at the wearable computing device 102. In some of such implementations, the wearable computing device 102 can implement the model trainer 160 to personalize the models 120 based on user-specific data.

[0093] FIG. 6C depicts a block diagram of an example computing device 600 that performs according to example embodiments of the present disclosure. The computing device 700 can be a wearable computing device or a server computing device.

[0094] The computing device 600 includes a number of applications (e.g., applications 1 through N). Each application contains its own machine learning library and machine-learned model(s). For example, each application can include a machine-learned model. Example applications include a biometric sensor application, a motion sensor application, a text messaging application, a browser application, etc.

[0095] As illustrated in FIG. 6C, each application of the computing device 600 can communicate with a number of other components of the computing device, such as, for example, one or more sensors, a context manager, a device state component, and/or additional components. In some implementations, each application can communicate with each device component using an API (e.g., a public API). In some implementations, the API used by each application is specific to that application.

[0096] FIG. 6D depicts a block diagram of an example computing device 700 that performs according to example embodiments of the present disclosure. The computing device 800 can be a wearable computing device or a server computing device.

[0097] The computing device 700 includes a number of applications (e.g., applications 1 through N). Each application is in communication with a central intelligence layer. Example applications include a biometric sensor application, a motion sensor application, a text messaging application, a browser application, etc. In some implementations, each application can communicate with the central intelligence layer (and model(s) stored therein) using an API (e.g., a common API across all applications).

[0098] The central intelligence layer includes a number of machine-learned models. For example, as illustrated in FIG. 6D, a respective machine-learned model can be provided for each application and managed by the central intelligence layer. In other implementations, two or more applications can share a single machine-learned model. For example, in some implementations, the central intelligence layer can provide a single model for all of the applications. In some implementations, the central intelligence layer is included within or otherwise implemented by an operating system of the computing device 700.

[0099] The central intelligence layer can communicate with a central device data layer. The central device data layer can be a centralized repository of data for the computing device 700. As illustrated in FIG. 6D, the central device data layer can communicate with a number of other components of the computing device, such as, for example, one or more sensors, a context manager, a device state component, and/or additional components. In some implementations, the central device data layer can communicate with each device component using an API (e.g., a private API).

[0100] Regardless of the particular machine-learned model used in step 412 of the first algorithm 400, if the first algorithm 400 detects a purported or suspected loss of pulse, then a second algorithm 450 is initiated. Referring to FIG. 7, the second algorithm 450 is semipassive and obtains sensor inputs 452 from the motion sensor(s) 124 and the PPG sensor(s) 126, including when the PPG sensor(s) 126 has its emitters 127 turned on and off (ambient) and when light is emitted by the emitters 127 over various wavelengths (green, red, infrared, etc.) The second algorithm 450, when on, can run in the low-power microcontroller unit (MCU) of the wearable computing device 102. Further, it should be understood that loss of pulse is confirmed by verifying that the sensor input data measured by the wearable computing device 102 is consistent with the physiological state of central or peripheral pulselessness.

[0101] Turning first to the green PPG signal data analysis in step 454, it is noted that during central (or peripheral) pulselessness, the green PPG signal does not have a pulsatile/ AC component due to lack of peripheral blood flow. As such, the second algorithm 450 measures this physiological state in the green PPG signal by estimating the signal -to-noise ratio (SNR) at frequencies corresponding to typical pulse rates and verifying that this SNR is low. The SNR at pulsatile frequencies is quantified by first computing the power spectral density of the green PPG signal (which is sampled at about 25 Hertz over 10 seconds) via the periodogram. The power spectral density quantifies the contributions of discrete frequency components (and integer multiples thereof [due to periodogram algorithm limitations]) to the overall signal.

[0102] If a signal contains a pulsatile component, then peaks would appear in the power spectral density at the frequency of the pulsatile component and integer multiples thereof (e.g., at 60, 120, 180 beats per minute (bpm), etc. for a signal with a true pulsatile component at 60 bpm. The SNR is defined as the difference between the signal of interest and the noise floor. The signal of interest is the maximum power at typical pulse rates (40 to 220 bpm), while the noise floor is approximated by the median power at these frequencies. If the SNR is 30 dB or less, then the signal is deemed to be pulseless. If the SNR is greater than 30 dB, then the signal is classified as pulsatile.

[0103] Next, during the infrared PPG signal data analysis in step 454, it is noted that the infrared PPG signal will also not have a pulsatile component in the absence of peripheral blood flow. The infrared PPG signal may capture pulsatility more sensitively than the green PPG due to its increased skin penetration depth of infrared photons compared to green photons, which reduces the impact of aggressors such as excessive pressure on the watch face artificially reducing the green PPG signal even when the user is pulsatile and due to less variation across users with diverse skin tones, in part due to a lower absorbance of the skin pigment melanin at infrared wavelengths than at green wavelengths. Infrared wavelengths have greater skin penetration depth than green wavelengths due to reduced optical scattering at higher wavelengths (for example, one mechanism of optical scattering is Rayleigh Scattering, which is proportional to wavelength to the negative fourth power; in other words, Rayleigh scattering at infrared wavelengths centered at 940 nm should be about a factor of 10 lower than at green wavelengths centered at 528 nm).

[0104] The SNR at pulsatile frequencies is calculated by estimating the power spectral density with a periodogram. The algorithm is the same as that for the green PPG, albeit with a different SNR threshold of 25 dB, where if the infrared SNR is 25 dB or less, then the signal is deemed to be pulseless, and if the SNR is greater than 25 dB, then the signal is classified as pulsatile.

[0105] Additionally, during the ambient infrared PPG signal data analysis in step 454, it is noted that possible loss of pulse events may be falsely detected if the user wears the wearable computing device 102 loosely or doffs the wearable computing device 102. The AC component of the PPG signal may also show artifacts from the user wearing the wearable computing device 102 loosely or from the user doffing the device. Loose wear may cause the wearable computing device 102 to intermittently decouple from the user’s skin. Similarly, doffing the wearable computing device 102 completely decouples the wearable computing device 102 from the user’s skin. In both cases, the change in optical coupling will manifest as a change in the measured PPG signal when the emitter 127 (e.g., LED) is off: i.e., the ambient PPG channel described in this section.

[0106] Possibly loose wear or doffing can be thus identified by computing the relative change of the ambient PPG channel over the 10 seconds before purported loss of pulse events via the equation below, where larger values indicate inconsistent ambient PPG measurements:

AfAmbient PPG channel] = [Ambient PPG]_max - [Ambient PPG]_min

[0107] If the relative change in the peak to peak amplitude of the ambient PPG channel is 300 units or greater, then the epoch is labeled as inconsistent and the purported loss of pulse event as being unreliable. On the other hand, if the relative change in the peak to peak amplitude of the ambient PPG channel is below 300 units, then the epoch is labeled as consistent and the purported loss of pulse event as warranting additional investigation. The ambient PPG channel associated with the infrared low latency off body (LLOB) PPG channel can be used for this second algorithm 450 component, though other ambient PPG channels may be used as well.

[0108] Lastly, in the second algorithm 450, motion sensor data can be analyzed in step 454. After an individual experiences a central loss of pulse event, the individual’s muscles may become “limp,” which manifests as a fall if the individual was previously upright, slumping or a fall if seated, or laying down if reclined. The person will remain in this motionless state unless they receive cardiopulmonary resuscitation (CPR) and/or defibrillation. The motion sensor 124 data analysis subcomponent in step 454 of the second algorithm 450 can detect the presence of this motionless state after purported loss of pulse events. This component includes taking the accelerometer motion components in the 3 orthogonal spatial axes (x, y, z) which are sampled at 100 Hertz, computing the relative change across all accelerometer axes over the past 30 seconds via the equation below, and verifying that the sum across all axes is below a threshold:

A[Motion SenSOr] (Xmax Xmin) + (ymax ymin) + (Zmax Zmin) [0109] The epoch duration of 30 seconds is sufficiently short to identify loss of pulse events that may occur in users who were moving just prior (e.g., walking, running, exercising, or otherwise engaged in everyday activities) while also being sufficiently long to minimize the risk of false positives from users doffing the wearable computing device 102. However, it is to be understood that the epoch duration may be lengthened or shortened to tune the sensitivity and specificity of the second algorithm 450. Relevant epoch durations could be as short at 1 second and as long as 120 seconds. If the relative accelerometer change is less than 0.3 g (2.943 m/s2), then we label the epoch as relatively still and the purported loss of pulse event as being a likely loss of pulse event.

[0110] Additional sensors such as the electrodermal activity (EDA) sensor, additional PPG channels, etc. may supplement these checks to improve the accuracy of the second algorithm 450. If, at step 456, a final decision of pulselessness is made with respect to the second algorithm 450, then the third algorithm 470 can be activated. While the first algorithm 400 and the second algorithm 450 are typically invisible to the wearer since the wearable computing device 102 performs passive and semi -passive checks to identify possible loss of pulse events automatically, without interaction from the wearer, the third algorithm 470 is considered active in that it initiates a subtle “check-in” with the wearer if the first algorithm 400 and the second algorithm 450 indicate loss of pulse.

[0111] At a high level, the third algorithm 470 attempts to confirm whether the wearer is truly pulseless, where, in step 472, the third algorithm 470 displays a notification to the wearer, after which, in step 474, the third algorithm provides a haptic stimulus to the wearer. Then, in step 476, a customized PPG sensor configuration can be initiated, and then in step 478, the third algorithm 470 can analyze the PPG sensor data and motion sensor data for signals or waveforms indicative of motion in the motion sensor 124 and/or pulsatility in the PPG sensor 124 in the vicinity of the haptic stimulus applied in step 474.

[0112] The third algorithm 470 design is guided by the pathophysiology of central pulseless events, during which individuals become unconscious (and thus does not show purposeful movement) and the limbs are no longer perfused (i.e., no pulsatility in the measured PPG signal). To improve the third algorithm 470’ s specificity and maximize the probability of detecting pulsatility if present (especially in scenarios of reduced peripheral perfusion, users with low Monk score, loose wear, or excessive pressure on the watch face), this algorithm component activates a custom PPG configuration in step 476 that uses multiple combinations of emitters and detectors (e.g., LEDs and photodiodes) at varied LED current and photodiode gain settings. If there is no response to the display notification in step 472 and the haptic stimulus in step 474 as determined by analyzing the sensor data, a final decision of pulselessness with respect to the third algorithm 470 can be made in step 480.

[0113] Then, the fourth algorithm 490 can be initiated. The fourth algorithm 490, unlike the previous three algorithms, does not consume sensor data and includes a step 492 where a notification is displayed to the wearer on the display screen 106 of the wearable computing device 102. The notification will indicate that emergency services (and/or a user’s emergency contacts) will be contacted unless the user dismisses the alert. A countdown timer will also be shown (with a typical duration of 20 seconds, though this could be lengthened or shortened to tune the latency and specificity, e.g. to 3 to 120 seconds) and started in step 494. If the user or wearer responds to the notification and provides input in step 496 before the timer started in step 494 expires, then the alert can be dismissed in step 498. Meanwhile, if the user or wearer does not respond to the notification and provide input in step 496, then a third party is contacted in step 499, where the third party can be an emergency contact, a first responder, emergency services, and the like.

[0114] In summary, the four algorithms described above can be utilized to avoid false positives in detecting pulselessness of a wearer of a wearable computing device to ensure that assistance is provided only when needed to help provide life-saving care efficiently and effectively.

Additional Disclosure

[0115] The technology discussed herein makes reference to servers, databases, software applications, and other computer-based systems, as well as actions taken and information sent to and from such systems. The inherent flexibility of computer-based systems allows for a great variety of possible configurations, combinations, and divisions of tasks and functionality between and among components. For instance, processes discussed herein can be implemented using a single device or component or multiple devices or components working in combination. Databases and applications can be implemented on a single system or distributed across multiple systems. Distributed components can operate sequentially or in parallel.

[0116] In addition, although the figures and description depict and describe steps performed in a particular order for purposes of illustration and discussion, the methods discussed herein are not limited to any particular order or arrangement. One skilled in the art, using the disclosures provided herein, will appreciate that various steps of the methods disclosed herein can be omitted, rearranged, combined, added, and/or adapted in various ways without deviating from the scope of the present disclosure. [0117] While the present subject matter has been described in detail with respect to various specific example embodiments thereof, each example is provided by way of explanation, not limitation of the disclosure. Those skilled in the art, upon attaining an understanding of the foregoing, can readily produce alterations to, variations of, and equivalents to such embodiments. Accordingly, the subject disclosure does not preclude inclusion of such modifications, variations and/or additions to the present subject matter as would be readily apparent to one of ordinary skill in the art. For instance, features illustrated or described as part of one embodiment can be used with another embodiment to yield a still further embodiment. Thus, it is intended that the present disclosure cover such alterations, variations, and equivalents.

Claims

WHAT IS CLAIMED IS:

1. A computing system to detect pulselessness in a wearer of a wearable computing device, the computing system comprising: one or more processors; and one or more non-transitory computer-readable media that collectively store instructions that, when executed by the one or more processors, cause the computing system to perform operations, the operations comprising, as part of a first algorithm:

(i) receiving, via the one or more processors, signal data from a motion sensor and a photoplethysmogram sensor associated with the wearable computing device;

(ii) performing, via the one or more processors, time domain analysis on a green light signal received from the photoplethysmogram sensor to quantify a decrease in an alternating current component of the green light signal, wherein if the decrease in the alternating current component of the green light signal is above a predetermined percentage over a predetermined time period, pulselessness is suspected, and wherein if the decrease in the alternating current component of the green light signal is less than or equal to the predetermined percentage over the predetermined time period, pulselessness is not suspected;

(iii) if pulselessness is suspected in step (ii), performing, via the one or more processors, gate-based analysis on an acceleration signal received from the motion sensor to quantify an acceleration change over a predetermined time period, wherein if the acceleration change is less than a predetermined value, pulselessness is suspected, and wherein if the acceleration change over the predetermined time period is greater than or equal to the predetermined value, then pulselessness is not suspected; and

(iv) if pulselessness is suspected in step (ii) and/or step (iii), implementing, via the one or more processors, a machine-learned model trained using a dataset of signal data from a prior predetermined amount of time to determine if the signal data received from the motion sensor, the photoplethysmogram sensor, or both is indicative of pulselessness.

2. The computing system of claim 1, the operations further comprising, as part of the first algorithm:

(iii)(a) if pulselessness is suspected in step (iii) and prior to performing step (iv), performing, via the one or more processors, gate-based analysis on an ambient light signal received from the photoplethysmogram sensor to quantify a change in a peak to peak amplitude of the ambient light signal, wherein if the change in the peak to peak amplitude of the ambient light signal is less than a predetermined level over a predetermined time period, pulselessness is suspected, and wherein if the change in the peak to peak amplitude of the ambient light signal is greater than or equal to a predetermined level over a predetermined time period, pulselessness is not suspected.

3. The computing system of claim 1, the operations further comprising, as part of the first algorithm:

(v) evaluating, via the one or more processors, the time domain analysis and the machine-learned model determinations from the signal data to predict whether or not pulselessness is suspected, and if pulselessness is suspected, to initiate further investigation by the computing system.

4. The computing system of claim 1, wherein the first algorithm is passive.

5. The computing system of claim 3, wherein pulselessness is suspected, the operations further comprising, as part of a second algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; and estimating, via the one or more processors, a signal-to-noise ratio of a green light signal received from the photoplethysmogram sensor at one or more predetermined frequencies, wherein if the signal-to-noise ratio is less than or equal to a predetermined value at one or more of the predetermined frequencies, pulselessness is suspected, and wherein if the signal-to-noise ratio is greater than the predetermined value at one or more of the predetermined frequencies, pulselessness is not suspected.

6. The computing system of claim 3, wherein pulselessness is suspected, the operations further comprising, as part of a second algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; and estimating, via the one or more processors, a signal-to-noise ratio of an infrared light signal received from the photoplethysmogram sensor at one or more predetermined frequencies, wherein if the signal-to-noise ratio is less than or equal to a predetermined value at one or more of the predetermined frequencies, pulselessness is suspected, and wherein if the signal-to-noise ratio is greater than the predetermined value at one or more of the predetermined frequencies, pulselessness is not suspected.

7. The computing system of claim 3, wherein pulselessness is suspected, the operations further comprising, as part of a second algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; and performing, via the one or more processors, gate-based analysis on an ambient light signal received from the photoplethysmogram sensor to quantify a change in a peak to peak amplitude of the ambient light signal, wherein if the change in the peak to peak amplitude of the ambient light signal is less than a predetermined level over a predetermined time period, pulselessness is suspected, and wherein if the change in the peak to peak amplitude of the ambient light signal is greater than or equal to a predetermined level over a predetermined time period, pulselessness is not suspected.

8. The computing system of claim 3, wherein pulselessness is suspected, the operations further comprising, as part of a second algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; and performing, via the one or more processors, gate-based analysis on an acceleration signal received from the motion sensor to quantify an acceleration change over a predetermined time period, wherein if the acceleration change is less than a predetermined value, pulselessness is suspected, and wherein if the acceleration change over the predetermined time period is greater than or equal to the predetermined value, then pulselessness is not suspected.

9. The computing system of claim 3, wherein pulselessness is suspected, the operations further comprising, as part of a third algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; generating, via the one or more processors, a first visual notification to provide an alert the wearer of the wearable computing device by controlling, via the one or more processors, a display screen of the wearable computing device to display the first visual notification; delivering, via the one or more processors, a haptic stimulus through the wearable computing device; and analyzing, via the one or more processors, the signal data received from the motion sensor and the photoplethysmogram sensor to determine if the signal data received during delivery of the haptic stimulus is indicative of motion and/or pulsatility, wherein if the signal data is not indicative of motion and/or pulsatility, then pulselessness is suspected, and wherein if the signal data is indicative of motion and/or pulsatility, then pulselessness is not suspected.

10. The computing system of claim 9, wherein pulselessness is suspected, the operations further comprising, as part of a fourth algorithm: generating, via the one or more processors, a second visual notification for a predetermined period of time to alert the wearer of the wearable computing device by controlling, via the one or more processors, the display screen of the wearable computing device to display the second visual notification; and if the wearer has not responded to the second visual notification after the predetermined period of time has passed, initiating an emergency alert communication, via the one or more processors, with a third party.

11. A computer-implemented method to detect pulselessness of a wearer of a wearable computing device, the method comprising, as part of a first algorithm:

(i) receiving, via one or more processors, signal data from a motion sensor and a photoplethysmogram sensor associated with the wearable computing device;

12. The computer-implemented method of claim 11, further comprising, as part of the first algorithm: (iii)(a) if pulselessness is suspected in step (iii) and prior to performing step (iv), performing, via the one or more processors, gate-based analysis on an ambient light signal received from the photoplethysmogram sensor to quantify a change in a peak to peak amplitude of the ambient light signal, wherein if the change in the peak to peak amplitude of the ambient light signal is less than a predetermined level over a predetermined time period, pulselessness is suspected, and wherein if the change in the peak to peak amplitude of the ambient light signal is greater than or equal to a predetermined level over a predetermined time period, pulselessness is not suspected.

13. The computer-implemented method of claim 11, further comprising, as part of the first algorithm:

(v) evaluating, via the one or more processors, the time domain analysis and the machine-learned model determinations from the signal data to predict whether or not pulselessness is suspected, and if pulselessness is suspected, to initiate further investigation by a computing system.

14. The computer-implemented method of claim 11, wherein the first algorithm is passive.

15. The computer-implemented method of claim 13, wherein pulselessness is suspected, further comprising, as part of a second algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; and estimating, via the one or more processors, a signal-to-noise ratio of a green light signal received from the photoplethysmogram sensor at one or more predetermined frequencies, wherein if the signal-to-noise ratio is less than or equal to a predetermined value at one or more of the predetermined frequencies, pulselessness is suspected, and wherein if the signal-to-noise ratio is greater than the predetermined value at one or more of the predetermined frequencies, pulselessness is not suspected.

16. The computer-implemented method of claim 13, wherein pulselessness is suspected, further comprising, as part of a second algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; and estimating, via the one or more processors, a signal-to-noise ratio of an infrared light signal received from the photoplethysmogram sensor at one or more predetermined frequencies, wherein if the signal-to-noise ratio is less than or equal to a predetermined value at one or more of the predetermined frequencies, pulselessness is suspected, and wherein if the signal-to-noise ratio is greater than the predetermined value at one or more of the predetermined frequencies, pulselessness is not suspected.

17. The computer-implemented method of claim 13, wherein pulselessness is suspected, further comprising, as part of a second algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; and performing, via the one or more processors, gate-based analysis on an ambient light signal received from the photoplethysmogram sensor to quantify a change in a peak to peak amplitude of the ambient light signal, wherein if the change in the peak to peak amplitude of the ambient light signal is less than a predetermined level over a predetermined time period, pulselessness is suspected, and wherein if the change in the peak to peak amplitude of the ambient light signal is greater than or equal to a predetermined level over a predetermined time period, pulselessness is not suspected.

18. The computer-implemented method of claim 13, wherein pulselessness is suspected, further comprising, as part of a second algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; and performing, via the one or more processors, gate-based analysis on an acceleration signal received from the motion sensor to quantify an acceleration change over a predetermined time period, wherein if the acceleration change is less than a predetermined value, pulselessness is suspected, and wherein if the acceleration change over the predetermined time period is greater than or equal to the predetermined value, then pulselessness is not suspected.

19. The computer-implemented method of claim 13, wherein pulselessness is suspected, further comprising, as part of a third algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; generating, via the one or more processors, a first visual notification to provide an alert the wearer of the wearable computing device by controlling, via the one or more processors, a display screen of the wearable computing device to display the first visual notification; delivering, via the one or more processors, a haptic stimulus through the wearable computing device; and analyzing, via the one or more processors, the signal data received from the motion sensor and the photoplethysmogram sensor to determine if the signal data received during delivery of the haptic stimulus is indicative of motion and/or pulsatility, wherein if the signal data is not indicative of motion and/or pulsatility, then pulselessness is suspected, and wherein if the signal data is indicative of motion and/or pulsatility, then pulselessness is not suspected.

20. The computer-implemented method of claim 19, wherein pulselessness is suspected, further comprising, as part of a fourth algorithm: generating, via the one or more processors, a second visual notification for a predetermined period of time to alert the wearer of the wearable computing device by controlling, via the one or more processors, the display screen of the wearable computing device to display the second visual notification; and if the wearer has not responded to the second visual notification after the predetermined period of time has passed, initiating an emergency alert communication, via the one or more processors, with a third party.

21. A wearable computing device to detect pulselessness of a wearer of the wearable computing device, the wearable computing device comprising: one or more processors; and one or more non-transitory computer-readable media that collectively store instructions that, when executed by the one or more processors, cause the wearable computing device to perform operations, the operations comprising, as part of a first algorithm:

22. The wearable computing device of claim 21, the operations further comprising, as part of the first algorithm:

23. The wearable computing device of claim 21, the operations further comprising, as part of the first algorithm:

(v) evaluating, via the one or more processors, the time domain analysis and the machine-learned model determinations from the signal data to predict whether or not pulselessness is suspected, and if pulselessness is suspected, to initiate further investigation by the wearable computing device.

24. The wearable computing device of claim 21, wherein the first algorithm is passive.

25. The wearable computing device of claim 23, wherein pulselessness is suspected, the operations further comprising, as part of a second algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; and estimating, via the one or more processors, a signal-to-noise ratio of a green light signal received from the photoplethysmogram sensor at one or more predetermined frequencies, wherein if the signal-to-noise ratio is less than or equal to a predetermined value at one or more of the predetermined frequencies, pulselessness is suspected, and wherein if the signal-to-noise ratio is greater than the predetermined value at one or more of the predetermined frequencies, pulselessness is not suspected.

26. The wearable computing device of claim 23, wherein pulselessness is suspected, the operations further comprising, as part of a second algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; and estimating, via the one or more processors, a signal-to-noise ratio of an infrared light signal received from the photoplethysmogram sensor at one or more predetermined frequencies, wherein if the signal-to-noise ratio is less than or equal to a predetermined value at one or more of the predetermined frequencies, pulselessness is suspected, and wherein if the signal-to-noise ratio is greater than the predetermined value at one or more of the predetermined frequencies, pulselessness is not suspected.

27. The wearable computing device of claim 23, wherein pulselessness is suspected, the operations further comprising, as part of a second algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; and performing, via the one or more processors, gate-based analysis on an ambient light signal received from the photoplethysmogram sensor to quantify a change in a peak to peak amplitude of the ambient light signal, wherein if the change in the peak to peak amplitude of the ambient light signal is less than a predetermined level over a predetermined time period, pulselessness is suspected, and wherein if the change in the peak to peak amplitude of the ambient light signal is greater than or equal to a predetermined level over a predetermined time period, pulselessness is not suspected.

28. The wearable computing device of claim 23, wherein pulselessness is suspected, the operations further comprising, as part of a second algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; and performing, via the one or more processors, gate-based analysis on an acceleration signal received from the motion sensor to quantify an acceleration change over a predetermined time period, wherein if the acceleration change is less than a predetermined value, pulselessness is suspected, and wherein if the acceleration change over the predetermined time period is greater than or equal to the predetermined value, then pulselessness is not suspected.

29. The wearable computing device of claim 23, wherein pulselessness is suspected, the operations further comprising, as part of a third algorithm: receiving, via the one or more processors, signal data from the motion sensor and the photoplethysmogram sensor; generating, via the one or more processors, a first visual notification to provide an alert the wearer of the wearable computing device by controlling, via the one or more processors, a display screen of the wearable computing device to display the first visual notification; delivering, via the one or more processors, a haptic stimulus through the wearable computing device; and analyzing, via the one or more processors, the signal data received from the motion sensor and the photoplethysmogram sensor to determine if the signal data received during delivery of the haptic stimulus is indicative of motion and/or pulsatility, wherein if the signal data is not indicative of motion and/or pulsatility, then pulselessness is suspected, and wherein if the signal data is indicative of motion and/or pulsatility, then pulselessness is not suspected.

30. The wearable computing device of claim 29, wherein pulselessness is suspected, the operations further comprising, as part of a fourth algorithm: generating, via the one or more processors, a second visual notification for a predetermined period of time to alert the wearer of the wearable computing device by controlling, via the one or more processors, the display screen of the wearable computing device to display the second visual notification; and if the wearer has not responded to the second visual notification after the predetermined period of time has passed, initiating an emergency alert communication, via the one or more processors, with a third party.