BR112020019815A2 - ANALYSIS AND PRESENTATION OF PHOTOPLETISMOGRAM DATA - Google Patents

Abstract

“Analysis and presentation of photoplethysmogram data” computer software systems, methods and products that can be used for analysis and presentation of photoplethysmogram data are disclosed. for example, photoplethysmogram (ppg) signal data is received as communicated by a ppg sensor from a wearable device worn by a user. a heart rate interval can be determined from at least the ppg signal data. in addition, an electrocardiogram (ecg) waveform covering the heartbeat interval can be generated in a graphical interface.

Description

“ANALYSIS AND PRESENTATION OF PHOTOPLETISMOGRAM DATA” RELATED ORDER (OR RELATED ORDERS)

[0001] This request claims the priority and benefit of Provisional Application No. US 62 / 649,527, filed on March 28, 2018, entitled “System and Methods for Heartbeat Detection”, of Provisional Application No. 62 / 649,530, filed on 28 March 2018, entitled “Systems and Methods for Detection of Atrial Fibrillation” and Patent Application No. US 62 / 649,533, filed on March 28, 2018, entitled “Systems and Methods For Photoplethysmogram Data Analysis and Presentation”, where the disclosures of each of them are incorporated by reference.

DESCRIPTION OF RELATED TECHNIQUE

[0002] The electrical and physiological characteristics of the human heart can be measured using, for example, sensors, such as electrocardiogram (ECG) sensors or photoplethysmograph (PPG) sensors. The signals from these sensors can then be analyzed to determine useful and informative information about a patient's condition, such as heart rates, specific heart rates and the like.

SUMMARY

[0003] The present invention discloses computer program systems, methods and products that can be used for analysis and presentation of photoplethysmogram data. For example, photoplethysmogram (PPG) signal data is received as it is communicated by a PPG sensor from a device usable close to the body worn by a user. A heart rate range can be determined from at least the PPG signal data. In addition, an electrocardiogram (ECG) waveform that covers the heartbeat interval can be generated in a graphical interface.

[0004] In some implantations, the ECG type waveform includes a PQRST waveform that includes features that represent the electrical activity in a heart. Based on at least signal data

PPG, it is possible to detect whether an atrial fibrillation rhythm is represented in the PPG signal data. In addition, a P wave displayed on the ECG-type waveform can be suppressed when atrial fibrillation rhythm is detected. In some implantations, based on at least the PPG signal data, it is possible to detect whether an atrial fibrillation rhythm is represented in the PPG signal data. A P wave can be displayed as part of the ECG type waveform when an atrial fibrillation rhythm is not detected.

[0005] In some deployments, a first heart rate interval can be calculated based on at least the PPG signal data. A second heartbeat interval, the second heartbeat interval occurring after the first heartbeat interval, can be calculated based on at least the PPG signal data. A graphical interface can generate a scatterplot that represents a variation of the heartbeat intervals, the scatterplot includes a graphic in a location determined by at least the first heartbeat interval and the second heartbeat interval. In some deployments, the graphical interface can display both the ECG waveform and the scatterplot.

[0006] The deployments of this subject may include, but are not limited to, methods consistent with the descriptions provided in this document, as well as articles comprising a machine-readable medium tangibly incorporated that has the purpose of making one or more machines (for example, computers, etc.) result in operations that implement one or more of the described resources. Similarly, computer systems that may include one or more processors and one or more memories coupled with one or more processors are also contemplated. A memory, which may include a computer-readable storage medium, may include, encode, store, or the like, one or more programs that cause one or more processors to perform one or more of the operations described herein. Computer-implemented methods consistent with one or more deployments of the present subject can be deployed by one or more data processors that reside in a single computing system or in multiple computing systems. These multiple computing systems may be connected and may exchange data and / or commands or other instructions, or the like, through one or more connections, including, but not limited to, a connection over a network (for example, the internet, a wireless wide area network, a local area network, a wide area network, a wired network, or the like), through a direct connection between one or more of the various computing systems, etc.

[0007] Details of one or more variations of the subject described in this document are presented in the attached drawings and in the description below. Other features and advantages of the subject described in this document will become evident from the description, drawings and claims. Although certain features of this disclosed matter are described for illustrative purposes in relation to specific deployments, it should be readily understood that these features are not intended to be limiting. The claims that accompany this disclosure are intended to define the scope of the protected matter.

BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The attached drawings, which are incorporated and constitute a part of this specification, show certain aspects of the matter disclosed in this document and, together with the description, help to explain some of the principles associated with the disclosed implantations.

[0009] Figure 1 illustrates an exemplary system that can provide monitoring of the health characteristics of the user and provide guidance related to health according to certain aspects of this disclosure.

[0010] Figure 2 illustrates an implantation of a device usable with the user's body according to certain aspects of this disclosure.

[0011] Figure 3 illustrates an implantation of a communication device according to certain aspects of the present disclosure.

[0012] Figure 4 illustrates a server deployment according to certain aspects of this disclosure.

[0013] Figure 5 illustrates an exemplary PPG signal and a corresponding PPG signal gradient according to certain aspects of the present disclosure.

[0014] Figure 6 illustrates an exemplary implementation of a method for determining heartbeat according to certain aspects of the present disclosure.

[0015] Figure 7 is a diagram that illustrates an exemplary process for determining the presence of atrial fibrillation according to certain aspects of the present disclosure.

[0016] Figure 8 is a diagram illustrating an exemplary scatterplot that represents a normal heart rate according to certain aspects of the present disclosure.

[0017] Figure 9 is a diagram that illustrates an exemplary moving window used in the generation of a scatter plot according to certain aspects of the present disclosure.

[0018] Figure 10 is a diagram that illustrates an exemplary method for determining an occupancy metric from a scatter plot that represents an irregular heart rate according to certain aspects of the present disclosure.

[0019] Figure 11 is a diagram illustrating an exemplary method for determining a distance metric from a scatter plot that represents an irregular heart rate according to certain aspects of the present disclosure.

[0020] Figure 12 is a diagram that illustrates an exemplary method for determining a variability metric from heartbeat interval data according to certain aspects of the present disclosure.

[0021] Figure 13 is a diagram that illustrates an exemplary method for determining a three-dimensional space and discrimination plans to determine a type of heart rate according to certain aspects of the present disclosure.

[0022] Figure 14 is a diagram that illustrates an exemplary decision tree for determining a type of heart rate according to certain aspects of this disclosure.

[0023] Figure 15 is a diagram illustrating an exemplary patient panel in accordance with certain aspects of the present disclosure.

[0024] Figure 16 is a diagram that illustrates an exemplary viewer according to certain aspects of the present disclosure.

[0025] Figure 17 is a diagram that illustrates an exemplary viewer that includes an expanded graph with a waveform that has ECG type characteristics according to certain aspects of the present disclosure.

[0026] Figure 18 is a diagram illustrating an exemplary viewer that includes an expanded graph with a waveform that has ECG-like characteristics and a suppressed P wave according to certain aspects of the present disclosure.

[0027] Figure 19 is a diagram illustrating an exemplary viewer that includes a scatter plot of points that represent changes in heartbeat intervals according to certain aspects of the present disclosure.

[0028] Figure 20 is a diagram that illustrates an exemplary method of generating an ECG type waveform according to certain aspects of the present disclosure.

DETAILED DESCRIPTION

[0029] The matter described in this document refers to systems, methods and software to monitor the health of a user and provide the user with health guidelines.

[0030] Figure 1 illustrates an example system 100 that can provide the monitoring of a user's health characteristics (for example, a human patient or another living organism) and can provide health guidance to the user based on the monitoring of health characteristics. Cheers.

[0031] In some deployments, the exemplary system 100 shown in Figure 1 may include elements, such as device (or devices) usable with the user body 108 (for example, a smart watch), communication devices 102, 104 and 106 ( eg a mobile phone or PC), user monitoring devices 110 and 112 (for example, a separate smart scale or blood glucose monitor), data analysis device (or devices) 114, server (or servers ) 116 (for example, including processor (or processors) 117 and bank (or databases) 118) and network (or networks) 120. The server (or servers) 116 and the devices illustrated in Figure 1 may include lines or ports communication to allow the exchange of information on a network (for example, network 120), or on other computing platforms using wired or wireless techniques (for example, Ethernet, fiber optics, coaxial cable, WiFi, Bluetooth , communication in near field or other technologies).

[0032] It should be noted that, although one or more operations are described in this document as being performed by specific components of system 100, these operations can, in some modalities, be performed by other components of system 100. For example, although one or more operations are described in this document as being performed by data analysis device components (or devices) 114, these operations may, in other modalities, be performed by device components (or devices) usable with the user body 108, by components of communication devices 102, 104 and 106 and / or by other components of system 100.

[0033] The device (or devices) usable with the user body 108 can be a smart watch (for example, Samsung Gear, Apple Watch, etc.) or any other device that a user can use. A device usable close to user body 108 may include one or more sensors that are integrated into the device. For example, a device usable close to user body 108 that is a smart watch can include motion sensors (eg accelerometers), bioimpedance sensors, ECG sensors, ballistic cardiogram sensors, acoustic sensors (eg ultrasound), photoplethysmography (PPG) sensors that use light-based technology to detect a blood flow rate, and other sensors. It can also be considered in the present document that the device usable next to the body 108 includes sensors that are used in the body of a user, but are not integrated into the portion usable next to the main body (for example, an ECG sensor used in the chest of a user who is not integrated with a smart watch, but who nevertheless communicates with the smart watch).

[0034] Figure 2 illustrates the device usable next to body 108, including processing circuits 202, sensor (or sensors) 204, user interface usable next to body 206, device application usable next to body 208 and memory 210. According to observed, the sensor (or sensors) 204 may include multiple sensors integrated into the usable portion next to the main body of the device and / or sensors located elsewhere in the user's body. The device application usable next to the body 208 and the signals from the sensor (or sensors) 204 can be stored in memory 210.

[0035] A user can interact with the user interface usable next to the body 206, for example, to enter data such as age, height, weight and sex, or to view measured or calculated metrics, such as heart rate, pulse rate variability , stress level, respiratory orientation and the like.

[0036] The device application usable next to body 208 can be run on processing circuits 202 and perform operations such as receiving signals from the sensor (or sensors) 204, calculating various health characteristics, issuing the information display, providing health guidelines to the user, etc.

[0037] The device usable next to the body 108 can be subjected to certain configurations during a calibration period. For example, a user can wear device 108 for a 24-hour calibration period during first use to allow the collection of user information from the sensor (or sensors) 204. For example, the collection of characteristics such as frequency of pulse or respiratory rate over a period of time can facilitate device calibration and provide useful information to the user in the future analysis of signals and in providing health guidance to the user. In some beneficial deployments, calibration can be performed while the user is using a single lead ECG or other sensor (or sensors) for reference purposes.

[0038] Communication devices 102, 104 and 106 can include any type of mobile or fixed device, for example, a desktop-type computer, a notebook-type computer, a smart phone, a tablet-type computer or other communication device . Users can, for example, use one or more communication devices 102, 104 and 106 to interact with each other, with one or more devices usable close to the body, one or more servers or other system components 100.

[0039] Figure 3 illustrates some components of an exemplary communication device 104, including processing circuits 302, memory 304, user interface 306 and communication device application 308. Processing circuits 302, memory 304 and the interface User 306 work similarly to processing circuitry 202, memory 210 and user interface 206, respectively, in Figure 2, although the application and user interface of a communication device generally have greater functionality than the of a device that can be used close to the body.

[0040] In some deployments, the 308 communication device application can be a mobile application (for example, a smart phone application) or a web application. The communication device application 308, in some deployments, can communicate with the device application usable with the user body 208 via Bluetooth (or any other wired or wireless communication method) and / or can transmit measurements for archiving and post-processing to a cloud-based database (for example, database (or databases) 118). The communication device application 308 can aggregate data from other sensors (for example, from user monitoring devices 110 and 112), perform pre-transmission processing locally, and transmit data for further processing or viewing.

[0041] In some deployments, user monitoring devices 110 and 112 may include a blood pressure monitoring device (for example, a blood pressure cuff), a weight monitoring device (for example, a scale), a blood glucose monitoring device, etc. User monitoring devices 110 and 112 can measure user health states different from the health states measured by the device (or devices) usable with the user's body 108.

[0042] The monitoring systems and methods and health guidelines detailed in this document normally use signals from one or more sensors that may be in contact with a user's body and that detect relevant information for the user. The sensors can be integrated with a wearable device close to the body, communicating with a wearable device close to the body, or, instead, they can be separated from a usable device close to the body, communicating with the system 100 through other components.

[0043] As further discussed in this document, system 100 may include components and methods for acquiring particular signals, for processing those signals (for example, noise reduction) and for modifying signal acquisition methods. Each of these activities can be performed by any of the components of the system

100.

[0044] In a deployment, a device usable next to user body 108 can capture an optical signal (for example, a pulse signal) from the optical sensor (or optical sensors) that use green and / or infrared wavelengths. The device usable close to the body 108 can also capture a motion signal that can be used to assess noise or interference resulting from the movement of a user using device 108 or to assess other parameters relevant to health analysis and guidance.

[0045] In some deployments, the optical signal and the motion signal can be buffered in a memory (for example, memory 210 in Figure 2) of the device (or devices) usable with the user body 108 for a period predetermined time and the optical signal and the motion signal can then be provided to other processors for processing those signals (for example, processing circuits 302 of the communication device 104 in Figure 3 or the device circuits (or devices ) of data analysis 114). In this way, the energy consumption of the device (or devices) usable with the user body 108 can be conserved or optimized. Alternatively, in some deployments, the processing circuits 202 of a device usable close to user body 108 can be used for processing the optical pulse signal and the motion signal captured by the device usable close to user body 108.

[0046] The collection or acquisition of signal from an optical sensor at a sampling frequency of 12-50 Hz can be used, especially under ideal conditions, such as the general absence of user movement combined with low levels of perfusion and low interference of ambient light. Several conditions can affect the sampling rate considered ideal, but one of the most striking conditions is the movement of a user using the device 108.

[0047] Signal processing challenges caused by user movement can be overcome by adjusting various parameters related to signal acquisition. For example, the performance of the optical sensor can be adjusted when the activity is detected by a motion sensor (for example, a three-axis accelerometer). In some deployments, if movement above a specific threshold is detected, any or all of the following acquisition parameters for an optical sensor can be adjusted to overcome the noise level and improve the accuracy of determining the health characteristic: (i ) sampling frequency, (ii) LED energy and / or (iii) pulses per sample. On the other hand, in some deployments, if movement below a specific threshold is detected, each of these acquisition parameters can be adjusted to maintain a specific level of performance and measurement accuracy, while also conserving energy.

[0048] In the general absence of user movement, a sampling frequency of approximately 20 Hz may be appropriate, but in more challenging environments, sampling can be increased to 100 or 200 Hz or, if necessary, up to 1,000 Hz or more to ensure that the signals received are useful for analyzing the health characteristics of the user. The use of other sampling frequencies is also contemplated.

[0049] Various health characteristics of a user can be determined using signals from the sensors discussed here. As an example, sensors associated with a device usable close to the user's body, such as a wristwatch, can be used to determine heart rate, pulse rate variability (PRV) or heart rate variability (HRV) of a user. Heart rate is generally described as the number of heartbeats per minute, while HRV and PRV refer to the variability in time intervals between beats. HRV typically refers to measurements of variability based on electrocardiography and can be derived from R-R intervals on the standard PQRS waveform. An HRV determination can use an ECG sensor on a user who can communicate with the device usable close to the body 108. PRV, on the other hand, usually refers to determinations of variability based on sensors placed close to peripheral arteries, as an optical sensor (or optical sensors) on a user's pulse that provides a peripheral pulse waveform absent from the morphological information seen in an ECG signal.

[0050] The health characteristics of the user can be determined through the signal analysis performed on the device usable next to the user body 108 or other components of the system 100, such as the communication device 102 or the data analysis device 114, or analysis can be performed on more than one component of system 100.

[0051] In some deployments, the received sensor signal can be an ECG signal and the time at which each heartbeat occurred can be determined, for example, from each R peak in the ECG waveform. Alternatively, the time that each heartbeat occurred can be determined from a PPG signal. In an exemplary implantation, the heartbeat times of a PPG signal can be determined using maximum points on a PPG gradient graph (see, for example, Figure 5). The improved resolution for these determinations can be obtained through a variety of methods, for example, spline interpolation, which is discussed in detail below.

[0052] After the received signals are analyzed and the precise heart beat times are determined (for example, in a sample time of 10 seconds), a heart rate in beats per minute can be determined.

[0053] Referring now to Figure 4, together with Figure 1, Figure 4 illustrates an exemplary deployment of server 116 in Figure 1. As shown in Figure 4, server 116 includes processors 117, database 118, code software 419, preprocessing module 420, decomposition module 422 and heart rate detection module

424. As shown further in Figure 4, server 116 can be configured to receive PPG signal 428 and transmit heart rate 430 over network 120 to any of the components of system 100. As noted above, the concepts described in this document they can be implanted in server 116, in the device usable next to the body 108, in the communication device 106, in the other elements of the system 100 or in combinations thereof.

[0054] In other exemplary deployments, the server module components 117 (modules 420, 422, 424) can be deployed in hardware (including, for example, FPGAs and ASICs), firmware, software and / or combinations thereof. As used in this document, the term “module” is not intended to be limited to a specific physical form. Based on the particular application, the modules can be deployed as firmware, software, hardware and / or combinations thereof. In an exemplary deployment, the modules can be deployed as dedicated circuits (for example, part of an ASIC). This reduces energy consumption with greater speed. In another exemplary deployment, the modules can be deployed as software, which runs on digital signal processors and / or general purpose processors. Various combinations can be implemented. In addition, different modules can share common components or be deployed by the same components. There may or may not be a clear boundary between each component of the module. Finally, the methods described in this document do not necessarily need to be implemented by modules. The modular structures described here are provided simply as deployments and potential examples to facilitate description.

[0055] Depending on the shape of the modules, the "communication" between the modules can also take different forms. Dedicated circuits can be coupled to each other by wiring or by accessing a common register or memory location, for example. Software “communication” can occur in several ways to pass information between modules (or between software and hardware, if applicable). The term “in communication” is intended to include all of this and is not intended to be limited to a permanent wired connection between two components. In addition, there may be intermediate elements. For example, when two elements are described as being "in communication", this does not mean that the elements are directly coupled together nor does it prevent the use of other elements between the two.

[0056] In an exemplary deployment, server 116 can be configured to receive, by processors 417, a photoplethysmographic signal (PPG) 428 communicated by a PPG sensor of the device usable next to body 108 used by an individual. Server 116 can also be configured to process at least a portion of the PPG signal 428 through frequency band filters to create band outputs corresponding to heart rate interval bands. The server 116 can further be configured to use a bandwidth output indicated between the bandwidth outputs to determine a first heartbeat and a second heartbeat. In addition, server 116 can be configured to determine heart rate 430 based on at least the first heartbeat and second heartbeat and cause heart rate 430 to be transmitted to at least one usable device near the body 108.

[0057] In an exemplary implantation, after receiving the signal

PPG 428, processor 117 can be configured to run software code 419 in order to detect the presence of noise in the PPG 428 signal and perform the exemplary deployments described herein. The PPG 428 signal may include a raw signal that may or may not have noise and artifact. By detecting the presence of noise in the PPG 428 signal, redundant or unnecessary noise removal operations can be avoided, thus achieving faster and more efficient heart rate determinations.

[0058] In an exemplary deployment, detecting the presence of noise and artifact may include determining an instantaneous amplitude estimate of the PPG 428 signal and forming a time-smoothed version of the PPG 428 signal. The instantaneous peak amplitude of the dominant component of the PPG 428 signal tends to be stable from heartbeat to heartbeat when the signal is noise free. However, in the presence of noise, the instantaneous amplitude tends to oscillate significantly. Therefore, large deviations from the time-smoothed version of the PPG 428 signal can correspond to noise bursts in the PPG 428 signal.

[0059] In an exemplary deployment, processor 117 can be configured to detect large deviations of instantaneous amplitude from the time-smoothed version of the PPG 428 signal. When such deviations of instantaneous amplitude exceed a predetermined amount, processor 117 can be configured to determine that noise and artifact are present in the PPG 428 signal.

[0060] In another exemplary implementation, when the instant amplitude deviations of the PPG signal smoothed in time 428 are less than a predetermined amount, processor 117 can be configured to determine that noise and artifact are not present in the PPG 428 signal. the PPG 428 signal is considered to be noise-free, an exemplary deployment may include not pre-processing the PPG 428 signal to remove noise, as discussed further below.

[0061] In yet another example implementation, the instantaneous amplitude of the PPG 428 signal is compared to a moving average of the instantaneous amplitude. For example, a 30-second moving average of the PPG signal

428. If the instantaneous amplitude is greater than the moving average by predetermined multiplicative factor, the PPG 428 signal portion receives noise signaling. In an exemplary deployment, the predetermined multiplicative factor can be a factor of 1.5. In other exemplary deployments, the predetermined multiplicative factor can be greater than 1.5 and / or less than 2.

[0062] In an exemplary deployment, by detecting large deviations of instantaneous amplitude, as discussed above, in order to guarantee complete coverage of a noisy portion of the PPG 428 signal, the pre-processing module 420 can be configured to extend the duration of the PPG 428 signal that received noise signaling.

[0063] Although the use of the time-smoothed version of the PPG 428 signal to determine instantaneous deviations from the PPG 428 signal is especially suitable for detecting the presence of noise, other methods of determining the presence of noise in the PPG 428 signal can be used without deviating from the scope and spirit of the exemplary deployments described here.

[0064] After processing the PPG signal 428 for noise detection, as described above, the server 116, through the pre-processing module 420, can be configured to perform the pre-processing of the PPG 428 signal.

[0065] In an exemplary deployment, the preprocessing module 420 can be configured to receive the PPG 428 signal and perform the preprocessing of the PPG 428 signal. In an exemplary deployment, the preprocessing of the PPG 428 signal can include the removal of high frequency noise and low frequency artifact. Preprocessing can also include the removal of gross final effects, including noise pedestal, baseline dispersion and DC offset in the PPG 428 signal.

[0066] In an exemplary deployment, the processing of the PPG 428 signal by the preprocessing module 420 may first include the removal of the gross final effects due to the large pedestal of DC noise that varies slowly in time inherent to the PPG 428 signal. This can be achieved through the implementation of basic straight-line trend reduction techniques. The reduction of the basic straight line trend can act as a mitigation of the final effects of the PPG 428 signal due to the large DC displacement that occurs at the ends of the raw input signals, that is, that occurs at the ends of a portion of the PPG 428 signal.

[0067] In an exemplary deployment, the removal of the gross final effect may further include the removal of the pedestal at each end of at least a portion of the PPG 428 signal. In an exemplary deployment, the pedestal at each end of at least a portion of the signal PPG 428 can be estimated by averaging the signal vector PPG 428 over nominally 2 seconds from each end. A trend line can then be implemented from the first sample time to the final sample time of the PPG 428 signal. The trend line then passes through the average values corresponding to the time of the first sample and the final time of the sample. sample. Subtracting the values resulting from this trend line calculation from the raw input signal, ie PPG 428 signal, results in the trend reduction of the emitted signal, ie, a filtered PPG 428 signal.

[0068] In an exemplary deployment, the removal of large artifacts may include the implementation of high-pass and low-pass filtering to suppress large artifacts. High-pass and low-pass filtering can include cascading low-pass and high-pass filters to suppress high and low frequency artifacts, respectively. In another exemplary implementation, the high-pass filter can be deployed by subtracting the output centered on the delay of an internal low-pass filter from the inlet of the low-pass filter.

[0069] In an exemplary deployment, pre-processing filtering can be deployed using linear phase filters to preserve the primary morphological features of the underlying PPG signal 428 and to align input-to-output time delays across all the frequencies. In another example implementation, pre-processing filtering may include the use of cascades of boxcar filters. In an exemplary deployment, the preprocessing of the PPG 428 signal may include the use of a bandpass filter with cutoff frequencies in a high bandwidth of 0.5 hertz and low bandwidth of 10 hertz.

[0070] In another exemplary deployment, the PPG 428 signal may experience additional input processing to remove large existing artifacts outside the desired bandwidth of the PPG 428 signal. In yet another exemplary deployment, other filter protocols to remove noise and artifact can be used, for example, blind source separation with the use of independent component analysis to discover signal components from an independent source and deduce linear mixtures of underlying sources. Other noise removal filtering processes can be implemented without deviating from the scope and spirit of the present deployments described here and have been fully contemplated.

[0071] In an exemplary deployment, in order to remove the intrinsic latency delay associated with the preprocessing PPG 428 signal, the output signal from the preprocessing module 420, ie, filtered PPG 428 signal, can have the same length as the input signal, ie the raw and time-aligned PPG 428 signal as well. In another exemplary deployment, the output signal may not be the same length as the input signal and the filtered PPG signal may not be aligned with time. While the preprocessing methods described above are especially suitable for deploying the exemplary deployments described here, other methods of separating time series signals into composite subcomponents using a signal reference can be deployed without deviating from scope and spirit of the present deployments and were fully contemplated here.

[0072] As used herein, the filtered PPG 428 signal can simply be called the PPG 428 signal. Any reference to the filtered PPG 428 signal or simply PPG 428 signal can thus refer to a filtered or unfiltered PPG signal and it should not be interpreted as limiting the PPG 428 signal to a specific implantation.

[0073] In an exemplary deployment, after performing the pre-processing of the PPG signal 428, the pre-processing module 420 can be configured to communicate the filtered PPG signal 428 to the decomposition module 422. The decomposition module 422 can then , perform operations on the filtered PPG 428 signal in preparation to detect heartbeat and determine heart rate.

[0074] The decomposition module 422 can be configured to decompose the PPG 428 signal. The decomposition can be implemented in a linear phase to preserve the morphological characteristics of the PPG 428 signal. The decomposition of the PPG 428 signal can also include the separation of the PPG signal 428 in a series of subsignals that together comprise substantially the signal PPG 428 (which means, substantially, that the difference may be insignificant). Therefore, the decomposition module 422 can be used to decompose the PPG signal 428 by separating from the PPG signal 428 a set of PPG subsignals.

[0075] The decomposition of the PPG 428 signal may include processing at least a portion of the PPG 428 signal through frequency band filters. In an exemplary deployment, frequency band filters can correspond to bandwidth ranges from 0.55 to 1.37 hertz, 0.78 to 2.33 hertz and 1.26 to 4.25 hertz, respectively. In another exemplary implementation, the frequency band filters are designed to be less than an octave wide. In other words, the bandwidth of the frequency band filters is chosen to avoid the inclusion of two or more successive harmonics of the PPG signal. In this way, only the first harmonic of the input signals appears substantially at the output for periodic input signals locally within the frequency band. Consequently, in the time around each heartbeat, the frequency band filter with a bandpass (ie bandwidth band) that covers the local fundamental pulse rate will be excited by the largest amount and therefore will display the greatest amplitude at the output.

[0076] In addition to having the largest amplitude, the largest amplitude output band of the frequency band filters will also exhibit the highest purity of the local sinusoid, since only the fundamental harmonic appears at the output of the frequency band filters. Consequently, the output of the frequency band with the greatest amplitude corresponds to the fundamental frequency of the input signal (i.e., the main component of the input signal).

[0077] In an exemplary deployment, frequency band filters can include a wavelet bank that uses a growing semi-diagonal cascade of low-pass / high-pass separations. In another exemplary implementation, decomposition module 422 can be configured to decompose the PPG signal 428 through the wavelet bank and produce an output. The wavelength bank outputs (ie frequency band outputs) can be aligned with delay and centralized in time on the input signal, signal PPG 428. In this way, the intrinsic delay due to the latency associated with the wavelength bank can be removed to ensure accurate heart rate determinations. Therefore, processing at least a portion of the PPG signal 428 through frequency band filters may include creating bandwidth outputs.

[0078] In an exemplary deployment, decomposition of the PPG 428 signal can include processing the PPG 428 signal through frequency band filters to create band outputs corresponding to a plurality of heartbeat interval bands. The heart rate range ranges may correspond to frequency ranges associated with the individual's expected heart rates. For example, the heart rate range ranges may correspond to a normal range, a tachycardia range and a bradycardia range. In other deployments, the heart rate range ranges may correspond to different ranges that may be useful in separating the PPG 428 signal into frequency ranges corresponding to the user's likely heart rate ranges.

[0079] Other signal processing techniques can be implemented to ensure that the frequency band filters produce outputs that are robust and free from signal degradation complexities. For example, in an exemplary deployment, the use of frequency band filters may include applying an input vector gain between each of the successive cascades to address signal attenuation and other signal degradation problems. Several other approaches can be used to filter the PPG 428 signal, including: combined filters, second-order derivatives, non-linear time scale decomposition, adaptive filtering, dynamic time synchronization, artificial neural networks or hidden Markov models.

[0080] The decomposition module 422 can be configured to communicate decomposed portions of the PPG signal 428 (for example, frequency band outputs) to the heart rate detection module 424. In an exemplary deployment, the frequency detection module cardiac 424 can be configured to track each of the communicated frequency band outputs and determine, by comparison, which of the frequency band outputs has the greatest amplitude. The higher bandwidth output can be indicative of the user's heart rate within that band, and that band can therefore be used to determine heart rate. Thus, the heart rate detection module 424 can be configured to determine amplitudes for the band outputs and to determine the widest band output between the band outputs,

the identification of a “indicated” band output corresponds to the bandwidth with the greatest amplitude. The indicated band output can then be used to determine a first heartbeat, a second heartbeat and a heart rate from the PPG 428 signal, as discussed further below.

[0081] In another deployment, the heart rate detection module 424 can be configured to use more than one frequency band output. For example, the heart rate detection module 424 can generate a continuous estimate of the amplitudes of the frequency band outputs and can perform additional processing on the band outputs with the two highest amplitudes. This method can provide faster heart rate determination / tracking in cases where heart rates are rapidly changing from beat to beat and perhaps moving from one frequency band to another. Therefore, in an exemplary deployment, in addition to determining the widest bandwidth output, analysis module 424 can be configured to determine a second widest bandwidth output and use the second widest bandwidth output to determine a “anticipated” heartbeat. The “anticipated” heartbeat can potentially be the appropriate heartbeat for analyzing whether the heart rate has changed to the new frequency band. The heart rate detection module 424 can therefore be further configured to use the anticipated heart rate in determining heart rate 430 when the indicated band output changes, to provide faster heart rate tracking and transmission to the usable device. next to the body 108.

[0082] For example, in an exemplary implantation, the largest amplitude band can correspond to the normal band range, while the output of the second largest amplitude band can correspond to the tachycardia heart rate range. Tracking both band outputs can then provide faster tracking in the event that the heart rate changes rapidly, as can occur with certain arrhythmias, such as atrial fibrillation.

[0083] In a deployment, the heart rate detection module 424 can be configured to "qualify" the second largest amplitude band for further processing (for example, determining early heartbeat). The qualification of the bandwidth with the second greatest amplitude may include, for example, determining whether the bandwidth with the second greatest amplitude is a sufficient fraction of the bandwidth of greatest amplitude. In the case where the output with the second largest amplitude is a sufficiently small fraction (for example, <0.85), the heart rate detection module 424 can forgo further processing of the bandwidth with the second largest amplitude and the determination of anticipated heart beats in this band.

[0084] The heart rate detection module 424 can be further configured to use the band outputs described above in determining heart rate (or, likewise, anticipated heart rate). Determining heartbeat, in this sense, generally refers to determining the times at which heartbeats occur. These times can then be used to determine your heart rate.

[0085] A technique for determining the heartbeat contemplated by this disclosure uses the rate of change of the PPG signal (also called gradient of the PPG signal). Figure 5 represents an exemplary PPG signal 428 (which has a first PPG signal peak 502 and a second PPG signal peak 504) and a PPG signal gradient 506 (which has maximum gradients 508A, 508B and 508C).

[0086] In an exemplary implantation, determining the first heartbeat and a second heartbeat may include determining the first heartbeat and the second heartbeat from maximum gradients 508 of the PPG signal 526.

[0087] As shown in Figure 5, the maximum gradients 508 represent the locations of the fastest inclination of the PPG signal 528, that is, the maximum values of the gradient of the PPG signal 506. These locations can be considered heartbeats (that is, the times when the heartbeat occurs). The heart rate can then be used to determine the heart rate.

[0088] The sampling rate limitations inherent in the acquisition of the PPG 428 signal can result in the PPG signal peaks (502 and 504) and the gradient peaks (508A, B, C) shown in Figure 5 and which appear to be at a location other than the actual peak (as would be represented if there was no limit on the sampling rate).

[0089] In order to overcome the sampling resolution limitations inherent in the PPG detection process and to increase the accuracy of heart rate determinations, the heart rate detection module 424 can be configured to use various methods to determine a peak more accurate gradient. This method can be explained with reference to Figure 6. Figure 6 represents the maximum gradient point 508A in Figure 5, along with two other gradient curve points (represented as gradient T-1 PPG 604 and gradient T + 1 PPG 606) that can be used, for example, with a spline interpolation 610 to determine a more accurate gradient peak 608. Therefore, heart rate detection module 424 can be configured to determine a heart rate using the maximum gradient 508A, at minus two other data points from the PPG 428 signal gradient and a mathematical method.

[0090] In an exemplary implementation, the mathematical model can understand spline interpolation. In addition, the at least two other points can include PPG signal gradient samples 506 at immediate intervals before and after the maximum gradient 508, that is,

gradient T-1 PPG 604 and gradient T + 1 PPG 606. As shown further in Figure 6, the most accurate gradient peak 608 can be determined using the three points (gradient T-1 604, gradient T + 1 PPG 606 and maximum gradient 508A ) and spline interpolation, or other method, thereby increasing the accuracy of heart rate determinations.

[0091] In a deployment, a state-based sequence detector can be used to determine the other two data points from the gradient of the PPG signal and to implement the mathematical method. Although a state-based sequence detector is especially suitable for deploying the methods described in this document, for determining other data points and deploying the mathematical method, other deployments can be used without departing from the scope and spirit of the deployments described in this document.

[0092] The processes described above can be repeated for other intervals that contain a heartbeat in the PPG 428 signal and the time intervals between heartbeats allow heart rate 430 to be determined. Consequently, heart rate detection module 424 can be configured to use the exemplary procedures described above to determine heart rate 430 based on at least a first heart beat and a second heart beat and to make heart rate 430 be transmitted at least to the device usable next to the body 108 and / or any other component of the system 100.

[0093] In an alternative method of determining heartbeat, PPG signal peaks are used instead of gradient peaks. For example, heart rate detection module 424 can be configured to use frequency band output signals to determine a first PPG signal peak 502 and a second PPG signal peak 504 from at least a portion of the PPG signal 428 and then determine a first heartbeat and a second heartbeat from the first peak of the PPG signal 502 and the second peak of the signal PPG 504. For example, the heartbeat can be considered to be the location of the peak signal PPG or at a peak offset. The heart rate detection module 424 can then determine heart rate 430 based on the first heartbeat and the second heartbeat.

[0094] In order to overcome the sampling resolution limitations inherent in the PPG detection process and to increase the accuracy of heart rate determinations, the heart rate detection module 424 can be configured to use various methods to determine peak measurements. more accurate PPG signal. This method can be explained in a similar way to the process described above with reference to Figure 6. Similarly, the heart rate detection module 424 can be configured to determine a heart rate using the PPG 502 signal peak, at least two other data points from the PPG 428 signal and a mathematical method.

[0095] In an exemplary implementation, the mathematical model can comprise spline interpolation, but other methods can be used, as noted above.

[0096] The at least two other points from the PPG 428 signal may, in a deployment, include a positive zero crossover and a negative zero crossover closer to the PPG 502 signal peak. These zero PPG signal crossings can be determined with the use of a zero crossing narrow band filter. The narrowband zero crossing filter may be included in the frequency band filters as described above. The narrowband nature of the zero crossover filter tends to produce output waveforms that are, on a local time basis per cycle, close to the sinusoidal shape in the waveform. Since these waveforms are substantially sinusoidal, they produce zero crossings that are well behaved and reliable. The narrowband zero crossing component of the PPG signal 428 is time aligned with the PPG signal 428. In this way, the zero crossings work to cover the peaks of the PPG signal 502, 504 in the PPG signal 428. A PPG signal peak plus accuracy can then be determined using the zero crossings with the original PPG signal peak and a mathematical method, as described above.

[0097] In a deployment, a state-based sequence detector can be used to determine the other two data points of the PPG signal and to deploy the mathematical method. Although a state-based sequence detector is especially suitable for deploying the methods described in this document and for determining other data points and deploying the mathematical method, other deployments can be used without departing from the scope and spirit of the deployments described in this document.

[0098] As discussed, the PPG signal peaks, or the more accurate PPG signal peaks, can be used to identify heartbeat at peak locations or in some displacement from them. These heartbeats can then be used to determine a user's heart rate, which can, for example, be transmitted to the wearable device next to the body 108 and displayed on the same or any other component of the system 100.

[0099] As described throughout this disclosure, a user can monitor his cardiac activity with a device usable close to the body (for example, a smart watch that includes a PPG sensor). System 100 can then be configured to alert the user, a healthcare provider, etc., when an irregular heart rate type is detected by analyzing the received cardiac data. In one example, an alert can be provided when atrial fibrillation (AF) is detected.

[0100] Figure 7 is a diagram that illustrates an exemplary basic process for determining the presence of atrial fibrillation according to certain aspects of the present disclosure. This determination can be made by executing one or more algorithms that analyze the PPG sensor data. In some deployments, a computer can, in 710, receive photoplethysmographic signal (PPG) data communicated by a PPG sensor from a device usable close to the body worn by a user. As discussed in this document, the algorithm can, in 720, determine heart rate from at least a portion of the PPG signal data and can, in 730, determine a type of heart rate based on at least heart rate. As discussed in detail below, the system can also, in 740, determine whether the type of heart rate includes atrial fibrillation (AF). Operations can also include, at 750, displaying, when AF is detected, an AF detection alert on the wearable device next to the body. Details of exemplary algorithms and analyzes to detect and identify various types of heart rhythm (for example, normal sinus rhythm or atrial fibrillation (AF)) are discussed below.

[0101] Certain exemplary systems, methods and software contemplated in this document can analyze and detect cardiac rhythms using a Poincaré space, a two-dimensional space that can be formed and visualized by tracing a current heart rate interval along an axis and the previous interval along the other axis. For normal cardiac rhythms, these points tend to be densely and narrowly distributed around the local mid-range. For AF, these points tend to be distributed more widely, with large jumps at intervals occurring with successive heartbeats.

[0102] In certain exemplary methods, heart rate determination can use metrics based on that space. For example, a metric based on the occupation of space (for example, how much space is occupied by the plotted points) and / or a metric based on the median of the distances measured from point to point. Other metrics can also be used, for example, a metric based on the variability of heartbeat intervals. These metrics can be calculated in a moving window of contiguous heartbeats (for example, 48 heartbeats), to balance a local time focus with the stability of the metrics. These metrics can be calculated and updated with each heartbeat, compared to the individual limits, and then a specific pattern can be declared if the metrics exceed certain limits. To improve the accuracy of certain heart rate determinations, a specific heart rate pattern can only be declared formally when enough patterns are declared in a local heartbeat group.

[0103] Figure 8 is a diagram illustrating an exemplary 810 scatter plot representing a normal heart rate according to certain aspects of the present disclosure. Scatter plot 810 contains 820 points that are determined to represent changes in the heart rate range over a period of time (or number of pulses, heart rate ranges, etc.). Dots 820 can represent a change between two adjacent (or local) heartbeat intervals. As used in this document, “adjacent” means that one heartbeat interval may be preceded or subsequent to the other heartbeat interval.

[0104] Although the description of such points 820 is made with reference to Figure 8, it is understood that the description applies equally to references to similar points in the present disclosure (for example, as in Figures 9, 10 and 11). The determination of an example point is further illustrated by the upper portion of Figure 8, where heartbeat 1 and heartbeat 2 can define heartbeat interval 1. Similarly, heartbeat 2 and heartbeat 3 can defining heart rate range 2 and, together, heart rate range 1 and heart rate range 2 can provide the X and Y coordinates for point 1.

[0105] Points 820 can be implanted, for example, as values stored in the computer's memory in the form of variables, arrays, vectors, matrices, objects or other data structures. Dots 820 can be expressed graphically or merely represented in the computer's memory. The points can be formatted, for example, to represent an absolute change in the heart rate range (for example, 0.5 s, at 0.1 s, etc.) or a ratio (for example, 0.9 - when a heart rate range is 90% of the adjacent heart rate range). Other expressions of 820 points that represent changes in heartbeat intervals over a period of time are also contemplated and considered within the scope of this disclosure.

[0106] In the additional description of the illustrative Figure 8, the horizontal axis of the 810 scatterplot refers to a heartbeat interval (for example, n interval) and the vertical axis refers to a second interval (for example, an interval preceding n-1). When a heart exhibits a normal heart rate type, the 820 points may not change significantly over time. Consequently, an 820 dot plot for a normal heart rate will generally be located in a limited region (as shown in Figure 8). In the example shown, points 820 were normalized by the average heart rate range (for example, for a heart rate range n that is 0.6 s and the heart rate range n-1 that is 0.4 s, the point would fall at point 1 on the example scatter plot). Therefore, as shown in Figure 8, when a heart rate interval increases, those 820 points will be below the 45 degree 830 line. When the heart rate interval decreases, these 820 points will be above the 45 degree 830 line.

[0107] Figure 9 is a diagram illustrating an exemplary movable window 910 based on the PPG signal data used in generating the exemplary scatter plot 810 in accordance with certain aspects of the present disclosure. The points 820 used for determining a type of heart rate can be determined for the movable window 910 (for example, the time period) containing some number of heartbeats or heartbeat intervals, etc. Consequently, the points used to generate the exemplary scatter plots and / or the points used to determine the metrics discussed here can be based on a 910 movable window. An exemplary illustration of a 910 movable window is referenced in the 920 range graph in the portion top of Figure 9. The movable window 910 contains a series of heartbeats or heartbeat intervals, including, in this example, the current heartbeat and the current heartbeat interval 930. Although discussed in relation to Figures 8 and 9, the movable window 910 can also be used to generate any of the scatter plots or data sets described here.

[0108] The movable window 910 can be designed to contain a predefined number of heartbeats before and / or including the current heartbeat. For example, in some deployments, the movable window 910 can include 50 heartbeats, 100 heartbeats, 192 heartbeats, 500 heartbeats, etc. In some deployments, a specific number of heartbeats can be identified as providing an optimization between recent heart rate data and accuracy in determining the type of heart rate. For example, 48 heartbeats can be a specific number.

[0109] Similarly, in some deployments, the movable window 910 can be based on a predefined or optimized time period (for example, 15 seconds, 30 seconds, one minute, five minutes, etc.). In other deployments, the movable window 910 can be based on a predefined number of heartbeat intervals (e.g., 10 heartbeat intervals, 20 heartbeat intervals, 48 heartbeat intervals, 100 heartbeat intervals, etc.). ) or can be designed to change the scope based on certain factors.

[0110] Figure 10 is a diagram illustrating an exemplary method of determining an occupancy metric from an exemplary 1010 scatter plot that represents an irregular heart rate in accordance with certain aspects of the present disclosure. The computational algorithms described in this document can generate quantitative metrics that can be used to accurately determine types of heart rate, such as AF. Comparing Figure 10 with Figures 8 or 9, it can be seen that the distribution of points 1020 in Figure 10 is greater. Therefore, in some deployments, the algorithm can determine an occupancy metric based, at least partially, on an interval dispersion analysis. Subsequently, and as discussed in more detail below, determining the type of heart rate can also include the use of the occupancy metric.

[0111] As used in this document, the term "interval dispersion analysis" means an analysis, performed by a computer, of changes in heartbeat intervals. Changes in heart rate intervals can be expressed or visualized, for example, points in the scatter plots in Figures 8, 9, 10, etc. Changes in heart rate intervals can also be represented in the computer's memory in the form of arrays, tables, etc., or calculated from heart rate data. In these deployments, interval dispersion analysis can optionally be performed without plot points on scatter plots. An interval dispersion analysis may include quantifying some characteristics of the distribution of changes in the heartbeat interval. For example, as seen in the example scatter plots. The characteristics (or metrics) can include an occupation associated with the points in the scatter plots, as described later in this document.

[0112] As used in this document, an “occupancy metric” means a metric (for example, a number, vector or some value) that quantifies the distribution of points in a defined space - that is, it provides a measure of the degree to which the space is “occupied” by one or more points. The space can be a scatterplot, an arrangement, a table or a graph, etc.

[0113] In an exemplary determination of an occupancy metric, scatter plot 1010 in Figure 10 can be subdivided into a series of 1030 areas. Areas 1030 that contain no points are shown with shading (for example, area 1040) . In the example in Figure 10, there are 36 areas 1030 shown. Due to the distribution of points in the 1010 scatterplot, 16 of the 1030 areas contain at least one point

1020. Consequently, in this example, the occupation of that 1020 point distribution is 16/36 = 0.444. As a comparison, although not shown, the occupancy metric of the 810 scatterplot in Figure 8 or 9 would be 6/36 = 0.166. This exemplary method, therefore, demonstrates that an occupancy metric can be calculated to result in a higher value when there is a greater variation in heartbeat intervals (and therefore a greater likelihood of PA). It is understood, however, that occupancy metrics can be calculated and used in alternative ways.

[0114] Although the example illustrated in Figure 10 is discussed in terms of areas of a scatter plot, computational solutions are contemplated that perform a similar analysis based, for example, on compartmentalization, partitioning, classification or similar. Some deployments of the current story may include defining compartments in which the points may fall. The occupancy metric can be determined by determining a fraction of compartments that contain at least one of the points.

[0115] As used in this document, a “compartment” means a section, area, volume or data structure that can potentially include one or more points. For example, a compartment, as performed by computer software, can be a specific area on a scatter plot (as shown in the example below), a portion of a data structure corresponding to the points, etc. Therefore, deployments can allow for such compartmentalization, and the like, using arrays, arrays, objects or other forms of partitioning of computer data to represent the heart rate intervals and areas 1030 illustrated in Figure 10. For example, an arrangement 1020 points can be partitioned or compartmented in the computer's memory according to where they would fall in a defined area (or range) representing a degree of change between adjacent intervals.

[0116] Figure 11 is a diagram illustrating the determination of an exemplary distance metric from an exemplary 1010 scatter plot that represents an irregular heart rate according to certain aspects of the present disclosure. Similar to the occupancy metric, an algorithm can determine a distance metric based, at least partially, on points that represent changes in heart rate intervals, which can be plotted on a scatter plot as shown in Figure 11. As before, determining the type of heart rate can also include using the distance metric.

[0117] As used in this document, a "distance metric" provides a measure of the degree of change in heartbeat intervals, usually calculated in terms of a "distance" between points that represent the change in heartbeat intervals. For example, a distance metric can be based on distances calculated between points in the example scatter plots (for example, Figure 11). A long distance metric indicates large changes in the heart rate range (possibly indicative of an irregular heart rate) and a short distance metric indicates small changes in the heart rate range (possibly indicative of a regular heart rate).

[0118] In the discussion below regarding the distance metric, the example in Figure 11 uses the same 1010 scatter plot and 1020 points shown in Figure 10. Similar to what is shown in Figure 8, the upper portion of Figure 11 it also contains an exemplary illustration of which points, intervals and heartbeat can be included in determining a distance metric.

[0119] In some deployments, the distance metric can be determined from a median of the distances between the points (for example, 1020 points). Similarly, the distance metric can be determined from an average of the distances between points. In the example in Figure 11, a first distance (Distance 1) 1120 can be calculated between consecutive points 1120 as shown. Similarly, a second distance (Distance 2) 1130 can be calculated between two other points 1130. Since the points used to calculate the distance metric in this example are consecutive, Distance 2 1130 is calculated with a common point - point 2 - with Distance 1 1120. The distance metric can then match the median of any number of those distances, for example, any or all 1020 points shown in the 1010 scatterplot or the points included within a specific moving window.

[0120] An interval variability metric can also be determined based on heartbeat intervals that are derived from a series of heartbeats, for example, the heartbeat included in a moving window. In some implantations of current material, determining the type of heart rate may also include the use of this interval variability metric.

[0121] As used in this document, a “range variability metric” represents a measure of the range of heartbeat intervals for any number of heartbeats. In one example, the range variability metric can be the standard deviation of the examined ranges.

[0122] In another example, the interval variability metric may involve examining changes in heartbeat intervals from one heartbeat to the next. For example, an interval defined by a first and a second heartbeat (for example, Interval 1 = 1,000 ms) can be compared to an interval of heartbeat defined by a second and third heartbeat (for example, Interval 2 = 900 ms). The change from interval 1 to interval 2 would then be one hundred milliseconds. These heartbeat to heartbeat interval changes can then be calculated over a period of time. The interval variability metric, then, in one example, can be a median of the absolute changes in heartbeat intervals. This example method can be illustrated with the following equation: 𝑀é𝑡𝑟𝑖𝑐𝑎 𝑑𝑒 𝑉𝑎𝑟𝑖𝑎𝑏𝑖𝑙𝑖𝑑𝑎𝑑𝑒 𝑑𝑒 𝐼𝑛𝑡𝑒𝑟𝑣𝑎𝑙𝑜 = 𝑚𝑒𝑑𝑖𝑎𝑛𝑎 (𝐴𝐵𝑆 (∆ (𝑖𝑛𝑡𝑒𝑟𝑣𝑎𝑙𝑜))) (1)

[0123] In another deployment, an interval variability metric can be calculated based on at least one current heart rate range 1210 and an average heart rate range 1230 (as shown in Figure 12). The heart rate intervals can be contained in the movable window 1220 (for example, 192 beats) or can be any other set of intervals desired. In the exemplary selection of heartbeat intervals shown in Figure 12, the average heartbeat interval 1230 for the movable window 1220 is illustrated by a horizontal line that extends on the right side of Figure 12. In some deployments, a Euclidean distance between each beat pair can be determined and a median of these distances can be used in this determination. In this example, the current heart rate range 1210 is shown by the longest range on the right side of the moving window 1220. Therefore, an example range variability metric would be quite large because the current heart rate range 1210 is significantly different from the range average heart rate.

[0124] In another deployment, the range variability metrics calculated above can be normalized by the number of heartbeat intervals (for example, the size of the array). In still other deployments, the range variability metric can be normalized to the average range. There can be any number of equivalent ways of calculating or normalizing the interval variability metric, this being explicitly contemplated in this disclosure.

[0125] Figure 13 is a diagram illustrating a three-dimensional space 1310 and discrimination plans (1330, 1340, 1350) that can be used in an exemplary method to determine a type of heart rate according to certain aspects of the present disclosure. The determination of the type of heart rate can include the use of the occupation metric, the distance metric and the interval variability metric, in relation to a three-dimensional space.

[0126] In certain deployments, the type of heart rate can be determined based on any combination of the above metrics. For example, the type of heart rate can be determined only based on the occupation metric, based only on the distance metric, based only on the interval variability metric, based on all three of the occupation metric, the distance metric and interval variability metric, or any combination thereof.

[0127] Figure 13 illustrates an exemplary implantation in which the determination of a type of heart rate is based on at least all three of an occupancy metric, a distance metric and an interval variability metric. In the illustrated deployment, points 1320 can be plotted in the three-dimensional space 1310 defined by at least these three metrics.

[0128] In this three-dimensional space, certain regions can be classified as corresponding to particular types of heart rhythm outlined by one or more discrimination plans (1330, 1340, 1350). For example, referring to the NSR - AF 1330 plane represented in Figure 13, when a point is on one side of the NSR - AF 1330 plane, that point 1320 can be classified as referring to a heartbeat that is classified as NSR (normal sinus rhythm). On the other hand, when a point is on the other side of the NSR - AF 1330 plane, that point can be classified as referring to the heartbeat which is a type of AF rhythm.

[0129] As shown in Figure 13, in general, a regular heart rate would have 1320 points grouped towards the origin of three-dimensional space (that is, with low occupation, distance and variability). Locations more distant from the origin may indicate some kind of irregular heart rhythm. The different exemplary symbols represented in Figure 13 indicate the types of heart rate determined for a given point. For example, open triangles refer to the heart rate types that are NSR, open squares refer to the heart rate types that are AF and the solid circles refer to the heart rate types that are NAF (that is, not AF) ).

[0130] Discrimination plans (1330, 1340, 1350) can be calculated automatically based on algorithms to define these plans. In other deployments, the discrimination plans (1330, 1340, 1350) can be determined from the computational optimization of plan parameters (for example, terms or coefficients in equations that define the discrimination plans (1330, 1340, 1350)) annotated data. For example, it can be established that certain collections of 1320 points correspond to a particular type of heart rhythm. The data can be annotated (for example, by a doctor who adds tags, markers or other metadata to points in the computer's memory) to specify the type of heart rate. So, the discrimination plans (1330, 1340, 1350)

they can be computationally determined so that they better discriminate points classified 1320. Similarly, these determinations can be made with machine learning.

[0131] Figure 14 is a diagram illustrating a decision tree 1410 for an exemplary method for determining a type of heart rate in accordance with certain aspects of the present disclosure. Since a given 1320 point can meet several criteria, as defined by the discrimination plans, other rules can be implemented to determine the type of heart rate. In general, implementing rules to classify that a point exhibits a specific type of heart rate may include determining, first, which side of a foreground discrimination the point is on, and then determining which side of a second discrimination plan the point is. For example, point 1320 can be on the NSR side of the NAF - NSR 1350 discrimination plan and can also be on the AF side of the NSR - AF 1330 discrimination plan. The exemplary decision tree 1440 shown in Figure 14 illustrates a deployment to establish a type of heart rate based on the above considerations. Decision tree 1410 starts at the NAF-NSR 1420 decision branch by first determining which side of the NAF-NSR 1350 plane the point is on. So, if the point was on the NSR side, a second determination is made in the decision branch NSR-AF 1430 to find out which side of the NSR-AF 1330 discrimination plan that point 1320 is on. When point 1320 is on the AF side, the heart rate type for that point 1320 is classified as AF. When point 1320 is on the NSR side, the type of heart rate for that point 1320 is classified as NSR.

[0132] Back to the NAF-NSR 1420 decision branch, when the point is on the NAF side of the NAF-NSR 1350 discrimination plan, a second determination can be made on the NF-AF 1440 decision branch if the point is classified as NAF or AF based on which side of the NAF-AF 1340 discrimination plan he is on.

[0133] This disclosure explicitly contemplates other combinations of decision branches to classify a point as AF, NSR or NAF. Similarly, other rules can be added to add conditions necessary for a specific classification, for example, heartbeats (and points) that have been marked as “noisy” are not used as part of a data set that serves as a basis for generate an AF alert.

[0134] In some deployments, the presence of a specific type of heart rate can be established when a sufficient number of heart rate assessments are determined based on a contiguous local heart rate group. In other words, the establishment of a specific type of heart rate can be based on a) a sufficient number of classification determinations and b) the need for the heart rate intervals that establish a particular type of heart rate to be contiguous. The term “contiguous” means that there is no need for an invalid heartbeat in the heartbeat intervals. An invalid heartbeat can, for example, be determined based on the beat detection algorithm described here or obtained during a period of noisy data (as detected by an accelerometer, another sensor, noise detection algorithm, etc.).

[0135] Certain deployments of the current story may include different methods of classifying a given point as having a specific type of heart rate. For example, instead of using discrimination plans, some implementations may include an algorithm that implements a series of conditional statements, limits, decision trees or other logical structures and equivalent flows to generate a one-point classification. For example, an implementation of this conditional statement may be AF = (occupancy metric greater than 0.1, distance metric greater than 20, metric for interval variability greater than 0.2). The previous example of a conditional statement implies three plans (or limits) of orthogonal discrimination. In other deployments, the discrimination plans need not be strictly flat, but can instead be a surface that can contain any degree of local curvature or other shape. In general, the present disclosure contemplates a general partition of a space (of arbitrary size) in regions that can, at least in part, determine a specific type of heart rate for a point in a specific region. These regions can be one-dimensional (for example, occupation metric, distance metric, or range variability metric greater than a predetermined value), two-dimensional (for example, an area specified in a two-dimensional space defined by the occupation metric and the metric of distance), three-dimensional (for example, as illustrated by the example in Figure 13), or with more dimensions (for example, when additional rules and metrics are implemented to create an arbitrary number of factors that determine the type of heart rate).

[0136] Some deployments of the current subject may include the generation of an output (for example, a visualization) corresponding to the points and graphs mentioned above and may include, for example, any combination of the occupation metric, the distance metric, the metric of interval variability, of an average heartbeat interval in a moving window, of a first heartbeat count to calculate the occupancy metric in the mobile window (optionally limited to qualified heartbeats), of a second heartbeat count for calculate the distance metric in the movable window (optionally limited to qualified heartbeat), a third heartbeat count to calculate the interval variability metric in the mobile window (optionally limited to qualified heartbeat), etc. This output can be generated and provided, for example, to a user or healthcare provider through an output device, such as a smart phone, smart watch, personal computer or the like. The output can be provided in the form of an array or a matrix (of arbitrary size), a data file, a graphical display, such as a graph (as illustrated by Figures 8 to 11) or a graph, three-dimensional view (as shown) Figure 13), etc.

[0137] When an alert (for example, indicating the presence of a particular type of heart rate, such as AF), these alerts can be sent and / or generated on any number or types of devices. Alerts can be sent to a wearable device, a computer or server, a smart phone (for example, when used by a doctor, a user, a caregiver, etc.) or the like. Alerts can be deployed as emails or text messages or they can be audible, graphic, sensory (for example, vibration), etc.

[0138] As described in this document, certain deployments of the current story may include providing an alert to a user based on the detection of a particular type of heart rate (for example, AF). However, in some deployments, the display of an AF detection alert can occur when AF is determined based on a number of heartbeats greater than an AF detection alert threshold. For example, in theory, three heartbeats may be sufficient to determine AF based on some of the implantations described here. However, it may be advantageous to rely on a larger set of data to obtain a more accurate determination of a particular type of heart rate. For example, in some deployments, the minimum number of heartbeats (or points) needed before generating an alert (for example, an AF detection alert threshold) can be set to a predetermined number, such as 50, 100, 200 , etc. In some deployments, this alert detection limit can be customizable (for example, by a user, a doctor or other healthcare professionals, etc., through a user interface as described in this document), so that a a single beat of AF may cause an alarm for one user, but 50 beats of AF, 100 beats of AF, etc., may be required for other users. In some deployments, the AF detection limit can be related to the amount of time a user experiences the AF beats and / or other information related to AF.

[0139] In other deployments of the current story, the system may require an (electronic) authorization before providing an AF detection alert to the user. In some deployments, this authorization may be provided electronically by an authorizing agent, such as a physician, technician, manufacturer, or other authorized person. The authorization can be stored, for example, as part of a data file on any computing device in communication with the usable device next to a user's body. The authorization may also include an association between authorization and the identification of a specific patient. For example, authorization may indicate that the authorizing agent has authorized a particular user (with a patient ID, such as a number, code, or the like) to receive alerts (for example, an AF detection alert). Correspondingly, this authorization can allow the device that can be used close to the body to generate an alert only when that authorization exists (as verified in the device that can be used close to the body or any other computer system in communication with the device that can be used close to the body). Therefore, these deployments may include receiving a patient identification communicated by the wearable device to the body and generating the AF detection alert when an authorization is received that indicates a permission to display AF detection alerts to the user.

[0140] As described above, portions of system 100 can receive PPG signal data communicated by a PPG sensor from a device usable close to the body worn by a user. The PPG signal data can then be used to derive information such as heart rate, heart rate, heart rate variability and the like. The PPG signal data can also be analyzed by a computer program to determine the heart rate intervals of at least the PPG signal data. A heartbeat interval can be expressed in terms of time (for example, 500 ms), although other expressions can also be used, such as frequency, beats per minute (BPM), etc.

[0141] It may be beneficial to provide systems, methods and graphical user interfaces implemented by computer programs that allow the user to select a desired set of PPG signal data for review or analysis. An example of this implementation may include a “patient panel” that allows the user to select, view, download, etc., the PPG signal or related data. Figure 15 is a diagram illustrating an exemplary patient panel 1510 in accordance with certain aspects of the present disclosure. The 1510 patient panel can be provided as part of a web portal, as part of standalone software or in other ways.

[0142] The 1510 patient panel can include user information such as name, height, weight, minimum heart rate, maximum heart rate, etc. Patient panel 1510 can include electronic medical record information (for example, received from an electronic medical record server). This information can be correlated with information collected during the 1520 data collection time periods. The patient panel 1510 can also include graphical displays of the 1520 data collection time periods, as shown in Figure

15.

[0143] 1520 data collection time periods can be and / or illustrate continuous data collection time periods. Individual data collection periods 1520 can be subsets of larger data collection periods and / or other time periods. In various deployments, the 1520 data collection time periods may represent time periods in the order of seconds in duration, minutes in duration and / or hours in duration. For example, 1520 data collection time periods can range from about 10 seconds in length to about eight hours in length. As another example, a first data collection time period 1520 can be about ten seconds long and be a subset of a second data collection time period which is about eight hours long.

[0144] The 1520 data collection time periods, as shown, can provide more details of the acquired PPG signal data, for example, date and time of collection, minimum and maximum heart rate, types of heart rates observed (for example , normal, AF, etc.), and the like. Data collection time periods 1520 can include a graphical indication 1530 if an AF rhythm is detected. Other deployments may include providing color coding for some part of the graphical interface, for example, highlighting displayed portions of PPG signal data in red, or some other color, to distinguish 1520 data collection time periods when AF have been detected. It should be noted that AF rhythms are used as examples. The detection of other cardiac rhythms (for example, apnea, premature atrial contraction, premature ventricular contraction, supraventricular tachycardia (for example, atrial vibration)) are contemplated.

[0145] Patient panel 1510 can also provide a means for a user to download PPG signal data corresponding to one or more 1520 data collection periods, for example, by pressing a download button on the panel of patient 1510. Panel 1510 can allow a user to similarly select a certain period of data collection time 1520, in order to view more details of the PPG signal and data analysis, for example, by pressing a button. view or by selecting a hyperlink embedded in the graphical display. In a deployment, additional details and analysis can be presented in a graphical user interface called a PPG viewer. Although the exemplary term “PPG viewer” is used in this document, it is understood that this interface can represent or provide details relating not only to PPG signals, but to additional signals, analysis results, etc. For example, additional signals may include signals related to a percentage of VO 2, accelerometer information, electrodermal activity (e.g., a galvanic skin response from an associated sensor), electromyography (EMG) data, electroencephalogram data ( EEG) and / or other sensor information that is included in the system (for example, in a wristwatch range) and / or that is configured to interact with the system.

[0146] Figure 16 is a diagram illustrating an exemplary PPG viewer 1610 in accordance with certain aspects of the present disclosure. The PPG 1610 viewer can be a graphical display (for example, a web page, etc.) that can include visualizations (for example, graphics, etc.) based, at least partially, on the PPG signal data over some time interval, for example, 1520 data collection periods as shown in Figure 15. The PPG 1610 viewer can be displayed on a display device, for example, a computer monitor, a smart phone screen, etc.

[0147] In some deployments, the PPG 1610 viewer may include, for example, graphical displays of accelerometer data (XL) 1620, PPG signal data (PPG) 1630, heart rate data (HR) 1640 and interval data heart rate (Intv) 1650. Accelerometer data 1620 can be acquired from an accelerometer that is part of the device that contains the PPG sensor (or sensors) or that is separate from that device. Heart rate data 1640 and heart rate interval data 1650 can be calculated, for example, using methods such as those described above. These types of data can be displayed with a common time base, as shown in Figure 16.

[0148] In the deployment shown in Figure 16, the PPG 1610 viewer can include an expanded or expanded graph 1660 for a specified period of time. The time period can be determined automatically or can be controlled by user input via an input device. For example, a user can select a start time 1670 and an end time 1680 (displayed graphically in Figure 16 by two vertical lines that extend vertically over the XL, PPG, HR and Intv charts). In response to such a selection, the display, for example, of PPG 1630 signal data can be automatically adjusted on the expanded graph 1660.

[0149] As described in this document, certain deployments of the present disclosure may include determining when heartbeats occur based on PPG signal data. The additional functionality of the PPG 1610 viewer can then include the display of heart rate markers 1690 along with PPG signal data 1630. For example, as shown in Figure 16, heart rate markers 1690 can be displayed as short vertical lines in the 1630 PPG signal data. In other deployments, the 1690 heartbeat markers may also include marking beats that have been determined to be noise. As shown in Figure 16, these 1690 heart rate markers can be indicated by an “X” (or any other graphic indication). In some deployments, 1690 heart rate markers can be configured to distinguish between an AF beat and other types of beats (for example, beats indicating normal sinus rhythm, beats indicating brachycardia, beats indicating tachycardia and / or other types beats). For example, 1690 heartbeat markers may have different colors, may comprise different background colors in a corresponding location on a screen, may comprise a lettered code (for example, a two-letter code) that indicates a type of heartbeat and / or be formed by any other type of visual indication that distinguishes the AF beats from other types of beats.

[0150] In some deployments, viewer 1610 can be configured so that data from the PPG 1630 signal can be displayed with markers 1690 and / or markers 1690 can be displayed by themselves, without 1630 data. Viewer 1610 can be configured to display a stabilized version of the 1630 data, for example, showing a beat time with a (horizontal) line of beat markers 1690 (removing the increasing / decreasing nature of typical markers that correspond to a typical PPG signal).

[0151] In some deployments, the 1610 viewer can facilitate the enlargement of a selected data range. For example, the 1610 viewer can facilitate the enlargement of a second range of data from a larger range (for example, ten seconds, one minute or more) of data. This magnification can cause corresponding changes in the corresponding data graphs (for example, see Figure 8).

[0152] Figure 17 is another example of the PPG viewer 1610 that includes different information in the expanded graph area 1660, for example, a waveform that resembles at least partially an ECG signal.

[0153] ECG signals can include PQRST waveform features, where each letter corresponds to a specific feature present in a typical heartbeat. The “R wave” portion is a peak most commonly associated with heartbeat time. The R wave can be bounded by a "Q wave" and an "S wave" to form what can be called a QRS complex (or QRS wave). The QRS complex is a group of waves (or a waveform) that represents ventricular depolarization. The QRS complex can include three distinct waves created by the passage of the cardiac electrical impulse through the ventricles and can occur at the beginning of each ventricular contraction. In some surface electrocardiograms, the R wave is the upward deflection, the first downward deflection represents a Q wave, and the final downward deflection is the S wave. The Q and S waves can be weak and sometimes absent. After the QRS complex, there may be a T wave that represents the repolarization (or recovery) of the ventricles. Before the QRS complex, there may be a P wave, which can be a positive deflection on a normal surface electrocardiogram produced by an excitation of the atria. The P wave can represent atrial depolarization, an intrinsic atrial event.

[0154] An implementation of the present disclosure may include a graphical output in the PPG 1610 viewer using PPG signal data, but which creates a graph that looks similar to ECG data. For example, as shown in Figure 17, certain implantations may include a 1720 electrocardiogram (ECG) waveform, with the 1720 ECG waveform spanning a heartbeat interval (range that can be determined from the PPG signal data). For example, in some implantations, the ECG 1720 waveform may include a PQRST waveform that includes features that represent electrical activity in the heart (as shown in Figure 17).

[0155] As used in this document, an "ECG type waveform" means a waveform with at least some characteristics that are indicative or suggestive of a typical ECG waveform. Consequently, ECG-type waveforms may be similar to an ECG, but they do not need to contain all the features normally found or described on an ECG. For example, as described below, an ECG waveform used in this document may include suppression of a P wave, to serve as an indication that an AF rhythm has been detected. As used herein, an ECG waveform can also include modifications to particular aspects of a PQRST waveform.

[0156] ECG waveforms can be based on a real detected ECG or can be “synthetic”. As used in this document, “synthetic” means a waveform that is artificially generated to look similar to an ECG, but not necessarily actual ECG data. For example, a standard ECG waveform can be stored in the computer's memory and retrieved to be displayed to cover heart rate intervals.

[0157] ECG waveforms can include any combination of P, Q, R, S and T waves. In one example, an ECG type waveform can include the display of an RSTPQR waveform scaled between two successive heartbeats. Here, the first R wave can be in the first heart beat and the second R wave (which can be a copy of the first R wave) can be in the second heart beat. The S, T, P and Q waves (and portions of the R waves) can then be scaled correspondingly to cover the heart rate range.

[0158] A repeated PQRST waveform can be similar to a repeated RSTPQR waveform, one difference being that the PQRST waveform is somewhat centered on a heartbeat (for example, the R wave), with scale appropriate applied to each side of the heartbeat. As shown in Figure 17, these ECG 1720 waveforms can be repeated any number of times, with the corresponding waveforms sized appropriately between heartbeats.

[0159] As used herein, when describing ECG waveforms as “spanning” a heartbeat interval, it means that the ECG waveform is graphically scaled to at least one dimension to cover or extend over a graphical indication of a heartbeat interval. It is contemplated that waveforms can be configured to cover any number of points within a heartbeat signal (for example, from R to R, P to P, P to T, PPG peak signal to PPG signal from peak, peak PPG gradient to peak PPG gradient, etc.). In addition, since heartbeat intervals may vary, ECG-type waveforms can be individually scaled to cover the corresponding heartbeat intervals.

[0160] In some deployments, an 1820 ECG waveform can be displayed without one or more features that may be present on an ECG. For example, since ECG signals during atrial fibrillation often exhibit little or no P waves, some implementations of the present disclosure may purposely suppress or eliminate the P 1730 wave from an ECG 1820 waveform when AF is detected, as shown in the example in Figure 18.

[0161] In some deployments, the ECG 1820 waveform may typically include a P wave, but the system may include detecting, based on at least PPG 1630 signal data, whether atrial fibrillation is represented in the signal data PPG 1630. The software can then suppress the P 1730 wave displayed in the ECG 1820 type waveform when atrial fibrillation is detected. As used in this document, when referring to a portion of the ECG type waveform as "suppressed", this may include the portion (for example, P 1730 wave) that is defined as a baseline value of the waveform type ECG 1820, replaced by a line segment that connects points where the part should be, reducing the amplitude of the portion or, otherwise, scaling the portion to make it appear smaller when displayed.

[0162] In another implantation, instead of modifying the P 1730 wave based on the detection of an AF rhythm, the ECG 1820 waveform may, by default, represent an AF rhythm and, therefore, not include the P wave

1730. This deployment may include detecting, based on at least the PPG signal data, whether an AF rhythm is represented in the PPG 1630 signal data. So, for example, the PPG 1610 viewer can display the P 1730 wave as part of the ECG 1720 waveform when an AF rhythm is not detected.

[0163] Computational analysis of PPG 1630 signal data, as described in this document, can benefit the user or a caregiver, allowing diagnostic indicators of an AF rhythm to be readily displayed in a familiar ECG-type waveform (1720 , 1820), but based on PPG signal data. These operations can have beneficial results, such as improved medical diagnosis of an AF rhythm. Therefore, the various deployments of the PPG 1610 viewer described here are not a routine or conventional graphical display of PPG 1630 signal data or ECG data, but hybrid deployments that use some features of both. In addition, the present disclosure provides specific examples of tangible machine operations (for example, sizing an ECG-type waveform (1720, 1820) to cover PPG-derived heartbeat intervals) that are performed with these deployments.

[0164] Figure 19 is a diagram illustrating an exemplary graphical interface that includes a 1920 dot 1910 scatter plot that represents changes in heart rate intervals according to certain aspects of the present disclosure.

[0165] The 1920 dot distribution on the 1910 scatter plot can assist a user in determining the presence of an AF rhythm (as discussed above with reference to Figures 8 to 11). Similar to other deployments described herein, the vertical axis can correspond to a heartbeat interval in the first moment and the horizontal axis can correspond to a heartbeat interval in a second time. For example, the second time may be the interval immediately after the first interval. In other words, the horizontal axis can display the i-th interval, while the vertical axis corresponds to the i-th interval - 1. The present disclosure also contemplates other methods of displaying heartbeat intervals, for example, the corresponding vertical axis at the i-th interval + 1. In general, the 1910 scatterplot can display a 1920 point determined by two consecutive heartbeat intervals.

[0166] Deployments may include, for example, calculating, based on at least PPG 1630 signal data, a first heart rate interval. In addition, some deployments may include calculating, based at least on the PPG 1630 signal data, of a second heartbeat interval, with the second heartbeat interval occurring after the first heartbeat interval. In the graphical interface, the 1910 scatterplot can be generated to represent a variation of heartbeat intervals and to include a graphic in a location determined by at least the first heartbeat interval and the second heartbeat interval. The graphic element can be, for example, points (as shown in Figure 19), or any other symbol, character or visual representation of the location corresponding to the point 1920.

[0167] As contemplated by this disclosure, graphic displays may display, in any combination or arrangement in the graphic display, any of the graphics described here, including the graphic interface that displays ECG waveforms (1720, 1820) and the 1910 scatter plot. In such deployments, as the start time 1670 and / or the end time 1680 are changed, in addition to the PPG signal data 1630 being rendered again in the expanded PPG graph 1660, the display of the 1920 points can also automatically be updated on the 1910 scatterplot to reflect the intervals that fall within the current adjustable time window.

[0168] Figure 20 is a diagram illustrating an exemplary method of generating an ECG type waveform according to certain aspects of the present disclosure. The computer program systems, methods and products contemplated in this document may include some or all of the following features, in any combination. In

2010, PPG 1630 signal data, communicated by the PPG sensor of a device usable close to the body worn by a user, can be received. In 2020, a heartbeat interval can be determined from at least PPG 1630 signal data. In 2030, the 1720 electrocardiogram (ECG) waveform can be generated on a graphical interface, the waveform being ECG type 1720 covers the heart rate range.

[0169] Although some modalities have been described in detail above, other modifications are possible. For example, the steps of the method depicted in Figure 7 and described in this document do not require the particular order shown, or sequential order, to achieve the desired results.

[0170] Current subject implementations may include, for example, performing calculations (for example, determining heartbeat, heartbeat intervals, processing signal data, scaling waveforms, etc.) or displaying any one graphics resources (for example, PPG charts, ECG waveforms, scatter charts, etc.) as described in this document, in any combination of computing systems and any combination of programmable processors distributed between computing systems. Such computing systems may include smart phones, smart watches, personal health monitoring devices, personal computers, laptop or tablet computers, cloud-based servers and network environments or the like. Thus, the present disclosure contemplates technical solutions that can be implanted in a single device, for example, entirely by a smart watch or personal health monitor. Other deployments of the technical solutions described here can be distributed on various devices, for example, which acquire sensor data from a sensor in contact with a user, and which transmit the sensor data from the detection device to a remote computer, such as a server via a smart phone, etc. The server can execute instructions to perform certain calculations on the data and then transmit data and / or commands to one or more target devices to cause the target devices to display a specific graphical output, generate an alert, etc.

[0171] The deployments of the present matter may include, but are not limited to, methods consistent with the descriptions provided in this document, as well as articles comprising a machine-readable medium tangibly incorporated that has the purpose of making one or more machines (for example, computers, etc.) result in operations that implement one or more of the described resources. Similarly, computer systems that may include one or more processors and one or more memories coupled with one or more processors are also contemplated. A memory, which may include a computer-readable storage medium, may include, encode, store, or the like, one or more programs that cause one or more processors to perform one or more of the operations described herein. Computer-implemented methods consistent with one or more deployments of the present subject can be deployed by one or more data processors that reside in a single computing system or in multiple computing systems. These multiple computing systems may be connected and may exchange data and / or commands or other instructions, or the like, through one or more connections, including, but not limited to, a connection over a network (for example, the internet, a wireless wide area network, a local area network, a wide area network, a wired network, or the like), through a direct connection between one or more of the various computing systems, etc.

[0172] Although certain features of this disclosed matter are described for illustrative purposes in relation to specific deployments, it should be readily understood that these features are not intended to be limiting. The claims that accompany this disclosure are intended to define the scope of the protected matter.

[0173] The present disclosure contemplates that the calculations disclosed in the modalities of this document can be performed in a series of ways, applying the same concepts taught in this document, and that such calculations are equivalent to the disclosed modalities.

[0174] One or more aspects or characteristics of the material described here can be performed in digital electronic circuits, integrated circuits, specially designed application-specific integrated circuits (ASICs), field programmable port arrangements (FPGAs), hardware, firmware, software computer and / or combinations thereof. These various aspects or features may include deployment to one or more computer programs that are executable and / or interpretable in a programmable system, including at least one programmable processor, which may be of special or general use, coupled to receive data and instructions , and to transmit data and instructions to a storage system, at least one input device and at least one output device. The programmable system or computing system can include clients and servers. A client and a server are usually far apart and usually interact through a communication network. The relationship between client and server arises because of computer programs that run on the respective computers and that have a client and server relationship between them.

[0175] These computer programs, which can also be called programs, software, software applications, applications, components or code, include machine instructions for a programmable processor and can be deployed in a high-level procedural language, a object-oriented programming language, a functional programming language, a logical programming language and / or a set / machine language. As used in this document, the term

“Machine-readable medium” (or “computer-readable medium”) refers to any computer program product, apparatus and / or device, such as magnetic disks, optical disks, memory and programmable logic devices (PLDs) , used to provide machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" (or "computer-readable signal") refers to any signal used to provide machine instructions and / or data to a programmable processor. The machine-readable medium can store these machine instructions in a non-transitory manner, for example, as would a non-transitory solid-state memory or a magnetic hard drive or any equivalent storage medium. The machine-readable medium may, alternatively or additionally, store such machine instructions in a transient manner, for example, as would a processor cache or other random access memory associated with one or more physical processor cores.

[0176] To provide interaction with a user, one or more aspects or characteristics of the material described here can be implanted in a computer with a display device, such as, for example, a cathode ray tube (CRT) or a crystal screen liquid (LCD) or a light-emitting diode (LED) monitor to display information to the user and a keyboard and pointing device, such as a mouse or trackball, through which the user can provide input to the computer. Other types of devices can also be used to provide user interaction. For example, the feedback provided to the user can be any form of sensory feedback, such as, for example, visual feedback, auditory feedback or tactile feedback; and user input can be received in any form, including, but not limited to, acoustic, voice or sensory input. Other possible input devices include, but are not limited to, touch screens or other touch sensitive devices, such as single or multi-point resistive or capacitive trackpads, voice recognition hardware and software, optical scanners, optical pointers, capture devices digital imaging and associated interpretation software and the like.

[0177] In the descriptions above and in the claims, expressions such as "at least one among" or "one or more among" may occur followed by a conjunctive list of elements or resources. The term "and / or" can also occur in a list of two or more elements or resources. Unless implicitly or explicitly contradicted by the context in which it is used, this expression means any of the elements or features listed individually or any of the elements or features cited in combination with any of the other elements or features cited. For example, the expressions “at least one of A and B”, “one or more of A and B” and “A and / or B” each mean “only A, only B or A and B together. " An interpretation similar to lists that include three or more items is also intended. For example, the expressions “at least one of A, B and C”, “one or more of A, B and C” and “A, B and / or C” each mean “only A, only B, only C, A and B together, A and C together, B and C together or A, B and C together ”. The use of the term “based on” above and in the claims means “based at least in part”, so that an unquoted feature or element is also allowed.

[0178] The material described in this document can be incorporated into systems, devices, methods, computer programs and / or articles, depending on the desired configuration. Any methods or logical flows represented in the attached figures and / or described in this document do not necessarily require the particular order shown, or sequential order, to achieve the desired results. The implantations established in the previous description do not represent all implantations consistent with the material described here. Instead, these are just a few examples consistent with aspects related to the subject described. Although some variations have been described in detail above, other modifications or additions are possible. In particular, other features and / or variations may be provided in addition to those set forth herein. The deployments described above may refer to various combinations and subcombination of the disclosed features and / or combinations and subcombination of other features mentioned above. In addition, the advantages described above are not intended to limit the application of any claims made to processes and structures that realize any or all of the advantages.

[0179] In addition, the section titles must not limit or characterize the invention (or inventions) established in any claims that may result from this disclosure. Specifically, and by way of example, although the titles refer to a “Field of the Technique”, these claims should not be limited by the language chosen in that title to describe the so-called field of the technique. In addition, the description of a technology in the "Background" should not be interpreted as an admission that the technology is prior to any invention (or inventions) in this disclosure. The “Summary” should also not be considered as a characterization of the invention (or inventions) established in the claims issued. In addition, any reference to this disclosure in general or the use of the word “invention” in the singular is not intended to imply any limitation in the scope of the claims set out below. Multiple inventions can be established according to the limitations of the multiple claims issued as of this disclosure, and those claims, accordingly, define the invention (or inventions), and their equivalents, which are protected by them.

Claims (16)

1. Computer program product, characterized by comprising a non-transitory machine-readable medium that stores instructions that, when executed by at least one programmable processor, cause at least one programmable processor to perform operations that include: receiving data from photoplethysmographic signal (PPG) communicated by a PPG sensor from a wearable device worn by a user; determining a plurality of heartbeats from at least a portion of the PPG signal data; determining a type of heart rate based on at least the plurality of heart beats; determine whether the type of heart rate includes Atrial Fibrillation (AF); and display, when FA is detected, an AF detection alert on the wearable device. 2. Computer program product, according to claim 1, characterized by the operations further comprising: determining an occupancy metric based at least partially on an interval dispersion analysis; determining the type of heart rate also includes the use of the occupation metric. 3. Computer program product, according to claim 2, characterized in that the occupancy metric is determined by operations that comprise: determining a plurality of points that represent changes in heartbeat intervals over a period of time; define boxes in which the plurality of points can fall; and determining the occupancy metric by determining a fraction of compartments that contain at least one of the plurality of points. 4. Computer program product according to claim 3, characterized in that the dots represent a change between two adjacent heartbeat intervals. 5. Computer program product according to claim 3, characterized in that the time period is based on a movable window. 6. Computer program product, according to claim 1, characterized by the operations further comprising: determining a distance metric based at least partially on points that represent changes in heartbeat interval; determining the type of heart rate also includes the use of the distance metric. 7. Computer program product, according to claim 6, characterized in that the distance metric is determined by operations that comprise: determining a plurality of points that represent changes in heartbeat interval over a period of time; and determining the metric distance by determining a median of distances between the plurality of points. 8. Computer program product according to claim 7, characterized in that the distances are determined between consecutive points. 9. Computer program product according to claim 7, characterized in that the time period is based on a movable window. 10. Computer program product, according to claim 1, characterized by the operations also comprising: determining an interval variability metric based on a plurality of heartbeat intervals derived from the plurality of heartbeats; in which the determination of the type of heart rate also includes the use of the interval variability metric. 11. Computer program product, according to claim 10, characterized in that the interval variability metric is a median of absolute changes in heartbeat intervals. Computer program product according to claim 10, characterized in that the plurality of heartbeat intervals is included in a movable window. 13. Computer program product according to claim 1, characterized by the fact that determining whether the type of heart rate includes Atrial Fibrillation (AF) comprises the use of an occupancy metric, a distance metric and a metric of interval variability. 14. Computer program product according to claim 1, characterized by the display of the AF detection alert occurring when AF is determined based on a number of heartbeats greater than an AF detection alert limit. 15. Computer program product, according to claim 1, characterized by the operations further comprising: receiving a patient identification communicated by the wearable device; and generate the AF detection alert when an authorization is received that indicates a permission to display AF detection alerts to the user. 16. System, characterized by the fact that it comprises: at least one programmable processor; and a non-transitory machine-readable medium that stores instructions that, when executed by the at least one programmable processor, cause the at least one programmable processor to perform operations comprising: receiving photoplethysmographic signal (PPG) data communicated by a PPG sensor from a wearable device worn by a user; determining a plurality of heartbeats from at least a portion of the PPG signal data; determining a type of heart rate based on at least the plurality of heart beats; determine whether the type of heart rate includes Atrial Fibrillation (AF); and display, when FA is detected, an AF detection alert on the wearable device.