Processor for heart rate measurement

Abstract

This document discloses an apparatus and method for heart rate calculation. The apparatus may include a processor configured to calculate heart rate information. The processor includes a heart rate calculator, which includes memory for storing photoplethysmogram (PPG) signals and a computing element. The computing element is coupled to the memory and configured to calculate heart rate values ​​and generate at least one quality check factor based on the PPG signals. The processor also includes a check element, a memory control element, and a multiplexer. The check element is configured to determine a validity index based on the at least one quality check factor. The memory control element is coupled to the memory and configured to access the memory to transmit PPG signals. The multiplexer is configured to output either the PPG signals accessed by the memory control element or the heart rate values ​​calculated by the computing element, based on the validity index.

Description

Technical Field

[0001] This invention relates to a heart rate measurement device, and more particularly to a heart rate measurement device. Prior Technology

[0002] Optical sensors are commonly used in many systems, such as smartphones, wearable electronics, robots, and autonomous vehicles, for proximity detection, 2D / 3D imaging, object recognition, image enhancement, material recognition, color fusion, health monitoring, and other related applications. Summary of the Invention

[0003] This disclosure discloses an electronic device with an optical sensing device, wherein the electronic device can calculate heart rate through a processor, thus enabling continuous monitoring of biometric information from heart rate with low power consumption. The optical sensing device can operate in different wavelength ranges, including visible light (e.g., wavelengths from 380 nm to 780 nm, or similar wavelength ranges defined by a specific application) and invisible light. Invisible light includes near-infrared (NIR, e.g., wavelengths from 780 nm to 1400 nm, or similar wavelength ranges defined by a specific application) and short-wavelength infrared (SWIR, e.g., wavelengths from 1400 nm to 3000 nm, or similar wavelength ranges defined by a specific application).

[0004] One aspect of this disclosure relates to a device including a processor configured to calculate heart rate information. The processor includes a heart rate calculator, which includes memory configured to store photoplethysmogram (PPG) signals and a computing element. The computing element is coupled to the memory and configured to calculate heart rate values ​​and generate at least one quality check factor based on the PPG signals. The processor also includes a check element, a memory control element, and a multiplexer. The check element is configured to determine a validity index based on at least one quality check factor. The memory control element is coupled to the memory and configured to access the memory to transmit PPG signals. The multiplexer is configured to output the PPG signals accessed through the memory control element or the heart rate values ​​calculated by the computing element, based on the validity index.

[0005] In some implementations, the computing element includes analysis circuitry configured to calculate the heart rate value by detecting the interval between the peak value of the PPG signal and the calculated peak value.

[0006] In some implementations, the computing element includes a matched filter configured to determine a quality metric as at least one quality check factor.

[0007] In some implementations, quality metrics are determined by calculating the average absolute value of the PPG signal.

[0008] In some implementations, the computing element includes a DC subtraction element and a bandpass filter. The DC subtraction element is configured to receive a PPG signal and output a PPG signal to remove the DC value of the AC signal. The bandpass filter is configured to receive an AC signal and output an AC signal to remove the bandpass signal of the out-of-frequency signal.

[0009] In some implementations, the processor also includes a skin detector configured to detect the presence of object skin.

[0010] In some implementations, the processor also includes a controller configured to operate the skin detector and heart rate calculator in an interleaved manner.

[0011] In some embodiments, the device further includes an optical receiver and a control device, the optical receiver including one or more light detectors, and the control device including a processor.

[0012] In some embodiments, one or more photodetectors are formed on a first substrate, and the wafer of the first substrate is bonded to a control device formed on a second substrate.

[0013] In some implementations, a bonding interface exists between one or more photodetectors and a processor.

[0014] In some implementations, one or more optical detectors are stacked on the control device.

[0015] In some implementations, the processor includes a controller configured to control the optical receiver.

[0016] In some embodiments, one or more photodetectors include a plurality of sensing regions deposited on a substrate, the plurality of sensing regions being made of a material different from that of the substrate.

[0017] In some implementations, the memory control element is configured to output the PPG signal in a first-in-first-out (FIFO) configuration.

[0018] Another aspect of this disclosure relates to a method for performing heart rate calculation via a device including a processor. This method includes obtaining a PPG signal configured within the device, calculating a heart rate value based on the PPG signal using a computing element configured within the processor, and generating at least one quality check factor via the computing element to a check element configured within the processor. This method also includes providing a validity index based on the at least one quality check factor using the check element, and determining whether to output the heart rate value based on the validity index.

[0019] In some implementations, this method includes determining whether to output a PPG signal based on an effectiveness metric.

[0020] In some implementations, this method includes detecting peak values ​​of the PPG signal and calculating the intervals between peak values ​​to calculate a heart rate value.

[0021] In some implementations, this method includes obtaining the average absolute value by a matched filter to generate a quality index as at least one quality check factor.

[0022] In some implementations, this method includes storing the PPG signal in memory.

[0023] In some implementations, the heart rate value can be obtained for each calculation cycle of the PPG signal.

[0024] Another aspect of this disclosure relates to a processor for receiving PPG signals, including a computing element and a check element coupled to the computing element. The computing element calculates a heart rate value based on the PPG signal and outputs a filtered PPG signal and a PPG signal interval. The check element is configured to determine a validity index of the PPG signal and includes a periodicity detector configured to: receive the filtered PPG signal; obtain a PPG sampling signal for a sampling period based on the filtered PPG signal; receive the PPG signal interval; obtain a normalized initial value related to the heart rate value based on the PPG signal interval; perform an autocorrelation operation on a local segment of the PPG sampling signal centered on the normalized initial value, the range of the local segment being smaller than the sampling period; and detect the maximum value of the autocorrelation operation result of the local segment. If the maximum value is greater than a preset threshold, the periodicity detector outputs a periodicity index to indicate that the PPG signal is periodic and to determine the validity index.

[0025] In some implementations, the periodic detector obtains the PPG sampling signal using a window function and a filtered PPG signal, wherein the window length of the window function is less than the sampling period.

[0026] In some implementations, the periodic detector obtains a normalized initial value by normalizing the PPG signal interval using the sampling frequency.

[0027] In some implementations, the range of a local segment is less than 1 / 10 of the sampling period.

[0028] In some implementations, the periodicity detector is configured to output a periodicity index to indicate that the PPG signal is not periodic if the maximum value is less than a preset threshold.

[0029] In some implementations, the PPG signal originates from an optical receiver.

[0030] In some implementations, the PPG signal interval is the time interval between PPG signal peaks.

[0031] In some implementations, the computing element includes a DC subtraction element and a bandpass filter, configured to eliminate noise in the PPG signal to obtain a filtered PPG signal.

[0032] In some implementations, the processor also includes a skin detector configured to detect the presence of object skin.

[0033] In some implementations, the skin detector and computing elements operate in an interleaved manner. Simple Explanation of the Diagram

[0034] Referring to the following detailed description, the foregoing features and numerous advantages of this application will become readily understood, and will be even more easily understood when viewed in conjunction with the accompanying drawings:

[0035] Figure 1 is a block diagram illustrating an optical sensing device according to an embodiment of the present disclosure.

[0036] Figure 2A is a block diagram illustrating a heart rate calculator of an optical sensing device according to an embodiment of the present disclosure.

[0037] Figure 2B is a block diagram of a heart rate calculator of an optical sensing device according to another embodiment of the present disclosure.

[0038] Figure 3A is a block diagram illustrating the calculation element of a heart rate calculator according to an embodiment of the present disclosure.

[0039] Figure 3B is a block diagram illustrating the checking element of a heart rate calculator according to an embodiment of the present disclosure.

[0040] Figure 4A is a timing diagram illustrating heart rate calculation according to an embodiment of the present disclosure.

[0041] Figure 4B is a timing diagram illustrating heart rate calculation according to another embodiment of the present disclosure.

[0042] Figure 5A is a flowchart illustrating heart rate calculation according to an embodiment of the present disclosure.

[0043] Figure 5B is a flowchart illustrating heart rate calculation according to another embodiment of the present disclosure.

[0044] Figure 5C is a flowchart illustrating heart rate calculation according to another embodiment of the present disclosure.

[0045] Figure 6 is a schematic diagram of an optical sensing device according to an embodiment of the present disclosure.

[0046] Figure 7 is a schematic diagram of an optical sensing device according to another embodiment of the present disclosure.

[0047] Figure 8 is a flowchart illustrating the determination process of a periodic detector according to another embodiment of the present disclosure. Implementation

[0048] The following embodiments, accompanied by drawings, illustrate the concepts of this disclosure. In the drawings or description, similar or identical parts use the same element symbols, and the shape, thickness, or height of elements may be reasonably enlarged or reduced in the drawings. The embodiments listed in this application are for illustrative purposes only and are not intended to limit the scope of this application. Any obvious modifications or changes to this application will not depart from the spirit and scope of this application.

[0049] Wearable electronic devices (such as earbuds, augmented reality / virtual reality (AR / VR) wearables, and wristbands) are typically worn by users for various activities, including listening to music, exercising, training, resting, daily living activities, and physical therapy. These devices, with their personal health monitoring capabilities, can provide users with biometric information during these activities. For example, photoplethysmography (PPG) is an optically obtained plethysmogram that can be used to determine various biometric data, such as heart rate, calories burned, skin moisture, blood oxygen saturation (SpO2), and / or blood pressure. In some cases, biometric information is calculated by the device's software, which consumes significant power, preventing the device from operating for extended periods without charging. Furthermore, software calculations may be slower than hardware calculations, and the software's computational processes can also slow down the operation of other software within the device. Therefore, developing electronic devices that provide accurate biometric information while reducing power consumption remains a crucial challenge. This would allow users to wear the devices for extended periods without frequent charging. Furthermore, this disclosure describes an optical sensing device that can calculate heart rate using hardware or software based on the quality of the PPG signal, thus providing more reliable measurements based on different operating conditions.

[0050] Figure 1 is a block diagram illustrating an optical sensing device 100 according to an embodiment of the present disclosure. The optical sensing device 100 may be located in an electronic device (not shown), which may be a wearable device or a portable device. The wearable device may be an earphone, wristband, watch, glasses, helmet, head-mounted device, or other wearable electronic device. The portable device may be a mobile phone, tablet computer, laptop computer, computer mouse, computer stylus, or other accessories. The optical sensing device 100 includes an optical transmitter 1, an optical receiver 2, a processor 10, a communication module 6, and other circuits 7.

[0051] The light transmitter 1 may include one or more light sources that emit light of the same or different wavelengths for object detection, biometric measurement, and / or indication. In one embodiment, the light transmitter 1 may include a first light source that emits NIR light and a second light source that emits SWIR light. In another embodiment, multiple light sources may emit SWIR light. The light transmitted by the light transmitter 1 may be absorbed and / or reflected by an object (not shown) near the electronic device. The light receiver 2 is configured to detect reflected light from an object near the electronic device. The light receiver 2 may include one or more photodetectors to receive light corresponding to the light emitted by the light transmitter 1 for object detection and / or biometric measurement. Biometric measurements may be calculated by the processor 10 based on the PPG signal received by the light receiver 2. In some embodiments, one or more photodetectors of the light receiver 2 may include photodetectors for three-dimensional (3D) depth sensing (such as i-TOF or d-TOF photodetectors), proximity sensing, spectral sensing, two-dimensional (2D) sensing (such as 2D IR imaging), or combinations thereof. Each photodetector can be implemented as a single photodetector or an array of photodetector pixels (as shown in Figures 6 and 7, which are 1D or 2D photodetector arrays).

[0052] Processor 10 may be implemented by a digital digital processor (DSP), a general-purpose processor, an application-specific integrated circuit (ASIC), digital circuitry, or any combination thereof. Processor 10 may include controller 3, heart rate (HR) calculator 4, and skin detector 5. Controller 3 is configured to control optical transmitter 1 and optical receiver 2. Heart rate calculator 4 is configured to receive PPG signals from optical receiver 2 and calculate heart rate values. Skin detector 5 is configured to detect the presence of skin on a target. In one example, if the user is not wearing the electronic device correctly, skin detector 5 will not recognize the presence of the target, and heart rate calculator 4 will not operate to calculate heart rate to save power consumption. In one embodiment, skin detector 5 and heart rate calculator 4 operate in an interleaved manner (e.g., measuring alternately over time) to precisely monitor biometric information. Communication module 6 is configured to transmit and receive electrical signals from processor 10 to one or more other devices and vice versa via one or more communication protocols (e.g., WiFi, Bluetooth, cellular systems). Other circuits 7 can be any circuit mounted on the optical sensing device 100 (such as charging circuits, additional processing circuits, memory, other sensors).

[0053] Figure 2A is a block diagram illustrating a heart rate calculator 4 of an optical sensing device 100 according to an embodiment of the present disclosure. The heart rate calculator 4 is configured to receive sensing signals from a light receiver 2 and provide output and validity indicators to an electronic device or other computing device, such as a smartphone, watch, or computer, for further signal processing or identification. The heart rate calculator 4 provides a heart rate value when the PPG signal quality is good. When the PPG signal quality is poor, the heart rate calculator 4 can switch to transmitting the PPG signal. In other words, the heart rate calculator 4 can output heart rate or PPG signals to an electronic device or other device based on validity indicators. Validity indicators represent the quality of the PPG signal.

[0054] The heart rate calculator 4 includes a calculation element 42, a memory 43, a check element 44, a memory control element 45 (such as a FIFO), a demultiplexer 41, and a multiplexer 47. When the skin detector 5 shown in Figure 1 detects the presence of the subject's skin, the sensing signal received from the light receiver 2 can be converted into a PPG signal by the processor 10, which is then transmitted to the calculation element 42 via the demultiplexer 41 and stored in the memory 43. The calculation element 42 is configured to calculate the heart rate value by detecting the peak value of the PPG signal and calculating the interval between the peak values. Furthermore, the calculation element 42 can output at least one quality check factor (Q check factor). The check element 44 is configured to receive at least one quality check factor from the calculation element 42 and determine a validity index to indicate the validity of the heart rate value. If the validity index indicates that the heart rate value is valid, a control signal is sent to the multiplexer 47 to control the heart rate calculator 4 to output the heart rate value through the multiplexer 47. If the validity indicator shows an invalid heart rate value, a control signal is sent to multiplexer 47 and demultiplexer 41 to control the heart rate calculator 4 to output a PPG signal via memory control element 45, memory 43, and multiplexer 47 for subsequent software processing. In another embodiment, when the validity indicator shows a valid heart rate value and the user has sufficiently accurate heart rate information, the control signal can also be sent to multiplexer 47 to control the heart rate calculator 4 to output a PPG signal for displaying other biometric information, such as calories, oxygen saturation (SpO2), and / or blood pressure. The control signal can come from hardware coupled to the validity indicator (such as a standalone controller), application-level software that can operate to issue control signals in response to the validity indicator, or user-provided command signals.

[0055] Memory 43 is configured to store PPG signals and is coupled to computing element 42 and memory control element 45 to provide PPG signals to computing element 42 or memory control element 45. PPG signals can be stored in memory 43 through computing element 42 or memory control element 45. Since computing element 42 and memory control element 45 share memory 43, the footprint of the heart rate calculator 4 can be reduced, and the chip size of processor 10 can be reduced.

[0056] In one embodiment, when the electronic device or other device receives an invalid validity indicator and PPG signal from the heart rate calculator 4, the electronic device or other device can transmit relevant information to adjust the measurement method, such as adjusting the power of the optical transmitter 1, the amplification gain of the optical receiver 2, and / or providing a notification to the user to adjust the wearing position of the electronic device.

[0057] Figure 2B is a block diagram illustrating a heart rate calculator 4 of an optical sensing device 100 according to another embodiment of the present disclosure. The heart rate calculator 4 includes a calculation element 42, a memory 43, a verification element 44, a memory control element 45, a demultiplexer 41, and a multiplexer 47. The heart rate calculator 4 can output a heart rate value or a PPG signal based on a validity indicator. If the validity indicator shows a valid heart rate value, the validity indicator controls the multiplexer 47 and the demultiplexer 41 to cause the heart rate calculator 4 to output a heart rate value through the multiplexer 47. If the validity indicator shows an invalid heart rate value, the validity indicator controls the multiplexer 47 and the demultiplexer 41 to cause the heart rate calculator 4 to output a PPG signal through the memory control element 45 and the multiplexer 47.

[0058] Figure 3A is a block diagram illustrating the calculation element 42 of a heart rate calculator 4 according to an embodiment of the present disclosure. The calculation element 42 is configured to receive PPG signals to generate at least one Q checksum factor and a heart rate value. The at least one Q checksum factor may include one or more PPG signals, filtered PPG signals, a quality index, and a PPG signal interval. The calculation element 42 includes a DC subtraction element 431, a bandpass filter 432, a matched filter 433, and an analysis circuit 434. The DC subtraction element 431 is configured to receive the PPG signals and output an alternating current (AC) signal having a DC value removed from the PPG signals. The bandpass filter 432 is configured to receive the AC signals and output a bandpass signal having out-of-band signals removed from the AC signals. The PPG signal is input to computing element 42 and processed by DC subtraction element 431 to remove the DC value of the PPG signal, and then processed by bandpass filter 432 to eliminate unwanted out-of-band noise, such as motion artifacts and electromagnetic interference, to obtain a filtered PPG signal with a higher signal-to-noise ratio (SNR). The filtered PPG signal is then stored in memory 43 and can be accessed by computing element 42 for further processing. In another embodiment, the PPG signal is input to computing element 42 and stored in memory 43 but is not processed by DC subtraction element 431 and bandpass filter 432. Matched filter 433 is coupled between memory 43 and analysis circuit 434 to move-average the periodic PPG signal according to its frequency and is configured to determine the quality index as a Q checksum factor. Analysis circuit 434 is coupled to matched filter 433 to generate PPG signal intervals as one of the Q checksum factors and heart rate values. The PPG signal interval can be obtained by calculating the interval between the PPG signal peaks.

[0059] In one embodiment, a quality metric, as a Q check factor, is transmitted to check element 44 to determine a validity metric. Matched filter 433 is configured to determine a quality metric representing the quality of the PPG signal. In one embodiment, matched filter 433 is configured to calculate the mean absolute value (MAVPPG) of the filtered PPG signal (or PPG signal), which is considered the quality metric. Matched filter 433 has multiple variables (such as tap size) to represent the ideal PPG profile of the object, and these variables can be dynamically adjusted during processing using feedback circuitry (not shown). Additionally, matched filter 433 is also configured to attenuate unwanted high-frequency noise from the filtered PPG signal (or PPG signal). Therefore, analysis circuitry 434 can receive low-noise PPG signals from matched filter 433 for subsequent processing. Analysis circuitry 434 is configured to determine a heart rate value, such as heartbeat, in real time by detecting the peak value of the low-noise PPG signal and calculating the interval between the peak values.

[0060] In one embodiment, referring to Figures 2A, 2B, and 3A, the check element 44 can be configured to compare a quality metric (such as MAVPPG) with a reference metric to determine a validity metric. When the optical sensing device 100 is activated, the optical transmitter 1 is inactive while the optical receiver 2 is activated. The average absolute value is obtained through a matched filter located in the computing element 42 or the check element 44 to represent background noise (MAVnoise). The background noise (MAVnoise) may originate from the circuitry of the optical receiver 2 or the package of the optical sensing device. In one embodiment, the reference metric can be set to C × MAVnoise, where C is an empirical constant. When the quality metric (such as MAVPPG) is greater than the reference metric (C × MAVnoise), the check element 44 indicates that the validity metric is valid. Conversely, the check element 44 indicates that the validity metric is invalid.

[0061] In another embodiment, the PPG signal, the filtered PPG signal, and / or the PPG signal interval are used as Q check factors to be transmitted to check element 44 to determine a validity index. Figure 3B is a block diagram illustrating the check element 44 of a heart rate calculator 4 according to an embodiment of the present disclosure, wherein the PPG signal, the filtered PPG signal, and / or the PPG signal interval are used as Q check factors. Check element 44 may include a stability detector 441, a signal strength detector 442, a noise floor detector 443, a periodicity detector 444, and decision logic 445. The stability detector 441 is configured to detect the stability of the PPG signal and generate a stability index based on the PPG signal. The signal strength detector 442 is configured to detect the signal strength of the PPG signal and generate a signal strength index based on the filtered PPG signal. The noise floor detector 443 is configured to detect the noise floor of the PPG signal and generate a noise floor index from the filtered PPG signal. The periodic detector 444 is configured to detect the periodicity of the PPG signal and generate a periodicity index from the filtered PPG signal and the PPG signal interval. Decision logic 445 can then determine an effectiveness index based on an evaluation stability index, a signal strength index, a noise floor index, and the periodicity index. In one embodiment, each index can be weighted with a corresponding weight value (predetermined or dynamically generated) for evaluation calculation. Each corresponding weight value can be different, or some corresponding weight values ​​can be different.

[0062] Figure 4A is a timing diagram illustrating the heart rate calculation, where the quality index is transmitted to check element 44 as a Q check factor to determine the validity index. Figure 4B is a timing diagram illustrating the heart rate calculation, where the PPG signal, filtered PPG signal, and / or PPG signal interval are transmitted to check element 44 as a Q check factor to determine the validity index. Each heart rate calculation can be performed after one calculation cycle of the PPG signal, and one calculation cycle has N samples. N is an integer, for example, N=50, 100, 150, and each calculation cycle can be 1 second. The first calculation requires multiple calculation cycles to accumulate enough samples for the matching filter to process; for example, the first calculation requires 6 calculation cycles or 6 seconds. After the first calculation, for accurate tracking, the heart rate calculator can obtain an updated heart rate every 1 calculation cycle (e.g., 1 second).

[0063] Figure 5A illustrates a heart rate calculation method 500 according to an embodiment of the present disclosure. Step S505 shows that when a skin detector (as shown in Figure 1, skin detector 5) detects the presence of skin on an object, an optical sensing device can obtain a PPG signal from a light receiver (as shown in Figure 1, light receiver 2) to operate the heart rate calculation. Step S510 shows that the calculation element of the heart rate calculator (as shown in Figures 2A to 2B, calculation element 42 of the heart rate calculator 4) is configured to calculate the heart rate value based on the PPG signal. Furthermore, the calculation element can also be configured to store the PPG signal in memory for access by other circuits or processors for other applications. Step S515 shows that the calculation element of the heart rate calculator can be configured to generate at least one quality check factor to the check element of the heart rate calculator. Step S520 shows that the check element of the heart rate calculator can be configured to provide a validity index based on at least one quality check factor. Step S525 shows that the heart rate calculator can decide whether to output a heart rate value based on the validity index. Furthermore, the heart rate calculator can select to output a PPG value based on the validity index.

[0064] Figure 5B illustrates steps S515 to S525 of heart rate calculation according to another embodiment of this disclosure. When at least one check factor includes a quality index as shown in Figure 3A, the calculation element of the heart rate calculator can include a matched filter to generate a quality index based on the PPG signal. The matched filter can process the PPG signal to generate a mean absolute value as the quality index. After step S515, step S521 shows that the check element of the heart rate calculator can be configured to provide a validity index based on a comparison of the quality index and a reference index. Next, step S525 shows that the heart rate calculator can decide whether to output a heart rate value based on the validity index.

[0065] Figure 5C illustrates steps S515 to S525 of heart rate calculation according to another embodiment of the present disclosure. When at least one check factor includes the PPG signal, the filtered PPG signal, and the heart rate value as shown in Figure 3B, after step S515, step S523 shows that the check element of the heart rate calculator can be configured to provide a validity index based on the evaluation of multiple indicators obtained from the PPG signal, the filtered PPG signal, and the heart rate value. Next, step S525 shows that the heart rate calculator can decide whether to output the heart rate value based on the validity index.

[0066] Figure 6 is a schematic diagram illustrating an optical sensing device 600 according to an embodiment of the present disclosure. The optical sensing device 600 includes a photodetector 601 and a control device 602 (such as a CMOS circuit) electrically coupled to the photodetector 601. The control device 602 may include the processor 10 shown in Figure 1. The photodetector 601 includes a first substrate 610 and a sensing region 612 deposited on the first substrate 610. The control device 602 includes a second substrate 630 and a circuit region 632 (such as a CMOS circuit) carried by the second substrate 630. In one embodiment, the first substrate 610 and the second substrate 630 include the same material, such as a silicon substrate. In one embodiment, the photodetector 601 and the control device 602 may be disposed on the same plane of a circuit board (not shown), and the photodetector 601 is electrically coupled to the control device 602 through (multiple) wirings 622 (such as wire bonding). In another embodiment, to reduce the size of the optical sensing device 600, the photodetector 601 can be stacked on the control device 602 and electrically coupled to the control device 602 via (multiple) wiring or conductive adhesive. The sensing region 612 includes a material that is different from (e.g., made of a heterogeneous material) or the same as (e.g., made of a homogeneous material) the first substrate 610. In one embodiment, the material of the sensing region 612 can include group III-V materials, such as phosphorus (P), nitrogen (N), gallium (Ga), indium (In), and aluminum (Al). In another embodiment, the material of the sensing region 612 can include group IV materials, such as germanium (Ge) and silicon (Si).

[0067] Figure 7 is a schematic diagram of an optical sensing device 700 according to another embodiment of the present disclosure. The optical sensing device 700 includes a photodetector 701, a control device 702, and a bonding interface 703. The control device 702 may include the processor 10 shown in Figure 1. The photodetector 701 and the control device 702 are wafer-bonded via the bonding interface 703 (such as oxide or other suitable material). The photodetector 701 includes a first substrate 710 and a plurality of sensing regions 712 deposited on the first substrate 710. The control device 702 includes a second substrate 730 and a plurality of corresponding circuit regions 732 carried by the second substrate 730. Each circuit region 732 is electrically coupled to a corresponding sensing region 712 through a conductive path 722 of the bonding interface 703. Both the first substrate 710 and the second substrate 730 may be silicon substrates. The sensing regions 712 may include materials different from (e.g., made of dissimilar materials) or the same as (e.g., made of homogeneous materials) the first substrate 710. In one embodiment, the material of the sensing region 712 may include group III-V materials, such as P, N, Ga, In, and Al. In another embodiment, the material of the sensing region 712 may include group IV materials, such as Ge and Si.

[0068] Referring to Figure 3B, one method for determining the periodicity of a PPG signal involves performing an autocorrelation operation on the PPG sampling signal throughout the entire sampling period and determining whether the PPG signal is periodic based on the peak value of the autocorrelation result. However, when the system's processor computing resources are limited, performing an autocorrelation operation on the PPG sampling signal throughout the entire sampling period may consume excessive processor resources, affecting the operation of other system functions. Figure 8 is a flowchart illustrating the determination process of a periodicity detector according to an embodiment of this disclosure. Flowchart 800 illustrates a method for determining the periodicity of a PPG signal using the periodicity detector 444 in the check element of Figure 3B. In particular, it describes a method that performs an autocorrelation operation on a local segment of a PPG sampling signal within a sampling period, thereby determining whether it is periodic using fewer processor resources. In step S801, the periodic detector 444 receives the filtered PPG signal X[n] and obtains a PPG sampling signal S[n] of length N based on the filtered PPG signal X[n] and a window function w[n]. S[n] = X[n]·w[n], 0≦n≦N; N is the sampling window length. Depending on the sampling frequency Fs (e.g., Fs = 50 Hz), the window length N can correspond to several calculation cycles (as shown in Figures 4A~4B, 1~6 seconds), N≧100 (e.g., N = 100, 150, 200, 250, 300). In step S802, the periodic detector 444 receives the PPG signal interval and obtains a normalized initial value Kini related to the heart rate value based on the PPG signal interval. The normalized initial value Kini can be used to set local segments in the PPG sampling signal of length N for autocorrelation calculation. The normalization calculation can be performed by normalizing the PPG signal interval using the sampling frequency Fs to obtain Kini. For example, the PPG signal interval is the time interval between PPG signal peaks, and the initial normalization value Kini = the time interval between PPG signal peaks × the sampling frequency (Fs). In step S803, the periodic detector 444 uses the initial normalization value Kini as the center of the local segment and performs autocorrelation calculation on the local segment [Kini - NR, Kini + NR] of S[n] to obtain the autocorrelation coefficient R[k]; Kini - NR ≦ k ≦ Kini + NR. The range 2NR of the local segment is less than the sampling period N, for example: 2NR ≦ Step S804: The periodicity detector 444 detects the segment maximum value Rmax of the autocorrelation coefficient R[k] in the local segment. Step S805: The periodicity detector 444 determines whether Rmax is greater than a preset threshold value RTH. Step S806: If Rmax≥RTH, the periodicity detector 444 outputs a periodicity indicator to indicate that the PPG signal has periodicity. Step S807: If Rmax<RTH, the periodicity detector 444 outputs a periodicity indicator to indicate that the PPG signal does not have periodicity.

[0069] Although the present disclosure has described the preferred embodiments by way of examples, it should be understood that the present disclosure is not limited thereto. On the contrary, it is intended to cover various modifications and approximate arrangements and procedures. Therefore, the appended patent claims should be construed in the broadest sense to cover all such modifications and approximate arrangements and procedures.

[0070] 1: Optical transmitter 10: Processor 100, 600, 700: Optical sensing device 2: Optical receiver 3: Controller 4: Heart rate calculator 41: Demultiplexer 42: Computing element 43: Memory 431: DC subtraction element 432: Bandpass filter 433: Matched filter 434: Analysis circuit 44: Verification element 441: Stability detector Y步驟S804,週期性偵測器444偵測局部區段中的自相關係數R[k]中的區段極大值Rmax。步驟S" 443: Noise floor detector 444: Periodicity detector 445: Decision logic 45: Memory control element 47: Multiplexer 5: Skin detector 500, 800: Method 6: Communication module {601, 701}: Photodetector 602, 702: Control device 610, "710: First substrate 612, 712: Sensing area 622: Wiring 630, 730: Second substrate 632, 732: Circuit Region 7: Other circuits 703: Interface 722: Conductive path 800: Flowchart S505, S510, S515, S520, S521, S523, S525: Steps S801, S802, S803, S804, S805, S806, S807, S808: Steps

Claims

1. A processor for receiving photoplethysmogram (PPG) signals, comprising: A computing element calculates a heart rate value based on the PPG signal and outputs a filtered PPG signal and a PPG signal interval. The system also includes a check element coupled to the computing element, configured to determine a validity index for the PPG signal. The check element includes a periodicity detector configured to: receive the filtered PPG signal; obtain a PPG sampling signal for a sampling period based on the filtered PPG signal; receive the PPG signal interval; obtain a normalized initial value related to the heart rate value based on the PPG signal interval; perform an autocorrelation operation on a local segment of the PPG sampling signal centered on the normalized initial value, the local segment being smaller than the sampling period; and detect a maximum value of the autocorrelation result of the local segment. If the maximum value is greater than a preset threshold, the periodicity detector outputs a periodicity index to indicate that the PPG signal is periodic, and to determine the validity index.

2. As in the processor of request item 1, where, The periodic detector obtains the PPG sampling signal using a window function and the filtered PPG signal, wherein the window length of the window function is less than the sampling period.

3. As in the processor of request item 1, where, The periodic detector obtains the initial value of the normalization by normalizing the PPG signal interval using a sampling frequency.

4. The processor of request item 1, wherein the range of the local segment is less than 1 / 10 of the sampling period.

5. As in the processor of request item 1, where, The periodicity detector is configured to output a periodicity index to indicate that the PPG signal is not periodic if the maximum value is less than the preset threshold value.

6. As in the processor of request item 1, where, The PPG signal comes from an optical receiver.

7. The processor for request item 1, wherein, The PPG signal interval is the time interval between the peak values ​​of the PPG signal.

8. The processor for request item 1, wherein, The computing element includes a DC subtraction element and a bandpass filter, configured to eliminate noise in the PPG signal to obtain the filtered PPG signal.

9. The processor of claim 1, wherein the processor further includes a skin detector configured to detect the presence of object skin.

10. The processor of claim 9, wherein the skin detector and the computing element operate in an interleaved manner.

Images