Data storage method and device, data reconstruction method and device, equipment and medium

Abstract

The invention provides a data storage method and device, a data reconstruction method and device, equipment and a medium, and the method comprises the steps: obtaining a plurality of groups of photoplethysmography (PPG) signals; dividing the plurality of groups of PPG signals into a group of reference signals and at least one group of candidate signals; for each group of candidate signals, target signals corresponding to the candidate signals are determined from the multiple groups of PPG signals; calculating a conversion relation between the target signal and the candidate signal; and storing the reference signal and the conversion relationship between each group of candidate signals and the corresponding target signal. Through the method, the storage efficiency can be improved while the integrity of the key information of the signal is ensured.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a data storage and data reconstruction method, device, equipment and medium. BACKGROUND

[0002] The photoplethysmography (PPG) signal is a physiological signal for non-invasive detection of the pulse wave change of the blood volume of the microvascular bed by using optical principles. The conventional collection method is to use a single-wavelength or multi-wavelength photoelectric sensor to contact the body surface, convert the transmitted or reflected light intensity signal into an electrical signal through a photoelectric converter, and then acquire and analyze the heart rate, blood oxygen saturation, blood vessel elasticity, respiratory rate and other physiological parameters.

[0003] However, the inventors have found in the in-depth research that, with the extension of continuous monitoring time and the expansion of application scenarios, the storage and transmission of massive PPG signal data have become a bottleneck restricting its large-scale application. The high sampling rate and large data volume of the original PPG signal have put forward too high requirements on the storage medium capacity, embedded system resources and wireless communication bandwidth, and significantly increased the hardware cost and system power consumption.

[0004] Therefore, it is necessary to provide a solution. SUMMARY

[0005] Therefore, the embodiments of the present application provide a data storage and data reconstruction method, device, equipment and medium to efficiently store multiple groups of PPG signals.

[0006] In a first aspect, the embodiments of the present application provide a data storage method, which comprises: acquiring multiple groups of photoplethysmography (PPG) signals; there is a partial difference between at least two groups of PPG signals; distinguishing the multiple groups of PPG signals into one group of reference signals and at least one group of candidate signals; the signal quality of the reference signals is higher than that of the candidate signals; for each group of candidate signals, determining a target signal corresponding to the candidate signal from the multiple groups of PPG signals; the similarity between the target signal and the candidate signal meets a preset requirement, and the target signal is the reference signal or a candidate signal for which the conversion relationship has been calculated; calculating the conversion relationship between the target signal and the candidate signal; the conversion relationship is used to reconstruct the candidate signal based on the target signal; storing the reference signals and the conversion relationship between each group of candidate signals and the target signal corresponding thereto.

[0007] In an implementable embodiment, the multiple sets of PPG signals are obtained by at least one acquisition device at different acquisition positions, and / or based on light of different colors; storing the reference signal, comprising: determining a core frequency component of each set of PPG signals, to obtain a target frequency band of the multiple sets of PPG signals; the target frequency band includes each core frequency component; band-pass filtering the reference signal based on the set of core frequency components; storing the reference signal after filtering.

[0008] In an implementable embodiment, the method further comprises: segmenting each set of PPG signals into multiple windows; there is at least partial overlap between any two windows; performing Fourier transform on data in each window to obtain a frequency domain signal; compressing each set of PPG signals based on the frequency domain signal to obtain multiple sets of compressed PPG signals; distinguishing the multiple sets of PPG signals into a set of reference signals and at least one set of candidate signals, comprising: distinguishing the multiple sets of compressed PPG signals into a set of reference signals and at least one set of candidate signals.

[0009] In an implementable embodiment, compressing each set of PPG signals based on the frequency domain signal comprises: for the frequency domain signal of each set of PPG signals, dynamically assigning a scaling factor to each frequency domain coefficient according to the amplitude of the frequency domain coefficient; the scaling factor is proportional to the amplitude; performing non-uniform scaling on the frequency domain signal based on the scaling factor corresponding to each frequency domain coefficient.

[0010] In an implementable embodiment, calculating the conversion relationship between the target signal and the candidate signal comprises: for each set of candidate signals, when the target signal is a reference signal, calculating the conversion relationship between the reference signal and the set of candidate signals based on a preset algorithm; the preset algorithm includes: generalized linear model, and / or cross-correlation analysis; determining a first signal obtained based on the conversion relationship and the reference signal as a candidate signal for which the conversion relationship has been calculated; or, for each set of candidate signals, when the target signal is a first signal, calculating the conversion relationship between the first signal and the set of candidate signals based on the preset algorithm; determining a second signal based on the conversion relationship and the first signal as the candidate signal for which the conversion relationship has been calculated; Alternatively, for each group of candidate signals, when the target signal is a second signal, a conversion relationship between the second signal and the group of candidate signals is calculated based on the preset algorithm.

[0011] In a feasible implementation, the method further includes: calculating a residual error between the target signal and the candidate signal.

[0012] determining a compression degree of the residual error based on a signal quality of the candidate signal; the compression degree is inversely proportional to the signal quality; compressing the residual error of the candidate signal based on the compression degree; storing the compressed residual error.

[0013] In a second aspect, the embodiments of the present application further provide a method for reconstructing data, which includes: obtaining a to-be-processed signal; the to-be-processed signal is obtained based on the method of any one of the first aspect; the to-be-processed signal at least includes a reference signal, and a conversion relationship between a target signal and a candidate signal, the target signal being the reference signal, or a candidate signal for which the conversion relationship has been calculated; reconstructing each group of candidate signals based on the target signal and the conversion relationship between the target signal and the candidate signal to be reconstructed.

[0014] In a third aspect, the embodiments of the present application further provide a data storage device, which includes: an obtaining module, configured to obtain multiple groups of photoplethysmography (PPG) signals; there is a partial difference between at least two groups of PPG signals; a distinguishing module, configured to distinguish the multiple groups of PPG signals into one group of reference signals and at least one group of candidate signals; a signal quality of the reference signal is higher than a signal quality of the candidate signal; a determining module, configured to determine, for each group of candidate signals, a target signal corresponding to the candidate signal from the multiple groups of PPG signals; a similarity between the target signal and the candidate signal meets a preset requirement, the target signal being the reference signal, or a candidate signal for which the conversion relationship has been calculated; a calculating module, configured to calculate a conversion relationship between the target signal and the candidate signal; the conversion relationship is used to reconstruct the candidate signal based on the target signal; a storage module, configured to store the reference signal, and a conversion relationship between each group of candidate signals and the target signal corresponding thereto.

[0015] In an implementable embodiment, the multiple groups of PPG signals are obtained by at least one acquisition device at different acquisition positions, and / or based on light of different colors; The storage module is configured to store the reference signal, and configured to: determine a core frequency component of each group of PPG signals to obtain a target frequency band of the multiple groups of PPG signals; the target frequency band includes each core frequency component; perform band-pass filtering on the reference signal based on the set of core frequency components; store the reference signal after filtering.

[0016] In an implementable embodiment, the device further comprises: The window module is configured to divide each group of PPG signals into multiple windows; and any two windows have at least partial overlap. The frequency domain conversion module is configured to perform Fourier transform on data in each window to obtain a frequency domain signal. The compression module is configured to compress each group of PPG signals based on the frequency domain signal to obtain multiple groups of compressed PPG signals. The distinguishing module is configured to distinguish the multiple groups of PPG signals into one group of reference signals and at least one group of candidate signals, and configured to: distinguish the multiple groups of compressed PPG signals into one group of reference signals and at least one group of candidate signals.

[0017] In an implementable embodiment, the compression module is configured to compress each group of PPG signals based on the frequency domain signal, and configured to: for the frequency domain signal of each group of PPG signals, dynamically assign a scaling ratio to each frequency domain coefficient according to the amplitude of the frequency domain coefficient; the scaling ratio is proportional to the amplitude; perform non-uniform scaling on the frequency domain signal based on the scaling factor corresponding to each frequency domain coefficient.

[0018] In an implementable embodiment, the calculation module is configured to calculate a conversion relationship between the target signal and the candidate signal, and configured to: for each group of candidate signals, when the target signal is a reference signal, calculate a conversion relationship between the reference signal and the group of candidate signals based on a preset algorithm; the preset algorithm includes a generalized linear model and / or cross-correlation analysis; determine a first signal obtained based on the conversion relationship and the reference signal as a candidate signal for which the conversion relationship has been calculated; Alternatively, for each group of candidate signals, when the target signal is the first signal, a conversion relationship between the first signal and the group of candidate signals is calculated based on the preset algorithm; A second signal obtained based on the conversion relationship and the first signal is determined as the candidate signal for which the conversion relationship has been calculated; Alternatively, for each group of candidate signals, when the target signal is the second signal, a conversion relationship between the second signal and the group of candidate signals is calculated based on the preset algorithm.

[0019] In one possible implementation, the computing module is further configured to: calculate a residual error between the target signal and the candidate signal; Further comprising: a first module configured to determine a compression degree of the residual error based on a signal quality of the candidate signal; the compression degree is inversely proportional to the signal quality; a second module configured to compress the residual error of the candidate signal based on the compression degree; The compression module is further configured to store the compressed residual error.

[0020] In a fourth aspect, the embodiments of the present application further provide a data reconstruction device, which comprises: a signal acquisition module configured to acquire a to-be-processed signal; the to-be-processed signal is obtained based on the method of any one of the first aspect; the to-be-processed signal at least comprises a reference signal, and a conversion relationship between a target signal and a candidate signal, the target signal being the reference signal or a candidate signal for which the conversion relationship has been calculated; a reconstruction module configured to reconstruct each group of candidate signals based on the target signal and the conversion relationship between the target signal and the candidate signal to be reconstructed.

[0021] In a fifth aspect, the embodiments of the present application further provide an electronic device, which comprises a processor, a storage medium and a bus, the storage medium stores machine readable instructions executable by the processor, when the electronic device is running, the processor and the storage medium communicate through the bus, and the processor executes the machine readable instructions to perform the steps of the method in any one of the first aspect or the second aspect.

[0022] In a sixth aspect, the embodiments of the present application further provide a computer readable storage medium, the computer readable storage medium stores a computer program, when the computer program is run by a processor, the steps of the method in any one of the first aspect or the second aspect are performed.

[0023] The method, device, equipment and medium for data storage and data reconstruction provided by the embodiment of the present application are as follows: after a plurality of groups of PPG signals are obtained, a group of reference signals with the best signal quality is determined, then a complete signal reference is constructed through the group of reference signals, and the rest of candidate signals are efficiently represented by establishing a conversion relationship with the reference signals. Specifically, the present application converts the storage content from a plurality of groups of complete PPG signals to a lightweight architecture of "one complete reference signal + a plurality of conversion relationships". When the original data needs to be restored, the system can accurately reconstruct each candidate signal through the reference signal and the corresponding conversion relationship. This design ensures the integrity of the key information of the signal and realizes a qualitative leap in storage efficiency.

[0024] In order to make the above objectives, characteristics and advantages of the present application more apparent, clear and easy to understand, the following will specifically describe a preferred embodiment, and the accompanying drawings will be described in detail as follows. BRIEF DESCRIPTION OF DRAWINGS

[0025] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be considered as a limitation to the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0026] Figure 1 A flow chart of a method for data storage provided by the embodiment of the present application is shown.

[0027] Figure 2 A flow chart of a method for data reconstruction provided by the embodiment of the present application is shown.

[0028] Figure 3 A structural schematic diagram of a data storage device provided by the embodiment of the present application is shown.

[0029] Figure 4 A structural schematic diagram of a data reconstruction device provided by the embodiment of the present application is shown.

[0030] Figure 5 A structural schematic diagram of an electronic device provided by the embodiment of the present application is shown. DETAILED DESCRIPTION

[0031] To make the purposes, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application and are not all the embodiments. The components of the embodiments of the present application generally described and shown in the drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work belong to the scope of protection of the present application.

[0032] Smartwatches and smart rings and other wearable devices commonly use photoplethysmography (PPG) technology for health monitoring. For example, they derive key physiological parameters such as heart rate, blood oxygen saturation, heart rate variability (HRV), etc. by analyzing PPG raw signals.

[0033] With the development of sensor technology, with the development of sensing technology, the acquisition of PPG signals is no longer limited to a single device or a single modality: Within a single high-end wearable device, multiple wavelength (such as green light, red light, infrared light) light sources and light receivers are integrated to form multiple acquisition channels. These PPG signals of different wavelengths carry complementary physiological information due to their specific absorption characteristics for blood components such as oxyhemoglobin and deoxyhemoglobin.

[0034] At a more macro level of personal health networks, users may simultaneously wear smartwatches, smart rings, ear-mounted monitoring devices, etc. These heterogeneous device groups constitute a distributed multi-channel PPG acquisition system that captures multi-dimensional information of cardiovascular activity from different parts of the body (such as the wrist, fingertips, and ear) at the same or different wavelengths, asynchronously or synchronously.

[0035] However, whether it is multiple photoelectric channels of a single device or an acquisition network composed of multiple devices, they all face the same core contradiction: the sharp increase in data dimensions and information density, and the huge gap between the limited storage, computing, and communication resources of terminal devices. Advanced algorithms such as deep learning crave all these raw data to train more robust and accurate models, but the physical limitations of hardware make it impractical to store and transmit all raw signals directly.

[0036] In this context, the embodiments of the present application provide a data storage method, which is briefly introduced as follows: For single-device multi-wavelength scenarios, the signal quality of a certain wavelength channel can be dynamically selected (e.g., the green light PPG with the highest signal-to-noise ratio in a motion scenario) as the reference signal, while other wavelengths (e.g., red and infrared light PPG for blood oxygen calculation) are used as candidate signals, and only the conversion relationship between them and the reference signal is stored. This is essentially a physiological and physical correlation between different wavelength signals, which is efficiently packaged using a mathematical model.

[0037] For multi-device collaborative scenarios, the signal collected by the device with the best acquisition location, least interference, and highest signal quality can be selected as the global reference signal from the distributed device group. Regardless of the wavelength or acquisition site, the signals collected by other devices are considered candidate signals and are characterized by calculating their conversion relationship with the global reference signal (or intermediate signals with established conversion relationships).

[0038] In this way, the massive and heterogeneous raw PPG data is compressed into an efficient data set consisting of "one core reference signal + a series of lightweight conversion relationships". It not only greatly relieves the storage and transmission pressure, but more importantly, it establishes a unified data representation paradigm, enabling PPG signals from different sources and characteristics to be organically integrated and managed.

[0039] When training data is needed for cloud AI models or advanced analysis is needed, the system can quickly and accurately reconstruct the original PPG waveform of any channel, any device at any time point based on this data structure.

[0040] The following will be described in detail. As shown in Figure 1 The following steps are included: Step 101, obtaining a plurality of groups of photoplethysmography PPG signals; at least two groups of PPG signals have some differences.

[0041] For example, the plurality of groups of PPG signals are obtained by a plurality of channels. At least two groups of PPG signals in the plurality of groups of PPG signals have some differences. For example, the differences can be different wavelengths.

[0042] In a typical application scenario, wearable devices such as smart watches integrate multi-wavelength photoelectric sensing systems. The system can emit light of different wavelengths (e.g., green, red, and infrared) almost simultaneously and receive reflected or transmitted light signals from subcutaneous tissues, forming multiple independent signal acquisition channels. The PPG signals obtained by the green, red, and infrared channels, although synchronized in time, all reflect the same pulse rhythm, but their waveform patterns, direct current components, alternating current component amplitudes, and degrees of influence by motion noise will have obvious and quantifiable differences.

[0043] For the multiple sets of PPG signals acquired in this case, if each set of PPG signals is directly stored, a large storage space is required.

[0044] Step 102, the multiple sets of PPG signals are divided into a set of reference signals and at least one set of candidate signals; the signal quality of the reference signals is higher than the signal quality of the candidate signals.

[0045] By analyzing the signal quality of each set of PPG signals, the multiple sets of PPG signals are divided into a set of reference signals with the best signal quality, and the remaining PPG signals are determined as candidate signals, which are the signals that need to be compressed and stored in this scheme.

[0046] The selection of the reference signal is crucial, because its quality will directly determine the overall fidelity of the subsequent signal reconstruction.

[0047] The following introduces a scheme for comprehensively evaluating signal quality by combining multiple quantitative indicators in time domain, frequency domain and cepstrum domain, to comprehensively and objectively evaluate the quality of PPG signals: (1) Cepstrum domain indicators: Cepstrum peak salience: Calculate the peak intensity in the cepstrum domain of the signal corresponding to the heart rate period. A significant and isolated peak indicates that the signal contains strong and regular pulse wave components, and the signal quality is high.

[0048] Cepstrum peak-to-mean ratio: The higher the ratio, the more prominent the heart rate component relative to background noise and artifacts, and the purer the signal.

[0049] (2) Frequency domain indicators: Spectral entropy: used to measure the uncertainty (or degree of disorder) of the signal in the frequency domain. A high-quality PPG signal should have its power spectrum energy concentrated around the heart rate and its harmonic frequencies, resulting in a lower spectral entropy value. On the contrary, a noisy signal has its spectral energy distributed, and the entropy value is higher.

[0050] Signal-to-noise ratio: By analyzing the signal power spectrum, estimate the ratio of the signal power in the narrow band centered on the heart rate fundamental frequency and its harmonics to the noise power in the entire frequency band. The higher the signal-to-noise ratio, the better the signal quality.

[0051] (3) Time domain indicators: Kurtosis: measures the shape of the signal amplitude distribution. A high-quality PPG pulse sequence usually has a more "sharp" amplitude distribution (i.e., higher kurtosis), because it contains regular pulse peaks and lower baseline fluctuations. The amplitude distribution of a noisy signal is closer to a normal distribution (kurtosis close to 3).

[0052] Coefficient of variation of peak-to-peak interval: the ratio of the standard deviation to the mean of the time interval between consecutive pulse wave peaks (corresponding to the heart rate interval). A too low value can imply a rigid signal (loss of physiological realism), a too high value can imply a heavily disturbed signal or a large amount of missed / falsely detected pulse waves. A moderate, physiologically plausible value of the coefficient of variation is a sign of a high quality signal.

[0053] In assessing the signal quality of each group of PPG signals, all or part of the above indicators are calculated for each PPG signal to be evaluated. Subsequently, a weighted sum or rule-based decision tree and mathematical model method can be used to obtain the comprehensive quality index of each group of PPG signals. Finally, the PPG signal with the highest comprehensive quality index is selected as the reference signal, and the remaining signals are automatically classified as candidate signals.

[0054] Step 103, for each group of candidate signals, determining a target signal corresponding to the candidate signal from the plurality of groups of PPG signals; the similarity between the target signal and the candidate signal meets a preset requirement, and the target signal is the reference signal or a candidate signal for which the conversion relationship has been calculated.

[0055] This step aims to find a most suitable "template" or "benchmark" for each group of candidate signals that need to be compressed, i.e. the target signal, so as to ensure that the conversion relationship calculated subsequently is the most efficient and concise.

[0056] The similarity between the target signal and the candidate signal can be determined according to a preset algorithm. One core idea is to calculate the correlation coefficient of the two signals in the time domain, frequency domain or feature space. The higher the correlation coefficient, the higher the similarity. Exemplarily, the target signal can be determined according to the similarity of wavelengths.

[0057] Among them, the "candidate signal for which the conversion relationship has been calculated" is a virtual signal generated by a mathematical process. It is not original data itself, but a high-precision copy reconstructed by combining the reference signal (or the last such signal) with the corresponding conversion relationship.

[0058] That is, it allows the subsequent candidate signal to no longer only refer to the original "gold standard" (i.e. the reference signal), but can choose a "predecessor" that is more similar to itself and has a closer wavelength as the target. In this way, the conversion relationship describing the difference between the two will be simpler, and the required storage space will be smaller, thereby achieving a very high overall compression efficiency. Essentially, it upgrades the compression process from a single "central radiation" mode to an efficient, chain-like or tree-like "pipeline" mode.

[0059] Step 104, calculating a conversion relationship between the target signal and the candidate signal; the conversion relationship is used to reconstruct the candidate signal based on the target signal.

[0060] In a feasible implementation, for each group of candidate signals, when the target signal is a reference signal, a conversion relationship between the reference signal and the candidate signals in the group is calculated based on a preset algorithm; the preset algorithm includes: a generalized linear model, and / or cross-correlation analysis; A first signal obtained based on the conversion relationship and the reference signal is determined as a candidate signal for which the conversion relationship has been calculated. Alternatively, for each group of candidate signals, when the target signal is a first signal, a conversion relationship between the first signal and the candidate signals in the group is calculated based on the preset algorithm. A second signal obtained based on the conversion relationship and the first signal is determined as a candidate signal for which the conversion relationship has been calculated. Alternatively, for each group of candidate signals, when the target signal is a second signal, a conversion relationship between the second signal and the candidate signals in the group is calculated based on the preset algorithm.

[0061] Specifically, the above description describes a chain or tree type signal processing and dependency relationship establishment process. The core idea is that not all candidate signals must be directly converted with the original reference signal, but can be converted with an "intermediate signal" that is more similar in characteristics, which is derived from the reference signal or the last layer of intermediate signal.

[0062] Generalized linear models (such as finding mapping coefficients from target signals to candidate signals through linear regression) or cross-correlation analysis (used to find the best alignment and similarity between signals) and other algorithms are mathematical tools for calculating the conversion relationship between signals.

[0063] The "first signal" and the "second signal" are not new original signals, but virtual signals reconstructed or estimated from the reference signal or the last layer of signal by applying the conversion relationship. They are marked as "candidate signals for which the conversion relationship has been calculated", which means that they can be used as "benchmarks" for calculating the conversion relationship of other candidate signals, that is, the target signal.

[0064] When the present scheme is applied to the scenario of multiple groups of PPG signals of multiple wavelengths, a more intelligent strategy based on spectral physical characteristics can be introduced, which is described below in conjunction with examples: Since the morphological changes of PPG signals of different wavelengths have continuity, the closer the wavelengths, the more similar the waveform characteristics of the signals.

[0065] The following is an example of a system that includes four wavelength bands: green light (526 nm), orange light (600 nm), red light (665 nm), and infrared light (940 nm). Two possible reference selection paths are described: (1) Green light channel as the reference signal: When the green light signal is selected as the global reference signal, the determination of the target signal can follow the order of wavelength from near to far: a. For the "orange light" candidate signal, since its wavelength is closest to green light, the green light reference signal can be selected as the target signal, and the conversion relationship between the orange light candidate signal and the green light reference signal is directly calculated.

[0066] b. For the "red light" candidate signal, the "near orange light" estimated signal (i.e., the first signal) generated by linear conversion from the green light signal can be selected as the target signal. The green light reference signal can be selected as the target signal, and the target signal is the first signal. The conversion relationship between the "near orange light" estimated signal and the red light candidate signal is calculated. The near red light generated by linear conversion (calculated conversion relationship) from the red light candidate signal is determined as the second signal.

[0067] c. For the infrared light candidate signal, the near red light (second signal) generated by linear conversion (calculated conversion relationship) from the red light candidate signal can be selected as the target signal. The conversion relationship between the near red light and the infrared light candidate signal is calculated.

[0068] (2) Red light channel as the reference signal: When the red light signal is selected as the global reference signal, the process can be adjusted as follows: a. The infrared light candidate signal selects the red light reference signal as the target signal, and the target signal is the reference signal. The conversion relationship between the infrared light candidate signal and the red light reference signal is calculated.

[0069] b. The orange light candidate signal selects the red light reference signal as the target signal, and the target signal is the reference signal. The conversion relationship between the red light reference signal and the orange light candidate signal is calculated.

[0070] c. For the green light candidate signal with the largest wavelength difference, the "near orange light" estimated signal (first signal) generated by conversion from the red light signal can be selected as the target signal, and the target signal is the first signal. The conversion relationship between the "near orange light" and the green light candidate signal is calculated.

[0071] Step 105, store the reference signal, and the conversion relationship between each set of candidate signals and their corresponding target signals.

[0072] At this time, instead of storing multiple sets of PPG signals, only one set of reference signals and the conversion relationship between each set of candidate signals and the corresponding target signal need to be stored.

[0073] In combination with the foregoing example of taking red light as the reference signal, finally we only need to store the following: Data 1: a complete red light PPG signal (reference signal).

[0074] Data 2: the conversion relationship of the other three sets of signals: the conversion relationship of infrared light relative to red light, the conversion relationship of orange light relative to red light, and the conversion relationship of green light relative to the near orange light generated by red light.

[0075] That is, the present scheme does not store the original signals of infrared, orange, and green light, but through the reference signal and the corresponding conversion relationship, all signals can be perfectly reconstructed when needed, thereby greatly saving storage space.

[0076] The method, device, equipment and medium for data storage and data reconstruction provided by the embodiments of the present application, after obtaining multiple sets of PPG signals, first determine a set of reference signals with the best signal quality, and then build a complete signal benchmark through the set of reference signals, and the rest of the candidate signals are efficiently represented by establishing the conversion relationship with the reference signal. Specifically, the present scheme converts the storage content from multiple complete sets of PPG signals to a lightweight architecture of “one complete reference signal + multiple conversion relationships”. When it is necessary to restore the original data, the system can accurately reconstruct each candidate signal through the reference signal and the corresponding conversion relationship. This design not only guarantees the integrity of the key information of the signal, but also realizes a qualitative leap in storage efficiency.

[0077] In an optional embodiment, the multiple sets of PPG signals are obtained by at least one acquisition device at different acquisition positions and / or based on light of different colors.

[0078] Storing the reference signal comprises: Determining the core frequency components of each set of PPG signals to obtain the target frequency bands of the multiple sets of PPG signals; the target frequency bands include each core frequency component; performing band-pass filtering on the reference signal based on the set of core frequency components; and storing the reference signal after filtering.

[0079] In this step, in order to further reduce the amount of data to be stored, the reference signal to be stored needs to be further filtered. That is, instead of storing the complete reference signal, only the essence of the reference signal is stored. The specific method is: first analyze all PPG signals to find the core pulse frequency range of each group of PPG signals, get the target frequency band, and then band-pass filter the reference signal to retain only the waveform in the frequency band, and finally store this filtered and simplified signal.

[0080] For example, if the analysis finds that the effective pulse components of all signals are in the range of 0.8-2.5 Hz, the system will filter the reference signal with this frequency band, and eliminate low-frequency drift and high-frequency noise. The waveform in the reference signal that falls within the range of 0.8-2.5 Hz is finally stored, which further compresses the data while retaining all the key features required for reconstruction.

[0081] Among them, for each group of PPG signals, the core frequency component can be determined by the following method: The core frequency component of the PPG signal is accurately locked by three-stage filtering. First, using the characteristics of PPG physiological signals (the effective pulse frequency is usually much lower than 10 Hz), a high-frequency cutoff valve of 10 Hz is set to directly eliminate high-frequency noise. Then, energy screening is performed in the frequency domain, and only significant frequency components with power exceeding 5% of the total energy are retained to filter out weak background interference. Finally, to ensure the integrity of the core physiological information, the remaining components are accumulated from high to low until they cover 98% of the total energy, thereby retaining all core frequency components representing heart rate and its harmonics. This combination strategy achieves the best balance between denoising and fidelity.

[0082] For each selected core frequency component, the index information of the core frequency component is recorded, so that the key core frequency components can be accurately found according to the index information during subsequent reconstruction, thereby reconstructing a better candidate signal.

[0083] In an optional embodiment, the method further comprises: For each group of PPG signals, the PPG signals are divided into multiple windows; there is at least partial overlap between any two windows; for each window, the data in the window is subjected to Fourier transform to obtain a frequency domain signal; based on the frequency domain signal, each group of PPG signals is compressed to obtain multiple groups of compressed PPG signals.

[0084] For each group of PPG signals, the continuous PPG signals are cut into sliding, partially overlapping time windows. This not only facilitates local spectral analysis (as the frequency characteristics of the signal can change over time), but also lays the foundation for real-time, streaming compression.

[0085] After obtaining the plurality of windows, the time-domain signals in each window are converted from time domain to frequency domain by fast Fourier transform. This conversion concentrates the signal energy on a few core frequency components, creating conditions for efficient compression.

[0086] Then, the frequency-domain coefficients in the frequency-domain signals obtained by Fourier transform are lossy compressed. For example, the frequency-domain signals can be compressed by the same amplitude.

[0087] Alternatively, different frequency-domain coefficients can be compressed by different amplitudes.

[0088] For example, for the frequency-domain signals of each group of PPG signals, a scaling ratio is dynamically assigned to each frequency-domain coefficient according to the amplitude of the frequency-domain coefficient; the scaling ratio is directly proportional to the amplitude; and the frequency-domain signals are non-uniformly scaled based on the scaling factors corresponding to the frequency-domain coefficients.

[0089] That is, a scaling ratio directly proportional to the amplitude of each frequency-domain coefficient is dynamically assigned to the frequency-domain coefficient. For high-amplitude coefficients (such as the direct current component determining the signal baseline and the core frequency representing the main pulse), a larger scaling ratio is assigned, which is equivalent to using a coarser “quantization step” to process the high-amplitude coefficients, thereby achieving “strong compression”. For low-amplitude coefficients, a smaller scaling ratio (i.e., a finer quantization step) is assigned, thereby “retaining more precision” when compressed. This “high-amplitude coarse quantization and low-amplitude fine quantization” strategy essentially optimally allocates limited data storage resources. It ensures that the main energy components of the signal are efficiently compressed with acceptable distortion, while retaining low-amplitude components containing detailed information, thereby maximizing the quality and usability of the reconstructed signal at an overall high compression rate.

[0090] At this time, step 102 divides the plurality of groups of PPG signals into a group of reference signals and at least one group of candidate signals, including: The compressed plurality of groups of PPG signals are divided into a group of reference signals and at least one group of candidate signals.

[0091] In this way, the overall storage and transmission burden of the plurality of groups of PPG signals can be reduced.

[0092] In an optional embodiment, the method further includes: calculating the residual error between the target signal and the candidate signal; determining the compression degree of the residual error based on the signal quality of the candidate signal; the compression degree is inversely proportional to the signal quality; compressing the residual error of the candidate signal based on the compression degree; and storing the compressed residual error.

[0093] Residual, which can be understood as the difference between the "candidate signal" and the "estimated value of the target signal after the conversion relationship processing", the formula is as follows: Candidate signal = (target signal x conversion relationship) + residual The smaller the residual, the more accurate the conversion relationship, and the more similar the candidate signal and the target signal.

[0094] Suppose the value of the target signal at a certain time is 100, and the conversion relationship is (a = 0.9, b = 5), then the estimated value is 0.9*100 + 5 = 95. If the true value of the candidate signal at this time is 97, then the residual is 97-95 = 2. When reconstructing, we can perfectly restore the original value by 95 + 2 = 97. When storing, we only need to store the conversion relationship (0.9, 5) and the residual 2, without storing the complete candidate signal.

[0095] Since the low-quality signal itself contains a large amount of noise, a large part of its residual is also noise, which is not worth saving with high accuracy. Therefore, the compression degree is set to be inversely proportional to the signal quality, and the residual of the low-quality signal can be compressed more aggressively.

[0096] For example, a calculated residual is 22.214.

[0097] If the candidate signal quality is high, it is stored as 22. If the candidate signal quality is low, it is quantized more coarsely and directly stored as 20.

[0098] Based on the same technical concept, the embodiments of the present application also provide a data reconstruction method. As shown in Figure 2 The method comprises the following steps: Step 201, obtaining a signal to be processed; the signal to be processed is obtained based on the data storage method described above; the signal to be processed at least comprises a reference signal, and a conversion relationship between a target signal and a candidate signal, the target signal being the reference signal or a candidate signal whose conversion relationship has been calculated.

[0099] Step 202, based on the target signal, the conversion relationship between the target signal and the candidate signal to be reconstructed, reconstructing each group of candidate signals.

[0100] The contents involved in the above steps 201-202 are actually the inverse process of the data storage method described above, and the core goal is to use the stored compressed data (reference signal, conversion relationship, etc.) to reconstruct the original waveform of all candidate signals with high fidelity.

[0101] The process of data reconstruction is described in an optimal embodiment: Reading selected frequency indices: First, determine which frequency components need to be reconstructed. During compression, only the core frequency components may be saved to remove noise. At this time, the system reads the indices of these saved frequency points, focuses on the core frequency components, and ignores irrelevant or noise frequency bands, laying the foundation for accurate reconstruction.

[0102] Loading stored spectral coefficients: Read the compressed frequency domain data for these key frequency points, mainly including amplitude and phase information. This is the "raw material" for reconstructing the signal.

[0103] Generating a complete base spectrum spectrum by zero padding: Since only part of the key frequency points are saved during storage, it is now necessary to fill in the discarded frequency points by zero padding to reconstruct a complete length spectrum. This step is to prepare for the subsequent inverse Fourier transform, to ensure correct conversion back to the time domain signal.

[0104] Reconstructing the spectra of other channels layer by layer using stored linear conversion parameters and residuals: This is the key to achieving efficient compression. Reconstruction is not done independently, but follows the dependency chain established during storage: starting from the complete spectrum of the reference signal. Apply the stored conversion relationship (such as the parameters of the linear model) to convert the reference signal spectrum into the estimated spectrum of the first candidate signal. Load the stored compressed residual for this candidate signal and add it to the estimated spectrum to accurately reconstruct the complete spectrum of the first candidate signal. Take this newly reconstructed candidate signal as the new "benchmark" and repeat the process, layer by layer, until the complete spectra of all channels are reconstructed. This greatly improves the overall compression efficiency.

[0105] Obtain and apply scaling factors to de-scale the spectrum: Reverse the normalization or quantization operations performed during compression to save storage space, restore the spectrum values to their original magnitude, and correct the amplitude scaling caused by compression.

[0106] Perform inverse Fourier transform to restore time domain signal: Convert each complete spectrum reconstructed in the above steps from frequency domain to time domain through inverse fast Fourier transform to obtain reconstructed signals for each time segment.

[0107] Discard overlapping parts and splice windows: Reverse the "overlapping segmentation" operation used during compression. Discard the overlapping parts at the beginning and end of each time segment, and only keep the most effective middle part. Then, smoothly splice these effective parts together to finally recover the original waveforms of all PPG channels completely and continuously.

[0108] In this way, the original signals of the stored multiple groups of PPG signals can be reconstructed.

[0109] Based on the same technical concept, the embodiments of the present application also provide a data storage device, such as Figure 3As shown, the device comprises: The acquisition module 301 is configured to acquire a plurality of groups of photoplethysmography (PPG) signals; there is a partial difference between at least two groups of PPG signals.

[0110] The distinguishing module 302 is configured to distinguish the plurality of groups of PPG signals into one group of reference signals and at least one group of candidate signals; the signal quality of the reference signals is higher than the signal quality of the candidate signals.

[0111] The determination module 303 is configured to determine, for each group of candidate signals, a target signal corresponding to the candidate signal from the plurality of groups of PPG signals; the similarity between the target signal and the candidate signal meets a preset requirement, and the target signal is the reference signal or a candidate signal for which a conversion relationship has been calculated.

[0112] The calculation module 304 is configured to calculate a conversion relationship between the target signal and the candidate signal; the conversion relationship is used to reconstruct the candidate signal based on the target signal.

[0113] The storage module 305 is configured to store the reference signal and the conversion relationship between each group of candidate signals and the target signal corresponding thereto.

[0114] In a feasible implementation, the plurality of groups of PPG signals are acquired by at least one acquisition device at different acquisition positions and / or based on different colors of light.

[0115] The storage module is configured to store the reference signal, and is configured to: Determine a core frequency component of each group of PPG signals to obtain a target frequency band of the plurality of groups of PPG signals; the target frequency band includes each core frequency component.

[0116] Band-pass filter the reference signal based on the set of core frequency components.

[0117] Store the reference signal after filtering.

[0118] In a feasible implementation, the device further comprises: The window module is configured to divide each group of PPG signals into a plurality of windows; there is at least partial overlap between any two windows.

[0119] The frequency domain conversion module is configured to, for each window, perform Fourier transform on data in the window to obtain a frequency domain signal.

[0120] The compression module is configured to compress each group of PPG signals based on the frequency domain signal to obtain a plurality of groups of compressed PPG signals.

[0121] The distinguishing module is configured to distinguish a plurality of groups of PPG signals into one group of reference signals and at least one group of candidate signals, for: The compressed plurality of groups of PPG signals are distinguished into one group of reference signals and at least one group of candidate signals.

[0122] In a feasible implementation, the compression module is configured to compress each group of PPG signals based on the frequency domain signal, for: For the frequency domain signal of each group of PPG signals, a scaling ratio is dynamically assigned to each frequency domain coefficient according to the amplitude of the frequency domain coefficient; the scaling ratio is directly proportional to the amplitude.

[0123] The frequency domain signal is non-uniformly scaled based on the scaling factor corresponding to each frequency domain coefficient.

[0124] In a feasible implementation, the calculation module is configured to calculate a conversion relationship between the target signal and the candidate signal, for: For each group of candidate signals, when the target signal is a reference signal, a conversion relationship between the reference signal and the group of candidate signals is calculated based on a preset algorithm; the preset algorithm includes a generalized linear model and / or cross-correlation analysis.

[0125] A first signal obtained based on the conversion relationship and the reference signal is determined as a candidate signal for which the conversion relationship has been calculated.

[0126] Alternatively, for each group of candidate signals, when the target signal is a first signal, a conversion relationship between the first signal and the group of candidate signals is calculated based on the preset algorithm.

[0127] A second signal obtained based on the conversion relationship and the first signal is determined as a candidate signal for which the conversion relationship has been calculated.

[0128] Alternatively, for each group of candidate signals, when the target signal is a second signal, a conversion relationship between the second signal and the group of candidate signals is calculated based on the preset algorithm.

[0129] In a feasible implementation, the calculation module is further configured to: Calculate a residual error between the target signal and the candidate signal.

[0130] Further comprising: The first module is configured to determine a compression degree of the residual error based on a signal quality of the candidate signal; the compression degree is inversely proportional to the signal quality.

[0131] The second module is configured to compress the residual error of the candidate signal based on the compression degree.

[0132] The compression module is further configured to store the compressed residual.

[0133] In one possible implementation, the embodiments of the present application also provide a data reconstruction device, which comprises Figure 4 As shown in the figure, the device comprises: A signal acquisition module 401 is configured to acquire a to-be-processed signal, wherein the to-be-processed signal is obtained by storing the method according to any one of the first aspect; and the to-be-processed signal comprises at least a reference signal and a conversion relationship between a target signal and a candidate signal, wherein the target signal is the reference signal or a candidate signal for which the conversion relationship has been calculated.

[0134] A reconstruction module 402 is configured to reconstruct each group of candidate signals based on the target signal and the conversion relationship between the target signal and the candidate signal to be reconstructed.

[0135] Figure 5 A structural diagram of an electronic device provided by the embodiments of the present application comprises a processor 501, a storage medium 502 and a bus 503, the storage medium 502 stores machine readable instructions executable by the processor 501, when the electronic device runs the method in the embodiments, the processor 501 and the storage medium 502 communicate through the bus 503, and the processor 501 executes the machine readable instructions to perform the steps in the embodiments.

[0136] In the embodiments, the storage medium 502 can also execute other machine readable instructions to perform the methods described in the embodiments, for the specific method steps and principles, refer to the description of the embodiments, and will not be described in detail here.

[0137] The embodiments of the present application also provide a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to perform the steps in the embodiments.

[0138] In the embodiments of the present application, the computer program executed by the processor can also execute other machine readable instructions to perform the methods described in the embodiments, for the specific method steps and principles, refer to the description of the embodiments, and will not be described in detail here.

[0139] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented by other manners. The above-described device embodiments are merely illustrative, for example, the division of the modules is merely a logical function division, and actual implementation can have another division manner, for example, a plurality of modules or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed modules can be indirect coupling or communication connection through some communication interfaces, devices or modules, which can be electrical, mechanical or other forms.

[0140] The modules described as separate components can or can not be physically separated, and the components displayed as modules can or can not be physical units, i.e., can be located in one place or distributed to a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the present embodiment.

[0141] In addition, the functional units in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit.

[0142] If the functions are realized in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer readable storage medium executable by a processor. Based on this understanding, the technical solutions of the present application or the part of the present application that essentially contributes to the prior art or the part of the technical solutions can be embodied in the form of a software product, which is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The foregoing storage medium includes: U disk, mobile hard disk, ROM, RAM, magnetic disk or optical disk, and various program code storage media.

[0143] The above is merely a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for data storage, characterized in that, The method includes: Acquire multiple sets of photoplethysmography (PPG) signals; at least two sets of PPG signals show some differences. Multiple PPG signals are divided into a reference signal and at least one candidate signal; the signal quality of the reference signal is higher than that of the candidate signal. For each set of candidate signals, a target signal corresponding to the candidate signal is determined from the multiple sets of PPG signals; the similarity between the target signal and the candidate signal meets a preset requirement, and the target signal is the reference signal or a candidate signal whose transformation relationship has been calculated; Calculate the conversion relationship between the target signal and the candidate signal; the conversion relationship is used to reconstruct the candidate signal based on the target signal. The reference signal and the conversion relationship between each group of candidate signals and their corresponding target signal are stored.

2. The method according to claim 1, characterized in that, The multiple sets of PPG signals are obtained by at least one acquisition device at different acquisition locations, and / or based on light of different colors; Storing the reference signal includes: The core frequency component of each group of PPG signals is determined to obtain target frequency bands for multiple groups of PPG signals; each target frequency band includes each core frequency component. The reference signal is bandpass filtered based on the core frequency component set; Store the filtered reference signal.

3. The method according to claim 1, characterized in that, The method further includes: For each group of PPG signals, it is divided into multiple windows; any two windows have at least partial overlap. For each window, perform a Fourier transform on the data in the window to obtain the frequency domain signal; Based on the frequency domain signal, each group of PPG signals is compressed to obtain multiple compressed groups of PPG signals. Multiple PPG signals are divided into a reference signal and at least one candidate signal, including: The compressed PPG signals are divided into a reference signal and at least one candidate signal.

4. The method according to claim 3, characterized in that, Based on the frequency domain signal, each group of PPG signals is compressed, including: For each group of PPG signals in the frequency domain, a scaling factor is dynamically assigned to it based on the amplitude of each frequency domain coefficient; the scaling factor is proportional to the amplitude. The frequency domain signal is non-uniformly scaled based on the scaling factor corresponding to each frequency domain coefficient.

5. The method according to claim 1, characterized in that, Calculating the transformation relationship between the target signal and the candidate signal includes: For each group of candidate signals, when the target signal is a reference signal, the transformation relationship between the reference signal and the group of candidate signals is calculated based on a preset algorithm; the preset algorithm includes: a generalized linear model and / or cross-correlation analysis; The first signal obtained based on the transformation relationship and the reference signal is determined as a candidate signal for which the transformation relationship has been calculated; Alternatively, for each group of candidate signals, when the target signal is the first signal, the conversion relationship between the first signal and the group of candidate signals is calculated based on the preset algorithm; The second signal obtained based on the transformation relationship and the first signal is determined as a candidate signal for which the transformation relationship has been calculated; Alternatively, for each group of candidate signals, when the target signal is the second signal, the conversion relationship between the second signal and the group of candidate signals is calculated based on the preset algorithm.

6. The method according to claim 5, characterized in that, The method further includes: Calculate the residual between the target signal and the candidate signal; Based on the signal quality of the candidate signal, the degree of compression of the residual is determined; the degree of compression is inversely proportional to the signal quality. Based on the compression level, the residual of the candidate signal is compressed; Store the compressed residual.

7. A method for data reconstruction, characterized in that, The method includes: Acquire a signal to be processed; the signal to be processed is stored based on the method described in any one of claims 1-6; the signal to be processed includes at least: a reference signal, and a conversion relationship between a target signal and a candidate signal, wherein the target signal is the reference signal, or a candidate signal whose conversion relationship has been calculated; Based on the target signal and the conversion relationship between the target signal and the candidate signal to be reconstructed, each group of candidate signals is reconstructed.

8. A data storage device, characterized in that, The device includes: The acquisition module is used to acquire multiple sets of photoplethysmography (PPG) signals; at least two sets of PPG signals have some differences between them. A differentiation module is used to differentiate multiple sets of PPG signals into a reference signal and at least one set of candidate signals; the signal quality of the reference signal is higher than that of the candidate signal. The determination module is used to determine the target signal corresponding to the candidate signal from the plurality of PPG signals for each candidate signal; the similarity between the target signal and the candidate signal meets a preset requirement, and the target signal is the reference signal or a candidate signal whose transformation relationship has been calculated; The calculation module is used to calculate the conversion relationship between the target signal and the candidate signal; the conversion relationship is used to reconstruct the candidate signal based on the target signal; A storage module is used to store the reference signal and the conversion relationship between each group of candidate signals and their corresponding target signal.

9. An electronic device, characterized in that, include: The device includes a processor, a storage medium, and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, and when the electronic device is in operation, the processor communicates with the storage medium via the bus, and the processor executes the machine-readable instructions to perform the steps of the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, performs the steps of the method as described in any one of claims 1 to 7.

Images