A Compression Algorithm for Restoring Signal Features

Abstract

The present application discloses a restorable signal feature compression algorithm for compressing the pulse wave signal collected by a wearable monitor, and the method specifically includes an initialization step, a signal collection and compression step. The present application also discloses a wearable monitor, including an acquisition module, a state judgment module, a compression module, a storage module and a processor; wherein the acquisition module is configured to collect the patient's pulse wave signal and acceleration information; the state judgment module determines whether the patient is in a sleeping state, a moving state or a resting state according to the pulse wave signal and the acceleration information; the compression module compresses the pulse wave signal and stores the compressed pulse wave signal in the storage module; the processor is connected to the acquisition module, the state judgment module, the compression module and the storage module, and the processor is configured to control the acquisition module, the state judgment module, the compression module and the storage module to execute the restorable signal feature compression algorithm.

Description

Technical Field

[0001] The present application relates to a compression algorithm in the field of pulse wave measurement, and in particular to a compression algorithm capable of restoring signal features, a wearable monitor, and a monitoring system. Background Art

[0002] Pulse waves are an important signal source for human health monitoring and have become a key monitoring target for health monitors. Current wearable monitors tend to integrate pulse wave signal monitoring and analysis functions, as well as complex compression algorithms (such as machine learning). However, the applicant believes that this is detrimental to the lightweight nature of wearable monitors. These functions, along with complex compression algorithms, require more complex processors, which in turn increase cost, power consumption, and battery requirements, leading to discomfort for patients.

[0003] At the same time, the existing pulse wave compression algorithm has the following problems: the compression algorithm cannot be adjusted for different patients, nor can it be adjusted for different states of the same patient, resulting in room for improvement in compression efficiency; it also cannot be dynamically adjusted according to the storage space during the compression process, and the storage space cannot be fully utilized to obtain more accurate compressed data; the monitor is prone to overheating during the compression process, thereby ignoring the patient's comfort. Summary of the Invention

[0004] The purpose of this application is to provide a signal feature compression algorithm that can be restored, a wearable monitor and a monitoring system to better solve the problems in the background technology.

[0005] To achieve the above objectives, this application provides the following technical solutions:

[0006] A signal feature compression algorithm capable of restoring signal characteristics is used to compress pulse wave signals collected by a wearable monitor, wherein the method specifically includes an initialization step, a signal collection step, and a compression step;

[0007] The initialization steps specifically include:

[0008] Step A1: The hospital's data processing device obtains basic information about the patient and generates a pulse wave feature library based on the patient's basic information; the pulse wave feature library includes a sleep state pulse wave feature sub-library, an exercise state pulse wave feature sub-library, and a rest state pulse wave feature sub-library; the data processing device obtains the patient's monitoring period T;

[0009] Step A2: The data processing device sends the waveform feature library and the total monitoring time T to the wearable monitor worn by the patient;

[0010] The signal acquisition and compression steps specifically include:

[0011] Step B1, the acquisition module of the wearable monitor acquires the patient's pulse wave signal and acceleration signal;

[0012] Step B2, the state judgment module judges the patient's state according to the pulse wave signal and acceleration signal;

[0013] Step B3, the compression module selects the corresponding pulse wave feature sub-library according to the patient's state, and compresses the pulse wave signal.

[0014] Step B4, the compression module stores the compressed pulse wave signal in the corresponding state area of the storage module;

[0015] Among them, the specific steps of compressing the pulse wave signal in step B3 further include the following steps:

[0016] Step S1, set the threshold Z = Z1, the first threshold Z1 is a preset value and is negatively correlated with the reduction degree of compression;

[0017] Step S2, perform component decomposition on the pulse wave signal Select the most matching feature from the standard pulse wave waveform features of the pulse wave feature sub-library corresponding to the patient's state and the pulse wave waveform features corresponding to the disease suffered Make the smallest, where ;

[0018] Step S3, judge whether <Z holds; if not, execute step S4; if so, execute step S5;

[0019] Step S4, select the most matching feature from the pulse wave component features of the pulse wave feature sub-library corresponding to the patient's state Make the smallest, where , represents the pulse wave component feature selected in the jth execution of step S4, and N is the total number of executions of step S4; jump to execute step S3;

[0020] Step S5, encode the features selected in each step as the compressed pulse wave signal.

[0021] Furthermore, the signal feature compression algorithm that can be restored further includes a parameter update step, which specifically includes:

[0022] Step C1, lower the second threshold Z2 by one level, and use the adjusted second threshold Z2 to replace Z1 in step S1;

[0023] Step C2, after the compression algorithm runs for a preset time, obtain the storage space speed v of the current compression algorithm, the current remaining storage space C, and the remaining monitoring period Tr;

[0024] Step C3, determine Is it established? If so, go to step C4;

[0025] In step C4 , the second threshold Z2 is increased by one level, and the adjusted second threshold Z2 is used to replace Z1 in step S1 .

[0026] The second threshold Z2 is a preset value and has multiple gears from small to large.

[0027] Furthermore, the restorable signal feature compression algorithm also includes a parameter updating step, specifically including:

[0028] Step D1, lowering the second threshold Z2 by one level, and using the adjusted second threshold Z2 to replace Z1 in step S1;

[0029] Step D2: After the compression algorithm runs for a certain period of time, the current contact temperature between the wearable monitor and the patient's body surface is obtained;

[0030] Step D3, determining whether the contact temperature exceeds a preset temperature threshold; if so, executing step D4;

[0031] In step D4 , the second threshold Z2 is adjusted higher by one level, and the adjusted second threshold Z2 is used to replace Z1 in step S1 .

[0032] The second threshold Z2 is a preset value and has multiple gears from small to large.

[0033] Furthermore, the patient's basic information includes age and diseases.

[0034] And a wearable monitor, characterized in that,

[0035] Wearable monitors do not have the function of analyzing whether the pulse wave signal is normal;

[0036] It includes an acquisition module, a state judgment module, a compression module, a storage module and a processor;

[0037] Among them, the acquisition module is configured to collect the patient's pulse wave signal and acceleration information; the state judgment module determines whether the patient is in a sleeping state, an exercise state or a resting state based on the pulse wave signal and acceleration information; the compression module compresses the pulse wave signal and stores the compressed pulse wave signal in the storage module; the processor is connected to the acquisition module, the state judgment module, the compression module, and the storage module, and the processor is configured to control the acquisition module, the state judgment module, the compression module, and the storage module to execute the above-mentioned restorable signal feature compression algorithm.

[0038] And a monitoring system, characterized in that,

[0039] Including the above-mentioned wearable monitor worn by the patient and the hospital's data processing device;

[0040] Among them, the data processing device is configured to receive the compressed pulse wave signal sent by the wearable monitor, restore it to a pulse wave signal, analyze whether the restored pulse wave signal is normal, update the pulse wave history database, and send the updated waveform feature library and the updated total monitoring time T to the wearable monitor worn by the patient for a new monitoring cycle.

[0041] Compared with the prior art, the innovations and beneficial effects of this application are:

[0042] 1. Separating the pulse wave signal analysis function of the monitoring system significantly reduces the power consumption and battery requirements of the wearable monitor, thereby improving the portability of the wearable monitor.

[0043] 2. Adjust the compression algorithm according to the patient's basic information to make it personalized.

[0044] 3. The compression algorithm is adjusted according to the patient's condition to further improve the compression efficiency.

[0045] 4. Dynamically adjust the compression process according to the storage space to avoid the compressed data amount exceeding the storage space, while making full use of the storage space to obtain higher-precision compressed data.

[0046] 5. Dynamically adjust the calculation amount of the compression process to avoid overheating of the monitor and improve patient comfort. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a schematic diagram. Implementation Method

[0048] To monitor patients, monitors are generally used to continuously monitor various physiological parameters of patients. For some patients who do not require hospitalization but still require monitoring, wearable monitors are often used for monitoring. Current wearable monitors tend to integrate pulse wave signal monitoring and analysis functions, as well as complex compression algorithms (such as machine learning). However, the applicant believes that this is not conducive to the lightweighting of wearable monitors. The pulse wave signal monitoring and analysis functions, as well as complex compression algorithms, require more complex processors, which in turn increase costs, power consumption, and battery requirements, resulting in discomfort for patients. Relying solely on simple compression algorithms cannot meet the needs of pulse wave signal compression.

[0049] Combine Figure 1As shown, the present application provides a wearable monitor, comprising an acquisition module, a state determination module, a compression module, a storage module, and a processor. The acquisition module is configured to acquire a patient's pulse wave signal and acceleration information; the state determination module determines whether the patient is in a sleeping state, an exercise state, or a resting state based on the pulse wave signal and acceleration information; the compression module compresses the pulse wave signal and stores the compressed pulse wave signal in the storage module; and the processor is connected to the acquisition module, the state determination module, the compression module, and the storage module, and is configured to control the acquisition module, the state determination module, the compression module, and the storage module to execute the following restorable signal feature compression algorithm.

[0050] The wearable monitor, the acquisition module is configured to acquire the patient's pulse wave signal, which can obtain the pulse wave signal through the common photoelectric volume pulse wave method, pressure measurement method, audio sensing and other methods in the prior art. The optional acquisition module can also acquire the patient's acceleration information, and / or electrocardiogram signal, and / or contact temperature with the patient's body surface. The acquisition module can collect the above signals by setting corresponding sensors. The state judgment module determines whether the patient is in a sleeping state, an exercise state or a resting state based on the pulse wave signal and acceleration information. The exercise state and the resting state are states when the patient is not sleeping, and the exercise state and the resting state can be distinguished by preset heart rate or acceleration information.

[0051] Furthermore, the wearable monitor lacks pulse wave signal analysis capabilities. That is, the wearable monitor does not need to analyze the patient's current pulse wave signal to determine if it is normal. Instead, upon the patient's return visit, the pulse wave data for the current monitoring cycle is transmitted to the hospital's higher-performance data processing device, which then analyzes the pulse wave signal. This significantly reduces the wearable monitor's power consumption and battery requirements, thereby improving its portability. However, this also requires the wearable monitor to perform better compression of the pulse wave signal and restore its signal characteristics to facilitate analysis by the data processing device.

[0052] In this regard, the present application also provides a restorable signal feature compression algorithm, which is specifically applied to the wearable monitor of the above-mentioned monitoring system to perform restorable signal feature compression on the pulse wave signal collected by the wearable monitor. The method specifically includes an initialization step, a signal acquisition and compression step.

[0053] The initialization steps specifically include:

[0054] Step A1: The hospital's data processing device obtains basic information about the patient and generates a pulse wave feature library based on the patient's basic information; the pulse wave feature library includes a sleep state pulse wave feature sub-library, an exercise state pulse wave feature sub-library, and a rest state pulse wave feature sub-library; the data processing device obtains the patient's monitoring period T;

[0055] In step A2, the data processing device sends the waveform feature library and the total monitoring time T to the wearable monitor worn by the patient.

[0056] Specifically, the data processing device obtains basic patient information, including age and illness, and forms a pulse wave feature library based on the patient's basic information and the pulse wave history database. This library can be generated by searching the pulse wave history database for pulse wave features that match the patient's basic information, or by modeling or machine learning based on the patient's basic information and the pulse wave history database. The pulse wave history database contains pulse wave data from the patient's previous monitoring cycles (if any), pulse wave data from other patients with the same illness, and standard pulse wave data for the patient's age. The amount of data contained in the pulse wave feature library can be adjusted based on the storage space available on the wearable monitor.

[0057] The pulse wave feature library includes a sleep state pulse wave feature sub-library, an exercise state pulse wave feature sub-library, and a rest state pulse wave feature sub-library. Each state pulse wave feature sub-library includes the standard pulse wave waveform characteristics of the age in the corresponding state, the pulse wave waveform characteristics corresponding to the disease suffered, and the pulse wave component characteristics, and each of the above features is multiple. Optionally, the pulse wave feature library includes at least the above three state pulse wave feature sub-libraries, and may also include more state pulse wave feature sub-libraries, and their number and state content can correspond to the state judgment module of the wearable monitor. Therefore, by adjusting the compression algorithm according to the patient's basic information, the compression algorithm is personalized, and by adjusting the compression algorithm according to the patient's state, the compression efficiency is further improved.

[0058] At the same time, the data processing device obtains the patient's total monitoring time T, which represents the duration of the current monitoring cycle, that is, the time until the next patient follow-up visit. The parameters of the subsequent compression steps are adjusted according to the monitoring cycle T.

[0059] The signal acquisition and compression steps specifically include:

[0060] Step B1, the acquisition module of the wearable monitor acquires the patient's pulse wave signal and acceleration signal;

[0061] Step B2: The status determination module determines the patient's status based on the pulse wave signal and the acceleration signal;

[0062] Step B3: The compression module selects the pulse wave feature sub-library corresponding to the patient's state and compresses the pulse wave signal.

[0063] Step B4: The compression module stores the compressed pulse wave signal in the corresponding state area of the storage module.

[0064] Specifically, when collecting the pulse wave signal and the acceleration signal, optionally, the collected signals are first subjected to conventional processing such as periodic segmentation, and / or filtering, and / or noise reduction, and / or normalization. Optionally, the acquisition module can also collect the patient's electrocardiogram signal to assist in the segmentation process. The state judgment module judges whether the patient is in a sleep state, a motion state or a resting state according to the pulse wave signal and the acceleration information. Optionally, the patient's state can be further subdivided according to actual needs, and the judgment of the patient's state can be made according to the conventional judgment method of the prior art. The compression module selects the pulse wave feature sub-library corresponding to the patient's state and compresses the pulse wave signal. The specific compression algorithm can adopt the commonly used compression algorithms in the prior art, but preferably, the applicant also proposes a signal feature reducible compression algorithm. The specific steps for compressing the pulse wave signal in step B3 further include the following steps:

[0065] Step S1: Set the threshold Z = Z1, where the first threshold Z1 is a preset value and is negatively correlated with the reduction degree of compression;

[0066] Step S2: Decompose the pulse wave signal Select the most matching feature from the standard pulse wave waveform features of the pulse wave feature sub-library corresponding to the patient's state and the pulse wave waveform features corresponding to the disease suffered such that is minimized, where ;

[0067] Step S3: Judge whether

[0068] < Z holds; if not, execute step S4; if so, execute step S5;

[0069] Step S4: Select the most matching feature from the pulse wave component features of the pulse wave feature sub-library corresponding to the patient's state such that is minimized, where and represents the pulse wave component feature selected in the j-th execution of step S4, and N is the total number of executions of step S4; then jump to execute step S3;

[0070] Step S5: Encode the features selected in each step as the compressed pulse wave signal.

[0071] Specifically, in step S1, a threshold Z=Z1 is set. The first threshold Z1 is a preset value and is negatively correlated with the degree of compression restoration. The compression process of the signal is controlled in combination with the first threshold Z1.

[0072] In step S2, the pulse wave signal is decomposed into components, and the detected pulse wave signal is expressed as , which can be expressed as Among them, K represents Decompose into K components , After decomposition into K components The residual value, it can be seen that the larger K is, the greater the residual value is. The smaller it is, the K components are used to represent The more accurate it is; Respectively represent the parameters of the amplitude, time, duration, frequency, and phase of the i-th component. In the same way, the pulse wave waveform features and pulse wave component features of the pulse wave feature library can also be expressed. The difference is that the number of components is fixed. The pulse wave waveform features are represented as the sum of multiple components, while the pulse wave waveform component features of the pulse wave feature library only contain one component. Moreover, since the features of the pulse wave feature library are known and quantified, they do not require residual values. Specifically, the standard pulse wave waveform features of the pulse wave feature library and the pulse wave waveform features corresponding to the disease can be represented as , the pulse wave waveform component characteristics of the pulse wave feature library can be represented as Therefore, a set of To represent a component, the standard pulse waveform features of the pulse wave feature library and the pulse waveform features corresponding to the disease can be represented by multiple groups. To represent and store, the pulse wave waveform component features of the pulse wave feature library can be represented by a set of to represent and store.

[0073] Select the most matching feature from the standard pulse waveform features of the pulse wave feature sub-library corresponding to the patient's condition and the pulse waveform features corresponding to the disease , here through To judge the degree of compression restoration, other conventional methods can also be used to judge the degree of compression restoration, among which After selecting the most matching feature, the storage location of the feature in the pulse wave feature library can be recorded. Further, optionally, fitting parameters representing amplitude, time, duration, and phase can be used. Pair Features Fitting is performed so that Each component All fit to However, in general, in order to reduce the computational complexity of the fitting process, only the fitting parameters After completing the fitting, you can also record the fitting parameters used.

[0074] Similarly, in step S4, the most matching feature is selected from the pulse wave component features of the pulse wave feature sub-library corresponding to the patient's state. Make Minimum, among which , represents the pulse wave component feature selected by executing step S4 for the jth time, and N is the total number of times step S4 is executed. After selecting the most matching feature, the storage location of the feature in the pulse wave feature library can be recorded. Further, optionally, the fitting parameters representing amplitude, time, duration, and phase can be used. Pair Features Fitting so that However, in general, in order to reduce the computational complexity of the fitting process, only the fitting parameters After completing the fitting, you can also record the fitting parameters used.

[0075] In step S5, the features selected in each step are encoded. Each selected feature can be encoded in a storage location within the pulse wave feature library. If fitting parameters are used, the storage location of each selected feature within the pulse wave feature library, along with the fitting parameters used for each feature, can be encoded. Common lossless encoding methods used in the prior art can be used for encoding, and the encoded data is the compressed pulse wave signal.

[0076] Optionally, the restorable signal feature compression algorithm may further include a parameter updating step, specifically including:

[0077] Step C1, lowering the second threshold Z2 by one level, and using the adjusted second threshold Z2 to replace Z1 in step S1;

[0078] Step C2: After the compression algorithm runs for a preset time, the storage space speed v, the current remaining storage space C, and the remaining monitoring period Tr of the current compression algorithm are obtained;

[0079] Step C3, determine Is it established? If so, go to step C4;

[0080] In step C4 , the second threshold Z2 is increased by one level, and the adjusted second threshold Z2 is used to replace Z1 in step S1 .

[0081] The second threshold Z2 is a preset value and has multiple gears from small to large. The value of the maximum gear should meet the restoration requirement of analyzing the pulse wave signal.

[0082] Specifically, to obtain an accurate pulse wave signal for subsequent analysis and detection, a lower threshold needs to be set in the compression algorithm. However, a lower threshold will increase the amount of data. If the threshold is always kept low, the data volume may exceed the storage space limit. At the same time, during the pulse wave signal compression process, for the same threshold, the compression ratio of the compression algorithm may also change due to changes in the pulse wave signal. Therefore, it is necessary to dynamically adjust the threshold in the compression algorithm after each certain period of operation. This operation time can be preset according to actual needs. Dynamic adjustments are made during the compression process based on the storage space to prevent the compressed data volume from exceeding the storage space, while fully utilizing the storage space to obtain more accurate compressed data.

[0083] In step C2, the storage space rate v of the current compression algorithm is obtained. This rate refers to the ratio of the storage space occupied by the compressed pulse wave signal during the preset compression algorithm execution time to the preset execution time, reflecting the increase in storage space occupied per unit time. The current remaining storage space C is obtained. This refers to the remaining storage space in the wearable monitor's storage module after the compression algorithm has executed for the preset execution time. The remaining monitoring period Tr is obtained. This refers to the difference between the total monitoring duration T and the monitored duration.

[0084] This application also proposes an alternative parameter updating step, which specifically includes:

[0085] Step D1, lowering the second threshold Z2 by one level, and using the adjusted second threshold Z2 to replace Z1 in step S1;

[0086] Step D2: After the compression algorithm runs for a certain period of time, the current contact temperature between the wearable monitor and the patient's body surface is obtained;

[0087] Step D3, determining whether the contact temperature exceeds a preset temperature threshold; if so, executing step D4;

[0088] In step D4 , the second threshold Z2 is adjusted higher by one level, and the adjusted second threshold Z2 is used to replace Z1 in step S1 .

[0089] The second threshold Z2 is a preset value and has multiple gears from small to large. The value of the maximum gear should meet the restoration requirement of analyzing the pulse wave signal.

[0090] Specifically, to obtain an accurate pulse wave signal for subsequent analysis and detection of the pulse wave signal, a lower threshold needs to be set in the compression algorithm. However, a lower threshold will increase the amount of computation, which will increase the power consumption of the wearable monitor, cause heat, and cause discomfort to the patient. At the same time, during the compression of the pulse wave signal, for the same threshold, the amount of computation required by the compression algorithm may also change due to changes in the pulse wave signal. Therefore, it is necessary to dynamically adjust the threshold in the compression algorithm after each certain period of operation. This running time can be preset according to actual needs. Dynamically adjusting the computational complexity of the compression process prevents the monitor from overheating and improves patient comfort.

[0091] In step D2, the current contact temperature between the wearable monitor and the patient's body surface is obtained. This can be achieved by installing a temperature sensor at the contact point between the wearable monitor and the patient's body surface. If the contact temperature exceeds a preset temperature threshold, the second threshold Z2 is increased.

[0092] The present application also provides a monitoring system comprising a wearable monitor worn by a patient and a data processing device in a hospital. The data processing device is configured to receive a compressed pulse wave signal transmitted by the wearable monitor, restore the compressed pulse wave signal to a pulse wave signal, analyze whether the restored pulse wave signal is normal, update a pulse wave history database, and transmit the updated waveform feature database and the updated total monitoring time T to the wearable monitor worn by the patient for a new monitoring cycle.

[0093] Among them, restoring the pulse wave signal is the inverse operation process of the above-mentioned restorable signal feature compression algorithm, that is, decoding the compressed pulse wave signal, obtaining the storage location of each feature in the pulse wave feature library and the fitting parameters used for each feature. Since the data processing device records the waveform feature library information sent to the wearable monitor, it can obtain the pulse wave signal by combining the storage location of each feature in the pulse wave feature library and the fitting parameters used for each feature. Components, after combination, can get the pulse wave signal with signal characteristics , combined with the inverse processing operations of period segmentation, and / or filtering, and / or noise reduction, and / or normalization, the patient's restored pulse wave signal is obtained.

[0094] In the prior art, wearable monitors often send pulse wave signals to data processing devices sequentially. Due to the large amount of data, doctors and patients often have to wait a long time for the data processing device to receive all the pulse wave signals and analyze them to obtain results. As another improvement to the present application, the monitoring system may optionally include the wearable monitor being configured to send a portion of the pulse wave signals for each of the sleep, exercise, and rest states during the entire monitoring cycle, and then send the remaining data for each day. For example, the wearable monitor first sends 20% of the sleep state pulse wave signals, 20% of the exercise state pulse wave signals, and 20% of the rest state pulse wave signals for the first day, and then sends 20% of the sleep state pulse wave signals, 20% of the exercise state pulse wave signals, and 20% of the rest state pulse wave signals for the second day. After sending 20% ​​of the pulse wave signals for each day, the wearable monitor sends the remaining 80% of the pulse wave signals. The data processing device analyzes the pulse wave signals while receiving them. This increases the chances of detecting abnormal signals earlier, saving time for both doctors and patients.

[0095] It can be understood that in the embodiments of the present application, the size of the serial numbers of the above-mentioned steps does not mean the order of execution. At the same time, the above-mentioned steps can also be combined arbitrarily. The execution order and combination of each step should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

Claims

1. A signal feature compression algorithm capable of restoring signal characteristics, used for compressing pulse wave signals collected by a wearable monitor, characterized in that: The method specifically includes an initialization step, a signal acquisition and compression step; The initialization steps specifically include: Step A1, the data processing device of the hospital obtains the basic information of the patient, and generates a pulse wave feature library according to the basic information of the patient; the pulse wave feature library includes a sleep state pulse wave feature sub-library, a motion state pulse wave feature sub-library, and a rest state pulse wave feature sub-library; the data processing device obtains the total monitoring time T of the patient; Step A2, the data processing device sends the pulse wave feature library and the total monitoring time T to the wearable monitor worn by the patient; The signal acquisition and compression steps specifically include: Step B1, the collection module of the wearable monitor collects the patient's pulse wave signal and acceleration signal; Step B2, the state determination module determines the state of the patient according to the pulse wave signal and the acceleration signal; Step B3, the compression module selects a pulse wave feature sub-library of a corresponding state according to the patient's state, and compresses the pulse wave signal; Step B4, the compression module stores the compressed pulse wave signal in the corresponding state area of ​​the storage module; The step B3 of compressing the pulse wave signal specifically includes the following steps: Step S1, setting a threshold Z=Z1, where the first threshold Z1 is a preset value and is negatively correlated with the degree of compression restoration; Step S2, decomposing the pulse wave signal x(t) into components; selecting the most matching feature xa(t) from the standard pulse wave waveform features of the pulse wave feature sublibrary corresponding to the patient's state and the pulse wave waveform features corresponding to the disease suffered by the patient so that minimum, where r(t) = x(t) - xa(t); Step S3, determine Is it true? If not, go to step S4; if true, go to step S5; Step S4, selecting the most matching feature xb from the pulse wave component features of the pulse wave feature sublibrary corresponding to the patient's state j (t) make Minimum, among which xb j (t) represents the pulse wave component feature selected by executing step S4 for the jth time, and N is the total number of times step S4 is executed; jump to execute step S3; Step S5, encoding the features selected in each step as a compressed pulse wave signal.

2. The restorable signal feature compression algorithm according to claim 1, characterized in that: The restorable signal feature compression algorithm also includes a parameter updating step, specifically including: Step C1, lowering the second threshold value Z2 by one level, and using the adjusted second threshold value Z2 to replace Z1 in step S1; Step C2, after the compression algorithm runs for a preset time, the storage space speed v, the current remaining storage space C, and the remaining monitoring period Tr of the current compression algorithm are obtained; Step C3, determine Is it established? If it is established, execute step C4; Step C4, raising the second threshold value Z2 by one level, and using the adjusted second threshold value Z2 to replace Z1 in step S1; Among them, the second threshold value Z2 is a preset value and has multiple gears from small to large; the storage space speed v represents the ratio of the storage space occupied by the compressed pulse wave signal during the preset time of running the compression algorithm to the preset time.

3. The restorable signal feature compression algorithm according to claim 1, characterized in that: The restorable signal feature compression algorithm also includes a parameter updating step, specifically including: Step D1, lowering the second threshold value Z2 by one level, and using the adjusted second threshold value Z2 to replace Z1 in step S1; Step D2, after the compression algorithm runs for a preset time, the current contact temperature between the wearable monitor and the patient's body surface is obtained; Step D3, determining whether the contact temperature exceeds a preset temperature threshold; if so, executing step D4; Step D4, raising the second threshold value Z2 by one level, and using the adjusted second threshold value Z2 to replace Z1 in step S1; The second threshold value Z2 is a preset value and has multiple gears from small to large.

4. The restorable signal feature compression algorithm according to any one of claims 1 to 3, characterized in that: The patient's basic information includes age and disease.

5. A wearable monitor, characterized in that: Wearable monitors do not have the function of analyzing whether the pulse wave signal is normal; It includes a collection module, a state judgment module, a compression module, a storage module and a processor; Among them, the acquisition module is configured to collect the patient's pulse wave signal and acceleration information; the state judgment module determines whether the patient is in a sleeping state, a moving state or a resting state based on the pulse wave signal and the acceleration information; the compression module compresses the pulse wave signal and stores the compressed pulse wave signal in the storage module; the processor is connected to the acquisition module, the state judgment module, the compression module, and the storage module, and the processor is configured to control the acquisition module, the state judgment module, the compression module, and the storage module to execute the restorable signal feature compression algorithm as described in any one of claims 1-4.

6. A monitoring system, characterized in that: A wearable monitor as claimed in claim 5, which is worn by a patient, and a data processing device of a hospital; Among them, the data processing device is configured to receive the compressed pulse wave signal sent by the wearable monitor, restore it into a pulse wave signal, analyze whether the restored pulse wave signal is normal, update the pulse wave history database, and send the updated pulse wave feature library and the updated total monitoring time T to the wearable monitor worn by the patient for a new monitoring cycle.

Images