Intelligent bone articular implant monitoring system based on piezoelectric-electromagnetic composite sensing

Abstract

The invention relates to the technical field of biomedical engineering, in particular to an intelligent bone articular implant monitoring system based on piezoelectric-electromagnetic composite sensing, which comprises the following contents: a signal acquisition module, a frequency spectrum analysis module, a characteristic evolution judgment module, a fusion evaluation module and an infection early warning module. According to the method, voltage and temperature signal sequences are synchronously collected and packaged, a signal set under a unified time index is constructed, a main peak frequency is extracted in a frequency domain processing link, and temperature fluctuation in multiple windows is analyzed in combination with a moving range mode, so that capturing of local abnormal jump is realized; the difference ratio analysis is combined with the second derivative to detect the frequency curvature change, the staging identification of the hopping frequency is established, and the frequency migration rate is further combined to construct the fusion health index for quantifying the continuous evolution characteristics, so that the time sequence consistency of signal processing, the fine granularity capability of fluctuation detection and the continuity of health trend expression are enhanced.

Description

Technical Field

[0001] This invention relates to the field of biomedical engineering technology, and in particular to an intelligent intra-articular plant monitoring system based on piezoelectric-electromagnetic composite sensing. Background Technology

[0002] The field of biomedical engineering technology involves the systematic application of engineering principles and design methods to medical diagnosis and physiological state monitoring, promoting the development of medical devices, implantable systems, and intelligent monitoring equipment. This includes the physical conversion and processing mechanisms of biological signals, biocompatibility research of implantable materials, energy harvesting and management technologies, miniaturized system integration design, and non-contact transmission and remote communication architecture for physiological parameters. It emphasizes the deep integration of engineering technology and life sciences, supporting accurate disease diagnosis, rehabilitation process assessment, and continuous monitoring of patient health status through multimodal sensing and system-level design.

[0003] Among them, the intelligent intra-articular implantation monitoring system based on piezoelectric-electromagnetic composite sensing refers to an integrated monitoring system that integrates piezoelectric sensing units and electromagnetic induction modules into an implantable structure. Through a multi-physics field signal acquisition and conversion mechanism, it realizes dynamic assessment of bone fusion process and infection risk. This includes using the mechanical deformation generated by piezoelectric materials during joint activity to convert external mechanical strain into electrical signals that can be used for monitoring and power supply; extracting specific frequency domain characteristic parameters through piezoelectric response to reflect the changing trend of bone tissue physiological state; integrating electromagnetic induction device to acquire temperature change signals in the implantation area; and relying on inductive coupling to realize the data transmission path from inside the body to outside the body. Through a combination of energy self-supply and signal feature extraction, it continuously senses and transmits multi-source physiological information, forming a fusion monitoring mechanism based on mechanical and thermal dimensions.

[0004] Existing technologies rely solely on physical sensing units to passively acquire electrical signals and sense temperature changes, transmitting these signals to an external terminal via coupling. This lacks a refined mechanism for structurally deconstructing the original physiological signals, making it difficult to reconstruct the physiological evolution path from the perspectives of frequency migration, temporal persistence, or multi-cycle stability. During processing, signals are often treated as raw streams, without identifying signals based on rate of change, jump density, or abnormal structures. This results in insufficient forward-looking judgment of potential pathological trends. Furthermore, it exhibits problems such as delayed response and low recognition in analyzing the nonlinear evolution of infection risks and bone tissue changes. In bone fusion monitoring, simply recording voltage peaks is insufficient to determine whether the fusion process has experienced phased stagnation or regression, and temperature changes only reflect problems in extreme anomalies, neglecting early signals in low-to-medium amplitude fluctuation sequences. This limits the system's application in continuous health management and real-time early warning. Summary of the Invention

[0005] To address the technical problems existing in the prior art, this invention provides an intelligent intra-articular vegetative monitoring system based on piezoelectric-electromagnetic composite sensing. The technical solution is as follows:

[0006] On the one hand, an intelligent intra-articular vegetative monitoring system based on piezoelectric-electromagnetic composite sensing is provided, which includes:

[0007] The signal acquisition module acquires voltage signal sequences through piezoelectric sensors, acquires temperature change sequences through electromagnetic temperature sensors, identifies and corrects abnormal data, packages the data with timestamps, obtains synchronization signal data groups, and transmits them to the spectrum analysis module.

[0008] The spectrum analysis module inputs the synchronization signal data group into a fast Fourier transform for frequency domain mapping, extracts the main peak frequency value as the main spectral feature, calculates the temperature peak-valley difference parameter using moving range analysis, obtains data feature extraction records, and transmits them to the feature evolution discrimination module and the infection early warning module.

[0009] The feature evolution discrimination module extracts records based on the data features, obtains the frequency variation rate by dividing the difference between the main peak frequency value and the previous period value, detects the curvature abnormal period by the second derivative method, evaluates the persistence, generates a fusion phased trigger signal, and transmits it to the fusion evaluation module.

[0010] The fusion assessment module calls the wavelet packet decomposition algorithm to convert the fusion phased trigger signal into a phased index, extracts the rate of change of the main peak of the spectrum by combining the low-amplitude periodic excitation, calculates the fusion health index based on the phased index and the frequency migration rate, and transmits it to the infection early warning module.

[0011] As a further embodiment of the present invention, the synchronization signal data group includes a voltage signal sequence, a temperature change sequence, and a timestamp index; the data feature extraction record includes the main peak frequency value, temperature peak-to-valley difference parameter, and frequency domain mapping processing data; the fusion phased trigger signal includes the frequency variation rate, curvature anomaly period identifier, and persistence evaluation result; and the fusion health index includes the phased index, the spectrum main peak change rate, and the frequency migration rate.

[0012] The frequency domain mapping processing data refers to a data set of frequency-amplitude correspondences obtained by performing frequency domain transformation on the synchronization signal, including a frequency-amplitude distribution matrix and corresponding timestamp indices.

[0013] As a further aspect of the present invention, the signal acquisition module includes:

[0014] The voltage extraction submodule acquires voltage signal sequences through piezoelectric sensors, records time index data according to the trigger time, divides the signal sequence and generates a frame sequence mapping between time points and voltage values, detects the difference range between each voltage value segment, extracts the voltage amplitude change between consecutive frames, and obtains the voltage fluctuation amplitude value.

[0015] The temperature monitoring submodule matches the temperature change sequence within the time index range of the electromagnetic temperature sensor based on the voltage fluctuation amplitude value, divides the temperature segments corresponding to the voltage change period, calculates the temperature change ratio within multiple segments and sets a temperature jump threshold, records the number of jump segments with a ratio greater than the threshold, and obtains the temperature jump frequency value.

[0016] The temperature mutation threshold is determined using a sliding window analysis method, which analyzes the temperature change sequence of the electromagnetic temperature sensor under voltage fluctuation amplitude, combined with historical data, engineering experience, and dynamic settings.

[0017] The data correction submodule calls the temperature jump frequency value, filters the voltage fluctuation segments that coincide with the corresponding time index, locates the jump segment in the corresponding voltage frame sequence, calculates the time average difference of voltage change within the jump segment, uses the mean to perform linear interpolation adjustment on the abnormal segment, and obtains the synchronization signal data group.

[0018] As a further aspect of the present invention, the spectrum analysis module includes:

[0019] The frequency domain conversion submodule acquires the synchronization signal data group, dynamically adjusts the fast Fourier transform sampling window based on the initially extracted main peak frequency value, performs frequency domain conversion processing on the data sequence, maps the time domain data points to the frequency domain, divides the frequency bands and allocates the amplitude, and generates frequency amplitude distribution values.

[0020] The main feature extraction submodule detects and locates the maximum amplitude point based on the amplitude sequence corresponding to multiple frequency bands in the frequency amplitude distribution value, extracts the corresponding frequency position as the main peak frequency, establishes the correspondence between frequency and amplitude, and obtains the main peak frequency value.

[0021] The peak-valley parameter calculation submodule calls the main peak frequency value, divides multiple data windows within the original data range of the synchronization signal with the main peak frequency as the period length, calculates the difference in signal amplitude within multiple windows, and obtains the temperature peak-valley difference parameter by summarizing the difference of each segment and calculating the moving range average. Combined with the main peak frequency value, the data feature extraction record is obtained.

[0022] As a further aspect of the present invention, the feature evolution discrimination module includes:

[0023] The frequency rate extraction submodule extracts records based on the data features, obtains the frequency difference between adjacent periods in the main peak frequency value sequence, and calculates the ratio with the previous period value to obtain the frequency variation rate value.

[0024] The preceding period value refers to the peak frequency value of the current period in the previous period of the time series, and is used to calculate the relative frequency change ratio between adjacent periods.

[0025] The curvature period detection submodule calculates the second derivative of each data point in the frequency variation ratio sequence based on the frequency variation rate value, and identifies the abrupt inflection point of the derivative value and marks the inflection point position in combination with the curvature anomaly threshold, detects the curvature anomaly period, and obtains the abnormal curvature period interval.

[0026] The aforementioned curvature anomaly threshold refers to comparing a preset second derivative judgment benchmark value with the actual derivative change, used to identify abrupt inflection points in frequency variation and define the curvature anomaly period.

[0027] The phased signal generation submodule calls the abnormal curvature period interval, evaluates the persistence of the abnormal curvature period, identifies the time point where the frequency change amplitude within the time period is greater than the jump judgment benchmark value, identifies the number of jump events within a unit time period and calculates the jump density, divides the time period into multiple frequency migration levels according to the jump density, records the time period distribution and jump density sequence under each level, and establishes a fused phased trigger signal.

[0028] The jump judgment benchmark value is used to clarify the definition standard of frequency change amplitude when assessing the periodicity of curvature anomaly. By setting this benchmark value, time points with large frequency change amplitudes can be distinguished from continuous changes, so as to identify and quantify jump events within a unit time period.

[0029] As a further aspect of the present invention, the fusion evaluation module includes:

[0030] The phased signal conversion submodule calls the fused phased trigger signal, and based on the frequency migration level and time period distribution, performs frequency domain reconstruction of the signal according to the wavelet packet decomposition algorithm, decomposes the time series signal of the phased state and performs normalization processing, extracts the wavelet coefficient change trend and analyzes the time series density change, and establishes the phased index value.

[0031] The frequency migration extraction submodule calls the phased index value to obtain the sequence of main peak frequency value changes under low amplitude periodic excitation in the spectrum data, calculates the variation amplitude and time span of the main peak frequency in the continuous interval, normalizes the variation amplitude according to the fixed time window length, obtains the variation rate of the main peak frequency per unit time, and generates the main peak frequency migration rate.

[0032] The health index generation submodule calls the main peak frequency migration rate, marks the frequency fluctuation level of the time period according to the correspondence between the frequency change rate and the stage index value, and calculates the fused health index based on the frequency fluctuation level value and the stage index value.

[0033] As a further aspect of the present invention, the specific formula for normalizing the variation amplitude based on a fixed time window length is as follows:

[0034] ;

[0035] Calculate the normalized characteristic value of the frequency variation amplitude;

[0036] in, Representing the Duan Di The normalized characteristic value of the frequency variation amplitude within the period. Representing the Duan Di Cycle number The main peak frequency value at each point This represents the magnitude weighting coefficient for the corresponding point. This represents the total offset value of the frequency difference within this period. This represents the time span within that period. Represents the stability constant of the denominator. The total number of sampling points in the current period is represented by , i is the segment number in the signal processing sequence, j is the index of the sampling point in the current period, and s is the data processing period number to which the current period belongs.

[0037] As a further aspect of the present invention, the system further includes:

[0038] The infection early warning module calls the data feature extraction record, extracts the temperature peak-valley difference parameter and performs continuous increase detection, identifies abnormal temperature increase events, calls the fused health index, and detects abnormal fluctuation events of the index by comparing it with the preset abnormal fluctuation range. By analyzing the continuity of abnormal fluctuations and combining the frequency of abnormal temperature increase events, the infection risk level is calculated and infection risk early warning information is output.

[0039] The infection risk warning information specifically includes the frequency of abnormal temperature increase events, abnormal fluctuation continuity indicators, and risk level quantification values.

[0040] As a further aspect of the present invention, the infection early warning module includes:

[0041] The temperature trend extraction submodule acquires the data feature extraction records, collects temperature values ​​within a continuous time period, extracts the peak and valley values ​​for each time period, calculates the extreme value difference and obtains the peak and valley difference sequence, extracts adjacent time periods with increasing peak and valley differences, identifies abnormal temperature increase events, and establishes temperature increase trend values.

[0042] The fluctuation range judgment submodule calls the fused health index and the temperature increase trend value, extracts the health index fluctuation sequence and temperature trend sequence at the corresponding time point, compares them with the preset abnormal fluctuation range, performs range matching and filtering on the health index value, identifies abnormal time periods of data fluctuation, detects abnormal fluctuation events of the index, and establishes fluctuation event records.

[0043] The abnormal fluctuation range refers to a continuous time period in which the health index value falls within a fixed upper and lower bound range, used to identify abnormal time segments with an infection trend.

[0044] The risk level generation submodule calls the fluctuation event records, analyzes the continuity of abnormal fluctuations, combines the frequency of abnormal temperature increase events, calculates the infection risk score in real time and identifies the infection risk level, and establishes infection risk early warning information.

[0045] As a further aspect of the present invention, the specific formula for calculating the infection risk score in real time by combining the frequency of abnormal temperature increase events is as follows:

[0046] ;

[0047] Calculate the infection risk score;

[0048] in, Represents the risk score of infection. This represents the total number of time steps within the time window. Representing the The frequency of abnormal temperature increase events at any given time. Representing the The temperature change at any given moment compared to the previous moment. Representing the The difference between the temperature change at any given moment and the average temperature change within the time window. This is the sequence number of the current time step.

[0049] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0050] By introducing the coordinated acquisition of piezoelectric voltage and electromagnetic temperature signals in the signal acquisition stage, and generating a synchronous signal data sequence through timestamp packaging, a data foundation for spatiotemporal synchronization is established, improving the accuracy of subsequent data alignment and analysis. This synchronous signal is input into a frequency domain mapping operation, and the main peak value of the spectrum is extracted through Fast Fourier Transform. Combined with a moving range algorithm based on frame sequence windows, multi-segment difference calculations are performed on temperature signal fluctuations, effectively forming a parameter extraction method oriented towards multi-period dynamic changes, enhancing sensitivity to physiological state anomalies. A frequency variation rate index is constructed by the difference ratio of the main peak frequency sequence changes. A curvature anomaly period identification method is constructed by introducing second-order derivative analysis, and jump density segmentation classification is used to achieve coordinated analysis of temporal continuity and jump structure, demonstrating responsiveness to nonlinear changes in physiological processes. The curvature period signal is converted into a periodic index, superimposed with the low-amplitude excitation response rate of the main spectral peak, forming a fused health index. This achieves integrated judgment of multi-source indicators in the evolutionary dimension, enabling continuous quantification of health trends and risk changes. Each execution stage, through the construction of process parameters and numerical evolution mechanisms, transforms discrete signal logic into a traceable and continuously observable evaluation framework, thereby enhancing the timeliness, sensitivity, and reliability of physiological state monitoring. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0052] Figure 1 This is a system flowchart of the present invention;

[0053] Figure 2 This is a system block diagram of the present invention. Detailed Implementation

[0054] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0055] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0056] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0057] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0058] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0059] This invention provides an intelligent intra-articular vegetative state monitoring system based on piezoelectric-electromagnetic composite sensing. Please refer to [link to relevant documentation]. Figures 1 to 2 This invention provides a technical solution for an intelligent intra-articular phytoreceptor monitoring system based on piezoelectric-electromagnetic composite sensing, comprising:

[0060] The signal acquisition module acquires voltage signal sequences through piezoelectric sensors, acquires temperature change sequences through electromagnetic temperature sensors, identifies and corrects abnormal data, packages the data with timestamps, obtains synchronization signal data groups, and transmits them to the spectrum analysis module.

[0061] The spectrum analysis module inputs the synchronization signal data group into the fast Fourier transform for frequency domain mapping, extracts the main peak frequency value as the main spectral feature, uses moving range analysis to calculate the temperature peak-valley difference parameter, obtains data feature extraction records, and transmits them to the feature evolution discrimination module and infection early warning module.

[0062] The feature evolution discrimination module extracts records based on data features, obtains the frequency variation rate by dividing the difference between the main peak frequency value and the previous period value, detects the curvature abnormal period by the second derivative method, evaluates its persistence, generates a fusion phased trigger signal, and transmits it to the fusion evaluation module.

[0063] The fusion assessment module calls the wavelet packet decomposition algorithm to convert the fusion phased trigger signal into a phased index, combines the low-amplitude periodic excitation to extract the rate of change of the main peak of the spectrum, calculates the fusion health index based on the phased index and the frequency migration rate, and transmits it to the infection early warning module.

[0064] The infection early warning module calls the data feature extraction records, extracts the temperature peak-valley difference parameter and performs continuous increase detection to identify abnormal temperature increase events, calls the fused health index, and detects abnormal fluctuation events of the index by comparing it with the preset abnormal fluctuation range. By analyzing the continuity of abnormal fluctuations and combining the frequency of abnormal temperature increase events, the infection risk level is calculated and infection risk early warning information is output.

[0065] The synchronization signal data set includes voltage signal sequences, temperature change sequences, and timestamp indexes. The data feature extraction records include main peak frequency values, temperature peak-to-valley difference parameters, and frequency domain mapping processing data. The fused phased trigger signals include frequency variation rate, curvature anomaly cycle identifier, and continuous assessment results. The fused health index includes phased index, spectrum main peak change rate, and frequency migration rate. The infection risk warning information specifically includes the frequency of abnormal temperature increase events, abnormal fluctuation continuity indicators, and risk level quantification values.

[0066] Frequency domain mapping data refers to a set of data on the correspondence between frequency and amplitude obtained by performing frequency domain transformation on the synchronization signal, including the frequency amplitude distribution matrix and the corresponding timestamp index.

[0067] Please see Figure 2 The signal acquisition module includes:

[0068] The voltage extraction submodule acquires voltage signal sequences through piezoelectric sensors, records time index data according to the trigger time, divides the signal sequence and generates a frame sequence mapping between time points and voltage values, detects the difference range between each voltage value segment, extracts the voltage amplitude change between consecutive frames, and obtains the voltage fluctuation amplitude value.

[0069] The voltage extraction submodule acquires voltage signal sequences using a piezoelectric sensor, records the time index data of the trigger moment, divides the signal sequence into multiple time periods, generates a frame sequence mapping between time points and voltage values, detects the difference interval between each voltage value segment, extracts the voltage amplitude change between consecutive frames, and obtains the voltage fluctuation amplitude value. The specific execution process is as follows: During each joint movement, the piezoelectric sensor converts mechanical strain into a voltage signal. The sampling frequency is set to 1kHz, and the acquisition time is 10 seconds, resulting in 10,000 voltage data points. The trigger moment is determined by detecting that the voltage signal exceeds a preset threshold (e.g., 0.5V), and the corresponding time index is recorded. The entire signal sequence is divided into multiple time periods according to the time index, for example, each segment contains 1,000 data points. For each data segment, a frame sequence mapping between time points and voltage values ​​is generated, forming a frame sequence. The difference interval between each voltage value segment is calculated, and the voltage amplitude change between consecutive frames is extracted. For example, in a certain segment, the voltage value changes from 0.5V to 1.2V, with an amplitude change of 0.7V. The amplitude changes of all segments are statistically analyzed to obtain the voltage fluctuation amplitude value.

[0070] The temperature monitoring submodule matches the temperature change sequence within the time index range of the electromagnetic temperature sensor based on the voltage fluctuation amplitude value, divides the temperature segments corresponding to the voltage change period, calculates the temperature change ratio within multiple segments and sets the temperature jump threshold, records the number of jump segments with a ratio greater than the threshold, and obtains the temperature jump frequency value.

[0071] The temperature monitoring submodule first identifies the fluctuation range between consecutive frames in the voltage signal curve based on the voltage fluctuation amplitude value. It then extracts continuous change segments exceeding a reference amplitude threshold. The reference amplitude threshold is set by referring to the voltage sequence fluctuation range of the same time period over the past 7 days, and is set as the historical average plus 1.5 times the standard deviation. For example, if the voltage fluctuation average in the reference sample is 0.18V and the standard deviation is 0.06V, then the voltage fluctuation amplitude threshold is set to 0.18 + 1.5 × 0.06 = 0.27V. When the voltage change value of the current frame segment is 0.31V, it is considered that this segment has [a voltage fluctuation range]. Significant voltage fluctuations occurred over a time period indexed from t3 to t6 seconds. The corresponding electromagnetic temperature sensor sampling data segment was then matched. Assuming a sampling frequency of 1Hz, temperatures of 36.2℃, 36.4℃, 36.7℃, and 36.5℃ were collected at t3, t4, t5, and t6 seconds, respectively. This temperature segment was recorded as ΔT1=[36.2,36.4,36.7,36.5]. The temperature sequence during this voltage change period was then divided into multiple segments, with a minimum granularity of a 2-second window. The first and last temperature values ​​within each segment were extracted, and the change was calculated. The ratio is calculated by dividing the difference in ratios by the previous value, i.e., (tail temperature - first temperature) ÷ first temperature. For the first segment (t3–t4 seconds), the ratio is (36.4–36.2) / 36.2 ≈ 0.0055; for the second segment (t4–t5 seconds), it is (36.7–36.4) / 36.4 ≈ 0.0082; and for the third segment (t5–t6 seconds), it is (36.5–36.7) / 36.7 ≈ -0.0054. Negative values ​​indicate a temperature drop. A temperature abrupt change threshold is then set to identify abrupt temperature changes. The threshold is set using the ratios of all segments within the past hour. The mean plus 1.75 times the standard deviation is used. If the mean ratio of segments within the sample is 0.004 and the standard deviation is 0.002, then the threshold is 0.004 + 1.75 × 0.002 = 0.0075. Segments greater than this threshold are selected as jump segments. In the results, only segment 2 has a ratio of 0.0082 > 0.0075, so one jump segment is recorded. The temperature jump frequency corresponding to this voltage fluctuation is 1 / 3, meaning that one of the three segments covered by voltage fluctuation is a temperature jump segment with a jump frequency of 0.333Hz. See the table below for complete data:

[0072] Table 1. Calculation of Temperature Segment Ratio During Voltage Period

[0073] Segment Number Start time (s) End time (s) Initial temperature (°C) Tail temperature (°C) Ratio value Jump icon 1 3 4 36.2 36.4 0.0055 no 2 4 5 36.4 36.7 0.0082 yes 3 5 6 36.7 36.5 -0.0054 no

[0074] As shown in Table 1, only segment 2 among the three segments satisfies the transition condition, with a final transition frequency of 0.333 Hz. This result indicates that there are certain temperature abrupt changes during voltage fluctuations, which is of analytical significance.

[0075] The temperature mutation threshold is determined using a sliding window analysis method. This involves analyzing the temperature change sequence of the electromagnetic temperature sensor under voltage fluctuations, combined with historical data, engineering experience, and dynamic settings.

[0076] The data correction submodule calls the temperature jump frequency value, filters the voltage fluctuation segments that coincide with the corresponding time index, locates the jump segment in the corresponding voltage frame sequence, calculates the time average difference of voltage change within the jump segment, uses the mean to perform linear interpolation adjustment on the abnormal segment, and obtains the synchronization signal data group.

[0077] Based on the temperature jump frequency, the segments whose time indices overlap with voltage fluctuation ranges are selected. Jump segments within these voltage fluctuation ranges are then located, i.e., those with abnormal voltage fluctuation amplitudes. For example, in a voltage fluctuation range, if the voltage value abruptly jumps from 0.5V to 1.5V and then falls back to 0.6V, this is identified as a jump segment. The average time difference of voltage changes within the jump segment is calculated using the following formula: ,in, The first jump segment Each voltage value This represents the number of data points in the transition segment. For example, if the voltage value in the transition segment is [0.5, 1.5, 0.6] V, then: The mean value is then used to perform linear interpolation adjustments on the abnormal segments. For example, the voltage value of the transition segment is adjusted to change linearly, smoothly transitioning from 0.5V to 0.6V. Through the above process, the synchronization signal data set is obtained.

[0078] Please see Figure 2 The spectrum analysis module includes:

[0079] The frequency domain conversion submodule acquires the synchronization signal data group, dynamically adjusts the fast Fourier transform sampling window based on the initially extracted main peak frequency value, performs frequency domain conversion processing on the data sequence, maps the time domain data points to the frequency domain, divides the frequency bands and assigns amplitudes, and generates frequency amplitude distribution values.

[0080] After acquiring the synchronization signal data group, it was defined as a single-channel voltage data sequence of length 1024, with a sampling frequency of 1000Hz and a recording interval of 1.024 seconds. First, data normalization was performed on the sequence, adjusting all voltage values ​​to a mean of 0 and an amplitude range of [-1, 1]. Then, frequency distribution transformation was applied to the sequence. By directly constructing a complex sequence and performing complex dot product expansion, the time-domain sampling points were mapped to their corresponding frequency positions. With a total of 1024 sampling points, the frequency resolution was 0.9766. The frequency range is divided into 512 frequency bands from 0 to 500 Hz. The corresponding complex modulus of each frequency band is matched as the amplitude, forming a frequency amplitude lookup table. In actual operation, the frequency index sequence is recorded as [0, 0.9766, 1.9531, ..., 499.02], and the corresponding amplitude sequence is as [0.02, 0.04, 0.08, ..., 0.01]. The amplitude sequence is labeled into the frequency band index array to build a complete mapping set. The frequency amplitude distribution value is formed by comparing the modulus of each frequency band.

[0081] The main feature extraction submodule detects and locates the maximum amplitude point based on the amplitude sequence corresponding to multiple frequency bands in the frequency amplitude distribution value, extracts the corresponding frequency position and uses it as the main peak frequency, establishes the correspondence between frequency and amplitude, and obtains the main peak frequency value.

[0082] After the frequency amplitude distribution values ​​are generated, they are imported as a linear array into the main frequency positioning process. From frequency band 1 to band 512, the amplitudes are compared pairwise. If the amplitude... If a frequency band is greater than or equal to the amplitude of the two preceding and following frequency bands, and its amplitude is greater than twice the mean of the entire amplitude sequence, then that frequency band is determined to be a candidate main peak band. The frequency position corresponding to the index of that frequency band is taken as the main peak frequency. When detecting an amplitude sequence in the amplitude array [0.02, 0.04, 0.08, 0.14, 0.10, 0.06, 0.03], the maximum amplitude is 0.14, corresponding to the 4th segment, and the frequency is... The frequency values ​​are recorded as the main peak frequencies, and a one-to-one mapping sequence table of frequency and amplitude is constructed as the main frequency reference data, as shown in Table 2:

[0083] Table 2 Main Frequency Positioning Mapping Table

[0084] Frequency band number Frequency value Amplitude 3 2.9298 0.08 4 3.9064 0.14 5 4.8828 0.10

[0085] As shown in Table 2, the frequency value of 3.9064Hz corresponds to the frequency with the largest amplitude in the current data segment, and is identified as the main peak frequency. This frequency value will be used subsequently in the feature sequence analysis and variant detection process.

[0086] The peak-valley parameter calculation submodule calls the main peak frequency value, uses the main peak frequency as the period length, divides the original data interval of the synchronization signal into multiple data windows, calculates the difference in signal amplitude within multiple windows, and obtains the temperature peak-valley difference parameter by summarizing each difference and calculating the moving range average. Combined with the main peak frequency value, it obtains the data feature extraction record.

[0087] The main peak frequency value is determined by performing spectral analysis on the raw synchronization signal data. The sampling rate is set to 1000Hz, and the sampling duration is 10 seconds, resulting in 10,000 data points. A Fast Fourier Transform (FFT) is then used to convert the time-domain signal to a frequency-domain signal with a frequency resolution of 0.1Hz. The maximum amplitude occurs at 5Hz, thus the main peak frequency is determined to be 5Hz. Data windows are divided with the main peak frequency as the period length, with a period of 0.2 seconds corresponding to 200 data points, resulting in 50 data windows. Within each data window, the maximum and minimum values ​​are extracted, and their difference is calculated. An array of amplitude differences is generated. For example, the maximum values ​​for windows 1 to 5 are 2.5, 2.7, 2.6, 2.8, and 2.9, respectively, and the minimum values ​​are 1.0, 1.2, 1.1, 1.3, and 1.4, respectively, corresponding to an amplitude difference of 1.5 units. Then, the moving window length is set to 5 windows. The range of amplitude differences for each group of 5 consecutive windows is calculated. The range for the first 5 windows is 1.5 - 1.5 = 0. This process is repeated by moving one window forward, resulting in 46 range values. These 46 range values ​​are then summed and their arithmetic mean is taken to obtain the moving range average, denoted as _____. ,in To move the window width, The total length of the amplitude difference sequence is given by the formula. Perform calculations, parameters Indicates the first The amplitude difference between the windows, The starting number for sliding. The local number within the current sliding window, symbol This indicates taking the maximum value. This indicates taking the minimum value. Represents absolute value. To express summation, This formula, representing division, is advantageous because it extracts the maximum and minimum amplitude differences within multiple sliding windows, performs difference operations, and then performs a global average, thus reflecting the central tendency of local fluctuations in the data. If the range is 0.3 in all sliding windows from the 1st to the 46th, then... The results show that the fluctuation level of the signal is relatively stable within the short-period sliding range. Further, the corresponding data feature extraction record was constructed by combining the main peak frequency value of 5Hz.

[0088] Please see Figure 2The feature evolution discrimination module includes:

[0089] The frequency rate extraction submodule extracts records based on data features, obtains the frequency difference between adjacent periods in the main peak frequency value sequence, and calculates the ratio with the previous period value to obtain the frequency variation rate value.

[0090] Frequency rate extraction is based on the time series of the main peak frequency value. First, the frequency array within the recording period is sorted chronologically. For example, if the main peak frequency sequence within the period is set as... When calculating the frequency difference, the differences between adjacent terms are taken sequentially: , , For each frequency difference, the variation ratio is calculated with respect to the corresponding previous period's main peak frequency value. , , The ratio values ​​under each period are combined as a frequency variation rate sequence, which serves as the basic input for the dynamic trend of frequency change.

[0091] The previous period value refers to the peak frequency value of the current period in the previous period of the time series, and is used to calculate the relative frequency change ratio between adjacent periods.

[0092] The curvature period detection submodule calculates the second derivative of each data point in the frequency variation ratio sequence based on the frequency variation rate value, and identifies the abrupt inflection point of the derivative value and marks the inflection point position in combination with the curvature anomaly threshold, thereby detecting the curvature anomaly period and obtaining the abnormal curvature period interval.

[0093] After the frequency variation rate sequence is formed, a three-point difference set is constructed by taking each data point in the sequence as the center and its two adjacent points before and after it to estimate the second derivative. The expression is as follows: If the three middle items above are Then there is The second derivative results of all data points are used to form a derivative sequence. A curvature anomaly threshold of 0.1 is set, and outliers are determined based on whether the absolute value of the derivative is greater than the threshold. These are marked as curvature jump points, and are grouped using a continuous indexing method to form curvature anomalous periodic intervals. For example, if the derivatives of the 3rd to 5th periods are all greater than 0.1, then this interval is an anomalous curvature segment, denoted as interval . ;

[0094] Combining curvature anomaly thresholds refers to comparing a preset second derivative judgment benchmark value with the actual derivative change, which is used to identify abrupt inflection points in frequency variation and define the curvature anomaly period.

[0095] The phased signal generation submodule calls the abnormal curvature cycle interval, evaluates the persistence of the abnormal curvature cycle, identifies the time point where the frequency change amplitude within the time period is greater than the jump judgment benchmark value, identifies the number of jump events within a unit time period and calculates the jump density, divides the time period into multiple frequency migration levels according to the jump density, records the time period distribution and jump density sequence under each level, and establishes a fused phased trigger signal.

[0096] After the abnormal curvature period interval is formed, the ratio of the amplitude of the main peak frequency change to the time interval is statistically analyzed to determine whether it exceeds the jump reference value. The reference value is set based on the 95% confidence interval of the frequency change amplitude in historical health data. Let's say the jump reference value is set to 0.8 Hz / s. If the main peak frequency rises by 2.6 Hz within 3 seconds, then the jump rate is... Hz / s, if higher than the reference value, is recorded as a jump event. The number of jump events within a unit time period is counted according to the sliding time window. The number of events per second is defined as the jump density. For example, if 4 jumps are observed within 5 seconds, the jump density is 0.8 times / s. The jump density is classified according to the segment rules. The density is below 0.4 as level 1, 0.4 to 0.7 as level 2, and above 0.7 as level 3, forming a frequency migration level label. The jump density and corresponding level of each segment are output and recorded as a frequency migration level sequence.

[0097] Table 3 Jump Density and Rank Distribution

[0098] Time period number Jump density (times / s) Frequency Shift Level 1 0.25 1 2 0.65 2 3 0.87 3

[0099] As shown in Table 3, the frequency migration level in the third time period is 3, indicating drastic changes and the formation of a high-intensity frequency migration trigger signal. This sequence was subsequently used in the fusion phased index generation process.

[0100] The jump judgment benchmark is used to clarify the definition standard of frequency change amplitude when assessing the periodicity of curvature anomalies. By setting this benchmark, time points with large frequency change amplitudes can be distinguished from continuous changes, so as to identify and quantify jump events within a unit time period.

[0101] Please see Figure 2 The integrated evaluation module includes:

[0102] The phased signal conversion submodule calls the fused phased trigger signal, reconstructs the signal in the frequency domain based on the frequency migration level and time period distribution, and performs normalization processing on the time-series signal of the phased state according to the wavelet packet decomposition algorithm, extracts the wavelet coefficient change trend and analyzes the time-series density change, and establishes the phased index value.

[0103] The fusion phased trigger signal records the frequency shift level sequence and jump density distribution within each time period. During the conversion process, the signal is first constructed into a discrete sequence according to each level segment, with the frequency shift level values ​​arranged sequentially as follows: The time intervals are all 1 second long. This sequence is input into a time-series signal sequence constructed based on level labels, with a sequence length of 8. Signal decomposition processing is used to expand the original sequence into a binary multi-level structure. Each subsequence constructed in each level is then folded and extracted in units of 4 lengths to construct a sub-band coefficient sequence. For example, the frequency shift level from second 1 to second 4 is... Summing the subbands yields a sum of 8, which is then normalized to... A wavelet coefficient density sequence is constructed using this sub-band sequence. Perform density slope change statistics on the sequence, and successively calculate the differences to obtain the change trend sequence. In each segment, determine the cumulative density rate of the upward trend, and calculate the average of the trend growth rates for the entire segment to obtain the average density of the frequency migration trend. This value serves as a representation of the temporal variation of wavelet density. After performing the same operation on all sub-segments, the trend value of each segment is obtained. Then, the staging index is calculated based on the average change level of the trend. The weighting coefficient is set to 1.5 for the upward trend segment, 0.7 for the downward trend segment, and 1.0 for the stable trend segment. The product of the trend within the segment and the coefficient is multiplied and summed. The normalized value is the final staging index.

[0104] The frequency migration extraction submodule calls the phased index value to obtain the sequence of changes in the main peak frequency value under low amplitude periodic excitation in the spectrum data, calculates the change amplitude and time span of the main peak frequency in the continuous interval, normalizes the change amplitude according to the fixed time window length, obtains the change rate of the main peak frequency per unit time, and generates the main peak frequency migration rate.

[0105] The specific formula for normalizing the variation amplitude based on a fixed time window length is as follows:

[0106] ;

[0107] Calculate the normalized characteristic value of the frequency variation amplitude;

[0108] in, Representing the Duan Di The normalized characteristic value of the frequency variation amplitude within the period. Representing the Duan Di Cycle number The main peak frequency value at each point This represents the magnitude weighting coefficient for the corresponding point. This represents the total offset value of the frequency difference within this period. This represents the time span within that period. Represents the stability constant of the denominator. The total number of sampling points in the current period is represented by , i is the segment number in the signal processing sequence, j is the sampling point index in the current period, and s is the data processing period number to which the current period belongs.

[0109] formula:

[0110] ;

[0111] Detailed explanation of the formula and its calculation derivation:

[0112] The formula is used to calculate the normalized characteristic value of the frequency variation amplitude within the i-th segment and s-th period. The result is used to evaluate the rate of change of the main peak frequency per unit time and generate the main peak frequency migration rate.

[0113] Parameter meanings and settings:

[0114] The peak frequency value of the j-th point in the i-th segment and s-th period is set to 10.

[0115] The amplitude weighting coefficient corresponding to the j-th point in the s-th period of the i-th segment is set to 1.0.

[0116] The total frequency difference offset formed by the cumulative differences between all main peak frequency values ​​and previous reference frequencies within the i-th segment and s-th period is set to 2.

[0117] : The time span between the start and end timestamps of sampling within the i-th segment and s-th period, set to 0.5;

[0118] The denominator stability constant is set to 0.01.

[0119] Total number of sampling points in the current period, set to 5;

[0120] Substitute the parameters into the formula to calculate:

[0121] ;

[0122] ;

[0123] ;

[0124] ;

[0125] ;

[0126] ;

[0127] The result of 39.92 indicates that the rate of change of the main peak frequency in the i-th segment and the s-th period is 39.92 Hz / s, reflecting the changing trend of the main peak frequency in this period, which helps to assess the health status of bones and joints.

[0128] The health index generation submodule calls the main peak frequency migration rate, marks the frequency fluctuation level of a time period according to the correspondence between the frequency change rate and the stage index value, and calculates the integrated health index based on the frequency fluctuation level value and the stage index value.

[0129] The main peak frequency shift rate, calculated in the previous stage by the ratio of amplitude change to time interval, is directly called in this module according to the period. The frequency change rate sequence is as follows: Hz / s. Levels are categorized by threshold: 0–0.4 is Level 1, 0.4–0.8 is Level 2, and above 0.8 is Level 3. The corresponding fluctuation level sequence is as follows: Calling the periodic index sequence value in each time period The health index is constructed by multiplying the volatility level value by the corresponding period index. The calculation method for the corresponding period value is: level weight multiplied by the period index. The weight is set to 0.9 for level 1, 1.0 for level 2, and 1.1 for level 3. For example, in the 4th segment, the volatility level is 3 and the period index is 1.36, then the integration index is... The calculation for the entire sequence is as follows:

[0130] Table 4. Calculation Table of Integrated Health Index

[0131] Time period Frequency rate level Phase Index Weighting coefficient Health Index 1 1 0.85 0.9 0.765 2 2 1.02 1.0 1.020 3 1 0.92 0.9 0.828 4 3 1.36 1.1 1.496 5 2 0.88 1.0 0.880 6 1 0.81 0.9 0.729 7 2 1.18 1.0 1.180 8 3 1.42 1.1 1.562

[0132] As shown in Table 4, the fusion health index in segment 8 reached 1.562, which is the maximum value in this cycle, indicating that the state fluctuation was the most drastic during this period. This sequence was subsequently used in the process of infection risk assessment and continuous fluctuation analysis.

[0133] Please see Figure 2 The infection early warning module includes:

[0134] The temperature trend extraction submodule acquires data feature extraction records, collects temperature values ​​within a continuous time period, extracts peak and valley values ​​for each time period, calculates extreme value differences and obtains peak and valley difference sequences, extracts adjacent time periods with increasing peak and valley differences, identifies abnormal temperature increase events, and establishes temperature increase trend values.

[0135] The temperature trend extraction process obtains the temperature value sequence from continuous sampling records of temperature sensors within the structure. The sampling period is 1 second, and the total duration is 10 seconds. The temperature data is recorded as follows: The data is measured in degrees Celsius and divided into 10 time periods per second. The maximum and minimum values ​​within each time period are extracted to form the peak-valley difference. For example, the data range in the first segment is 36.8 to 37.2, with an extreme value difference of 0.4; the difference in the second segment is 0.6; and the difference in the third segment is 0.9. The extreme value differences for all time periods are calculated sequentially to form the peak-valley difference sequence. In this sequence, combinations of increasing differences between adjacent time periods are identified. For example, if the difference between segments 1 and 3 increases, and the difference between segments 5 and 7 also increases, a jump judgment condition is set: if the difference between two consecutive time periods increases positively and the increase is greater than 0.2, then this segment is determined to be an abnormal temperature increase event, and the occurrence frequency is recorded as 2. The temperature increase trend value is obtained by dividing the occurrence frequency by the total number of time periods. The current value is... .

[0136] The fluctuation range judgment submodule calls the fusion health index and temperature increase trend value to extract the health index fluctuation sequence and temperature trend sequence at the corresponding time point. By comparing with the preset abnormal fluctuation range, the health index value is range matched and filtered to identify abnormal time periods of data fluctuation, detect abnormal fluctuation events of the index, and establish fluctuation event records.

[0137] The fluctuation range judgment submodule first calls the fusion health index and temperature increase trend value to extract the health index fluctuation sequence and temperature trend sequence at the corresponding time points, and then processes the health index sequence... and temperature trend series Based on their respective upper and lower limits, the data are subjected to minimum-maximum normalization and uniformly converted into dimensionless intervals. Using the normalization formula , The lower and upper limits of the normalization are set by selecting extreme values ​​from historical monitoring data. The normalized values ​​facilitate subsequent comparison and judgment. If the normalized health index value at a certain time point is... If the value exceeds the upper threshold of 0.85, or if any value in three consecutive sampling points exceeds this upper threshold, then that moment and the next two time points are classified as the preliminary abnormal fluctuation detection interval, and then further analyzed based on the temperature trend sequence. To determine whether there is a significant increasing temperature trend within the same time range, the difference between any two adjacent points in three consecutive time points is used. When all values ​​exceed the set jump amplitude value of 0.05, a temperature increase is considered to have occurred during that period. This jump amplitude value is determined by the 95th percentile variation in the sample statistical analysis. The temperature values ​​before normalization are collected from an electromagnetic temperature sensor in the joint area, with a sampling frequency set to 1Hz. The reference range for the actual data in the recent period is 36.1℃ to 36.9℃, therefore the setting is... , Therefore, when the temperatures from t1 to t3 are 36.3, 36.5, and 36.7℃ respectively, the normalized sequences are 0.25, 0.5, and 0.75, which meet the jump requirement and are identified as temperature increasing segments. Subsequently, a matching function is called to integrate the normalized health index with the temperature trend value, and the intersection time points are filtered to form an abnormal period index. A fluctuation event record is established, with the timestamp as the primary key, storing four fields: health index value, temperature value, normalized value, and status label. After normalization, the two sequences have the same numerical range, facilitating simultaneous comparison and trend analysis. If the original value needs to be restored after normalization, it is done using the inverse transformation formula. and The original value is obtained, such as t2 normalized temperature is 0.5, corresponding to an actual temperature of 36.5℃. The consistency between normalization and inverse normalization can be verified. Finally, each identified event point and its matching information are synchronously saved to the system database as a basis for subsequent infection warning judgment.

[0138] Abnormal fluctuation range refers to a continuous time period in which the health index value falls within a fixed upper and lower limit range, used to identify abnormal time segments with an infection trend.

[0139] The risk level generation submodule calls the fluctuation event record, analyzes the continuity of abnormal fluctuations, combines the frequency of abnormal temperature increase events, calculates the infection risk score in real time and identifies the infection risk level, and establishes infection risk early warning information.

[0140] The specific formula for calculating the infection risk score in real time, based on the frequency of abnormal temperature increase events, is as follows:

[0141] ;

[0142] Calculate the infection risk score;

[0143] in, Represents the risk score of infection. This represents the total number of time steps within the time window. Representing the The frequency of abnormal temperature increase events at any given time. Representing the The temperature change at any given moment compared to the previous moment. Representing the The difference between the temperature change at any given moment and the average temperature change within the time window. The sequence number of the current time step;

[0144] formula:

[0145] ;

[0146] Detailed explanation of the formula and its calculation derivation:

[0147] The formula is used to calculate the infection risk score. It is used to assess the level of infection risk that an individual may face due to abnormal fluctuations in body temperature within a specific time window;

[0148] Parameter meanings and settings:

[0149] The total number of time steps within the time window is set to 6;

[0150] For the first The frequency of abnormal temperature increase events at time t, based on continuous body temperature monitoring data acquired by a temperature sensor, is defined as the frequency at the time t. The number of times body temperature rises by more than 0.5°C in the hour prior to the event, with the following settings:

[0151] , ;

[0152] The temperature difference between adjacent time points collected by the temperature sensor, i.e., the first... The body temperature change value compared to the previous moment is set as follows:

[0153] , ;

[0154] For the first The difference between the change in body temperature at any given time and the average change in body temperature within the time window, calculated as follows: ,in This represents the average value of body temperature changes within the time window.

[0155] calculate :

[0156] ;

[0157] calculate :

[0158] , , , , , ;

[0159] Substitute the parameters into the formula to calculate:

[0160] Calculate the fractional term for each time step:

[0161] Step 1:

[0162] ;

[0163] Step 2:

[0164] ;

[0165] Step 3:

[0166] ;

[0167] Step 4:

[0168] ;

[0169] Step 5:

[0170] ;

[0171] Step 6:

[0172] ;

[0173] Summing and calculating the average:

[0174] ;

[0175] A result of 0.8922 indicates that within this time window, the overall infection risk score reflected by an individual's body temperature changes is 0.8922. This value can be used to further determine the infection risk level, and, in conjunction with preset risk level classification criteria, to decide whether appropriate medical intervention measures are needed.

[0176] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0177] It should be understood that the term "and / or" in this article merely describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0178] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0179] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process 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 invention.

[0180] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0181] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0182] In the embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0183] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0184] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0185] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0186] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An intelligent bone joint implant monitoring system based on piezoelectric- electromagnetic composite sensing, characterized in that, The system comprises: The signal acquisition module acquires a voltage signal sequence through a piezoelectric sensor, acquires a temperature change sequence according to an electromagnetic temperature sensor, identifies abnormal data and corrects, packs in combination with a time stamp, obtains a synchronous signal data group, and transmits to a frequency spectrum analysis module; The frequency spectrum analysis module inputs the synchronous signal data group into a fast Fourier transform for frequency domain mapping, extracts a main peak frequency value as a frequency spectrum main feature, calculates a temperature peak-valley difference parameter by using a moving range analysis, obtains a data feature extraction record, and transmits to a feature evolution discrimination module and an infection early warning module; The feature evolution discrimination module, based on the data feature extraction record, calculates a frequency variation rate according to the main peak frequency value by dividing the previous period value, detects a curvature abnormal period by a second derivative method, evaluates persistence, generates a fusion staging trigger signal, and transmits to a fusion evaluation module; The fusion evaluation module converts the fusion staging trigger signal into a staging index by calling a wavelet packet decomposition algorithm, extracts a frequency spectrum main peak change rate in combination with a low amplitude period excitation, calculates a fusion health index according to the staging index and the frequency migration rate, and transmits to the infection early warning module.

2. The piezoelectric-electromagnetic composite sensor based intelligent bone joint implant monitoring system as claimed in claim 1, wherein, The synchronous signal data group comprises a voltage signal sequence, a temperature change sequence, and a time stamp index, the data feature extraction record comprises a main peak frequency value, a temperature peak-valley difference parameter, and frequency domain mapping processing data, the fusion staging trigger signal comprises a frequency variation rate, a curvature abnormal period identifier, and a persistence evaluation result, and the fusion health index comprises a staging index, a frequency spectrum main peak change rate, and a frequency migration rate; The frequency domain mapping processing data refers to a data set of the corresponding relationship between frequency and amplitude obtained by frequency domain transformation of the synchronous signal, comprising a frequency amplitude distribution matrix and a corresponding time stamp index.

3. The piezoelectric-electromagnetic composite sensor based intelligent bone joint implant monitoring system as claimed in claim 1, wherein, The signal acquisition module comprises: The voltage extraction submodule acquires a voltage signal sequence through a piezoelectric sensor, divides the signal sequence according to the time index data recorded at the trigger time, generates a frame sequence mapping of time points and voltage values, detects the difference interval between each voltage value, extracts the voltage amplitude change between consecutive frames, and obtains the voltage fluctuation amplitude value; The temperature monitoring submodule matches the temperature change sequence in the time index range collected by the electromagnetic temperature sensor according to the voltage fluctuation amplitude value, divides the temperature segments corresponding to the voltage change period, calculates the temperature change rate in multiple segments and sets a temperature mutation threshold, records the number of jump segments with a rate greater than the threshold, and obtains a temperature jump frequency value; The temperature mutation threshold is analyzed by using a sliding window analysis method, by analyzing the temperature change sequence of the electromagnetic temperature sensor under the voltage fluctuation amplitude, in combination with historical data, engineering experience and dynamic setting; The data correction submodule calls the temperature jump frequency value, filters the voltage fluctuation segments with the same time index, locates the jump segments in the corresponding voltage frame sequence, calculates the time average difference of voltage change in the jump segment, uses the mean value to adjust the abnormal segment by linear interpolation, and obtains the synchronous signal data group.

4. The piezoelectric-electromagnetic composite sensor-based intelligent bone joint implant monitoring system as claimed in claim 3, wherein, The frequency spectrum analysis module comprises: The frequency domain conversion submodule obtains the synchronization signal data set, dynamically adjusts a fast Fourier transform sampling window based on a preliminarily extracted main peak frequency value, performs frequency domain conversion processing on a data sequence, maps time domain data points to a frequency domain, divides a frequency band and allocates an amplitude value, and generates a frequency amplitude distribution value; The main feature extraction submodule detects and locuses a maximum amplitude point according to an amplitude sequence corresponding to multiple frequency bands in the frequency amplitude distribution value, extracts a corresponding frequency position and takes the corresponding frequency position as a main peak frequency, establishes a corresponding relationship between the frequency and the amplitude, and obtains a main peak frequency value; The peak-valley parameter calculation submodule calls the main peak frequency value, divides multiple data windows in a synchronization signal original data interval with the main peak frequency as a period length, calculates a difference value of signal amplitudes in the multiple windows, obtains a temperature peak-valley difference parameter by summarizing each difference value and calculating a moving range average, and obtains a data feature extraction record in combination with the main peak frequency value.

5. The piezoelectric-electromagnetic composite sensor based intelligent bone joint implant monitoring system as claimed in claim 4, wherein, The feature evolution discrimination module comprises: The frequency rate extraction submodule obtains a frequency difference value of adjacent periods in a main peak frequency value sequence based on the data feature extraction record, and calculates a ratio with a previous period value to obtain a frequency variation rate value; The previous period value refers to a main peak frequency value of a previous period in a time sequence corresponding to a current period main peak frequency, and is used to calculate a frequency relative change ratio between adjacent periods; The curvature period detection submodule calculates a second derivative of each data point in a frequency variation ratio sequence according to the frequency variation rate value, identifies a mutation inflection point of the derivative value in combination with a curvature anomaly threshold, labels an inflection point position, detects a curvature anomaly period, and obtains an abnormal curvature period interval; The curvature anomaly threshold refers to comparing a preset second derivative judgment reference value with an actual derivative change, and is used to identify a mutation inflection point in frequency variation and define a curvature anomaly period; The staging signal generation submodule calls the abnormal curvature period interval, evaluates the persistence of the curvature anomaly period, identifies a time point at which a frequency change amplitude is greater than a jump judgment reference value in a time period, identifies a number of jump events in a unit time period and calculates a jump density, divides the time period into multiple frequency migration levels according to the jump density, records a time period distribution and a jump density sequence in each level, and establishes a fusion staging trigger signal; The jump judgment reference value is used to clearly define a standard for the frequency change amplitude when evaluating the persistence of the curvature anomaly period, and by setting the reference value, the time point at which the frequency change amplitude is large is distinguished from continuous change, so as to identify and quantify the jump events in a unit time period.

6. The piezoelectric-electromagnetic composite sensor based intelligent bone joint implant monitoring system as claimed in claim 5, wherein, The fusion evaluation module comprises: The staging signal conversion submodule calls the fusion staging trigger signal, performs frequency domain reconstruction on a signal based on a frequency migration level and a time period distribution according to a wavelet packet decomposition algorithm, decomposes a time sequence signal of a decomposition level state and performs normalization processing, extracts a wavelet coefficient change trend and analyzes a time sequence density change, and establishes a staging index value. The frequency migration extraction submodule calls the staging index value, obtains a main peak frequency value change sequence under a low-amplitude periodic excitation in the spectrum data, calculates a variation amplitude and a time span of the main peak frequency in a continuous interval, performs normalization processing on the variation amplitude according to a fixed time window length, obtains a variation rate of the main peak frequency per unit time, and generates a main peak frequency migration rate; The health index generation submodule calls the main peak frequency migration rate, labels a frequency fluctuation level of a time period according to a corresponding relationship between the frequency variation rate and the staging index value, and calculates a fusion health index according to the frequency fluctuation level value and the staging index value.

7. The smart bone joint implant monitoring system based on piezoelectric- electromagnetic composite sensing of claim 6, wherein, The specific formula for performing normalization processing on the variation amplitude according to the fixed time window length is: ； Calculate a frequency variation amplitude normalization characteristic value; in, Representing the Duan Di The normalized characteristic value of the frequency variation amplitude within the period. Representing the Duan Di Cycle number The main peak frequency value at each point This represents the magnitude weighting coefficient for the corresponding point. This represents the total offset value of the frequency difference within this period. This represents the time span within that period. Represents the stability constant of the denominator. The total number of sampling points in the current period is represented by , i is the segment number in the signal processing sequence, j is the index of the sampling point in the current period, and s is the data processing period number to which the current period belongs.

8. The piezoelectric-electromagnetic composite sensor based intelligent bone joint implant monitoring system as claimed in claim 1, wherein, The system further includes: The infection early warning module calls the data feature extraction record, extracts a temperature peak-valley difference parameter and performs continuous incremental frequency detection, identifies an abnormal temperature incremental event, calls the fusion health index, detects an abnormal fluctuation event of the index by comparison with a preset abnormal fluctuation interval, calculates an infection risk level by analyzing the continuity of the abnormal fluctuation and combining the abnormal temperature incremental event frequency, and outputs infection risk early warning information; The infection risk early warning information specifically includes an abnormal temperature incremental event frequency, an abnormal fluctuation continuity index, and a risk level quantization value.

9. The piezoelectric-electromagnetic composite sensor-based intelligent bone joint implant monitoring system as claimed in claim 8, wherein, The infection early warning module includes: The temperature trend extraction submodule obtains the data feature extraction record, collects temperature values in a continuous time period, extracts peak-valley values of each time period, calculates an extreme value difference and obtains a peak-valley difference sequence, extracts adjacent time periods of the peak-valley difference increment, identifies an abnormal temperature incremental event, and establishes a temperature incremental trend value; The fluctuation interval judgment submodule calls the fusion health index and the temperature incremental trend value, extracts a health index fluctuation sequence and a temperature trend sequence at a corresponding time point, performs interval matching screening on the health index value by comparison with a preset abnormal fluctuation interval, identifies a data fluctuation abnormal time period, detects an abnormal fluctuation event of the index, and establishes a fluctuation event record; The abnormal fluctuation interval refers to a continuous time period in which the health index value falls within a fixed upper and lower limit range, and is used to identify an abnormal time sequence section with an infection trend; The risk level generation submodule calls the fluctuation event record, analyzes the continuity of the abnormal fluctuation, combines the abnormal temperature incremental event frequency, calculates an infection risk score in real time and identifies an infection risk level, and establishes infection risk early warning information.

10. The piezoelectric-electromagnetic composite sensor-based intelligent bone joint implant monitoring system as claimed in claim 9, wherein, The specific formula for calculating the infection risk score in real time is: ； Calculate an infection risk score; wherein, represents an infection risk score, represents the total number of time steps within the time window, represents the abnormal temperature increasing event frequency at the time point, represents the temperature change value at the time point compared to the previous time point, represents the difference between the temperature change value at the time point and the average temperature change value within the time window, is the serial number of the current time step.

Images