Multi-parameter human body fat real-time data monitoring method, equipment and medium

Abstract

The invention discloses a multi-parameter human body fat real-time data monitoring method and device and a medium, and relates to the technical field of bioelectrical impedance analysis, and the method comprises the steps: obtaining a multi-frequency impedance amplitude and an impedance phase according to a fixed beat, continuously recording the change of impedance along with time at a reference frequency point, and forming a data fragment sequence sorted according to time; performing fragment shaping on data fragments in the data fragment sequence in sequence to obtain an effective multi-frequency amplitude-phase sequence and an effective continuous time sequence; extracting multi-frequency spectrum form representation and time form representation on the effective multi-frequency amplitude-phase sequence and the effective continuous time sequence, and performing fractional suppression coupling and logic bounded mapping to obtain a body fat percentage; and performing baseline deduction and triangular adjacent smoothing on the body fat percentage to obtain a continuous body fat percentage curve, freezing the body fat percentage when abnormal segments continuously appear, and updating the continuous body fat percentage curve after recovery. And the smoothness and the stability of a body fat percentage curve are ensured through triangular adjacent smoothness.

Description

Technical Field

[0001] This invention relates to the field of bioelectrical impedance analysis technology, and in particular to a method, device and medium for real-time monitoring of multi-parameter human body fat data. Background Technology

[0002] Human body fat monitoring technology is widely used in health management, sports science, and clinical medicine. Traditional body fat measurement methods include bioelectrical impedance analysis (BIA), subcutaneous fat interlayer method, and DXA (dual-energy X-ray absorptiometry). Bioelectrical impedance analysis is one of the most common body fat monitoring methods. Its principle is based on applying a small current to the human body and estimating the body fat percentage by measuring the impedance of the current through the body. The content of body fat, muscle, and water can be inferred by the conduction characteristics of the current in different tissues. In recent years, with the advancement of technology, bioelectrical impedance analysis equipment has been widely used, making body fat measurement more convenient and allowing for real-time reflection of the body's physiological state.

[0003] Existing technologies typically rely on measurements at a single frequency or a limited number of frequency points, leading to significant differences in accuracy among individuals. Furthermore, when processing multi-frequency signals, existing technologies lack systematic integration and optimization of data from different frequencies, making it difficult to reflect complex physiological changes in real time and efficiently, especially changes under dynamic conditions. Moreover, they generally rely on simplified models, which may contain certain biases, affecting their accuracy and real-time performance under complex conditions. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a multi-parameter real-time human body fat data monitoring method to solve the problems of accuracy and real-time performance in multi-frequency signal integration and dynamic change monitoring.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a method for real-time monitoring of multi-parameter human body fat data, comprising, The multi-frequency impedance amplitude and impedance phase are acquired at a fixed beat, and the impedance changes over time are continuously recorded at a reference frequency point to form a data segment sequence sorted by time. The data segments in the data segment sequence are sequentially segmented to obtain an effective multi-frequency amplitude-phase sequence and an effective continuous time sequence; Multispectral and temporal morphological representations are extracted from effective multi-frequency amplitude-phase sequences and effective continuous time sequences, and fractional suppression coupling and logical bounded mapping are performed to obtain body fat percentage. Baseline subtraction and triangular proximity smoothing are applied to the body fat percentage to obtain a continuous body fat percentage curve. When abnormal segments appear continuously, the body fat percentage is frozen, and the continuous body fat percentage curve is updated after recovery.

[0007] As a preferred embodiment of the multi-parameter real-time human body fat data monitoring method of the present invention, the steps of acquiring multi-frequency impedance amplitude and impedance phase at a fixed beat, and continuously recording impedance changes over time at a reference frequency point to form a time-ordered data segment sequence are as follows: Set a fixed beat and the frequency measurement sequence within each beat. When the beat begins, measure all frequency points within the beat in sequence according to the frequency measurement sequence, and obtain the impedance amplitude and impedance phase of each frequency point respectively. Select a reference frequency point and record the real part time series of the reference frequency point. Using a fixed-length time window on the time axis, the multi-frequency impedance amplitude and phase of non-reference frequency points within each time window are collected and written into data segments according to the beat number and position. At the same time, the real part time series of reference frequency points is collected synchronously and bound to the data segments. The bound data segments are sorted by time to form a time-sorted data segment sequence, and a spatial global multi-frequency amplitude and phase sequence and a spatial global continuous time series are established.

[0008] As a preferred embodiment of the multi-parameter real-time human body fat data monitoring method of the present invention, the specific steps of performing segment shaping on the data segments in the data segment sequence are as follows: Within a data segment in the data segment sequence, select any non-reference frequency point, and take the first impedance phase of the current frequency point in the data segment as the starting value of the expanded phase of the current frequency point in the data segment; In chronological order, the impedance phase of the current frequency point in the data segment is processed sequentially. For each new moment, the difference between the impedance phase at the new moment and the impedance phase at the previous moment is calculated. When the difference is between the negative half-phase period and the positive half-phase period, the difference is considered as the actual change. When the difference is greater than the positive half-phase period, a complete phase period is deleted from the difference. When the difference is less than the negative half-phase period, a complete phase period is superimposed on the difference. The sum of the expanded phase and the corrected difference from the previous moment is taken as the expanded phase at the current moment. This process is repeated frequency-by-frequency, and the expanded phase and the corrected difference are continuously accumulated until the last impedance phase of the current frequency in the data segment is processed.

[0009] As a preferred embodiment of the multi-parameter real-time human body fat data monitoring method of the present invention, the specific steps for obtaining the effective multi-frequency amplitude-phase sequence and the effective continuous time series are as follows: Within a data segment, taking each beat as a boundary, check whether the overall change direction of the real part time series of the reference frequency point within the beat is consistent with the overall change direction of the unfolded phase of the lowest frequency point within the beat. Data segments that pass the check are marked as consistent segments, and data segments that fail the check are marked as inconsistent segments and splicing is stopped. For a consistent segment, the impedance amplitude and expanded phase of each non-reference frequency point are appended to the end of the global effective multi-frequency amplitude and phase sequence in the order of time stamps, while the real part time sequence of the reference frequency point is appended to the end of the global effective continuous time sequence.

[0010] As a preferred embodiment of the multi-parameter real-time human body fat data monitoring method of the present invention, the specific steps for extracting multi-spectral shape characteristics and temporal shape characteristics from effective multi-frequency amplitude and phase sequences and effective continuous time sequences are as follows: Within a fixed-length time window, low-frequency points, high-frequency points, and characteristic frequency points are selected from the frequency point list. The impedance amplitude and expanded phase of the low-frequency points, high-frequency points, and characteristic frequency points are read from the effective multi-frequency amplitude-phase sequence. The real parts of the impedance at low and high frequencies are calculated. The ratio of the difference between the real parts of the low-frequency impedance and the real parts of the high-frequency impedance to the real part of the low-frequency impedance is used as the normalization ratio of the low-frequency impedance difference. The phase shape quantity is obtained by tangent mapping of the expanded phase of the characteristic frequency point in the effective multi-frequency amplitude-phase sequence. The normalization ratio of the low-frequency impedance difference and the phase shape quantity are the multi-spectral shape characterization. Within the same time window, the real part time series of the reference frequency is taken from the effective continuous time series, and mean-reduction is performed within the time window. The amplitude envelope of the real part time series of the reference frequency after mean-reduction is calculated, and normalized by the average absolute amplitude of the real part time series of the reference frequency after mean-reduction within the time window to obtain the modulation intensity. Within the same time window, the local average value of the real part of the reference frequency impedance is calculated, and the natural logarithm of the ratio of the local average value of the real part of the reference frequency impedance to the real part of the reference frequency impedance is taken to obtain the baseline normalization. The modulation intensity and the baseline normalization are the time shape characteristics.

[0011] As a preferred embodiment of the multi-parameter real-time human body fat data monitoring method of the present invention, the specific steps for obtaining the body fat percentage by performing fractional suppression coupling and logically bounded mapping are as follows: Under the same time window and unified time mark, the normalized ratio of the low-frequency impedance difference is summed with the phase morphology quantity, and the coupling is suppressed by the modulation intensity. The intermediate quantity is obtained by summing the fractional suppression coupled quantity with the baseline normalized quantity. The body fat mapping quantity is obtained through logically bounded mapping and a personal baseline is established. The body fat percentage is obtained by multiplying the ratio of the current body fat mapping quantity and the personal baseline body fat mapping quantity with the personal baseline clinical reference body fat percentage.

[0012] As a preferred embodiment of the multi-parameter real-time human body fat data monitoring method of the present invention, wherein: The steps for performing baseline subtraction and triangular proximity smoothing on the body fat percentage to obtain a continuous body fat percentage curve are as follows: Select a baseline time window of fixed length, and calculate the arithmetic mean of all body fat percentages within the continuous time period covered by the baseline time window to obtain the individual baseline body fat percentage. At each uniform time marker where a body fat percentage has been generated, the difference between the body fat percentage and the individual baseline body fat percentage is arranged over time to obtain the body fat fluctuation sequence after baseline subtraction. Using the current unified time marker as the center marker, take the effective neighboring markers within half the support time range on both sides of the center marker, sort and number them according to their time distance from the center marker, determine the number of times to be repeated, and record the body fat fluctuation value after baseline subtraction in the temporary lists on both sides according to the number of repetitions. The temporary lists generated on both sides are merged, and the arithmetic mean of the merged temporary lists is taken to obtain the smoothed body fat fluctuation value at the current unified time mark. The sum of the smoothed body fat fluctuation value and the individual baseline body fat percentage is taken as the continuous body fat percentage at the current unified time mark. The continuous body fat percentage of each unified time mark is sorted sequentially by time to obtain the continuous body fat percentage curve.

[0013] As a preferred embodiment of the multi-parameter real-time human body fat data monitoring method of the present invention, the steps of freezing the body fat percentage when abnormal segments occur consecutively, and updating the continuous body fat percentage curve after recovery are as follows: If no body fat percentage is generated at the end of two adjacent time windows at the unified time mark, it is determined to be a continuous abnormal segment, the most recent valid output is retained and the unified time mark range is frozen; When body fat percentage is generated again in subsequent time windows, the freeze is lifted, the continuous body fat percentage at the same time mark is recalculated using effective neighboring markers, and the continuous body fat percentage curve is updated.

[0014] In a second aspect, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the multi-parameter real-time human body fat data monitoring method as described in the first aspect of the present invention.

[0015] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the multi-parameter real-time human body fat data monitoring method as described in the first aspect of the present invention.

[0016] The beneficial effects of this invention are as follows: By using baseline subtraction and triangular proximity smoothing, fluctuations caused by changes in body posture and breathing are effectively reduced, ensuring the smoothness and stability of the body fat percentage curve. The application of fractional suppression coupling and logically bounded mapping makes real-time body fat calculation unaffected by short-term interference, improving the accuracy of the calculation. When abnormal segments occur continuously, the most recently stable body fat percentage is frozen and continuously updated after recovery, avoiding fluctuations in results caused by temporary data loss and ensuring the continuity and reliability of the monitoring process. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the 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.

[0018] Figure 1 This is a flowchart of a method for real-time monitoring of multi-parameter human body fat data.

[0019] Figure 2 This is a flowchart for fragment shaping and data validity checks.

[0020] Figure 3 Flowchart for calculating body fat percentage.

[0021] Figure 4 This is a flowchart of a triangle proximity smoothing process. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] Reference Figures 1-4 As one embodiment of the present invention, this embodiment provides a method for real-time monitoring of multi-parameter human body fat data, including the following steps: S1. Obtain the multi-frequency impedance amplitude and impedance phase at a fixed beat, and continuously record the impedance change over time at the reference frequency point to form a data segment sequence sorted by time.

[0026] Skin cleaning and electrode attachment are performed at the same pair of surface electrode locations. Once stable contact is confirmed, the positions are locked in place. A global clock is established as a unified time reference, and fixed beats and frequency measurement sequences within each beat are set.

[0027] A global clock is a unique, continuous, and monotonically increasing time reference throughout the entire acquisition process.

[0028] The frequency measurement sequence refers to measuring all frequency points in ascending order from low to high within each fixed beat, until the highest frequency point.

[0029] When a beat begins, all frequency points within the beat are measured in strict order according to the frequency measurement sequence to obtain the impedance amplitude and impedance phase of each frequency point. After all frequency points of the beat have been measured, the next beat begins.

[0030] A reference frequency is selected and the impedance change over time is continuously recorded. The real part time series of the reference frequency is recorded. The reference frequency remains unchanged throughout the acquisition process and is synchronized with the global clock. This ensures that the reference frequency and the multi-frequency impedance amplitude and impedance phase obtained in each clock cycle are naturally aligned. No interpolation is performed for short-time interference; only time stamping is performed.

[0031] The reference frequency point refers to a single frequency point selected from all the frequency points used in this sampling, and the impedance change over time is continuously recorded at the selected single frequency point.

[0032] The real part time series of the reference frequency refers to a series of values ​​of the real part of the impedance (resistance component) as a function of time, obtained by continuously sampling the same pair of body surface electrodes at a fixed reference frequency and according to a uniform clock.

[0033] Using the start time of each beat as the reference time of the beat, the beat number and position within the beat are added to the impedance amplitude and impedance phase measurements of each frequency point within the beat, and the actual sampling time of the sample is added to each sample of the reference frequency point. The correspondence between the multi-frequency impedance amplitude and impedance phase measurements within the same beat and the reference frequency points on the same time axis is established, forming an alignment table.

[0034] A sample refers to a single impedance real part reading continuously recorded at a reference frequency point according to a unified time base.

[0035] The actual sampling time refers to the precise time scale at which the sample is actually measured and latched under the global clock.

[0036] Using a fixed-length time window (e.g., 12 seconds, since resting breathing is typically between 12 and 20 breaths per minute, 12 seconds can completely cover at least two to three breathing fluctuations) sliding from front to back on the same time axis, the start and end points of the time window are determined by a unified time marker. For each time window, the multi-frequency impedance amplitude and impedance phase of all non-reference frequency points within the time range are collected and written into data segments according to the beat number and position within the beat. Simultaneously, the real part time series of reference frequency points within the same time range are collected and bound to the data segments. A unique data segment number is assigned to each data segment, and the start time marker, end time marker, beat number range covered, frequency point list, and number of reference frequency point samples are recorded.

[0037] A frequency list is an ordered directory of frequency measurement points other than all reference frequencies within a data segment where impedance amplitude and impedance phase measurements were actually performed.

[0038] The obtained data segments are sorted according to the start time marker to form a time-sorted data segment sequence. If the time window slides to the end and is less than a complete time window, the last data segment is generated according to the actual time range covered and placed at the end of the sequence. An empty global multi-frequency amplitude and phase sequence and an empty global continuous time series are also established.

[0039] S2. Perform segment shaping on the data segments in the data segment sequence in sequence to obtain the effective multi-frequency amplitude and phase sequence and the effective continuous time sequence.

[0040] Within a data segment in the data segment sequence, select any non-reference frequency point, and take the first impedance phase of the current frequency point in the data segment as the starting value of the expanded phase of the current frequency point in the data segment.

[0041] The impedance phase of the current frequency point within the data segment is processed sequentially according to the time sequence. For each new moment, the phase difference between the impedance phase at the new moment and the impedance phase at the previous moment is calculated.

[0042] A new moment refers to the next moment immediately following the previous moment when the time stamps of the global clock are arranged in chronological order within the same data segment.

[0043] The phase difference between adjacent moments refers to the difference between the impedance phase at the new moment and the impedance phase at the previous moment.

[0044] When the phase difference between adjacent moments is between the negative half-phase period and the positive half-phase period, the phase difference between adjacent moments is considered as the actual change. When the phase difference between adjacent moments is greater than the positive half-phase period, a complete phase period is removed from the phase difference between adjacent moments to bring the phase difference between adjacent moments back to between the negative half-phase period and the positive half-phase period. When the phase difference between adjacent moments is less than the negative half-phase period, a complete phase period is superimposed on the phase difference between adjacent moments to bring the phase difference between adjacent moments back to between the negative half-phase period and the positive half-phase period.

[0045] Phase period refers to the period of a complete revolution. Radius (360°).

[0046] A positive half-phase period refers to Radius (+180°).

[0047] The negative half-phase period refers to Radius (−180°).

[0048] The sum of the expanded phase of the previous moment and the phase difference of the adjacent moment after correction is taken as the expanded phase of the current moment. This process is repeated for each frequency point, and the expanded phase and the phase difference of the adjacent moment after correction are continuously accumulated until the last impedance phase of the current frequency point in the data segment is processed, forming the expanded phase sequence of the current frequency point in the data segment. The remaining non-reference frequency points in the data segment are then expanded one by one.

[0049] When processing the next data segment, for the same non-reference frequency point, the expanded phase at the end of the previous data segment is used as the new starting value, and frequency-by-frequency processing continues in the new data segment. If there is a very short time gap between two adjacent data segments, no interpolation is performed, and only the time mark is continued.

[0050] Within a data segment, using each beat as a boundary, check whether the overall direction of change of the real part time series of the reference frequency point within the beat is consistent with the overall direction of change of the expanded phase of the lowest frequency point (the non-reference frequency point with the smallest value in the frequency list) within the beat. If the overall direction of change of the real part time series of the reference frequency point within the beat and the overall direction of change of the expanded phase of the lowest frequency point are both upward or downward, the data segment is marked as a consistent segment. If the overall direction of change of the real part time series of the reference frequency point within the beat and the overall direction of change of the expanded phase of the lowest frequency point are different, the data segment is marked as an inconsistent segment and splicing stops. If either the overall direction of change of the real part time series of the reference frequency point within the beat and the overall direction of change of the expanded phase of the lowest frequency point are flat, then it is not counted.

[0051] The overall direction of change of the real part time series of the reference frequency point within a beat refers to taking the real part of the reference frequency point at the earliest and latest unified time mark within the beat, comparing the sign of the arithmetic difference between the real part of the latest reference frequency point and the real part of the earliest reference frequency point. When the arithmetic difference is greater than zero, it is recorded as upward; when the arithmetic difference is less than zero, it is recorded as downward; and when the arithmetic difference is equal to zero, it is recorded as flat.

[0052] The overall direction of change of the unfolded phase of the lowest frequency point refers to the sign of the arithmetic difference between the unfolded phase of the lowest frequency point in the current beat and the current beat value. When the arithmetic difference is greater than zero, it is recorded as upward; when the arithmetic difference is less than zero, it is recorded as downward; and when the arithmetic difference is equal to zero, it is recorded as flat.

[0053] For a consistent segment, the impedance amplitude and expanded phase of each non-reference frequency point are appended to the end of the global multi-frequency amplitude and phase sequence in a unified time-marking order. At the same time, the real part time sequence of the reference frequency point is appended to the end of the global continuous time sequence. After all data segments are processed and spliced, the global multi-frequency amplitude and phase sequence arranged with the unified time mark is the effective multi-frequency amplitude and phase sequence, and the global continuous time sequence of the reference frequency point is the effective continuous time sequence.

[0054] S3. Extract multispectral shape representation and temporal shape representation from effective multi-frequency amplitude and phase sequences and effective continuous time sequences, and perform fractional suppression coupling and logical bounded mapping to obtain body fat percentage.

[0055] Using a fixed-length time window, processing is performed window by window from front to back. For each time window, the list of frequency points covered by the time window is located from the effective multi-frequency amplitude and phase sequence, and at the same time, the real part time sequence of the reference frequency point within the same time range is extracted from the effective continuous time sequence. If there is no last impedance amplitude or impedance phase available for any non-reference frequency point in the time window at the end of the time window, the body fat percentage is not generated at the end of the time window and the next time window is processed.

[0056] In the frequency point list within the time window, the smallest non-reference frequency point is taken as the low frequency point, the largest non-reference frequency point is taken as the high frequency point, and the non-reference frequency point at the position in the frequency point list is taken as the characteristic frequency point. According to the unified time mark of the time window and the beat number and position within the beat corresponding to the frequency point, the impedance amplitude and expanded phase of the low frequency point, high frequency point and characteristic frequency point are read from the effective multi-frequency amplitude and phase sequence. The impedance amplitude and expanded phase of the low frequency point and high frequency point are calculated by the real part of the impedance, and the real parts of the low frequency impedance and high frequency impedance are obtained respectively. The ratio of the difference between the real parts of the low frequency impedance and the real parts of the high frequency impedance to the real part of the low frequency impedance is used as the normalization ratio of the low and high frequency impedance difference. The phase shape quantity is obtained by tangent mapping of the expanded phase of the characteristic frequency point in the effective multi-frequency amplitude and phase sequence.

[0057] Within the same time window, the real part time series of the reference frequency is taken from the effective continuous time series, and mean-reduction is performed within the time window so that the real part time series of the reference frequency fluctuates around zero. The amplitude envelope of the real part time series of the reference frequency is calculated, and normalized by the average absolute amplitude of the real part time series of the reference frequency within the time window to obtain the modulation intensity. Within the same time window, the local average value of the real part of the reference frequency impedance is calculated, and the natural logarithm of the ratio of the local average value of the real part of the reference frequency impedance to the real part of the reference frequency impedance is taken to obtain the baseline normalization.

[0058] Under the same time window and unified time marker, the normalized ratio of the low-to-high frequency impedance difference is summed with the phase morphology quantity, and fractional suppression coupling is performed through modulation intensity to automatically reduce the impact of increased respiration and body posture fluctuations on the instantaneous estimation. The spectral characterization after fractional suppression coupling is summed with the baseline normalization quantity to obtain an intermediate quantity. A dimensionless body fat mapping quantity is obtained through logically bounded mapping. A personal baseline time period is selected, and the clinical reference body fat percentage is measured. The clinical reference body fat percentage of the personal baseline time period is used as the personal baseline percentage, and the baseline body fat mapping quantity of the personal baseline time period is recorded simultaneously. The body fat percentage at any unified time marker thereafter is the product of the ratio of the current body fat mapping quantity to the baseline body fat percentage and the personal baseline percentage. When the baseline body fat mapping quantity is missing or zero, the body fat percentage is not output. The expression is: ; ; in, This indicates the percentage of body fat aligned to individual baselines. This represents the reference body fat percentage obtained using clinically comparable methods within an individual's baseline period. This represents the dimensionless body fat mapping quantity after fractional suppression and logically bounded mapping. This represents the baseline body fat mapping value, which is the arithmetic mean of the body fat mapping values ​​over an individual's baseline time period. This represents a logical function that maps real numbers to bounded intervals. This represents the normalized proportion of the impedance difference between low and high frequencies. Represents phase shape quantity. Indicates modulation intensity. Indicates the baseline normalization quantity. It represents a unified time stamp.

[0059] S4. Perform baseline subtraction and triangular proximity smoothing on the body fat percentage to obtain a continuous body fat percentage curve. When abnormal segments appear continuously, freeze the body fat percentage and update the continuous body fat percentage curve after recovery.

[0060] Select a fixed-length baseline time window, which is composed of several adjacent and non-overlapping time windows. The body fat percentage should have been successfully generated at each uniform time mark within the baseline time window. If there are any missing points, extend the entire baseline time window forward until it is continuously usable. Calculate the arithmetic mean of all body fat percentages within the continuous time period covered by the baseline time window to obtain the individual baseline body fat percentage.

[0061] For each uniform time stamp of the generated body fat percentage, the difference between the body fat percentage and the individual baseline body fat percentage is arranged by time to obtain the body fat fluctuation sequence after baseline subtraction, which reflects the instantaneous shift relative to the individual baseline.

[0062] Centered on the current unified time marker, effective neighboring markers within half the support time range on both sides of the current unified time marker are selected. The linear relationship of greater influence as the distance increases is reflected by the replication count. The effective neighboring markers on both sides are sorted and numbered sequentially from the nearest to the farthest time distance from the center, starting from 1 and increasing. The number of times each side is recorded is equal to the difference between the total number of effective neighboring markers on both sides and the sorting number of the unified time marker plus one. For each effective neighboring marker, the body fat fluctuation value after baseline subtraction corresponding to the effective neighboring marker is repeatedly recorded in a temporary list. Unified time markers outside the half support time range on both sides are not recorded.

[0063] Half-support duration refers to the maximum time offset that can be included in the smoothing calculation, centered on any uniform time marker, when performing triangular proximity smoothing. The value range is positive time greater than zero and does not exceed the baseline time window.

[0064] Effective neighbor markers refer to a set of unified time markers that participate in the smoothing calculation of the current unified time marker.

[0065] A temporary list is a list of baseline-subtracted body fat fluctuation values ​​temporarily constructed for triangular proximity smoothing at any uniform time point.

[0066] The temporary lists generated on both sides are merged, and the arithmetic mean of the merged temporary lists is taken to obtain the smoothed body fat fluctuation value at the current unified time mark. The sum of the smoothed body fat fluctuation value and the individual baseline body fat percentage is taken as the continuous body fat percentage at the current unified time mark. The continuous body fat percentage of each unified time mark is sorted sequentially according to the unified time mark to obtain the continuous body fat percentage curve.

[0067] When no body fat percentage is generated at the end of two adjacent time windows, a continuous abnormal segment is determined to begin from the second time window without output. During the continuous abnormal segment, the continuous body fat percentage curve remains at the value of the most recent output before the occurrence of the abnormal segment. The range of the unified time mark corresponding to the continuous abnormal segment is frozen in the result sequence. When the body fat percentage is generated again at the end of the unified time mark in the subsequent time window, the freeze is lifted, and the continuous body fat percentage at the unified time mark is recalculated and output with the available effective neighboring markers, and the continuous body fat percentage curve is updated.

[0068] Each continuous body fat percentage is bound to a uniform time stamp and written sequentially into the result sequence, which is then used as the final continuous output.

[0069] This embodiment also provides a computer device applicable to the multi-parameter real-time human body fat data monitoring method, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the multi-parameter real-time human body fat data monitoring method proposed in the above embodiment.

[0070] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0071] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the method for real-time monitoring of multi-parameter human body fat data as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0072] In summary, this invention effectively reduces fluctuations caused by changes in body posture and respiration by using baseline subtraction and triangular proximity smoothing, ensuring the smoothness and stability of the body fat percentage curve. By applying fractional suppression coupling and logically bounded mapping, real-time body fat calculation is not affected by short-term interference, thus improving the accuracy of the calculation. When abnormal segments occur continuously, the most recently stable body fat percentage is frozen and continuously updated after recovery, avoiding fluctuations in results caused by temporary data loss and ensuring the continuity and reliability of the monitoring process.

[0073] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for real-time monitoring of multi-parameter human body fat data, characterized in that: include, The multi-frequency impedance amplitude and impedance phase are acquired at a fixed beat, and the impedance changes over time are continuously recorded at a reference frequency point to form a data segment sequence sorted by time. The data segments in the data segment sequence are sequentially segmented to obtain an effective multi-frequency amplitude-phase sequence and an effective continuous time sequence; Multispectral and temporal morphological representations are extracted from effective multi-frequency amplitude-phase sequences and effective continuous time sequences, and fractional suppression coupling and logical bounded mapping are performed to obtain body fat percentage. Baseline subtraction and triangular proximity smoothing are applied to the body fat percentage to obtain a continuous body fat percentage curve. When abnormal segments appear continuously, the body fat percentage is frozen, and the continuous body fat percentage curve is updated after recovery.

2. The method for real-time monitoring of multi-parameter human body fat data as described in claim 1, characterized in that: The specific steps for forming a time-sorted sequence of data segments are as follows: Set a fixed beat and the frequency measurement sequence within each beat. When the beat begins, measure all frequency points within the beat in sequence according to the frequency measurement sequence, and obtain the impedance amplitude and impedance phase of each frequency point respectively. Select a reference frequency point and record the real part time series of the reference frequency point. Using a fixed-length time window on the time axis, the multi-frequency impedance amplitude and phase of non-reference frequency points within each time window are collected and written into data segments according to the beat number and position. At the same time, the real part time series of reference frequency points is collected synchronously and bound to the data segments. The bound data segments are sorted by time to form a time-sorted data segment sequence, and a spatial global multi-frequency amplitude and phase sequence and a spatial global continuous time series are established.

3. The method for real-time monitoring of multi-parameter human body fat data as described in claim 2, characterized in that: The specific steps for sequentially shaping the data segments in the data segment sequence are as follows: Within a data segment in the data segment sequence, select any non-reference frequency point, and take the first impedance phase of the current frequency point in the data segment as the starting value of the expanded phase of the current frequency point in the data segment; In chronological order, the impedance phase of the current frequency point within the data segment is processed sequentially. For each new moment, the phase difference between the impedance phase at the new moment and the impedance phase at the previous moment is calculated. When the phase difference between adjacent moments is between the negative half-phase period and the positive half-phase period, the phase difference between adjacent moments is regarded as the actual change. When the phase difference between adjacent moments is greater than the positive half-phase period, a complete phase period is deleted from the phase difference between adjacent moments. When the phase difference between adjacent moments is less than the negative half-phase period, a complete phase period is superimposed on the phase difference between adjacent moments. The sum of the expanded phase from the previous moment and the phase difference between the next corrected moment is taken as the expanded phase at the current moment. This process is repeated point by point, and the expanded phase and the phase difference between the next corrected moment are continuously accumulated until the last impedance phase of the current frequency point in the data segment is processed.

4. The method for real-time monitoring of multi-parameter human body fat data as described in claim 3, characterized in that: The specific steps to obtain the effective multi-frequency amplitude-phase sequence and the effective continuous time sequence are as follows: Within a data segment, taking each beat as a boundary, check whether the overall change direction of the real part time series of the reference frequency point within the beat is consistent with the overall change direction of the unfolded phase of the lowest frequency point within the beat. Data segments that pass the check are marked as consistent segments, and data segments that fail the check are marked as inconsistent segments and splicing is stopped. By splicing consistent segments according to a unified time mark, the impedance amplitude and expanded phase of non-reference frequency points are added to the global multi-frequency amplitude and phase sequence, and the real part time sequence of reference frequency points is added to the global continuous time sequence. After all data segments are processed and spliced, the effective multi-frequency amplitude and phase sequence and the effective continuous time sequence are obtained.

5. The method for real-time monitoring of multi-parameter human body fat data as described in claim 4, characterized in that: The specific steps for extracting multi-spectral shape representations and temporal shape representations from effective multi-frequency amplitude-phase sequences and effective continuous time sequences are as follows: Within a fixed-length time window, low-frequency points, high-frequency points, and characteristic frequency points are selected from the frequency point list. The impedance amplitude and expanded phase of the low-frequency points, high-frequency points, and characteristic frequency points are read from the effective multi-frequency amplitude-phase sequence. The real parts of the impedance at low and high frequencies are calculated. The ratio of the difference between the real parts of the low-frequency impedance and the real parts of the high-frequency impedance to the real part of the low-frequency impedance is used as the normalization ratio of the low-frequency impedance difference. The phase shape quantity is obtained by tangent mapping of the expanded phase of the characteristic frequency point in the effective multi-frequency amplitude-phase sequence. The normalization ratio of the low-frequency impedance difference and the phase shape quantity are the multi-spectral shape characterization. Within the same time window, the real part time series of the reference frequency is taken from the effective continuous time series, and mean-reduction is performed within the time window. The amplitude envelope of the real part time series of the reference frequency after mean-reduction is calculated, and normalized by the average absolute amplitude of the real part time series of the reference frequency after mean-reduction within the time window to obtain the modulation intensity. Within the same time window, the local average value of the real part of the reference frequency impedance is calculated, and the natural logarithm of the ratio of the local average value of the real part of the reference frequency impedance to the real part of the reference frequency impedance is taken to obtain the baseline normalization. The modulation intensity and the baseline normalization are the time shape characteristics.

6. The method for real-time monitoring of multi-parameter human body fat data as described in claim 5, characterized in that: The steps for performing fractional suppression coupling and logically bounded mapping to obtain body fat percentage are as follows: Under the same time window and unified time mark, the normalized ratio of the low-frequency impedance difference is summed with the phase morphology quantity, and the coupling is suppressed by the modulation intensity. The intermediate quantity is obtained by summing the fractional suppression coupled quantity with the baseline normalized quantity. The body fat mapping quantity is obtained through logically bounded mapping and a personal baseline is established. The body fat percentage is obtained by multiplying the ratio of the current body fat mapping quantity and the personal baseline body fat mapping quantity with the personal baseline clinical reference body fat percentage.

7. The method for real-time monitoring of multi-parameter human body fat data as described in claim 6, characterized in that: The steps for performing baseline subtraction and triangular proximity smoothing on the body fat percentage to obtain a continuous body fat percentage curve are as follows: Select a baseline time window of fixed length, and calculate the arithmetic mean of all body fat percentages within the continuous time period covered by the baseline time window to obtain the individual baseline body fat percentage. At each uniform time marker where a body fat percentage has been generated, the difference between the body fat percentage and the individual baseline body fat percentage is arranged over time to obtain the body fat fluctuation sequence after baseline subtraction. Using the current unified time marker as the center marker, take the effective neighboring markers within half the support time range on both sides of the center marker, sort and number them according to their time distance from the center marker, determine the number of times to be repeated, and record the body fat fluctuation value after baseline subtraction in the temporary lists on both sides according to the number of repetitions. The temporary lists generated on both sides are merged, and the arithmetic mean of the merged temporary lists is taken to obtain the smoothed body fat fluctuation value at the current unified time mark. The sum of the smoothed body fat fluctuation value and the individual baseline body fat percentage is taken as the continuous body fat percentage at the current unified time mark. The continuous body fat percentage of each unified time mark is sorted sequentially by time to obtain the continuous body fat percentage curve.

8. The method for real-time monitoring of multi-parameter human body fat data as described in claim 7, characterized in that: When consecutive abnormal segments appear, the body fat percentage is frozen, and the continuous body fat percentage curve is updated after recovery. The specific steps are as follows: If no body fat percentage is generated at the end of two adjacent time windows at the unified time mark, it is determined to be a continuous abnormal segment, the most recent valid output is retained and the unified time mark range is frozen; When body fat percentage is generated again in subsequent time windows, the freeze is lifted, the continuous body fat percentage at the same time mark is recalculated using effective neighboring markers, and the continuous body fat percentage curve is updated.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the multi-parameter real-time human body fat data monitoring method according to any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the multi-parameter real-time human body fat data monitoring method according to any one of claims 1 to 8.

Images