Calibration method of human body composition analyzer

Abstract

The invention discloses a calibration method of a human body composition analyzer. The method comprises the following steps: measuring through a human body composition analyzer to obtain a human body impedance spectrum of a measured object; fitting according to the human body impedance spectrum to obtain a target Cole-Cole model; determining a fitting residual error between the module value of the target Cole-Cole model and the human body impedance spectrum; and if the fitting residual error is smaller than or equal to a preset residual error threshold value, determining human body composition data according to the target Cole-Cole model. According to the technical scheme provided by the invention, the Cole-Cole model is fitted, and the fitting model is directly used for replacing the measurement value in the allowable fitting residual range, so that measurement abnormity caused by various interference factors can be effectively discriminated and automatically corrected, and the accuracy and robustness of the human body composition analyzer in different use scenes are improved.

Description

Technical Field

[0001] This invention relates to the field of bioelectrical impedance analysis technology, and more particularly to a calibration method for a human body composition analyzer. Background Technology

[0002] Current body composition analyzers generally apply a small AC signal to the human body using a fixed-frequency or multi-frequency current and estimate body composition such as water, muscle, and fat by measuring the impedance response. However, due to interference factors such as electrode contact quality, wire reactance, and the condition of the human skin, the measurement results of current body composition analyzers may have significant errors during the measurement process. Summary of the Invention

[0003] This invention provides a calibration method for a body composition analyzer to achieve highly accurate and robust body composition measurement.

[0004] This invention provides a calibration method for a body composition analyzer, the method comprising:

[0005] The human impedance spectrum of the subject was obtained by measuring the human body composition analyzer.

[0006] The target Cole-Cole model is obtained by fitting the human body impedance spectrum.

[0007] Determine the fitting residual between the target Cole-Cole model modulus and the human body impedance spectrum;

[0008] If the fitting residual is less than or equal to a preset residual threshold, then the human body composition data are determined based on the target Cole-Cole model.

[0009] Optionally, after determining the fitting residual between the target Cole-Cole model magnitude and the human body impedance spectrum, the method further includes:

[0010] If the fitting residual is greater than the preset residual threshold, the user is prompted to remeasure.

[0011] Optionally, the step of obtaining the human impedance spectrum of the subject by measuring the human body composition analyzer includes:

[0012] Multi-frequency excitation signals are applied to the subject through the electrodes of the body composition analyzer, and the electrical signals on the electrodes are sampled.

[0013] The human body impedance spectrum is calculated based on the multi-frequency excitation signal and the sampled electrical signal.

[0014] Optionally, the multi-frequency excitation signal includes sinusoidal current excitation signals of different frequencies.

[0015] Optionally, the frequency range of the sinusoidal current excitation signal is 1kHz-1MHz.

[0016] Optionally, the multi-frequency excitation signal can be applied sequentially with a single frequency or applied by multi-frequency synthesis.

[0017] Optionally, the method further includes:

[0018] The measured impedance spectrum of the standard bioimpedance simulation fixture was obtained by measuring with a human body composition analyzer. The standard bioimpedance simulation fixture is used to simulate multi-frequency bioimpedance spectrum.

[0019] The measured impedance spectrum is compared with the theoretical impedance spectrum of the standard bioimpedance simulation fixture to correct the error parameters of the internal hardware pathway of the human body composition analyzer based on the comparison results.

[0020] Optionally, the standard bioimpedance simulation fixture includes multiple sets of parallel target RC modules, the component parameters of which are designed based on the fitting results of the Cole-Cole model of human body impedance.

[0021] Optionally, the standard bioimpedance simulation device specifically includes RC module combinations corresponding to different populations; correspondingly, obtaining the measured impedance spectrum of the standard bioimpedance simulation device through a body composition analyzer includes:

[0022] The target RC module selected in the standard bioimpedance simulation fixture is switched according to the target population corresponding to the tested object.

[0023] Optionally, the different groups of people may be divided based on one or more of the following: age, gender, BMI, health status, and exercise status.

[0024] This invention provides a calibration method for a body composition analyzer. First, the body impedance spectrum of the subject is measured using the body composition analyzer. Then, a target Cole-Cole model is fitted based on the body impedance spectrum. Next, the fitting residual between the target Cole-Cole model magnitude and the body impedance spectrum is determined. If the obtained fitting residual is less than or equal to a preset residual threshold, the body composition data is determined based on the target Cole-Cole model. This calibration method for a body composition analyzer, by fitting a Cole-Cole model and directly using the fitted model to replace the measured values ​​within an allowable fitting residual range, can effectively identify and automatically correct measurement anomalies caused by various interference factors, thereby improving the accuracy and robustness of the body composition analyzer in different usage scenarios. Attached Figure Description

[0025] Figure 1A flowchart illustrating the calibration method for a body composition analyzer provided in an embodiment of the present invention;

[0026] Figure 2 This is a schematic diagram showing the connection of the measuring instrument electrodes to the human body in an embodiment of the present invention. Detailed Implementation

[0027] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, the accompanying drawings show only the parts relevant to the present invention, and not all of the structures.

[0028] Before discussing the exemplary embodiments in more detail, it should be noted that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe the steps as sequential processes, many of these steps can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the steps can be rearranged. The process can be terminated when its operation is complete, but may also have additional steps not included in the figures. The process can correspond to a method, function, procedure, subroutine, subroutine, etc.

[0029] Figure 1 This is a flowchart illustrating a calibration method for a body composition analyzer provided in an embodiment of the present invention. This embodiment is applicable to situations where equipment calibration is performed when using a body composition analyzer for measurements. Figure 1 As shown, the method specifically includes the following steps:

[0030] S11. The human body impedance spectrum of the subject is obtained by measuring the human body composition analyzer.

[0031] S12. Obtain the target Cole-Cole model by fitting the human body impedance spectrum.

[0032] S13. Determine the fitting residual between the target Cole-Cole model modulus and the human body impedance spectrum.

[0033] S14. If the fitting residual is less than or equal to a preset residual threshold, then the human body composition data is determined according to the target Cole-Cole model.

[0034] Specifically, the measurement can be initiated after the user wears the electrodes of the body composition analyzer normally. Any existing measurement method can be used to measure the subject, specifically an eight-electrode method to reduce the influence of contact resistance. During measurement, each electrode is worn on a designated part of the body, such as one pair of electrodes connected to each hand and one pair to each foot. Four electrodes on different parts of the body serve as driving electrodes for current injection, while the other four electrodes serve as sensing electrodes for signal measurement. The connection is as follows: Figure 2 As shown, taking the torso as an example, the current flow direction is HIL(IOUT3)→FIL(IOUT1), and the measurement points include FVR(VSENSE1) and HVR(VSENSE2). The body composition analyzer can synchronously sample the voltage on each sensing electrode through a high input impedance amplifier and an ADC module, and then calculate the impedance based on the injected current and the sampled voltage to obtain the human body impedance spectrum.

[0035] Optionally, the step of measuring the human impedance spectrum of the subject using a human body composition analyzer includes: applying a multi-frequency excitation signal to the subject through the electrodes of the human body composition analyzer and sampling the electrical signal on the electrodes; and calculating the human impedance spectrum based on the multi-frequency excitation signal and the sampled electrical signal.

[0036] Specifically, the body composition analyzer can generate multi-frequency excitation signals using a constant current source, which can be injected into the human body through the aforementioned driving electrodes to measure the impedance response of the subject under multi-frequency excitation. Optionally, the multi-frequency excitation signals include sinusoidal current excitation signals of different frequencies; further optionally, the frequency range of the sinusoidal current excitation signals is 1kHz-1MHz to accommodate changes in human body impedance. Based on the above technical solution, the multi-frequency excitation signals can optionally be applied sequentially (swept-sine) or synthesized (e.g., a composite signal obtained by superimposing multiple sinusoids). Simultaneously with the application of the multi-frequency excitation signals, the voltage on the aforementioned sensing electrodes can be sampled synchronously. Correspondingly, if a multi-frequency synthesis application method is used, the sampled voltages can be frequency-domain separated using methods such as Fast Fourier Transform (FFT) to obtain the electrical signal values ​​corresponding to each frequency. After sampling the voltage values ​​corresponding to the multi-frequency excitation signal, for each frequency f, the impedance value can be calculated according to Ohm's law Z(f) = V(f) / I(f), where Z(f) is the impedance value at frequency f, V(f) is the voltage value at the sampled frequency f, and I(f) is the current value at frequency f in the multi-frequency excitation signal, thus obtaining the human body impedance spectrum. Simultaneously, impedance is usually expressed as a complex number Z(f) = R(f) + jX(f), so the magnitude |Z| and phase angle θ at each frequency f can be further extracted for subsequent calculations.

[0037] After obtaining the human body impedance spectrum, a nonlinear fitting of the Cole-Cole model can be performed based on the human body impedance spectrum to obtain the model parameters and the fitted target Cole-Cole model. The Cole-Cole model formula is:

[0038]

[0039] Where R0 represents the low-frequency limiting resistance (in Ω), R∞ represents the high-frequency limiting resistance (in Ω), τ represents the time constant, α represents the Cole relaxation coefficient, and ω = 2πf, where f represents the frequency. Specifically, the Trust Region Reflective (TRF) algorithm can be used for fitting. The fitting objective is to find the four parameters (R0, R∞, τ, α) in the Cole-Cole model based on the human impedance spectrum (fi, |Zi|) such that the model's magnitude |Zmodel(fi; R0, R∞, τ, α)| is closest to the true value. Fitting can begin by initializing the model parameters based on a set frequency range and empirical values, and then iterating until convergence (e.g., when the error change is small or the maximum number of iterations is reached). In each iteration, the human impedance spectrum is first substituted into the calculation of the current model's magnitude |Zmodel|, then the residual vector ri = |Zmodel(fi)| - |Zmeasured(fi)| between the current model's magnitude and the corresponding measured impedance value is calculated, and the Jacobian matrix is ​​constructed. That is, the partial derivative of the residual vector with respect to each model parameter, where pi is the model parameter vector. Then, within the predefined trust region, a linear subproblem is solved: min δ ||J·δ+r|| 2 subjectto||δ||≤Δ, where δ is the correction to the model parameters, and Δ represents the confidence radius (i.e., the range currently allowed to be searched in the parameter space). After obtaining the optimal δ, try to use δ to correct the original model parameters, i.e., p new =p old The algorithm increments by δ and calculates the new residuals. If the residuals decrease, the revised model parameters are accepted, and the confidence radius can be appropriately increased. Otherwise, the revision is rejected, the confidence radius is reduced, and the next iteration begins. After completing the iteration process, the optimal model parameters are obtained, thus determining the desired target Cole-Cole model.

[0040] After obtaining the target Cole-Cole model, the fitting residual between the final model modulus and the measured human impedance spectrum can be determined, and this fitting residual can be compared with a preset residual threshold. If the fitting residual is less than or equal to the preset residual threshold, it can be considered that only a weak interference exists, the fitting result is reliable, and the fitting result can absorb the influence of interference. In this case, the original human impedance spectrum can be skipped, and the fitting result can be directly used as the final measurement result, thereby improving robustness. Even if the contact state is imperfect, the fitting result can correct the error. Furthermore, based on the fitted target Cole-Cole model, human body composition data, such as water percentage, muscle mass, fat mass, and other core BIA indicators, can be calculated, and a human body composition report can be generated to provide feedback to the user to complete the measurement. In addition, with long-term use, it is possible to combine historical parameters to perform trend analysis, intervention suggestions, and other intelligent function expansions.

[0041] Optionally, after determining the fitting residual between the target Cole-Cole model magnitude and the human body impedance spectrum, the method further includes: if the fitting residual is greater than the preset residual threshold, prompting the user to remeasure. Specifically, if the fitting residual is greater than the preset residual threshold, it can be considered that there is significant interference, and both the fitting result and the original measurement result are unreliable. In this case, the user can be prompted to remeasure, for example, by indicating possible interference such as poor electrode contact, poor skin condition, or component damage, and prompting the user to adjust the electrodes, wipe the skin, replace the component, etc. The measurement process can then be repeated until the new fitting residual is less than or equal to the preset residual threshold, and the human body composition data is determined based on the finally fitted target Cole-Cole model. This residual control ensures data reliability and improves the reliability of the measurement results.

[0042] In an optional embodiment, the method further includes: measuring the measured impedance spectrum of a standard bioimpedance simulation fixture using a body composition analyzer, the standard bioimpedance simulation fixture being used to simulate multi-frequency bioimpedance spectrum; comparing the measured impedance spectrum with the theoretical impedance spectrum of the standard bioimpedance simulation fixture, so as to correct the error parameters of the internal hardware pathway of the body composition analyzer based on the comparison result.

[0043] Specifically, initial calibration can be performed each time the body composition analyzer is started. First, a standard bioimpedance simulation fixture is connected to the electrodes of the body composition analyzer, and the measured impedance spectrum is obtained through the analyzer. The measurement process is similar to the measurement process of the tested object, measuring its impedance response by applying multi-frequency excitation signals. The standard bioimpedance simulation fixture can simulate a multi-frequency human impedance spectrum through physical circuitry, serving as a standard signal source for instrument self-calibration. Similar to actual user measurements, the standard bioimpedance simulation fixture can provide interfaces for simulating the connection positions of the hands and feet, respectively, and can be connected to the electrodes of the body composition analyzer to simulate measurements of different parts of the human body. Correspondingly, the standard bioimpedance simulation fixture may include circuitry for simulating the impedance of different parts of the human body.

[0044] After obtaining the measured impedance spectrum of the standard bioimpedance simulation fixture, this measured impedance spectrum can be compared with the theoretical impedance spectrum of the standard bioimpedance simulation fixture. Based on the comparison results, the body composition analyzer can be physically calibrated. The calibration process is similar to the calibration of a weighing scale using weights, thereby automatically correcting error parameters such as gain, phase, ADC, electrode bias, and nonlinearity of the internal hardware pathways, ensuring the stability of the instrument's measurement reference. The theoretical impedance spectrum can be calculated based on the parameters of each component in the fixture and the different frequencies used for measurement. Furthermore, this physical calibration process can be combined with the aforementioned algorithmic calibration process to achieve a dual calibration mechanism, ensuring high accuracy from hardware to algorithm. It also features a traceable calibration chain, supports long-term stable operation, meets medical-grade standards, has a simple structure, low implementation cost, and can be used in portable or home-based devices.

[0045] Optionally, the standard bioimpedance simulation fixture includes multiple sets of parallel-connected target RC modules. The component parameters of these target RC modules are designed based on the fitting results of a Cole-Cole model of human impedance. Each set of target RC modules includes a resistor and a capacitor unit connected in series. Multiple sets can have their resistor and capacitor units connected in parallel to simulate human impedance data. After the target RC modules are connected in parallel, they can be connected in series with a high-frequency limiting resistance R∞ simulating the human body. A voltage input terminal Vin and a voltage output terminal Vout are connected to the two ends of the series connection to simulate impedance behavior under applied electrical stimulation by connecting an external power source. By using RC modules to simulate impedance values, the impedance behavior characteristics of a specified object at different frequencies can be accurately simulated. Furthermore, by designing based on the fitting results of the Cole-Cole model, rather than relying on empirical values, the accuracy of the simulated impedance can be further improved.

[0046] Specifically, the design process begins by acquiring impedance test data of the sample object within a specified frequency range (e.g., 1kHz-1MHz). Then, a reference Cole-Cole model is fitted based on this impedance test data; the fitting process can refer to the fitting process of the target Cole-Cole model described above. After determining the reference Cole-Cole model, the equivalent physical network can be constructed. First, N groups of target RC modules are selected and connected in parallel. The component parameters (Ri, Ci) of each target RC module correspond to a time constant τi = RiCi. The total parallel response of all target RC modules is then:

[0047]

[0048] R∞ can be directly adopted from the fitted reference Cole-Cole model. Then, optimization algorithms such as TRF and genetic algorithms can be used to optimize the component parameters of each target RC module with the goal of minimizing the amplitude frequency or phase frequency error between the parallel total response and the reference Cole-Cole model, ultimately achieving an amplitude frequency / phase frequency relative error of <0.5%.

[0049] Further optionally, the standard bioimpedance simulation device specifically includes combinations of RC modules corresponding to different populations; correspondingly, the step of obtaining the measured impedance spectrum of the standard bioimpedance simulation device by measuring the body composition analyzer includes: switching the target RC module selected in the standard bioimpedance simulation device according to the target population corresponding to the tested object.

[0050] Specifically, there may be significant differences in impedance values ​​among different population groups. Impedance matching can be achieved by combining different numbers and parameters of RC modules for different population groups, thereby improving the accuracy of the simulated impedance. Optionally, the different population groups can be categorized based on one or more of age, gender, BMI, health status, and activity level. Therefore, the standard bioimpedance simulation fixture can pre-provide the required RC module combinations for different population groups. The RC parameters for each RC module combination can be designed by referring to the component parameter design process of the target RC modules. Furthermore, the standard bioimpedance simulation fixture can switch between different RC module combinations using one or more of DIP switches, jumper caps, and programmable analog switches, thus achieving adjustability. During the calibration of the body composition analyzer, the standard bioimpedance simulation fixture can be switched to the corresponding target RC module based on the target population of the subject to be measured, before performing the calibration process described above, to improve calibration accuracy. Furthermore, the resistor units and capacitor units in each RC module can be adjustable resistor units and adjustable capacitor units, making adjustment more convenient.

[0051] The technical solution provided in this invention first measures the human impedance spectrum of the subject using a human body composition analyzer. Then, a target Cole-Cole model is fitted based on this impedance spectrum. Next, the fitting residual between the target Cole-Cole model's modulus and the human impedance spectrum is determined. If the obtained fitting residual is less than or equal to a preset residual threshold, the human body composition data is determined based on the target Cole-Cole model. By fitting a Cole-Cole model and directly using the fitted model to replace the measured values ​​within the allowable fitting residual range, measurement anomalies caused by various interference factors can be effectively identified and automatically corrected, thereby improving the accuracy and robustness of the human body composition analyzer in different usage scenarios.

[0052] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.

Claims

1. A calibration method for a body composition analyzer, characterized in that, include: The human impedance spectrum of the subject was obtained by measuring the human body composition analyzer. The target Cole-Cole model is obtained by fitting the human body impedance spectrum. Determine the fitting residual between the target Cole-Cole model modulus and the human body impedance spectrum; If the fitting residual is less than or equal to a preset residual threshold, then the human body composition data are determined based on the target Cole-Cole model.

2. The calibration method for the body composition analyzer according to claim 1, characterized in that, After determining the fitting residual between the target Cole-Cole model magnitude and the human body impedance spectrum, the method further includes: If the fitting residual is greater than the preset residual threshold, the user is prompted to remeasure.

3. The calibration method for the body composition analyzer according to claim 1, characterized in that, The method of obtaining the human body impedance spectrum of the subject by measuring the human body composition analyzer includes: Multi-frequency excitation signals are applied to the subject through the electrodes of the body composition analyzer, and the electrical signals on the electrodes are sampled. The human body impedance spectrum is calculated based on the multi-frequency excitation signal and the sampled electrical signal.

4. The calibration method for the body composition analyzer according to claim 3, characterized in that, The multi-frequency excitation signal includes sinusoidal current excitation signals of different frequencies.

5. The calibration method for the body composition analyzer according to claim 4, characterized in that, The frequency range of the sinusoidal current excitation signal is 1kHz-1MHz.

6. The calibration method for the body composition analyzer according to any one of claims 3-5, characterized in that, The multi-frequency excitation signal is applied sequentially at a single frequency or synthesized from multiple frequencies.

7. The calibration method for the body composition analyzer according to claim 1, characterized in that, The method further includes: The measured impedance spectrum of the standard bioimpedance simulation fixture was obtained by measuring with a human body composition analyzer. The standard bioimpedance simulation fixture is used to simulate multi-frequency bioimpedance spectrum. The measured impedance spectrum is compared with the theoretical impedance spectrum of the standard bioimpedance simulation fixture to correct the error parameters of the internal hardware pathway of the human body composition analyzer based on the comparison results.

8. The calibration method for the body composition analyzer according to claim 7, characterized in that, The standard bioimpedance simulation fixture includes multiple sets of parallel target RC modules, the component parameters of which are designed based on the fitting results of the Cole-Cole model of human body impedance.

9. The calibration method for the body composition analyzer according to claim 8, characterized in that, The standard bioimpedance simulation device specifically includes RC module combinations corresponding to different populations; correspondingly, the measured impedance spectrum of the standard bioimpedance simulation device obtained by measuring with a body composition analyzer includes: The target RC module selected in the standard bioimpedance simulation fixture is switched according to the target population corresponding to the tested object.

10. The calibration method for the body composition analyzer according to claim 9, characterized in that, The different groups are categorized based on one or more of the following: age, gender, BMI, health status, and exercise status.

Images