A waist circumference measuring instrument based on bioelectrical impedance and method thereof

Abstract

The present invention discloses a waist circumference measuring instrument based on bioelectrical impedance and a method thereof. First, a waist circumference measurement model based on a BP neural network is constructed. The input vectors of the input layer of the model include height, weight, age, and abdominal impedance values ​​at three different frequencies; the output vectors of the output layer include waist circumference; and the waist circumference measurement model based on the BP neural network is trained using a training sample set. Before the formal measurement, two precision resistors are used to calibrate the system parameters of the waist circumference measuring instrument based on bioelectrical impedance, namely the slope and offset; during the measurement, the abdominal voltage value is collected and sent to the trained waist circumference measurement model based on the BP neural network to obtain the waist circumference of the tester. The present invention is mainly used for waist circumference measurement. Compared with the measurement method of a transmission mechanical tape measure, the measurement method based on electrical impedance can eliminate the errors caused by different tension forces and different positions during the measurement of the mechanical tape measure.

Description

Technical Field

[0001] The present invention relates to the technical field of human health measurement, and in particular to a waist circumference measuring instrument based on bioelectrical impedance and a method thereof. Background Art

[0002] As people's living standards gradually improve, obesity has become a serious problem in modern society. People with abdominal obesity, in particular, are prone to vascular hardening, general fatigue, abdominal distension, indigestion, poor mental state, and poor sleep quality. They are also more likely to develop cardiovascular and cerebrovascular diseases such as hypertension, hyperlipidemia, diabetes, fatty liver, and coronary heart disease. In severe cases, they may develop myocardial infarction, cerebral infarction, stroke, hemiplegia, cirrhosis, and many other serious diseases. Therefore, specialized instruments and equipment are needed to test and analyze the composition of the human abdomen, so that people can understand their own physical fitness development. Summary of the Invention

[0003] The present invention provides a waist circumference measuring instrument based on bioelectrical impedance and a method thereof, which can accurately measure the waist circumference of a human body based on bioelectrical impedance.

[0004] To solve the above problems, the present invention is achieved through the following technical solutions:

[0005] A waist circumference measurement method based on bioelectrical impedance comprises the following steps:

[0006] Step 1: Construct a waist circumference measurement model based on a BP neural network. The waist circumference measurement model based on a BP neural network consists of an input layer, a hidden layer, and an output layer. The input vector of the input layer includes height, weight, age, and abdominal impedance values ​​at three different frequencies. The output vector of the output layer includes waist circumference.

[0007] Step 2: Collect a training sample set, which consists of sample data from different populations. Each sample data includes height, weight, age, abdominal impedance values ​​at three different frequencies, and waist circumference.

[0008] Step 3: inputting the training sample set obtained in step 2 into the waist measurement model based on the BP neural network constructed in step 1, training and learning the waist measurement model based on the BP neural network, and obtaining a trained waist measurement model based on the BP neural network;

[0009] Step 4: embedding the waist circumference measurement model based on the BP neural network trained in step 3 into the microcontroller of the waist circumference measurement instrument based on bioelectrical impedance;

[0010] Step 5. Before the formal measurement, the system parameters of the bioelectrical impedance-based waist circumference measurement instrument, namely the slope and offset, are calibrated using two precision resistors. Specifically, the excitation electrode and the measuring electrode of the bioelectrical impedance-based waist circumference measurement instrument are connected to the two ends of the precision resistor. The bioelectrical impedance-based waist circumference measurement instrument sends an excitation signal to the precision resistor through the excitation electrode, and the measuring electrode collects the voltage value fed back by the precision resistor. At this time, the microcontroller of the bioelectrical impedance-based waist circumference measurement instrument calculates the slope and offset of the system based on the returned voltage value.

[0011] ε=R1-K×ADC1;

[0012] Step 6: During measurement, enter the subject's height, weight, and age into the bioelectrical impedance waist measurement instrument. Attach the excitation electrodes, measurement electrodes, and temperature sensor of the bioelectrical impedance waist measurement instrument to the subject's abdomen. Ensure that the subject stands normally and does not touch any external conductive objects during the measurement.

[0013] Step 7: The bioelectrical impedance-based waist measurement instrument applies three excitation signals of different frequencies to the abdomen of the test subject through the excitation electrodes. The measuring electrodes collect the abdominal voltage values ​​at these three different frequencies and send them to the microcontroller.

[0014] Step 8: The microcontroller of the bioelectrical impedance-based waist measurement instrument calculates the abdominal impedance values ​​at three different frequencies based on the slope and offset of the system determined in step 5; wherein:

[0015] Z = K × ADC + ε;

[0016] Step 9: The microcontroller of the abdominal health comprehensive detection and analysis instrument inputs the obtained height, weight, age, and abdominal impedance values ​​at three different frequencies of the test subject into a waist circumference measurement model based on a BP neural network. The waist circumference measurement model based on a BP neural network predicts the waist circumference of the test subject.

[0017] The three different frequencies in the above steps 1, 2, 7, 8 and 9 correspond to each other and are all within the frequency range of 1 KHz to 1 MHz.

[0018] A waist circumference measuring instrument based on bioelectrical impedance consists of a constant current source generating circuit, a signal acquisition circuit, a signal amplifying circuit, a signal conditioning circuit, an amplitude and phase detection circuit, an AD data acquisition circuit, a microprocessor and a data processing platform; the control end of the microprocessor is connected to the constant current source generating circuit; the output end of the constant current source generating circuit, i.e., the excitation electrode, contacts the human abdomen, the input end of the signal acquisition circuit, i.e., the measuring electrode, contacts the human abdomen, and the output end of the constant current source generating circuit is a certain distance away from the input end of the signal acquisition circuit; the output end of the signal acquisition circuit is connected to the input end of the microprocessor after passing through the signal amplifying circuit, the signal conditioning circuit, the amplitude and phase detection circuit and the AD data acquisition circuit in sequence, and the output end of the microprocessor is connected to the data processing platform.

[0019] In the above solution, the output end of the constant current source generation circuit outputs three sinusoidal excitation signals within the frequency range of 1KHz to 1MHz.

[0020] In the above scheme, the constant current source generating circuit is composed of a DSS chip, an amplifying and filtering circuit and a voltage-controlled constant current source circuit; wherein the DSS chip is further composed of a phase accumulator, a waveform memory, a digital-to-analog converter and a low-pass filter; the control end of the microprocessor is connected to the input end of the phase accumulator, as well as the clock control end of the phase accumulator and the digital-to-analog converter; the output end of the phase accumulator is connected to the input end of the waveform memory, the output end of the waveform memory is connected to the input end of the digital-to-analog converter, and the output end of the digital-to-analog converter is connected to the input end of the low-pass filter; the output end of the low-pass filter is connected to the input end of the amplifying and filtering circuit, the output end of the amplifying and filtering circuit is connected to the input end of the voltage-controlled constant current source circuit, and the output end of the voltage-controlled constant current source circuit forms the output end of the constant current source generating circuit.

[0021] In the above scheme, the signal conditioning circuit consists of a notch filter and a bandpass filter; the input end of the notch filter forms the input end of the signal conditioning circuit, the output end of the notch filter is connected to the input end of the bandpass filter, and the output end of the bandpass filter forms the output end of the signal conditioning circuit.

[0022] The present invention is mainly used for waist circumference measurement. Compared with the measurement method of a transmission mechanical tape measure, the measurement method based on electrical impedance can eliminate the errors caused by different tensioning forces and different positions during mechanical tape measure measurement. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 This is a diagram of the single-cell impedance model.

[0024] Figure 2 for Figure 1 Simplified model diagram of .

[0025] Figure 3 This is the frequency characteristic diagram of biological tissue.

[0026] Figure 4 This is a system block diagram of the present invention.

[0027] Figure 5 Generates a flow chart for the signal.

[0028] Figure 6 This is the working principle diagram of DDS.

[0029] Figure 7 This is the voltage-controlled constant current circuit diagram.

[0030] Figure 8 This is the signal conditioning schematic.

[0031] Figure 9 This is the amplitude-frequency curve of the notch filter.

[0032] Figure 10 This is the amplitude-frequency response diagram of the bandpass filter.

[0033] Figure 11 Figure 1 is the schematic diagram of the amplitude and phase detection circuit, (a) is the amplitude detection circuit, and (b) is the phase detection circuit.

[0034] Figure 12 (a) is the output corresponding voltage relationship, (b) is the output corresponding voltage relationship of the phase detection output.

[0035] Figure 13 This is the frequency response diagram of the resistor-capacitor series circuit.

[0036] Figure 14 This is a low-pass filter diagram.

[0037] Figure 15 This is a data queue diagram.

[0038] Figure 16 It is a human waist training model.

[0039] Figure 17 This is a flow chart of the overall design of the present invention.

[0040] Figure 18 It is a four-electrode measurement method.

[0041] Figure 19 This is a data comparison chart. DETAILED DESCRIPTION

[0042] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to specific examples.

[0043] 1. Human body model:

[0044] Biological tissues are composed of cells, and the main components of cells are cell membranes and cell fluids. Different cells are separated by extracellular fluids and intercellular matrix. From an electrochemical perspective, the human body can be equivalent to an impedance model. Figure 1 shown.

[0045] When the frequency is less than 1 MHZ, the equivalent resistance of the cell membrane can be regarded as a short circuit, so Figure 1 The model is simplified as Figure 2 shown.

[0046] The simplified expression is as follows:

[0047]

[0048] Where w is the angular frequency, Cm is the cell membrane capacitance, Ri is the intracellular fluid resistance, and Re is the extracellular fluid resistance. The impedance modulus is:

[0049]

[0050] The phase angle is:

[0051]

[0052] The purpose of the present invention is to measure the above two parameters.

[0053] 2. Frequency Scattering Theory

[0054] The electrical properties of biological tissues show different regular characteristics in different frequency bands. Figure 3 The following is a diagram of the frequency characteristics of biological tissues. In the α frequency band, the electrical characteristics of biological tissues are mainly related to the characteristics of cell membranes. The β frequency band reflects the characteristics of cell fluids (including intracellular and extracellular fluids), and when it is greater than 1MHz, it enters the γ band, which reflects the status of water molecules. In clinical practice, most pathological conditions only occur in the α and β bands. In order to ensure that the excitation can penetrate the cell membrane, the selected frequency must be greater than 1KHz. Currently, most fat measuring meters use a single 50KHz frequency signal. Due to the single frequency and single input parameters, it cannot more accurately reflect the status of various tissues in the human body. Therefore, three frequencies of 1KHz to 1MHz are selected here.

[0055] 3. Measurement principle

[0056] like Figure 4As shown, a waist circumference measuring instrument based on bioelectrical impedance is composed of a constant current source generating circuit, a signal acquisition circuit, a signal amplifying circuit, a signal conditioning circuit, an amplitude and phase detection circuit, an AD data acquisition circuit, a microprocessor and a data processing platform; the control end of the microprocessor is connected to the constant current source generating circuit; the output end of the constant current source generating circuit contacts the human abdomen, the input end of the signal acquisition circuit contacts the human abdomen, and the output end of the constant current source generating circuit is separated from the input end of the signal acquisition circuit by a certain distance; the output end of the signal acquisition circuit is connected to the input end of the microprocessor after passing through the signal amplifying circuit, the signal conditioning circuit, the amplitude and phase detection circuit and the AD data acquisition circuit in sequence, and the output end of the microprocessor is connected to the data processing platform.

[0057] The microprocessor generates a constant current excitation source z through a control circuit. The excitation source applies a signal to the measured part of the body. This generates a corresponding electric field at the measured part of the body. The measured signal is conditioned, amplitude and phase detected, and transmitted to the ADC acquisition module, which then sends it to the microprocessor. Figure 4 The upper part of the module is the signal acquisition terminal. After the data is collected, it enters the microprocessor for data processing and calculates the corresponding body composition content. How to correctly collect data is the problem that the acquisition module should solve. The following describes the specific process from signal generation to acquisition.

[0058] 3.1 Generation of constant current source signal

[0059] The constant current source is generated by a constant current source generating circuit. The constant current source generating circuit includes a DSS chip, an amplifying filter circuit and a voltage-controlled constant current source circuit. Figure 5 The DSS chip is further composed of a phase accumulator, a waveform memory, a digital-to-analog converter, and a low-pass filter, as shown in Figure 6 As shown. The control end of the microprocessor is connected to the input end of the phase accumulator, as well as the clock control end of the phase accumulator and the digital-to-analog converter; the output end of the phase accumulator is connected to the input end of the waveform memory, the output end of the waveform memory is connected to the input end of the digital-to-analog converter, and the output end of the digital-to-analog converter is connected to the input end of the low-pass filter; the output end of the low-pass filter is connected to the input end of the amplifying and filtering circuit, the output end of the amplifying and filtering circuit is connected to the input end of the voltage-controlled constant current source circuit, and the output end of the voltage-controlled constant current source circuit forms the output end of the constant current source generating circuit. In this embodiment, the output end of the constant current source generating circuit outputs three sinusoidal excitation signals in the frequency range of 1 kHz to 1 MHz.

[0060] exist Figure 6In the figure, fclk is the reference clock of the DDS, which outputs a point after each clock cycle. The frequency control code is used to set the jump step size and control the frequency of the signal. The phase accumulator consists of an N-bit adder and an N-bit accumulation register. It accumulates the phase value according to the frequency control code and inputs the accumulated value into the waveform memory. The frequency output formula is: fout = fclk*k / 2^n, where k is the frequency control word. The resolution of the output frequency is: fclk / 2^n. The waveform memory uses the value of the phase accumulator as the current address, searches for the signal data corresponding to the phase value, and outputs it to the digital-to-analog converter. The digital-to-analog converter converts the digital value output by the waveform memory into the corresponding analog value. Due to the quantization error of the digital-to-analog converter, there is aliasing in the output waveform, which requires filtering with a low-pass filter at the output to improve the output performance of the signal.

[0061] The signal size output by DDS is not necessarily the amplitude we want. In addition, there may be bias voltage, so an amplifier and filter circuit must be added to the DDS post-stage to remove the bias and noise rate. At this point, the desired frequency and amplitude signal has been obtained, but how to ensure constant current output of the signal. Finally, a voltage-controlled constant current circuit is added to ensure the load capacity of the excitation signal. Figure 7 The figure shows a voltage-controlled constant current circuit.

[0062] The error of the constant current excitation circuit output under different load conditions according to this circuit design is shown in Table 1 below:

[0063] Table 1 Output signal error under different loads

[0064]

[0065] As can be seen from the table above, the signal output by the voltage-controlled constant current circuit has very little error under different load conditions. This indicates that the output signal of the voltage-controlled constant current circuit meets our requirements. Although the error is very small, showing no more than 2% in the table, it shows a trend of correlation with increasing impedance. This error can be corrected again in the program, further ensuring the accuracy of the impedance measurement.

[0066] 3.2 Signal Conditioning Circuit Design

[0067] The signal conditioning circuit consists of a notch filter and a bandpass filter; the input end of the notch filter forms the input end of the signal conditioning circuit, the output end of the notch filter is connected to the input end of the bandpass filter, and the output end of the bandpass filter forms the output end of the signal conditioning circuit.

[0068] In fact Figure 4The signal amplification, signal conditioning and amplitude and phase detection in the upper part of the figure can all be classified as signal conditioning. After specific acquisition, the signal conditioning flow chart is as follows Figure 8 As shown. The differential amplifier circuit converts the sampled differential two-ended signal into a single-ended signal. The purpose of the notch filter is to avoid the influence of the 50Hz mains power supply and mainly filters the 50Hz frequency. The amplitude-frequency response curve of the circuit is shown as follows: Figure 9 As shown in the figure, it can be seen that there is obvious attenuation around 50Hz frequency, which ensures that only the power frequency signal is filtered out without affecting signals of other frequencies. This is the main function of the notch filter. The bandpass filter can filter out high-frequency noise and some DC bias. Its amplitude-frequency curve is as follows Figure 10 As shown. For the low-frequency part, in addition to filtering out the DC component, it is also necessary to consider the subtle fluctuations of the signal caused by the pulsation of blood flow. Here, in order to obtain a more accurate and stable impedance value, it is necessary to remove the signal changes in this part. For the high-frequency part, the main part is still the noise part, and the main thing to filter out is external interference noise.

[0069] 3.3 Amplitude and Phase Detection

[0070] The amplitude and phase detection part mainly detects the amplitude and phase of the signal. It uses a logarithmic detector to perform amplitude and phase detection. It inputs two measurement signals through INPA and INPB, and outputs the decibel value of the ratio of the two signal powers and the phase between them through VMAG and VPHS. Figure 11 shown.

[0071] According to the logarithmic detection principle, the amplitude and phase calculation formulas are:

[0072] V MAG =(R F I SLP / 20)(P INA -P INB )+V CP

[0073] V PHS =-R F I Φ (|Φ(V INA )-Φ(V INB )|-90°)+V CP

[0074] In the above formula, PINA and PINB are the equivalent power in logarithmic units of VINA and VINB at the specified reference impedance. For the gain function, use R F I SLPThe slope is 600mv / decade, divided by 20dB / decade is 30mv / dB. With Vcp=900mV as the midpoint, the voltage value corresponding to -30dB to +30dB is 0-1.8V. The slope of the phase function represents R F I Φ Indicates 10mV / degree, with 900mV as the midpoint, corresponding to 90 degrees. 0-180 degrees corresponds to 1.8V-0V, and the voltage range corresponding to 0 to -180 degrees is the same, but the slope is opposite. Where φ is the phase degree corresponding to each signal. Figure 12 shown.

[0075] 4. Impedance Correction Process

[0076] The traditional method of calculating impedance based on voltage and current can cause the excitation signal applied to the impedance being measured to differ from the calculated value due to signal attenuation inherent in the circuit system. Table 1 also shows that the output current accuracy can vary slightly with different loads. Furthermore, since the voltage sampled by the ADC and the ADC value do not necessarily correspond to their zero crossings, there may be a certain offset. Therefore, using the traditional method of calculating impedance based on sampled voltage and known current can result in significant errors.

[0077] Since the relationship between impedance and voltage is linear, and the errors in Table 1 are also positively correlated, the present invention uses a method of direct calculation based on ADC values ​​to calculate the impedance. The calculation formula is:

[0078] Z=K×ADC+ε

[0079] Where K and ε are the slope and offset respectively.

[0080] Due to component production processes, even devices of the same model from the same manufacturer will have varying degrees of performance. To ensure the accuracy of impedance measurements, system parameters (K, ε) must be calibrated before formal measurement. This calibration method involves using two precision resistors as calibration resistors, measuring their voltages separately, and then performing calibration using the least squares method. The specific calibration method is as follows:

[0081] The voltages of the two resistors R1 and R2 are measured and sampled by ADC, and the values ​​obtained are ADC1 and ADC2 respectively. The slope relationship between the voltage and the resistance and the intercept can be obtained:

[0082] ε=R1-K×ADC1

[0083] The calibrated resistance measurement results can avoid resistance differences caused by differences in chip production processes.

[0084] After calibration, six standard through-hole resistors are selected and the resistance values ​​of the standard resistors are compared with the resistance values ​​measured by the digital multimeter. The error is no more than 2% within the specific calibration range.

[0085] 5. Digital Signal Processing

[0086] 5.1 Low-pass filtering

[0087] Following the above steps, the measured impedance value can be read out via a microprocessor. However, due to the complexity of the human body, direct readout can present some issues. Considering the human body's resistance as a system, the human body also has its own frequency response. Therefore, due to the body's inherent frequency response, the impedance value obtained at the moment of stimulus application may not directly reflect the body's true condition. Only when the body's frequency response reaches a steady state can the impedance value be meaningful.

[0088] like Figure 13 As shown, the impedance response curve obtained by measuring the resistor-capacitor series circuit with the above circuit is sent to the host computer via the serial port.

[0089] The red line is the ADC value measured by the circuit, the white line is the value after digital low-pass filtering, and the green line is the calculated impedance value. As can be seen from the figure, the values ​​read only after the system reaches stability are meaningful; otherwise, the values ​​read during the dynamic response period are not representative and cannot meet consistency requirements.

[0090] Now it seems that how to judge whether the response curve has entered a steady state is the key. Here we use a digital algorithm to implement the principle of a low-pass filter.

[0091] It is known that the transfer function of a first-order passive low-pass filter in an analog signal is:

[0092]

[0093] This is the transfer function in the S domain, which must be converted to a discrete signal before it can be applied to a computer. Converting a continuous signal to a discrete signal can be achieved through the Z transform, which gives:

[0094]

[0095] Where T is the sampling period. In this way, the above equation can be transformed into a differential equation:

[0096]

[0097] Among them, Y(n) is the output of this filter, X(n) is the current sampling value, and Y(n-1) is the output value of the previous filter. It is not difficult to see from the above formula that T / (T+RC) and RC / (T+RC) are complementary, so the digital filter algorithm can be simplified to:

[0098] Y(n)=a*X(n)+(1-a)*Y(n-1)

[0099] Where a is the filter coefficient. After adjusting this parameter to a suitable value, the following can be obtained: Figure 14 As can be seen from the figure, the dotted line is the result after digital low-pass filtering, which is easier to judge when the signal enters a steady state than the solid line without digital low-pass filtering.

[0100] 5.2 Steady-state determination

[0101] When entering the steady state, the sampled data remains basically unchanged. Based on this idea, a linear queue list is created through the data structure algorithm. It is judged that when the error approaches zero one by one, the data is in the steady state period. Figure 15 As shown, this is the process of data looping into the queue:

[0102] The formula for determining steady state can be defined by the minimum mean square error (LMS). For ease of calculation, the minimum absolute error (LAD) is used to define the steady state determination function. The conditions for determining steady state are:

[0103]

[0104] In the above formula, n is the number of data in the queue, ξ is the defined minimum steady-state error, and the accuracy of steady-state judgment can be adjusted by changing n and ξ.

[0105] 6. Model Training

[0106] 6.1 Human waist measurement method based on BP neural network

[0107] In the present invention, after measuring the human body impedance, a relationship between the impedance value and waist circumference is established through a machine learning algorithm based on weight, height, age, etc. The present invention uses a BP neural network as a training algorithm. Figure 16 The figure shows a waist circumference training model based on BP neural network, which consists of three parts: input layer, hidden layer and output layer.

[0108] exist Figure 16 In the middle, X1, X2, X3...X m is the input vector, which represents height, weight, age, and impedance values ​​at three different frequencies. Output Y is the output vector, which represents the output waist circumference index.

[0109] for Figure 16 The input-output relationship of the three-layer neural network shown is:

[0110]

[0111] In the formula, l is the number of neurons in the hidden layer, m is the number of neurons in the hidden layer, and w is the training weight. is the activation function and b is the bias.

[0112] d is the waist circumference learning sample measured by a ruler, that is, the output expected value. The present invention defines the cost function of the output layer as J based on the least mean square algorithm (LMS):

[0113]

[0114] In the formula, the w vector is the weight of the output layer to be trained. The purpose of training is to minimize the J value. According to the steepest descent method, the training algorithm of the output layer weight is described as:

[0115]

[0116] Where η is the learning rate.

[0117] For the hidden layer, there is no expected value for the corresponding neurons. Therefore, the error of the hidden layer cannot be calculated directly, but must be obtained by reverse recursion through the error signals of the neurons directly connected to the hidden neurons. The details will not be explained here.

[0118] 7. Implementation Method

[0119] 7.1 Overall Implementation Process

[0120] According to the above model and learning algorithm, after continuous iterative learning and finding the specific model weights, a model of the relationship between impedance and waist circumference can be established. By detecting the impedance value obtained, the human waist circumference can be calculated based on this model. The overall design flow chart is as follows Figure 17 shown.

[0121] 7.2 Measurement method

[0122] The four-electrode measurement method adopted by the present invention is as follows: Figure 18 As shown. When using the four-electrode test method, the position of the electrodes also has a certain impact on the measurement results. Figure 5 As shown, if the AC excitation electrode and the voltage measurement electrode are too close together, the current density is too high, leading to instability and drift. If the two voltage measurement electrodes are too close together, the difference in impedance between different body types will be too small. If the distance is too large, the instrument design will be too large. This article aims to minimize the instrument size while ensuring measurement accuracy.

[0123] The data obtained by measuring the electrode position is machine-learned to output the comparison results of the human body components with those measured by standard medical equipment, such as Figure 19 As shown. Figure 19 The data in the paper show that the measurement error is less than 10%.

[0124] Based on the above analysis, the present invention designs a waist circumference measurement method based on bioelectrical impedance, which includes the following steps:

[0125] Step 1: Construct a waist circumference measurement model based on a BP neural network. The waist circumference measurement model based on a BP neural network consists of an input layer, a hidden layer, and an output layer. The input vector of the input layer includes height, weight, age, and abdominal impedance values ​​at three different frequencies. The output vector of the output layer includes waist circumference.

[0126] Step 2: Collect a training sample set, which consists of sample data from different populations. Each sample data includes height, weight, age, abdominal impedance values ​​at three different frequencies, and waist circumference.

[0127] Step 3: inputting the training sample set obtained in step 2 into the waist measurement model based on the BP neural network constructed in step 1, training and learning the waist measurement model based on the BP neural network, and obtaining a trained waist measurement model based on the BP neural network;

[0128] Step 4: embedding the waist circumference measurement model based on the BP neural network trained in step 3 into the microcontroller of the waist circumference measurement instrument based on bioelectrical impedance;

[0129] Step 5. Before the formal measurement, the system parameters of the bioelectrical impedance-based waist circumference measurement instrument, namely the slope and offset, are calibrated using two precision resistors. Specifically, the excitation electrode and the measuring electrode of the bioelectrical impedance-based waist circumference measurement instrument are connected to the two ends of the precision resistor. The bioelectrical impedance-based waist circumference measurement instrument sends an excitation signal to the precision resistor through the excitation electrode, and the measuring electrode collects the voltage value fed back by the precision resistor. At this time, the microcontroller of the bioelectrical impedance-based waist circumference measurement instrument calculates the slope and offset of the system based on the returned voltage value.

[0130] ε=R1-K×ADC1;

[0131] Step 6: During measurement, enter the subject's height, weight, and age into the bioelectrical impedance waist measurement instrument. Attach the excitation electrodes, measurement electrodes, and temperature sensor of the bioelectrical impedance waist measurement instrument to the subject's abdomen. Ensure that the subject stands normally and does not touch any external conductive objects during the measurement.

[0132] Step 7: The bioelectrical impedance-based waist measurement instrument applies three excitation signals of different frequencies to the abdomen of the test subject through the excitation electrodes. The measuring electrodes collect the abdominal voltage values ​​at these three different frequencies and send them to the microcontroller.

[0133] Step 8: The microcontroller of the bioelectrical impedance-based waist measurement instrument calculates the abdominal impedance values ​​at three different frequencies based on the slope and offset of the system determined in step 5; wherein:

[0134] Z = K × ADC + ε;

[0135] Step 9: The microcontroller of the abdominal health comprehensive detection and analysis instrument inputs the obtained height, weight, age, and abdominal impedance values ​​at three different frequencies of the test subject into a waist circumference measurement model based on a BP neural network. The waist circumference measurement model based on a BP neural network predicts the waist circumference of the test subject.

[0136] The three different frequencies in the above steps 1, 2, 7, 8 and 9 correspond to each other and are all within the frequency range of 1 KHz to 1 MHz.

[0137] A waist circumference measuring instrument based on bioelectrical impedance for implementing the above method is composed of a constant current source generating circuit, a signal acquisition circuit, a signal amplification circuit, a signal conditioning circuit, an amplitude and phase detection circuit, an AD data acquisition circuit, a microprocessor and a data processing platform; the control end of the microprocessor is connected to the constant current source generating circuit; the output end of the constant current source generating circuit, i.e., the excitation electrode, contacts the human abdomen, the input end of the signal acquisition circuit, i.e., the measuring electrode, contacts the human abdomen, and the output end of the constant current source generating circuit is separated from the input end of the signal acquisition circuit by a certain distance; the output end of the signal acquisition circuit is connected to the input end of the microprocessor after passing through the signal amplification circuit, the signal conditioning circuit, the amplitude and phase detection circuit and the AD data acquisition circuit in sequence, and the output end of the microprocessor is connected to the data processing platform.

[0138] It should be noted that although the embodiments of the present invention described above are illustrative, they are not intended to limit the present invention. Therefore, the present invention is not limited to the above-mentioned specific embodiments. Without departing from the principles of the present invention, any other embodiments obtained by those skilled in the art under the guidance of the present invention are deemed to be within the protection of the present invention.

Claims

1. A waist circumference measuring instrument based on bioelectrical impedance, characterized in that: The device is composed of a constant current source generating circuit, a signal acquisition circuit, a signal amplifying circuit, a signal conditioning circuit, an amplitude and phase detection circuit, an AD data acquisition circuit, a microprocessor, and a data processing platform; the control end of the microprocessor is connected to the constant current source generating circuit; the output end of the constant current source generating circuit, i.e., the excitation electrode, contacts the abdomen of the human body, the input end of the signal acquisition circuit, i.e., the measuring electrode, contacts the abdomen of the human body, and the output end of the constant current source generating circuit is separated from the input end of the signal acquisition circuit by a certain distance; the output end of the signal acquisition circuit is connected to the input end of the microprocessor after passing through the signal amplifying circuit, the signal conditioning circuit, the amplitude and phase detection circuit, and the AD data acquisition circuit in sequence, and the output end of the microprocessor is connected to the data processing platform; The constant current source generating circuit is composed of a DSS chip, an amplifying and filtering circuit, and a voltage-controlled constant current source circuit; wherein the DSS chip is further composed of a phase accumulator, a waveform memory, a digital-to-analog converter, and a low-pass filter; the control end of the microprocessor is connected to the input end of the phase accumulator, as well as the clock control end of the phase accumulator and the digital-to-analog converter; the output end of the phase accumulator is connected to the input end of the waveform memory, the output end of the waveform memory is connected to the input end of the digital-to-analog converter, and the output end of the digital-to-analog converter is connected to the input end of the low-pass filter; The output end of the low-pass filter is connected to the input end of the amplifying filter circuit, the output end of the amplifying filter circuit is connected to the input end of the voltage-controlled constant current source circuit, and the output end of the voltage-controlled constant current source circuit forms the output end of the constant current source generating circuit; When the frequency response of the human body reaches a steady state, the microprocessor reads the measured impedance value. That is, through the data structure algorithm, a linear queue list is created to determine when the error approaches zero. The data is in a steady state period. The conditions for determining the steady state are: Where LAD is the minimum absolute error, n is the number of data in the queue, and ξ is the defined minimum steady-state error; The waist measurement method implemented by the waist measurement instrument is as follows: Step 1: Construct a waist circumference measurement model based on a BP neural network. The waist circumference measurement model based on a BP neural network consists of an input layer, a hidden layer, and an output layer. The input vector of the input layer includes height, weight, age, and abdominal impedance values ​​at three different frequencies. The output vector of the output layer includes waist circumference. Step 2: Collect a training sample set, which consists of sample data from different populations. Each sample data includes height, weight, age, abdominal impedance values ​​at three different frequencies, and waist circumference. Step 3: inputting the training sample set obtained in step 2 into the waist measurement model based on the BP neural network constructed in step 1, training and learning the waist measurement model based on the BP neural network, and obtaining a trained waist measurement model based on the BP neural network; Step 4: embedding the waist circumference measurement model based on the BP neural network trained in step 3 into the microcontroller of the waist circumference measurement instrument based on bioelectrical impedance; Step 5. Before the formal measurement, the system parameters of the bioelectrical impedance-based waist circumference measurement instrument, namely the slope and offset, are calibrated using two precision resistors. Specifically, the excitation electrode and the measuring electrode of the bioelectrical impedance-based waist circumference measurement instrument are connected to the two ends of the precision resistor. The bioelectrical impedance-based waist circumference measurement instrument sends an excitation signal to the precision resistor through the excitation electrode, and the measuring electrode collects the voltage value fed back by the precision resistor. At this time, the microcontroller of the bioelectrical impedance-based waist circumference measurement instrument calculates the slope and offset of the system based on the returned voltage value. Step 6: During measurement, enter the subject's height, weight, and age into the bioelectrical impedance waist measurement instrument. Attach the excitation electrodes, measurement electrodes, and temperature sensor of the bioelectrical impedance waist measurement instrument to the subject's abdomen. Ensure that the subject stands normally and does not touch any external conductive objects during the measurement. Step 7: The bioelectrical impedance-based waist measurement instrument applies three excitation signals of different frequencies to the abdomen of the test subject through the excitation electrodes. The measuring electrodes collect the abdominal voltage values ​​at these three different frequencies and send them to the microcontroller. Step 8: The microcontroller of the bioelectrical impedance-based waist measurement instrument calculates the abdominal impedance values ​​at three different frequencies based on the slope and offset of the system determined in step 5; wherein: Z = K × ADC + ε; Step 9: The microcontroller of the abdominal health comprehensive detection and analysis instrument inputs the obtained height, weight, age, and abdominal impedance values ​​at three different frequencies of the test subject into a waist circumference measurement model based on a BP neural network. The waist circumference measurement model based on a BP neural network predicts the waist circumference of the test subject.

2. The waist circumference measuring instrument based on bioelectrical impedance according to claim 1, characterized in that: The output end of the constant current source generation circuit outputs three sinusoidal excitation signals in the frequency range of 1KHz to 1MHz.

3. The waist circumference measuring instrument based on bioelectrical impedance according to claim 1, characterized in that: The signal conditioning circuit consists of a notch filter and a bandpass filter; the input end of the notch filter forms the input end of the signal conditioning circuit, the output end of the notch filter is connected to the input end of the bandpass filter, and the output end of the bandpass filter forms the output end of the signal conditioning circuit.

Images