Children trunk fat percentage prediction model

Abstract

The invention discloses a children trunk fat percentage prediction model, which relates to the technical field of fat measurement and comprises a data acquisition module, a data processing module, a terminal module and a power supply module. Performing infrared measurement on subcutaneous fat of a tested child to obtain infrared electric signal data, performing ultrasonic measurement on visceral fat of the tested child to obtain ultrasonic electric signal data, judging data acquisition quality of the near-infrared acquisition unit and data acquisition quality of the ultrasonic acquisition unit, and judging whether re-measurement is needed or not; and calculating the infrared electric signal data to obtain the subcutaneous fat content of the tested child, calculating the ultrasonic electric signal data to obtain the visceral fat content of the tested child, and finally calculating the child trunk fat percentage prediction value by combining the tested child information. The accuracy of data acquisition is improved through multi-modal acquisition and an acquisition quality judgment mechanism, and the accuracy of model prediction is improved by extracting the deep characteristics of the trunk fat content through data processing.

Description

Technical Field

[0001] The invention relates to the technical field of fat measurement, and in particular to a prediction model for trunk fat percentage of children. Background Art

[0002] In recent years, with the continuous improvement of my country's material living conditions, the problem of childhood obesity has become increasingly prominent. On the one hand, high-calorie snacks and beverages are popular among children, both in cities and towns. On the other hand, due to the rise of electronic products and short video applications, children's exercise volume has declined, resulting in a trend of younger and more universal childhood obesity. Childhood is a critical stage of physical development, and is closely related to the occurrence of various chronic diseases such as cardiovascular disease and diabetes in adulthood. Expressing trunk fat content in percentage can make the measurement results more standardized and intuitive. Children of different ages, genders and body shapes can assess their obesity by comparing trunk fat percentages to avoid inaccurate assessments due to differences in weight, height and other factors.

[0003] At present, the Chinese invention patent application with application number CN202110084841.5 discloses a visceral fat measurement method, device, equipment and storage medium, which includes: measuring the height of the child being tested; measuring the weight of the subject and the weight of the limbs and trunk; measuring the skin fat thickness of the limbs and trunk; measuring the lean body impedance of the left arm, the lean body impedance of the right arm, the lean body impedance of the left leg, the lean body impedance of the right leg and the lean body impedance of the trunk; calculating the lean body weight of the left arm, the lean body of the right arm, the lean body of the left leg, the lean body of the right leg and the lean body of the trunk; calculating the subcutaneous fat weight of the left arm, the right arm, the left leg, the right leg and the trunk; calculating the visceral fat weight, and finally detecting the visceral fat amount conveniently and quickly. However, when the application quantifies the trunk fat of children, the weight of the limbs and trunk is used as a fixed proportion of the body weight, which does not fully consider the problem of large differences in physique among different children. Summary of the invention

[0004] The technical problem solved by the present invention is that when the traditional method performs quantitative calculation on the trunk fat of children, the data collected is limited in scope, the professional requirements on the user are high, and the problem of great differences in physique among different children is not fully considered.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] A child trunk fat percentage prediction model includes: a data acquisition module, a data processing module and a terminal module;

[0007] The data acquisition module includes a near-infrared acquisition unit, an ultrasonic acquisition unit and an acquisition judgment unit, which are used to perform infrared measurement on the subcutaneous fat of the test child and ultrasonic measurement on the visceral fat of the test child, to obtain infrared electrical signal data by emitting near-infrared light of a preset wavelength combination to the test child at each infrared measurement point by an infrared light emitting diode probe, to obtain ultrasonic electrical signal data by emitting preset low-frequency output ultrasonic waves at each ultrasonic point by an ultrasonic probe, to obtain first judgment results on the data acquisition quality of the near-infrared acquisition unit at each infrared measurement point, to obtain second judgment results on the data acquisition quality of the ultrasonic acquisition unit at each ultrasonic measurement point, and to calculate and determine whether re-measurement is required based on the first judgment results and the second judgment results;

[0008] The data processing module includes an infrared calculation unit, an ultrasonic calculation unit and a comprehensive prediction unit, which are used to calculate the infrared electrical signal data to obtain the subcutaneous fat content of the tested child, calculate the ultrasonic electrical signal data to obtain the visceral fat content of the tested child, and calculate the predicted value of the child's trunk fat percentage based on the age, gender, predicted subcutaneous fat thickness and predicted visceral fat content of the tested child;

[0009] The terminal module includes an interaction unit and a display unit, which are used to receive operation instructions input by the user and issue sound and light reminders, and provide the user with an operation interface and visual results.

[0010] As a preferred solution of the child trunk fat percentage prediction model described in the present invention, the near-infrared acquisition unit is used to perform infrared measurement on the subcutaneous fat of the test child to obtain infrared electrical signal data, and the logic of the infrared measurement includes: the infrared light emitting diode probe in the near-infrared acquisition unit sequentially emits near-infrared light of a preset wavelength combination to the test child at each infrared measurement point position, the near-infrared light of the preset wavelength combination irradiates the test child to generate near-infrared back-reflected light, and the photoelectric sensor in the near-infrared acquisition unit receives and converts the near-infrared back-reflected light to obtain infrared electrical signal data;

[0011] The infrared light emitting diode probe comprises K infrared light emitting diodes, each of which is started in sequence and emits near-infrared light corresponding to a preset wavelength for T milliseconds;

[0012] The infrared measurement point positions include a merged point position and an infrared additional point position. The merged point position includes 2 cm directly below the navel of the tested child, the left waist side, the right waist side, the left costal arch lower edge and the right costal arch lower edge. The infrared additional point positions include the left inferior scapular angle area, the left inferior scapular angle area, the left infraclavicular fossa and the right infraclavicular fossa.

[0013] As a preferred embodiment of the child trunk fat percentage prediction model described in the present invention, the ultrasonic acquisition unit is used to perform ultrasonic measurement on the visceral fat of the test child to obtain ultrasonic electrical signal data, and the logic of the ultrasonic measurement includes: transmitting a preset low-frequency output ultrasonic wave to the test child at each ultrasonic measurement point in turn through an ultrasonic probe, and receiving an ultrasonic echo signal through a piezoelectric crystal in the ultrasonic probe to obtain ultrasonic electrical signal data;

[0014] The ultrasonic measurement point positions include a merged point position and an ultrasonic additional point position. The merged point position includes A cm below the navel of the tested child, the left waist side, the right waist side, the lower edge of the left costal arch and the lower edge of the right costal arch. The ultrasonic additional point positions include the xiphoid process, the posterior left kidney, the posterior right kidney and B cm above the navel of the tested child.

[0015] As a preferred solution of the child trunk fat percentage prediction model described in the present invention, the acquisition judgment unit is used to perform a first judgment on the data acquisition quality of the near-infrared acquisition unit at each infrared measurement point to obtain a first judgment result, and perform a second judgment on the data acquisition quality of the ultrasonic acquisition unit at each ultrasonic measurement point to obtain a second judgment result;

[0016] The first judgment result includes a successful acquisition and an unsuccessful acquisition, and the second judgment result includes a successful acquisition and an unsuccessful acquisition;

[0017] When the first judgment result is that the acquisition is successful, the measurement at the corresponding infrared measurement point position is stopped; when the first judgment result is that the acquisition is unsuccessful, the infrared acquisition unit needs to re-measure at the corresponding infrared measurement point position;

[0018] When the second judgment result is that the acquisition is successful, the measurement of the corresponding ultrasonic measurement point position is stopped, and when the first judgment result is that the acquisition is unsuccessful, the ultrasonic acquisition unit needs to re-measure at the corresponding ultrasonic measurement point position.

[0019] As a preferred solution of the child trunk fat percentage prediction model described in the present invention, the logic of the one-time judgment includes: when the infrared acquisition unit measures the subcutaneous fat of the test child at each infrared measurement point position in turn, the pressure data is obtained by the pressure sensor on the infrared light emitting diode probe, and the minimum pressure and the average pressure are calculated based on the pressure data. If and only if the minimum pressure is greater than the first pressure threshold and the average pressure is greater than the second pressure threshold, it is judged that the acquisition is successful, otherwise it is judged that the acquisition is unsuccessful;

[0020] The logic of the secondary judgment includes: when the ultrasonic acquisition unit measures the visceral fat of the tested child at each ultrasonic measurement point in turn, the ultrasonic electrical signal data is obtained through the piezoelectric crystal in the ultrasonic acquisition unit and the received echo time is recorded, the ultrasonic electrical signal data obtains the maximum echo intensity and the minimum echo intensity, and the ultrasonic time difference is calculated based on the ultrasonic emission time and the received echo time. The acquisition is judged to be successful if and only if the ultrasonic time difference is greater than the time difference threshold, the maximum echo intensity is less than the first echo intensity threshold, and the minimum echo intensity is greater than the second echo intensity threshold; otherwise, the acquisition is judged to be unsuccessful.

[0021] As a preferred embodiment of the child trunk fat percentage prediction model described in the present invention, the infrared calculation unit is used to calculate the infrared electrical signal data to obtain the subcutaneous fat content of the tested child, the infrared light signal is denoised by a bandpass filter to obtain a smooth infrared signal, the Nyquist frequency is calculated based on the smooth infrared signal, the smooth infrared signal is discretized with N times the Nyquist frequency as the sampling frequency to obtain an infrared light discrete sequence, the infrared light discrete sequence is vectorized to obtain an infrared light discrete vector, the infrared light discrete vectors are sorted in order from low to high according to the wavelengths corresponding to the infrared light discrete vectors and vector splicing is performed to obtain an infrared light discrete matrix, a prediction regression equation group of the infrared light discrete matrix and the subcutaneous fat predicted thickness is established by partial least squares method, the ridge regression equation coefficient and the least squares residual square are calculated based on the preset regularization parameter and the prediction regression equation group, the regression parameter is calculated based on the ridge regression equation coefficient, the least squares residual square and the ridge regression polynomial, the score matrix is ​​calculated by the regression parameter, the prediction regression equation group and the preset regularization parameter, and the subcutaneous fat predicted thickness is calculated based on the score matrix and the prediction regression equation group.

[0022] As a preferred embodiment of the child trunk fat percentage prediction model described in the present invention, the ultrasonic calculation unit is used to calculate the ultrasonic electrical signal data to obtain the visceral fat content of the tested child, perform high-frequency noise removal processing on the ultrasonic electrical signal data through Butterworth low-pass filtering to obtain a denoised ultrasonic signal, calculate the ultrasonic attenuation coefficient of the preset low-frequency output ultrasonic wave and the denoised ultrasonic signal through the cross-correlation method, establish an ultrasonic attenuation grayscale image based on the ultrasonic acoustic attenuation coefficient by grayscale level mapping, perform adaptive histogram equalization and bilinear interpolation processing on the ultrasonic attenuation grayscale image to obtain an enhanced ultrasonic image, perform texture feature extraction on the enhanced ultrasonic image based on a preset Markov random method to obtain a feature ultrasonic image, obtain a feature prediction image through image corrosion processing and preset binary segmentation, and obtain the predicted visceral fat content based on the ratio of grayscale pixels in the feature prediction image to all pixels.

[0023] As a preferred embodiment of the child trunk fat percentage prediction model described in the present invention, the comprehensive prediction unit is used to calculate the predicted value of the child trunk fat percentage, calculate the ternary body mass index based on the height and weight of the tested child, calculate the trunk coefficient based on the ternary body mass index, and calculate the predicted value of the child trunk fat percentage based on the weight, height, trunk coefficient, predicted subcutaneous fat thickness and predicted visceral fat content of the tested child. The calculation expression is:

[0024] TBMI=m / h 3

[0025] λ=TBMI / (1+TBMI)

[0026] PRV=αPFC+(1 / mλ)πρβh[δ(ATF) 2 +2ATFe ελ ]

[0027] Among them, TBMI represents ternary body mass index, PRV represents the predicted value of trunk fat percentage of children, PFC represents the predicted content of visceral fat, α represents the predicted content coefficient of visceral fat, m represents the weight of the tested child, λ represents the trunk coefficient, π represents pi, ρ represents fat density, β represents the predicted content coefficient of subcutaneous fat, h represents the height of the tested child, ATF represents the predicted thickness of subcutaneous fat, δ represents the predicted thickness coefficient of subcutaneous fat, e represents the natural constant, and ε represents the index parameter.

[0028] As a preferred solution of a child trunk fat percentage prediction model described in the present invention, wherein: the interactive unit is used to receive operation instructions and issue sound and light reminders, the operation instructions include a start detection instruction, a stop detection instruction, a test child information input instruction, a prediction result query instruction, a history record data query instruction, a first pressure threshold setting, a second pressure threshold setting, a time difference threshold setting, a first echo intensity threshold setting, and a second echo intensity threshold setting, and the sound and light reminder includes an operation instruction receiving prompt sound, a successful acquisition sound and light reminder, and an unsuccessful acquisition sound and light reminder;

[0029] The display unit is used to provide the user with an operation interface and visualization results, including operation instruction display, data collection result display, prediction result display and historical record data display.

[0030] The beneficial effects of the present invention are as follows: the subcutaneous fat and visceral fat of children are measured respectively by combining a near-infrared acquisition unit and an ultrasonic acquisition unit, which is conducive to multi-modal data acquisition and avoids the limitations of a single measurement method. Near-infrared measurement and ultrasonic measurement are performed at multiple points at the same time, and fat data is obtained from different angles and parts, thereby improving the comprehensiveness and accuracy of data acquisition. The acquisition judgment mechanism is introduced through the acquisition judgment unit, and when it is determined that the acquisition is unsuccessful, an audible and visual reminder is issued to the user to make the user re-measure, thereby ensuring the validity of the collected data and reducing the requirements for professional operation. Even if the user does not have a professional medical operation level, the data acquisition can be completed with high quality, thereby improving the applicability of the model. The predicted thickness of subcutaneous fat is calculated by combining partial least squares method and ridge regression, and the ultrasonic electrical signal data is processed and converted into an image for calculation, thereby facilitating the extraction of deep features of fat content from the collected data, abstracting the specific and complex fat content problem, and introducing the ternary body mass index to replace the BMI in the traditional method for calculation, thereby improving the applicability and accuracy of the prediction of the percentage of trunk fat in children. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 A schematic diagram of the basic framework of a child trunk fat percentage prediction model provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0032] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0033] Example 1, reference Figure 1 , as an embodiment of the present invention, provides a child trunk fat percentage prediction model, including: a data acquisition module, a data processing module, a terminal module and a power supply module;

[0034] The data acquisition module includes a near-infrared acquisition unit, an ultrasonic acquisition unit and an acquisition judgment unit, which are used to perform infrared measurement on the subcutaneous fat of the test child and ultrasonic measurement on the visceral fat of the test child, to obtain infrared electrical signal data by emitting near-infrared light of a preset wavelength combination to the test child at each infrared measurement point by an infrared light emitting diode probe, to obtain ultrasonic electrical signal data by emitting preset low-frequency output ultrasonic waves at each ultrasonic point by an ultrasonic probe, to obtain first judgment results on the data acquisition quality of the near-infrared acquisition unit at each infrared measurement point, to obtain second judgment results on the data acquisition quality of the ultrasonic acquisition unit at each ultrasonic measurement point, and to calculate and determine whether re-measurement is required based on the first judgment results and the second judgment results;

[0035] The data processing module includes an infrared calculation unit, an ultrasonic calculation unit and a comprehensive prediction unit, which are used to calculate the infrared electrical signal data to obtain the subcutaneous fat content of the tested child, calculate the ultrasonic electrical signal data to obtain the visceral fat content of the tested child, and calculate the predicted value of the child's trunk fat percentage based on the age, gender, predicted subcutaneous fat thickness and predicted visceral fat content of the tested child;

[0036] The terminal module includes an interaction unit and a display unit, which are used to receive operation instructions input by the user and issue sound and light reminders, and provide the user with an operation interface and visual results;

[0037] The power supply module provides electric energy for the data acquisition module, the data processing module and the terminal module.

[0038] In this embodiment, the near-infrared acquisition unit is used to perform infrared measurement on the subcutaneous fat of the test child to obtain infrared electrical signal data. The logic of the infrared measurement includes: emitting near-infrared light of a preset wavelength combination to the test child at each infrared measurement point in turn through the infrared light emitting diode probe in the near-infrared acquisition unit, the near-infrared light of the preset wavelength combination irradiates the test child to generate near-infrared back-reflected light, and the photoelectric sensor in the near-infrared acquisition unit receives and converts the near-infrared back-reflected light to obtain infrared electrical signal data;

[0039] The infrared light emitting diode probe comprises K infrared light emitting diodes, each of which is started in sequence and emits near-infrared light corresponding to a preset wavelength for T milliseconds;

[0040] The infrared measurement point positions include a merged point position and an infrared additional point position. The merged point position includes 2 cm directly below the navel of the tested child, the left waist side, the right waist side, the left costal arch lower edge and the right costal arch lower edge. The infrared additional point positions include the left inferior scapular angle area, the left inferior scapular angle area, the left infraclavicular fossa and the right infraclavicular fossa.

[0041] The infrared light emitting diode probe emits near-infrared light with wavelengths of 725nm, 750nm, 785nm, 820nm, 855nm, 890nm, 925nm, 1100nm and 1300nm. When collecting data, a collection judgment unit is added to determine whether the user has successfully collected valid data, which reduces the requirements for professional operation when using infrared light probes and ultrasonic probes and improves applicability. At the same time, by collecting data at multiple points, the error tolerance range of the measurement position when collecting data on the child's torso is improved.

[0042] In this embodiment, the ultrasonic acquisition unit is used to perform ultrasonic measurement on the visceral fat of the test child to obtain ultrasonic electrical signal data. The logic of the ultrasonic measurement includes: transmitting a preset low-frequency output ultrasonic wave to the test child at each ultrasonic measurement point in turn through an ultrasonic probe, and receiving an ultrasonic echo signal through a piezoelectric crystal in the ultrasonic probe to obtain ultrasonic electrical signal data;

[0043] The ultrasonic measurement point positions include a merged point position and an ultrasonic additional point position. The merged point position includes A cm below the navel of the tested child, the left waist side, the right waist side, the lower edge of the left costal arch and the lower edge of the right costal arch. The ultrasonic additional point positions include the xiphoid process, the posterior left kidney, the posterior right kidney and B cm above the navel of the tested child.

[0044] In this embodiment, the acquisition judgment unit is used to perform a first judgment on the data acquisition quality of the near-infrared acquisition unit at each infrared measurement point to obtain a first judgment result, and perform a second judgment on the data acquisition quality of the ultrasonic acquisition unit at each ultrasonic measurement point to obtain a second judgment result;

[0045] The first judgment result includes a successful acquisition and an unsuccessful acquisition, and the second judgment result includes a successful acquisition and an unsuccessful acquisition;

[0046] When the first judgment result is that the acquisition is successful, the measurement at the corresponding infrared measurement point position is stopped; when the first judgment result is that the acquisition is unsuccessful, the infrared acquisition unit needs to re-measure at the corresponding infrared measurement point position;

[0047] When the second judgment result is that the acquisition is successful, the measurement of the corresponding ultrasonic measurement point position is stopped, and when the first judgment result is that the acquisition is unsuccessful, the ultrasonic acquisition unit needs to re-measure at the corresponding ultrasonic measurement point position.

[0048] In this embodiment, the logic of the one-time judgment includes: when the infrared acquisition unit measures the subcutaneous fat of the test child at each infrared measurement point position in turn, the pressure data is obtained by the pressure sensor on the infrared light emitting diode probe, and the minimum pressure and the average pressure are calculated based on the pressure data. If and only if the minimum pressure is greater than the first pressure threshold and the average pressure is greater than the second pressure threshold, it is judged that the acquisition is successful, otherwise it is judged that the acquisition is unsuccessful;

[0049] The logic of the secondary judgment includes: when the ultrasonic acquisition unit measures the visceral fat of the tested child at each ultrasonic measurement point in turn, the ultrasonic electrical signal data is obtained through the piezoelectric crystal in the ultrasonic acquisition unit and the received echo time is recorded, the ultrasonic electrical signal data obtains the maximum echo intensity and the minimum echo intensity, and the ultrasonic time difference is calculated based on the ultrasonic emission time and the received echo time. The acquisition is judged to be successful if and only if the ultrasonic time difference is greater than the time difference threshold, the maximum echo intensity is less than the first echo intensity threshold, and the minimum echo intensity is greater than the second echo intensity threshold; otherwise, the acquisition is judged to be unsuccessful.

[0050] Among them, the pressure sensor is an infrared light emitting diode probe and four pressure sensors are evenly distributed on the contact surface of the test child. Due to the limited penetration of near-infrared light, the probe of the infrared acquisition unit is required to be close to the skin when collecting at the measuring point. The user is given sound and light prompts through the first judgment, which is conducive to quantitative evaluation of the user's collection quality and improves the accuracy of the collected data.

[0051] In this embodiment, the infrared calculation unit is used to calculate the infrared electrical signal data to obtain the subcutaneous fat content of the tested child, perform denoising processing on the infrared light signal through a bandpass filter to obtain a smooth infrared signal, calculate the Nyquist frequency based on the smooth infrared signal, discretize the smooth infrared signal with N times the Nyquist frequency as the sampling frequency to obtain an infrared light discrete sequence, vectorize the infrared light discrete sequence to obtain an infrared light discrete vector, sort the infrared light discrete vectors in order from low to high according to the wavelengths corresponding to them and perform vector splicing processing to obtain an infrared light discrete matrix, establish a prediction regression equation group for the infrared light discrete matrix and the predicted thickness of subcutaneous fat through the partial least squares method, calculate the ridge regression equation coefficient and the least squares residual square based on the preset regularization parameter and the prediction regression equation group, calculate the regression parameter based on the ridge regression equation coefficient, the least squares residual square and the ridge regression polynomial, calculate the score matrix through the regression parameter, the prediction regression equation group and the preset regularization parameter, and calculate the predicted thickness of subcutaneous fat based on the score matrix and the prediction regression equation group.

[0052] In this embodiment, the ultrasonic calculation unit is used to calculate the ultrasonic electrical signal data to obtain the visceral fat content of the tested child, perform high-frequency noise removal processing on the ultrasonic electrical signal data through Butterworth low-pass filtering to obtain a denoised ultrasonic signal, calculate the ultrasonic attenuation coefficient of the preset low-frequency output ultrasonic wave and the denoised ultrasonic signal through the cross-correlation method, establish an ultrasonic attenuation grayscale image based on the ultrasonic acoustic attenuation coefficient by grayscale level mapping, perform adaptive histogram equalization and bilinear interpolation processing on the ultrasonic attenuation grayscale image to obtain an enhanced ultrasonic image, perform texture feature extraction on the enhanced ultrasonic image based on a preset Markov random method to obtain a feature ultrasonic image, obtain a feature prediction image through image corrosion processing and preset binary segmentation, and obtain the predicted visceral fat content based on the ratio of grayscale pixels in the feature prediction image to all pixels.

[0053] In this embodiment, the comprehensive prediction unit is used to calculate the predicted value of the child's trunk fat percentage, calculate the ternary body mass index based on the height and weight of the tested child, calculate the trunk coefficient based on the ternary body mass index, and calculate the predicted value of the child's trunk fat percentage based on the weight, height, trunk coefficient, predicted subcutaneous fat thickness and predicted visceral fat content of the tested child. The calculation expression is:

[0054] TBMI=m / h 3

[0055] λ=TBMI / (1+TBMI)

[0056] PRV=αPFC+(1 / mλ)πρβh[δ(ATF) 2 +2ATFe ελ ]

[0057] Among them, TBMI represents ternary body mass index, PRV represents the predicted value of trunk fat percentage of children, PFC represents the predicted content of visceral fat, α represents the predicted content coefficient of visceral fat, m represents the weight of the tested child, λ represents the trunk coefficient, π represents pi, ρ represents fat density, β represents the predicted content coefficient of subcutaneous fat, h represents the height of the tested child, ATF represents the predicted thickness of subcutaneous fat, δ represents the predicted thickness coefficient of subcutaneous fat, e represents the natural constant, and ε represents the index parameter.

[0058] In this embodiment, the interactive unit is used to receive operation instructions and issue sound and light reminders, the operation instructions include start detection instructions, stop detection instructions, test child information input instructions, prediction result query instructions, historical record data query instructions, first pressure threshold setting, second pressure threshold setting, time difference threshold setting, first echo intensity threshold setting, and second echo intensity threshold setting, and the sound and light reminders include operation instruction reception prompt sound, acquisition success sound and light reminder, and acquisition failure sound and light reminder;

[0059] The display unit is used to provide the user with an operation interface and visualization results, including operation instruction display, data collection result display, prediction result display and historical record data display.

[0060] Among them, the scheme in this embodiment adopts a combination of a near-infrared acquisition unit and an ultrasonic acquisition unit to measure the subcutaneous fat and visceral fat of children respectively, which is conducive to multi-modal data acquisition and avoids the limitations of a single measurement method. Near-infrared measurement and ultrasonic measurement are performed at multiple points at the same time to obtain fat data from different angles and parts, thereby improving the comprehensiveness and accuracy of data acquisition. The acquisition judgment mechanism is introduced through the acquisition judgment unit. When it is determined that the acquisition is unsuccessful, an audible and visual reminder is issued to the user to make the user re-measure, thereby ensuring the validity of the collected data and reducing the requirements for professional operation. Even if the user does not have a professional medical operation level, the data acquisition can be completed with high quality, thereby improving the applicability of the model. The predicted thickness of subcutaneous fat is calculated by combining partial least squares and ridge regression, and the ultrasonic electrical signal data is processed and converted into an image for calculation, which is conducive to extracting deep features of fat content from the collected data, abstracting the specific and complex fat content problem, and improving the applicability and accuracy of predicting the percentage of trunk fat in children.

[0061] It should be understood by those skilled in the art that embodiments of the present invention may provide methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program codes. Among them, the storage medium may be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic memory, flash memory, disk or optical disk. These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0062] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A model for predicting trunk fat percentage in children, characterized in that: include: Data acquisition module, data processing module, terminal module and power supply module; The data acquisition module includes a near-infrared acquisition unit, an ultrasonic acquisition unit and an acquisition judgment unit, which are used to perform infrared measurement on the subcutaneous fat of the test child and ultrasonic measurement on the visceral fat of the test child, to obtain infrared electrical signal data by emitting near-infrared light of a preset wavelength combination to the test child at each infrared measurement point by an infrared light emitting diode probe, to obtain ultrasonic electrical signal data by emitting preset low-frequency output ultrasonic waves at each ultrasonic point by an ultrasonic probe, to obtain first judgment results on the data acquisition quality of the near-infrared acquisition unit at each infrared measurement point, to obtain second judgment results on the data acquisition quality of the ultrasonic acquisition unit at each ultrasonic measurement point, and to calculate and determine whether re-measurement is required based on the first judgment results and the second judgment results; The data processing module includes an infrared calculation unit, an ultrasonic calculation unit and a comprehensive prediction unit, which are used to calculate the infrared electrical signal data to obtain the subcutaneous fat content of the tested child, calculate the ultrasonic electrical signal data to obtain the visceral fat content of the tested child, and calculate the predicted value of the child's trunk fat percentage based on the age, gender, height, weight, predicted subcutaneous fat thickness and predicted visceral fat content of the tested child; The terminal module includes an interaction unit and a display unit, which are used to receive operation instructions input by the user and issue sound and light reminders, and provide the user with an operation interface and visual results; The power supply module provides electric energy for the data acquisition module, the data processing module and the terminal module.

2. A child trunk fat percentage prediction model as claimed in claim 1, characterized in that: The near-infrared acquisition unit is used to perform infrared measurement on the subcutaneous fat of the test child to obtain infrared electrical signal data. The logic of the infrared measurement includes: emitting near-infrared light of a preset wavelength combination to the test child at each infrared measurement point in turn through the infrared light emitting diode probe in the near-infrared acquisition unit, the near-infrared light of the preset wavelength combination irradiating the test child to generate near-infrared back-reflected light, and receiving and converting the near-infrared back-reflected light through the photoelectric sensor in the near-infrared acquisition unit to obtain infrared electrical signal data; The infrared light emitting diode probe comprises K infrared light emitting diodes, each of which is started in sequence and emits near-infrared light corresponding to a preset wavelength for T milliseconds; The infrared measurement point positions include a merged point position and an infrared additional point position. The merged point position includes 2 cm directly below the navel of the tested child, the left waist side, the right waist side, the left costal arch lower edge and the right costal arch lower edge. The infrared additional point positions include the left inferior scapular angle area, the left inferior scapular angle area, the left infraclavicular fossa and the right infraclavicular fossa.

3. A child trunk fat percentage prediction model as claimed in claim 1, characterized in that: The ultrasonic acquisition unit is used to perform ultrasonic measurement on the visceral fat of the test child to obtain ultrasonic electrical signal data. The logic of the ultrasonic measurement includes: transmitting a preset low-frequency output ultrasonic wave to the test child at each ultrasonic measurement point in turn through an ultrasonic probe, and receiving an ultrasonic echo signal through a piezoelectric crystal in the ultrasonic probe to obtain ultrasonic electrical signal data; The ultrasonic measurement point positions include a merged point position and an ultrasonic additional point position. The merged point position includes A cm below the navel of the tested child, the left waist side, the right waist side, the lower edge of the left costal arch and the lower edge of the right costal arch. The ultrasonic additional point positions include the xiphoid process, the posterior left kidney, the posterior right kidney and B cm above the navel of the tested child.

4. A child trunk fat percentage prediction model as claimed in claim 1, characterized in that: The acquisition judgment unit is used to make a first judgment on the data acquisition quality of the near-infrared acquisition unit at each infrared measurement point to obtain a first judgment result, and to make a second judgment on the data acquisition quality of the ultrasonic acquisition unit at each ultrasonic measurement point to obtain a second judgment result; The first judgment result includes a successful acquisition and an unsuccessful acquisition, and the second judgment result includes a successful acquisition and an unsuccessful acquisition; When the first judgment result is that the acquisition is successful, the measurement at the corresponding infrared measurement point position is stopped; when the first judgment result is that the acquisition is unsuccessful, the infrared acquisition unit needs to re-measure at the corresponding infrared measurement point position; When the second judgment result is that the acquisition is successful, the measurement of the corresponding ultrasonic measurement point position is stopped, and when the first judgment result is that the acquisition is unsuccessful, the ultrasonic acquisition unit needs to re-measure at the corresponding ultrasonic measurement point position.

5. A child trunk fat percentage prediction model as claimed in claim 4, characterized in that: The logic of the one-time judgment includes: when the infrared acquisition unit measures the subcutaneous fat of the test child at each infrared measurement point position in turn, the pressure data is obtained by the pressure sensor on the infrared light emitting diode probe, and the minimum pressure and the average pressure are calculated based on the pressure data, and the acquisition is judged to be successful when and only when the minimum pressure is greater than the first pressure threshold and the average pressure is greater than the second pressure threshold, otherwise it is judged that the acquisition is unsuccessful; The logic of the secondary judgment includes: when the ultrasonic acquisition unit measures the visceral fat of the tested child at each ultrasonic measurement point in turn, the ultrasonic electrical signal data is obtained through the piezoelectric crystal in the ultrasonic acquisition unit and the received echo time is recorded, the ultrasonic electrical signal data obtains the maximum echo intensity and the minimum echo intensity, and the ultrasonic time difference is calculated based on the ultrasonic emission time and the received echo time. The acquisition is judged to be successful if and only if the ultrasonic time difference is greater than the time difference threshold, the maximum echo intensity is less than the first echo intensity threshold, and the minimum echo intensity is greater than the second echo intensity threshold; otherwise, the acquisition is judged to be unsuccessful.

6. A child trunk fat percentage prediction model as claimed in claim 1, characterized in that: The infrared calculation unit is used to calculate the infrared electrical signal data to obtain the subcutaneous fat content of the tested child, perform denoising processing on the infrared light signal through a bandpass filter to obtain a smooth infrared signal, calculate the Nyquist frequency based on the smooth infrared signal, discretize the smooth infrared signal with N times the Nyquist frequency as the sampling frequency to obtain an infrared light discrete sequence, vectorize the infrared light discrete sequence to obtain an infrared light discrete vector, sort the infrared light discrete vectors in order from low to high according to the wavelengths corresponding to them and perform vector splicing processing to obtain an infrared light discrete matrix, establish a prediction regression equation group of the infrared light discrete matrix and the predicted thickness of subcutaneous fat through the partial least squares method, calculate the ridge regression equation coefficient and the least squares residual square based on the preset regularization parameter and the prediction regression equation group, calculate the regression parameter based on the ridge regression equation coefficient, the least squares residual square and the ridge regression polynomial, calculate the score matrix through the regression parameter, the prediction regression equation group and the preset regularization parameter, and calculate the predicted thickness of subcutaneous fat based on the score matrix and the prediction regression equation group.

7. A child trunk fat percentage prediction model as claimed in claim 1, characterized in that: The ultrasonic calculation unit is used to calculate the ultrasonic electrical signal data to obtain the visceral fat content of the tested child, perform high-frequency noise removal processing on the ultrasonic electrical signal data through Butterworth low-pass filtering to obtain a denoised ultrasonic signal, calculate the ultrasonic attenuation coefficient of the preset low-frequency output ultrasonic wave and the denoised ultrasonic signal through the cross-correlation method, establish an ultrasonic attenuation grayscale image based on the ultrasonic acoustic attenuation coefficient by grayscale level mapping, perform adaptive histogram equalization and bilinear interpolation processing on the ultrasonic attenuation grayscale image to obtain an enhanced ultrasonic image, perform texture feature extraction on the enhanced ultrasonic image based on a preset Markov random method to obtain a feature ultrasonic image, obtain a feature prediction image through image corrosion processing and preset binary segmentation, and obtain the predicted visceral fat content based on the ratio of grayscale pixels in the feature prediction image to all pixels.

8. A child trunk fat percentage prediction model as claimed in claim 1, characterized in that: The comprehensive prediction unit is used to calculate the predicted value of the child's trunk fat percentage, calculate the ternary body mass index based on the height and weight of the tested child, calculate the trunk coefficient based on the ternary body mass index, and calculate the predicted value of the child's trunk fat percentage based on the weight, height, trunk coefficient, predicted subcutaneous fat thickness and predicted visceral fat content of the tested child. The calculation expression is: TBMI=m / h 3 λ=TBMI / (1+TBMI) PRV=αPFC+(1 / mλ)πρβh[δ(ATF) 2 +2ATFe ελ ] Among them, TBMI represents ternary body mass index, PRV represents the predicted value of trunk fat percentage of children, PFC represents the predicted content of visceral fat, α represents the predicted content coefficient of visceral fat, m represents the weight of the tested child, λ represents the trunk coefficient, π represents pi, ρ represents fat density, β represents the predicted content coefficient of subcutaneous fat, h represents the height of the tested child, ATF represents the predicted thickness of subcutaneous fat, δ represents the predicted thickness coefficient of subcutaneous fat, e represents the natural constant, and ε represents the index parameter.

9. A child trunk fat percentage prediction model as claimed in claim 1, characterized in that: The interactive unit is used to receive operation instructions and issue sound and light reminders, the operation instructions include start detection instructions, stop detection instructions, test child information input instructions, prediction result query instructions, historical record data query instructions, first pressure threshold setting, second pressure threshold setting, time difference threshold setting, first echo intensity threshold setting, and second echo intensity threshold setting, and the sound and light reminders include operation instruction reception prompt sound, acquisition success sound and light reminder, and acquisition failure sound and light reminder; The display unit is used to provide the user with an operation interface and visualization results, including operation instruction display, data collection result display, prediction result display and historical record data display.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the child trunk fat percentage prediction model described in any one of claims 1-9 is implemented.

Images