Body composition analysis system and method based on bioelectrical impedance analysis technology for small animals

Abstract

The present invention relates to a body composition measurement system and method based on bioelectrical impedance analysis technology for small animals, comprising: a bioelectrical impedance measuring device for applying a current to the sole of a small animal and measuring a resulting bioelectrical impedance, reactance, and phase angle; a frequency optimization module for determining an optimal frequency for each individual by analyzing measurement data per frequency collected through the bioelectrical impedance measuring device; a body composition analysis module for, on the basis of data measured at the optimal frequency, predicting the body composition of the small animal using a neural network-based machine learning model; and an output module for providing body composition information calculated through the body composition analysis module to a user.

Description

Body composition analysis system and method based on bioimpedance analysis technology for small animals

[0001] The present invention relates to a system and method for analyzing body composition using bioimpedance analysis (BIA) technology for small animals. More specifically, the present invention relates to a technology that automatically applies individual bioadaptive frequencies and precisely analyzes body composition using a neural network-based machine learning model.

[0002]

[0003] Body composition analysis is a crucial element in assessing the health of small animals. Previously, body composition was primarily estimated based on weight measurement or visual assessment. However, these methods had limitations in accurately reflecting individual physiological differences. Therefore, bioelectrical impedance analysis (BIA) technology was introduced for more precise body composition analysis. BIA measures bioimpedance by passing a small electric current through the body, and then analyzes body composition based on this data. Research and application of BIA in human subjects are actively underway.

[0004] However, most of the currently used BIA technologies are based on models designed based on the human body, so they have limitations in sufficiently reflecting the physical characteristics of small animals, and there is a problem in that the frequency range optimized for the physiological characteristics and body structure of small animals is not clear.

[0005] In addition, the conventional BIA method has difficulty in reflecting individual characteristics by using a fixed frequency or a simple multi-frequency method, and there is a lack of a process for analyzing individual signal data in detail or determining the optimal frequency to accurately analyze body composition, so there is a limit to securing reliable data. Therefore, there is a need for the development of technology that can automatically search for the optimal frequency according to the individual characteristics of small animals and perform more precise body composition analysis.

[0006]

[0007] [Prior Art Literature]

[0008] [Patent Document]

[0009] Korean Patent Publication No. 10-2012-0012258 (published on February 9, 2012)

[0010] Korean Patent Publication No. 10-2024-0080423 (published on June 7, 2024)

[0011]

[0012] The present invention aims to address the aforementioned issues and relates to a body composition analysis system and method based on bioimpedance analysis (BIA) technology for small animals. More specifically, the present invention provides a body composition measurement system and method based on bioimpedance analysis technology for small animals, which automatically applies individual bioadaptive frequencies and precisely analyzes body composition using a neural network-based machine learning model.

[0013] The problem to be solved by this specification is not limited to what has been described above, and can be expanded to various matters that can be derived from the embodiments of the invention described below.

[0014]

[0015] A body composition analysis system (100) based on bioimpedance analysis technology for small animals according to one embodiment of the present invention may include a bioimpedance measurement device (110) that applies a microcurrent to the sole of a small animal and measures bioimpedance, reactance, and phase angle accordingly, a frequency optimization module (120) that analyzes frequency-based measurement data collected through the bioimpedance measurement device (110) to determine an optimal frequency for each individual, a body composition analysis module (130) that predicts the body composition of the small animal using a neural network-based machine learning model based on data measured at the optimal frequency, and an output module (140) that provides body composition information calculated through the body composition analysis module (130) to a user.

[0016] The bioimpedance measuring device (110) according to one embodiment of the present invention includes a measurement space (111) into which both front legs and both hind legs of a small animal are inserted, and a plurality of electrodes that contact the soles of each of the front legs and both hind legs of the small animal and apply microcurrents are arranged in the measurement space (111), and the measurement space (111) can be formed in a size that is universally compatible with small animals of various sizes.

[0017]

[0018] The frequency optimization module (120) according to one embodiment of the present invention can scan frequencies within a preset range, analyze impedance change patterns at specific frequencies, and automatically search for optimal frequencies for each individual in real time using a machine learning model based on a recurrent neural network (RNN).

[0019] The body composition analysis module (130) according to one embodiment of the present invention includes a body composition prediction model based on a multi-layer perceptron (MLP), and inputs individual impedance (Z), reactance (R), phase angle (φ), body length, body height, and body weight as input variables into the body composition prediction model, and then analyzes lean body mass (FFM), etc. within a preset error compared to the DEXA standard through deep learning operations.

[0020] According to another embodiment of the present invention, a body composition measurement method based on bio-impedance analysis technology for a small animal includes a step of applying a microcurrent to the sole of the small animal through a bio-impedance measurement device including a measurement space (111) into which both forelegs and both hind legs of the animal are inserted, measuring impedance (Z), reactance (R), and phase angle (φ) within a preset frequency range, and analyzing a change pattern of the corresponding values, a step of automatically searching for an optimal frequency for each individual in real time through an RNN-based machine learning model using the analyzed data, and a multilayer perceptron-based body composition prediction model based on the data measured at the optimal frequency, wherein the body composition prediction model may include a step of inputting impedance (Z), reactance (R), phase angle (φ), body length, body height, and body weight for each individual as input variables, and calculating fat-free mass (FFM) and the like within a preset error compared to a DEXA standard through a deep learning operation, and a step of providing the calculated body composition information to a user.

[0021]

[0022] According to one embodiment of the present invention, by applying a microcurrent to the sole of a small animal to measure bioimpedance, and then automatically searching for the optimal frequency for each individual in real time using an RNN-based machine learning model, it has the advantage of enabling more precise body composition analysis than the conventional BIA method.

[0023] In addition, according to one embodiment of the present invention, by applying an MLP-based body composition prediction model and utilizing individual impedance (Z), reactance (R), phase angle (φ), body length, body height, and body weight as input variables, fat-free mass (FFM) can be predicted within a preset error compared to the DEXA standard, thereby having the advantage of improving the reliability of body composition analysis compared to the existing method.

[0024] In addition, according to one embodiment of the present invention, by designing the measurement space of the bioimpedance measuring device to be universally compatible with small animals of various sizes, there is an advantage of being able to obtain consistent measurement results for small animals of various body shapes.

[0025] In addition, according to one embodiment of the present invention, noise is minimized by applying optimized frequency scan and data analysis techniques, and body composition analysis reflecting individual physiological characteristics is possible, thereby providing an advantage in more reliably evaluating the health status of small animals.

[0026] It should be understood that the effects of this specification are not limited to the matters described above, but can be expanded to various contents that can be derived from the detailed description of the embodiments of the invention below.

[0027]

[0028] FIG. 1 is a drawing for explaining a body composition analysis system (100) based on a bioimpedance analysis technology for small animals according to one embodiment of the present invention.

[0029] FIG. 2 is a drawing schematically showing the overall shape of a bioimpedance measuring device (110) according to one embodiment of the present invention.

[0030] FIG. 3 is a drawing for explaining a process of determining the arrangement position between electrodes placed in a measurement space (111) of a bioimpedance measuring device (110) according to one embodiment of the present invention.

[0031] Figure 4 is a flowchart for more specifically explaining the process of predicting the body composition of a small animal in a body composition analysis module (130) based on data measured through a bioimpedance measuring device (110).

[0032] FIG. 5 is a flowchart showing a method for measuring the body composition of a small animal in a series of sequential steps using a body composition analysis system (100) based on a bioimpedance analysis technology for small animals according to one embodiment of the present invention.

[0033] FIG. 6 is a drawing for explaining a body composition analysis system (100) based on a bioimpedance analysis technology for small animals according to another embodiment of the present invention.

[0034] Figure 7 is a drawing schematically showing the overall shape of a bioimpedance measuring device (110) according to another embodiment of the present invention.

[0035]

[0036] A bioimpedance measuring device (110) that applies a microcurrent to the sole of a small animal and measures the resulting impedance, reactance, and phase angle;

[0037] A frequency optimization module (120) that analyzes frequency-specific measurement data collected through the above bioimpedance measuring device (110) to determine the optimal frequency for each individual;

[0038] A body composition analysis module (130) that predicts the body composition of a small animal using a neural network-based machine learning model based on data measured at the above optimal frequency; and

[0039] An output module (140) that provides the user with body composition information calculated through the body composition analysis module (130);

[0040] Body composition measurement system based on bioimpedance analysis technology for small animals.

[0041]

[0042] In describing the embodiments of this specification, if a detailed description of a known configuration or function is judged to obscure the gist of the embodiments of this specification, a detailed description thereof will be omitted. In addition, parts of the drawings that are not related to the description of the embodiments of this specification have been omitted, and similar parts have been designated with similar drawing reference numerals.

[0043] In the embodiments of this specification, when a component is said to be "connected," "coupled," or "connected" to another component, this may include not only a direct connection, but also an indirect connection in which another component exists in between. Furthermore, when a component is said to "include" or "have" another component, unless otherwise specifically stated, this does not exclude the other component, but rather implies that the other component may be included.

[0044] In the embodiments of this specification, the terms first, second, etc. are used only for the purpose of distinguishing one component from another component, and do not limit the order or importance between components unless specifically stated otherwise. Therefore, within the scope of the embodiments of this specification, a first component in an embodiment may be referred to as a second component in another embodiment, and similarly, a second component in an embodiment may be referred to as a first component in another embodiment.

[0045] In the embodiments of this specification, distinct components are used to clearly illustrate their respective characteristics and do not necessarily imply separation. That is, multiple components may be integrated into a single hardware or software unit, or a single component may be distributed into multiple hardware or software units. Therefore, even if not specifically mentioned, such integrated or distributed embodiments are also included within the scope of the embodiments of this specification.

[0046] In this specification, the term "network" may encompass both wired and wireless networks. In this case, the term "network" may refer to a communications network that enables data exchange between devices, systems, and devices, and is not limited to a specific network.

[0047] Embodiments described herein may be entirely hardware, partially hardware and partially software, or entirely software. As used herein, "unit," "device," or "system" refers to a computer-related entity such as hardware, a combination of hardware and software, or software. For example, a unit, module, device, or system as used herein may be, but is not limited to, a running process, a processor, an object, an executable, a thread of execution, a program, and / or a computer. For example, both an application running on a computer and the computer itself may correspond to a unit, module, device, or system as used herein.

[0048] Additionally, in this specification, a device may be a mobile device such as a smartphone, tablet PC, wearable device, or HMD (Head Mounted Display), as well as a fixed device such as a PC or home appliance with display functions. Furthermore, as an example, a device may be an in-vehicle cluster or an IoT (Internet of Things) device. In other words, in this specification, a device may refer to any device capable of operating an application, and is not limited to a specific type. For convenience of explanation, the device on which an application operates is referred to as a device below.

[0049] In this specification, the network communication method is not limited, and connections between each component may not be made using the same network method. The network may include not only communication methods utilizing communication networks (e.g., mobile communication networks, wired Internet, wireless Internet, broadcasting networks, satellite networks, etc.), but also short-range wireless communication between devices. For example, the network may include all communication methods that enable objects to network with each other, and is not limited to wired communication, wireless communication, 3G, 4G, 5G, or other methods. For example, wired and / or networks include Local Area Network (LAN), Metropolitan Area Network (MAN), Global System for Mobile Network (GSM), Enhanced Data GSM Environment (EDGE), High Speed ​​Downlink Packet Access (HSDPA), Wideband Code Division Multiple Access (W-CDMA), Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Bluetooth, Zigbee, Wi-Fi, VoIP (Voice over Internet Protocol), LTE Advanced, IEEE802.16m, WirelessMAN-Advanced, HSPA+, 3GPP Long Term Evolution (LTE), Mobile WiMAX (IEEE 802.16e), UMB (formerly EV-DO Rev. C), Flash-OFDM, iBurst and MBWA (IEEE 802.20) It may refer to a communication network using one or more communication methods selected from the group consisting of systems, HIPERMAN, Beam-Division Multiple Access (BDMA), Wi-MAX (World Interoperability for Microwave Access), and ultrasonic communication, but is not limited thereto.

[0050] The components described in various embodiments are not necessarily essential components, and some may be optional. Therefore, embodiments comprising a subset of the components described in the embodiments are also included within the scope of the embodiments of the present disclosure. Furthermore, embodiments including other components in addition to the components described in various embodiments are also included within the scope of the embodiments of the present disclosure.

[0051]

[0052] According to the present invention, the invention is based on a bio-impedance measuring device for companion animals disclosed in Korean Patent Publication No. 10-2024-0080423 (publication date: 2024.06.07.) filed by the inventor of the present invention.

[0053] More specifically, the present invention relates to a body composition analysis system and method based on bioimpedance analysis technology for small animals, which is an improvement on the existing invention, by improving the foot-fixing structure of a conventional bioimpedance measurement device for companion animals, thereby enabling more stable control of the companion animal's movements, and by dynamically applying a frequency suitable for each individual's biological response through an automatic frequency optimization system, thereby improving the accuracy of body composition analysis.

[0054]

[0055] Hereinafter, embodiments of the present specification will be described in detail with reference to the drawings.

[0056]

[0057] FIG. 1 is a drawing for explaining a body composition analysis system (100) based on a bioimpedance analysis technology for small animals according to one embodiment of the present invention.

[0058]

[0059] The small animals mentioned in the present invention are not particularly limited, but preferably refer to dogs and cats.

[0060]

[0061] Referring to FIG. 1, a body composition analysis system (100) based on a bioimpedance analysis technology for a small animal according to one embodiment of the present invention is largely configured to include a bioimpedance measurement device (110), a frequency optimization module (120), a body composition analysis module (130), and an output module (140).

[0062] The bioimpedance measuring device (110) applies a microscopic current to the sole of a small animal and measures the resulting impedance (Z), reactance (X), and phase angle (φ). A more detailed look at this is as follows.

[0063]

[0064] Fig. 2 is a drawing schematically showing the overall shape of a bio-impedance measuring device (110), and Fig. 3 is a drawing for explaining a process of determining the arrangement positions between electrodes placed in a measurement space (111) of a bio-impedance measuring device (110).

[0065]

[0066] First, looking at Fig. 2, the bioimpedance measuring device (110) is a device for precisely analyzing the bioimpedance of a small animal. It applies a microcurrent to the sole of the small animal and measures the resulting impedance (Z), reactance (X), and phase angle (φ). This bioimpedance measuring device (110) is configured to be applicable to small animals of various sizes, and has a structure that allows the small animal to be positioned in the measurement space (111) in a natural posture.

[0067] More specifically, a measurement space (111) is formed in the center of the bioimpedance measuring device (110) into which the two forelegs and two hind legs of a small animal are inserted. The measurement space (111) is formed in a universal size so that it can be applied to small animals of various sizes, and has a structure that allows the small animal to be stably supported according to its body size. This allows the small animal to maintain a natural posture without unnecessary movement during body composition measurement, thereby improving the accuracy of body composition analysis.

[0068] In addition, a number of electrodes are arranged on the floor inside the measurement space (111) to apply a microscopic current to the soles of the feet of the small animal and measure bioimpedance. The electrodes are arranged to contact the floors of both front legs and both hind legs of the small animal, respectively, and are configured to cover the entire area of ​​the floor inside the measurement space (111) so as to be able to respond to differences in body shape of each individual.

[0069]

[0070] Additionally, the bioimpedance measuring device (110) may include a multi-contact electrode structure. By arranging the electrodes in two or more layers, stable signal acquisition is ensured through primary and secondary contacts, thereby minimizing signal errors due to changes in body contact conditions. Furthermore, a function for automatically compensating for changes in contact resistance according to the skin condition of the paws of small animals (dryness, moisture, keratin, etc.) may be added. This enables real-time measurement of the contact resistance of the electrodes and signal compensation reflecting this.

[0071]

[0072] In addition, in one embodiment, the bioimpedance measuring device (110) may include a signal processing module for processing and transmitting the measured impedance (Z), reactance (X), and phase angle (φ) data to an analysis system. The signal processing module collects bioelectrical signals measured from the body of a small animal and then converts them into digital data. The converted data may be transmitted to a frequency optimization module (120) and a body composition analysis module (130) described below, so that additional signal analysis and body composition prediction can be performed.

[0073]

[0074] In addition, in one embodiment, the bioimpedance measuring device (110) measures impedance (Z), reactance (R), and phase angle (φ) within a preset frequency range and analyzes the change pattern of the corresponding values. The measured data is linked to a frequency optimization module (120) that automatically searches for the optimal frequency for each individual in real time, and through this, the present invention can automatically search for the optimal frequency suitable for each individual in real time to perform more precise body composition analysis.

[0075] Additionally, a FSR sensor may be placed within the bioimpedance measurement device (110) to automatically detect the foot position of the subject. This automatically detects whether the subject is correctly positioned in the measurement space (111), and data is recorded only when accurate signal measurement is possible, thereby improving measurement reliability.

[0076]

[0077] Looking at Figure 3, the arrangement positions between electrodes placed in the measurement space (111) of the bioimpedance measuring device (110) can be determined through the following processes.

[0078] First, a step is performed to determine the placement of electrodes by considering the body structure and contact points of the small animal inserted into the measurement space (111). Accordingly, the electrodes are placed so that they can make stable contact with both forelimbs and both hindlimbs of the small animal, and a process is included to adjust the spacing and placement of the electrodes to accommodate individual body shape differences.

[0079] Additionally, to improve the search algorithm of the frequency optimization module (120), an adaptive frequency search algorithm can be added instead of the existing simple sequential frequency application method. The adaptive frequency search algorithm can improve accuracy while reducing the amount of computation by performing tests at three representative frequencies initially and then performing detailed optimization near the frequency with the best response.

[0080] Next, a step is performed to optimize the distance between electrodes to maximize measurement accuracy. Since the bioimpedance of small animals varies depending on individual body type, muscle mass, and water content, this step determines the electrode placement to ensure consistent signal transmission while maintaining a consistent body contact area. This improves the reliability of measurement data and minimizes unnecessary fluctuations.

[0081] Additionally, the electrode placement is adjusted to take into account the measurement environment of the small animal. When the limbs of the small animal are inserted into the measurement space (111), the electrodes are positioned to ensure stable contact with the body to prevent contact imbalance that may occur due to movement. The structural design of the bioimpedance measurement device (110) enables universal compatibility with small animals of various sizes, thereby reducing errors in body composition measurement.

[0082]

[0083] Returning to Figure 1, the frequency optimization module (120) analyzes the impedance (Z), reactance (X), and phase angle (φ) data measured through the bioimpedance measuring device (110) to determine the optimal frequency for each individual.

[0084] More specifically, the frequency optimization module (120) first performs a frequency scan step that sequentially applies various frequencies within a preset range to measure the bioimpedance of a small animal. Since the electrical properties of biological tissues generally respond differently at low and high frequencies, a process of acquiring data by scanning multiple frequencies rather than a single frequency is necessary.

[0085] To this end, the frequency optimization module (120) sequentially applies various frequencies between 250 Hz and 1 MHz, collecting impedance, reactance, and phase angle data measured at each frequency. During this process, the frequency scan does not end with a single measurement, but rather performs at least five or more repeated measurements to secure highly reliable data.

[0086] In addition, the frequency optimization module (120) analyzes the bioimpedance change pattern according to frequency using the collected data. Bioimpedance varies depending on the body composition of an individual (water content, fat content, muscle mass, etc.), and it is important to evaluate whether the response is prominent at a specific frequency. To this end, the frequency optimization module (120) analyzes the changes in impedance, reactance, and phase angle data according to frequency, and evaluates the signal-to-noise ratio (SNR) to select a frequency suitable for body composition analysis. In addition, it removes measurement noise by applying a filtering technique (e.g., low-pass filter, wavelet transform) and derives a frequency optimized for body composition analysis.

[0087] Thereafter, the frequency optimization module (120) automatically determines the most suitable frequency for each individual based on the analyzed data. In particular, during this process, the frequency optimization module (120) applies a machine learning model based on a recurrent neural network (RNN) to search for the optimal frequency.

[0088] The RNN model learns individual frequency scan data, enabling real-time determination of the optimal frequency for body composition analysis based on each individual's physiological and physical characteristics. Once the optimal frequency is determined, it is applied during subsequent body composition analysis processes to achieve more accurate impedance measurements.

[0089]

[0090] The body composition analysis module (130) analyzes data collected from the body composition measurement device (110) and the frequency optimization module (120) to predict individual body composition. In particular, the body composition analysis module (130) reflects the physical characteristics of small animals and performs multivariate analysis including the cell correction factor (CCF) to increase the reliability of body composition prediction.

[0091] The body composition analysis module (130) uses impedance (Z), reactance (X), and phase angle (φ) data measured at the optimal frequency determined by the frequency optimization module (120) as input values. Additionally, it precisely predicts the lean body mass (FFM) of small animals by inputting body weight, body length, body height, and cell correction factor (CCF).

[0092] This module utilizes a neural network model based on the Multi-Layer Perceptron (MLP). The optimized structure, based on experimental analysis, includes three hidden layers, each with 300 neurons. The Rectified Linear Unit (ReLU) activation function effectively learns nonlinearities and prevents the vanishing gradient problem, thereby enhancing the performance of the neural network.

[0093] Additionally, the Adam optimizer was applied for model optimization to improve learning speed and maximize body composition prediction performance. The Adam optimizer effectively adjusts weight updates for body composition measurement data, contributing to the increased accuracy of small animal body composition analysis.

[0094]

[0095] Meanwhile, in the body composition analysis process, the neural network model operates as follows.

[0096] - Input Layer

[0097] In the input layer, a total of six variables are received as input values: impedance (Z, X1), reactance (X, X2), phase angle (φ, X3), body weight (X4), body length (X5), and body height (X6).

[0098] The variables entered in this way are connected to 300 neurons, and the ReLU (Rectified Linear Unit) activation function is applied.

[0099]

[0100] - Hidden Layers Operation Process

[0101] This neural network model consists of a total of three hidden layers, and the computational process of each hidden layer is as follows.

[0102] First hidden layer: H1 = ReLU(W1 · [X1, X2, X3, X4, X5, X6] + b1)

[0103] Second hidden layer: H2 = ReLU(W2 · H1 + b2)

[0104] Third hidden layer: H3 = ReLU(W3 · H2 + b3)

[0105]

[0106] - Output Layer

[0107] Finally, the fat-free mass (FFM) is predicted, and its formula is defined as follows:

[0108] FFM(kg)=W4·H3+b4

[0109] In the output layer, no activation function is applied, and the mean squared error (MSE) is used as the loss function for optimization.

[0110]

[0111] The body composition analysis module (130) of the present invention shows improved prediction performance compared to the existing method, and the evaluation results of the neural network model are as follows.

[0112] Prediction accuracy (R 2 ) = 0.9933

[0113] Mean absolute error (MAE) = 0.0621 kg

[0114] Root mean square error (RMSE) = 0.0984 kg

[0115] Through this, it can be confirmed that the body composition analysis module (130) of the present invention can predict body composition with high reliability and perform precise body composition analysis reflecting the physical characteristics of each individual.

[0116]

[0117] The process of predicting the body composition of a small animal in the body composition analysis module (130) is examined in more detail as follows.

[0118] Figure 4 is a flowchart for more specifically explaining the process of predicting the body composition of a small animal in a body composition analysis module (130) based on data measured through a bioimpedance measuring device (110).

[0119] Referring to Figure 4, a current is first applied to the sole of a small animal's foot via a bioimpedance measuring device (110), and the resulting measured impedance, reactance, and phase angle data are collected (S401). The data measured during this process are acquired at various frequencies and subsequently used for analysis during the frequency optimization process.

[0120] Next, the frequency optimization module (120) automatically searches for the optimal frequency for each individual based on the collected impedance, reactance, and phase angle data (S402). This process analyzes changes in impedance, reactance, and phase angle data for each frequency within a preset frequency range, and determines the most appropriate frequency. The determined optimal frequency is then transmitted to the body composition analysis module (130).

[0121] Next, the body composition analysis module (130) inputs data measured at the optimal frequency determined by the frequency optimization module (120) as input values ​​(S403). At this time, the input variables include impedance, reactance, phase angle, body length, and body height data.

[0122] Next, the body composition analysis module (130) performs a step of processing the input data using a neural network model based on a multilayer perceptron (MLP) (S404). In this process, the body composition analysis module (130) inputs bioimpedance, phase angle, reactance, body length, body height, body weight of the individual, and cell correction constant values ​​into the input layer, performs neural network operations through multiple hidden layers, and then predicts the body composition (lean body mass (FFM) and fat mass (FM)) of the small animal in the final output layer.

[0123] Next, the neural network model performs a process to finally produce the calculated body composition analysis results (S405). During this process, the body composition analysis module (130) can review whether the predicted body composition values ​​fall within a preset standard error range and perform a process to correct the resulting values ​​to ensure a highly reliable analysis result.

[0124] In addition to the function of analyzing body composition based on the MLP neural network model, the body composition analysis module (130) can introduce a hybrid neural network model based on a CNN (Convolutional Neural Network) and an RNN (Recurrent Neural Network) for more precise analysis. This allows for predicting body composition changes reflecting temporal patterns and more accurately analyzing individual biological change trends. Furthermore, a function can be added to analyze the correlation between exercise volume and body composition changes by learning individual physical activity data together.

[0125]

[0126] Returning to Figure 1, the output module (140) serves to provide the user with body composition data predicted by the body composition analysis module (130). In particular, the output module (140) performs the function of outputting the analyzed body composition information in a more clear and usable manner.

[0127] To this end, the output module (140) processes body composition data such as lean body mass (FFM) and fat mass (FM) derived from the body composition analysis module (130) and converts them into a form that is easy for the user to understand. At this time, the analysis results can be configured to be compared with preset reference values ​​to determine whether they are within the normal range, and the output module (140) can also perform the role of analyzing and providing changes in body composition for each individual.

[0128] Additionally, the output module (140) can output body composition analysis data not only as simple numerical values, but also provide graphical or visual information to enable users to more intuitively interpret the analysis results. This can be useful for tracking changes in body composition in small animals and providing personalized health management for each individual.

[0129] Additionally, the output module (140) may store measured body composition data or provide a function to record and manage body composition information by linking with an external system. This allows users to compare past measurement data for long-term health management and more precisely analyze body composition change patterns.

[0130] Additionally, the output module (140) can provide graph and data comparison functions to enable visual analysis of body composition change data. The output module (140) can be improved to store and manage individual body composition change histories by linking with a mobile application via Bluetooth and Wi-Fi connections.

[0131]

[0132] Next, we will examine in a sequential manner a method for measuring the body composition of a small animal using a body composition analysis system (100) based on the previously discussed small animal bioimpedance analysis technology.

[0133] FIG. 5 is a flowchart showing a method for measuring the body composition of a small animal in a series of sequential steps using a body composition analysis system (100) based on a bioimpedance analysis technology for small animals according to one embodiment of the present invention.

[0134] Referring to Figure 5, first, there is a step (S501) of placing a small animal in a measurement space (111) of a bioimpedance measuring device (110). The measurement space (111) is configured so that both front legs and both hind legs of the small animal can be inserted. In particular, the measurement space (111) can be formed in a size that is universally compatible with small animals of various sizes.

[0135] Next, the bioimpedance measurement device (110) applies a microcurrent to the sole of the small animal's foot and measures the resulting impedance, reactance, and phase angle data (S501). The data measured through this process is used as basic data for body composition analysis of the small animal and is then transmitted to the frequency optimization module (120).

[0136] Next, based on the measured bioimpedance data, the frequency optimization module (120) searches for the optimal frequency for each individual (S503). In this step, the frequency optimization module (120) sequentially applies various frequencies within a preset range and analyzes the impedance change at each frequency. Thereafter, the frequency optimization module (120) evaluates the signal response for each frequency and automatically determines the frequency most suitable for body composition analysis for each individual, and the determined optimal frequency is transmitted to the body composition analysis module (130).

[0137] Next, the body composition analysis module (130) receives data measured at the optimal frequency as input values ​​(S504). At this time, input variables include impedance, reactance, phase angle, body weight, body length, body height, etc., and the collected data is ready to be processed in a neural network model for body composition analysis.

[0138] Next, the body composition analysis module (130) performs a step of predicting body composition based on the input data (S505). In this process, the body composition analysis module (130) analyzes body composition by utilizing a neural network-based multilayer perceptron (MLP) model. In addition, bioimpedance, phase angle, reactance, body length, body height, body weight, and cell correction constant values ​​are input in the input layer, and neural network operations are performed through multiple hidden layers, and finally, lean body mass (FFM) and fat mass (FM) of the small animal are predicted in the output layer.

[0139] Next, a step is performed to derive body composition analysis results using neural network operations (S506). The body composition data analyzed through this process can be compared to pre-established reference values ​​to determine whether the results are within the normal range. This process may include necessary corrections to ensure the results are within the reliable analysis results.

[0140] Next, the output module (140) performs a step of providing the body composition analysis results to the user (S507). The analyzed data is converted and processed for easy user understanding and output. Furthermore, body composition information can be provided in graphical or numerical form, and the output module (140) can provide a function for tracking changes in body composition for each individual.

[0141] Additionally, the output module (140) controls the body composition measurement device, and the output can be practically performed on a cloud-based dashboard.

[0142]

[0143] FIG. 6 is a drawing for explaining a body composition analysis system (100) based on a bioimpedance analysis technology for small animals according to another embodiment of the present invention.

[0144] Looking at Fig. 6, it includes all of the body composition analysis system (100) based on the bioimpedance analysis technology for small animals described above, and further includes a body length and body height measurement module (150).

[0145] In other words, a body composition analysis system (100) based on a bioimpedance analysis technology for a small animal according to another embodiment of the present invention is configured to include a bioimpedance measurement device (110), a frequency optimization module (120), a body composition analysis module (130), an output module (140), and a body length and height measurement module (150).

[0146] Here, the body length and height measurement module (150) is formed by attaching a vision camera to the back of the touch panel of the output module (140), and can analyze and provide the body length and height of the measured animal. (See Fig. 7)

[0147] In addition, the above-mentioned vision camera can perform a preliminary calibration on the measurement object by utilizing six or more calibration points on the measurement space (111), and can measure the body length and height of a small animal to derive a value.

[0148]

[0149] The present invention is not limited to the specific preferred embodiments described above, and anyone with ordinary skill in the art to which the invention pertains can make various modifications without departing from the gist of the present invention claimed in the claims, and as long as it relates to technical ideas forming such modifications, it is within the scope of the claims.

[0150]

[0151] [Explanation of symbols]

[0152] 100: Body composition measurement system based on bioimpedance analysis technology for small animals

[0153] 110: Bioimpedance measuring device

[0154] 111: Measurement space

[0155] 120: Frequency Optimization Module

[0156] 130: Body Composition Analysis Module

[0157] 140: Output module

[0158] 150: Body length and height measurement module

Claims

1. A bioimpedance measuring device (110) that applies a microcurrent to the sole of a small animal and measures the resulting impedance, reactance, and phase angle; A frequency optimization module (120) that analyzes frequency-specific measurement data collected through the above bioimpedance measuring device (110) to determine the optimal frequency for each individual; A body composition analysis module (130) that predicts the body composition of a small animal using a neural network-based machine learning model based on data measured at the above optimal frequency; and An output module (140) that provides the user with body composition information calculated through the body composition analysis module (130); Body composition measurement system based on bioimpedance analysis technology for small animals.

2. In paragraph 1, The above bioimpedance measuring device (110) is It includes a measurement space (111) into which both front legs and both hind legs of a small animal are inserted, and a plurality of electrodes that contact the soles of each of the front legs and both hind legs of the small animal and apply microcurrent are arranged in the measurement space (111), and the measurement space (111) is formed in a size that is universally compatible with small animals of various sizes. Body composition measurement system based on bioimpedance analysis technology for small animals.

3. In paragraph 1, The above frequency optimization module (120) is Scanning the frequency within a preset range, analyzing the impedance change pattern at a specific frequency, and continuously adjusting the optimal frequency for each individual through data feedback using an RNN-based machine learning model. Body composition measurement system based on bioimpedance analysis technology for small animals.

4. In paragraph 1, The above body composition analysis module (130) A body composition measurement system based on bioimpedance analysis technology for small animals, comprising a body composition prediction model based on a multi-layer perceptron (MLP), wherein the body composition prediction model inputs individual impedance (Z), reactance (X) and phase angle (φ), body length, body height, and body weight as input variables, and then learns to maintain the mean absolute error (MAE) and root mean square error (RMSE) below a specific threshold by comparing the fat-free mass (FFM) with the DEXA measurement value through deep learning operation.

5. A step of applying a microcurrent to the sole of a small animal's foot through a bioimpedance measuring device including a measurement space (111) into which both front legs and both hind legs of the animal are inserted, and measuring the resulting impedance (Z), reactance (R), and phase angle (φ); A step of collecting measurement data by frequency through the above bioimpedance measuring device, scanning the frequency within a preset range, and analyzing the impedance, reactance, and phase angle change pattern at a specific frequency; A step of learning the frequency-dependent bioimpedance change using an RNN-based machine learning model and determining the optimal frequency for each individual; A step of predicting lean body mass (FFM) in a range where the mean absolute error (MAE) and root mean square error (RMSE) compared to the DEXA standard are maintained below a specific value using a multilayer perceptron-based body composition prediction model based on data measured at the optimal frequency, inputting impedance (Z), reactance (R), phase angle (φ), body length, body height, and body weight as input variables, and performing deep learning operations; and comprising a step of providing predicted body composition information to a user; Body composition measurement method based on bioimpedance analysis technology for small animals.

Images