JP2006231077A - Body composition estimation device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a body composition estimation device by which the accuracy of estimation of body fat weight, body fat rate, or the like are improved. <P>SOLUTION: The body composition estimation device is equipped with a means for measuring the amount of bioelectricity to generate probe currents of one or more frequencies generated for frequencies and put the generated probe current in the subject's body to measure a voltage applied to the living body and a current flowing in the living body, an electric admittance/electric impedance calculation means to calculate electric admittance or electric impedance of the the subject's body when the frequency is infinite and is 0 based on a result of measurement from the means for measuring the amount of bioelectricity, and an estimation means for extracellular fluid volume ECF, intracellular fluid volume ICF, or total body water TBW to estimate at least one of an extracellular fluid volume, an intracellular fluid volume, and total body water using a specific formula. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention mainly relates to a body composition estimation apparatus for estimating a body fat state of a subject such as a bone removal weight / fat weight, a fat weight, a body fat percentage, and the like based on a bioelectrical impedance method.

  Precise measurement of human lean body mass LBM, fat mass FAT, etc. mainly uses X-rays represented by the dual energy X-ray absorption method (hereinafter simply referred to as DXA). The measurement device used is used or the underwater body weight method is used. However, in the measurement by X-ray, the device becomes large and there is a risk of exposure. In addition, the underwater weight method is very complicated because the subject must enter the bathtub.

  Therefore, body composition estimation devices using a bioelectric impedance method have become widespread as simple devices that can be handled without exposure. In this bioelectrical impedance method, the electrical characteristics of a living body vary significantly depending on the type of tissue or organ. For example, in the case of humans, the electrical resistivity of blood is around 150 Ω · cm, whereas bone and fat It is used that the electrical resistivity is 1-5 kΩ · cm.

Bioelectrical impedance consists of the body's resistance to the current carried by ions in the body and the reactance associated with various types of polarization processes created by cell membranes, tissue interfaces, or non-ionized tissues .
As shown in FIG. 12, the cells 1, 1,... Are surrounded by cell membranes 2, 2,..., Which are electrically connected to capacitors having a large capacitance (capacitance). Can see. Therefore, as shown in FIG. 13, the bioelectric impedance is an intracellular fluid consisting of an extracellular fluid impedance consisting only of the extracellular fluid resistance 1 / Ye and a series connection of the intracellular fluid resistance 1 / Yi and the cell membrane capacitance Cm. It can be considered as a parallel combined impedance with impedance.

  By the way, in the conventional body composition estimation apparatus, the frequency of the sinusoidal alternating current to be passed between the surface electrodes of the limbs is fixed to about 50 kHz that is a frequency (characteristic frequency) when the electric phase angle φ is maximum. Since it is configured to measure the bioelectric impedance of the subject, the extracellular fluid resistance 1 / Ye and the intracellular fluid resistance 1 / Yi cannot be obtained separately, and the extracellular fluid impedance and the intracellular Since the body fat state of the subject was estimated based on the parallel synthetic impedance with the liquid impedance and on the basis of the bioelectrical impedance including the capacitive component of the cell membrane, estimation of lean body weight LBM, fat weight FAT, etc. There was a drawback that the accuracy was not so good.

  As a means to eliminate this drawback, a multi-frequency probe current is generated, and the multi-frequency probe current is generated and applied to the subject's body, and the current and the voltage between the subject's voltage measurement surface parts are measured. By measuring the bioelectrical impedance of the subject, the extracellular fluid resistance 1 / Ye and the intracellular fluid resistance 1 / Yi were determined as pure resistance values (not including the cell membrane capacitance Cm), and the obtained cells There is provided a body composition estimation apparatus for estimating the body fat state of a subject based on the external fluid resistance 1 / Ye and the intracellular fluid resistance 1 / Yi (Japanese Patent Laid-Open No. 10-14899 related to the applicant's application). No. publication etc.).

  However, according to the multi-frequency impedance method according to the above application, the estimation accuracy is clearly improved as compared with the impedance method using a single frequency (50 kHz), and the fat weight estimated by the multi-frequency impedance method is recognized. However, this is not sufficient for the calculation of the body fat amount that takes the difference between the body weight and the lean body weight, although the correlation coefficient between the body weight and the fat weight measured with DXA is as high as 0.918.

Japanese Patent Laid-Open No. 10-14899

The present invention has been made in view of the above circumstances, and is capable of accurately estimating the deboning weight and fat weight, and can improve the estimation accuracy of fat weight and body fat percentage. It is the first purpose to provide.
In addition, the present invention provides a body composition estimation device capable of accurately estimating the deboning weight / fat weight and estimating the bone weight so as to improve the estimation accuracy of the fat weight and the body fat percentage. This is the second purpose.

According to the first aspect of the present invention, a probe current having one or more frequencies is generated for each frequency, and the generated probe current is input to the body of the subject, and the voltage applied to the living body and the living body at that time Bioelectric quantity measuring means for measuring the flowing current;
An electrical admittance / electrical impedance calculation means for calculating an electrical admittance or an electrical impedance at a frequency of infinity and a frequency of 0:00 of the subject's body based on a measurement result from the bioelectricity measurement means;
Using formula (2), the amount of extracellular fluid, the amount of intracellular fluid, or the body that estimates at least one of the amount of extracellular fluid ECF, the amount of intracellular fluid ICF, or the amount of body water TBW of the subject It is characterized by comprising water content estimation means.
F = Rm / {Rinf ⁻¹ −Ro ⁻¹ }
Ricf = {- Rm + (Rm 2 + 4F) 0.5} / 2
Recf = 1 / {Rinf ⁻¹ −Ricf ⁻¹ }
LEAN = A ・ H ² / Recf + B ・ H ² / Ricf + C ・ H ^E
ECF = A ・ H ² / Recf + D ・ LEAN
ICF = B ・ H ² / Ricf + 2D ・ LEAN
TBW = ECF + ICF (2)
Ricf: Resistance of intracellular fluid
Recf: Resistance of extracellular fluid
Rinf: electrical impedance when the frequency is infinite Ro: electrical impedance when the frequency is 0 LEAN: bone removal weight / fat weight of the subject's body
H: Height of subject
A, B, C, D, E, Rm: constant (Rm corresponds to resistance of cell membrane)
F: Intermediate variable The invention according to claim 2 relates to the body composition estimation apparatus according to claim 1, wherein the constants A, B, C, D, E, and Rm are determined according to gender, and the electrodes are arranged on the same side of the subject. When attached to the wrist and ankle, it is characterized by 0.5 to 2 times the value shown in Table 4.

  As described above, according to the configuration of the present invention, when estimating the bone removal weight / fat weight LEAN defined as “weight obtained by excluding fat weight and bone weight from the body weight of the subject”, Since the term up to the third power of height H is considered, the body fat state of the subject can be estimated more accurately.

In addition, since the fat weight of the subject is determined in consideration of the bone weight BMC in which individual differences are recognized, the state of the body fat of the subject can be estimated more accurately.
In addition, according to the configuration of claim 1, the subject's extracellular fluid amount ECF, intracellular fluid amount ICF, or body water amount TBW is determined in consideration of the finite cell membrane resistance. The state of body fat can be estimated more accurately.

Embodiments of the present invention will be described below with reference to the drawings. The description will be made specifically with reference to examples.
FIG. 1 is a block diagram showing an electrical configuration of a body composition estimation apparatus according to an embodiment of the present invention, FIG. 2 is a schematic diagram schematically showing a use state of the body composition estimation apparatus, and FIG. 3 is a human body. FIG. 4 is an electrical equivalent circuit diagram of human tissue, FIG. 5 is an electrical equivalent circuit diagram of human tissue at an infinite frequency, and FIG. 6 is an operation procedure of the body composition estimation apparatus. FIG. 7 is a view showing a display example of a display in the body composition estimation apparatus, and FIG. 8 is obtained using the bone removal weight / fat weight LEAN obtained by DXA and the body composition estimation apparatus. FIG. 9 is a correlation diagram between the bone weight BMC obtained by DXA and the bone weight BMC obtained by using the body composition estimation apparatus, FIG. , Fat weight FA determined by DXA FIG. 11 is a correlation diagram between the fat weight FAT obtained using the body composition estimation apparatus and the fat weight FAT obtained using DXA and the fat weight obtained using the conventional body composition estimation apparatus. It is a correlation diagram with FAT.

  The body composition estimation apparatus 4 in this example relates to an apparatus for estimating the body fat state of the subject, such as bone removal weight / fat weight, fat weight, body fat percentage, etc., as shown in FIG. 1 and FIG. A signal output circuit 5 for causing a multi-frequency probe current Ib to flow through the body E as a measurement signal for each frequency; a current detection circuit 6 for detecting a multi-frequency probe current Ib flowing through the body E of the subject; A voltage detection circuit 7 for detecting the voltage Vp between the limbs, a keyboard 8 as an input device, a display 9 as an output device, and a CPU (central processing unit) that controls each part of the device and performs various arithmetic processes. Apparatus) 10, a ROM 11 for storing a processing program of the CPU 10, a RAM 12 in which a data area for temporarily storing various data and a work area of the CPU 10 are set, and a back of the subject at the time of measurement Four surface electrodes Hp to be pasted to conductively to the skin surface of the Ha and instep portion Le, Hc, Lp, is schematically composed of a Lc.

  The keyboard 8 includes a numeric keypad and function keys for inputting a subject's height, age, weight, and the like, a start / end switch for an operator (or subject) to instruct measurement start / measurement end, and the like. It is configured. Height data, age data, and weight data supplied from the keyboard 8 are converted into key codes by a key code generation circuit (not shown) and supplied to the CPU 10. The CPU 10 temporarily stores various operation signals, height data, and the like, which have been code-input, in the data area of the RAM 12.

  The signal output circuit 5 includes a PIO (parallel interface) 51, a measurement signal generator 52, and an output buffer 53. When the signal generation instruction signal SG is supplied from the CPU 10 via the PIO 51 at a predetermined sweep cycle, the measurement signal generator 52 changes in frequency in a range of, for example, 1 kHz to 400 kHz and at a frequency interval of 15 kHz. The measurement signal (current) Ia to be generated is repeatedly generated over a predetermined number of sweeps N and input to the output buffer 53. The output buffer 53 sends the input measurement signal Ia to the surface electrode Hc as a multi-frequency probe current Ib while maintaining a constant current state. At the time of measurement, the surface electrode Hc is attached to the subject's back Ha so as to be conductive, whereby a current Ib in the range of 100 to 800 μA flows through the body E (FIG. 2) of the subject.

  The current detection circuit 6 is roughly composed of an I / V converter (current / voltage converter) 61, a BPF (band pass filter) 62, an A / D converter 63 and a sampling memory 64. The I / V converter 61 is a multi-channel that flows between the body electrode E of the subject, that is, the surface electrode Hc attached to the back part Ha (FIG. 2) of the subject and the surface electrode Lc attached to the back part Le. The frequency probe current Ib is detected and converted to the voltage Vb, and the voltage Vb obtained by the conversion is supplied to the BPF 62. The BPF 62 supplies the A / D converter 63 through only the voltage signal in the band of approximately 1 kHz to 800 kHz among the input voltage Vb.

  The A / D converter 63 converts the analog input voltage Vb into a digital voltage signal Vb in accordance with a digital conversion instruction issued by the CPU 10, and then uses the digitized voltage signal Vb as current data Vb for each sampling period. Each frequency of the measurement signal Ia is stored in the sampling memory 64. Further, the sampling memory 64 is composed of SRAM, and sends digital current data Vb temporarily stored for each frequency of the measurement signal Ia to the CPU 10 in response to the request of the CPU 10.

The voltage detection circuit 7 includes a differential amplifier 71, a BPF (band pass filter) 72, an A / D converter 73, and a sampling memory 74. The differential amplifier 71 detects a voltage (potential difference) between the body electrode E of the subject, that is, the surface electrode Hp attached to the back Ha of the subject and the surface electrode Lp attached to the back Le. The BPF 72 supplies the voltage to the A / D converter 73 through only the voltage signal in the band of approximately 1 kHz to 800 kHz among the input voltage Vp.
The A / D converter 73 converts the analog input voltage Vp into a digital voltage signal Vp in accordance with a digital conversion instruction issued by the CPU 10, and then uses the digitized voltage signal Vp as voltage data Vp for each sampling period. Each frequency of the measurement signal Ia is stored in the sampling memory 74. Further, the sampling memory 74 is composed of an SRAM, and sends digital voltage data Vp temporarily stored for each frequency of the measurement signal Ia to the CPU 10 in response to a request from the CPU 10.
The CPU 10 issues digital conversion instructions to the two A / D converters 63 and 73 at the same timing.

The ROM 11 is a processing program for the CPU 10, in addition to the main program, for example, an impedance calculation subprogram, an impedance locus calculation subprogram, an admittance determination subprogram when the frequency is 0 and an infinite frequency (hereinafter, simply referred to as an admittance determination subprogram). , A bone removal weight / fat weight estimation subprogram, a bone weight estimation subprogram, a fat weight estimation subprogram, a body fat percentage estimation subprogram, and the like are stored.
Various programs are read from the ROM 11 to the CPU 10 and control the operation of the CPU 10. The recording medium for recording these subprograms is not limited to a semiconductor memory such as the ROM 11, but a magnetic memory such as a magnetic disk or a magnetic tape, a semiconductor memory such as a ROM or RAM, a magneto-optical memory such as a CD-ROM, It may be an optical memory or other recording medium.

Here, the impedance calculation subprogram sequentially reads out current data and voltage data for each frequency stored in the sampling memories 64 and 74 to the CPU 10 according to the algorithm described in the subprogram, and the subject's living body for each frequency. A process for calculating electrical impedance is executed.
As described in the “Prior Art” column, the cell membranes 2, 2,... Can be regarded as capacitors having a large capacity. Therefore, when the frequency of the externally applied current is low, a solid line A, FIG. As indicated by A,..., Only the extracellular fluid 3 flows. However, as the frequency increases, the current flowing through the cell membranes 2, 2,... Increases, and when the frequency becomes very high, as shown by broken lines B, B,. To flow through.

The impedance trajectory calculation subprogram, according to the algorithm described, causes the CPU 10 to calculate the frequency according to the calculation method of the least square method based on the bioelectric impedance of the subject for each frequency obtained under the control of the impedance calculation subprogram. Extrapolation is performed until 0 o'clock and when the frequency is infinite, and processing for calculating an impedance locus from frequency 0 to frequency infinite is executed.
In the “prior art” column, the human body tissue is represented by a simple electrical equivalent circuit (FIG. 13). However, in an actual human body tissue, cells of various sizes are irregularly arranged. The actual impedance trace of the human body is an arc whose center is higher than the real axis as shown by the solid line D in FIG. 3, and the electrical equivalent circuit has a time constant τ = Cmk / Yik distributed as shown in FIG. It is expressed by a distributed constant circuit. In the figure, 1 / Ye represents the extracellular fluid resistance, 1 / Yik represents the intracellular fluid resistance of each cell, and Cmk represents the cell membrane capacity of each cell.

  The admittance determination subprogram, based on the impedance locus obtained by the CPU 10 under the control of the impedance locus calculation subprogram, in accordance with the described algorithm, is the bioelectric admittance at frequency 0 (hereinafter also referred to as frequency 0 admittance). ) A process of determining Y (0) and bioelectric admittance (hereinafter also referred to as frequency infinity admittance) Y (∞) of the subject at infinite frequency is executed.

  Here, since the bioelectrical impedance at the frequency 0 is equivalent to the extracellular fluid resistance, the obtained frequency 0 admittance Y (0) is the reciprocal of the extracellular fluid resistance. At an infinite frequency, as shown in FIG. 5, the cell membrane loses its capacity, and the bioelectrical impedance is equivalent to the combined resistance of the intracellular fluid resistance and the extracellular fluid resistance (FIG. 5). Therefore, the admittance Y (∞) when the frequency is infinite is the reciprocal of the combined resistance of the intracellular fluid resistance and the extracellular fluid resistance.

  In accordance with the algorithm described in the subprogram, the deboning weight / fat weight estimation subprogram causes the CPU 10 to detect the frequency 0 admittance Y (0) obtained under the control of the admittance determination subprogram and the frequency infinity. Based on the admittance Y (∞) and the height H of the subject input via the keyboard 8, specifically, using the formula (3), “the bone weight and the fat weight are removed from the body weight of the subject. A process of estimating the bone removal weight / fat weight LEAN defined as “weight” is executed.

LEAN = aY (0) H ² + bY (∞) H ² + cH ³ + d (3)
Y (0): Extrapolated frequency admittance [1 / Ω]
Y (∞): Extrapolated frequency admittance
H: Subject's height [cm]
a, b, c, d: constants

Equation (3) is a multiple regression equation of deboning weight / fat weight LEAN obtained as a result of conducting a sample survey on a large number of subjects in advance, and the constants a, b, and c are divided by sex measured by DXA. The bone weight / fat weight LEAN was obtained by multiple regression analysis with three explanatory variables Y (0) H ² , Y (∞) H ² , H ³ . Is given by a = 98.26 ± 20%, b = 306.07 ± 20%, c = 0.629 ± 20%, and for women, a = 185.69 ± 20%, b = 169 .41 ± 20%, c = 0.030 ± 20%.
On the other hand, the constant d is obtained from the difference between the average value of the left side of Equation (3) and the average value of the sum of the first to third terms on the right side.

  Further, the bone weight estimation subprogram is based on the age AGE, height H, and weight W of the subject input to the CPU 10 via the keyboard 8 according to the algorithm described in the subprogram, specifically, A process for estimating the bone weight BMC of the subject is executed using Expression (4).

BMC = eAGE + fH + gW + h (4)
AGE: Subject's age H: Subject's height [cm]
W: Subject's weight [kg]
e, f, g, h: constant Equation (4) is a multiple regression equation of bone weight BMC obtained as a result of conducting a sample survey on a large number of subjects in advance, and constants e, f, g are measured by DXA. The bone weight BMC was determined by multiple regression analysis using three explanatory variables AGE, H, and W. For men, e = 2.13 ± 20%, f = 22.65 ± 20 %, G = 46.11 ± 20%, and for women, e = −11.37 ± 20%, f = 21.03 ± 20%, g = 20.98 ± 20% . On the other hand, the constant h is obtained from the difference between the average value of the left side of Expression (4) and the average value of the sum of the first to third terms on the right side.

  The fat weight estimation subprogram is obtained under the control of the subject's body weight W input to the CPU 10 via the keyboard 8 and the deboning weight / fat weight estimation subprogram according to the algorithm described in the subprogram. Based on the bone removal weight / fat weight weight LEAN and the bone weight BMC under the control of the bone weight estimation subprogram, specifically, the fat weight FAT of the subject is estimated using Equation (5) To execute the process.

FAT = W− (LEAN + BMC) (5)
W: Subject's weight [kg]
LEAN: Bone weight and fat weight [kg] of the subject's body
BMC: Bone weight of subject [kg]

  In addition, the body fat percentage estimation subprogram, according to the algorithm described in the subprogram, the fat weight FAT estimated under the control of the fat weight estimation subprogram and the body weight input via the keyboard 8 Based on W, specifically, the process of calculating the body fat percentage% FAT of the subject is executed using Expression (6).

    % FAT = 100 FAT / W (6)

  The data area of the RAM 12 further includes a bioelectrical impedance storage area for storing the bioelectrical impedance of the subject obtained by the impedance calculation subprogram for each frequency, an admittance Y at frequency 0 (0), and an admittance at infinite frequency. An admittance storage area for storing Y (∞), a height data storage area for storing subject height data input via the keyboard 8, and an age data storage for storing the age of the subject input via the keyboard 8 The fat weight obtained by the area, the body weight data storage area for storing the body weight of the subject input via the keyboard 8, the bone removal weight / fat weight storage area, the BMC storage area, and the fat weight estimation subprogram Obtained by fat weight storage area to be stored and body fat percentage estimation subprogram Body fat rate storage area for storing the body fat ratio is set.

  The display 9 is composed of, for example, a liquid crystal display panel capable of color display, and input data from the keyboard 8, for example, the height of the subject, calculation results of the CPU 10, for example, bone removal weight / fat weight, fat weight, body Displays fat percentage, impedance trajectory, etc.

Next, the operation of this example will be described.
First, prior to measurement, as shown in FIG. 2, the operator (or subject) has two surface electrodes Hc, Hp on the subject's back Ha and two surface electrodes Lp, Lc on the same side of the subject. Attached to each of the instep part Le, for example, through a conductive cream. Hereinafter, the surface electrode will be described. Among the surface electrodes to be attached, the surface electrodes Hc and Lc are used as electrodes in which the current inflow surface electrode and the current outflow surface electrode are switched for each period in the multi-frequency probe current. Hp and Lp are electrodes for extracting a voltage generated at a voltage detection site by applying a multi-frequency probe current from the surface electrodes Hc and Lc to the living body. When one of the two surface electrodes Hc, Lc becomes the inflow surface electrode, the surface electrodes Hp, Lp are attached to the living body surface portion that is upstream of the inflow surface electrode and close to the center of the human body. The other electrode in the charged state is an electrode that is attached to a living body surface portion that is downstream of the current outflow surface electrode and close to the center of the human body. An operator (or the subject himself / herself) operates the keyboard 8 of the body composition estimation device 4 to input the height, age, and weight of the subject and store them in the height data storage area, age data storage area, and weight data storage area of the RAM 12. Let

  Next, when the operator (or the subject himself / herself) turns on the start / end switch of the keyboard 8, the CPU 10 starts operation according to the processing flow shown in FIG. First, in step SP10, the CPU 10 supplies a signal generation instruction signal SG to the measurement signal generator 52 of the signal output circuit 5. When the measurement signal generator 52 receives the signal generation instruction signal SG from the CPU 10, the measurement signal generator 52 starts driving, and the frequency is in a range of 1 kHz to 400 kHz and a frequency of 15 kHz with a predetermined sweep period during the entire measurement time. Measurement signals Ia that change stepwise at frequency intervals are sequentially generated and input to the output buffer 53. The output buffer 53 maintains the input measurement signal Ia in a constant current state (a constant value in a range of 100 to 800 μA), and sequentially supplies the surface electrode Hc as a multi-frequency probe current Ib generated at different frequencies. Send it out. Thereby, the constant current Ib flows through the body E of the subject from the surface electrode Hc, and measurement is started.

  When the current Ib is supplied to the body E of the subject, the I / V converter 61 of the current detection circuit 6 detects the multi-frequency probe current Ib flowing between the limbs to which the surface electrodes Hc and Lc are attached, After being converted to an analog voltage signal Vb, it is supplied to the BPF 62. In the BPF 62, only the voltage signal component in the band of 1 kHz to 800 kHz is allowed to pass from the input voltage signal Vb and is supplied to the A / D converter 63. In the A / D converter 63, the supplied analog voltage signal Vb is converted into a digital voltage signal Vb and stored in the sampling memory 64 as current data Vb for each predetermined sampling period and for each frequency of the measurement signal Ia. Is done. In the sampling memory 64, the stored digital current data Vb is sent to the CPU 10 in response to a request from the CPU 10.

  On the other hand, in the differential amplifier 71 of the voltage detection circuit 7, the voltage Vp generated between the limbs to which the surface electrodes Hp and Lp are attached is detected and supplied to the BPF 72. In the BPF 72, only the voltage signal component in the band of 1 kHz to 800 kHz is allowed to pass from the input voltage signal Vp and supplied to the A / D converter 73. In the A / D converter 73, the supplied analog voltage signal Vp is converted into a digital voltage signal Vp and stored in the sampling memory 74 as voltage data Vp for each predetermined sampling period and for each frequency of the measurement signal Ia. Is done. In the sampling memory 74, the stored digital voltage data Vp is sent to the CPU 10 in response to a request from the CPU 10. The CPU 10 repeats the specified number of sweeps N for the number of sweeps of the probe current Ia.

  When the number of sweeps reaches the specified number, the CPU 10 performs control to stop the measurement, and then proceeds to step SP11 shown in FIG. 6. From this, first, the impedance calculation subprogram is started, and both sampling memories are started. Current data and voltage data for each frequency stored in 64 and 74 are sequentially read, and the bioelectric impedance (average value of N sweeps) of the subject for each frequency is calculated. The calculated bioelectric impedance is stored in the bioelectric impedance storage area of the RAM 12. The calculation of the bioelectrical impedance includes calculation of its components (resistance and reactance). Next, the CPU 10 activates the impedance trajectory calculation subprogram and uses the least square method based on the bioelectrical impedance of the subject and its components (resistance and reactance) for each frequency obtained by the impedance calculation subprogram. According to the curve fitting technique, an impedance locus from frequency 0 to frequency infinity is calculated. The impedance locus calculated in this way is an arc whose center is higher than the real axis, as shown in FIGS.

  Next, in accordance with the admittance determination subprogram, the CPU 10 determines the frequency 0 admittance Y (0) and the frequency infinity admittance Y (∞) based on the impedance locus obtained under the control of the impedance locus calculation subprogram. Ask for. The CPU 10 stores the obtained admittance Y at frequency 0 (0) and admittance Y (∞) at infinity in the admittance storage area of the RAM 12.

Next, under the control of the bone removal weight / fat weight estimation subprogram, the CPU 10 calculates the frequency 0 admittance Y (0) and the frequency infinity admittance Y (∞) calculated under the control of the admittance determination subprogram. ) And the height H of the subject input via the keyboard 8 are substituted into Equation (3) to estimate the subject's deboning weight / fat weight LEAN, and the subject's deboning weight / fat weight is estimated. LEAN is stored in the bone removal weight / fat weight storage area of the RAM 12 (step SP12).
FIG. 8 shows the correlation between the deboned weight / fat weight LEAN estimated in this way and the deboned weight / fat weight LEAN measured by DXA. The correlation coefficient r was 0.802.

  Further, under the control of the bone weight estimation subprogram, the CPU 10 substitutes the age AGE, the height H, and the body weight W input via the keyboard 8 into the equation (4) to calculate the bone weight BMC of the subject. The estimated bone weight BMC is stored in the BMC storage area of the RAM 12. FIG. 9 shows the correlation between the bone weight BMC estimated in this way and the bone weight BMC measured by DXA. The correlation coefficient r was 0.91.

  Next, under the control of the fat weight estimation subprogram, the CPU 10 executes the bone removal weight / fat weight estimation subprogram of the subject obtained by starting the bone removal weight / fat weight estimation subprogram, and the bone weight estimation subprogram. The bone weight BMC obtained by activation and the body weight W of the subject input via the keyboard 8 are substituted into the equation (5) to estimate the subject's fat weight, and the estimated fat weight is stored in the RAM 12. Is stored in the fat weight storage area (step SP13).

FIG. 10 shows the correlation (correlation coefficient r = 0.937) between the fat weight FAT estimated in this way and the fat weight FAT measured by DXA. For comparison, the correlation (correlation coefficient r = 0.937) between the fat weight FAT estimated using the conventional body composition estimation apparatus and the fat weight FAT measured by DXA is shown in FIG. Shown in As can be seen from comparison between FIG. 10 and FIG. 11, the correlation coefficient r for fat weight FAT is 0.918 in the conventional apparatus, whereas the correlation coefficient r according to the apparatus of this example is It improves to 0.937.
Further, the CPU 10 calculates the fat weight obtained by the fat weight estimation subprogram and the body weight W of the subject input through the keyboard 8 under the control of the body fat percentage estimation subprogram, using Equation (6). By substituting, the body fat percentage of the subject is estimated, and the estimated body fat percentage is stored in the body fat percentage storage area of the RAM 12.

Next, the CPU 10 proceeds to step SP14 and displays the estimated values obtained by the above-described processes, that is, the deboning weight / fat weight, bone weight, fat weight, body fat percentage, etc. of the subject on the display 9. Let Then, a series of processing ends.
Thus, according to the configuration of this example, when estimating the bone removal weight / fat weight LEAN, the term of the subject's height H to the third power is taken into account, so that the state of the subject's body fat can be more accurately determined. Can be estimated.
In addition, since the fat weight of the subject is determined in consideration of the bone weight BMC in which individual differences are recognized, the state of the body fat of the subject can be estimated more accurately.
The model of the above example assumes that the resistance of the cell membrane is infinite. This is based on the fact that the resistivity of the cell membrane is sufficiently large.

Certainly, this assumption can be correct for cells that have a nearly spherical cell shape, but this assumption does not necessarily work for elongated cells such as muscle fiber cells. Because the side surface of the fiber cell has a large surface area, for example, even if the resistivity of the cell membrane is large, it is considered that the resistance of the cell membrane on the side surface itself decreases (resistance = resistivity × film thickness / area). That is, the current passes through the cell membrane even at a frequency of zero. This is shown in FIG. A thick line indicates a current path at a frequency of 0, and a dotted line indicates a current path at an infinite frequency. FIG. 14 (a) shows a case of a normal cell, and when the frequency is 0, no current passes through the cell, so the resistance is exclusively of the extracellular fluid. On the other hand, FIG. 14 (b) shows the case of fiber cells, and even when the frequency is 0, the resistance of the cell membrane takes a finite value, and current flows through the cells. Synthesis with external liquid. FIG. 15 shows an equivalent circuit of bioelectrical impedance. The capacitor Cm is bypassed by the cell membrane resistance Rm with respect to the equivalent circuit shown in FIG. In the impedance meter of the type in which electrodes are attached to the limbs, the contribution to the impedance is overwhelmingly larger in the limbs (muscles of the limbs) than in the trunk. Therefore, in the above-described model, it is considered that the current flowing inside the cell at the frequency 0 is also regarded as the current flowing outside the cell, and the ECF is overestimated.
Therefore, a finite cell membrane resistance Rm is considered so that ECF and ICF are statistically 1: 2. Here, since the cell membrane resistance Rm is simple, there is only a gender difference and a fixed value is assumed.

The intracellular resistance Ricf and the extracellular resistance Recf can be derived according to the equation (7) based on the cell membrane resistance Rm, the electrical impedance Rinf at the infinite frequency, and the electrical impedance Ro at the frequency 0.
F = 1 / {Rinf ⁻¹ −Ro ⁻¹ }
Ricf = {- Rm + (Rm 2 + 4F) 0.5} / 2
Recf = 1 / {Rinf ⁻¹ −Ricf ⁻¹ } (7)
F: Intermediate variable The estimation formula of the deboning weight / fat weight LEAN is as shown in Expression (8).
LEAN = A · H ² / Recf + B · H ² / Ricf + C · H ^E (8)
H: Subject's height [cm]
A, B, C, E: Constant

Here, E is determined so that a constant term is not included in the LEAN estimation formula. If E is determined in this way, adaptability to a relatively small body such as a child can be achieved.
In Equation (8), when the first term is the extracellular fluid amount ECF and the second term is the intracellular fluid amount ICF, the body water amount TBW is the first term + the second term, but the extracellular fluid amount ECF If there is a component that is independent of the first term, the amount of extracellular fluid ECF will be underestimated. The same applies to the intracellular fluid amount ICF.
Therefore, it is assumed that moisture independent of these resistances is present in LEAN at a certain ratio, and that they are divided into ECF and ICF at a ratio of 1: 2. The ratio is determined so that the average of TBW (= ECF + ICF) / FFM (Fat Free Mass: lean body mass = body weight-fat weight) is 0.73. That is,

ECF [g] = A ・ H ² / Recf + D ・ LEAN
ICF [g] = B · H ² / Ricf + 2D · LEAN (9)
The constants A, B, C, D, E, and Rm are determined for each gender, and when the electrodes are attached to the wrist and ankle on the same side of the subject, 0.5 to 2 times the values shown in Table 5 are appropriate. It is.

The embodiment of the present invention has been described in detail with reference to the drawings. However, the specific configuration is not limited to this embodiment, and the present invention can be changed even if there is a design change or the like without departing from the gist of the present invention. include.
For example, numerical values such as regression coefficients of various body composition estimation formulas are not limited to those of the above-described embodiments, and can be changed as necessary.

Further, in the above-described embodiment, of the four surface electrodes Hc, Hp, Lc, and Lp, two surface electrodes Hc and Hp are provided on the back part Ha of the subject E, and the remaining two surface electrodes Lc and Lp are provided. Although it affixed on the instep part Le of the test subject E, it is not restricted to this, For example, you may make it attach to a right-and-left leg or a hand 2 each.
Further, the frequency range of the measurement signal (current) Ia is not limited to 1 kHz to 400 kHz. Similarly, in the case of the first embodiment, the number of frequencies is arbitrary as long as it is plural. In addition, the example of calculating the bioelectric admittance after calculating the bioelectric impedance has been shown, but the bioelectric admittance may be calculated directly. Calculation is used.

  Further, in the first embodiment described above, the bioelectrical impedance at the frequency of 0 and infinity is obtained using the curve fitting method by the least square method. When the influence of noise can be avoided by other means, for example, a probe current of two frequencies (a low frequency of 5 kHz or less and a high frequency of 200 kHz or more) is generated and input to the subject, and the subject's body is at a low frequency. The bioelectrical impedance / admittance of the subject may be regarded as the bioelectrical impedance / admittance at a frequency of 0, and the bioelectrical impedance / admittance at a high frequency of the subject's body may be regarded as the bioelectrical impedance / admittance at an infinite frequency. .

The output device is not limited to a display device, and a printer may be used.
Moreover, although the example in which the probe current is input to the living body by the constant current method has been described for the measurement of the bioelectric impedance in the above-described embodiment, the probe current may be input to the living body by the constant voltage method.
Further, a plurality of frequencies may be generated not only sequentially, but also generated simultaneously and separated by a filter or the like for measurement.

It is a block diagram which shows the electrical constitution of the body composition estimation apparatus which is one Example of this invention. It is a schematic diagram which shows typically the use condition of the body composition estimation apparatus. It is a figure which shows the impedance locus | trajectory of a human body. It is an electrical equivalent circuit diagram of a human body tissue. It is an electrical equivalent circuit diagram of the human body tissue when the frequency is infinite. It is a flowchart which shows the operation | movement process sequence of the body composition estimation apparatus. It is a figure which shows the example of a display of the impedance locus | trajectory of the indicator in the body composition estimation apparatus. FIG. 6 is a correlation diagram of the deboned weight / fat weight LEAN determined by DXA and the deboned weight / fat weight LEAN determined using the body composition estimation apparatus. It is a correlation diagram of the bone weight BMC calculated | required by DXA, and the bone weight BMC calculated | required using the body composition estimation apparatus. It is a correlation diagram of fat weight FAT calculated | required by DXA, and fat weight FAT calculated | required using the body composition estimation apparatus. It is a correlation diagram of fat weight FAT calculated | required using the conventional body composition estimation apparatus, and fat weight FAT calculated | required by DXA. It is the schematic diagram (the 1) which shows a structure | tissue of a biological body typically. It is the electrical equivalent circuit schematic (1) of a biological tissue. It is the schematic diagram (the 2) which shows a structure | tissue of a biological body typically. FIG. 2 is an electrical equivalent circuit diagram (part 2) of a living tissue.

Explanation of symbols

4 Body composition estimation device 5 Signal output circuit (part of bioelectric quantity measuring means)
52 Measurement Signal Generator 53 Output Buffer 6 Voltage Detection Circuit (Part of Bioelectric Measurement Unit)
61 I / V converter 62 BPF
63 A / D converter 64 Sampling memory 7 Voltage detection circuit (part of bioelectricity measuring means)
10 CPU (part of bioelectricity measurement means, electric admittance / electrical impedance calculation means, deboning weight / fat weight estimation means, fat weight / body fat ratio estimation means)
11 ROM
12 RAM
Hc, Hp, Lc, Lp Surface electrode E Subject's body Ha Subject's back part Le Subject's back part Ia Measurement signal Ib Multi-frequency probe current Vp Voltage between the subject's limbs

Claims (2)

A bioelectric quantity for generating a probe current of one or more frequencies for each frequency, and inputting the generated probe current into the body of a subject and measuring a voltage applied to the living body and a current flowing through the living body at that time Measuring means;
An electrical admittance / electrical impedance calculating means for calculating an electrical admittance or an electrical impedance at a frequency of infinity and a frequency of 0:00 of the subject's body based on a measurement result from the bioelectric quantity measuring means;
Using formula (1), the amount of extracellular fluid, the amount of intracellular fluid, or the body that estimates at least one of the amount of extracellular fluid ECF, the amount of intracellular fluid ICF, or the amount of body water TBW of the subject A body composition estimation device comprising a moisture amount estimation means.
F = Rm / {Rinf ⁻¹ −Ro ⁻¹ }
Ricf = {- Rm + (Rm 2 + 4F) 0.5} / 2
Recf = 1 / {Rinf ⁻¹ −Ricf ⁻¹ }
LEAN = A ・ H ² / Recf + B ・ H ² / Ricf + C ・ H ^E
ECF = A ・ H ² / Recf + D ・ LEAN
ICF = B ・ H ² / Ricf + 2D ・ LEAN
TBW = ECF + ICF (1)
Ricf: Resistance of intracellular fluid
Recf: Resistance of extracellular fluid
Rinf: Electrical impedance at infinite frequency
Ro: electrical impedance at frequency 0
LEAN: Bone weight and fat weight of the subject's body
H: Height of subject
A, B, C, D, E, Rm: constant (Rm corresponds to resistance of cell membrane)
F: Intermediate variable The constants A, B, C, D, E, and Rm are determined for each gender and are 0.5 to 2 times the values shown in Table 2 when the electrodes are attached to the wrist and ankle on the same side of the subject. The body composition estimation apparatus according to claim 1.

Images