Concentration calculation device and blood processing system

Abstract

The concentration calculation device of the present invention comprises: an irradiation unit, which is used to irradiate excitation light to the dialysate; a detection unit, which detects the fluorescence generated from the dialysate; a classification unit, which classifies the spectrum of the fluorescence detected by the detection unit into one of a plurality of predetermined groups; and a concentration calculation unit, which calculates the concentration of albumin contained in the dialysate based on a standard model corresponding to the group into which the fluorescence spectrum is classified by the classification unit, among a plurality of standard models pre-set corresponding to the plurality of groups, and the fluorescence intensity of a plurality of wavelengths in a predetermined wavelength range of the spectrum of the fluorescence detected by the detection unit.

Description

Technical Field

[0001] This application is based on Japanese Patent Application No. 2019-118622 filed on June 26, 2019 and Japanese Patent Application No. 2019-118624 filed on June 26, 2019, and the contents thereof are cited herein.

[0002] The invention relates to a concentration calculating device and a blood processing system. Background Art

[0003] Hemodialysis and hemodiafiltration are currently widely used treatment methods for patients with renal failure. In this treatment method, metabolites in the blood flowing into the hemodialyzer (filter) diffuse into the fresh dialysate flowing into the hemodialyzer through the dialysis membrane and are filtered and discharged.

[0004] Many patients with renal failure complain of symptoms such as pruritus, anxiety, and bone and joint pain. In order to improve such symptoms, α1-microglobulin (MG) is actively removed as an indicator substance. However, since the size of albumin required by the body is close to that of α1-MG, it is inevitable that albumin required by the body will leak into the dialysis fluid due to the active removal of α1-MG. Compared with dialysis treatment three times a week, the current general method of managing albumin is to measure the patient's serum albumin concentration 1 to 2 times a month to grasp the albumin level.

[0005] Under the circumstances described above, a method of grasping the concentration of albumin in the dialysis effluent in real time is required at the site of dialysis treatment.

[0006] As a method for measuring the concentration of a target substance such as albumin in a dialysate, Patent Documents 1 and 2 disclose a method of measuring the concentration of the substance based on fluorescence intensity by irradiating ultraviolet rays.

[0007] As a method for calculating the concentration of a target substance in a dialysate, Patent Document 3 discloses a method of irradiating ultraviolet light to calculate the concentration of the target substance based on the fluorescence intensity after eliminating the influence of impurities (such as indoxyl).

[0008] In addition, as a method for calculating the concentration of a target substance, Patent Document 4 discloses the following method: irradiating linearly polarized light, detecting the intensity of the emitted fluorescence on a first polarization plane, and detecting the intensity of the fluorescence on a second polarization plane different from the first polarization plane to determine the anisotropy of these fluorescences, and calculating the concentration of one side based on both the anisotropy and the intensity.

[0009] In addition, as a method for calculating albumin concentration, Patent Document 5 discloses the following method: after measuring the absorbance of the dialysis effluent, it is separated by a filter as an albumin separation unit, the absorbance is measured again, and the albumin concentration is calculated based on the change in absorbance before and after passing through the filter.

[0010] Prior art literature

[0011] Patent Literature

[0012] Patent Document 1: European Patent No. 2579910

[0013] Patent Document 2: European Patent No. 2579911

[0014] Patent Document 3: Japanese Patent Application No. 2014-518517

[0015] Patent Document 4: Japanese Patent Application No. 2015-521492

[0016] Patent Document 5: Japanese Patent Application Publication No. 2015-146837 Summary of the invention

[0017] Problem that the invention aims to solve

[0018] However, in the methods disclosed in Patent Documents 1 and 2, in the dialysis discharge containing various mixed substances, if the fluorescence intensities between the fluorescent substances are difficult to be additive, there will be a fluorescence resonance energy transfer phenomenon and substances that absorb the fluorescence of the target substance. Therefore, it cannot be said that the correlation between the fluorescence intensity and the concentration of the target substance is high, and the substance concentration cannot be measured with high accuracy at the excitation wavelength of 280nm in Patent Documents 1 and 2.

[0019] In the method disclosed in Patent Document 3, in the case of albumin, the fluorescence of albumin is generated by tryptophan residues, and tryptophan as an amino acid monomer is also present in the dialysis effluent. The fluorescence of albumin at excitation wavelengths of 280 nm and 295 nm is extremely similar to the fluorescence of tryptophan. In the method of Patent Document 3, linear superposition does not hold and separation cannot be performed. Therefore, the albumin concentration cannot be calculated with high accuracy.

[0020] In the fluorescence polarization method disclosed in Patent Document 4, anisotropy varies depending on the molecular weight of the fluorescent substance, and therefore, a rotating unit is provided to detect the fluorescence, which tends to make the device complicated and expensive. In addition, most proteins of fluorescent substances have at least one of tryptophan residues, tyrosine residues, and phenylalanine residues, and since low molecular weight proteins contained in the dialysis effluent have similar molecular weights, and their excitation wavelengths and fluorescence wavelengths are also similar, it is difficult to classify them based only on anisotropy.

[0021] Furthermore, in the method disclosed in Patent Document 5, since a filter must be used in the albumin separation section, in addition to the issues of the frequency of filter replacement and the price of the filter, there is a need for more accurate control of the filter flow rate / filtration flow rate, the complication of the device, etc., so it cannot be denied that there are many inappropriate aspects in terms of practical application.

[0022] The present invention has been made in view of the above-mentioned points, and one of its objects is to provide a concentration calculation device and a blood processing system that can calculate the concentration of a substance such as albumin with high accuracy without making the device complicated.

[0023] Solutions for solving problems

[0024] As a result of in-depth research, the inventors of the present invention found that the above-mentioned problem can be solved by classifying the specified factors that affect the standard model used for concentration calculation into one of a plurality of predetermined groups, and using the standard model corresponding to the group into which the specified factors are classified to calculate the concentration of the substance, thereby completing the present invention.

[0025] That is, the present invention includes the following aspects.

[0026] (1) A concentration calculation device, comprising: an irradiation unit for irradiating an object for concentration calculation with excitation light; a detection unit for detecting fluorescence generated from the object for concentration calculation; a classification unit for classifying specified factors affecting concentration calculation into one of a plurality of predetermined groups; and a concentration calculation unit for calculating the concentration of a substance contained in the object for concentration calculation based on a standard model corresponding to the group into which the specified factor is classified by the classification unit, among a plurality of standard models pre-set respectively corresponding to the plurality of groups, and the fluorescence intensity of a plurality of wavelengths in a predetermined wavelength range of the spectrum of the fluorescence detected by the detection unit.

[0027] (2) In the concentration calculation device described in (1), the element is a spectrum of the fluorescence detected by the detection unit.

[0028] (3) In the concentration calculation device described in (2), the plurality of groups are divided based on feature quantities obtained by performing a principal component analysis on the spectrum of the fluorescence.

[0029] (4) In the concentration calculation device described in any one of (1) to (3), the concentration calculation object is dialysate discharged during dialysis treatment.

[0030] (5) In the concentration calculation device described in (1), the object of concentration calculation is dialysate discharged during dialysis treatment, and the elements are any one or more of the spectrum of the fluorescence, the time elapsed since the start of the dialysis treatment, the patient who is the subject of the dialysis treatment, the treatment method for the patient, the type of filter of the dialyzer used in the dialysis treatment, the intensity of the fluorescence detected by the detection unit, the maximum value of the spectrum of the fluorescence within a specified wavelength range, and the setting parameters used for the detection of the detection unit.

[0031] (6) In the concentration calculation device described in (4) or (5), the substance is albumin in the dialysate.

[0032] (7) In the concentration calculation device described in any one of (4) to (6), parameter values ​​related to concentration calculation are set in such a manner that the relationship between the percentage error A of the substance concentration allowed in any time period in the first half of the dialysis treatment and the percentage error B of the measured concentration allowed in any time period in the second half of the dialysis treatment is A≤B.

[0033] (8) In the concentration calculation device described in any one of (4) to (7), the actual time t n The previous time t n-1 The estimated substance concentration C(t n ) The estimated difference E(t n ) divided by the actual time t expressed by (3) n-1 The estimated difference E(t n-1 )RMSD(t n-1 ) is the estimated differential rate EF(t n ),

[0034] When t<60 minutes of dialysis time, the estimated differential rate EF(t n ) satisfies -1.0≤EF(t n )≤1.0, the actual time t n The calculated substance concentration,

[0035] When the dialysis time is 60 minutes ≤ t < 120 minutes, the estimated differential rate EF (t n ) satisfies -1.2≤EF(t n )≤1.2, the actual time t n The calculated substance concentration,

[0036] When the dialysis time is 120 minutes ≤ t, the estimated differential rate EF (t n ) satisfies -1.5≤EF(t n )≤1.5, the actual time t n The calculated substance concentration,

[0037] C(t)=A×t^(-B)···(1)

[0038] E(t n )=(at actual time t n Calculated concentration)-C(t n )···(2)

[0039] RMSD(t n-1 )={[Σ(E(t i )^2)] / (n-1)}^0.5···(3)

[0040] EF(t n )=E(t n ) / RMSD(t n-1 )···(4)

[0041] Where C: estimated substance concentration in the dialysate [mg / dL], A: constant, t: time [min], B: constant, E: estimated difference, suffix i: the i-th calculated concentration, suffix n: the n-th calculated concentration.

[0042] (9) In the concentration calculation device described in any one of (1) to (8), there is also a standard model preparation unit, which prepares the multiple standard models by performing multivariate analysis on the fluorescence spectrum for each of the multiple groups.

[0043] (10) In the concentration calculation device described in (9), the multivariate analysis is any one of partial least squares (PLS) regression analysis, principal component regression analysis, multivariate regression analysis, support vector machine regression analysis, and machine learning analysis.

[0044] (11) The concentration calculation device described in any one of (1) to (10) further includes a storage unit configured to store a plurality of substance concentrations calculated by the concentration calculation unit at a plurality of times.

[0045] (12) The concentration calculation device described in (11) further includes an integral value calculation unit that calculates a time integral value of the product of each substance concentration corresponding to a plurality of time points stored in the storage unit and a flow rate of a concentration calculation object.

[0046] (13) In the concentration calculation device described in (11) or (12), there is also a change rate calculation unit, which calculates the change rate dC / dt of each substance concentration corresponding to multiple moments stored in the storage unit relative to time, where C is the substance concentration and t is the time.

[0047] (14) The concentration calculation device described in any one of (11) to (13) further includes a difference calculation unit that calculates the difference between the concentration of each substance corresponding to a plurality of time points stored in the storage unit and a predetermined value.

[0048] (15) In the concentration calculation device described in any one of (11) to (14), there is also an estimated value calculation unit, which calculates a time-dependent formula for the concentration of the substance based on the concentrations of each substance corresponding to multiple time points stored in the storage unit, and calculates an estimated value of the total amount of the substance contained in the concentration calculation object based on the time-dependent formula.

[0049] (16) The concentration calculation device according to any one of (1) to (15), further comprising a display unit that displays information on the substance concentration calculated by the concentration calculation unit.

[0050] (17) In the concentration calculation device described in any one of (1) to (16), the concentration calculation object irradiated with the excitation light is a continuous flow with a flow rate of 10 mL / min or more and 1000 mL / min or less.

[0051] (18) A blood processing system comprising the concentration calculation device according to any one of (1) to (17).

[0052] (19) A concentration calculation device comprises: an irradiation unit for irradiating an irradiation object containing albumin with excitation light; a detection unit for detecting fluorescence generated from the irradiation object; and a concentration calculation unit for calculating the concentration of albumin contained in the irradiation object based on the fluorescence intensity of multiple wavelengths in a predetermined wavelength range of the spectrum of the fluorescence detected by the detection unit and a preset standard model, wherein the excitation light includes light with a wavelength between 300 nm and 400 nm for exciting fluorescence that becomes a secondary peak of albumin.

[0053] (20) In the concentration calculation device described in (19), the excitation light includes light having a wavelength between 310 nm and 380 nm.

[0054] (21) In the concentration calculation device described in (19) or (20), the predetermined wavelength range of the spectrum of the fluorescence includes a range from a wavelength obtained by adding 10 nm to the lower limit of the wavelength of the excitation light to a wavelength obtained by adding 450 nm to the upper limit of the wavelength of the excitation light.

[0055] (22) In the concentration calculation device described in (19) or (20), the predetermined wavelength range of the spectrum of the fluorescence includes a range from a wavelength obtained by adding 10 nm to the lower limit of the wavelength of the excitation light to a wavelength obtained by adding 400 nm to the upper limit of the wavelength of the excitation light.

[0056] (23) In the concentration calculation device described in any one of (19) to (22), it also includes a standard model preparation unit, which prepares the standard model by performing multivariate analysis.

[0057] (24) In the concentration calculation device described in (23), the multivariate analysis is any one of partial least squares (PLS) regression analysis, principal component regression analysis, multivariate regression analysis, support vector machine regression analysis and machine learning analysis.

[0058] (25) In the concentration calculation device described in any one of (19) to (24), the concentration of albumin is calculated by the following formula:

[0059] C=a1×I1+a2×I2+···+a n ×I n +K

[0060] Where, C: calculated concentration of albumin, a n : Coefficient, I n : fluorescence intensity, K: constant, suffix n: a natural number assigned to each wavelength of the spectrum obtained by fluorescence from shortest to longest wavelength.

[0061] (26) In the concentration calculation device described in any one of (19) to (25), the irradiation object is dialysate discharged during dialysis treatment.

[0062] (27) The concentration calculation device described in (26) further includes a storage unit for storing a plurality of albumin concentrations calculated by the concentration calculation unit during dialysis treatment.

[0063] (28) The concentration calculation device described in (27) further includes an integral value calculation unit that calculates an integral value of the product of the albumin concentration and the dialysate discharge flow rate stored in the storage unit since the start of the dialysis treatment.

[0064] (29) The concentration calculation device described in (27) or (28) further comprises a change rate calculation unit, which calculates the change rate dC / dt of the albumin concentration stored in the storage unit with respect to time, wherein C is the albumin concentration and t is the time.

[0065] (30) The concentration calculation device according to any one of (27) to (29), further comprising a difference calculation unit that calculates a difference between the albumin concentration stored in the storage unit and a predetermined value.

[0066] (31) In the concentration calculation device described in any one of (27) to (30), there is also an estimated value calculation unit, which calculates a time-dependent formula for the albumin concentration after the start of the dialysis treatment based on the albumin concentration stored in the storage unit, and calculates an estimated value of the total amount of albumin in the dialysate from the start of the dialysis treatment to the end of the dialysis treatment based on the time-dependent formula.

[0067] (32) The concentration calculation device according to any one of (26) to (31), further comprising a display unit that displays information on the concentration of albumin calculated by the concentration calculation unit.

[0068] (33) In the concentration calculation device described in any one of (26) to (32), parameter values ​​related to concentration calculation are set in such a manner that the relationship between the percentage error A of the albumin concentration allowed in any time period in the first half of the dialysis treatment and the percentage error B of the measured concentration allowed in any time period in the second half of the dialysis treatment is A≤B.

[0069] (34) In the concentration calculation device described in any one of (26) to (33), the actual time t of the albumin concentration in the dialysate is used. n The previous time t n-1 The estimated albumin concentration C(t n ) The estimated difference E(t n ) divided by the actual time t expressed by (3) n-1 The estimated difference E(t n-1 )RMSD(t n-1 ) is the estimated differential rate EF(t n ),

[0070] When t<60 minutes of dialysis time, the estimated differential rate EF(t n) satisfies -1.0≤EF(t n )≤1.0, the actual time t n The calculated albumin concentration,

[0071] When the dialysis time is 60 minutes ≤ t < 120 minutes, the estimated differential rate EF (t n ) satisfies -1.2≤EF(t n )≤1.2, the actual time t n The calculated albumin concentration,

[0072] When the dialysis time is 120 minutes ≤ t, the estimated differential rate EF (t n ) satisfies -1.5≤EF(t n )≤1.5, the actual time t n The calculated albumin concentration,

[0073] C(t)=A×t^(-B)···(1)

[0074] E(t n )=(at actual time t n Calculated concentration)-C(t n )···(2)

[0075] RMSD(t n-1 )={[Σ(E(t i )^2)] / (n-1)}^0.5···(3)

[0076] EF(t n )=E(t n ) / RMSD(t n-1 )···(4)

[0077] Where, C: estimated albumin concentration in the dialysate [mg / dL], A: constant, t: time [min], B: constant, E: estimated difference, suffix i: the i-th calculated concentration, suffix n: the n-th calculated concentration.

[0078] (35) In the concentration calculation device described in any one of (26) to (34), the dialysate irradiated with the excitation light is a continuous flow with a flow rate of 10 mL / min or more and 1000 mL / min or less.

[0079] (36) A blood processing system comprising the concentration calculation device described in any one of (19) to (35).

[0080] Effects of the Invention

[0081] According to the present invention, it is possible to provide a concentration calculation device and a blood processing system capable of calculating the concentration of a substance with high accuracy without making the device complicated. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] Figure 1 It is an explanatory diagram showing an example of the configuration of the dialysis system in the first embodiment.

[0083] Figure 2 This is an explanatory diagram showing an example of the structure of an optical monitor.

[0084] Figure 3 This is a graph showing an excitation wavelength at which the fluorescence characteristic of albumin exhibits a secondary peak.

[0085] Figure 4 A diagram for explaining classification of a plurality of groups.

[0086] Figure 5 is a block diagram of a concentration calculation device.

[0087] Figure 6 This is a block diagram of the numerical calculation unit.

[0088] Figure 7 It is a graph for explaining the integrated value of the leakage amount of albumin.

[0089] Figure 8 It is a graph for explaining the rate of change (dC / dt) of albumin concentration with respect to time.

[0090] Fig. 9 It is a graph for explaining the difference between the albumin concentration and a predetermined value.

[0091] Fig.10 This is a graph for explaining the estimated curve of the albumin leakage amount.

[0092] Fig.11 It is a graph showing the wavelength of the fluorescence spectrum.

[0093] Fig.12 It is a graph for explaining the error in the calculated concentration of albumin.

[0094] Fig.13 This is a graph showing changes in albumin concentration in general dialysis effluent.

[0095] Fig.14 It is a graph showing the fluorescence spectrum.

[0096] Fig.15 It is a graph showing the calculated concentration and the measured concentration of albumin.

[0097] Fig.16This is a diagram showing the results of principal component analysis of the fluorescence spectrum.

[0098] Fig.17 It is an explanatory diagram showing an example of the configuration of the dialysis system in the second embodiment.

[0099] Fig.18 This is an explanatory diagram showing an example of the structure of an optical monitor.

[0100] Fig.19 This is a graph showing an excitation wavelength at which the fluorescence characteristic of albumin exhibits a secondary peak.

[0101] Fig. 20 is a block diagram of a concentration calculation device.

[0102] Fig.21 This is a block diagram of the numerical calculation unit.

[0103] Fig. 22 It is a graph for explaining the integrated value of the leakage amount of albumin.

[0104] Fig.23 It is a graph for explaining the rate of change (dC / dt) of albumin concentration with respect to time.

[0105] Fig.24 It is a graph for explaining the difference between the albumin concentration and a predetermined value.

[0106] Fig.25 This is a graph for explaining the estimated curve of the albumin leakage amount.

[0107] Fig.26 It is a graph showing the wavelength of the fluorescence spectrum.

[0108] Fig. 27 : is a table showing the fluorescence wavelength range, correlation coefficient, and root mean square error.

[0109] Fig.28 It is a graph for explaining the error in the calculated concentration of albumin.

[0110] Fig.29 This is a graph showing changes in albumin concentration in general dialysis effluent.

[0111] Fig.30 It is a graph showing the fluorescence spectrum.

[0112] Fig.31 Graph showing the actual concentration and estimated concentration of albumin in Example 1.

[0113] Fig.32 It is a graph showing the actual concentration and estimated concentration of albumin in Comparative Example.

[0114] Fig.33 is a table showing the excitation wavelength and the root mean square error.

[0115] Fig.34 It is a graph showing the change of albumin concentration over time. DETAILED DESCRIPTION

[0116] Below, an example of a preferred embodiment of the present invention is described with reference to the accompanying drawings. In addition, as long as there is no special description, the positional relationship of the upper, lower, left, right, etc. of the accompanying drawings is assumed to be a positional relationship based on the positional relationship shown in the accompanying drawings. The dimensional ratios of the accompanying drawings are not limited to the ratios shown in the drawings. In addition, the following embodiments are examples for illustrating the present invention and are not intended to limit the present invention to only these embodiments. In addition, the present invention can be subjected to various modifications as long as it does not deviate from its main purpose.

[0117] First embodiment

[0118] <Dialysis System>

[0119] Figure 1 1 is an explanatory diagram schematically showing the configuration of a dialysis system 1 as a blood processing system equipped with the concentration calculation device according to the present embodiment.

[0120] The dialysis system 1 includes, for example, a dialyzer 10 , a blood circuit 11 , a dialysate circuit 12 , a drainage circuit 13 , a fluid replacement circuit 14 , a control device 15 , a concentration calculation device 16 , and the like.

[0121] The dialyzer 10 is, for example, a hollow fiber module with a hollow fiber membrane built therein, which can separate useless components from the blood. The dialyzer 10 has a cylindrical container 20, and a plurality of hollow fiber membranes 21 are arranged inside the cylindrical container 20 along the long side direction of the cylindrical container 20. The hollow fiber membrane 21 can separate useless components from the blood. An inlet 22 and an outlet 23 connected to the inner space (blood side) of the hollow fiber membrane 21 are provided at the upper and lower parts of the cylindrical container 20, and two inlets and outlets 24, 25 connected to the outer space (dialysis fluid side) of the hollow fiber membrane 21 are provided at the side part of the cylindrical container 20.

[0122] The blood circuit 11 includes, for example, a blood removal line 31 connecting the blood removal unit 30 and the dialyzer 10, and a blood return line 33 connecting the dialyzer 10 and the blood return unit 32. The blood removal line 31 and the blood return line 33 are mainly composed of soft tubes. The blood removal line 31 is connected to the inlet 22 of the dialyzer 10, and the blood return line 33 is connected to the outlet 23 of the dialyzer 10.

[0123] For example, a blood pump 40 is provided on the blood vessel removal line 31. In addition, a dripping funnel 41 is connected to the blood vessel removal line 31. The dripping funnel 41 may not be provided in some cases.

[0124] The dialysate circuit 12 is connected from a dialysate supply source (not shown) to an inlet and outlet 25 of the dialyzer 10. The drainage circuit 13 is connected from an inlet and outlet 24 of the dialyzer 10 to a drainage unit (not shown). The dialysate circuit 12 and the drainage circuit 13 are provided with a supply and drainage pump (not shown) for supplying the dialysate to the dialyzer 10 through the dialysate circuit 12 and for discharging the dialysate from the dialyzer 10 through the drainage circuit 13.

[0125] The fluid replacement circuit 14 is connected to the dripping funnel 41 (blood circuit 11) from the dialysate circuit 12, for example. In the absence of the dripping funnel 41, the fluid replacement circuit 14 is directly connected to the blood removal line 31. The fluid replacement circuit 14 is provided with a fluid replacement pump 50.

[0126] The control device 15 is, for example, a computer, and can control the operation of the blood pump 40 and the infusion pump 50 by, for example, executing a program stored in a storage unit by a CPU, thereby performing a dialysis treatment for dialysis treatment. In addition, the communication between the control device 15 and various devices (blood pump 40, infusion pump 50) can be performed by a wired method such as a communication cable, or by a wireless method such as Bluetooth (registered trademark).

[0127] In dialysis treatment, in the blood circuit 11, the patient's blood is transported from the blood removal section 30 to the tube space of the hollow fiber membrane 21 of the dialyzer 10, and returns to the patient from the blood return section 32 after passing through the dialyzer 10. At this time, the dialysate is transported to the tube space outside the hollow fiber membrane 21 of the dialyzer 10 through the dialysate circuit 12, and then discharged through the drainage circuit 13. In the dialyzer 10, the blood flowing in the tube space of the hollow fiber membrane 21 is mainly useless components that flow out to the tube space outside (dialysate side) through the hollow fiber membrane 21 and are discharged together with the dialysate. The infusion (dialysate) of the dialysate circuit 12 is supplied to the blood circuit 11 through the infusion circuit 14 to supplement the blood with the specified components. In addition, there are hemodialysis filtration with supplementation of infusion and hemodialysis without supplementation of infusion.

[0128] <Concentration Calculation Device>

[0129] The concentration calculation device 16 includes an optical monitor 70 for irradiating excitation light to a dialysate as a concentration calculation object and detecting fluorescence generated from the dialysate; a classification unit 71 for classifying the spectrum of fluorescence acquired by the optical monitor 70 as a predetermined factor affecting the concentration calculation into one of a plurality of predetermined groups; a concentration calculation unit 72 for calculating the concentration of albumin as a substance contained in the dialysate based on a standard model corresponding to the group into which the spectrum of fluorescence is classified by the classification unit 71 and the fluorescence intensity of a plurality of wavelengths in a predetermined wavelength range of the spectrum of fluorescence detected by the optical monitor 70; a display unit 73; and an input unit 74. In addition, the communication among the optical monitor 70, the classification unit 71, the concentration calculation unit 72, the display unit 73, and the input unit 74 may be performed by a wired method such as a communication cable or by a wireless method such as Bluetooth (registered trademark).

[0130] For example Figure 2 As shown, the optical monitor 70 includes an irradiation unit 80 for irradiating the dialysate in the pipe of the drainage circuit 13 with excitation light, and a detection unit 81 for detecting fluorescence generated from the dialysate.

[0131] The irradiation unit 80 can irradiate Figure 3 The light having a wavelength of 300 nm to 400 nm is used to excite the fluorescence of albumin to form a secondary peak. The light source of the irradiation unit 80 is not particularly limited, but may be, for example, a halogen lamp, a xenon lamp, a deuterium lamp, an LED, etc. The wavelength of the excitation light irradiated from the irradiation unit 80 is preferably 310 nm to 380 nm, and more preferably 330 nm to 350 nm.

[0132] The detection unit 81 detects the fluorescence generated from the dialysis drainage to obtain a fluorescence spectrum. The detection unit 81 can detect fluorescence in a wavelength range of 310 nm to 850 nm. The fluorescence wavelength detected by the detection unit 81 is preferably 310 nm to 700 nm, more preferably 320 nm to 650 nm. In addition, the irradiation unit 80 and the detection unit 81 are, for example, arranged in directions that are perpendicular to each other with respect to the piping of the drainage circuit 13, and the fluorescence is obtained at an angle that is perpendicular to the excitation light. In addition, the configuration of the irradiation unit 80 and the detection unit 81, and the shape of the drainage circuit 13 are not limited to this.

[0133] The classification unit 71 is, for example, a computer, and classifies the spectrum of fluorescence detected by the detection unit 81 into one of a plurality of predetermined groups. Figure 4As shown, the division is based on the feature quantity obtained by principal component analysis of the fluorescence spectrum in advance, and the multiple groups are divided into four groups A, B, C, and D according to the value of the first component (i component) and the value of the second component (k component) obtained by principal component analysis, wherein the first component (i component) is the component factor that best represents the characteristics of the fluorescence spectrum, and the second component (k component) is the component factor that secondly represents the characteristics of the fluorescence spectrum. Specifically, in the figure with the first component as the horizontal axis and the second component as the vertical axis, the case where the first component and the second component are positive is set as group A, the case where the first component is negative and the second component is positive is set as group B, the case where the first component is positive and the second component is negative is set as group C, and the case where the first component and the second component are negative is set as group D. In addition, the classification unit 71 may not be assembled in the main body of the concentration calculation device 16. It is also possible to communicate with respect to the fluorescence spectrum data detected by the detection unit 81, to classify on the cloud, and to communicate to the concentration calculation unit 72.

[0134] Figure 5 , a block diagram of the concentration calculation unit 72 is shown in FIG. The concentration calculation unit 72 is, for example, a computer. The concentration calculation unit 72 includes: a concentration calculation unit 90, which calculates the concentration of albumin in the dialysis discharge fluid using, for example, a standard model corresponding to the group into which the spectrum of fluorescence is classified by the classification unit 71, among a plurality of standard models pre-set corresponding to a plurality of groups A to D, and the fluorescence intensity of a plurality of wavelengths in a predetermined wavelength range of the spectrum of fluorescence acquired by the optical monitor 70; a storage unit 91, which stores the albumin concentration obtained by the concentration calculation unit 90; a numerical calculation unit 92, which uses the albumin concentration stored in the storage unit 91 to calculate various numerical values ​​described later; and a standard model preparation unit 93, which prepares a plurality of standard models for each of the groups A to D.

[0135] The standard model preparation unit 93 obtains information on the fluorescence spectrum of each group A to D from the optical monitor 70 in advance, and performs multivariate analysis on the information to prepare standard models Ma, Mb, Mc, and Md corresponding to each group A to D. The multivariate analysis used to prepare the standard models Ma to Md can also be any one of partial least squares (PLS) regression analysis, principal component regression analysis, multivariate regression analysis, support vector machine regression analysis, and machine learning analysis. An example of the standard models Ma to Md is shown below. The standard models Ma to Md can be expressed as a formula for calculating the concentration C by multiplying the fluorescence intensity of each wavelength of the obtained fluorescence spectrum by a coefficient and taking their sum. Specifically, the standard models Ma to Md are expressed as formula (5).

[0136] C=a1×I1+a2×I2+···+a n ×I n +K···(5)

[0137] Where, C: calculated concentration of albumin, a n : Coefficient, I n : fluorescence intensity, K: constant, suffix n: a natural number assigned to each wavelength of the acquired fluorescence spectrum in ascending order of wavelength (e.g., if the wavelength is 300nm, 310nm, 320nm,..., 400nm, the suffixes are 1, 2, 3,..., 11).

[0138] The standard models Ma to Md produced by the standard model production unit 93 are stored in the storage unit 91, for example, and are used as parameters in the calculation process in the concentration calculation unit 90 and the numerical calculation unit 92. Alternatively, a pre-calculated standard model may be stored in the storage unit 91 and used.

[0139] The concentration calculation unit 90 obtains the fluorescence spectrum of the dialysate from the optical monitor 70 during the actual dialysis treatment, and calculates the albumin concentration using the fluorescence intensities of multiple wavelengths in a predetermined wavelength range of the fluorescence spectrum and the standard models Ma to Md corresponding to the groups A to D classified by the classification unit 71.

[0140] The numerical calculation unit 92 obtains the albumin concentration value stored in the storage unit 91 to calculate various numerical values ​​during the actual time of the dialysis treatment. Figure 6 As shown, the numerical calculation unit 92 includes: an integral value calculation unit 100, which calculates, for example, an integral value of the product of the albumin concentration and the dialysate discharge flow rate since the start of the dialysis treatment; a change rate calculation unit 101, which calculates the change rate of the albumin concentration; a difference calculation unit 102, which calculates the difference between the albumin concentration and a predetermined value; and an estimated value calculation unit 103, which calculates an estimated value of the total amount of albumin in the dialysate from the start of the dialysis treatment to the end of the dialysis treatment.

[0141] like Figure 7 As shown, the integral value calculation unit 100 calculates the albumin leakage at that time based on the product of the albumin concentration value and the dialysis drainage flow rate since the start of the dialysis treatment, and calculates the integral value of the albumin leakage based on the time curve of the albumin leakage. The integral value is the albumin leakage from the start of the dialysis treatment to that time point ( Figure 7 of the shading).

[0142] like Figure 8 As shown, the change rate calculation unit 101 makes the slope of the time change curve of the albumin concentration the change rate of albumin with respect to time (dC / dt).

[0143] like Fig. 9As shown, the other predetermined value in the difference calculation unit 102 is, for example, an albumin concentration calculated based on a preset allowable albumin leakage amount, and the albumin concentration becomes a threshold value for indicating excessive albumin leakage when the albumin concentration is exceeded. In this case, the difference calculation unit 102 calculates the difference between the albumin concentration and the threshold value.

[0144] The estimated value calculation unit 103 first obtains A and B of the formula shown in formula (1) based on at least three or more calculation points of albumin concentration. Fig.10 The albumin leakage estimation curve is shown in FIG. Then, the albumin leakage after the start of the dialysis treatment is estimated based on the formula (1), and the estimated value of the total amount of albumin (total leakage) per treatment is calculated during the dialysis treatment.

[0145] C(t)=A×t^(-B)···(1)

[0146] Where, C: concentration in dialysis effluent [mg / dL], A: constant, t: time [min], B: constant

[0147] Figure 5 The display unit 73 shown is, for example, a flat panel display, and displays the albumin concentration calculated by the concentration calculation unit 90, various numerical values ​​obtained by the numerical value calculation unit 92, or various warnings. For example, a warning is issued when the albumin concentration calculated by the concentration calculation unit 90 exceeds a prescribed range, when the integral value of the albumin leakage calculated by the integral value calculation unit 100 exceeds a prescribed range, when the rate of change of the albumin concentration calculated by the rate of change calculation unit 101 exceeds a prescribed range, when the difference between the albumin concentration calculated by the difference calculation unit 102 and the threshold value thereof exceeds a prescribed range, when the estimated value of the total amount of albumin per treatment calculated by the estimated value calculation unit 103 exceeds a prescribed range, etc.

[0148] The input unit 74 has the following functions: it is used to input from the outside information required for the classification performed by the classification unit 71, the calculation of the albumin concentration performed by the concentration calculation unit 72, the preparation of the standard model, and the information required for various numerical calculations. In addition, the above-mentioned control device 15, the classification unit 71, the concentration calculation unit 72, the display unit 73, and the input unit 74 can also be methods and functions realized by the same computer.

[0149] Next, the operation of the concentration calculation device 16 configured as above will be described.

[0150] The concentration calculation device 16 continuously or intermittently calculates the albumin concentration in the dialysis effluent of the effluent circuit 13 in real time during the dialysis treatment. The dialysis effluent is a continuous flow with a flow rate of 10 mL / min or more and 1000 mL / min or less.

[0151] Specifically, the irradiation unit 80 of the optical monitor 70 irradiates the drainage circuit 13 with excitation light, and the detection unit 81 detects fluorescence generated from the dialysis drainage. The excitation light irradiated from the irradiation unit 80 includes light having a wavelength of 300 nm to 400 nm that excites fluorescence of albumin as a secondary peak.

[0152] like Fig.11 As shown, the fluorescence spectrum detected by the detection unit 81 is, for example, in the wavelength range of 310 nm to 850 nm. In the detection unit 81, the parameters related to the measurement are adjusted based on the value of the maximum intensity value of the wavelength range of the fluorescence spectrum obtained between 380 nm and 480 nm, and the fluorescence spectrum is measured within the upper limit of the measurement of the sensor of the detection unit 81.

[0153] Next, the classification unit 71 classifies the spectrum of the fluorescence detected by the detection unit 81 into one of the predetermined plurality of groups A to D. Specifically, principal component analysis is performed on the spectrum of the fluorescence of the detection unit 81 to obtain the values ​​of the first component (i component) and the second component (k component), and the fluorescence spectrum of the detection unit 81 is classified into a group among the groups A to D that matches the values ​​of the first component (i component) and the second component (k component).

[0154] The concentration calculation unit 90 of the concentration calculation unit 72 calculates the concentration of albumin contained in the dialysate based on the standard model corresponding to the group into which the fluorescence spectrum is classified by the classification unit 71, among the four standard models Ma to Md pre-set corresponding to the four groups A to D, and the fluorescence intensity of multiple wavelengths in a predetermined wavelength range of the spectrum of fluorescence detected by the detection unit 81. For example, when the fluorescence spectrum of the detection unit 81 is classified into group A, the concentration calculation unit 90 calculates the concentration of albumin using the standard model Ma stored in the storage unit 91. The concentration calculation unit 90 continuously or intermittently calculates the albumin concentration in the dialysate during the actual time of the dialysis treatment. The albumin concentration calculated by the concentration calculation unit 90 is stored in the storage unit 91 each time. The albumin concentration is displayed on the display unit 73, for example, in real time.

[0155] Before the start of dialysis treatment, the standard models Ma to Md are prepared in advance in the standard model preparation unit 93 and stored in the storage unit 91. The standard model preparation unit 93 obtains fluorescence spectra of dialysis effluents with different albumin concentrations from the optical monitor 70, for example, and prepares each standard model Ma to Md by multivariate analysis.

[0156] The numerical calculation unit 92 obtains a plurality of albumin concentrations at each time calculated by the concentration calculation unit 90 from the storage unit 91 to calculate various numerical values. The integral value calculation unit 100 calculates the time integral value of the product of the albumin concentration and the dialysate discharge flow rate since the start of the dialysis treatment, the change rate calculation unit 101 calculates the change rate of the albumin concentration, the difference calculation unit 102 calculates the difference between the albumin concentration and a predetermined value, and the estimated value calculation unit 103 calculates the estimated value of the total amount of albumin in the dialysate discharge from the start of the dialysis treatment to the end of the dialysis treatment. These values ​​are displayed on the display unit 73, for example. In addition, when the albumin concentration, the integral value of the albumin leakage amount, the change rate of the albumin concentration, the difference between the albumin concentration and the threshold value, and the estimated amount of albumin leakage in one treatment are not within an appropriate range, a warning is displayed on the display unit 73.

[0157] According to the present embodiment, the classifier 71 classifies the spectrum of fluorescence detected by the detector 81 into one of the plurality of predetermined groups A to D, and the concentration calculator 72 calculates the concentration of albumin contained in the dialysis discharge fluid based on the standard model corresponding to the group into which the fluorescence spectrum is classified by the classifier 71 among the plurality of standard models Ma to Md pre-set corresponding to the plurality of groups A to D, and the fluorescence intensity of the plurality of wavelengths in the predetermined wavelength range of the spectrum of fluorescence detected by the detector 81. As a result, the concentration of albumin can be calculated with high accuracy without complicating the device.

[0158] The plurality of groups A to D are divided based on the characteristic quantities obtained by principal component analysis of the fluorescence spectrum, so that the concentration of albumin can be calculated with high accuracy. In addition, in the present embodiment, as the characteristic quantities obtained by principal component analysis of the fluorescence spectrum, Figure 4 As shown, the first component factor (i component) and the second component factor (k component) are used, but as long as the characteristic quantities of the fluorescence spectrum of the obtained dialysate can be classified, the first component factor and the second component factor can be replaced by other component factors obtained by principal component analysis.

[0159] The concentration calculation device 16 further includes a standard model creation unit 93 that creates standard models Ma to Md by performing multivariate analysis on the fluorescence spectrum for each of the plurality of groups A to D. This makes it possible to easily create highly reliable standard models Ma to Md.

[0160] The multivariate analysis is any one of partial least squares (PLS) regression analysis, principal component regression analysis, multiple regression analysis, support vector machine regression analysis, and machine learning analysis, and therefore standard models Ma to Md can be easily and accurately prepared.

[0161] The concentration calculation device 16 includes a storage unit 91 for storing a plurality of albumin concentrations calculated at a plurality of times after the start of the dialysis treatment. Therefore, for example, the calculated albumin concentrations can be used to perform various numerical calculations to analyze the status of the dialysis treatment.

[0162] The concentration calculation device 16 includes an integral value calculation unit 100 that calculates the time integral value of the product of each albumin concentration corresponding to a plurality of time points stored in the storage unit 91 and the flow rate of the dialysis effluent since the start of the dialysis treatment. The integral value represents the amount of albumin leakage at that time point, so that the amount of albumin leakage can be compared with a preset albumin permissible leakage amount during the dialysis treatment, thereby determining whether the amount of albumin leakage is appropriate.

[0163] The concentration calculation device 16 includes a change rate calculation unit 101 that calculates the change rate of each albumin concentration corresponding to a plurality of time points stored in the storage unit 91. Figure 8 As shown in Example A, where the absolute value of the rate of change of the initial albumin with respect to time (dC / dt) is large and changes rapidly to a low value, it can be understood that the clogging of the hollow fiber membrane during the dialysis treatment is progressing rapidly.

[0164] The concentration calculation device 16 includes a difference calculation unit 102 that calculates the difference between each albumin concentration corresponding to a plurality of time points stored in the storage unit 91 and a predetermined value. Fig. 9 As shown, when the difference between the calculated concentration and the predetermined value becomes large, it can be understood that the amount of albumin leakage is larger than expected.

[0165] The concentration calculation device 16 includes an estimated value calculation unit 103, which calculates a temporal expression of the albumin concentration after the start of the dialysis treatment based on the albumin concentrations corresponding to the plurality of time points stored in the storage unit 91, and calculates an estimated value of the total amount of albumin in the dialysate from the start of the dialysis treatment to the end of the dialysis treatment based on the temporal expression. Thus, the estimated total leakage amount of albumin can be compared with the preset allowable total leakage amount of albumin during the dialysis treatment, thereby determining whether the dialysis treatment is appropriate.

[0166] The concentration calculation device 16 includes a display unit 73 that displays information on the concentration of albumin calculated by the concentration calculation unit 72 . Therefore, the user can understand the concentration of albumin in real time during dialysis treatment.

[0167] The dialysis drainage irradiated with the excitation light is a continuous flow with a flow rate of more than 10mL / min and less than 1000mL / min. In the clinical field of dialysis treatment, there are many equipments such as dialysis devices and beds, and there is basically no space for setting up an optical monitor. Therefore, it is preferred that the optical monitor does not need equipment for storing dialysis drainage, equipment for separation such as membranes or chromatography, and equipment for obtaining a part of the dialysis drainage using a pump. Therefore, it is preferred that the optical monitor 70 is directly connected to the drainage circuit 13 or connected to the optical monitor 70 on the branch of the drainage circuit 13 to measure with a continuous flow, and it is preferred that the flow rate of the continuous flow is more than 10mL / min and less than 1000mL / min.

[0168] In the above embodiment, the classification unit 71 divides the fluorescence spectrum acquired from the detector 81 into a plurality of groups A to D based on the feature quantity obtained by principal component analysis. However, in addition to principal component analysis, the classification method may also use factor analysis, discriminant analysis, cluster analysis, correspondence analysis, KNN analysis, SIMCA analysis, PLS-DA, etc. In addition, the concentration calculation device 16 may also include a group creation unit that creates a plurality of groups related to the fluorescence spectrum that affects the concentration calculation.

[0169] In the above-mentioned embodiment, in the classification unit 71, the spectrum of fluorescence detected by the detection unit 81 is classified into one of the predetermined multiple groups as a prescribed factor affecting the concentration calculation, but the prescribed factor affecting the concentration calculation is not limited to the fluorescence spectrum, and may be other factors, or may be factors including other factors. For example, the prescribed factor affecting the concentration calculation may also be the fluorescence spectrum, the elapsed time since the start of dialysis treatment, the patient as the object of dialysis treatment, the treatment method for the patient, the type of the filter of the dialyzer used in the dialysis treatment, the intensity of fluorescence detected by the detection unit 81, the maximum value of the spectrum of fluorescence within the prescribed wavelength range, and the setting parameters for the detection of the detection unit 81. In this case, in addition to being divided based on the fluorescence spectrum, the multiple groups classified by the classification unit 71 may also be divided based on the elapsed time since the start of dialysis treatment, the patient as the object of dialysis treatment, the treatment method for the patient, the type of the filter of the dialyzer used in the dialysis treatment, the intensity of fluorescence detected by the detection unit 81, the maximum value of the spectrum of fluorescence within the prescribed wavelength range, and the setting parameters for the detection of the detection unit 81.

[0170] For example, the elapsed time from the start of dialysis treatment is divided into four time periods, for example, 0 minutes to 20 minutes, 20 minutes to 60 minutes, 60 minutes to 120 minutes, and 120 minutes to 240 minutes, and is divided into four groups, and standard models corresponding to the groups are prepared. Furthermore, the classifier 71 classifies the elapsed time, which is a prescribed element, into one of the four groups each time during dialysis treatment, and the concentration calculator 72 calculates the concentration of albumin using the standard model of the group into which the time is classified.

[0171] For example, 10 patients who are the subjects of dialysis treatment are divided into four groups of 2, 2, 3, and 3, and standard models corresponding to the respective groups are prepared. Furthermore, the classifier 71 classifies the patients who are the prescribed elements in the dialysis treatment into one of the four groups, and the concentration calculator 72 calculates the concentration of albumin using the standard model of the group into which the patients are classified.

[0172] The treatment methods for patients are divided into three groups, for example, hemodialysis, hemodiafiltration (low Alb leakage), and hemodiafiltration (high Alb leakage), and standard models corresponding to each group are prepared. In addition, the classification unit 71 classifies the treatment method as a prescribed element in the dialysis treatment into one of the three groups, and the concentration calculation unit 72 calculates the concentration of albumin using the standard model of the group to which it is classified.

[0173] The types of filters of the dialyzer used in the dialysis treatment are divided into four groups, for example, dialyzers, dialysis filters with little albumin leakage, dialysis filters with low albumin leakage, and dialysis filters with high albumin leakage, and standard models corresponding to the respective groups are prepared. Furthermore, the classifying unit 71 classifies the type of filter as a prescribed element in the dialysis treatment into one of the four groups, and the concentration calculating unit 72 calculates the concentration of albumin using the standard model of the group into which the filter is classified.

[0174] The intensity of the fluorescence detected by the detection unit 81 is divided into four groups, for example, in descending order of fluorescence intensity, and standard models corresponding to the respective groups are prepared. Furthermore, the classification unit 71 classifies the intensity of the fluorescence as a predetermined factor detected by the detection unit 81 into one of the four groups during dialysis treatment, and the concentration calculation unit 72 calculates the concentration of albumin using the standard model of the group into which the fluorescence intensity is classified.

[0175] For the maximum value in the wavelength range of the fluorescence spectrum, for example, a wavelength range of 380 nm or more and 480 nm or less, the maximum value is divided into four groups in descending order, and a standard model corresponding to each group is prepared. In addition, the classification unit 71 classifies the maximum value of the wavelength range of the fluorescence spectrum as a specified element detected by the detection unit 81 into one of the four groups during dialysis treatment, and the concentration calculation unit 72 calculates the concentration of albumin using the standard model of the group to which the classification is made. In addition, the wavelength range of fluorescence of 380 nm or more and 480 nm or less corresponds to the excitation wavelength at which the fluorescence characteristic of albumin presents a secondary peak, and is a range with a high correlation with the concentration of albumin.

[0176] For the setting parameters for detection of the detection unit 81, for example, the integration time (equivalent to the exposure time of the camera) as a setting parameter is divided into 4 groups, and standard models corresponding to each group are prepared respectively. In addition, the classification unit 71 classifies the integration time of the detection unit as a prescribed element detected by the detection unit 81 into one of the 4 groups during dialysis treatment, and the concentration calculation unit 72 calculates the concentration of albumin using the standard model of the group to which it is classified. In particular, in the low concentration area (about 1 mg / dL or less), the fluorescence intensity is small, so when the integration time as the setting parameter of the detection unit is lengthened, the fluorescence intensity can be increased. However, in the high concentration area (about 30 mg / dL or more), when the integration time is lengthened in the same way, the upper limit of the detection value of the detector is sometimes exceeded. Therefore, it is necessary to automatically switch the integration time according to the concentration, in fact, according to the fluorescence intensity and the dialysis elapsed time. Therefore, by classifying into groups according to the integration time as a setting parameter and preparing a standard model for each group, the accuracy of the concentration calculation can be improved.

[0177] The classification of the predetermined factors affecting the concentration calculation, such as the patient to be treated with dialysis, the treatment method for the patient, and the type of filter of the dialyzer used in the dialysis treatment, is performed using the necessary information input by the user from the input unit 74 .

[0178] In addition, the albumin concentration calculated at the actual time in the actual dialysis treatment for the patient (clinical) may produce errors due to various factors originating from the organism. Assuming this situation, it is necessary to determine whether to allow (adopt) the calculated albumin concentration as an appropriate value. At this time, it is also possible to set the parameter value related to the concentration calculation in such a way that the relationship between the percentage error A of the albumin concentration allowed in any time period of the first half of the dialysis treatment and the percentage error B of the measured concentration allowed in any time period of the second half of the dialysis treatment is A≤B.

[0179] For example, Fig.12The estimated difference E shown is set as an error. That is, the albumin concentration in the dialysis effluent is used until the actual time t n The previous time t n-1 The estimated curve of albumin concentration represented by formula (1) is obtained by calculating at least three or more calculation points up to the calculation point of the albumin concentration. Fig.13 The estimated albumin concentration curve was obtained using five calculation points. n ) The estimated difference E(t n ) divided by the actual time t expressed by (3) n-1 The estimated difference E(t n-1 )RMSD(t n-1 ) is the estimated differential rate EF(t n ),

[0180] When t<60 minutes of dialysis time, the estimated differential rate EF(t n ) satisfies -1.0≤EF(t n )≤1.0, the actual time t n The calculated substance concentration,

[0181] When the dialysis time is 60 minutes ≤ t < 120 minutes, the estimated differential rate EF (t n ) satisfies -1.2≤EF(t n )≤1.2, the actual time t n The calculated substance concentration,

[0182] When the dialysis time is 120 minutes ≤ t, the estimated differential rate EF (t n ) satisfies -1.5≤EF(t n )≤1.5, the actual time t n Calculated substance concentration.

[0183] C(t)=A×t^(-B)···(1)

[0184] E(t n )=(at actual time t n Calculated concentration)-C(t n )···(2)

[0185] RMSD(t n-1 )={[Σ(E(t i )^2)] / (n-1)}^0.5···(3)

[0186] EF(t n )=E(tn ) / RMSD(t n-1 )···(4)

[0187] Where, C: estimated albumin concentration in the dialysate [mg / dL], A: constant, t: time [min], B: constant, E: estimated difference, suffix i: the i-th calculated concentration, suffix n: the n-th calculated concentration.

[0188] exist Fig.12 In FIG. 1 , the calculated concentration at the sixth calculation point is the calculated concentration at a dialysis time of 40 minutes (t=40), and is therefore adopted when the EF value obtained from the estimated difference E of the calculated concentration is greater than or equal to −1.0 and less than or equal to 1.0.

[0189] like Fig.13 As shown, the concentration variation curve of albumin in the dialysis discharge can explain that, in the case of 4 hours of dialysis treatment, about 80% of the total amount of albumin in the dialysis discharge is discharged in 1 hour at the beginning of dialysis treatment. In addition, it can be known that the initial dialysis concentration of albumin is about more than 20 times of the concentration in the final stage of dialysis. Therefore, when trying to obtain the total leakage of albumin, for example, according to the calculated albumin concentration, even if the percentage error from the latter half of dialysis treatment to the final stage is set to be larger than the percentage error in the initial stage of dialysis treatment, the total leakage of albumin will not be caused to produce a large error. On the other hand, generally speaking, the lower the concentration of the concentration calculation object, the more expensive the optical monitor, etc., are in order to measure with high accuracy, but the concentration calculation device can be designed more cheaply by the consideration method based on the above-mentioned allowable error.

[0190] The preferred embodiments of the present invention have been described above with reference to the accompanying drawings, but the present invention is not limited to the above examples. Those skilled in the art will clearly know that within the scope of the concept described in the claims, various variations or modifications can be thought of, and it is understood that these variations or modifications also naturally belong to the technical scope of the present invention.

[0191] For example, in the above embodiment, the concentration calculation object irradiated with excitation light is a dialysate (protein solution) containing albumin, and the concentration calculation device 16 calculates the concentration of albumin contained in the dialysate. However, the present invention can also be applied to calculate the concentration of substances other than albumin (e.g., protein) in the dialysate. In addition, the present invention can also irradiate excitation light to other concentration calculation objects other than the dialysate, and calculate the substances contained in the concentration calculation object.

[0192] In addition, the structure of the dialysis system 1 is not limited to the structure for performing hemodialysis filtration as in the above-mentioned embodiment. For example, when hemodialysis is performed in the dialysis system 1, the infusion pump 50 and the infusion circuit 14 may not be used. The infusion circuit 14 may also be connected to the return blood line 33 instead of being connected to the blood removal line 31. In addition, the present invention is not limited to being applied to the dialysis system, but can also be applied to a blood treatment system for performing other blood treatments. The present invention can also be applied to a blood treatment system for performing, for example, plasma exchange therapy, leukocyte removal therapy, continuous slow blood filtration therapy, etc.

[0193] (Example)

[0194] (Example 1)

[0195] A dialysis patient with a four-year dialysis history was treated with an online dialysis filtration therapy using a pre-replacement fluid volume of 45 L using an MFX-21M eco (manufactured by Nipro) filter. Fig.14 FIG. 8 shows the fluorescence spectrum obtained by irradiating the dialysate with ultraviolet light using an optical monitor. Fig.15 4 shows the albumin concentration (calculated concentration) calculated by the method of the above-mentioned concentration calculation device 16 based on the obtained fluorescence spectrum. Fig.16 The classification of groups A to D used in calculating the concentration of albumin is shown in FIG. The classification of the groups is performed based on the feature quantity obtained by principal component analysis of the fluorescence spectrum. Specifically, first, Fig.14 The fluorescence spectra of 76 samples of dialysis effluent obtained from multiple dialysis patients detected by an optical monitor as shown in the figure are subjected to principal component analysis. Each fluorescence spectrum is obtained in the form of a 51-dimensional vector. The intrinsic matrix transformed into the first component factor, the second component factor, ..., the fifty-second component factor is obtained by principal component analysis, and the factor score is obtained based on the measured fluorescence spectrum and the intrinsic matrix. Next, a matrix with the first component as the horizontal axis and the second component as the vertical axis is prepared. Fig.16 The graph shown in FIG. 1 is divided into four groups A to D, wherein the first component is the component factor that best represents the characteristics of the fluorescence spectrum, and the second component is the component that second best represents the characteristics of the fluorescence spectrum. Then, based on the fluorescence spectra of the samples corresponding to each group A to D, a standard model is prepared for each group A to D. The albumin concentration is calculated by classifying the fluorescence spectrum obtained from the optical monitor into one of the four groups A to D and using the standard model corresponding to the group to which it is classified. Fig.15 The calculated concentration and the measured concentration are shown in . It can be seen that the calculated concentration is well consistent with the measured concentration. It can be confirmed that the concentration can be calculated with high accuracy even in clinical dialysis effluent mixed with various substances derived from organisms.

[0196] (Example 2)

[0197] An ABH-21P filter (produced by Asahi Kasei Medical Corporation) was used to perform online dialysis filtration therapy with a pre-replacement fluid volume of 16L for a dialysis patient with a four-year dialysis history. The dialysis effluent was irradiated with ultraviolet light to obtain a fluorescence spectrum. The obtained fluorescence spectrum was classified into one of a plurality of groups, and the albumin concentration was calculated according to the standard model of the corresponding group. The classification of the groups was performed using principal component analysis in the same manner as in Example 1. The total albumin leakage estimated by the method of the above-mentioned estimated value calculation unit 103 until 60 minutes and estimated using formula (1) was 1.3g. On the other hand, the actual total albumin leakage obtained from the inspection data was 1.2g, and it can be seen that the estimated total albumin leakage was well consistent with the actual total albumin leakage. It can be confirmed that the estimated total albumin leakage obtained by the concentration calculation device of the present invention is consistent with the measured value with high precision.

[0198] (1) Feature quantities obtained by principal component analysis of the spectrum

[0199] When the 76 samples of dialysis drainage in Example 1 were not classified into groups and a standard model was made and verified, the correlation coefficient (R) and root mean square error (RMSE) of the 19 samples of the verification data were 0.729 and 3.77 mg / dL, respectively. On the other hand, if the groups were classified by principal component analysis (Example 1), and a standard model was made for each group, and the verification was performed using the verification data, the correlation coefficient (R) and root mean square error (RMSE) were 0.972 and 0.966 mg / dL, respectively. The correlation coefficient (R) increased, the root mean square error (RMSE) decreased, and the concentration calculation accuracy was improved.

[0200] (2) Time elapsed since the start of dialysis treatment

[0201] When the 76 samples of the dialysis drainage of Example 1 were not classified into groups and a standard model was made and verified, the correlation coefficient (R) and root mean square error (RMSE) of the 19 samples of the verification data were 0.729 and 3.77 mg / dL, respectively. On the other hand, if the samples were classified into 4 groups according to the elapsed time (0 minutes to 20 minutes, 20 minutes to 60 minutes, 60 minutes to 120 minutes, and 120 minutes to 240 minutes), and a standard model was made for each group, and the verification was performed using the verification data, the correlation coefficient (R) and root mean square error (RMSE) were 0.986 and 0.790 mg / dL, respectively. The correlation coefficient (R) increased, the root mean square error (RMSE) decreased, and the concentration calculation accuracy was improved.

[0202] (3) Intensity of fluorescence detected by the detection unit

[0203] When the 76 samples of the dialysis drainage of Example 1 were not classified into groups and a standard model was made and verified, the correlation coefficient (R) and root mean square error (RMSE) of the 19 samples of the verification data were 0.729 and 3.77 mg / dL, respectively. On the other hand, if the fluorescence intensity was classified into 4 groups in the order of fluorescence intensity, and a standard model was made for each group, and the verification was performed using the verification data, the correlation coefficient (R) and root mean square error (RMSE) were 0.967 and 1.57 mg / dL, respectively. The correlation coefficient (R) increased and the root mean square error (RMSE) decreased, improving the accuracy of concentration calculation.

[0204] (4) Setting parameters for detection by the detection unit

[0205] Especially in the low concentration area (about 1 mg / dL or less), the fluorescence intensity is small, so if the integration time as a setting parameter of the detection unit is lengthened, the fluorescence intensity can be increased. However, in the high concentration area (about 30 mg / dL or more), when the integration time is also lengthened, the upper limit of the detection value of the detector is sometimes exceeded. Therefore, it is necessary to automatically switch the integration time according to the concentration, in fact, according to the fluorescence intensity and the dialysis elapsed time. Therefore, by classifying into groups according to the integration time as a setting parameter and making a standard model for each group, the accuracy of the concentration calculation can be improved.

[0206] (5) Patients who are candidates for dialysis treatment

[0207] When the 76 samples of dialysis drainage in Example 1 were not classified into groups and a standard model was made and verified, the correlation coefficient (R) and root mean square error (RMSE) of the 19 samples of the verification data were 0.729 and 3.77 mg / dL, respectively. On the other hand, if the patients (6 people) were classified into groups and a standard model was made for each group and verified by the verification data, the correlation coefficient (R) and root mean square error (RMSE) were 0.888 and 1.90 mg / dL, respectively. The correlation coefficient (R) increased and the root mean square error (RMSE) decreased, which improved the accuracy of concentration calculation.

[0208] (6) Treatment methods for patients

[0209] When the 76 samples of the dialysis drainage of Example 1 were not classified into groups and a standard model was made and verified, the correlation coefficient (R) and root mean square error (RMSE) of the 19 samples of the verification data were 0.729 and 3.77 mg / dL, respectively. On the other hand, if the dialysis treatment (hemodialysis, hemodiafiltration Alb leakage is low, and hemodiafiltration Alb leakage is high) is classified into 3 groups, and a standard model is made for each group, and the verification is performed by the verification data, the correlation coefficient (R) and root mean square error (RMSE) are 0.954 and 1.25 mg / dL, respectively. The correlation coefficient (R) increases, the root mean square error (RMSE) decreases, and the concentration calculation accuracy is improved.

[0210] (7) Types of filters used in dialysis treatment

[0211] When the 76 samples of the dialysis drainage of Example 1 were not classified into groups and a standard model was created and verified, the correlation coefficient (R) and root mean square error (RMSE) of the 19 samples of the verification data were 0.729 and 3.77 mg / dL, respectively. On the other hand, if the filters (4 types) were classified into 4 groups and a standard model was created for each group and verified by the verification data, the correlation coefficient (R) and root mean square error (RMSE) were 0.981 and 0.880 mg / dL, respectively. The correlation coefficient (R) increased and the root mean square error (RMSE) decreased, which improved the accuracy of concentration calculation.

[0212] Second embodiment

[0213] <Dialysis System>

[0214] Fig.17 1 is an explanatory diagram schematically showing the configuration of a dialysis system 1 as a blood processing system equipped with the concentration calculation device according to the present embodiment.

[0215] The dialysis system 1 includes, for example, a dialyzer 10 , a blood circuit 11 , a dialysate circuit 12 , a drainage circuit 13 , a fluid replacement circuit 14 , a control device 15 , a concentration calculation device 16 , and the like.

[0216] The dialyzer 10 is, for example, a hollow fiber module with a hollow fiber membrane built therein, which can separate useless components from the blood. The dialyzer 10 has a cylindrical container 20, and a plurality of hollow fiber membranes 21 are arranged inside the cylindrical container 20 along the long side direction of the cylindrical container 20. The hollow fiber membrane 21 can separate useless components from the blood. An inlet 22 and an outlet 23 connected to the inner space (blood side) of the hollow fiber membrane 21 are provided at the upper and lower parts of the cylindrical container 20, and two inlets and outlets 24, 25 connected to the outer space (dialysis fluid side) of the hollow fiber membrane 21 are provided at the side part of the cylindrical container 20.

[0217] The blood circuit 11 includes, for example, a blood removal line 31 connecting the blood removal unit 30 and the dialyzer 10, and a blood return line 33 connecting the dialyzer 10 and the blood return unit 32. The blood removal line 31 and the blood return line 33 are mainly composed of soft tubes. The blood removal line 31 is connected to the inlet 22 of the dialyzer 10, and the blood return line 33 is connected to the outlet 23 of the dialyzer 10.

[0218] For example, a blood pump 40 is provided on the blood vessel removal line 31. In addition, a dripping funnel 41 is connected to the blood vessel removal line 31. The dripping funnel 41 may not be provided in some cases.

[0219] The dialysate circuit 12 is connected from a dialysate supply source (not shown) to an inlet and outlet 25 of the dialyzer 10. The drainage circuit 13 is connected from an inlet and outlet 24 of the dialyzer 10 to a drainage unit (not shown). The dialysate circuit 12 and the drainage circuit 13 are provided with a supply and drainage pump (not shown) for supplying the dialysate to the dialyzer 10 through the dialysate circuit 12 and for discharging the dialysate from the dialyzer 10 through the drainage circuit 13.

[0220] The fluid replacement circuit 14 is connected to the dripping funnel 41 (blood circuit 11) from the dialysate circuit 12, for example. In the absence of the dripping funnel 41, the fluid replacement circuit 14 is directly connected to the blood removal line 31. The fluid replacement circuit 14 is provided with a fluid replacement pump 50.

[0221] The control device 15 is, for example, a computer, and can control the operation of the blood pump 40 and the infusion pump 50 by, for example, executing a program stored in a storage unit by a CPU, thereby performing a dialysis treatment for dialysis treatment. In addition, the communication between the control device 15 and various devices (blood pump 40, infusion pump 50) can be performed by a wired method such as a communication cable, or by a wireless method such as Bluetooth (registered trademark).

[0222] In dialysis treatment, in the blood circuit 11, the patient's blood is transported from the blood removal section 30 to the tube space of the hollow fiber membrane 21 of the dialyzer 10, and returns to the patient from the blood return section 32 after passing through the dialyzer 10. At this time, the dialysate is transported to the tube space outside the hollow fiber membrane 21 of the dialyzer 10 through the dialysate circuit 12, and then discharged through the drainage circuit 13. In the dialyzer 10, the blood flowing in the tube space of the hollow fiber membrane 21 is mainly useless components that flow out to the tube space outside (dialysate side) through the hollow fiber membrane 21 and are discharged together with the dialysate. The infusion (dialysate) of the dialysate circuit 12 is supplied to the blood circuit 11 through the infusion circuit 14 to supplement the blood with the specified components. In addition, there are hemodialysis filtration with supplementation of infusion and hemodialysis without supplementation of infusion.

[0223] <Concentration Calculation Device>

[0224] The concentration calculation device 16 includes: an optical monitor 70 for irradiating the dialysate as an irradiation target (protein solution) containing albumin with excitation light and detecting fluorescence generated from the dialysate; a concentration calculation unit 71 for calculating the concentration of albumin contained in the dialysate based on the fluorescence intensity of a plurality of wavelengths in a predetermined wavelength range of the spectrum of the fluorescence acquired by the optical monitor 70 and a preset standard model; and a display unit 72. In addition, the communication among the optical monitor 70, the concentration calculation unit 71, and the display unit 72 may be performed by a wired method such as a communication cable or by a wireless method such as Bluetooth (registered trademark).

[0225] For example Fig.18 As shown, the optical monitor 70 includes an irradiation unit 80 for irradiating the dialysate in the pipe of the drainage circuit 13 with excitation light, and a detection unit 81 for detecting fluorescence generated from the dialysate.

[0226] The irradiation unit 80 can irradiate Fig.19 The light shown in the figure is light with a wavelength between 300 nm and 400 nm, which excites the fluorescence of albumin as a secondary peak. The light source of the irradiation unit 80 is not particularly limited, but for example, it can be a halogen lamp, a xenon lamp, a deuterium lamp, an LED, etc. The wavelength of the excitation light irradiated from the irradiation unit 80 is preferably between 310 nm and 380 nm, and more preferably between 330 nm and 350 nm. In addition, the excitation light irradiated from the irradiation unit 80 can be either an excitation light with a continuous wavelength or an excitation light with a wavelength of a certain value (for example, 340 nm).

[0227] The detection unit 81 detects the fluorescence generated from the dialysis drainage to obtain a fluorescence spectrum. The detection unit 81 can detect fluorescence in a wavelength range of 310 nm to 850 nm. Preferably, the fluorescence wavelength detected by the detection unit 81 is in a range between 850 nm and above the wavelength obtained by adding 10 nm to the wavelength (excitation wavelength) that excites the fluorescence (in the case where the excitation wavelength has a range, it is the lower limit of the excitation wavelength range + 10 nm), and more preferably, in a range between 800 nm and below the wavelength obtained by adding 10 nm to the excitation wavelength (in the case where the excitation wavelength has a range, it is the lower limit of the excitation wavelength range + 10 nm). In addition, the irradiation unit 80 and the detection unit 81 are, for example, arranged in a direction that is perpendicular to each other relative to the piping of the drainage circuit 13, and the fluorescence is obtained at an angle perpendicular to the excitation light. In addition, the configuration of the irradiation unit 80 and the detection unit 81 is not limited to this.

[0228] The concentration calculation unit 71 is a computer. Fig. 20As shown, the concentration calculation unit 71 includes: a concentration calculation unit 90, which uses information on fluorescence intensities of multiple wavelengths in a predetermined wavelength range of the fluorescence spectrum acquired by the optical monitor 70 and a standard model obtained by multivariate analysis to calculate the concentration of albumin in the dialysis discharge; a storage unit 91, which stores the albumin concentration obtained by the concentration calculation unit 90; and a numerical calculation unit 92, which uses the albumin concentration stored in the storage unit 91 to calculate various numerical values ​​described later; and a standard model preparation unit 93, which prepares a standard model in advance.

[0229] The standard model preparation unit 93 obtains the fluorescence spectra of the simulated dialysis drainage fluid with known and different albumin concentrations from the optical monitor 70, and prepares the standard model M in advance through multivariate analysis. The multivariate analysis used to prepare the standard model M may also be any one of partial least squares (PLS) regression analysis, principal component regression analysis, multivariate regression analysis, support vector machine regression analysis, and machine learning analysis. The prepared standard model M is stored in the storage unit 91, for example, and is used as a parameter in the calculation process in the concentration calculation unit 90 and the numerical calculation unit 92. In addition, the pre-calculated standard model may be stored in the storage unit 91 and used.

[0230] An example of the standard model M is shown below. The standard model M can be expressed as a formula for obtaining the calculated concentration C of albumin by multiplying the fluorescence intensity at each wavelength of the obtained fluorescence spectrum by a coefficient and taking the sum. Specifically, the standard model M is expressed as formula (6).

[0231] C=a1×I1+a2×I2+···+a n ×I n +K···(6)

[0232] Where, C: calculated concentration of albumin, a n : Coefficient, I n : fluorescence intensity, K: constant, suffix n: a natural number assigned to each wavelength of the acquired fluorescence spectrum in ascending order of wavelength (e.g., if the wavelength is 300nm, 310nm, 320nm,..., 400nm, the suffixes are 1, 2, 3,..., 11).

[0233] The concentration calculator 90 acquires the fluorescence spectrum of the dialysate from the optical monitor 70 during the actual dialysis treatment, and calculates the albumin concentration based on the fluorescence intensities at multiple wavelengths in a predetermined wavelength range of the fluorescence spectrum and a standard model M prepared in advance.

[0234] The numerical calculation unit 92 obtains the albumin concentration value stored in the storage unit 91 to calculate various numerical values ​​during the actual time of the dialysis treatment. Fig.21 As shown, the numerical calculation unit 92 includes: an integral value calculation unit 100, which calculates, for example, an integral value of the product of the albumin concentration and the dialysate discharge flow rate since the start of the dialysis treatment; a change rate calculation unit 101, which calculates the rate of change of the albumin concentration with respect to time (dC / dt); a difference calculation unit 102, which calculates the difference between the albumin concentration and a predetermined value; and an estimated value calculation unit 103, which calculates an estimated value of the total amount of albumin in the dialysate from the start of the dialysis treatment to the end of the dialysis treatment.

[0235] like Fig. 22 As shown, the integral value calculation unit 100 calculates the albumin leakage at that moment based on the product of the albumin (Alb) concentration value since the start of the dialysis treatment and the dialysis drainage flow rate, and calculates the integral value of the albumin leakage based on the time curve of the albumin leakage. The integral value is the albumin leakage from the start of the dialysis treatment to that moment ( Fig. 22 of the shading).

[0236] like Fig.23 As shown, the change rate calculation unit 101 makes the slope of the time change curve of the albumin concentration the change rate of albumin with respect to time (dC / dt).

[0237] like Fig.24 As shown, the other predetermined value in the difference calculation unit 102 is, for example, an albumin concentration calculated based on a preset allowable albumin leakage amount, and the albumin concentration becomes a threshold value for indicating excessive albumin leakage when the albumin concentration is exceeded. In this case, the difference calculation unit 102 calculates the difference between the albumin concentration and the threshold value.

[0238] The estimated value calculation unit 103 first obtains A and B of the formula shown in formula (1) based on at least three or more calculation points of albumin concentration. Fig.25 The albumin leakage estimation curve is shown in FIG. Then, the albumin leakage after the start of the dialysis treatment is estimated based on the formula (1), and the estimated value of the total amount of albumin (total leakage) per treatment is calculated during the dialysis treatment.

[0239] C(t)=A×t^(-B)···(1)

[0240] Where, C: concentration in dialysis effluent [mg / dL], A: constant, t: time [min], B: constant

[0241] Fig. 20The display unit 72 shown is, for example, a flat panel display, and displays the albumin concentration calculated by the concentration calculation unit 90, various numerical values ​​obtained by the numerical calculation unit 92, or various warnings. For example, a warning is issued when the albumin concentration calculated by the concentration calculation unit 90 exceeds a prescribed range, when the integral value of the albumin leakage calculated by the integral value calculation unit 100 exceeds a prescribed range, when the rate of change of the albumin concentration calculated by the rate of change calculation unit 101 exceeds a prescribed range, when the difference between the albumin concentration calculated by the difference calculation unit 102 and its threshold value exceeds a prescribed range, when the estimated value of the total albumin leakage for each treatment calculated by the estimated value calculation unit 103 exceeds a prescribed range, etc. In addition, the above control device 15, concentration calculation unit 71, and display unit 72 may also be methods and functions implemented by the same computer.

[0242] Next, the operation of the concentration calculation device 16 configured as above will be described.

[0243] The concentration calculation device 16 continuously or intermittently calculates the albumin concentration in the dialysis effluent of the effluent circuit 13 in real time during the dialysis treatment. The dialysis effluent is a continuous flow with a flow rate of 10 mL / min or more and 1000 mL / min or less.

[0244] Specifically, the irradiation unit 80 of the optical monitor 70 irradiates the excitation light, and the detection unit 81 detects the fluorescence generated from the dialysis effluent. The excitation light irradiated from the irradiation unit 80 includes light with a wavelength of 300 nm to 400 nm that excites the fluorescence of albumin to become a secondary peak. In addition, the excitation light may also include light with a wavelength of 310 nm to 380 nm, light with a wavelength of 320 nm to 350 nm, light with a wavelength of 330 nm to 350 nm, and light with a wavelength of 340 nm to 350 nm.

[0245] like Fig.26 As shown, the fluorescence spectrum detected by the detection unit 81 can be a range of wavelengths greater than or equal to the wavelength obtained by adding 10 nm to the lower limit of the wavelength of the excitation light and less than or equal to 450 nm to the upper limit of the wavelength of the excitation light, or a range of wavelengths greater than or equal to the wavelength obtained by adding 10 nm to the lower limit of the wavelength of the excitation light and less than or equal to 400 nm to the upper limit of the wavelength of the excitation light. The wavelength range of the fluorescence detected by the detection unit 81 is, for example, greater than or equal to 310 nm and less than or equal to 850 nm. In the detection unit 81, the parameters related to the measurement are adjusted based on the numerical value of the maximum intensity value of the wavelength range of the obtained fluorescence spectrum greater than or equal to 380 nm and less than or equal to 480 nm, and the fluorescence spectrum is measured within the upper limit of the measurement of the sensor of the detection unit 81.

[0246] The concentration calculation unit 90 of the concentration calculation unit 71 calculates the albumin concentration contained in the dialysate based on the fluorescence intensity of a plurality of wavelengths in a predetermined wavelength range of the fluorescence spectrum detected by the detection unit 81 and the standard model M. The concentration calculation unit 90 continuously or intermittently calculates the albumin concentration in the dialysate during the actual time of the dialysis treatment. The albumin concentration calculated by the concentration calculation unit 90 is stored in the storage unit 91 each time. The albumin concentration is displayed on the display unit 72, for example, in real time.

[0247] In addition, before the start of dialysis treatment, a standard model M is prepared in advance in the standard model preparation unit 93 and stored in the storage unit 91. For example, fluorescence spectra of simulated dialysis effluents with different albumin concentrations are obtained from the optical monitor 70, and the standard model M is prepared by multivariate analysis. Multivariate analysis can use, for example, partial least squares (PLS) regression analysis, principal component regression analysis, multivariate regression analysis, support vector machine regression analysis, or machine learning analysis.

[0248] The numerical calculation unit 92 obtains the albumin concentration calculated by the concentration calculation unit 90 from the storage unit 91 to calculate various numerical values. The integral value calculation unit 100 calculates the integral value of the product of the albumin concentration and the dialysate discharge flow rate since the start of the dialysis treatment, the change rate calculation unit 101 calculates the change rate of the albumin concentration, the difference calculation unit 102 calculates the difference between the albumin concentration and a predetermined value, and the estimated value calculation unit 103 calculates the estimated value of the total amount of albumin in the dialysate from the start of the dialysis treatment to the end of the dialysis treatment. These values ​​are displayed on the display unit 72, for example. In addition, when the albumin concentration, the integral value of the albumin leakage amount, the change rate of the albumin concentration, the difference between the albumin concentration and the threshold value, and the estimated amount of albumin leakage in one treatment are not within an appropriate range, a warning is displayed on the display unit 72.

[0249] According to the present embodiment, the excitation light irradiated to the dialysate includes light having a wavelength of 300 nm to 400 nm that excites fluorescence of albumin to form a secondary peak, and the concentration of albumin contained in the dialysate is calculated based on the fluorescence intensities of a plurality of wavelengths in a predetermined wavelength range of the spectrum of fluorescence generated from the dialysate by irradiation with the excitation light and the standard model M. As a result, the concentration of albumin can be calculated with high accuracy without complicating the device.

[0250] In this regard, regarding the detection of substances present in mixed solutions such as dialysis effluent using autofluorescence and the calculation of the amount of substances, if the description of paragraph 0048 of Patent Document 4 is cited, it is generally believed that "the emission spectrum of each fluorescent pigment molecule of interest in pathological analysis overlaps with a specific irradiation wavelength to a large extent and exists in the same ultraviolet region. In addition to this disadvantage, since the fluorescence spectrum of the above-mentioned molecules is relatively wide, it is difficult to deconvolve the measured spectrum, or even if it can be performed, it will be accompanied by a large measurement error. Therefore, based on general fluorescence spectroscopy, it is impossible to obtain the accurate concentration and accurate ratio of each substance in the dialysate. In other words, it is impossible to quantitatively determine the concentration with high reliability." This has also been confirmed by the research of the inventors of the present application. In the mixed solution, in the research conducted near the representative excitation wavelength of 280nm to 296nm for proteins such as albumin, the influence of the mixed substance cannot be excluded. Therefore, an excitation wavelength of 300 nm to 400 nm that is not easily affected by mixed substances is found, and then a multivariate analysis is performed on the fluorescence spectrum of 310 nm to 850 nm obtained at an excitation wavelength of 300 nm to 400 nm to obtain a standard model. The albumin concentration is calculated based on the newly obtained fluorescence spectrum based on the standard model, and the albumin concentration can be calculated with high accuracy.

[0251] By making the excitation light include light having a wavelength of 310 nm to 380 nm, the albumin concentration can be calculated with higher accuracy.

[0252] Fig. 27 An experimental example of the correlation coefficient R and the root mean square error RMSE of the calculated concentration of albumin and the actual albumin concentration when the wavelength range in the fluorescence spectrum is changed is shown in FIG. The excitation wavelength includes wavelengths between 300 nm and 400 nm. When the fluorescence start wavelength (minimum wavelength) is set to 310 nm, the error (RMSE) becomes smaller. By changing the fluorescence start wavelength (minimum wavelength) from 310 nm to 340 nm, the error (RMSE) becomes smaller, but the error becomes slightly larger at 350 nm. By changing the fluorescence end wavelength (maximum wavelength) from 850 nm to 750 nm, the error becomes smaller. By setting the fluorescence wavelength to a range from a wavelength obtained by adding 10 nm to the wavelength of the excitation light to a wavelength obtained by adding 450 nm to the wavelength of the excitation light, the concentration of albumin with high accuracy can be achieved. In addition, by setting the fluorescence wavelength to a range from a wavelength obtained by adding 10 nm to the wavelength of the excitation light to a wavelength obtained by adding 400 nm to the wavelength of the excitation light, the concentration of albumin with high accuracy can be calculated.

[0253] The concentration calculation device 16 further includes a standard model creation unit 93 that performs multivariate analysis to create a standard model M. This allows the standard model M to be easily created with high reliability.

[0254] The multivariate analysis is any one of partial least squares (PLS) regression analysis, principal component regression analysis, multiple regression analysis, support vector machine regression analysis, and machine learning analysis, so the standard model M can be produced simply and with high accuracy.

[0255] The concentration calculation device 16 further includes a storage unit 91 for storing a plurality of albumin concentrations calculated after the start of the dialysis treatment. This allows, for example, the calculated albumin concentrations to be used to perform various numerical calculations and analyze the status of the dialysis treatment.

[0256] The concentration calculation device 16 includes an integral value calculation unit 100 that calculates the integral value of the product of the albumin concentration and the dialysate discharge flow rate stored in the storage unit 91 since the start of the dialysis treatment. The integral value is a value indicating the amount of albumin leakage up to that point in time, so that the amount of albumin leakage can be compared with a preset albumin permissible leakage amount during the dialysis treatment, thereby determining whether the amount of albumin leakage is appropriate.

[0257] The concentration calculation device 16 includes a change rate calculation unit 101 that calculates the change rate (dC / dt) of the albumin concentration stored in the storage unit 91 with respect to time. Fig.23 As in the case of Example A where the absolute value of the rate of change (dC / dt) of the slope of the albumin time change curve shown is large, it can be understood that the clogging of the hollow fiber membrane during the dialysis treatment is progressing rapidly.

[0258] The concentration calculation device 16 includes a difference calculation unit 102 that calculates the difference between the albumin concentration stored in the storage unit 91 and a predetermined value. Fig.24 As shown, when the difference between the calculated concentration and the predetermined value becomes large, it can be understood that the amount of albumin leakage is larger than expected.

[0259] The concentration calculation device 16 includes an estimated value calculation unit 103, which calculates a temporal expression of the albumin concentration after the start of the dialysis treatment based on the albumin concentration stored in the storage unit 91, and calculates an estimated value of the total amount of albumin in the dialysate from the start of the dialysis treatment to the end of the dialysis treatment based on the temporal expression. In this way, the estimated total leakage amount of albumin can be compared with the preset allowable total leakage amount of albumin during the dialysis treatment, so as to determine whether the dialysis treatment is appropriate.

[0260] The concentration calculation device 16 includes a display unit 72 that displays information on the albumin concentration calculated by the concentration calculation unit 71 . Therefore, the user can understand the albumin concentration in real time during the dialysis treatment.

[0261] The dialysis drainage irradiated with the excitation light is a continuous flow with a flow rate of more than 10mL / min and less than 1000mL / min. In the clinical field of dialysis treatment, there are many equipments such as dialysis devices and beds, and there is basically no space for setting up an optical monitor. Therefore, it is preferred that the optical monitor does not need equipment for storing dialysis drainage, equipment for separation such as membranes or chromatography, and equipment for obtaining a part of the dialysis drainage using a pump. Therefore, it is preferred that the optical monitor 70 is directly connected to the drainage circuit 13 or connected to the optical monitor 70 on the branch of the drainage circuit 13 to measure with a continuous flow, and it is preferred that the flow rate of the continuous flow is more than 10mL / min and less than 1000mL / min.

[0262] In addition, the albumin concentration in the dialysis effluent calculated at the actual time in the actual dialysis treatment for the patient (clinical) may produce errors due to various factors originating from the organism. Assuming this situation, it is necessary to determine whether to allow (adopt) the calculated albumin concentration as an appropriate value. At this time, it is also possible to set the parameter value related to the concentration calculation in such a way that the relationship between the percentage error A of the albumin concentration allowed in any time period of the first half of the dialysis treatment and the percentage error B of the measured concentration allowed in any time period of the second half of the dialysis treatment is A≤B.

[0263] For example, Fig.28 The estimated difference E shown is set as an error. That is, the albumin concentration in the dialysis effluent is used until the actual time t n The previous time t n-1 The estimated curve of albumin concentration represented by formula (1) is obtained by calculating at least three or more calculation points up to the calculation point of the albumin concentration. Fig.28 The estimated albumin concentration curve was obtained using five calculation points. n ) The estimated difference E(t n ) divided by the actual time t expressed by (3) n-1 The estimated difference E(t n-1 )RMSD(t n-1 ) is the estimated differential rate EF(t n ),

[0264] When t<60 minutes of dialysis time, the estimated differential rate EF(t n) satisfies -1.0≤EF(t n )≤1.0, the actual time t n The calculated albumin concentration,

[0265] When the dialysis time is 60 minutes ≤ t < 120 minutes, the estimated differential rate EF (t n ) satisfies -1.2≤EF(t n )≤1.2, the actual time t n The calculated albumin concentration,

[0266] When the dialysis time is 120 minutes ≤ t, the estimated differential rate EF (t n ) satisfies -1.5≤EF(t n )≤1.5, the actual time t n Calculated albumin concentration.

[0267] C(t)=A×t^(-B)···(1)

[0268] E(t n )=(at actual time t n Calculated concentration)-C(t n )···(2)

[0269] RMSD(t n-1 )={[Σ(E(t i )^2)] / (n-1)}^0.5···(3)

[0270] EF(t n )=E(t n ) / RMSD(t n-1 )···(4)

[0271] Where, C: estimated albumin concentration in the dialysate [mg / dL], A: constant, t: time [min], B: constant, E: estimated difference, suffix i: the i-th calculated concentration, suffix n: the n-th calculated concentration.

[0272] exist Fig.28 In FIG. 1 , the calculated concentration at the sixth calculation point is the calculated concentration at a dialysis time of 40 minutes (t=40), and is therefore adopted when the EF value obtained from the estimated difference E of the calculated concentration is greater than or equal to −1.0 and less than or equal to 1.0.

[0273] like Fig.29As shown, the concentration variation curve of albumin in the dialysis discharge can illustrate that, in the case of 4 hours of dialysis treatment, about 80% of the total amount of albumin in the dialysis discharge is discharged in 1 hour at the beginning of dialysis treatment. In addition, it can be known that the initial dialysis concentration of albumin is about more than 20 times of the concentration in the dialysis final stage. Therefore, when trying to obtain the total leakage of albumin according to the calculated albumin concentration, for example, even if the percentage error from the latter half of dialysis treatment to the final stage is set to be larger than the percentage error in the initial stage of dialysis treatment, the total leakage of albumin will not be caused to produce a large error. On the other hand, generally speaking, the lower the concentration of the object, the more expensive the optical monitor, etc., are for measuring with high accuracy, but the concentration calculation device can be designed more cheaply by the consideration method based on the above-mentioned allowable error.

[0274] The preferred embodiments of the present invention have been described above with reference to the accompanying drawings, but the present invention is not limited to the above examples. Those skilled in the art will clearly know that within the scope of the concept described in the claims, various variations or modifications can be thought of, and it is understood that these variations or modifications also naturally belong to the technical scope of the present invention.

[0275] For example, in the above embodiment, the irradiation object irradiated with the excitation light is the dialysate containing albumin, and the concentration calculation device 16 calculates the concentration of albumin contained in the dialysate. However, the present invention can also irradiate other irradiation objects containing albumin and calculate the albumin concentration of the irradiation object.

[0276] In addition, the structure of the dialysis system 1 is not limited to the structure for performing hemodialysis filtration as in the above-mentioned embodiment. For example, when hemodialysis is performed in the dialysis system 1, the infusion pump 50 and the infusion circuit 14 may not be used. The infusion circuit 14 may also be connected to the return blood line 33 instead of being connected to the blood removal line 31. In addition, the present invention is not limited to being applied to the dialysis system, but can also be applied to a blood treatment system for performing other blood treatments. The present invention can also be applied to a blood treatment system for performing, for example, plasma exchange therapy, leukocyte removal therapy, continuous slow blood filtration therapy, etc.

[0277] (Example)

[0278] (Example 1)

[0279] (Calculation example of albumin concentration)

[0280] The components of the simulated dialysate are shown in the following Table 1. A phosphate buffer solution (pH 7.4) was used as the solvent.

[0281] [Table 1]

[0282] Table 1 Simulated dialysis drainage

[0283]

[0284] The prepared simulated dialysis effluent was allowed to flow through a monitor at a flow rate of 600 mL / min and irradiated with an LED having a peak wavelength of 340 nm (half-peak width of 10 nm) to acquire a fluorescence spectrum. Fig.30 An example of the fluorescence spectrum obtained is shown in FIG. CA12880MA (produced by Hamamatsu Photonics Co., Ltd.) is used as the detector for acquiring the fluorescence. The output of the LED is set to 350 mA, and the integration time of the detector is set to 0.5 s, so that the maximum value of the fluorescence spectrum of 380 nm to 480 nm is within the upper limit of the measurement of the sensor of the detection unit. A standard model is prepared in advance based on the obtained fluorescence spectrum. According to the standard model, a simulated dialysis discharge with a known albumin concentration is used, and the albumin concentration is calculated using the concentration calculation device of the present invention. The consistency between the estimated concentration and the actual concentration is verified, and the accuracy of the standard model is verified. Fig.31 The results are shown in . The results show that the albumin concentration can be calculated with high accuracy using the standard model.

[0285] (Comparative Examples 1 and 2)

[0286] Using the simulated dialysis drainage prepared in the same manner as in Example 1, a spectrofluorophotometer (FP-8500: manufactured by JASCO Corporation) was used to irradiate the simulated dialysis drainage with excitation light of 280 nm (Comparative Example 1) and excitation light of 410 nm (Comparative Example 2), and fluorescence spectra were obtained. A standard model was prepared in advance based on the obtained fluorescence spectra. Based on the standard model, a simulated dialysis drainage with a known albumin concentration was used, and in addition to calculating the wavelength of the excitation light, the albumin concentration was calculated in the same manner as the concentration calculation device of the present invention, and the consistency between the estimated concentration and the actual concentration was verified, and the accuracy of the standard model was verified. Fig.32 The results when the excitation wavelength is 280 nm are shown in . From these results, it can be seen that even at the same actual concentration (known concentration), the estimated concentration has a large deviation, and it is difficult to calculate the albumin concentration with high accuracy.

[0287] (Examples 2 to 7)

[0288] As in Comparative Examples 1 and 2, the excitation light is set to 310 nm (Example 2), 320 nm (Example 3), 330 nm (Example 4), 340 nm (Example 5), 350 nm (Example 6), and 380 nm (Example 7), and irradiated to the simulated dialysis drainage to obtain a fluorescence spectrum. A standard model is prepared in advance based on the obtained fluorescence spectrum. According to the standard model, a simulated dialysis drainage with a known albumin concentration is used to calculate the albumin concentration using the concentration calculation device of the present invention, and the consistency between the estimated concentration and the actual concentration is verified, and the accuracy of the standard model is verified. The root mean square error (RMSE) obtained from the difference between the actual concentration and the estimated concentration at this time is calculated according to the following formula (5).

[0289] RMSE={[Σ((Ce i -Ck i )^2)] / n}^0.5···(5)

[0290] Among them, RMSE: root mean square error, Ce: estimated concentration, Ck: actual concentration,

[0291] Suffix i: the i-th sample, n: the number of samples.

[0292] Fig.33 ] The relationship between the excitation wavelength and RMSE obtained in Comparative Examples 1 and 2 and Examples 2 to 7 is shown in FIG. It is understood that when the excitation wavelength is 300 nm or more and 400 nm or less, the RMSE is small, and the accuracy of the estimated concentration is high.

[0293] (Albumin concentration changes)

[0294] If the response of the concentration calculation device is slow relative to the change in albumin concentration, it will not be able to follow the significant albumin concentration changes in the early stage of dialysis, and thus it will not be possible to monitor in real time, which is not practical. Therefore, the following study was conducted: Using a simulated dialysis effluent prepared in the same manner as in Example 1, the albumin concentration was continuously changed to confirm the response followability and practicality of the concentration calculation device. 750 mL of simulated dialysis effluent was allowed to flow at a flow rate of 500 mL / min and monitored by the concentration calculation device. Phosphate buffer was added to the pool at the same flow rate of 500 mL / min, and mixed / diluted in the simulated dialysis effluent pool. The albumin concentration was calculated based on a pre-made standard model. Fig.34 The concentration change of albumin is shown in . It can be confirmed that the concentration change of albumin is also substantially consistent with the theoretical curve, and the response follow-up property of the concentration calculation device is excellent, and there is no problem in practical use.

[0295] Industrial Applicability

[0296] The present invention is useful in providing a concentration calculation device and a blood processing system that can calculate the concentration of a substance with high accuracy without making the device complicated.

[0297] Description of Reference Numerals

[0298] 1: dialysis system; 16: concentration calculation device; 70: optical monitor; 71: classification unit; 72: concentration calculation unit; 73: display unit; 80: irradiation unit; 81: detection unit; 91: storage unit.

Claims

1. A concentration calculation device, characterized in that: have: an irradiation unit for irradiating an object for concentration calculation with excitation light; a detection unit that detects fluorescence generated from the concentration calculation object; a classification unit that classifies the spectrum of the fluorescence detected by the detection unit, which affects the concentration calculation, into one of a plurality of predetermined groups; as well as a concentration calculation unit that calculates the concentration of the substance included in the concentration calculation object based on a standard model corresponding to the group into which the spectrum of the fluorescence is classified by the classification unit, among a plurality of standard models pre-set corresponding to the plurality of groups, and the fluorescence intensity of a plurality of wavelengths in a predetermined wavelength range of the spectrum of the fluorescence detected by the detection unit, The standard model is created by performing multivariate analysis on the spectra of a plurality of fluorescences corresponding to each of the plurality of groups.

2. The concentration calculation device according to claim 1, characterized in that: The plurality of groups are divided based on feature quantities obtained by performing a principal component analysis on the spectrum of the fluorescence.

3. The concentration calculation device according to claim 1 or 2, characterized in that: The object for concentration calculation is dialysate discharged during dialysis treatment.

4. The concentration calculation device according to claim 3, characterized in that: The substance in question is albumin in the dialysate.

5. The concentration calculation device according to claim 3, characterized in that: Use the substance concentration in the dialysis effluent until the actual time t n The previous time t n-1 The estimated substance concentration C(t n ) The estimated difference E(t n ) divided by the actual time t expressed by (3) n-1 The estimated difference E(t n-1 )RMSD(t n-1 ) is the estimated differential rate EF(t n ), When t<60 minutes of dialysis time, the estimated differential rate EF(t n ) satisfies -1.0≤EF(t n )≤1.0, the actual time t n The calculated substance concentration, When the dialysis time is 60 minutes ≤ t < 120 minutes, the estimated differential rate EF (t n ) satisfies -1.2≤EF(t n )≤1.2, the actual time t n The calculated substance concentration, When the dialysis time is 120 minutes ≤ t, the estimated differential rate EF (t n ) satisfies -1.5≤EF(t n )≤1.5, the actual time t n The calculated substance concentration, C(t)=A×t^(-B)···(1) E(t n )=(at actual time t n The concentration calculated by the concentration calculation unit)-C(t n )···(2) RMSD(t n-1 )={[Σ(E(t i )^2)] / (n-1)}^0.5···(3) EF(t n )=E(t n ) / RMSD(t n-1 )···(4) Where C: estimated substance concentration in the dialysate [mg / dL], A: constant, t: time [min], B: constant, E: estimated difference, suffix i: the i-th calculated concentration, suffix n: the n-th calculated concentration.

6. The concentration calculation device according to claim 1, characterized in that: The multivariate analysis is any one of partial least squares (PLS) regression analysis, principal component regression analysis, multiple regression analysis, support vector machine regression analysis and machine learning analysis.

7. The concentration calculation device according to claim 1 or 2, characterized in that: A storage unit is further provided for storing a plurality of substance concentrations calculated by the concentration calculation unit at a plurality of times.

8. The concentration calculation device according to claim 7, characterized in that: The storage device further includes an integrated value calculation unit that calculates a time-integrated value of a product of each substance concentration corresponding to a plurality of time points stored in the storage unit and a flow rate of a concentration calculation target object.

9. The concentration calculation device according to claim 7, characterized in that: The storage device further includes a change rate calculation unit for calculating a change rate dC / dt of each substance concentration corresponding to a plurality of time points stored in the storage unit with respect to time, where C is the substance concentration and t is the time.

10. The concentration calculation device according to claim 7, characterized in that: The method further includes a difference calculation unit that calculates a difference between each substance concentration corresponding to a plurality of time points stored in the storage unit and a predetermined value.

11. The concentration calculation device according to claim 7, characterized in that: It also includes an estimated value calculation unit that calculates a temporal formula for substance concentration based on the concentrations of each substance corresponding to multiple time points stored in the storage unit, and calculates an estimated value of the total amount of substance contained in the concentration calculation object based on the temporal formula.

12. The concentration calculation device according to claim 1 or 2, characterized in that: A display unit is further provided for displaying information on the substance concentration calculated by the concentration calculation unit.

13. The concentration calculation device according to claim 1 or 2, characterized in that: The concentration calculation object irradiated with the excitation light is a continuous flow having a flow rate of 10 mL / min or more and 1000 mL / min or less.

14. A blood processing system, characterized in that: A concentration calculation device according to any one of claims 1 to 13 is provided.

15. A concentration calculation device, characterized in that: have: an irradiation unit configured to irradiate an irradiation object including albumin with excitation light; a detection unit that detects fluorescence generated from the irradiation object; and a concentration calculation unit that calculates the concentration of albumin contained in the irradiated object based on the fluorescence intensities at a plurality of wavelengths in a predetermined wavelength range of the spectrum of the fluorescence detected by the detection unit and a preset standard model, The excitation light includes light having a wavelength of 300 nm to 400 nm for stimulating fluorescence of albumin to form a secondary peak. The standard model was prepared by performing multivariate analysis.

16. The concentration calculation device according to claim 15, characterized in that: The excitation light includes light with a wavelength of 310 nm to 380 nm.

17. The concentration calculation device according to claim 15 or 16, characterized in that: The predetermined wavelength range of the fluorescence spectrum includes a range from a wavelength obtained by adding 10 nm to a lower limit of the wavelength of the excitation light to a wavelength obtained by adding 450 nm to an upper limit of the wavelength of the excitation light.

18. The concentration calculation device according to claim 15 or 16, characterized in that: The predetermined wavelength range of the fluorescence spectrum includes a range from a wavelength obtained by adding 10 nm to a lower limit of the wavelength of the excitation light to a wavelength obtained by adding 400 nm to an upper limit of the wavelength of the excitation light.

19. The concentration calculation device according to claim 15, characterized in that: The multivariate analysis is any one of partial least squares (PLS) regression analysis, principal component regression analysis, multiple regression analysis, support vector machine regression analysis and machine learning analysis.

20. The concentration calculation device according to claim 15 or 16, characterized in that: The concentration of albumin is calculated by the following formula, C=a1×I1+a2×I2+···+a n ×I n +K Where, C: calculated concentration of albumin, a n : Coefficient, I n : fluorescence intensity, K: constant, suffix n: a natural number assigned to each wavelength of the spectrum obtained by fluorescence from shortest to longest wavelength.

21. The concentration calculation device according to claim 15 or 16, characterized in that: The irradiation object is the dialysate discharged during the dialysis treatment.

22. The concentration calculation device according to claim 21, characterized in that: The method further includes a storage unit for storing a plurality of albumin concentrations calculated by the concentration calculation unit during dialysis treatment.

23. The concentration calculation device according to claim 22, characterized in that: The method further includes an integrated value calculation unit that calculates an integrated value of the product of the albumin concentration and the dialysate discharge flow rate stored in the storage unit since the start of the dialysis treatment.

24. The concentration calculation device according to claim 22 or 23, characterized in that: The storage device further includes a change rate calculation unit that calculates a change rate dC / dt of the albumin concentration stored in the storage unit with respect to time, where C is the albumin concentration and t is the time.

25. The concentration calculation device according to claim 22 or 23, characterized in that: The device further includes a difference calculation unit that calculates a difference between the albumin concentration stored in the storage unit and a predetermined value.

26. The concentration calculation device according to claim 22 or 23, characterized in that: The method further comprises an estimated value calculation unit, which calculates a temporal equation of the albumin concentration after the start of the dialysis treatment based on the albumin concentration stored in the storage unit, and calculates an estimated value of the total amount of albumin in the dialysate from the start of the dialysis treatment to the end of the dialysis treatment based on the temporal equation.

27. The concentration calculation device according to claim 21, characterized in that: A display unit is further provided for displaying information on the concentration of albumin calculated by the concentration calculation unit.

28. The concentration calculation device according to claim 21, characterized in that: The albumin concentration in the dialysis effluent was used until the actual time t n The previous time t n-1 The estimated albumin concentration C(t n ) The estimated difference E(t n ) divided by the actual time t expressed by (3) n-1 The estimated difference E(t n-1 )RMSD(t n-1 ) is the estimated differential rate EF(t n ), When t<60 minutes of dialysis time, the estimated differential rate EF(t n ) satisfies -1.0≤EF(t n )≤1.0, the actual time t n The calculated albumin concentration, When the dialysis time is 60 minutes ≤ t < 120 minutes, the estimated differential rate EF (t n ) satisfies -1.2≤EF(t n )≤1.2, the actual time t n The calculated albumin concentration, When the dialysis time is 120 minutes ≤ t, the estimated differential rate EF (t n ) satisfies -1.5≤EF(t n )≤1.5, the actual time t n The calculated albumin concentration, C(t)=A×t^(-B)···(1) E(t n )=(at actual time t n The concentration calculated by the concentration calculation unit)-C(t n )···(2) RMSD(t n-1 )={[Σ(E(t i )^2)] / (n-1)}^0.5···(3) EF(t n )=E(t n ) / RMSD(t n-1 )···(4) Where, C: estimated albumin concentration in the dialysate [mg / dL], A: constant, t: time [min], B: constant, E: estimated difference, suffix i: the i-th calculated concentration, suffix n: the n-th calculated concentration.

29. The concentration calculation device according to claim 21, characterized in that: The dialysate irradiated with the excitation light is a continuous flow at a flow rate of 10 mL / min or more and 1000 mL / min or less.

30. A blood processing system, characterized in that: A concentration calculation device according to any one of claims 15 to 29 is provided.

Images