Alzheimer's dementia onset time prediction device and Alzheimer's dementia onset time prediction method

Abstract

To predict when subjects with mild cognitive impairment will develop Alzheimer's disease. The Alzheimer's disease onset prediction device (1) calculates the hippocampal volume [mm ] of a subject with mild cognitive impairment relative to the Alzheimer's Disease Assessment Scale-Cognition score of the subject. 3 ]; a memory unit 20 that stores a cumulative distribution function determined by a predetermined n and a predetermined λ; a predicted onset time information calculation unit 30 that calculates predicted onset time information indicating the time when the subject is predicted to develop Alzheimer's disease based on the volume score value acquired by the acquisition unit 10 and the probability distribution function stored in the memory unit 20; and an output unit 40 that outputs the predicted onset time information calculated by the predicted onset time information calculation unit 30 to the outside.

Description

[Technical Field]

[0001] The present disclosure relates to an apparatus and method for predicting when a subject with mild cognitive impairment (MCI) will develop Alzheimer's disease. [Background technology]

[0002] Conventionally, a technique for classifying each of a plurality of subjects with mild cognitive impairment into one of a plurality of groups according to the risk of progression to Alzheimer's disease in the future is known (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. 2023 / 157447 Summary of the Invention [Problem to be solved by the invention]

[0004] However, while conventional techniques can classify multiple subjects with mild cognitive impairment into one of multiple groups based on the risk of progressing to Alzheimer's disease in the future, it is difficult to predict when a particular subject with mild cognitive impairment will develop Alzheimer's disease.

[0005] Therefore, the present disclosure aims to provide an Alzheimer's disease onset prediction device and an Alzheimer's disease onset prediction method that can calculate predicted onset information indicating the time when a subject with mild cognitive impairment is predicted to develop Alzheimer's disease. [Means for solving the problem]

[0006] The device for predicting the onset time of Alzheimer's disease dementia according to one embodiment of the present disclosure is a device for predicting the hippocampal volume [mm ] of a subject with mild cognitive impairment relative to the Alzheimer's Disease Assessment Scale-Cognition score of the subject. 3 ] and a cumulative distribution function 1-G(x;n,λ) for a variable x, which is determined by an integer n and a positive real number λ.

number

[0007] A method for predicting the onset time of Alzheimer's disease according to one embodiment of the present disclosure includes: calculating a cumulative distribution function 1-G(x;n,λ) for a variable x, where the cumulative distribution function 1-G(x;n,λ) is determined by an integer n and a positive real number λ;

number

[0008] According to one embodiment of the Alzheimer's disease onset prediction device and Alzheimer's disease onset prediction method of the present invention, it is possible to calculate predicted onset information indicating the time when a subject with mild cognitive impairment is predicted to develop Alzheimer's disease. [Brief explanation of the drawings]

[0009] [Figure 1] Figure 1 shows the frequency distribution table of the number of people who developed Alzheimer's disease in 0.8-year intervals among subjects diagnosed with mild cognitive impairment. [Figure 2] FIG. 2 is a graph showing the cumulative distribution difference functions defined by (n, λ)=(6, 5.75), (n, λ)=(13, 5.75), and (n, λ)=(26, 5.75). [Figure 3] FIG. 3 is a graph showing the value indicated by (Equation 6) according to the present disclosure superimposed on the number of cases in group G1 according to the present disclosure. [Figure 4] FIG. 4 is a graph showing the value indicated by (Equation 8) according to the present disclosure superimposed on the number of cases in group G2 according to the present disclosure. [Figure 5] FIG. 5 is a graph showing the value indicated by (Equation 10) according to the present disclosure superimposed on the number of cases in group G3 according to the present disclosure. [Figure 6] FIG. 6 is a graph showing the values ​​indicated by (Equation 11) according to the present disclosure superimposed on the numbers of cases in groups G1, G2, and G3 according to the present disclosure. [Figure 7] FIG. 7 is a block diagram showing an example of the configuration of an Alzheimer's disease onset prediction device according to an embodiment. [Figure 8] FIG. 8 is a flowchart of a first Alzheimer's dementia onset time prediction process according to the embodiment. [Figure 9]FIG. 9 is a block diagram showing an example of the configuration of an Alzheimer's disease onset prediction device according to a modified example. [Figure 10] FIG. 10 is a flowchart of a second Alzheimer's disease onset prediction process according to a modified example. DETAILED DESCRIPTION OF THE INVENTION

[0010] (How one aspect of the present invention was achieved) First, the process by which the inventors arrived at one aspect of the present disclosure will be described.

[0011] The inventors conceived the idea that if a mathematical function could be found that applies to a subject with mild cognitive impairment and that indicates the probability density of the period from when the subject is diagnosed with mild cognitive impairment to when the subject develops Alzheimer's disease, it might be possible to predict when the subject will develop Alzheimer's disease based on that function.

[0012] Based on this idea, the inventors conducted extensive research using the results of a follow-up study on the progression from mild cognitive impairment to Alzheimer's disease, which was conducted on multiple subjects with mild cognitive impairment, in order to find the above function that applies to each of the multiple subjects who were the subject of the follow-up study.

[0013] That is, the inventors first classified each of the multiple subjects with mild cognitive impairment who were the subject of a follow-up survey into one of multiple groups according to the risk of progressing to Alzheimer's disease in the future.

[0014] Then, for each of these multiple groups, the inventors searched for a mathematical function (hereinafter also referred to as a "search function") that indicated the probability density of the period from when a subject belonging to that group was diagnosed with mild cognitive impairment to when the subject developed Alzheimer's disease, which applied to the subjects belonging to that group.

[0015] In this case, the inventors devised a pathological model in which a subject with mild cognitive impairment develops Alzheimer's disease, in which small changes occur in each brain cell of the subject at an average incidence rate λ, and each of these small changes does not individually cause Alzheimer's disease, but rather the accumulation of these small changes n times leads to the onset of Alzheimer's disease.

[0016] Based on this pathological model, the inventors performed the above search under the idea that the above search function should be expressed using the Erlang distribution cumulative distribution function f(x;n,λ) defined by the following (Equation 1) using an integer n and a positive real number λ.

[0017] f(x;n,λ)=1-G(x;n,λ) (Formula 1)

[0018] Here, G(x;n,λ) is a function defined as follows:

[0019]

number

[0020] As a result, the inventors compared the hippocampal volume [mm ] of each of a plurality of subjects with mild cognitive impairment with that of the subject relative to the subject's ADAS-cog (Alzheimer's Disease Assessment Scale-cognitive subscale) score. 3 We found that by classifying subjects into one of several groups based on their risk of progressing to Alzheimer's disease in the future, based on the volume score value that indicates the ratio of [ ], it is possible to find a search function that applies to subjects belonging to each of the multiple groups.

[0021] Here, the Alzheimer's Disease Assessment Scale-Cognition score is the score obtained as a result of administering the Alzheimer's Disease Assessment Scale-Cognition to a subject, which is a widely used clinical psychological test for assessing cognitive impairment due to Alzheimer's disease dementia.

[0022] Hereinafter, specific examples of findings obtained by the inventors will be described.

[0023] First, based on the results of a follow-up study on the progression from mild cognitive impairment to Alzheimer's disease, the inventors classified each of 266 subjects with mild cognitive impairment who subsequently developed Alzheimer's disease into one of five groups, G1 to G5, according to the risk of future progression to Alzheimer's disease, as determined by the volume score value (Cog_Vol) defined by the following (Equation 2):

[0024] Cog_Vol = (hippocampal volume [mm 3 ]) / (Alzheimer's Disease Assessment Scale-Cognitive score) (Formula 2)

[0025] That is, the inventors classified each of the 266 subjects who subsequently developed Alzheimer's disease into one of five groups: G1, where the volume score value (Cog_Vol) was less than 101; G2, where the volume score value (Cog_Vol) was between 101 and 160; G3, where the volume score value was between 160 and 266; G4, where the volume score value was between 266 and 340; and G5, where the volume score value was 340 or more.

[0026] Next, based on the results of the follow-up survey, the inventors calculated the frequency distribution of the number of people who developed Alzheimer's disease (hereinafter also referred to as the "number of cases") every 0.8 years for each of the five groups, G1 to G5, during the period from when a subject in each group was diagnosed with mild cognitive impairment to when they developed Alzheimer's disease (hereinafter also referred to as the "onset period").

[0027] FIG. 1 is a frequency distribution table showing the above frequency distribution calculated by the inventors.

[0028] Next, since the difference between the Erlang distribution cumulative distribution functions f(x; n, λ) at time t2 and time t1, that is, the probability density of the Erlang distribution during the period from time t1 to time t2 is f(t2; n, λ) - f(t1; n, λ), the inventors defined the cumulative distribution difference function F(n, λ, k) represented by natural numbers k from 1 to 9 as in the following (Equation 3).

[0029] F(n, λ, k) = G(0.8(k - 1); n, λ) - G(0.8k; n, λ) (Equation 3)

[0030] Then, for group G1, from the frequency distribution table shown in FIG. 1, the number of onset cases g1(k) for k = 1 to 9 are respectively g1(1) = 13 (g1(1) is the number of onset cases during the onset period of 0.8 years or less), g1(2) = 25 (g1(2) is the number of onset cases during the onset period greater than 0.8 years and 1.6 years or less), g1(3) = 6 (g1(3) is the number of onset cases during the onset period greater than 1.6 years and 2.4 years or less), g1(4) = 2 (g1(4) is the number of onset cases during the onset period greater than 2.4 years and 3.2 years or less), g1(5) = 0 (g1(5) is the number of onset cases during the onset period greater than 3.2 years and 4.0 years or less), g1(6) = 3 (g1(6) is the number of onset cases during the onset period greater than 4.0 years and 4.8 years or less), g1(7) = 1 (g1(7) is the number of onset cases during the onset period greater than 4.8 years and 5.6 years or less), g1(8) = 0 (g1(8) is the number of onset cases during the onset period greater than 5.6 years and 6.4 years or less), g1(9) = 0 (g1(9) is the number of onset cases during the onset period greater than 6.4 years and 7.2 years or less). Therefore, using these values of g1(1) to g1(9), the inventors sought l, m, n, x1, y1, z1, λ that minimize the value of the following (Equation 4).

[0031] Here, l, m, n, x1, y1, z1 are natural numbers that satisfy l < m < n and x1 + y1 + z1 = 50.

[0032]

number

[0033] In order to find l, m, n, x1, y1, z1, and λ that minimize the value of (Equation 4), the inventors first found l, a, b, x1, and λ that minimize the value of the following (Equation 5).

[0034] |x1F(l,λ,1)-g1(1)| 2 +|x1F(l,λ,2)-g1(2)-a| 2 +|x1F(l,λ,3)-g1(3)-b| 2 (Formula 5)

[0035] Here, a and b are

number

[0036] The inventors used Mathematica 13.0 to find l, a, b, x1, and λ that minimize the value of (Equation 5), and obtained the following two candidates for l, a, b, x1, and λ.

[0037] (l,a,b,x1:λ)=(6,1,2,42:5.75189…) (l,a,b,x1:λ)=(6,1,2,41:5.75804…)

[0038] Therefore, the inventors calculated (l, a, b) to be (6, 1, 2), and for simplicity, calculated λ to be 5.75.

[0039] Next, using these calculated values, the inventors calculated m and n that minimize the value of (Equation 4) as (m, n)=(13, 26).

[0040] Figure 2 is a graph showing the cumulative distribution difference functions defined by (n,λ)=(6,5.75), (n,λ)=(13,5.75), and (n,λ)=(26,5.75). In Figure 2, each circular point corresponds to the first cumulative distribution difference function F(x;6,5.75) defined by (n,λ)=(6,5.75), each square point corresponds to the second cumulative distribution difference function F(x;13,5.75) defined by (n,λ)=(13,5.75), and each diamond point corresponds to the third cumulative distribution difference function F(x;26,5.75) defined by (n,λ)=(26,5.75).

[0041] Next, the inventors obtained x1, y1, and z1 that minimize (Equation 4) using Mathematica 13.0 based on the conditions l < m < n, x1 + y1 + z1 = 50, and (l,m,n:λ)=(6,13,26:5.75).

[0042] And the inventors calculated (x1,y1,z1)=(42,4,4).

[0043] As a result, the inventors obtained the mathematical formula shown in the following (Equation 6).

[0044] 42F(k;6,5.75)+4F(k;13,5.75)+4F(k;26,5.75) (Equation 6)

[0045] Figure 3 is a graph showing the values indicated by (Equation 6) when k = 1, 2, 3, 4, 5, 6, 7, 8, 9 and the number of onset cases in the actual group G1 superimposed thereon.

[0046] In Figure 3, the horizontal axis indicates 0.8× years, the solid line indicates the values indicated by (Equation 6) when k = 1, 2, 3, 4, 5, 6, 7, 8, 9, and the dashed line indicates the number of onset cases in group G1.

[0047] Here, the mean squared error between (Equation 6) and the actual number of cases in group G1 when k = 1, 2, 3, 4, 5, 6, 7, 8, 9 is 4.23323..., the coefficient of determination is 0.992524..., and the adjusted coefficient of determination is 0.970095...

[0048] As shown in FIG. 3, it can be seen that the formula (6) is a formula that fits well to the frequency distribution of the actual number of cases in group G1.

[0049] In this way, the inventors found the function (first function) expressed by the above (Equation 6) as a search function applicable to subjects classified into group G1.

[0050] Next, the inventors determined from the frequency distribution table shown in FIG. 1 that for group G2, the number of onsets g2(k) for k=1 to 9 was g2(1)=19 (g2(1) is the number of onsets in an onset period of 0.8 years or less), g2(2)=38 (g2(2) is the number of onsets in an onset period of more than 0.8 years and less than 1.6 years), g2(3)=29 (g2(3) is the number of onsets in an onset period of more than 1.6 years and less than 2.4 years), g2(4)=16 (g2(4) is the number of onsets in an onset period of more than 2.4 years and less than 3.2 years), and g2(5)=7 (g2(5) is the number of onsets in an onset period of more than 3.2 years and less than 4.4 years). g2(6) = 5 (g2(6) is the number of cases with an onset period of more than 4.0 years and less), g2(7) = 2 (g2(7) is the number of cases with an onset period of more than 4.8 years and less), g2(8) = 0 (g2(8) is the number of cases with an onset period of more than 5.6 years and less), g2(9) = 0 (g2(9) is the number of cases with an onset period of more than 6.4 years and less), and g2(10) = 0 (g2(10) is the number of cases with an onset period of more than 6.4 years and less). Therefore, using the values ​​of g2(1) to g2(9), we decided to find x2, y2, and z2 that minimize the value of the following (Equation 7).

[0051] Here, x2, y2, and z2 are natural numbers that satisfy the equation x2+y2+z2=116.

[0052]

number

[0053] As in the case of group G1, the inventors used Mathematica 13.0 to calculate x2, y2, and z2 that minimize the value of (Equation 7) as (x2, y2, z2) = (55, 49, 12).

[0054] As a result, the inventors obtained the following formula (Formula 8).

[0055] 55F(k;6,5.75)+49F(k;13,5.75)+12F(k;26,5.75) (Formula 8)

[0056] FIG. 4 is a graph showing the values ​​indicated by (Equation 8) when k=1, 2, 3, 4, 5, 6, 7, 8, and 9 superimposed on the actual number of cases in group G2.

[0057] In FIG. 4, the horizontal axis represents 0.8×years, the solid lines represent the values ​​indicated by (Equation 8) when k=1, 2, 3, 4, 5, 6, 7, 8, 9, and the dashed lines represent the number of incidents in group G2.

[0058] Here, the mean square error between (Equation 8) and the actual number of cases in group G2 when k = 1, 2, 3, 4, 5, 6, 7, 8, 9 is 6.22818..., the coefficient of determination is 0.995806..., and the coefficient of determination adjusted for degrees of freedom is 0.983222....

[0059] As shown in FIG. 4, it can be seen that the formula (8) is a formula that fits well to the frequency distribution of the actual number of cases in group G2.

[0060] In this way, the inventors found the function (second function) expressed by the above (Equation 8) as a search function applicable to subjects classified into group G2.

[0061] Next, the inventors determined from the frequency distribution table shown in FIG. 1 that for group G3, the number of onsets g3(k) for k=1 to 9 was g3(1)=10 (g3(1) is the number of onsets in an onset period of 0.8 years or less), g3(2)=25 (g3(2) is the number of onsets in an onset period of more than 0.8 years and less than 1.6 years), g3(3)=28 (g3(3) is the number of onsets in an onset period of more than 1.6 years and less than 2.4 years), g3(4)=15 (g3(4) is the number of onsets in an onset period of more than 2.4 years and less than 3.2 years), and g3(5)=3 (g3(5) is the number of onsets in an onset period of more than 3.2 years and less than 4.4 years). g3(6) = 3 (g3(6) is the number of cases with an onset period of more than 4.0 years and less than 4.8 years), g3(7) = 0 (g3(7) is the number of cases with an onset period of more than 4.8 years and less than 5.6 years), g3(8) = 1 (g3(8) is the number of cases with an onset period of more than 5.6 years and less than 6.4 years), g3(9) = 0 (g3(9) is the number of cases with an onset period of more than 6.4 years and less than 7.2 years). Therefore, using the values ​​of g3(1) to g3(9), we decided to find x3, y3, and z3 that minimize the value of the following (Equation 9).

[0062] Here, x3, y3, and z3 are natural numbers that satisfy the equation x3+y3+z3=85.

[0063]

number

[0064] As in the cases of groups G1 and G2, the inventors used Mathematica 13.0 to calculate x3, y3, and z3 that minimize the value of (Equation 9) as (x3, y3, z3) = (31, 50, 4).

[0065] As a result, the inventors obtained the following mathematical formula (Formula 10).

[0066] 31F(k;6,5.75)+50F(k;13,5.75)+4F(k;26,5.75) (Formula 10)

[0067] FIG. 5 is a graph showing the values ​​indicated by (Equation 10) when k=1, 2, 3, 4, 5, 6, 7, 8, and 9 superimposed on the actual number of cases in group G3.

[0068] In FIG. 5, the horizontal axis represents 0.8×years, the solid lines represent the values ​​indicated by (Equation 10) when k=1, 2, 3, 4, 5, 6, 7, 8, 9, and the dashed lines represent the number of incidents in group G3.

[0069] Here, the mean square error between (Equation 10) and the actual number of cases in group G2 when k = 1, 2, 3, 4, 5, 6, 7, 8, 9 is 5.6465..., the coefficient of determination is 0.994058..., and the coefficient of determination adjusted for the degrees of freedom is 0.976231....

[0070] As shown in FIG. 5, it can be seen that the formula (Formula 10) is a formula that fits well to the frequency distribution of the actual number of cases in group G3.

[0071] In this way, the inventors found the function (third function) expressed by the above (Equation 10) as a search function applicable to subjects classified into group G3.

[0072] Next, for groups G1, G2, and G3, based on (Equation 6), (Equation 8), and (Equation 10), x1+x2+x3=128, y1+y2+y3=103, and z1+z2+z3=20, the inventors obtained the following formula (Equation 11) by superimposing these three formulas.

[0073] 128F(k;6,5.75)+103F(k;13,5.75)+20F(k;26,5.75) (Formula 11)

[0074] FIG. 6 is a graph showing the values ​​indicated by (Equation 11) when k=1, 2, 3, 4, 5, 6, 7, 8, 9 superimposed on the frequency distribution of the actual number of incidents in groups G1, G2, and G3.

[0075] In FIG. 6, the horizontal axis represents 0.8×years, the solid lines represent the values ​​given by (Equation 11) when k=1, 2, 3, 4, 5, 6, 7, 8, 9, and the dashed lines represent the number of cases in groups G1, G2, and G3.

[0076] Here, the mean square error between (Equation 11) when k = 1, 2, 3, 4, 5, 6, 7, 8, 9 and the actual number of cases in groups G1, G2, and G3 is 21.0938..., the coefficient of determination is 0.997295..., and the coefficient of determination adjusted for the degrees of freedom is 0.989178....

[0077] As shown in FIG. 6, it can be seen that the formula (11) is a formula that fits well to the frequency distribution of the actual number of cases in groups G1, G2, and G3.

[0078] In this way, the inventors found the function (fourth function) expressed by the above (Equation 11) as a search function that applies to subjects classified into groups G1, G2, and G3.

[0079] Based on the specific findings thus obtained, the inventors conducted further research and came up with an Alzheimer's dementia onset prediction device and an Alzheimer's dementia onset prediction method according to one embodiment of the present disclosure described below.

[0080] The device for predicting the onset time of Alzheimer's disease dementia according to one embodiment of the present disclosure is a device for predicting the hippocampal volume [mm ] of a subject with mild cognitive impairment relative to the Alzheimer's Disease Assessment Scale-Cognition score of the subject. 3 ] and a cumulative distribution function 1-G(x;n,λ) for a variable x, which is determined by an integer n and a positive real number λ.

number

[0081] The above-described Alzheimer's disease onset prediction device can calculate predicted onset time information indicating the time when a subject with mild cognitive impairment is predicted to develop Alzheimer's disease.

[0082] Here, the memory unit may store multiple cumulative distribution functions, and the predicted onset time information calculation unit may calculate the predicted onset time information based on a first function obtained by weighting and adding multiple cumulative distribution differential functions based on each of the multiple cumulative distribution functions with a predetermined first weighting coefficient when the volume score value indicates a value in a first range, and may calculate the predicted onset time information based on a second function obtained by weighting and adding the multiple cumulative distribution differential functions with a predetermined second weighting coefficient different from the first weighting coefficient when the volume score value indicates a value in a second range that does not overlap with the first range.

[0083] Here, the plurality of cumulative distribution functions stored in the storage unit may include a first cumulative distribution function determined by n=6 and λ=5.75, a second cumulative distribution function determined by n=13 and λ=5.75, and a third cumulative distribution function determined by n=26 and λ=5.75.

[0084] Here, when the volume score value indicates a value in a third range that does not overlap with the first range and the second range, the predicted onset time information calculation unit further calculates the predicted onset time information based on a third function obtained by weighting and adding the plurality of cumulative distribution difference functions with a predetermined third weighting coefficient different from the first weighting coefficient and the second weighting coefficient, wherein the first range is a range less than 101, the second range is a range not less than 101 and not more than 160, and the third range is a range not less than 160 and not more than 266, and the first weighting coefficient is a first cumulative distribution function based on the first cumulative distribution function. The weighting coefficient may be a weighting coefficient that weights and adds together a distribution difference function, a second cumulative distribution difference function based on the second cumulative distribution function, and a third cumulative distribution difference function based on the third cumulative distribution function in a ratio of 42:4:4, the second weighting coefficient may be a weighting coefficient that weights and adds together the first cumulative distribution difference function, the second cumulative distribution difference function, and the third cumulative distribution difference function in a ratio of 55:49:12, and the third weighting coefficient may be a weighting coefficient that weights and adds together the first cumulative distribution difference function, the second cumulative distribution difference function, and the third cumulative distribution difference function in a ratio of 31:50:4.

[0085] Here, the cumulative distribution functions stored in the memory unit include a first cumulative distribution function determined by n=6 and λ=5.75, a second cumulative distribution function determined by n=13 and λ=5.75, and a third cumulative distribution function determined by n=26 and λ=5.75, and when the volume score value is in a range less than 266, the predicted onset time information calculation unit may calculate the predicted onset time information based on a fourth function obtained by weighting and adding a first cumulative distribution difference function based on the first cumulative distribution function, a second cumulative distribution difference function based on the second cumulative distribution function, and a third cumulative distribution difference function based on the third cumulative distribution function in a ratio of 128:103:20.

[0086] A method for predicting the onset time of Alzheimer's disease according to one embodiment of the present disclosure includes: calculating a cumulative distribution function 1-G(x;n,λ) for a variable x, where the cumulative distribution function 1-G(x;n,λ) is determined by an integer n and a positive real number λ;

number

[0087] According to the method for predicting the onset time of Alzheimer's disease having the above configuration, it is possible to calculate predicted onset time information indicating the time when a subject with mild cognitive impairment is predicted to develop Alzheimer's disease.

[0088] A specific example of an Alzheimer's disease onset prediction device according to one aspect of the present disclosure will be described below with reference to the drawings. Each embodiment shown here represents a specific example of the present disclosure. Therefore, the numerical values, shapes, components, the arrangement and connection of the components, steps (processes), and the order of steps shown in the following embodiments are merely examples and are not intended to limit the present disclosure. Furthermore, each figure is a schematic diagram and is not necessarily an exact illustration. In each figure, substantially identical components are assigned the same reference numerals, and redundant explanations will be omitted or simplified.

[0089] (Embodiment) An Alzheimer's disease onset prediction device according to an embodiment is described below. The Alzheimer's disease onset prediction device calculates predicted onset information indicating when a subject with mild cognitive impairment is predicted to develop Alzheimer's disease, and outputs the calculated predicted onset information to an external device.

[0090] <Configuration> FIG. 7 is a block diagram showing an example of the configuration of an Alzheimer's disease onset prediction device 1 according to the embodiment.

[0091] The Alzheimer's disease onset prediction device 1 is realized, for example, by a computer device having a CPU (Central Processing Unit), memory, a storage device, various interfaces, etc., in which the CPU executes a program loaded into the memory.

[0092] As shown in FIG. 7, the Alzheimer's disease onset prediction device 1 includes an acquisition unit 10, a storage unit 20, a predicted onset time information calculation unit 30, and an output unit 40.

[0093] The acquisition unit 10 calculates the hippocampal volume [mm ] of a subject with mild cognitive impairment relative to the Alzheimer's Disease Assessment Scale-Cognition score of the subject. 3 ] to obtain a volume score value (Cog_Vol) that indicates the ratio of

[0094] Here, the volume score value (Cog_Vol) acquired by the acquisition unit 10 is a value defined by (Equation 2).

[0095] The volume of the subject's hippocampus can be measured, for example, by performing an MRI examination on the subject and performing image processing on the MRI image of the head obtained as a result of the MRI examination.

[0096] The acquisition unit 10 may, for example, be provided with a communication interface and acquire the volume score value (Cog_Vol) from an external device via the communication interface, or may, for example, be provided with a memory interface that reads data from a portable memory (e.g., a USB memory) and acquire the volume score value (Cog_Vol) from the portable memory via the memory interface, or may, for example, be provided with an input / output interface that accepts operations by a user using the Alzheimer's disease onset prediction device 1 and acquire the volume score value (Cog_Vol) by accepting data input operations by the user via the input / output interface.

[0097] Alternatively, the acquisition unit 10 may acquire the volume score value (Cog_Vol) by, for example, acquiring the Alzheimer's Disease Assessment Scale-cognitive score and the hippocampal volume from outside, instead of directly acquiring the volume score value (Cog_Vol) from outside, and calculating the volume score value (Cog_Vol) from the acquired Alzheimer's Disease Assessment Scale-cognitive score and the hippocampal volume.

[0098] The storage unit 20 stores a cumulative distribution function 1-G(x;n,λ) for a variable x, which is determined by an integer n and a positive real number λ.

number

[0099] Here, the description will be given assuming that the storage unit 20 stores a plurality of cumulative distribution functions.

[0100] More specifically, the cumulative distribution functions stored in the memory unit 20 will be described as including a first cumulative distribution function 1-G(x;6,5.75) determined by n=6 and λ=5.75, a second cumulative distribution function 1-G(x;13,5.75) determined by n=13 and λ=5.75, and a third cumulative distribution function 1-G(x;26,5.75) determined by n=26 and λ=5.75.

[0101] The predicted onset time information calculation unit 30 calculates predicted onset time information indicating the time when the subject is predicted to develop Alzheimer's disease based on the volume score value (Cog_Vol) acquired by the acquisition unit 10 and the cumulative distribution functions stored in the memory unit 20 (here, multiple cumulative distribution functions including a first cumulative distribution function 1-G(x;6,5.75) determined when n=6 and λ=5.75, a second cumulative distribution function 1-G(x;13,5.75) determined when n=13 and λ=5.75, and a third cumulative distribution function 1-G(x;26,5.75) determined when n=26 and λ=5.75).

[0102] More specifically, when the volume score value (Cog_Vol) acquired by the acquisition unit 10 indicates a value in a first range, the predicted onset time information calculation unit 30 calculates predicted onset time information based on a first function obtained by weighting and adding, with a predetermined first weighting coefficient, a plurality of cumulative distribution difference functions based on each of a plurality of cumulative distribution functions stored in the storage unit 20; and when the volume score value (Cog_Vol) acquired by the acquisition unit 10 indicates a value in a second range that does not overlap with the first range, the predicted onset time information calculation unit 30 calculates predicted onset time information based on a first function obtained by weighting and adding, with a predetermined first weighting coefficient, a plurality of cumulative distribution difference functions based on each of a plurality of cumulative distribution functions stored in the storage unit 20. Predicted onset time information is calculated based on a second function obtained by weighting and adding the functions with a predetermined second weighting coefficient different from the first weighting coefficient, and when the volume score value (Cog_Vol) acquired by the acquisition unit 10 indicates a value in a third range that does not overlap with the first range and the second range, predicted onset time information is calculated based on a third function obtained by weighting and adding the above-mentioned multiple cumulative distribution differential functions with a predetermined third weighting coefficient different from the first weighting coefficient and the second weighting coefficient.

[0103] Here, the first range is a range less than 101, the second range is a range equal to or greater than 101 and less than 160, and the third range is a range equal to or greater than 160 and less than 266, and the first weighting coefficient is a weighting coefficient that weights a first cumulative distribution difference function F(x; 6, 5.75) based on the first cumulative distribution function 1-G(x; 6, 5.75), a second cumulative distribution difference function F(x; 13, 5.75) based on the second cumulative distribution function 1-G(x; 13, 5.75), and a third cumulative distribution difference function F(x; 26, 5.75) based on the third cumulative distribution function 1-G(x; 26, 5.75) in a ratio of 42:4:4. The first weighting coefficient is a weighting coefficient for weighting and adding the first cumulative distribution difference function F(x; 6, 5.75), the second cumulative distribution difference function F(x; 13, 5.75), and the third cumulative distribution difference function F(x; 26, 5.75) in a ratio of 55:49:12, and the third weighting coefficient is a weighting coefficient for weighting and adding the first cumulative distribution difference function F(x; 6, 5.75), the second cumulative distribution difference function F(x; 13, 5.75), and the third cumulative distribution difference function F(x; 26, 5.75) in a ratio of 31:50:4.

[0104] As an example, the predicted onset time information calculation unit 30 may (1) calculate the predicted onset time information based on the formula shown in (Formula 6) when the volume score value (Cog_Vol) acquired by the acquisition unit 10 is in the range of less than 101, (2) calculate the predicted onset time information based on the formula shown in (Formula 8) when the volume score value (Cog_Vol) acquired by the acquisition unit 10 is in the range of 101 or more and less than 160, and (3) calculate the predicted onset time information based on the formula shown in (Formula 10) when the volume score value (Cog_Vol) acquired by the acquisition unit 10 is in the range of 160 or more and less than 266.

[0105] In this case, for example, the predicted onset time information calculation unit 30 calculates a function F1(k) shown in the following (Equation 12) which is obtained by normalizing the formula shown in (Equation 6), a function F2(k) shown in the following (Equation 13) which is obtained by normalizing the formula shown in (Equation 8), and a function F3(k) shown in the following (Equation 14) which is obtained by normalizing the formula shown in (Equation 10).

[0106] F1(k)=(1 / 50)×{42F(k;6,5.75)+4F(k;13,5.75)+4F(k;26,5.75)} (Formula 12)

[0107] F2(k)=(1 / 116)×{55F(k;6,5.75)+49F(k;13,5.75)+12F(k;26,5.75)} (Formula 13)

[0108] F3(k)=(1 / 85)×{31F(k;6,5.75)+50F(k;13,5.75)+4F(k;26,5.75)} (Formula 14)

[0109] Then, the predicted onset time information calculation unit 30 (1) calculates predicted onset time information indicating that the predicted probability of developing Alzheimer's disease between ((k-1) × 0.8) years and (k × 0.8) years is F1(k) for each of k = 1, 2, 3, 4, 5, 6, 7, 8, 9 when the volume score value (Cog_Vol) acquired by the acquisition unit 10 is in the range of less than 101, and (2) calculates predicted onset time information indicating that the predicted probability of developing Alzheimer's disease between ((k-1) × 0.8) years and (k × 0.8) years is F1(k) for each of k = 1, 2, 3, 4, 5, 6, 7, 8, 9 when the volume score value (Cog_Vol) acquired by the acquisition unit 10 is in the range of 101 or more and less than 160. (3) if the volume score value (Cog_Vol) acquired by the acquisition unit 10 is in the range of 160 or more and less than 266, for each of k=1, 2, 3, 4, 5, 6, 7, 8, 9, predicted onset time information is calculated indicating that the predicted probability of developing Alzheimer's disease between ((k-1) x 0.8) years and (k x 0.8) years is the value indicated by F3(k).

[0110] Alternatively, the predicted onset time information calculation unit 30 (1) calculates predicted onset time information indicating that the probability density of the time when Alzheimer's disease is predicted to occur is F1(k) when the volume score value (Cog_Vol) acquired by the acquisition unit 10 is in the range of less than 101, (2) calculates predicted onset time information indicating that the probability density of the time when Alzheimer's disease is predicted to occur is F2(k) when the volume score value (Cog_Vol) acquired by the acquisition unit 10 is in the range of 101 or more and less than 160, and (3) calculates predicted onset time information indicating that the probability density of the time when Alzheimer's disease is predicted to occur is F3(k) when the volume score value (Cog_Vol) acquired by the acquisition unit 10 is in the range of 160 or more and less than 266.

[0111] The output unit 40 outputs the predicted onset time information calculated by the predicted onset time information calculation unit 30 to the outside.

[0112] The output unit 40 may, for example, be provided with a communication interface and output the predicted onset time information to an external device via the communication interface, or may, for example, be provided with a memory interface for writing data to a portable memory (e.g., a USB memory) and write the predicted onset time information to the portable memory via the memory interface, or may, for example, be provided with a display device (e.g., a display) and display an image based on the predicted onset time information on the display device.

[0113] <Operation> Hereinafter, the operation of the Alzheimer's disease onset prediction device 1 configured as described above will be described with reference to the drawings.

[0114] The Alzheimer's disease onset prediction device 1 performs a first Alzheimer's disease onset prediction process for predicting the time when a subject with mild cognitive impairment will develop Alzheimer's disease.

[0115] FIG. 8 is a flowchart of the first Alzheimer's dementia onset prediction process performed by the Alzheimer's dementia onset prediction device 1.

[0116] The first Alzheimer's disease onset prediction process is started, for example, when a user using the Alzheimer's disease onset prediction device 1 performs an operation on the Alzheimer's disease onset prediction device 1 to start the first Alzheimer's disease onset prediction process.

[0117] When the first Alzheimer's disease onset prediction process is started, the acquisition unit 10 calculates the hippocampal volume [mm 3 A volume score value (Cog_Vol) indicating the ratio of [ ] is obtained (step S10).

[0118] When the volume score value (Cog_Vol) is acquired by the acquisition unit 10, the predicted onset time information calculation unit 30 calculates predicted onset time information indicating the time when the subject is predicted to develop Alzheimer's disease based on the volume score value (Cog_Vol) acquired by the acquisition unit 10 and the cumulative distribution function stored in the memory unit 20 (step S20).

[0119] More specifically, when the volume score value (Cog_Vol) acquired by the acquisition unit 10 indicates a value in a first range, the predicted onset time information calculation unit 30 calculates predicted onset time information based on a first function obtained by weighting and adding together multiple cumulative distribution differential functions based on each of the multiple cumulative distribution functions stored in the memory unit 20 with a predetermined first weighting coefficient; when the volume score value (Cog_Vol) acquired by the acquisition unit 10 indicates a value in a second range, the predicted onset time information calculation unit 30 calculates predicted onset time information based on a second function obtained by weighting and adding together the multiple cumulative distribution differential functions with a predetermined second weighting coefficient; and when the volume score value (Cog_Vol) acquired by the acquisition unit 10 indicates a value in a third range, the predicted onset time information calculation unit 30 calculates predicted onset time information based on a third function obtained by weighting and adding together the multiple cumulative distribution differential functions with a predetermined third weighting coefficient.

[0120] Here, the first range is a range less than 101, the second range is a range equal to or greater than 101 and less than 160, and the third range is a range equal to or greater than 160 and less than 266, and the first weighting coefficient is a weighting coefficient that weights a first cumulative distribution difference function F(x; 6, 5.75) based on the first cumulative distribution function 1-G(x; 6, 5.75), a second cumulative distribution difference function F(x; 13, 5.75) based on the second cumulative distribution function 1-G(x; 13, 5.75), and a third cumulative distribution difference function F(x; 26, 5.75) based on the third cumulative distribution function 1-G(x; 26, 5.75) in a ratio of 42:4:4. The first weighting coefficient is a weighting coefficient for weighting and adding the first cumulative distribution difference function F(x; 6, 5.75), the second cumulative distribution difference function F(x; 13, 5.75), and the third cumulative distribution difference function F(x; 26, 5.75) in a ratio of 55:49:12, and the third weighting coefficient is a weighting coefficient for weighting and adding the first cumulative distribution difference function F(x; 6, 5.75), the second cumulative distribution difference function F(x; 13, 5.75), and the third cumulative distribution difference function F(x; 26, 5.75) in a ratio of 31:50:4.

[0121] When the predicted onset time information calculation section 30 calculates the predicted onset time information, the output section 40 outputs the predicted onset time information calculated by the predicted onset time information calculation section 30 to the outside (step S30).

[0122] When the process of step S30 ends, the Alzheimer's dementia onset prediction device 1 ends the first Alzheimer's dementia onset prediction process.

[0123] <Consideration> The Alzheimer's disease onset prediction device 1 configured as described above can calculate predicted onset time information indicating the time when a subject with mild cognitive impairment is predicted to develop Alzheimer's disease.

[0124] Therefore, by using the Alzheimer's disease onset prediction device 1 configured as described above, it is possible to predict when a subject with mild cognitive impairment will develop Alzheimer's disease.

[0125] (Variation) Hereinafter, a description will be given of an Alzheimer's disease onset prediction device according to a modified example, which is configured by partially changing the configuration of the Alzheimer's disease onset prediction device 1 according to the embodiment.

[0126] In the modified Alzheimer's dementia onset prediction device, components that are similar to those in the Alzheimer's dementia onset prediction device 1 have already been explained, so they will be assigned the same symbols and their detailed explanations will be omitted, and the explanation will focus on the differences from the Alzheimer's dementia onset prediction device 1.

[0127] <Configuration> FIG. 9 is a block diagram showing an example of the configuration of an Alzheimer's disease onset prediction device 1A according to a modified example.

[0128] As shown in Figure 9, the Alzheimer's disease onset prediction device 1A is configured by replacing the predicted onset time information calculation unit 30 of the Alzheimer's disease onset prediction device 1 of the embodiment with a predicted onset time information calculation unit 30A.

[0129] Similar to the predicted onset time information calculation unit 30, the predicted onset time information calculation unit 30A calculates predicted onset time information indicating when a subject is predicted to develop Alzheimer's disease based on the volume score value (Cog_Vol) acquired by the acquisition unit 10 and the cumulative distribution functions (here, a plurality of cumulative distribution functions including a first cumulative distribution function 1-G(x;6,5.75) determined when n=6, λ=5.75, a second cumulative distribution function 1-G(x;13,5.75) determined when n=13, λ=5.75, and a third cumulative distribution function 1-G(x;26,5.75) determined when n=26, λ=5.75) stored in the storage unit 20. However, the predicted onset time information calculation unit 30A differs from the predicted onset time information calculation unit 30 in the method of calculating the predicted onset time information.

[0130] That is, more specifically, when the volume score value (Cog_Vol) acquired by the acquisition unit 10 indicates a value in the fourth range, the predicted onset time information calculation unit 30A calculates the predicted onset time information based on a fourth function obtained by weighting and adding multiple cumulative distribution difference functions based on each of the multiple cumulative distribution functions stored in the memory unit 20 with a predetermined fourth weighting coefficient.

[0131] Here, the fourth range is a range less than 266, and the fourth weighting coefficient is a weighting coefficient that weights and adds the first cumulative distribution difference function F(x; 6, 5.75) based on the first cumulative distribution function 1-G(x; 6, 5.75), the second cumulative distribution difference function F(x; 13, 5.75) based on the second cumulative distribution function 1-G(x; 13, 5.75), and the third cumulative distribution difference function F(x; 26, 5.75) based on the third cumulative distribution function 1-G(x; 26, 5.75) in a ratio of 128:103:20.

[0132] As an example, the predicted onset time information calculation unit 30A may calculate the predicted onset time information based on the formula shown in (Equation 11) when (1) the volume score value (Cog_Vol) acquired by the acquisition unit 10 is in a range less than 266.

[0133] In this case, for example, the predicted onset time information calculation unit 30A calculates a function F4(k) shown in the following (Equation 15), which is obtained by normalizing the mathematical expression shown in (Equation 11).

[0134] F4(k)=(1 / 251)×{128F(k;6,5.75)+103F(k;13,5.75)+20F(k;26,5.75)} (Formula 15)

[0135] Then, when the volume score value (Cog_Vol) acquired by the acquisition unit 10 is in the range of less than 266, the predicted onset time information calculation unit 30A calculates predicted onset time information indicating that the predicted probability of onset of Alzheimer's disease between ((k-1) x 0.8) years and (k x 0.8) years is F4(k), for each of k = 1, 2, 3, 4, 5, 6, 7, 8, 9.

[0136] Alternatively, when the volume score value (Cog_Vol) acquired by the acquisition unit 10 is in a range less than 266, the predicted onset time information calculation unit 30A calculates predicted onset time information indicating that the probability density of the time when Alzheimer's disease is predicted to occur is F4(k).

[0137] <Operation> Hereinafter, the operation of the Alzheimer's disease onset prediction device 1A configured as described above will be described with reference to the drawings.

[0138] The Alzheimer's dementia onset prediction device 1A performs a second Alzheimer's dementia onset prediction process in which some of the processing has been changed from the first Alzheimer's dementia onset prediction process, instead of the first Alzheimer's dementia onset prediction process performed by the Alzheimer's dementia onset prediction device 1 of the embodiment.

[0139] FIG. 10 is a flowchart of the second Alzheimer's dementia onset prediction process performed by Alzheimer's dementia onset prediction device 1A.

[0140] As shown in FIG. 10, the second Alzheimer's dementia onset prediction process is a process in which the process of step S20 in the first Alzheimer's dementia onset prediction process is changed to the process of step S20A.

[0141] Therefore, the processing of step S20A will be mainly described here.

[0142] When the processing of step S10 is completed, the predicted onset time information calculation unit 30A calculates predicted onset time information indicating the time when the subject is predicted to develop Alzheimer's disease based on the volume score value (Cog_Vol) acquired by the acquisition unit 10 and the cumulative distribution function stored in the memory unit 20 (step S20A).

[0143] More specifically, when the volume score value (Cog_Vol) acquired by the acquisition unit 10 indicates a value in the fourth range, the predicted onset time information calculation unit 30A calculates predicted onset time information based on a fourth function obtained by weighting and adding multiple cumulative distribution difference functions based on each of the multiple cumulative distribution functions stored in the memory unit 20 with a predetermined fourth weighting coefficient.

[0144] Here, the fourth range is a range less than 266, and the fourth weighting coefficient is a weighting coefficient that weights and adds the first cumulative distribution difference function F(x; 6, 5.75) based on the first cumulative distribution function 1-G(x; 6, 5.75), the second cumulative distribution difference function F(x; 13, 5.75) based on the second cumulative distribution function 1-G(x; 13, 5.75), and the third cumulative distribution difference function F(x; 26, 5.75) based on the third cumulative distribution function 1-G(x; 26, 5.75) in a ratio of 128:103:20.

[0145] When the process of step S20A ends, the second Alzheimer's disease onset time prediction process proceeds to the process of step S30.

[0146] <Consideration> The Alzheimer's disease onset prediction device 1A having the above configuration can calculate predicted onset time information indicating the time when a subject with mild cognitive impairment is predicted to develop Alzheimer's disease.

[0147] Therefore, by using the Alzheimer's disease onset prediction device 1A configured as described above, it is possible to predict when a subject with mild cognitive impairment will develop Alzheimer's disease.

[0148] (supplement) As described above, the present disclosure has been described based on the embodiments and modifications as examples of the technology disclosed in the present application. However, the present disclosure is not limited to these embodiments and modifications. As long as they do not deviate from the spirit of the present disclosure, various modifications that would occur to a person skilled in the art to these embodiments or modifications, or forms constructed by combining components of different embodiments or modifications, may also be included within the scope of one or more aspects of the present disclosure.

[0149] A general or specific aspect of the present disclosure may be realized as a system, an apparatus, a method, an integrated circuit, a program, or a non-transitory recording medium such as a computer-readable CD-ROM, or may be realized as any combination of a system, an apparatus, a method, an integrated circuit, a program, and a non-transitory recording medium. [Industrial Applicability]

[0150] The present disclosure is widely applicable to devices and methods for predicting when a subject with mild cognitive impairment is expected to develop Alzheimer's disease. [Explanation of symbols]

[0151] 1. 1A Alzheimer's disease onset prediction device 10 Acquisition Department 20 Memory section 30, 30A Predicted onset time information calculation section 40 Output section

Claims

1. The hippocampal volume [mm ] of subjects with mild cognitive impairment versus the Alzheimer's Disease Rating Scale-cognition score of the subjects. ３ an acquisition unit that acquires a volume score value indicating a ratio of The cumulative distribution function 1-G(x; n, λ) for a variable x, which is determined by an integer n and a positive real number λ, is [Equation 1] a storage unit that stores the cumulative distribution function determined by a predetermined n and a predetermined λ, a predicted onset time information calculation unit that calculates predicted onset time information indicating a time when the subject is predicted to develop Alzheimer's disease, based on the volume score value acquired by the acquisition unit and the cumulative distribution function stored in the storage unit; and an output unit that outputs the predicted onset time information calculated by the predicted onset time information calculation unit to an outside, the storage unit stores a plurality of cumulative distribution functions, the predicted onset time information calculation unit calculates the predicted onset time information based on a first function obtained by weighting and adding a plurality of cumulative distribution difference functions based on each of the plurality of cumulative distribution functions with a predetermined first weighting coefficient when the volume score value indicates a value in a first range, and calculates the predicted onset time information based on a second function obtained by weighting and adding the plurality of cumulative distribution difference functions with a predetermined second weighting coefficient different from the first weighting coefficient when the volume score value indicates a value in a second range that does not overlap with the first range, the plurality of cumulative distribution functions stored in the storage unit include a first cumulative distribution function determined by n=6 and λ=5.75, a second cumulative distribution function determined by n=13 and λ=5.75, and a third cumulative distribution function determined by n=26 and λ=5.75; the predicted onset time information calculation unit further calculates the predicted onset time information based on a third function obtained by weighting and adding the plurality of cumulative distribution difference functions with a predetermined third weighting coefficient different from the first weighting coefficient and the second weighting coefficient when the volume score value indicates a value in a third range that does not overlap the first range and the second range; the first range is less than 101; the second range is equal to or greater than 101 and less than 160; the third range is equal to or greater than 160 and less than 266; the first weighting coefficient is a weighting coefficient for weighting and adding a first cumulative distribution difference function based on the first cumulative distribution function, a second cumulative distribution difference function based on the second cumulative distribution function, and a third cumulative distribution difference function based on the third cumulative distribution function in a ratio of 42:4:4; the second weighting coefficient is a weighting coefficient for weighting and adding the first cumulative distribution difference function, the second cumulative distribution difference function, and the third cumulative distribution difference function in a ratio of 55:49:12; The third weighting coefficient is a weighting coefficient for weighting and adding the first cumulative distribution difference function, the second cumulative distribution difference function, and the third cumulative distribution difference function in a ratio of 31:50:

4. A device for predicting the onset of Alzheimer's disease.

2. An acquisition unit; The cumulative distribution function 1-G(x; n, λ) for a variable x, which is determined by an integer n and a positive real number λ, is [Equation 2] a storage unit that stores the cumulative distribution function determined by a predetermined n and a predetermined λ, A predicted onset time information calculation unit; an output unit; and an Alzheimer's dementia onset prediction method performed by an Alzheimer's dementia onset prediction device comprising: The acquisition unit acquires a hippocampal volume [mm ] of a subject with mild cognitive impairment relative to the Alzheimer's Disease Assessment Scale-cognition score of the subject. ３ an acquisition step of acquiring a volumetric score value indicative of a predicted onset time information calculation step in which the predicted onset time information calculation unit calculates predicted onset time information indicating the time when the subject is predicted to develop Alzheimer's disease, based on the volume score value acquired in the acquisition step and the cumulative distribution function stored in the storage unit; and an output step in which the output unit outputs the predicted onset time information calculated by the predicted onset time information calculation step to an outside, the storage unit stores a plurality of cumulative distribution functions, In the predicted onset time information calculation step, when the volume score value indicates a value in a first range, the predicted onset time information is calculated based on a first function obtained by weighting and adding a plurality of cumulative distribution difference functions based on each of the plurality of cumulative distribution functions with a predetermined first weighting coefficient, and when the volume score value indicates a value in a second range that does not overlap with the first range, the predicted onset time information is calculated based on a second function obtained by weighting and adding the plurality of cumulative distribution difference functions with a predetermined second weighting coefficient different from the first weighting coefficient, the plurality of cumulative distribution functions stored in the storage unit include a first cumulative distribution function determined by n=6 and λ=5.75, a second cumulative distribution function determined by n=13 and λ=5.75, and a third cumulative distribution function determined by n=26 and λ=5.75; In the predicted onset time information calculation step, when the volume score value indicates a value in a third range that does not overlap with the first range and the second range, the predicted onset time information is calculated based on a third function obtained by weighting and adding the plurality of cumulative distribution difference functions with a third weighting coefficient that is different from the first weighting coefficient and the second weighting coefficient, the first range is less than 101; the second range is equal to or greater than 101 and less than 160; the third range is equal to or greater than 160 and less than 266; the first weighting coefficient is a weighting coefficient for weighting and adding a first cumulative distribution difference function based on the first cumulative distribution function, a second cumulative distribution difference function based on the second cumulative distribution function, and a third cumulative distribution difference function based on the third cumulative distribution function in a ratio of 42:4:4; the second weighting coefficient is a weighting coefficient for weighting and adding the first cumulative distribution difference function, the second cumulative distribution difference function, and the third cumulative distribution difference function in a ratio of 55:49:12; The third weighting coefficient is a weighting coefficient for weighting and adding the first cumulative distribution difference function, the second cumulative distribution difference function, and the third cumulative distribution difference function in a ratio of 31:50:

4. A method for predicting the onset of Alzheimer's disease.

Images