Method for determining cognitive function level of subject using prediction model and system for performing same
The cognitive function prediction model processes biomarker and demographic data to accurately classify cognitive function levels, addressing the need for objective assessment and aiding in the diagnosis and management of cognitive decline.
Patent Information
- Application Number
- PCT/KR2024/006637
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-05-30
- Filing Date
- 2024-05-16
- Publication Date
- 2024-12-05
AI Technical Summary
Current methods for determining cognitive function are not objective or accurate enough, particularly in distinguishing between age-related decline and cognitive function-related diseases like mild cognitive impairment and dementia, necessitating a more precise assessment.
A method and system using a cognitive function prediction model that processes biomarker information, cognitive function test history, and demographic data to predict cognitive function levels, incorporating biomarkers such as tau proteins, amyloid beta, and Neurofilament Light Chain, and applying regression analysis to classify cognitive function indices into predefined levels.
This approach enables detailed and accurate determination of cognitive function, supporting appropriate drug treatments and interventions by providing a probabilistic classification of cognitive function levels, thereby aiding in early diagnosis and management of cognitive decline.
Smart Images

Figure KR2024006637_05122024_PF_FP_ABST
Abstract
Description
Method for determining the level of cognitive function of a subject using a predictive model and a system for performing the same
[0001] The embodiments relate to a method for determining the level of cognitive function of a subject using a predictive model and a system for performing the same.
[0002] Cognitive function refers to the ability of the human brain to efficiently receive, store, and manipulate knowledge and information, resulting in behavior or thought. Cognitive function encompasses virtually all functions the brain can perform, including memory, language, judgment, spatial and temporal perception, concentration, and abstract thinking.
[0003] Declining human cognitive function is a natural phenomenon that occurs with aging. However, excessive cognitive decline may not be due to aging, but rather to cognitive disorders such as mild cognitive impairment or dementia. Therefore, accurately assessing cognitive function levels is crucial for diagnosing cognitive disorders.
[0004] Processes for assessing cognitive function include in-depth interviews, clinical scale assessments, and neurocognitive testing. However, a more objective and accurate assessment of a subject's cognitive function is needed.
[0005] To solve the above-described problem, the present application seeks to provide a method for determining the level of cognitive function of a subject and a system for performing the same.
[0006] A method for determining the level of cognitive function of a box according to one aspect of the present application may be performed by a computing system. The method may include: a step of acquiring at least one piece of information from among biomarker information of the subject and cognitive function test history information of the subject; a step of acquiring demographic information of the subject, wherein the demographic information of the subject includes the subject's age and gender; and a step of inputting at least some of the biomarker information and cognitive function test history information of the subject, and the demographic information of the subject into a pre-learned cognitive function prediction model, thereby calculating a predicted value of the level of cognitive function of the subject.
[0007] In one embodiment, the biomarker information of the subject may include one or more of first biomarker information describing the tau protein of the subject, second biomarker information describing the amyloid beta protein of the subject, and third biomarker information describing the blood concentration of NfL (Neurofilament Light Chain) of the subject.
[0008] In one embodiment, in the first biomarker information, the tau protein of the subject is one or more proteins selected from the group consisting of phosphorated-tau 181, phosphorated-tau 199, phosphorated-231, and total-tau. The second biomarker information may include one or more of information describing the blood concentration of amyloid beta 42 of the subject, the ratio between the blood concentration of amyloid beta 42 of the subject and the blood concentration of amyloid beta 40, and information describing the proportion of amyloid beta 38 among the blood concentration of total amyloid beta (total-Aβ) of the subject.
[0009] In one embodiment, the cognitive function prediction model is a model configured to calculate a probability that the cognitive function index of the subject will be classified into a plurality of preset cognitive function levels and to predict the cognitive function level of the subject based on the calculated classification probability. The step of calculating a predicted value of the cognitive function level of the subject may include the step of inputting an input value vector of the subject and calculating a probability that the cognitive function index of the subject will be classified into a plurality of preset cognitive function levels in the cognitive function prediction model; and the step of outputting the cognitive function level having the highest calculated probability as the predicted value of the cognitive function level of the subject.
[0010] In one embodiment, the probability of each cognitive function index of the subject being classified into multiple cognitive function levels is calculated based on a linear combination of a pre-learned regression coefficient vector and the input value vector. In the cognitive function prediction model, the multiple cognitive function levels are defined by respective level-specific boundary values. The regression coefficient vector and boundary values are learned through a regression analysis method.
[0011] In one embodiment, in the cognitive function prediction model, a linear combination of the regression coefficient vector and the input value vector is expressed as an inner product of the i-dimensional regression coefficient vector and the input value vector, and the input value vector may include one or more of the biomarker information and cognitive function test history information, demographic information, and a preset constant.
[0012] In one embodiment, when the multiple cognitive function levels are set to three or more levels, the probability that the subject's cognitive function index is classified into the lowest level with the least cognitive function decline (Lmin), the highest level with the most severe cognitive function decline (Lmax), and one or more intermediate levels (Lj) between the lowest level and the highest level is calculated using the following mathematical formulas.
[0013] [Mathematical formula]
[0014]
[0015] [Mathematical formula]
[0016]
[0017] [Mathematical formula]
[0018]
[0019] Here, ki is the boundary value for the cognitive function index that distinguishes the cognitive function level of stage i, and i.min, j, and i.max are natural numbers such as i.min <j<i.max의 관계를 가지며, β'는 회귀계수 벡터이고, x는 상기 대상자의 입력 값 벡터이다.
[0020] In one embodiment, the cognitive function prediction model is composed of a plurality of prediction models, each trained to predict the level of cognitive function of the subject by type of cognitive function. The step of calculating the predicted value of the level of cognitive function of the subject may include inputting the input value vector of the subject into each of the plurality of prediction models trained in advance to predict the level of each type of cognitive function, in order to calculate the predicted value of the level of cognitive function of the subject by type. Here, the type of cognitive function may be one or more types of cognitive function selected from the group consisting of memory, visuospatial sense construction, motor coordination, attention evaluating the level of arousal, language ability, perception ability, and comprehensive cognitive function that synthesizes the abilities of each type thereof.
[0021] In one embodiment, at least some of the biomarker information of the subject is obtained through a blood test method. The blood test method comprises the steps of: flowing a fluid sample containing the analyte over a substrate having a microchamber array formed thereon, on the surface of which a first capture agent that specifically binds to the analyte is fixed; binding the analyte to the first capture agent within each microchamber of the microchamber array; flowing a second capture agent that specifically binds to the analyte and binds to a signal-generating agent over the substrate to cause the analyte to react with the second capture agent; flowing the signal-generating agent over the substrate to cause the signal-generating agent to bind to the second capture agent; flowing a substrate solution that reacts with the signal-generating agent to generate a fluorescent signal over the substrate; flowing a hydrophobic solvent over the substrate to cause the signal-generating agent and the substrate solution to react inside the microchamber, and removing the substrate solution outside the microchamber. It may be obtained by a method for detecting an analyte present in a trace amount in a fluid sample, including a step of counting and detecting the number of microchambers in which the fluorescent signal is generated.
[0022] According to another aspect of the present application, a computer-readable recording medium can record a program that causes a method of determining the level of cognitive function of a subject according to the above-described embodiments.
[0023] The system according to the embodiments of the present application can determine the level of cognitive function of a subject in detail and accurately by using the subject's biomarker information and demographic information.
[0024] As a result, the system supports providing a solution for applying an appropriate drug treatment depending on the subject's cognitive function level.
[0025] The effects that can be obtained from the present invention are not limited to the effects mentioned above, and other effects not mentioned can be clearly understood by a person having ordinary skill in the art to which the present disclosure pertains from the description below.
[0026] To more clearly explain the technical solutions of embodiments of the present invention or prior art, the drawings required for the description of the embodiments are briefly introduced below. It should be understood that the drawings below are for the purpose of illustrating embodiments of the present specification and are not intended to be limiting. Furthermore, for clarity of explanation, some elements may be depicted with various modifications, such as exaggeration or omission, in the drawings below.
[0027] Figure 1 is a flowchart of a method for determining the level of cognitive function of a subject according to one aspect of the present application.
[0028] FIG. 2 is a schematic diagram illustrating five levels of cognitive function according to cognitive function indicators according to various embodiments of the present application.
[0029] FIG. 3 is a flowchart of a process for setting independent variables to be included as components in the input value vector according to various embodiments of the present application.
[0030] FIG. 4 illustrates a computer system (30) capable of performing the method of FIG. 1 according to various embodiments of the present application.
[0031] FIG. 5 is a schematic diagram of a device for detecting particles of an analyte substance present in trace amounts in a fluid sample according to a first embodiment of the present invention.
[0032] Figure 6 illustrates in more detail the analysis chip (300) according to the first embodiment of the present invention.
[0033] Figure 7 conceptually illustrates a process for detecting a target substance for analysis according to the first embodiment of the present invention.
[0034] FIG. 8 is a flowchart of a method for detecting particles of an analyte substance present in trace amounts in a fluid sample according to a first embodiment of the present invention.
[0035] FIG. 9 is a schematic diagram of a device for detecting particles of an analyte substance present in trace amounts in a fluid sample according to a second embodiment of the present invention.
[0036] Figure 10 illustrates in more detail the analysis chip (301) according to the second embodiment of the present invention.
[0037] FIG. 11 illustrates a combined form of a microparticle, a first capture material, an analyte, and a signal generating material according to a second embodiment of the present invention.
[0038] Figure 12 shows in detail the state in which fine particles are captured in a microchamber arranged in a detection space according to the second embodiment of the present invention.
[0039] Figure 13 conceptually illustrates a process for detecting a substance to be analyzed according to a second embodiment of the present invention.
[0040] Figures 14 and 15 illustrate a preprocessing process of a method for detecting particles of an analysis target material according to a second embodiment of the present invention.
[0041] Figure 16 is a flowchart of a method for detecting particles of an analyte substance present in trace amounts in a fluid sample according to a second embodiment of the present invention.
[0042] Figure 17 is a schematic diagram of a device for detecting particles of an analyte substance present in trace amounts in a fluid sample according to a third embodiment of the present invention.
[0043] Figure 18 illustrates in more detail the analysis chip (302) according to the third embodiment of the present invention.
[0044] Figure 19 illustrates in more detail the microwell microparticles according to the third embodiment of the present invention.
[0045] Figure 20 conceptually illustrates a process for detecting a substance to be analyzed according to a third embodiment of the present invention.
[0046] Figure 21 is a flowchart of a method for detecting particles of an analyte substance present in trace amounts in a fluid sample according to a third embodiment of the present invention.
[0047] FIG. 22 is a drawing showing a microchamber containing microparticles and a microchamber generating a fluorescent signal according to a second embodiment of the present invention.
[0048] Figure 23 is a graph showing the relationship between the concentration (pg / mL) of phosphorylated tau 181 and the signal (au).
[0049] Hereinafter, embodiments of the present application will be described in detail with reference to the drawings.
[0050] However, this disclosure is not intended to limit the present disclosure to a specific embodiment, but should be understood to encompass various modifications, equivalents, and / or alternatives of the embodiments of the present disclosure. In connection with the description of the drawings, similar reference numerals may be used for similar components.
[0051] In this specification, expressions such as “has,” “may have,” “includes,” or “may include” indicate the presence of a feature (e.g., a component such as a number, function, operation, step, part, element, and / or component), and do not exclude the presence or addition of additional features.
[0052] When a component is referred to as being "connected" or "connected" to another component, it should be understood that it may be directly connected or connected to that other component, but that there may be other components intervening. Conversely, when a component is referred to as being "directly connected" or "connected" to another component, it should be understood that there are no other components intervening.
[0053] The terms "first," "second," "first," or "second" used in various embodiments may describe various components, regardless of order and / or importance, and do not limit the components. These terms may be used to distinguish one component from another. For example, "first component" and "second component" may represent different components, regardless of order or importance.
[0054] The embodiments of the singular expression used in this specification also include embodiments of the plural expressions, unless the phrases related to the singular expression clearly indicate a contrary meaning.
[0055] The expression "configured to" as used herein can be used interchangeably with, for example, "suitable for," "having the capacity to," "designed to," "adapted to," "made to," or "capable of." The term "configured to" does not necessarily mean something is "specifically designed to" in terms of hardware. Instead, in some contexts, the expression "a device configured to" can mean that the device, together with other devices or components, is "capable of." For example, the phrase "a processor configured (or set) to perform A, B, and C" may mean a dedicated processor (e.g., an embedded processor) for performing those operations, or a general-purpose processor (e.g., a CPU or application processor) that can perform those operations by executing one or more software programs stored in a memory device.
[0056]
[0057] Figure 1 is a flowchart of a method for determining the level of cognitive function of a subject according to one aspect of the present application.
[0058] The method for determining the cognitive function level of the subject in Fig. 1 can be performed by the system described below.
[0059] Referring to FIG. 1, a method for determining a cognitive function level of a subject includes a step (S100) of acquiring at least one piece of information from among biomarker information and cognitive function test history information of the subject; a step (S200) of acquiring demographic information of the subject; and a step (S400) of inputting at least some of the biomarker information and cognitive function test history information of the subject, and the demographic information of the subject into a pre-learned cognitive function prediction model, thereby calculating a predicted value of the cognitive function level of the subject. In addition, in some embodiments, the method of FIG. 1 may further include a step (S600) of generating an analysis report including the cognitive function level of the subject predicted in step (S400).
[0060] Additionally, in some embodiments, the method of FIG. 1 may further include a step (S500) of generating solution information for the subject based on the predicted level of cognitive function. The solution information may include information regarding pharmacological and / or non-pharmacological treatments to alleviate or prevent cognitive decline in the subject. Then, the analysis report in step (S600) may further include the solution information for the subject generated in step (S500).
[0061]
[0062] Subject information
[0063] The biomarker information obtained in the above step (S100) represents the blood concentration of a biomarker that can determine cognitive function decline.
[0064] In various embodiments of the present application, the biomarker information of the subject may include one or more of first biomarker information describing the blood concentration of tau protein of the subject, second biomarker information describing the blood concentration of amyloid beta protein of the subject, and third biomarker information describing the blood concentration of NfL (Neurofilament Light Chain) of the subject.
[0065] In some embodiments, the biomarker information of the subject may further include one or more of the following biomarker information: fourth biomarker information describing the Glial fibrillary acidic protein (GFAP) of the subject, fifth biomarker information describing the blood concentration of Neurogranin protein of the subject, and sixth biomarker information describing the blood concentration of sTREM2 of the subject.
[0066] Additionally, in some embodiments, the biomarker information of the subject may further include seventh biomarker information describing the hippocampal volume of the subject, and / or eighth biomarker information describing the centriole score of the subject.
[0067] The above first biomarker information and second biomarker information can be expressed as one or more values.
[0068] In the first biomarker information, the tau protein of the subject may be one or more proteins selected from the group consisting of phosphored-tau 181, phosphored-tau 199, phosphored-tau 217, phosphored-tau 231, phosphored-tau 235, phosphored-tau 396-404, and total tau. The variable representing the first biomarker may be expressed as one or more detailed variables according to the tau protein that the blood concentration indicates. For example, when n proteins are treated as the tau protein of the subject, the first biomarker information may include 1-1 biomarker information representing the blood concentration of the first tau protein, 1-2 biomarker information representing the blood concentration of the second tau protein, and 1-n biomarker information representing the blood concentration of the nth tau protein.
[0069] In various embodiments of the present application, the tau protein of the subject in the first biomarker information included in the biomarker information of the subject may be one or more proteins selected from the group consisting of phosphorated-tau 181, phosphorated-tau 199, phosphorated-231, and total tau. The first biomarker information may include one or more values of the blood concentration of each of the one or more tau proteins and the total blood concentration.
[0070] In some embodiments, the subject's tau protein may be selected based on the subject's demographic information. The subject's tau protein may be a specific tau protein selected from the group of tau proteins described above that is pre-associated with a combination of parameters within the subject's demographic information.
[0071] The second biomarker information may include one or more values among the blood concentrations of each of one or more amyloid beta proteins, and one or more values among the transformed values based on at least a portion of the blood concentrations of a plurality of different amyloid beta proteins. The variable representing the second biomarker may be expressed as one or more sub-variables according to the amyloid beta protein associated with the variable value. For example, the second biomarker information may include the blood concentration of one amyloid beta protein as the 2-1 biomarker information, and / or the transformed values based on the blood concentrations of different amyloid beta proteins as the 2-2 biomarker information.
[0072] In various embodiments of the present application, the second biomarker information may include one or more of information describing the blood concentration of amyloid beta 42 of the subject, information describing the ratio between the blood concentration of amyloid beta 42 and the blood concentration of amyloid beta 40 of the subject, and information describing the proportion of amyloid beta 38 among the blood concentration of total amyloid beta (total-Aβ) of the subject.
[0073] The second biomarker information may include a blood concentration value that directly represents a blood concentration of a specific amyloid beta, such as a blood concentration of amyloid beta 42. Alternatively, the second biomarker information may include a conversion value, such as a ratio between the blood concentration of amyloid beta 42 and the blood concentration of amyloid beta 40 of the subject, or a proportion of amyloid beta 38 among the total blood concentration of amyloid beta of the subject.
[0074] The above total amyloid beta refers to amyloid beta that can be detected in the blood of the subject, and is not limited to a specific amyloid beta used as biomarker information. The above total amyloid beta may be two or more proteins selected from the group consisting of amyloid beta 37, amyloid beta 38, amyloid beta 39, amyloid beta 40, and amyloid beta 42.
[0075] Additionally, in some embodiments, when the second biomarker information includes information about amyloid beta 40 of the subject, the second biomarker information may include information that indirectly describes the blood concentration of amyloid beta 40 of the subject, excluding information that directly describes the blood concentration of amyloid beta 40 of the subject. That is, the second biomarker information is not expressed as describing only the blood concentration of amyloid beta 40 of the subject. The second biomarker information may include information related to amyloid beta 40 of the subject in the form of a ratio between the blood concentration of amyloid beta 42 and the blood concentration of amyloid beta 40 of the subject.
[0076] The genetic information of the subject may include ApoE information describing the genotype of ApoE of the subject and whether the subject possesses the ApoE4 gene.
[0077] In the above step (S100), the biomarker information of the subject can be obtained through various testing methods. Specifically, the various methods may be a blood test, magnetic resonance elastography (MRE), or amyloid PET. When the results of the testing method applied to the subject are analyzed, the analysis results describing the aforementioned biomarker can be input into the system as the subject's biomarker information (S100). Here, the blood test may be an ELISA, a digital ELISA, or any other testing method capable of obtaining the aforementioned biomarker information.
[0078] In various embodiments of the present application, the biomarker information of the subject acquired in the step (S100) may be information acquired by a method for detecting an analyte present in a trace amount in a fluid sample to quantitatively and accurately detect a biomarker present in blood at a concentration of several pg / ml. A blood test method for detecting an analyte present in a trace amount in a fluid sample to quantitatively and accurately detect a biomarker present in blood at a concentration of several pg / ml is described in more detail with reference to FIGS. 5 to 23 below.
[0079] In the above step (S100), the subject's cognitive function test history information represents the results of cognitive function tests performed by the subject in the past prior to the method of FIG. 1. The past cognitive function tests may be tests using the cognitive function prediction model below or cognitive function tests performed at a medical institution.
[0080] The results of the above-mentioned past cognitive function tests may include the test date and the level of cognitive function tested for each past test date. The above-mentioned level of cognitive function may be expressed as a normal state without cognitive impairment, or as a cognitive function-related disease diagnosed when cognitive function declines by a corresponding level compared to the above-mentioned normal state.
[0081] In the above step (S200), demographic information is the personal information of the subject for demographic purposes. The demographic information includes gender and age.
[0082] In various embodiments of the present application, the subject's demographic information may further include medical history information. This medical history information may include past medical history information describing conditions the subject has suffered but recovered from, and current medical history information describing conditions the subject is currently suffering from but has not yet recovered from at the time the demographic information is provided. Such medical history information may include, but is not limited to, hyperlipidemia and hypertension.
[0083] In various embodiments of the present application, the demographic information of the subject may further include information on the subject's education level, and one or more pieces of family member information describing one or more of the total number of family members of the subject and the number of family members of the subject with a disease associated with cognitive decline. The disease associated with cognitive decline may be Alzheimer's disease.
[0084] The higher the educational level of the subject, the lower the likelihood of developing Alzheimer's disease. The higher the proportion of Alzheimer's disease patients among the population, the higher the likelihood of developing Alzheimer's disease.
[0085] In the above step (S200), such demographic information can be entered as the demographic information of the subject.
[0086] In the above step (S400), the cognitive function prediction model is configured to receive biomarker information of the subject, at least one piece of information from the subject's cognitive function test history, and demographic information as input value vectors. In addition, the cognitive function prediction model has model parameters learned in advance to produce a cognitive function level corresponding to the subject's cognitive function index, which indicates the possibility of cognitive function decline of the subject, as a predicted value of the subject's cognitive function level.
[0087]
[0088] Cognitive function indicators
[0089] The cognitive function prediction model is modeled using a cognitive function index (U) that indicates the possibility of cognitive function decline.
[0090] The above cognitive function indicators and the likelihood of cognitive decline have a linear relationship, either proportional or inverse. For example, the higher the cognitive function indicator, the higher the likelihood of cognitive decline and the higher the likelihood of dementia or mild cognitive impairment. The lower the cognitive function indicator, the lower the likelihood of cognitive decline and the higher the likelihood of the subject remaining in a normal state, free of dementia or mild cognitive impairment.
[0091] In the above cognitive function prediction model, the cognitive function index (U) is a dependent variable for the input values of the cognitive function prediction model. In the cognitive function prediction model, the final output value is the probability that the subject indicated by the input values belongs to each cognitive function level, and the probability of the subject at each level is calculated through the dependent variable.
[0092] The input values of the above cognitive function prediction model are independent variables for the cognitive function index (U). The independent variables are detailed information comprising the biomarker information, demographic information, and cognitive function test history information. For example, the input values may include the biomarker information of step (S100) and the statistical information.
[0093] The cognitive function test history information of the subject included as an independent variable in the above input value can indicate the test date and the past cognitive function level of the subject for each test date.
[0094] The above input values can be expressed in vector form and may be referred to as input value vectors in this specification.
[0095] In various embodiments of the present application, in the regression analysis-based cognitive function prediction model, the input value vector may include one or more of the biomarker information and cognitive function test history information, demographic information, and a preset constant. For example, the input value vector is composed of independent variables composed of biomarker information (e.g., first to third biomarker information), demographic information (e.g., gender, age), cognitive function test history information, and a preset constant. The constant is expressed as x0, and the biomarker information and demographic information are expressed as x n When expressed as {x0, x1, .., x n}, where n is the number of biomarker information and demographic information details. In some embodiments, the constant has a value of 1. Then, the input value vector is {1, x1, .., x n} can be expressed as .
[0096] The above cognitive function prediction model may be generated through various modeling methods capable of producing the output values from the aforementioned input value vector. The various modeling methods may be statistical-based modeling methods or artificial neural network-based modeling methods.
[0097] The above statistical-based modeling method may be a regression analysis-based modeling method or another statistical analysis-based modeling method.
[0098] If the above cognitive function prediction model is modeled using a regression analysis-based modeling method, the cognitive function index (U) may be defined based on a linear combination of regression coefficient vectors corresponding to the input value vectors. Specifically, in the above cognitive function prediction model, the cognitive function index (U) may be based on a linear combination of an input value vector and a regression coefficient vector for the input value vector, and an error term.
[0099] The above regression coefficient vector is a vector composed of regression coefficient values and is a model parameter of the cognitive function prediction model.
[0100] The above regression coefficient vector includes a regression coefficient β0 corresponding to a constant. The regression coefficient component β0 corresponding to a constant in the above regression coefficient vector is learned as a value for correcting errors in regression analysis. If the constant of the input value vector is preset to 1, the linear combination of the input value vector and the regression coefficient vector includes a constant for correcting errors in regression analysis.
[0101] In the above regression coefficient vector, each regression coefficient value corresponds to an input value of the same dimension in the input value vector. The input value vector is {1, x1, .., x n}, the regression coefficient vector is {β0, β1, .., β n} can be expressed in the form of .
[0102] The remaining components excluding the constant in the above input value vector, specifically, the regression coefficient components β1, .., β corresponding to biomarker information, demographic information, and / or cognitive function test history information n It shows the influence of biomarker information, demographic information, and cognitive function test history information on the presence or absence of cognitive function decline.
[0103] The above regression coefficient vector and the above input value vector can be expressed as an i-dimensional vector. Here, i is the number of constant components, one or more information components of biomarker information and cognitive function test history information, and demographic information components. If n1 biomarker information components, n2 cognitive function test history information components, and n3 demographic information components are included in the input value vector, i can be 1+n1+n2+n3. As described above, since the cognitive function test history information is expressed as the past cognitive function level according to the test date and the test performed on the test date, n2 can be 2.
[0104] The linear combination value of the input value vector and the regression coefficient vector can be expressed as the inner product of the i-dimensional input value vector and the regression coefficient vector.
[0105] The above error term represents an unobservable attribute that influences the determination of cognitive decline. This error term indicates the influence of latent variables, not used in learning, on the determination of cognitive decline, in addition to the aforementioned independent variables. This error term can be referred to as a disturbance, meaning it disrupts the stable relationship between the independent and dependent variables.
[0106] In some embodiments, the cognitive function index (U) may be defined by the following mathematical formula.
[0107]
[0108] Here, β' is the regression coefficient vector, x is the input value vector, and ε represents the error term.
[0109] The value of the cognitive function index (U) defined in this way depends on the value of the learned regression coefficient vector.
[0110]
[0111] Prediction of cognitive function level based on cognitive function indicators
[0112] The cognitive function prediction model is configured to calculate the probability that the cognitive function index (U) of the subject will be classified into a plurality of preset cognitive function levels, and to predict the cognitive function level of the subject based on the calculated classification probability. The cognitive function prediction model is generated based on a linear combination of a regression coefficient vector and the input value vector.
[0113] The final output value of the above cognitive function prediction model may be any one of a plurality of preset cognitive function levels.
[0114] The above-described multiple cognitive function levels can be preset to two or more levels. The above-described cognitive function prediction model is a model including model parameters that are pre-learned to calculate the probability that the cognitive function index (U) of a subject having the input value vector is classified into a cognitive function level predicted to be possessed by the subject.
[0115] In the above cognitive function prediction model, Stage 1 represents the initial stage, a normal state without cognitive impairment. As stages increase from Stage 1, the level of cognitive decline increases. For example, Stage 2 represents a more severe decline in cognitive function than Stage 1, and Stage 3 represents a further decline in cognitive function than Stage 2.
[0116] The above multiple cognitive function levels can be defined by boundary values based on a cognitive function index (U). Adjacent levels can share boundary values. Specifically, the minimum boundary value of a particular level is the maximum boundary value of the previous level, and the maximum boundary value of a particular level is the minimum boundary value of the next level.
[0117] If there are three or more cognitive function levels, the definition of each cognitive function level can be expressed as follows.
[0118]
[0119] Here, L min The level of cognitive decline is the lowest, and Lmax is the most severe level of cognitive decline. Subjects in the normal state are those in the L min It corresponds to .
[0120] ki is the boundary value for the cognitive function index (U) that distinguishes the level of cognitive function of stage i, and i is the number of boundaries, k i.max is the boundary value of the highest level, i.e. the maximum boundary value, k i.min is the boundary value of the lowest step, i.e. the minimum boundary value, k jrepresents the intermediate boundary value(s) between the maximum and minimum boundary values. i.min, j, i.max are natural numbers, and i.min <j<i.max의 관계를 가진다. 예를 들어, 전체 인지기능 수준이 5단계이고, 정상 상태를 의미하는, 인지기능 저하가 가장 낮은 최저 단계는 1단계로 설정되면, i.min은 1, j는 2, 3, i.max는 4일 수 있다.
[0121] In various embodiments of the present application, the error term (ε) in the mathematical expression 1 may be modeled as an extreme value distribution in the form of a normal distribution. For example, the error term (ε) may be modeled as a Type II extreme value distribution. Then, the cognitive function prediction model may be configured to calculate, for each of the plurality of cognitive function levels for which the preset boundary values are defined, a probability that the cognitive function index (U) having the error term (ε) modeled as an extreme value distribution belongs to one of the predetermined levels of cognitive function, based on a linear combination of the regression coefficient vector of the cognitive function prediction model and the input value vector, and a maximum value of a boundary among the boundary values defining individual cognitive function levels. In the cognitive function prediction model, the probability that the cognitive function index (U) belongs to the highest level with the most severe cognitive function decline, the probability that it belongs to the lowest level with the least cognitive function decline, and the probability that it belongs to a level between the highest level and the lowest level may each be calculated using different mathematical expressions.
[0122] In some embodiments, the mathematical formulas of the cognitive function prediction model, which calculate the probability that a cognitive function index (U) having an error term (ε) modeled as an extreme value distribution belongs to each of a plurality of predefined cognitive function levels, may be expressed by the following mathematical formulas 3 to 5, respectively.
[0123]
[0124]
[0125]
[0126] As mentioned above, β' is a regression coefficient vector and x is an input value vector.
[0127] Then, the cognitive function prediction model is the lowest level (L) of the cognitive function index (U). min ) is calculated using mathematical formula 3, and the cognitive function index (U) is the highest level (L max ) can be calculated using mathematical formula 5, and the probability that the cognitive function index (U) belongs to the intermediate stage (Lj) can be calculated using mathematical formula 3. The cognitive function prediction model can be configured so that the sum of the probabilities of being classified by cognitive function level, calculated using mathematical formulas 3 to 5, is 1.
[0128] FIG. 2 is a schematic diagram illustrating five levels of cognitive function according to cognitive function indicators according to various embodiments of the present application.
[0129] Referring to Figure 2, the cognitive function levels can be set to a total of five levels (i.e., i=1, 2, 3, 4). Level 1 represents the normal state, and level 5 represents the most severe level of cognitive decline. Then, using Equation 2, the cognitive function levels of each of the five levels can be expressed using the following mathematical equation.
[0130]
[0131] Referring to the above mathematical formulas 3 to 5, the cognitive function prediction model can calculate the probability that the subject's cognitive function index (U) falls into the cognitive function level of the first to fifth stages, respectively, through the following mathematical formulas.
[0132]
[0133]
[0134]
[0135]
[0136]
[0137] Meanwhile, the step values indicating the aforementioned cognitive function levels (step 1, step 2, etc.) are merely exemplary and may be modified in various forms. Depending on the embodiments, the step values indicating the cognitive function levels may be set in reverse. In some alternative embodiments, the Lmax step (e.g., step 5) may represent a normal state without cognitive impairment, and a gradually decreasing step from the Lmax step may represent a higher level of cognitive function decline. For example, step 2 may represent a more severe level of cognitive function decline than step 3, and step 1 may represent a more severe level of cognitive function decline than step 2, and may be set to represent the most severe level of cognitive function decline.
[0138] The step of inputting the cognitive function index calculated in the above step (S400) into a pre-learned cognitive function prediction model to calculate a predicted value of the cognitive function level of the subject may include a step of calculating the probability that the cognitive function index (U) of the subject will be classified into a plurality of preset cognitive function levels in the cognitive function prediction model; and a step of outputting the cognitive function level with the highest calculated probability as the predicted value of the cognitive function level of the subject.
[0139] For example, if the cognitive function index (U) of the subject is most likely to belong to stage 1, the cognitive function prediction model can predict the level of cognitive function of the subject as stage 1.
[0140] In various embodiments of the present application, the cognitive function can be predicted by type. The cognitive function can be classified into memory, visuospatial awareness, motor coordination, attention assessing arousal levels, language ability, perception, or a comprehensive cognitive function that combines the abilities of each of these types.
[0141] The above cognitive function prediction model may be composed of a plurality of prediction models, each trained to predict the level of cognitive function of the subject by type of cognitive function. Then, the step (S400) may be to input at least some of the biomarker information and cognitive function test history information of the subject, and the demographic information of the subject as input value vectors to each of the plurality of prediction models trained in advance to predict the level of each type of cognitive function, in order to produce a prediction value of the level of cognitive function of the subject by type.
[0142] In the above step (S400), the types of cognitive functions that can be predicted may be one or more types of cognitive functions selected from the group consisting of memory, visuospatial sense construction ability, motor coordination ability, attention that evaluates the level of arousal, language ability, perception ability, and comprehensive cognitive functions that synthesize abilities of each type. Then, the number of multiple prediction models in the above step (S400) is prepared equal to the number of types of cognitive functions that can be predicted.
[0143] Specifically, the step (S400) includes a step of inputting the subject's input value vector into a pre-trained first prediction model to predict the subject's first type of cognitive function level; and a step of inputting the subject's input value vector into a pre-trained second prediction model to predict the subject's second type of cognitive function level.
[0144] The above first prediction model is a cognitive function level prediction model that has been pre-trained to predict the level of the first type of cognitive function. The first prediction model is trained using the truth value for the level of the first type of cognitive function of the sample subject.
[0145] The second prediction model is a cognitive function level prediction model that has been pre-trained to predict the level of the second type of cognitive function. The second prediction model is trained using the truth value for the level of the second type of cognitive function of the sample subject.
[0146] For example, the first type of cognitive function may be a comprehensive cognitive function. Then, the second type of cognitive function may be each of the remaining types of cognitive functions selected from the group consisting of memory, visuospatial sense organization, motor coordination, attention assessing arousal levels, language ability, perception ability, and a comprehensive cognitive function that synthesizes each of these types of abilities.
[0147]
[0148] Training (building and validating) a predictive model
[0149] The cognitive function prediction model used in the above step (S400), the component values in the regression coefficient vector in the cognitive function prediction model (e.g., β0, β1, .., β) n ) has values learned using the training data set. In addition, the boundary values (k) that define the cognitive function level in the cognitive function prediction model i ) also has values learned using the above training data set.
[0150] The above training data set is composed of a plurality of training samples, and each training sample may include pairs of sample input values and truth values corresponding to the sample input values.
[0151] The above sample input values are a vector of input values of sample subjects for model learning.
[0152] The truth value corresponding to the above sample input value may be the true cognitive function level of the sample subject who provided the sample input value. For example, the truth value may be expressed as a value representing the cognitive function level corresponding to the sample subject among the cognitive function levels of steps 1 to 5 illustrated in FIG. 2.
[0153] The above training data set contains a vector of regression coefficients and boundary values (k) defining the level of cognitive function. i) and the first data set used to estimate the estimated regression coefficient vector, boundary values (k i ) may include a second data set used to verify the first data set. The sample subjects of the first data set and the sample subjects of the second data set may be at least partially different.
[0154] The process of creating a cognitive function prediction model using the above training data set may include a process of creating a cognitive function prediction model using a first data set; and a process of verifying the created cognitive function prediction model using a second data set.
[0155] For pairs of sample input values and corresponding truth values of the first data set, a coefficient that can connect the sample input values and the corresponding truth values is estimated through regression analysis to generate a regression coefficient vector, and the generated regression coefficient vector is applied to mathematical equations 3 to 5 for predicting a cognitive function level to generate a cognitive function prediction model. Specifically, for the predicted value output by the cognitive function prediction model processing the sample input values, and the truth value for the predicted value, that is, the truth value corresponding to the sample input values, a regression coefficient vector and boundary values (k) are generated by connecting the sample values and the output predicted values through regression analysis. i ) is estimated, and a cognitive function prediction model can be created.
[0156] The above regression coefficient vector, boundary values (k i ) can be estimated through maximum likelihood estimation, but is not limited thereto and can be estimated through other regression coefficient estimation methods.
[0157] The second data set also includes pairs of validation sample input values and their corresponding truth values. Regression coefficient vector, boundary values (k i) is estimated, the prediction accuracy of the model can be analyzed using the second data set to verify the generated cognitive function prediction model. Specifically, the verification process inputs sample input values in the second data set into the previously generated cognitive function prediction model, and the regression coefficient vector and the estimated boundary values (k) estimated in the model generation process i ), the estimated values of the probability that the cognitive function index (U) will be classified are respectively obtained, and the accuracy of the model is verified by comparing the estimated value with the highest probability among the estimated values of the calculated probabilities with the truth value corresponding to the verification sample input value. The estimated values of the probability that the cognitive function index (U) will be classified may be probability values produced by inputting the verification sample input value vector into the cognitive function prediction model modeled by the above mathematical expressions 3 to 5 (for example, the above mathematical expressions 6 to 11). The truth value compared with the estimated value is a value paired with the sample input value that produces the estimated value.
[0158] When the input value vector based on the above steps (S100 to S200) is input into the cognitive function prediction model that has completed learning in this way, the probability of each being classified into a plurality of preset cognitive function levels is calculated, and the cognitive function level with the highest probability is predicted as the cognitive function level of the subject, and can be output from the cognitive function prediction model.
[0159] Specifically, after the values of the regression coefficient vector and the boundary values are determined through learning, at least one of the biomarker information and the cognitive function test history information of the subject to be judged as having cognitive function decline, and demographic information are acquired (S100, S200), an input value vector of the subject may be formed from the demographic information along with at least one of the biomarker information and the cognitive function test history information of the subject, and the formed input value vector of the subject may be input into the learned cognitive function prediction model (S400). For example, an input value vector x based on gender and age in the demographic information, first biomarker information (e.g., blood concentration of phosphored-tau 181), second biomarker information (e.g., ratio of amyloid beta 42 / 40) in the biomarker information, and a preset constant (e.g., 1) may be generated and input into the cognitive function prediction model. Then, the cognitive function level of the subject is calculated (S400).
[0160]
[0161] In various embodiments of the present application, one or more other cognitive function levels that exhibit cognitive function decline compared to a cognitive function level that exhibits a normal state among the preset cognitive function levels may be expressed as a cognitive function-related disease that is diagnosed when cognitive function declines by that level compared to the normal state.
[0162] In Figure 2, the fifth stage, which is a normal state level without cognitive impairment, can be designated as a normal stage.
[0163] Meanwhile, Stage 1, the most severe level of cognitive decline, can be designated as severe dementia. Then, in the cognitive function prediction model, the cognitive function index (U) of a subject who is either a severe dementia patient or whose cognitive function level is similar to that of a severe dementia patient based on normal conditions is calculated to have the highest probability of being classified as Stage 1, and ultimately, the subject's cognitive function level can be output as severe dementia.
[0164] Additionally, Stage 2, which is a level of cognitive function that is relatively less impaired compared to Stage 1, can be designated as mild dementia. Then, in the cognitive function prediction model, the cognitive function index (U) of a subject who is either a mild dementia patient or whose cognitive function level is similar to that of a mild dementia patient based on the normal state is calculated to have the highest probability of being classified as Stage 2, and ultimately, the subject's cognitive function level can be output as a mild dementia level.
[0165] Additionally, Stage 3, which is a level of cognitive function that is relatively less impaired compared to Stage 2, can be designated as Mild Cognitive Impairment. Then, in the cognitive function prediction model, the cognitive function index (U) of a subject who is a patient with mild cognitive impairment or whose cognitive function level is equally impaired as a patient with mild cognitive impairment based on the normal state is calculated to have the highest probability of being classified as Stage 3, and ultimately, the subject's cognitive function level can be output as Mild Cognitive Impairment.
[0166] In addition, the fourth stage, which is a level of cognitive function that is relatively less impaired compared to the third stage, can be designated as Subjective Cognitive Decline (SCD). Then, in the cognitive function prediction model, the cognitive function index (U) of a subject who is a patient with subjective cognitive decline or whose cognitive function level is reduced to the same level as a patient with mild cognitive impairment based on the normal state is calculated to have the highest probability of being classified as the third stage, and ultimately, the cognitive function level of the subject can be output as the level of mild cognitive impairment.
[0167] In some embodiments, the multiple cognitive function levels may be set to two levels. Then, the first level is considered the lowest level, and the second level is considered the highest level. The probability of being classified as a cognitive function level of the first level is calculated using Equation 3, and the probability of being classified as a cognitive function level of the second level is calculated using Equation 5. In this case, the first level may be designated as a level of cognitive decline, including mild cognitive impairment and dementia levels as shown in FIG. 2. The second level may be designated as a level of cognitive normality.
[0168]
[0169] Referring back to Figure 1, the core of the solution provided in relation to the treatment of cognitive decline in step (S500) is whether a treatment can be applied to the subject. In step (S500), solution information for the subject with the predicted level of cognitive function is generated using the prediction results of step (S400).
[0170] The above step (S500) may be to generate information on whether drug treatment is applicable to the subject having the predicted level of cognitive function, or to generate information on whether cognitive rehabilitation treatment is applicable.
[0171] Information on whether the above drug treatment is applicable may include the level of cognitive function of the subject predicted in the above step (S400) and the name of the drug treatment to be applied.
[0172] The step (S500) may determine a drug treatment that matches the cognitive function level of the subject predicted in the step (S400) from a list of drug treatment agents according to the pre-stored cognitive function level. In some embodiments, multiple drug treatment agents may be associated with each cognitive function level. In the drug treatment list, each drug treatment agent may be associated with a combination of a cognitive function level, one or more pieces of information from among suitable biomarker information and cognitive function test history information, and demographic information. When a plurality of drug treatment agents that match the cognitive function level of the subject are searched, a specific drug treatment agent that matches the biomarker information of the subject in the step (S100) and one or more pieces of information from the past cognitive function test history information, and the demographic information of the subject in the step (S200), among the multiple searched drug treatment agents, may be determined as the target drug treatment agent to be applied to the subject.
[0173] The information regarding the applicability of the above cognitive rehabilitation treatment indicates a specific type of cognitive rehabilitation treatment suitable for the subject, based on the predicted level of cognitive function, among the preset types of cognitive rehabilitation treatment. The type of cognitive rehabilitation treatment may include brain activity activation, exercise, infrared irradiation, electrical stimulation, or other treatment methods. The brain activity activation treatment involves solving puzzles or math problems or other quizzes.
[0174] In the above step (S600), the analysis report is a report that supports the early diagnosis of cognitive decline in the subject. The analysis report can be generated by applying information to be inserted into a pre-saved report template that defines the layout structure of the analysis report. The information to be inserted into the report template is preset based on the report template.
[0175] The analysis report includes the predicted level of cognitive function. Furthermore, in some embodiments, the analysis report may further include one or more of the subject's biomarker information and cognitive function test history information, as well as the subject's demographic information. Furthermore, in some embodiments, the analysis report may further include information regarding the subject's potential for drug therapy.
[0176]
[0177] Setting the components of the input value vector
[0178] In some embodiments, the independent variables to be included as components in the input value vector may be preset.
[0179] As mentioned above, the input value vector is {1, x1, .., x n}, it may include one or more of the biomarker information and past cognitive function test history information, demographic information, and preset constants. In the cognitive function prediction model, variable components of the input value vector x1, .., x n can be pre-specified before step (S400).
[0180] For example, the input value vector may be set to include first biomarker information and may be set to have information describing a blood concentration of a specific tau protein(s) selected from the group consisting of phosphored-tau 181, phosphored-tau 199, phosphored-tau 217, phosphored-tau 231, phosphored-tau 235, phosphored-tau396-404, and total-tau as the first biomarker information. Then, even if the blood information includes a tau protein other than the specific tau protein, only the specific tau protein may be included in the input value vector and input to the cognitive function prediction model.
[0181] FIG. 3 is a flowchart of a process for setting independent variables to be included as components in the input value vector according to various embodiments of the present application.
[0182] Referring to FIG. 3, the process of setting independent variables to be included as components in the input value vector may include a step (S310) of defining a plurality of input variable combinations as independent variables that can be included in the biomarker information, past cognitive function test history information, and demographic information; a step (S320) of generating a combination-specific prediction model by estimating a combination-specific regression coefficient vector and boundary values for a plurality of preset cognitive function levels using a combination-specific first data set composed of information corresponding to each input variable combination; a step (S330) of verifying the prediction performance of the generated combination-specific prediction model for the cognitive function level using a combination-specific second data set; and a step (S340) of determining the input variable combination with the best prediction performance as the independent variables to be included in the input value vector based on the verification result for each combination-specific prediction model.
[0183] As previously described, various variables based on biomarker information, details contained in cognitive function test history information, and details contained in demographic information can be included as independent variables in the input value vector. All possible combinations of variables generated using these details can be defined as input variable combinations (S310).
[0184] In the above step (S310), different input variable combinations have different variable combinations. Different variable combinations mean that the combinations of sub-variables for the same variable are different. Different combinations (or sub-combinations) mean that the entire combination components are inconsistent.
[0185] For example, since detailed information that can be included in demographic information is gender and age, combinations of input variables such as {gender age, biomarker information}, {gender age, cognitive function test history information}, and {gender age, biomarker information, cognitive function test history information} can be defined.
[0186] Meanwhile, the detailed information that may be included in the biomarker information may be one or more of the first biomarker information to the eighth biomarker information. The first biomarker information and the second biomarker information may also include detailed information. Then, in the above example, a plurality of detailed variable combinations including the first biomarker information may be defined, and the detailed variable combinations included in the plurality of input variable combinations may include a first-first detailed variable combination (phosphorated-tau 181), a first-second detailed variable combination (phosphorated-tau 181, phosphorated-tau 199), a first-third detailed variable combination (phosphorated-tau 181, phosphorated-tau 199, phosphorated-231), and a first-fourth detailed variable combination (phosphorated-tau 181, phosphorated-tau 199, phosphorated-231, and total-tau).
[0187] The first data set for each combination consists of data for multiple sample subjects, and the data for each sample subject consists of pairs of sample values for the corresponding input variable combination; and the true value of the sample subject for which the corresponding sample value was measured.
[0188] In the above example, the first data set for each combination may be a first data set for the first combination, which is composed of a sample value for the first combination {biomarker, gender, age}; and a truth value of the sample subject whose sample value was measured; a first data set for the second combination, which is composed of a sample value for the second combination {cognitive function test history information, gender, age}; and a truth value of the sample subject whose sample value was measured; a first data set for the third combination, which is composed of a sample value for the third combination {biomarker, cognitive function test history information, gender, age}; and a truth value of the sample subject whose sample value was measured; etc. (S320).
[0189] In the above step (S320), the cognitive function prediction model processes the sample input values in the first data set for each combination, outputs a predicted value, and estimates the regression coefficient values in the regression coefficient vector and boundary values for a plurality of preset cognitive function levels through regression analysis on the truth value corresponding to the sample input value, and generates a cognitive function prediction model for each combination using the linear combination of the estimated regression coefficient vector and the input variable combinations for each combination and the boundary values for a plurality of preset cognitive function levels.
[0190] The second data set also includes pairs of sample input values for verification and their corresponding truth values.
[0191] In the above step (S330), a verification sample input value is input into the generated combination-specific cognitive function prediction model, and the estimated value of the corresponding combination-specific prediction model is output, and the accuracy of the model is verified by comparing the estimated value with the true value corresponding to the estimated value. Specifically, the sample input value in the second data set for each combination is input into the generated combination-specific prediction model, and the regression coefficient vector and the estimated boundary values (k) estimated in the model generation process are compared. i) are calculated based on the estimated values of the probability that the cognitive function index of the subject will be classified, and the estimated value with the highest probability among the calculated estimated values of the probability is compared with the true value corresponding to the verification sample input value to verify the model. The verification result may be, but is not limited to, a logloss value.
[0192] If the verification result is analyzed as a logloss value (S330), the input variable combination in the prediction model with the largest logloss value can be set as the final independent variables of the input value vector to be input in the above step (S400). For example, if the verification result (logloss) of the prediction model by combination is 0.223 to 0.693, the input variable combination for the smallest value, 0.223, can be set as the independent variables of the input value vector to be input in the above step (S400).
[0193] In this way, the cognitive function prediction model can provide more accurate risk prediction results by selectively utilizing some of the total information that can be obtained from the input subject's biomarker information.
[0194]
[0195] Predicting the subject's level of cognitive function at a future point in time
[0196] In the above-described embodiments, a cognitive function prediction model that outputs the level of cognitive function predicted to be possessed by the subject at the time when the subject information of steps (S100, S200) was collected was learned and utilized.
[0197] In alternative embodiments, a cognitive function prediction model that outputs a level of cognitive function predicted to be possessed by the subject at a future point in time after a certain period of time from the time at which the subject information of the above steps (S100, S200) was collected can be learned and utilized.
[0198] The structure of the cognitive function prediction model (hereinafter, future prediction model) that predicts the level of cognitive function of the subject at the above-mentioned future point in time is identical to the structure of the cognitive function prediction model (hereinafter, real-time prediction model) that predicts the level of cognitive function of the subject at the time of acquisition of the subject information, which is explained with reference to mathematical equations 1 to 11, and there are differences in some of the configurations and learning processes, and these differences will be mainly described.
[0199] The predicted value output from the above future prediction model is a predicted value at a future point in time that is a specific period of time after the acquisition time of the subject information (S100, S200) constituting the input value vector. For example, if the acquisition time of the subject information in steps (S100, S200) is 2023 and the specific period is 2 years, the subject's cognitive function level is predicted in 2025.
[0200] In the above future prediction model, the cognitive function index (U) indicates the possibility of cognitive function decline of the subject at a future point in time from the time of information acquisition of the subject.
[0201] The above future prediction model, like the real-time prediction model, is a linear combination of the regression coefficient vector and the input value vector and boundary values (k) that define the level of cognitive function. i ) is configured to calculate the probability that the subject's cognitive function index (U) will be classified at the future point in time. Here, the regression coefficient vector and the boundary values for a plurality of preset cognitive function levels have values learned for prediction at the future point in time.
[0202] The training data set of the above future prediction model and the training data set of the above real-time prediction model may also be different.
[0203] As described above, the real-time cognitive function prediction model is trained using a training data set, wherein the training data set is composed of a plurality of training samples, and each training sample may include pairs of sample input values and truth values corresponding to the sample input values. The training data set may include a regression coefficient vector, boundary values (k) defining the cognitive function level, i ) and the first data set used to estimate the estimated regression coefficient vector, boundary values (k i ) is used to verify the second data set.
[0204] Unlike the training data set for the aforementioned real-time prediction model, the individual training samples in the training data set for the future prediction model may each include a sample input value of a sample subject acquired at a first time point and a truth value of the sample subject acquired at a second time point. Here, the sample input value acquired at the first time point refers to the time point of acquisition of subject information providing the sample input value. The first time point may be a blood test or an amyloid PET test time point.
[0205] The second point in time is a future point in time after a certain period of time has elapsed from the first point in time. The point in time at which the truth value is acquired, i.e., the second point in time, may be the point in time at which the true cognitive function level of the subject who provided the sample input value is determined. Therefore, the training samples for the future prediction model can be defined as a combination of the first and second points in time.
[0206] In the training data set for the above-mentioned future prediction model, some sample input values may be acquired at different times than other sample input values. For example, certain sample input values may have been acquired in 2018, while others may have been acquired in 2019.
[0207] Additionally, in the training data set for the aforementioned future prediction model, the acquisition times of some truth values may differ from those of other truth values. For example, a particular truth value may have been acquired in 2020, while another truth value may have been acquired in 2021.
[0208] In various embodiments of the present application, the specific period may be any one of 12 months to 48 months.
[0209] Additionally, the training data set can be subdivided into multiple subsets based on the combination of acquisition points of sample input values and acquisition points of truth values within the training samples. Each subset consists of pairs of truth values acquired at the same second time point, similar to input value vectors acquired at the same first time point.
[0210] For example, the training data set may include subsets corresponding to a first combination and a second combination. The subset corresponding to the first combination has pairs of input value vectors acquired at time point 1-1 and truth values acquired at time point 2-1 as training samples. The subset corresponding to the second combination has pairs of input value vectors acquired at time point 1-2 and truth values acquired at time point 2-2 as training samples.
[0211] Additionally, the training data set may include training samples with different specific periods for each individual training sample. Specifically, the time interval between the sample input value and the acquisition time of the truth value of a sample subject of some subsets among the plurality of subsets may be different from the time interval between the sample input value and the acquisition time of the truth value of a sample subject of other subsets. For example, the training data set may include a first subset having a time interval of 12 months, a second subset having a time interval of 18 months, and a third subset having a time interval of 24 months.
[0212] In various embodiments of the present application, the deviation between specific periods defining each of the subsets may be set to be less than a preset threshold deviation. In some embodiments, the threshold deviation may be 12 months. Then, the training data set for the future prediction model is composed of subsets in which the specific periods between each subset have a deviation of less than 12 months from each other. For example, the deviation between specific periods of each of the subsets may be less than 12 months. If the training data set has as a subset a first set of training samples in which the independent variables (i.e., the input value vector of the sample subject) were acquired in 2019 and the dependent variable (i.e., the truth value of the sample subject) was acquired in 2022, the same training data set may also include as a subset a second set of training samples in which the independent variables were acquired in 2017 and the dependent variable was acquired in 2020. This is because the difference between a specific period (3 years) in the first set and a specific period (3 years) in the second set is less than 12 months.
[0213] On the other hand, the same training data set above does not include a third set of training samples, in which the independent variable was acquired in 2017 and the dependent variable was acquired in 2018, as a subset. This is because a specific period (1 year) of the third set has a deviation (24 months) exceeding 12 months from the specific periods of the first and second sets.
[0214] The process of generating a future prediction model using such a training data set may include a process of generating a future prediction model using a first data set within the training data set; and a process of verifying the generated cognitive function prediction model using a second data set.
[0215] For pairs of sample input values and corresponding truth values of the first data set for a future prediction model, a coefficient that can connect the sample input values and corresponding truth values is estimated through regression analysis to generate a regression coefficient vector, and the generated regression coefficient vector is applied to mathematical equations 3 to 5 for predicting a cognitive function level to generate a cognitive function prediction model. The values of the regression coefficient vector, boundary values (k i ) is learned as an estimated value that can calculate the level of cognitive function of the subject at a future point in time after a certain period of time has passed from the time of acquisition of the input value vector.
[0216] The second data set also includes pairs of validation sample input values and their corresponding truth values. Regression coefficient vector, boundary values (k i ) is estimated, the prediction accuracy of the model can be analyzed using the second data set to verify whether the generated cognitive function prediction model accurately predicts the level of cognitive function of the subject at a future point in time after the specific period of time has elapsed.
[0217] When the information of the subject in the above steps (S100, S200) is input into the future prediction model learned in this way, the level of cognitive function of the subject predicted to have in the future after a certain period of time has passed since the time of acquisition of the information of the subject can be calculated (S300).
[0218]
[0219] System
[0220] FIG. 4 illustrates a computer system (30) capable of performing the method of FIG. 1 according to various embodiments of the present application.
[0221] One or more computer systems (30) perform one or more steps of one or more methods described or illustrated herein. The one or more computer systems (30) provide functionality described or illustrated herein. Software executing on the one or more computer systems (30) performs one or more steps of one or more methods described or illustrated herein, or provides functionality described or illustrated herein. In some embodiments, the computer systems (30) include one or more portions of one or more computer systems. In this specification, reference to a computer system may include a computing device, where appropriate. Furthermore, reference to a computer system may include one or more computer systems, where appropriate.
[0222] This disclosure contemplates any suitable number of computer systems (30). This disclosure contemplates computer systems (30) taking any suitable physical form. By way of example and not limitation, the computer system (30) may be an embedded computer system, a system-on-a-chip (SOC), a single-board computer system (SBC) (e.g., a computer-on-module (COM) or a system-on-module (SOM)), a desktop computer system, a laptop or notebook computer system, an interactive kiosk, a mainframe, a mesh of computer systems, a mobile phone, a personal digital assistant (PDA), a server, a tablet computer system, or a combination of two or more thereof. Where appropriate, the computer system (30) may include one or more computer systems (30); may be integral or distributed; may span multiple locations; may span multiple machines; may span multiple data centers; or may reside in a cloud, which may include one or more cloud components in one or more networks. Where appropriate, one or more computer systems (30) may perform one or more steps of one or more of the methods described or illustrated herein without substantial spatial or temporal limitations. For example, and without limitation, one or more computer systems (30) may perform one or more steps of one or more of the methods described or illustrated herein in real time or in batch mode. Where appropriate, one or more computer systems (30) may perform one or more steps of one or more of the methods described or illustrated herein at different times or locations.
[0223] In various embodiments of the present application, a computer system (30) includes a processor (31), memory (32), an input / output (I / O) interface (33), a communication interface (34), and a bus (35). Although this specification describes and illustrates a particular computer system having a particular number of particular components in a particular arrangement, this specification contemplates any suitable computer system having any suitable number of any suitable components in any suitable configuration.
[0224] In various embodiments of the present application, the processor (31) includes hardware for executing instructions, such as instructions constituting a computer program. For example, and without limitation, to execute the instructions, the processor (31) may retrieve (or fetch) the instructions from an internal register, an internal cache, or memory (32); decode and execute the instructions; and write one or more results to the internal register, internal cache, or memory (32). In some embodiments, the processor (31) may include one or more internal caches for data, instructions, or addresses. This disclosure contemplates a processor (31) including any suitable number of any suitable internal caches, where appropriate. For example, and without limitation, the processor (31) may include one or more instruction caches, one or more data caches, and one or more translation lookaside buffers (TLBs). Instructions in the instruction cache may be copies of instructions in the memory (32), and the instruction cache may speed up the retrieval of such instructions by the processor (31). Data within the data cache may be the results of previous instructions executed by the processor (31) for access by subsequent instructions executed by the processor (31) or for writing to memory (32); or may be a copy of data within the memory (32) for instructions executed by the processor (31) to operate on other suitable data. The data cache may speed up read or write operations by the processor (31). The TLBs may speed up virtual address translation by the processor (31). In various embodiments of the present application, the processor (31) may include one or more internal registers for data, instructions, or addresses. This disclosure contemplates a processor (31) including any suitable number of any suitable internal registers, where appropriate.Where appropriate, the processor (31) may include one or more arithmetic logic units (ALUs); may be a multi-core processor; or may include one or more processors (31). Although this specification describes and illustrates a particular processor, this specification contemplates any suitable processor.
[0225] In various embodiments of the present application, the memory (32) includes a main memory for storing instructions to be executed by the processor (31) or data to be operated by the processor (31). The processor (210) may load instructions from the memory (32) into an internal register or an internal cache. To execute the instructions, the processor (31) may retrieve the instructions from the internal register or the internal cache and decode them. During or after execution of the instructions, the processor (31) may write one or more results (which may be intermediate results or final results) to the internal register or the internal cache. Thereafter, the processor (31) may write one or more of these results to the memory (32). One or more memory buses (which may each include an address bus and a data bus) may connect the processor (31) to the memory (32). As described below, the bus (35) may include one or more memory buses. In various embodiments of the present application, one or more memory management units (MMUs) reside between the processor (31) and the memory (32) and facilitate access to the memory (32) requested by the processor (31).
[0226] In various embodiments of the present application, the memory (32) can store a program for the operation of the processor (31), and can also temporarily or permanently store data generated through input / output and operation processing (e.g., blood information, demographic information, probability of cognitive decline, regression coefficients, etc.). The memory (32) can include at least one type of storage medium among a flash memory type, a hard disk type, a multimedia card micro type, a card type memory (e.g., SD or XD memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, and an optical disk. The description of the above-described memory is merely an example, and the present disclosure is not limited thereto.
[0227] In various embodiments of the present application, the I / O interface (33) comprises hardware, software, or both that provide one or more interfaces for communication between a computer system (30) and one or more I / O devices. Where appropriate, the computer system (30) may include one or more of these I / O devices. One or more of these I / O devices may enable communication between a person and the computer system (30). By way of example and without limitation, the I / O devices may include a keyboard, a keypad, a joystick, a microphone, a monitor, a mouse, a printer, a scanner, a speaker, a still camera, a stylus, a tablet, a touch screen, a touch panel, a trackball, a video camera, another suitable I / O device, or a combination of two or more thereof. The I / O devices may include one or more sensors. This disclosure contemplates any suitable I / O device and a suitable I / O interface (33) therefor. Where appropriate, the I / O interface (33) may include one or more devices or software drivers that enable the processor (31) to drive one or more of these I / O devices. Where appropriate, the I / O interface (33) may include one or more I / O interfaces (33). Although this specification describes and illustrates a particular I / O interface, this specification contemplates any suitable I / O interface.
[0228] In various embodiments of the present application, the communication interface (34) comprises hardware, software, or both that provides one or more interfaces for communication (e.g., packet-based communication) between the computer system (30) and one or more other computer systems (30) or one or more networks. For example, and without limitation, the communication interface (34) may include a network interface controller (NIC) or network adapter for communicating over an Ethernet or other wired-based network, or a wireless NIC (WNIC) or wireless adapter for communicating over a wireless network, such as a WI-FI network. This disclosure contemplates any suitable network and any suitable communication interface (34) therefor. For example, and without limitation, the computer system (30) may communicate over one or more portions of an ad hoc network, a personal area network (PAN), a local area network (LAN), a wide area network (WAN), a metropolitan area network (MAN), the Internet, or a combination of two or more such networks. One or more portions of one or more such networks may be wired or wireless. By way of example, the computer system (30) may communicate over a wireless PAN (WPAN) (e.g., a BLUETOOTH WPAN), a WI-FI network, a WI-MAX network, a cellular phone network (e.g., a Global System for Mobile Communication (GSM) network), another suitable wireless network, or a combination of two or more such networks. Where appropriate, the computer system (30) may include any suitable communication interface (34) for any of these networks. Where appropriate, the communication interface (34) may include one or more communication interfaces (34). Although this specification describes and illustrates a particular communication interface, this specification contemplates any suitable communication interface.
[0229] In various embodiments of the present application, the bus (35) comprises hardware, software, or both that interconnect components of the computer system (30). For example, and without limitation, the bus (35) may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a front-side bus (FSB), a HYPERTRANSPORT (HT) interconnect, an Industry Standard Architecture (ISA) bus, an INFINIBAND interconnect, a low-pin-count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCIe) bus, a serial advanced technology attachment (SATA) bus, a Video Electronics Standard Association local (VLB) bus, another suitable bus, or a combination of two or more such buses. Where appropriate, the bus (35) may include more than one bus (35). Although this specification describes and illustrates a particular bus, this specification contemplates any suitable bus or interconnect.
[0230] The subject's blood information and demographic information can be input through the I / O interface (33). The subject's blood information can be at least a portion of the results analyzed by a device for detecting particles of an analyte present in trace amounts in a fluid sample, or by other diagnostic kits.
[0231] The processor (31) can process a training data set to learn a cognitive function prediction model and store it in the memory (32). The processor (31) can call the cognitive function prediction model stored in the memory (32) to calculate the probability of cognitive function decline. In addition, the processor (31) can generate solution information and an analysis report based on the probability of cognitive function decline.
[0232] Additionally, the predicted cognitive function level of the subject, solution information for the subject with the predicted cognitive function level (e.g., information on the applicability of a drug treatment), or an analysis report may be output through the I / O interface (33). Additionally, the predicted cognitive function level of the subject, solution information for the subject, or an analysis report may be transmitted to an external device (e.g., a user's terminal device, other computing device) through the communication interface (34).
[0233] In addition, the system (30) may further include a database that manages analysis information related to the cognitive function level of the subject. The database may be integrated within the computing system (30) or implemented as an external database server and connected to the computing system (30). In some embodiments, the information stored in the database may be structured according to a specific data structure. The database may be a relational, columnar, relational, or other suitable database. Although this specification describes or illustrates a particular type of database, this specification contemplates any suitable type of database. In some embodiments, the system (30) may provide an interface that allows for managing, retrieving, modifying, adding, or deleting information stored in the database via the I / O interface (33) or a separate interface device. The system (30) may store the subject's blood information, demographic information, and Alzheimer's disease analysis results based thereon in the database. The Alzheimer's disease analysis results may include information on the probability of cognitive decline and / or the applicability of Alzheimer's disease drug treatment.
[0234] Additionally, the system (30) can add additional biomarker(s) based on information stored in the database. Additionally, the system (30) can adjust the values of model parameters (e.g., regression coefficient vector) of a cognitive function prediction model based on information stored in the database.
[0235] Additionally, in some embodiments, the system (30) may be connected to a user's terminal device equipped with an application related to the subject's cognitive function level.
[0236]
[0237] Detection of analyte substances present in trace amounts in fluid samples
[0238] In the above step (S100), at least some of the biomarker information of the subject can be obtained by a method of detecting an analyte present in a trace amount in a fluid sample, which will be described below with reference to FIGS. 5 to 23.
[0239] In various embodiments of the present application, the system (30) of FIG. 4 can receive the detection results of the analysis target substance analyzed using the device of FIG. 5, FIG. 9, or FIG. 17 as biomarker information of the subject (S100).
[0240] FIG. 5 is a schematic diagram of a device for detecting particles of an analyte substance present in trace amounts in a fluid sample according to a first embodiment of the present invention.
[0241] FIG. 5(a) is a plan view of a device for detecting particles of an analysis target substance according to a first embodiment of the present invention, and FIG. 5(b) is a cross-sectional view of a device for detecting particles of an analysis target substance according to a first embodiment of the present invention.
[0242] Referring to FIG. 5, a device for detecting particles of an analyte substance present in a trace amount in a fluid sample according to a first embodiment of the present invention is configured to include a cartridge (100).
[0243] The cartridge (100) is configured to include a fluid sample chamber (200), a waste chamber (210), a washing solution chamber (230), a second capture material chamber (240), a signal generating material chamber (250), a substrate solution chamber (260), and an analysis chip (300).
[0244] It is preferable that the upper surface of the cartridge (100) be sealed with an air-impermeable film using heat sealing, UV sealing, ultrasonic sealing, tape, or the like to prevent the plurality of chambers and the analysis chip (300) from being contaminated with each other or external substances.
[0245] The fluid sample chamber (200) is configured to accommodate a certain amount of fluid sample. The fluid sample chamber (200) may be configured to have a volume ranging from 50 μL to 10 mL. The fluid sample includes body fluids such as blood, plasma, serum, and cerebrospinal fluid.
[0246] The waste chamber (210) is configured to provide a receiving space for receiving waste generated when the fluid sample, cleaning solution, second capture material, signal generating material, and substrate solution introduced into the analysis chip (300) are discharged. This waste chamber (210) may be configured to have a capacity of 1 to 15 mL. The waste chamber (210) of FIG. 1 is included in the cartridge (100), but the waste chamber (210) may be provided in a separate space outside the cartridge (100).
[0247] The washing solution chamber (230) is configured to receive a washing solution for washing the detection space (330) having the microchamber (30) of the analysis chip (300). The washing solution may be a solution such as phosphate buffered saline (PBS), Tris-buffered saline (TBS), PBS-T, TBS-T, or ultrapure water.
[0248] As an example, the cleaning solution of the cleaning solution chamber (230) can be injected into the injection port (310) to wash away the fluid sample, second capture material, signal generating material, and substrate solution remaining after the reaction in the detection space (330).
[0249] The second capture material chamber (240) is configured to accommodate a second capture material that specifically binds to the analyte and binds to the signal-generating material. The second capture material has the characteristic of specifically binding to both the analyte and the signal-generating material, so that the signal-generating material binds to the second capture material that specifically binds to the analyte and generates a fluorescent signal proportional to the amount of the analyte. For example, the second capture material may be a secondary antibody and may be provided in a form that is pre-bound to the signal-generating material.
[0250] Since the second capture material binds to the substance to be analyzed, and the first capture material fixed to the plurality of microchambers (30) also binds to the substance to be analyzed, it is preferable that the first capture material and the second capture material have different portions that bind to the substance to be analyzed.
[0251] The signal generating substance chamber (250) is configured to accommodate a signal generating substance that specifically binds to the second capture substance bound to the substance to be analyzed. The signal generating substance may be, for example, an enzyme.
[0252] The signal generating substance is, for example, a color source, a fluorescent, or a chemiluminescent enzyme precursor, which, when contacted with a substrate solution, is converted into a detectable indicator.
[0253] The signal generating material and the second capturing material may be accommodated in a combined state in either the second capturing material chamber (240) or the signal generating material chamber (250).
[0254] The signal generating substance reacts with a substrate solution to generate a detectable indicator such as a fluorescent signal, and the enzyme component may include β-galactosidase, horseradish peroxidase, or alkaline phosphatase.
[0255] The substrate solution chamber (260) is configured to receive a substrate solution that reacts with the signal generating substance to generate a signal. The substrate solution reacts with the signal generating substance to generate a detectable indicator, such as a fluorescent signal, and may include Resorufin ß-D-galactopyranoside, etc.
[0256] The analysis chip (300) is configured to include an injection port (310) and a detection space (330).
[0257] The injection port (310) is used to inject a fluid sample, a washing solution, a second capture material, a signal generating material, a substrate solution, or a mixed solution thereof.
[0258] The inlet (310) may further include a filter to separate and supply plasma from whole blood. The filter filters out blood cells in the whole blood and allows a fluid sample containing plasma components to pass through. The filter may be formed of various pore sizes or materials to filter out biological cells, inorganic particles, organic particles, etc. The filter of the present embodiment preferably has pores capable of filtering out cells or foreign substances larger than the substance to be analyzed, and is preferably formed with pores capable of filtering out blood cells having a size of approximately 5 to 10 μm. In addition, the filter may be formed of a biologically inert material so that it can be applied to biological samples.
[0259] In the process of supplying a fluid sample to the inlet (310), the output surface of the filter can be pre-wetted by wetting it with a buffer solution before filter filtration. Accordingly, the plasma of the whole blood supplied to the filter input surface can pass through the filter pores without surface resistance and be smoothly filtered to the filter output surface.
[0260] The detection space (330) is preferably provided with a plurality of micro chambers (30), and a plurality of first capture substances are fixed inside each micro chamber.
[0261] The micro chambers (30) are arranged in a form of a plurality of micro chambers at regular intervals on one side of the detection space (330) of the analysis chip (300), and each micro chamber is characterized by having the same size. It is preferable that the number of micro chambers is 10,000 or more, and the number may vary depending on the range of the concentration of the analyte to be measured. The inner surface of the micro chamber (30) is characterized in that a first capture substance that specifically binds to the analyte is immobilized. In addition, in order to prevent nonspecific reactions, the remaining surface without the first capture substance is preferably coated with a blocking substance such as BSA (bovine serum albumin).
[0262] When a fluid sample is injected through the injection port (310), the micro-
[0263] The first capture material fixed in the internal space of the chamber (30) and the analysis target material contained in the fluid sample are combined.
[0264] Afterwards, when the washing solution is injected through the injection port (310), the remaining fluid sample after the first capture material and the analysis target material are combined can be removed and washed.
[0265] Meanwhile, when the second capture material contained in the second capture material chamber (240) that binds to the analysis target material and the signal generating material is injected through the injection port (310), the analysis target material bound to the first capture material and the second capture material bind to each other, and thereafter, the remaining second capture material can be removed and washed through the washing solution injected through the injection port (310).
[0266] Next, when a signal generating substance is injected through the injection port (310), the signal generating substance and the second capturing substance combine.
[0267] Finally, a substrate solution is injected through the injection port (310) to cause a reaction between the signal generating substance and the substrate solution, thereby generating a fluorescent signal, and a hydrophobic solvent is allowed to flow through the injection port (310) to remove the substrate solution remaining after reacting with the signal generating substance. The hydrophobic solvent is preferably an oil or a gas.
[0268] Figure 6 illustrates in more detail the analysis chip (300) according to the first embodiment of the present invention.
[0269] A fluid sample, a cleaning solution, a second capture material, a signal generating material, a substrate solution, or a mixed solution thereof are sequentially injected through the injection port (310). The second capture material is a capture material for binding a signal generating material that generates a detectable indicator such as a fluorescent signal and an analysis target material, and may be provided combined with the signal generating material, which may simplify the procedure compared to when the second capture material and the signal generating material are provided separately.
[0270] When the fluid sample, washing solution, second capture material, signal generating material, and substrate solution or a mixed solution thereof are injected into the injection port (310), they diffuse and flow into the detection space (330) and flow through the microchamber (30).
[0271] It is preferable that a first capture material that binds to the analysis target material contained in the fluid sample is fixed in the microchamber (30).
[0272] Figure 7 conceptually illustrates a process for detecting a target substance for analysis according to the first embodiment of the present invention.
[0273] Fig. 7(a) illustrates that a microchamber (30) is concavely formed on the lower portion of the substrate, and a first capture material (40) is fixed to the microchamber (30). The size of the microchamber (30) is 0.5 to 15 μm, which is the minimum size for which optical measurement is possible.
[0274] The microchambers (30) are arranged on the bottom surface along the detection space (330) with a gap of 0.5 ㎛ or more between each other, and can be designed so that at least 10,000 microchambers are arranged in a uniform distribution.
[0275] The microchamber (30) is formed in a groove shape on the bottom surface of the detection space (330), and the first capture material (40) is fixed therein to specifically capture the analysis target material (50).
[0276] In one embodiment of the present invention, the microchamber (30) may be formed in a direction perpendicular to the fluid flow direction on the bottom surface of the detection space (330). Accordingly, the microchamber (30) and the fluid flow are arranged at a right angle. Accordingly, particles inserted into the microchamber (30) cannot easily escape from the microchamber (30) along the fluid flow.
[0277] In this way, the first capture material (40) is stably inserted and fixed within the microchamber (30) so that it can maintain a uniform arrangement state in the detection space (330) regardless of the fluid flow.
[0278] When the first capture material is an antibody particle, the size of the antibody is generally 10 to 50 nm, so it is preferable that a large number of antibodies are fixed in the microchamber (30).
[0279] FIG. 7(b) illustrates that the analyte (50) is captured by the first capture material (40) as the fluid sample flows over the microchamber (30). It is preferable that the remaining fluid sample, including the analyte that is not captured, be washed with a washing solution.
[0280] As can be seen in Fig. 7(b), since a plurality of first capture substances are fixed to the microchamber (30), the number of analyte targets (50) fixed to the first capture substances in one microchamber (30) may be such that one analyte target is evenly captured in each microchamber or the number of analyte targets captured in each microchamber may vary depending on the number of analyte targets and the material properties. For example, when the number of analyte targets is less than 60% of the number of microchambers, the probability that one analyte target is equally captured in each microchamber is high, but when the number of analyte targets is 60% or more of the number of microchambers, the number of analyte targets captured in each microchamber may vary in each microchamber. Therefore, when the number of analyte substances captured in each microchamber is the same, it is desirable to infer the number of analyte substances by calculating the number of microchambers in which a detectable indicator signal, such as a fluorescent signal, is generated, and when the number of analyte substances captured in each microchamber is different, it is desirable to infer the number of analyte substances by calculating not only the number of microchambers in which a detectable indicator signal, such as a fluorescent signal, is generated, but also the intensity of the detectable indicator signal, such as a fluorescent signal, generated in the microchamber.
[0281] Except for the analyte substance inserted and fixed or captured in the microchamber (30) by the washing liquid, the remaining substances are removed from the detection space (330). During the washing process, the substances inserted and fixed or captured in the microchamber are inserted and fixed in the microchamber (30) formed deep in a direction perpendicular to the flow direction of the washing liquid, so that they do not escape from the microchamber (30) due to the flow of the washing liquid and can remain fixed. Accordingly, only the remaining substances that are not fixed by combining with the first capture substance inserted and fixed in the microchamber (30) within the detection space (330) flow out from the detection space (330) along the flow of the washing liquid and are removed.
[0282] Figure 7(c) illustrates a state in which the analysis target substance (50) and the second capture substance (60) combine and react when the second capture substance (60) is allowed to flow over the microchamber (30). It is preferable that the remaining second capture substance (60) be washed away with a washing solution.
[0283] Figure 7(d) illustrates that when a signal generating substance (70) is allowed to flow over a microchamber (30), the signal generating substance (70) and the second capturing substance (60) combine and react. Afterwards, it is preferable to wash away the remaining signal generating substance (70) with a washing solution.
[0284] Figure 7(e) illustrates that a substrate solution is allowed to flow over a microchamber (30), and a signal generating substance (70) and the substrate solution react to generate a detectable indicator signal, such as a fluorescent signal. It is preferable that the remaining substrate solution be washed away with a washing solution.
[0285] In Fig. 7(b), since the number of analysis target substances (50) fixed to the first capture substance in one microchamber (30) is different, the number of signal generating substances (70) that generate a detectable indicator signal such as a fluorescent signal also differs for each microchamber.
[0286] Thereafter, a hydrophobic solvent is allowed to flow through the substrate to remove the substrate solution outside the plurality of microchambers (30), and the number of microchambers where the fluorescent signal is generated is counted and detected. The inside of the plurality of microchambers (30) is coated with a hydrophilic solvent, and the outside of the plurality of microchambers (30) is coated with a hydrophobic solvent, so that when the hydrophobic solvent is allowed to flow through the substrate, the substrate solution outside the plurality of microchambers (30) can be removed.
[0287] In order to detect the analyte contained in a fluid sample more accurately, it is desirable to correspond the number of particles of the analyte to the brightness of the signal of a detectable indicator, such as a fluorescent signal, and to count the number of particles of the analyte by considering the brightness of the signal of a detectable indicator, such as a fluorescent signal, of each microchamber.
[0288] FIG. 8 is a flowchart of a method for detecting particles of an analyte substance present in trace amounts in a fluid sample according to a first embodiment of the present invention.
[0289] At step 400, a fluid sample containing the substance to be analyzed is allowed to flow through a substrate having a microchamber array formed on the surface of which a first capture material that specifically binds to the substance to be analyzed is fixed.
[0290] At step 410, the substance to be analyzed is bound to the first capture substance in each microchamber of the microchamber (30) array.
[0291] In step 420, the remaining fluid sample after combining the first capture material and the analyte is removed and washed.
[0292] In step 430, a second capture material that specifically binds to the analyte and binds to the signal generating material is allowed to flow across the substrate to cause the analyte to react with the second capture material.
[0293] In step 440, the above-mentioned analysis target substance and the second capture substance are reacted, and the remaining second capture substance is removed and washed.
[0294] At step 450, the signal generating material is allowed to flow through the substrate so as to bind to the second capturing material.
[0295] At step 460, a substrate solution that reacts with the signal generating substance to generate a detectable indicator signal, such as a fluorescent signal, is allowed to flow over the substrate.
[0296] In step 470, a hydrophobic solvent is allowed to flow through the substrate, so that the signal generating substance and the substrate solution are combined inside the microchamber, and the substrate solution outside the microchamber is removed.
[0297] It is preferable that the interior of the microchamber is hydrophilic and the exterior of the microchamber is hydrophobic.
[0298] At step 480, the number of microchambers where a signal of a detectable indicator such as the fluorescent signal is generated is counted and detected.
[0299] FIG. 9 is a schematic diagram of a device for detecting particles of an analyte substance present in trace amounts in a fluid sample according to a second embodiment of the present invention.
[0300] FIG. 9(a) is a plan view of a device for detecting particles of an analysis target substance according to a second embodiment of the present invention, and FIG. 9(b) is a cross-sectional view of a device for detecting particles of an analysis target substance according to a second embodiment of the present invention.
[0301] Referring to FIG. 9, a device for detecting particles of an analyte substance present in a trace amount in a fluid sample according to a second embodiment of the present invention is configured to include a magnetic device (150) and a cartridge (101).
[0302] The magnetic device (150) causes separation between the magnetic particles, which are one example of fine particles within the reaction chamber (205), and the residual solution by moving closer to or further away from the reaction chamber (205). Afterwards, after removing the residual solution, the cleaning solution contained in the cleaning solution chamber (231) is injected into the reaction chamber (205), thereby cleaning the wall surface of the reaction chamber (205) where the magnetic particles are located.
[0303] The cartridge (101) is configured to include a fluid sample chamber (201), a reaction chamber (205), a waste chamber (211), a microparticle chamber (221), a washing solution chamber (231), a second capture material chamber (241), a signal generating material chamber (251), a substrate solution chamber (261), and an analysis chip (301).
[0304] It is preferable that the upper surface of the cartridge (101) be sealed with an air-impermeable film using heat sealing, UV sealing, ultrasonic sealing, tape, or the like to prevent the plurality of chambers and the analysis chip (301) from being contaminated with each other or external substances.
[0305] The fluid sample chamber (201) is configured to accommodate a certain amount of fluid sample. The fluid sample chamber (201) may be configured to have a volume ranging from 50 μL to 10 mL. The fluid sample includes body fluids such as blood, plasma, serum, and cerebrospinal fluid.
[0306] The reaction chamber (205) is configured to receive and mix a predetermined amount of a fluid sample and microparticles having a first capture material fixed thereon. The reaction chamber (205) may be configured to have a capacity ranging from 50 μL to 15 mL.
[0307] The waste chamber (211) is configured to provide a receiving space for receiving waste generated when fluid samples, microparticles, cleaning solutions, second capture materials, signal generating materials, and substrate solutions introduced into the analysis chip (301) are discharged. This waste chamber (211) may be configured to have a capacity of 1 to 30 mL. The waste chamber (211) of FIG. 9 is included in the cartridge (101), but the waste chamber (211) may be provided in a separate space outside the cartridge.
[0308] The microparticle chamber (221) is configured to accommodate microparticles having a first capture material fixed thereon, and the microparticles are injected into the reaction chamber (205). The microparticles may be polymer or magnetic particles. The magnetic particles preferably have a diameter of between 600 nm and 10 μm, and include a magnetic material such as iron oxide. The microparticles may be spherical magnetic particles or donut-shaped microchamber particles.
[0309] It is preferable that the microparticles accommodated in the microparticle chamber (221) have a first capture material bound to their surface that binds to the substance to be analyzed. The first capture material may include an antibody, a nucleic acid molecule, a peptide, or the like.
[0310] The washing solution chamber (231) is configured to receive a washing solution for washing the wall surface of the reaction chamber (205) or washing the detection space (331) having the microchamber (31) of the analysis chip (301). The washing solution may be a solution such as PBS (phosphate buffered saline), TBS (Tris-buffered saline), PBS-T, TBS-T, or ultrapure water.
[0311] As an example, the cleaning solution of the cleaning solution chamber (231) can be injected into the injection port (311) to wash away the fluid sample, fine particles, second capture material, signal generating material, and substrate solution remaining after reaction in the detection space (331).
[0312] In another embodiment, when the magnetic device (150) is brought into proximity with the reaction chamber (205) and separation occurs between the magnetic particles, which are one example of fine particles within the reaction chamber (205), and the residual solution, after removing the residual solution, the cleaning solution contained in the cleaning solution chamber (231) is injected into the reaction chamber (205), thereby cleaning the wall surface of the reaction chamber (205) where the magnetic particles are located.
[0313] The second capture material chamber (241) is configured to accommodate a second capture material that specifically binds to the analyte and binds to the signal generating material. The second capture material has the property of specifically binding to both the analyte and the signal generating material, so that the signal generating material binds to the second capture material that has specifically bound to the analyte and generates a fluorescent signal proportional to the amount of the analyte. The capture material may be, for example, a secondary antibody, and may be provided in a form that is pre-bound to the signal generating material.
[0314] Since the second capture material binds to the analyte, and the first capture material fixed to the microparticle also binds to the analyte, it is preferable that the first capture material and the second capture material have different portions that bind to the analyte.
[0315] The signal generating substance chamber (251) is configured to accommodate a signal generating substance that specifically binds to the second capture substance bound to the substance to be analyzed. The signal generating substance may be, for example, an enzyme.
[0316] The signal generating substance is, for example, a color source, a fluorescent, or a chemiluminescent enzyme precursor, which, when contacted with a substrate solution, is converted into a detectable indicator.
[0317] The signal generating material and the second capturing material may be accommodated in a combined state in either the second capturing material chamber (241) or the signal generating material chamber (251).
[0318] The signal generating substance reacts with a substrate solution to generate a detectable indicator such as a fluorescent signal, and the enzyme component may include β-galactosidase, horseradish peroxidase, or alkaline phosphatase.
[0319] The substrate solution chamber (261) is configured to receive a substrate solution that reacts with the signal generating substance to generate a signal. The substrate solution reacts with the signal generating substance to generate a detectable indicator, such as a fluorescent signal, and may include Resorufin ß-D-galactopyranoside, etc.
[0320]
[0321] The analysis chip (301) is configured to include an injection port (311) and a detection space (331).
[0322] Since a certain amount of fluid sample and the first capture material fixed microparticles are received and mixed in the reaction chamber (205), the substance to be analyzed is bound to the first capture material fixed to the microparticles.
[0323] The injection port (311) is used to inject fluid samples, microparticles, washing solutions, second capture materials, signal generating materials, and substrate solutions or mixed solutions.
[0324] The inlet (311) may further include a filter to separate and supply particles larger than fine particles from a mixed solution of fine particles and a fluid sample. The filter may be formed with pores of various sizes or made of various materials to filter biological cells, inorganic particles, organic particles, etc. The filter of the second embodiment is preferably formed with pores capable of filtering particles larger than fine particles or foreign substances. In addition, the filter may be formed of a biologically inert material so that it can be applied to biological samples.
[0325] In the process of supplying a fluid sample to the inlet (311), the output surface of the filter can be pre-wetted by wetting it with a buffer solution before filter filtration. Accordingly, the mixed solution of fine particles and fluid sample supplied to the filter input surface can pass through the filter pores without surface resistance and be smoothly filtered to the filter output surface.
[0326] The detection space (331) has a plurality of microchambers (31) arranged therein, and when a mixed solution of microparticles and a fluid sample is injected through the injection port (311), the microparticles are captured in the internal space of the microchambers (31) within the detection space (331). By arranging a magnetic body (15) below the microchambers (31), if the microparticles are magnetic particles, they can be easily captured in the internal space of the microchambers (31).
[0327] The microparticles captured in the microchamber (31) are a mixture of microparticles captured by the first capture substance and microparticles that are not captured. The size of the microchamber (31) is preferably larger than the diameter of the microparticles and smaller than twice the diameter of the microparticles.
[0328] Afterwards, when the washing solution is injected through the injection port (311), the remaining fluid sample after the first capture material and the analysis target material are combined can be removed and washed.
[0329] Meanwhile, when the second capture substance contained in the second capture substance chamber (241) that binds to the analysis target substance and the signal generating substance is injected through the injection port (311), the analysis target substance bound to the first capture substance and the second capture substance bind to each other, and the remaining second capture substance can then be removed and washed using a washing solution.
[0330] Next, when a signal generating substance is injected through the injection port (311), the signal generating substance and the second capturing substance combine.
[0331] Finally, a substrate solution is injected through the injection port (311) to cause a reaction between the signal generating substance and the substrate solution, thereby generating a detectable indicator signal such as a fluorescent signal, and a hydrophobic solvent is caused to flow through the injection port (311) to remove the substrate solution remaining after reacting with the signal generating substance. The hydrophobic solvent is preferably an oil or a gas.
[0332] Figure 10 illustrates in more detail the analysis chip (301) according to the second embodiment of the present invention.
[0333] A mixed solution of a fluid sample and microparticles, a washing solution, a second capture material, a signal generating material, and a substrate solution or a mixed solution thereof are sequentially injected through the injection port (311).
[0334] The second capture material is a capture material for binding a signal generating material that generates a detectable indicator such as a fluorescent signal and a substance to be analyzed, and can be provided in combination with the signal generating material, which can simplify the procedure compared to when the second capture material and the signal generating material are provided separately.
[0335] When the above fluid sample and the mixed solution of fine particles, the washing solution, the second capture material, the signal generating material, and the substrate solution or the mixed solution thereof are injected into the injection port (311), they diffuse and flow into the detection space (331) and flow through the microchamber (31).
[0336] The size of the microchamber (310) is preferably larger than the diameter of the microparticles and smaller than twice the diameter of the microparticles.
[0337]
[0338] FIG. 11 illustrates a combined form of a microparticle, a first capture material, an analyte, and a signal generating material according to a second embodiment of the present invention.
[0339] Referring to Fig. 11(a), it is illustrated that a first capture material (41) is fixed to a microparticle (1), and referring to Fig. 11(b), it is illustrated that an analyte (51) is bound to the first capture material (41), a second capture material (61) is bound to the analyte (51), and a signal generating material (71) is bound to the second capture material (61). When a substrate solution reacts with the signal generating material (71), a detectable indicator signal such as a fluorescent signal is generated.
[0340]
[0341] Figure 12 shows in detail the state in which fine particles are captured in a microchamber arranged in a detection space according to the second embodiment of the present invention.
[0342] Referring to Fig. 12, a magnetic body (15) is placed below the internal space of a microchamber (31) to form a magnetic field, and as an example of a microparticle (1) in a mixed solution, a state of capturing a magnetic particle is illustrated.
[0343]
[0344] Figure 13 conceptually illustrates a process for detecting a substance to be analyzed according to a second embodiment of the present invention.
[0345] Figure 13(a) illustrates a state in which magnetic particles (1) having a first capture material (41) fixed thereto are injected into the injection port (311) while being mixed with a fluid sample in a reaction chamber (205) and reacted, and the magnetic particles (1) are captured by a magnetic body (15) in each microchamber (31). The first capture material (41) is fixed to the magnetic particles (1).
[0346] As the mixed solution coming out of the injection port (311) flows through the plurality of microchambers (31), magnetic particles are captured in each receiving portion by the magnetic body placed under the plurality of microchambers (31).
[0347] The microchambers (31) are arranged at intervals on the bottom surface along the detection space (331) and can be designed to be arranged in a uniform distribution.
[0348] In one embodiment of the present invention, the depth of the microchamber (31) may be d < h < 2d. Here, h = the depth of the microchamber (31), and d = the diameter of the microparticle. The depth (h) of the microchamber (31) is the distance from the bottom surface of the microchamber (31) to the inner upper edge of the microchamber (31). In addition, the width of the microchamber (31) may be d < w < 2d. Here, w = the width of the microchamber (30), and d = the diameter of the microparticle.
[0349] When the depth (h) of the microchamber (31) is smaller than the diameter (d) of the microparticle, the fixation of the microparticle is unstable, and the microparticle escapes from the microchamber (31) due to the flow of the fluid sample flowing along the detection space (331). When the depth (h) of the microchamber (31) exceeds twice the diameter (d) of the microparticle, the probability of two or more microparticles being captured in one chamber increases, which causes a problem of a decrease in analysis accuracy and a problem of a problem of a long diffusion of the cleaning solution for cleaning the microparticles and the detection space (331).
[0350] In addition, in one embodiment of the present invention, the microchamber (31) may be formed in a direction perpendicular to the fluid flow direction on the bottom surface of the detection space (331). Accordingly, when the microchamber cleaning liquid is injected, after the first capture material (41) and the analysis target material (51) are combined, the fluid sample including the analysis target material that is not captured can be removed and washed. The magnetic particles (1) are not washed because they are fixed to each receiving portion of the microchamber (31) by the magnetic body.
[0351] Except for the magnetic particles inserted and fixed or captured in the microchamber (31) by the cleaning solution, the remaining substances are removed from the detection space (331). The substances inserted and fixed or captured in the microchamber during the cleaning process are inserted into the microchamber (31) which is formed deep in the direction perpendicular to the flow direction of the cleaning solution, so that they can remain fixed without being removed from the microchamber (31) by the flow of the cleaning solution. Accordingly, only the remaining substances that were inserted into the microchamber (31) and were not combined and fixed with the substance to be analyzed are removed by flowing out of the detection space (331) along the flow of the cleaning solution.
[0352] As can be seen in Fig. 13(b), when the second capture substance (61) contained in the second capture substance chamber (241) that binds to the analysis target substance and the signal generating substance is injected through the injection port (311), the analysis target substance (51) bound to the first capture substance (41) and the second capture substance (61) bind together, and thereafter, the remaining second capture substance (61) can be removed and washed using a washing solution.
[0353] In Fig. 13(c), when a signal generating substance (71) is injected through the injection port (311), the signal generating substance (71) and the second capturing substance (61) combine. Thereafter, it is preferable to wash away the remaining signal generating substance (71) with a washing solution.
[0354] In Fig. 13(d), when a substrate solution is injected through the injection port (311), the signal generating substance (71) and the substrate solution react to generate a detectable indicator signal such as a fluorescent signal, and a hydrophobic solvent is allowed to flow through the injection port (311) to remove the substrate solution remaining after reacting with the signal generating substance (71).
[0355] By causing a hydrophobic solvent to flow on the substrate, the substrate solution outside the plurality of microchambers (31) is removed, and the number of microchambers where a signal of a detectable indicator such as the fluorescent signal is generated is counted and detected. The inside of the plurality of microchambers (31) is coated with a hydrophilic solvent, and the outside of the plurality of microchambers (30) is coated with a hydrophobic solvent, so that when the hydrophobic solvent is flowed on the substrate, the substrate solution outside the plurality of microchambers (31) can be easily removed.
[0356] In order to detect the analyte contained in a fluid sample more accurately, the number of particles of the analyte corresponds to the brightness of the signal of a detectable indicator, such as a fluorescent signal, and the number of particles of the analyte can be counted by considering the brightness of the signal of a detectable indicator, such as a fluorescent signal, of each micro chamber.
[0357] Referring again to Fig. 13(d), it can be confirmed that the number of signal generating substances that generate a detectable indicator signal, such as a fluorescent signal, is different for each microchamber, and the brightness of the fluorescent signal is different accordingly.
[0358] Therefore, when the number of signal-generating substances captured in each microchamber is the same, it is desirable to infer the number of target substances by calculating the number of microchambers where fluorescence signals are generated, and when the number of signal-generating substances captured in each microchamber is different, it is desirable to infer the number of target substances by calculating not only the number of microchambers where fluorescence signals are generated but also the intensity of the fluorescence signals generated in the microchambers.
[0359] According to another embodiment of the present invention, a first capture material fixed to a microparticle and an analyte in a fluid sample are reacted within a reaction chamber (200), a magnetic device is brought close to the reaction chamber (200) to separate the microparticle from the fluid sample, and then the fluid sample that does not bind to the microparticle can be removed and washed.
[0360] Thereafter, a second capture material that specifically binds to the analyte bound to the first capture material can be injected into the reaction chamber (200) and reacted, and a signal generating material that specifically binds to the second capture material can be injected into the reaction chamber (200) and reacted.
[0361] Next, a substrate solution that reacts with the signal generating material to generate a signal is injected into the reaction chamber (205) and mixed, and magnetic particles in the mixed substrate solution are captured in a micro chamber (31) within the detection space (331) using a magnetic body (15).
[0362] Lastly, it is preferable that the particle detection device of the target substance be detected by counting the number of micro chambers (31) in which a signal is generated by mixing the signal generating substance and the substrate solution.
[0363] Figures 14 and 15 illustrate a preprocessing process of a method for detecting particles of an analysis target material according to a second embodiment of the present invention.
[0364] First, as shown in (a) of Fig. 14, the magnetic device driving unit (160) can move the magnetic device (150) to contact the reaction chamber (205) for 20 minutes. The reaction chamber (205) contains a mixed solution (S) containing a certain amount of a fluid sample and magnetic particles with a first capture material fixed thereon.
[0365] As a result, as shown in (b) of Fig. 14, separation between the magnetic particles (S1) and the residual solution (W) can occur in the reaction chamber (205).
[0366] Next, as illustrated in (a) of FIG. 15, the residual solution (W) is removed from the reaction chamber (205), and the cleaning solution contained in the cleaning solution chamber (231) is injected into the reaction chamber (205), thereby cleaning the wall surface of the reaction chamber (205) where the magnetic particles (S1) are located. Thereafter, the magnetic device driving unit (160) can move the magnetic device (150) again to separate the magnetic device (150) from the reaction chamber (205).
[0367] Then, as shown in (a) of Fig. 15, magnetic particles (S1) remain in the reaction chamber (205), and as shown in (b) of Fig. 15, they are preferably placed in the injection port (311).
[0368]
[0369] Figure 16 is a flowchart of a method for detecting particles of an analyte substance present in trace amounts in a fluid sample according to a second embodiment of the present invention.
[0370] At step 1200, a first capture agent is mixed with the microparticles so that the first capture agent is fixed to the microparticles. The first capture agent may include an antibody, a nucleic acid molecule, a peptide, or the like. Preferably, the microparticles are magnetic particles.
[0371] The magnetic particles have a diameter of between 600 nm and 10 μm, and preferably include a magnetic material such as iron oxide. According to one embodiment of the present invention, the mixing may be performed using a method such as vibration, ultrasound, or pipetting.
[0372] At step 1210, the first capture material fixed to the microparticles and the analyte in the fluid sample are reacted in the reaction chamber (200).
[0373] The above fluid samples include body fluids such as blood, plasma, serum, and cerebrospinal fluid.
[0374] At step 1220, the magnetic device is brought close to the reaction chamber (200) to separate the microparticles from the fluid sample, and the fluid sample that does not bind to the microparticles is removed and washed.
[0375] After removing the fluid sample that does not bind to the above-mentioned fine particles, a washing solution is injected and mixed with the above-mentioned fine particles, and then the magnetic body is brought close to the reaction chamber (200) to separate the magnetic particles from the washing solution, and the washing solution is removed, thereby performing washing of the fine particles.
[0376] At step 1230, a second capture agent that specifically binds to the analyte bound to the first capture agent is injected into the reaction chamber (200) and reacted. The second capture agent that does not bind to the analyte is removed and washed.
[0377] Since the second capture material binds to the analyte, and the first capture material fixed to the microparticle also binds to the analyte, it is preferable that the first capture material and the second capture material have different portions that bind to the analyte.
[0378] At step 1240, a signal-generating substance that specifically binds to the second capture substance is injected into the reaction chamber (200) and reacted. The signal-generating substance that does not bind to the second capture substance is removed and washed. The signal-generating substance may be, for example, an enzyme.
[0379] The signal generating substance is, for example, a color source, a fluorescent, or a chemiluminescent enzyme precursor, which, when contacted with a substrate solution, is converted into a detectable indicator.
[0380] The signal generating substance reacts with a substrate solution to generate a detectable indicator signal such as a fluorescent signal, and the enzyme component may include β-galactosidase, horseradish peroxidase, or alkaline phosphatase.
[0381] In steps 1230 and 1240, the particle detection device of the target substance is cleaned of fine particles in the same manner as in step 1220.
[0382] At step 1250, the particle detection device for the target substance injects a substrate solution that reacts with the signal generating substance to generate a signal into the reaction chamber (205) and mixes the substrate solution. It is preferable to dry the fine particles remaining in the reaction chamber (205) and then mix them with the substrate solution.
[0383] The above substrate solution reacts with the signal generating substance to generate a detectable indicator signal such as a fluorescent signal, and may include Resorufin ß-D-galactopyranoside, etc.
[0384] At step 1260, the particle detection device of the target substance injects the mixed substrate solution into the injection port (301) and captures the magnetic particles in the mixed substrate solution into the micro chamber (31) in the detection space (331) using the magnetic body (15).
[0385] In order to separate the substances in the microchamber (31) placed in the detection space (331) of the analysis chip (301) from each other so that they do not communicate with each other, a fluid having a property of not mixing with the substrate solution can be injected, and it is preferable that the fluid contains oil or air, etc. It is preferable that the microchamber is larger than the size of the microparticles and has a size in which one or less magnetic particles can be contained.
[0386] According to another embodiment of the present invention, instead of reacting the microparticles with the second capture material in the reaction chamber (200), after the first capture material fixed to the microparticles binds to the analyte, the microparticles can be injected into the injection port (311) so that the microparticles are captured in each microchamber (31).
[0387] Afterwards, when the second capture material (61) that binds to the analysis target substance and the signal generating substance is injected through the injection port (311), the analysis target substance (51) bound to the first capture material (41) and the second capture material (61) bind together, and as a next step, when the signal generating substance (71) is injected through the injection port (311), the signal generating substance (71) and the second capture material (61) bind together.
[0388] In addition, when a substrate solution is injected through the injection port (311), the signal generating substance (71) and the substrate solution react to generate a detectable indicator signal such as a fluorescent signal, and by allowing a hydrophobic solvent to flow through the injection port (311), the substrate solution remaining after reacting with the signal generating substance (71) can be removed.
[0389] At step 1270, the particle detection device of the target substance for analysis counts and detects the number of micro chambers (31) in which a signal is generated by mixing the signal generating substance and the substrate solution.
[0390] As an example, a first count value is calculated by recognizing and counting a microchamber containing the magnetic particles, and a second count value is calculated by recognizing and counting a microchamber in which a signal is generated by mixing the signal generating material and the substrate solution.
[0391] The amount of the analyte in the fluid sample can be estimated by calculating the ratio of the second count value to the first count value.
[0392] As another embodiment, an area including a plurality of micro chambers may be photographed using a microscope, a digital camera, or the like, and a first count value may be calculated by counting the number of micro chambers containing the magnetic particles through image processing, and a second count value may be calculated by counting the number of micro chambers in which a signal of a detectable indicator such as a fluorescent signal is generated through image processing.
[0393] As an example of estimating the amount of particles of the above-mentioned target substance, the amount can be estimated by calculating the ratio of the number of micro-chambers in which the fluorescent signal is generated to the number of micro-chambers in which the magnetic particles are contained, and the higher the amount of the target substance in the fluid sample, the higher the ratio, and the lower the amount of the target substance, the lower the ratio.
[0394] As another embodiment of estimating the amount of particles of the above-mentioned target substance, a particle detection device for the target substance of analysis according to an embodiment of the present invention may include three or more standard substances whose concentrations are known in advance in a cartridge, generate a calibration curve indicating the concentration of the standard substances, and then correct the concentration calculation of the target substance of analysis using the generated calibration curve.
[0395] Figure 17 is a schematic diagram of a device for detecting particles of an analyte substance present in trace amounts in a fluid sample according to a third embodiment of the present invention.
[0396] FIG. 17(a) is a plan view of a device for detecting particles of an analysis target substance according to a third embodiment of the present invention, and FIG. 17(b) is a cross-sectional view of a device for detecting particles of an analysis target substance according to a third embodiment of the present invention.
[0397] Referring to FIG. 17, a device for detecting particles of an analyte substance present in a trace amount in a fluid sample according to a third embodiment of the present invention is configured to include a magnetic device (152) and a cartridge (102).
[0398] The magnetic device (152) causes separation between the microwell microparticles (4) and the residual solution within the reaction chamber (206) by moving closer to or further away from the reaction chamber (206). Afterwards, after removing the residual solution, the cleaning solution contained in the cleaning solution chamber (232) is injected into the reaction chamber (206), thereby cleaning the wall surface of the reaction chamber (206) where the microwell microparticles (4) are located. It is preferable that a magnetic material is formed at the bottom of the microwell microparticles (4).
[0399] The cartridge (102) is configured to include a fluid sample chamber (202), a reaction chamber (206), a waste chamber (212), a microwell microparticle chamber (222), a washing solution chamber (232), a second capture material chamber (242), a signal generating material chamber (252), a substrate solution chamber (262), and an analysis chip (302).
[0400] It is preferable that the cartridge (102) be sealed using a film that is impermeable to air, such as heat-sealing, UV-sealing, ultrasonic sealing, or tape, to prevent the multiple chambers and the analysis chip (302) from being contaminated.
[0401] The fluid sample chamber (202) is configured to accommodate a certain amount of fluid sample. The fluid sample chamber (202) may be configured to have a volume ranging from 50 μL to 10 mL. The fluid sample includes body fluids such as blood, plasma, serum, and cerebrospinal fluid.
[0402] The waste chamber (212) is configured to provide a receiving space for receiving waste generated when the fluid sample, microwell microparticles, washing solution, second capture material, signal generating material, and substrate, signal generating material, and substrate solution, etc., injected into the analysis chip (302) are discharged. This waste chamber (212) may be configured to have a capacity of 1 to 15 mL. The waste chamber (212) of FIG. 17 is included in the cartridge (102), but the waste chamber (212) may be provided in a separate space outside the cartridge.
[0403] The microwell microparticle chamber (222) is configured to accommodate microwell microparticles having a first capture material fixed thereon, and the microwell microparticles are injected into the reaction chamber (206).
[0404] It is preferable that the microwell microparticles accommodated in the microwell microparticle chamber (222) have a first capture substance bound to the surface that binds to the substance to be analyzed. The first capture substance may include an antibody, a nucleic acid molecule, a peptide, or the like.
[0405] The washing solution chamber (232) is configured to receive a washing solution for washing the wall surface of the reaction chamber (206) or the detection space (332) of the analysis chip (302). The washing solution may be a solution such as phosphate buffered saline (PBS), Tris-buffered saline (TBS), PBS-T, TBS-T, or ultrapure water.
[0406] As an example, the cleaning solution of the cleaning solution chamber (232) can be injected into the injection port (312) to wash away the fluid sample, fine particles, second capture material, signal generating material, and substrate solution remaining after the reaction in the detection space (332).
[0407] In another embodiment, when the magnetic device (152) is brought into proximity with the reaction chamber (206) and separation occurs between the microwell microparticles and the residual solution within the reaction chamber (206), after removing the residual solution, the washing solution contained in the washing solution chamber (232) is injected into the reaction chamber (206), thereby washing the wall surface of the reaction chamber (205) where the microwell microparticles are located. In this case, it is preferable that a magnetic material is formed at the bottom of the microwell microparticles.
[0408] The second capture material chamber (242) is configured to accommodate a second capture material that specifically binds to the analyte and binds to the signal generating material. The second capture material has the property of specifically binding to both the analyte and the signal generating material, so that the signal generating material binds to the second capture material that has specifically bound to the analyte and generates a fluorescent signal proportional to the amount of the analyte. The capture material may be, for example, a secondary antibody, and may be provided in a form that is pre-bound to the signal generating material.
[0409] Since the second capture material binds to the analyte, and the first capture material fixed to the microwell microparticle also binds to the analyte, it is preferable that the first capture material and the second capture material have different portions that bind to the analyte.
[0410] The signal generating substance chamber (252) is configured to receive a signal generating substance that specifically binds to the second capture substance bound to the substance to be analyzed. The signal generating substance may be, for example, an enzyme.
[0411] The signal generating substance is, for example, a color source, a fluorescent, or a chemiluminescent enzyme precursor, which, when contacted with a substrate solution, is converted into a detectable indicator.
[0412] The signal generating material and the second capturing material may be accommodated in a combined state in either the second capturing material chamber (242) or the signal generating material chamber (252).
[0413] The signal generating substance reacts with a substrate solution to generate a detectable indicator such as a fluorescent signal, and the enzyme component may include β-galactosidase, horseradish peroxidase, or alkaline phosphatase.
[0414] The substrate solution chamber (262) is configured to receive a substrate solution that reacts with the signal generating substance to generate a signal. The substrate solution reacts with the signal generating substance to generate a detectable indicator, such as a fluorescent signal, and may include Resorufin ß-D-galactopyranoside, etc.
[0415] The analysis chip (302) is configured to include an injection port (312) and a detection space (332).
[0416] In the reaction chamber (206), a certain amount of fluid sample and the microwell microparticles on which the first capture material is fixed are received and mixed, so that the substance to be analyzed is bound to the first capture material fixed to the microwell microparticles.
[0417] The injection port (312) is used to inject a fluid sample, microwell microparticles, a washing solution, a second capture material, a signal generating material, and a substrate solution or mixed solution.
[0418] The inlet (312) may further include a filter to separate and supply particles larger than the microwell microparticles from the mixed solution of the microwell microparticles and the fluid sample. The filter may be formed of various sized pores or various materials to filter biological cells, inorganic particles, organic particles, etc. The filter of the third embodiment is preferably formed of pores capable of filtering particles larger than the microwell microparticles. In addition, the filter may be formed of a biologically inert material so that it can be applied to biological samples.
[0419] In the process of supplying a fluid sample to the inlet (312), the output surface of the filter can be pre-wetted by wetting it with a buffer solution before filter filtration. Accordingly, the mixed solution of microwell fine particles and fluid sample supplied to the filter input surface can pass through the filter pores without surface resistance and be smoothly filtered to the filter output surface.
[0420] When a mixed solution of microwell microparticles and a fluid sample is injected into the detection space (332) through the injection port (312), the microwell microparticles are placed in the internal space of the detection space (332). By placing a magnetic body (16) under the microwell microparticles (4), they can be easily fixed in the internal space of the detection space (332).
[0421] Thereafter, when the washing solution is injected through the injection port (312), the microwell microparticles that are not fixed to the internal space of the detection space (332), the first capture material that is not fixed to the microwell microparticles (4), and the fluid sample that is not bound to the first capture material can be removed and washed.
[0422] Meanwhile, when the second capture material contained in the second capture material chamber (242) that combines with the analysis target material and the signal generating material is injected through the injection port (312), the analysis target material combined with the first capture material and the second capture material combine, and thereafter, the remaining second capture material can be removed and washed using a washing solution.
[0423] Next, when a signal generating substance is injected through the injection port (312), the signal generating substance and the second capturing substance combine.
[0424] Finally, a substrate solution is injected through the injection port (312) to cause a reaction between the signal generating substance and the substrate solution, thereby generating a detectable indicator signal such as a fluorescent signal, and a hydrophobic solvent is caused to flow through the injection port (312) to remove the substrate solution remaining after reacting with the signal generating substance. The hydrophobic solvent is preferably an oil or a gas.
[0425] Figure 18 illustrates in more detail the analysis chip (302) according to the third embodiment of the present invention.
[0426] Through the injection port (312), a fluid sample, a mixed solution of microwell microparticles, a washing solution, a second capture material, a signal generating material, and a substrate solution or a mixed solution thereof are sequentially injected. The second capture material is a capture material for detecting a signal of a detectable indicator such as a fluorescent signal, and may be provided in combination with a signal generating material, which may simplify the procedure compared to when the second capture material and the signal generating material are provided separately.
[0427] When the fluid sample and the mixed solution of microwell microparticles (4), washing solution, second capture material, signal generating material, and substrate solution or the mixed solution thereof are injected into the injection port (312), they diffuse and flow into the detection space (332), and a filler (5) is formed at one end of the detection space (332) so that the microwell microparticles (4) can gather into the detection space (332).
[0428] As shown in Fig. 18, microwell microparticles (4) are gathered in the detection space (302), and microwell microparticles that are bound to the first capture material fixed in the internal space of the microwell microparticles (4) and the analyte in the fluid sample and microwell microparticles that are not bound are mixed.
[0429] Figure 19 illustrates in more detail the microwell microparticles according to the third embodiment of the present invention.
[0430] Referring to Fig. 19, the microwell microparticle (4) has a hexagonal upper surface, and the receiving portion (6), which is the internal space of the microwell microparticle (4), may have one side open or may have a donut-shaped shape with both sides open and penetrating.
[0431] The receiving portion (6) of the microwell microparticle (4) can be a cylindrical shape with one side closed (Fig. 19(b)) or a cylindrical shape with both sides open (Fig. 19(c)).
[0432] Meanwhile, the first capture material is fixed to the receiving portion (6) of the microwell microparticle (4) and can bind to the substance to be analyzed.
[0433] The inner space of the microwell microparticle (4), which is the receiving portion (6) and the first capture material, is coated with a hydrophilic solvent, and the outer space (7) of the microwell microparticle (4) is coated with a hydrophobic solvent, so that when the hydrophobic solvent is flowed onto the substrate, the substrate solution in the outer space (7) of the microwell microparticle (4) can be removed.
[0434] Comparing Fig. 11 and Fig. 19, the difference is that in Fig. 19, the first capture material is fixed to the internal space of the microwell microparticle (4), whereas in Fig. 10, the first capture material is fixed to the microparticle.
[0435] Figure 20 conceptually illustrates a process for detecting a substance to be analyzed according to a third embodiment of the present invention.
[0436] A mixture of a fluid sample and microwell microparticles (4), a washing solution, a second capture material, a signal generating material, and a substrate solution or a mixture thereof are sequentially injected through the injection port (312).
[0437] Figure 20(a) illustrates a state in which each microwell microparticle (4) is positioned and fixed by a magnetic body (16) when the microwell microparticle (4) to which the first capture material (42) is fixed is injected into the injection port (312) while being mixed with a fluid sample in a reaction chamber (206) and reacted. The first capture material (42) is fixed to the microwell microparticle (4) to specifically capture the analysis target material (52).
[0438] A mixture of a fluid sample and microwell microparticles (4) can be mixed in a reaction chamber (206). The internal space of the microwell microparticles (4) has a first capture substance fixed therein, and the substance to be analyzed is combined with the first capture substance.
[0439] When the mixture of the fluid sample and the micro well micro particles (4) is injected into the injection port (312), it diffuses and flows into the detection space (332). The mixture of the fluid sample and the micro well micro particles (4) that passes through the detection space (332) is blocked by a plurality of fillers (5) formed at one end opposite the injection port (312) and is collected in the detection space (332).
[0440] Thereafter, when the washing solution is injected through the injection port (312), the fluid sample including the binding between the first capture material and the analysis target material or the analysis target material not captured by the first capture material can be removed and washed.
[0441] As can be seen in Fig. 20(b), when the second capture substance (62) contained in the second capture substance chamber (242) that binds to the analyte and the signal generating substance is injected through the injection port (312), the analyte (52) bound to the first capture substance (42) and the second capture substance (62) bind together, and thereafter, the remaining second capture substance (62) can be removed and washed using a washing solution.
[0442] In Fig. 20(c), when a signal generating substance (72) is injected through the injection port (312), the signal generating substance (72) and the second capturing substance (62) combine. Thereafter, it is preferable that the remaining signal generating substance (72) be washed away with a washing solution.
[0443] In Fig. 20(d), when a substrate solution is injected through the injection port (312), the signal generating substance (72) and the substrate solution react to generate a detectable indicator signal such as a fluorescent signal, and a hydrophobic solvent is allowed to flow through the injection port (312) to remove the substrate solution remaining after reacting with the signal generating substance (72).
[0444] By causing a hydrophobic solvent to flow on the substrate, the substrate solution outside the microwell microparticles (4) is removed, and the number of microwell microparticles (4) that generate a detectable indicator signal, such as a fluorescent signal, is counted and detected. The inside (6) of the microwell microparticles (4) is coated with a hydrophilic solvent, and the outside (7) of the microwell microparticles (4) is coated with a hydrophobic solvent, so that when the hydrophobic solvent is flowed on the substrate, the substrate solution outside (7) of the microwell microparticles (4) can be easily removed.
[0445] In order to detect the analyte contained in a more accurate fluid sample, the number of particles of the analyte corresponds to the brightness of the signal of a detectable indicator such as a fluorescent signal, and the number of particles of the analyte can be counted by considering the brightness of the fluorescent signal of each microwell microparticle (4).
[0446] Referring again to Fig. 20(d), it can be confirmed that the number of signal generating substances that generate a detectable indicator signal, such as a fluorescent signal, is different for each microwell microparticle (4), and the brightness of the fluorescent signal is different accordingly.
[0447]
[0448] Figure 21 is a flowchart of a method for detecting particles of an analyte substance present in trace amounts in a fluid sample according to a third embodiment of the present invention.
[0449] At step 1700, the fluid sample is mixed with the microwell microparticles (4) on which the first capture material that specifically binds to the substance to be analyzed is fixed in the reaction chamber (206).
[0450] The mixed fluid sample and micro well micro particles (4) are injected into the injection port (312), and the first capture material (42) is fixed in the internal space (6) of the micro well micro particles (4).
[0451] At step 1710, a mixed solution of microwell microparticles (4) and a fluid sample is flowed into the detection space (332). The microwell microparticles (4) flow into the detection space (332) while the analyte (52) in the fluid sample is combined with the first capture material (42).
[0452] At step 1720, the remaining fluid sample after the first capture material (42) and the analyte (52) are combined is removed and washed.
[0453] At step 1730, a second capture material (62) that specifically binds to the analyte (52) and binds to the signal generating material (72) is allowed to flow through the microwell microparticles (4) to cause the analyte (52) and the second capture material (62) to react.
[0454] At step 1740, the analysis target substance (52) and the second capture substance (62) are reacted, and the remaining second capture substance (62) is removed and washed.
[0455] At step 1750, the signal generating substance (72) is allowed to flow through the microwell microparticles (4) so as to bind to the second capturing substance (62).
[0456] At step 1760, a substrate solution that reacts with a signal generating substance (72) to generate a detectable indicator signal such as a fluorescent signal is allowed to flow through the microwell microparticles (4).
[0457] At step 1770, a hydrophobic solvent is allowed to flow through the microwell microparticle (4), so that the signal generating substance (72) and the substrate solution are combined inside the microwell microparticle (4), and the substrate solution outside the microwell microparticle (4) is removed.
[0458] It is preferable that the inside (6) of the microwell microparticle (4) is hydrophilic and the outside (7) of the microwell microparticle (4) is hydrophobic.
[0459] At step 1780, the number of microwell microparticles (4) that generate a detectable indicator signal such as the fluorescent signal is counted and detected.
[0460] As an example, a first count value is calculated by recognizing and counting the microwell microparticle (4) and a second count value is calculated by recognizing and counting the microwell microparticle (4) that generates a signal by mixing the signal generating substance and the substrate solution.
[0461] The amount of the analyte in the fluid sample can be estimated by calculating the ratio of the second count value to the first count value.
[0462] As another example, an area containing the microwell microparticles (4) may be photographed using a microscope, a digital camera, or the like, and a first count value may be calculated by counting the number of microwell microparticles (4) by distinguishing them through image processing, and a second count value may be calculated by counting the number of microwell microparticles (4) that generate the fluorescence signal by distinguishing them through image processing.
[0463] As an example of estimating the amount of particles of the above-mentioned target substance, the amount can be estimated by calculating the ratio of the number of micro chambers (4) in which a signal of a detectable indicator such as a fluorescent signal is generated compared to the number of the entire micro well micro particles (4), and the ratio increases as the amount of the target substance in the fluid sample increases, and the ratio decreases as the amount of the target substance decreases.
[0464] As another embodiment of estimating the amount of particles of the above-mentioned target substance, a particle detection device for the target substance of analysis according to an embodiment of the present invention may include three or more standard substances whose concentrations are known in advance in a cartridge, generate a calibration curve indicating the concentration of the standard substances, and then correct the concentration calculation of the target substance of analysis using the generated calibration curve.
[0465] FIG. 22 is a drawing showing a microchamber containing microparticles and a microchamber generating a fluorescent signal according to a second embodiment of the present invention.
[0466] The microchamber containing microparticles that are not bound to the analyte is represented as a dark cell, and the microchamber where a fluorescent signal is generated is represented as a bright cell.
[0467] Instead of counting the number of cells, in order to calculate the ratio of the number of microchambers containing microparticles that are not bound to the analyte and the number of microchambers where the fluorescence signal is generated, by collecting bright cells and then calculating the area of the bright cell region or collecting dark cells and then calculating the area of the dark cell region, the amount of the analyte in the fluid sample can be estimated by calculating the ratio of the second count value to the first count value.
[0468] When the first capture material is a microbead (a protein such as a small bead) coated with an antibody as a microparticle according to the second embodiment of the present invention, the first capture material captures the analyte and the second capture material bound to the desired protein (analyte) in the serum or plasma sample and the bead that generates a signal by the enzyme (signal generating material) bound to the second capture material may be called an "on" bead.
[0469] In determining the concentration of the substance to be analyzed according to the second embodiment of the present invention, the concentration of the enzyme (AEB) is determined using a digital and analog analysis system, and it is preferable that the concentration is determined by the number of “on” beads (activated proteins).
[0470] The concentration of a desired protein (analyte) in a serum or plasma sample can be obtained by converting the ratio of active beads to the concentration of enzyme (AEB) using the Poisson distribution when the ratio of enzyme (signal generating substance) to beads (microparticles) is low or when there are no beads that generate signals by the enzyme.
[0471] The Poisson distribution can describe the probability of multiple events occurring when the mean number of events is known. If the expected mean number of events is μ, the probability of an exact υ occurring can be expressed as follows (Equation 12) (υ is a nonnegative integer, υ = 0, 1, 2, 3, ...).
[0472]
[0473] Here, μ is equal to the fraction of protein bound to the beads, and υ is equal to the number of enzyme-labeled proteins carried by each subpopulation of beads (i.e., 0, 1, 2, 3, etc.).
[0474] According to a second embodiment of the present invention, μ, which is equal to the concentration of the enzyme (AEB), can be determined and used as a quantitative parameter to determine the protein concentration.
[0475] Beads associated with one enzyme are indistinguishable from beads associated with two enzymes, and the population of beads associated with one enzyme appears to have a distribution up to sevenfold during the activation phase. This wide distribution allows the identification of a fraction of "off" beads (inactivated proteins) when υ = 0 (Pμ(0)).
[0476] Using Equation 12 to determine Pμ(0), and the fact that the fraction of "off" beads is equal to the fraction of "on" beads minus 1, we can determine AEBdigital (digitally determined AEB) from fon (fraction of "on" beads, or % activity).
[0477]
[0478] The protein (substance to be analyzed) concentration of a plasma sample can be accurately measured using the above mathematical formula 13.
[0479] Concentration quantification using the Poisson distribution is well measured when the activated beads are less than 70%, but has a limitation in that when the proportion of activated beads exceeds approximately 70%, the change in AEB according to concentration decreases, which increases the inaccuracy of concentration measurement.
[0480] To overcome this, AEB uses the average fluorescence intensity value of the activated beads ( ) and single enzyme( ) can be determined from the average fluorescence intensity generated by the array. The AEB of the array in the analog range (AEBanalog) is defined as in Equation 14.
[0481]
[0482] In order to determine the fluorescence intensity, it is desirable to make the AEB values of Equations 13 and 14 the same in the ratio of activated beads.
[0483] At this time, it is desirable to select an array of "on" beads <0.1 that satisfies the criterion of mathematical expression 15.
[0484]
[0485] Therefore, the enzyme concentration (AEB) can be configured for both digital and analog ranges.
[0486]
[0487] Figure 23 is a graph showing the relationship between the concentration (pg / mL) of phosphorylated tau 181 and the signal (au).
[0488] Phosphorylated tau 181 protein is a protein used in the diagnosis of Alzheimer's disease and is a protein corresponding to the target substance of the analysis of the present invention.
[0489] A second capture material binds to the phosphored-tau 181 protein, and after an enzyme (signal generating material) binds to the second capture material, a signal is generated by a substrate solution. Figure 23 shows the relationship between the concentration of phosphored-tau 181 protein and the intensity of signal generation.
[0490] Referring to Figure 23, it can be confirmed that the concentration of phosphored-tau 181 protein and the signal generation intensity are proportional.
[0491] The analysis target material that indicates the aforementioned biomarker, which can be obtained in the above step (S100), is combined with the above second capture material, so that biomarker information of the subject can be provided through the embodiments according to FIGS. 5 to 23.
[0492] The analyte to which the second capture material binds is not limited to phosphored-tau 181 protein. The analyte to which the second capture material binds may be a material corresponding to one or more biomarker information from among the first to sixth biomarker information and / or genetic information.
[0493] Specifically, the analyte to which the second capture material is bound may be one or more proteins selected from the group consisting of phosphored-tau 181, phosphored-tau 199, phosphored-tau 217, phosphored-tau 231, phosphored-tau 235, phosphored-tau 396-404, and total tau. Then, the value of the blood concentration of each tau protein in the first biomarker information can be obtained through the devices of FIGS. 5 to 23.
[0494] In addition, the analyte bound to the second capture material may be one or more proteins selected from the group consisting of amyloid beta 37, amyloid beta 38, amyloid beta 39, amyloid beta 40, and amyloid beta 42. Then, from the second biomarker information, the blood concentration of amyloid beta 42 of the subject, the ratio between the blood concentration of amyloid beta 42 of the subject and the blood concentration of amyloid beta 40, and the proportion of amyloid beta 38 among the blood concentration of total amyloid beta (total-Aβ) of the subject can be obtained through the devices of FIGS. 5 to 23.
[0495] In some embodiments, even if amyloid beta 40 is detected as a target substance for analysis, the blood concentration of amyloid beta 40 alone may be excluded from the second biomarker information and not input into the prediction model. Information about amyloid beta 40 is input into the prediction model as an indirect conversion value, such as the ratio of amyloid beta 40 to amyloid beta 42.
[0496] In addition, the analyte bound to the second capture material may be NfL (Neurofilament Light Chain), GFAP (Glial fibrillary acidic protein), Neurogranin protein, and / or sTREM2. Then, one or more values of the third biomarker information to the sixth biomarker information may be obtained through the devices of FIGS. 5 to 23.
[0497]
[0498] Using methods for assessing a subject's cognitive function level according to these examples, cognitive function level can be accurately assessed. The results of this assessment can be useful in determining whether the subject has mild cognitive impairment, Alzheimer's disease, or dementia, and in preventing or alleviating the progression of cognitive decline.
[0499] For example, it's known that without specific treatment, 10-15% of patients with mild cognitive impairment will develop dementia within one year, and 80% within six years. Accurately assessing cognitive levels to identify patients with mild cognitive impairment early and administer appropriate medications to them can prevent the progression to dementia.
[0500]
[0501] It will be apparent to those skilled in the art that a system for performing a method for assessing the cognitive function level of a subject may include other components. For example, a system for performing a method for assessing the cognitive function level of a subject may include other hardware elements necessary for the operations described herein, including an input device for data entry and an output device for printing or other data display.
[0502] When implementing an embodiment of the present invention using hardware, ASICs (application specific integrated circuits) or DSPs (digital signal processors), DSPDs (digital signal processing devices), PLDs (programmable logic devices), FPGAs (field programmable gate arrays), etc. configured to perform the present invention may be provided in the processor of the present invention.
[0503] Meanwhile, the above-described method can be written as a program that can be executed on a computer, and can be implemented on a general-purpose digital computer that runs the program using a computer-readable medium. In addition, the structure of the data used in the above-described method can be recorded on a computer-readable storage medium through various means. It should be understood that the program storage devices that can be used to describe a storage device including executable computer code for performing various methods of the present invention do not include transient objects such as carrier waves or signals. The computer-readable storage medium includes storage media such as magnetic storage media (e.g., ROM, floppy disk, hard disk, etc.) and optical reading media (e.g., CD-ROM, DVD, etc.).
[0504] The embodiments described above are combinations of components and features of the present invention in a predetermined form. Each component or feature should be considered optional unless explicitly stated otherwise. Each component or feature may be implemented without being combined with other components or features. Furthermore, it is also possible to form an embodiment of the present invention by combining some components and / or features. The order of operations described in the embodiments of the invention may be changed. Some components or features of one embodiment may be included in another embodiment or may be replaced with corresponding components or features of another embodiment. It is self-evident that claims that are not explicitly cited in the scope of the patent may be combined to form an embodiment or may be incorporated as a new claim through a post-application amendment.
[0505] It will be apparent to those skilled in the art that the present invention can be embodied in other forms without departing from the technical spirit and essential characteristics of the present invention. Therefore, the above embodiments should be considered illustrative rather than restrictive in all respects. The scope of the present invention should be determined by a reasonable interpretation of the appended claims and all possible variations within the scope of equivalents of the present invention.
[0506]
[0507] The system according to the embodiments is expected to have high industrial applicability because it supports the provision of a solution for applying an appropriate drug treatment according to the level of cognitive function of the subject by accurately and in detail determining the level of cognitive function of the subject using the subject's biomarker information and demographic information.
Claims
1. A method for determining the level of cognitive function of a subject, performed by a computing system, A step of acquiring at least one piece of information from among biomarker information of the subject and cognitive function test history information of the subject; A step of obtaining demographic information of the subject - the demographic information of the subject includes the subject's age and gender; and A step of inputting at least some of the biomarker information and cognitive function test history information of the subject, and the demographic information of the subject into a pre-learned cognitive function prediction model, thereby calculating a predicted value of the cognitive function level of the subject; including; method.
2. In paragraph 1, The biomarker information of the subject is characterized in that it includes at least one biomarker information among first biomarker information describing the tau protein of the subject, second biomarker information describing the amyloid beta protein of the subject, and third biomarker information describing the blood concentration of NfL (Neurofilament Light Chain) of the subject. method.
3. In paragraph 2, In the above first biomarker information, the subject's tau protein is at least one protein selected from the group consisting of phosphorated-tau 181, phosphorated-tau 199, phosphorated-231, and total-tau. The second biomarker information is characterized in that it includes at least one of information describing the blood concentration of amyloid beta 42 of the subject, information describing the ratio between the blood concentration of amyloid beta 42 of the subject and the blood concentration of amyloid beta 40, and information describing the proportion of amyloid beta 38 among the blood concentration of total amyloid beta (total-Aβ) of the subject. method.
4. In paragraph 1, The above cognitive function prediction model is a model configured to calculate the probability that the cognitive function index of the subject is classified into a plurality of preset cognitive function levels and to predict the cognitive function level of the subject based on the calculated classification probability. The step of calculating the predicted value of the cognitive function level of the subject is as follows: A step of inputting the input value vector of the subject and calculating the probability that the subject's cognitive function indicator is classified into a plurality of preset cognitive function levels in the cognitive function prediction model; and A step of outputting the cognitive function level with the highest probability as a predicted value of the cognitive function level of the subject; characterized in that it includes; method.
5. In paragraph 4, The probability that the cognitive function indicator of the subject is classified into multiple cognitive function levels is calculated based on a linear combination of the pre-learned regression coefficient vector and the input value vector. In the above cognitive function prediction model, multiple cognitive function levels are defined by boundary values for each level, The above regression coefficient vector and boundary values are characterized in that they are learned through a regression analysis method. method.
6. In paragraph 5, In the cognitive function prediction model, the linear combination of the regression coefficient vector and the input value vector is expressed as an inner product of the i-dimensional regression coefficient vector and the input value vector, and the input value vector is characterized in that it includes at least one of the biomarker information and cognitive function test history information, demographic information, and a preset constant. method.
7. In paragraph 4, When the above multiple cognitive function levels are set to 3 or more levels, the probability that the subject's cognitive function index is classified into the lowest level (Lmin) with the least cognitive function decline, the highest level (Lmax) with the most severe cognitive function decline, and one or more intermediate levels (Lj) between the lowest level and the highest level is calculated using the following mathematical formulas, [Mathematical formula] [Mathematical formula] [Mathematical formula] Here, ki is the boundary value for the cognitive function index that distinguishes the cognitive function level of stage i, and i.min, j, and i.max are natural numbers such as i.min <j<i.max의 관계를 가지며, β'는 회귀계수 벡터이고, x는 상기 대상자의 입력 값 벡터인 것을 특징으로 하는, method.
8. In paragraph 4, The above cognitive function prediction model is composed of multiple prediction models, each trained to predict the level of cognitive function of the subject according to the type of cognitive function. The step of calculating the predicted value of the cognitive function level of the subject is as follows: In order to produce a predicted value of the level of cognitive function for each type of the subject, the input value vector of the subject is input into each of multiple prediction models that have been pre-trained to predict the level of each type of cognitive function. The cognitive function type is characterized by one or more types of cognitive functions selected from the group consisting of memory, visuospatial sense organization, motor coordination, attention that evaluates the level of arousal, language ability, perception ability, and comprehensive cognitive function that synthesizes the abilities of each type. method.
9. In paragraph 1, At least some of the biomarker information of the above subject was obtained through a blood test method, The above blood test method is, A step of flowing a fluid sample containing a substance to be analyzed across a substrate having a microchamber array formed thereon, on the surface of which a first capture substance that specifically binds to the substance to be analyzed is fixed; a step of binding the substance to the first capture substance within each microchamber of the microchamber array; A step of causing a second capture material that specifically binds to the substance to be analyzed and binds to a signal generating substance to flow through the substrate to react the substance to be analyzed with the second capture material; A step of allowing the signal generating material to flow through the substrate and bind to the second capturing material; A step of flowing a substrate solution that reacts with the signal generating substance to generate a fluorescent signal onto the substrate; A step of allowing a hydrophobic solvent to flow through the substrate so that the signal generating material and the substrate solution react inside the microchamber, and then removing the substrate solution outside the microchamber; and A method for detecting an analyte present in a trace amount in a fluid sample, characterized in that it is obtained by including a step of counting and detecting the number of microchambers where the fluorescent signal is generated. method.
10. A computer-readable recording medium having recorded thereon a program that causes a method for determining the level of cognitive function of a subject according to any one of claims 1 to 9.
Citation Information
Patent Citations
Method and apparatus for detecting target substance in blood
KR101899733B1
Machine learning-based parking support device and method
KR1020210155657A
Heating device for clothing management
KR102705471B1
Cancer classification using patch convolutional neural networks
US20210327534A1
