Depression diagnosis support system

Abstract

To provide a depression diagnosis support system.SOLUTION: In a depression diagnosis support system, a depression diagnosis support device 10 comprises: a storage part 12 storing a knowledge base; and an information searching part 15 for searching at least one of, a link which indicates, a creature medical care node being a gene included in diagnosed person information, DNA (deoxyribo nucleic acid), protein, disease, medical drug, or a mental disease diagnosis standard being a node corresponding to a fact related to depression, other creature medical care node related to the same creature medical care node, and a relationship between these creature medical care nods, a human being identifying node corresponding to a human being who has information related to records of diseases other than current or past depression included in the diagnosed person information or has personal data information corresponding to information related to a living body of the diagnosed person, and a personal data node linked to the human being identifying node.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a depression diagnosis support system that supports the diagnosis of depression. [Background technology]

[0002] Conventionally, electronic medical records contain electronic medical data in various formats. Non-Patent Document 1 discloses a technology that extracts information on patients, diseases, and medicines from such electronic medical data in various formats and associates this information with existing biomedical knowledge graphs.

[0003] In addition, although the mechanism of depression onset has not yet been elucidated, there are studies showing that the amount of monoamines such as serotonin and norepinephrine is related to the onset of depression, that polymorphisms of the HTTLPR (5-hydroxytryptamine transporter-linked polymorphic region) in the promoter region of the SLC6A4 gene are related to depression, and that methylation of CpG islands around the first exon of the SLC6A4 gene is related to depression. [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] Meng Wang, Jiaheng Zhang, Jun Liu, Wei Hu, Sen Wang, Xue Li, and Wenqiang Liu, “PDD Graph: Bridging Electronic Medical Records and Biomedical Knowledge Graphs via Entity Linking”, [online], 2017, arXiv.org, [Retrieved September 14, 2021], Internet<https: / / arxiv.org / abs / 1707.05340>

[0005] [Non-Patent Document 2] Masaru Ogata, Tenpei Ikegame, Miki Bunto, Kiyoto Kasai, Kazuya Iwamoto, "1. DNA methylation of the serotonin transporter and psychiatric disorders," Journal of the Japanese Society of Biological Psychiatry, Vol. 26, No. 1, 2015, pp. 3-6 Summary of the Invention [Problem to be solved by the invention]

[0006] However, conventional technology only discloses a method to extract medical entities from MIMIC-III, an open data source of information on patients, diseases, and medicines, and link them to a biomedical knowledge graph.

[0007] For this reason, the conventional technology was not able to systematically use data at each level, ranging from higher-level knowledge to lower-level knowledge in each related field such as diagnosis and treatment of depression, biomedicine, or personal health checkups, with a focus on information on a specific disease. For this reason, the conventional technology was not sufficiently practical for supporting the diagnosis of a specific disease. [Means for solving the problem]

[0008] Therefore, as a means for solving the above-mentioned problems, the depression diagnostic support system according to the present invention is a depression diagnostic support system that supports the diagnosis of depression using a computer, and comprises a memory unit in which a knowledge base is stored, an input receiving unit that receives input of subject information on a subject who is to be diagnosed with depression, an information retrieval unit that searches for predetermined information from the knowledge base, and an output unit that outputs the information retrieved by the information retrieval unit, and the knowledge base is represented by a semantic network so that depression knowledge, which is knowledge related to depression, can be used in a systematic manner, and the knowledge base includes a plurality of biomedical nodes that represent the respective entities related to depression, such as genes, DNA (deoxyribonucleic acid) or proteins, diseases, medicines, and mental disorder diagnostic criteria; the information search unit searches for and retrieves information on genes, DNA (deoxyribonucleic acid), proteins, diseases, medicines, or mental illness diagnostic criteria included in the diagnosed person information, and the biomedical node corresponding to the facts related to depression. Search for , This searched biomedical node is linked to another said biomedical node; and retrieving relationships represented by the links connecting these biomedical nodes; and The human identification node corresponding to a person having the personal data information corresponding to information on a record of a current or past illness other than depression included in the diagnosed person information, or information on the living body of the diagnosed person. Search for the personal data node linked to the human identification node, Search for The system is characterized by comprising a search means.

[0009] In this case, the depression diagnosis support system receives input of the diagnosed person's diagnosed person information, and generates a corresponding biomedical node Search for , Link to this searched biomedical nodeOther Biomedical Nodes and retrieving the relationships represented by the links connecting these biomedical nodes; , Diagnosed person information The human identification node corresponding to Search for Personal data nodes linked to this human identification node Search for. This makes it possible to search for depression knowledge that is useful for diagnosing depression and related to the information of the person being diagnosed according to the attributes of the person being diagnosed, and that is represented by a plurality of nodes and links connecting the plurality of nodes.

[0010] "Biomedicine" refers to medical care based on a biological approach. This term includes biomedicine, biomedical, Western medicine, and personalized medicine.

[0011] In addition, the semantic network may be configured such that nodes indicating the same fact, among the biomedical nodes, the human identification nodes, or the personal data nodes, have a unified definition of the fact and are represented by an ontology.

[0012] In this case, nodes that indicate the same fact among the biomedical node, the human identification node, or the personal data node are represented by an ontology with a unified definition of the fact, and information is linked between the biomedical node, the human identification node, and the personal data node with the same fact as the contact point. This unifies the concepts between nodes in the semantic network, making it easier to search across multiple facts, and further improving the practicality of the depression diagnosis support system.

[0013] In addition, the semantic network may be configured to set synonyms corresponding to facts represented by specific nodes, and the information search unit may search for a node representing the fact corresponding to the synonym when the input receiving unit receives input of the diagnosed person information represented by the synonym.

[0014] In this case, it becomes possible to control terms using synonyms, and even if synonyms other than the expressions used in the nodes in the semantic network are input, it becomes possible to search for knowledge or facts that are adequate for diagnosing depression.

[0015] The depression knowledge may also include the probability of depression occurring together with another disease.

[0016] In this case, since the depression knowledge includes the probability of depression occurring together with other illnesses, it becomes possible to diagnose depression based on the probability of depression occurring together with other illnesses, taking into account the probability of the person being diagnosed's current or past illnesses, etc., occurring together with depression. Such other illnesses are not limited to other mental illnesses that are known to often occur together with depression, but may be other non-mental illnesses.

[0017] In addition, when the questionnaire or the diagnostic result information is assigned to the human identification node searched for by the information search unit, the output unit may output the questionnaire or the diagnostic result information.

[0018] The questionnaire may be, for example, a questionnaire related to the diagnosis of depression, such as the Patient Health Questionnaire (PHQ), or responses of the person to each question in the questionnaire, or an evaluation by a diagnoser. Further, the diagnosis result refers to the result of diagnosis as to whether or not a given person suffers from depression.

[0019] In addition, the personal data information represented by each of the plurality of personal data nodes may include at least one of information regarding gene expression, DNA (deoxyribonucleic acid) methylation, or single nucleotide polymorphisms of the human identified by the linked human identifying node.

[0020] Information on gene expression, DNA methylation, and single nucleotide polymorphisms can be extracted from data obtained, for example, by NGS (next generation sequencing).

[0021] In addition, the personal data information represented by each of the multiple personal data nodes may include at least one of medications prescribed for depression or treatment methods for depression, and the information search unit may search for medications or treatment methods prescribed for a person diagnosed with depression that contain the same personal data information as the specified information, based on information regarding current or past records of illnesses other than depression contained in the diagnosed person information, or information regarding each human body including information regarding gene expression, DNA (deoxyribonucleic acid) methylation, or single nucleotide polymorphisms.

[0022] In this case, if a patient is diagnosed with depression in the depression diagnosis, it will be possible to search for medicines or treatment methods previously prescribed for a person who has a record of an illness other than depression or biological information. This allows for the selection of medications or treatments based on past medication or treatment data for depressed patients with similar individual characteristics.

[0023] In this invention, a "fact" includes actual matters, events, and matters that are generally considered to be correct, and corresponds to an individual, specific part of the knowledge or concept in relation to "knowledge" or "concept."

[0024] Here, "personal data information" refers to information relating to data that represents the attributes of an individual (that is, personal data).

[0025] "Synonym" refers collectively to synonyms, similar words, paraphrases, or different spellings resulting from hiragana, katakana, or kanji, domestic or foreign languages, or other identifying marks. Effect of the Invention

[0026] The depression diagnostic support system of the present invention accepts input of diagnosed person information and performs at least one of searching for a corresponding biomedical node, other biomedical nodes related to this biomedical node, and links representing the relationships between these biomedical nodes, or searching for a corresponding human identification node and a personal data node linked to this human identification node. This makes it possible to search for depression knowledge that is useful for diagnosing depression and related to the information of the person being diagnosed according to the attributes of the person being diagnosed, and that is represented by a plurality of nodes and links connecting the plurality of nodes. [Brief description of the drawings]

[0027] [Figure 1] FIG. 2 is a simplified conceptual diagram of a knowledge base according to the present embodiment. [Diagram 2] FIG. 2 is a hardware configuration diagram of the depression diagnostic support system according to the present embodiment. [Diagram 3] FIG. 2 is a configuration diagram of a knowledge base explained in the present embodiment. [Figure 4] FIG. 4 is a diagram for explaining an index of the knowledge base of FIG. [Diagram 5] FIG. 2 is a flow diagram of the depression diagnostic support system according to the present embodiment. [Figure 6] FIG. 11 is a diagram for explaining a node search step S3 in this embodiment. [Figure 7] FIG. 11 is a diagram for explaining a relation node search step S3 in this embodiment. [Figure 8] FIG. 11 is a diagram for explaining a relation node search step in the present embodiment. [Figure 9] FIG. 13 is a diagram showing an example of a knowledge base using an ontology according to another embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0028] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. In this specification and the drawings, components having substantially the same functional configuration are denoted by the same reference numerals, and duplicated explanations will be omitted. Note that the following description shows an example of an embodiment of the present invention, and does not limit the technical scope of the present invention.

[0029] (Overview of the embodiment) First, an overview of the depression diagnostic support system according to this embodiment will be described.

[0030] The depression diagnosis support system can be used, for example, to support the diagnosis of depression in psychiatry etc. In particular, the depression diagnosis support system receives an input of diagnosed person information, which is information about a diagnosed person, searches for predetermined information from a knowledge base stored in the storage unit 12, and outputs the search results, thereby supporting the diagnosis of depression.

[0031] The design of the hardware elements constituting the depression diagnostic support system can be changed as appropriate, but typically, it can be composed of a diagnostician terminal 20 which is a client device selected from tablet terminals, smartphones, personal computers, etc. used in psychiatry, etc., and a depression diagnostic support device 10 which is a server device that stores a knowledge base and is configured to be able to communicate with the client device. Note that the knowledge base may be stored in the diagnostician terminal 20 itself, and the depression diagnostic support system may be composed of the diagnostician terminal 20 alone. In other words, the distribution and concentration of the hardware constituting the depression diagnostic support system can be changed as appropriate.

[0032] 1 is a simplified conceptual diagram of a knowledge base in the depression diagnosis support system in this embodiment. The knowledge base is stored in a storage unit 12 of the depression diagnosis support system. The knowledge base stores depression knowledge, which is knowledge related to depression, in a semantic network so that it can be used in a systematic manner. Depression knowledge includes information on depression, such as genes, DNA, or proteins, diseases, medicines, and mental illness diagnostic criteria, or personal data information such as information on a person who has been diagnosed with depression and their questionnaire and diagnosis results, information on the person's current or past records of diseases other than depression, or information on the person's living body.

[0033] As shown in the simplified conceptual diagram of Figure 1, the knowledge base in this embodiment holds a plurality of biomedical nodes representing the respective entities of genes, DNA, and proteins, diseases, medicines, and mental illness diagnostic criteria related to depression, and a link connecting two biomedical nodes and representing the relationship between the two biomedical nodes. The knowledge base further holds a person identification node that identifies each of a plurality of people who have been diagnosed with depression, and to which information on a depression diagnosis questionnaire and diagnosis result corresponding to the person is assigned. The knowledge base further holds a plurality of personal data nodes linked to the person identification nodes, each representing information on each person's current or past records of diseases other than depression, or personal data information including information on each person's biological body.

[0034] Here, an example of a knowledge base storing genes, DNA, and proteins will be described, but it is not necessary to store all of the facts about genes, DNA, and proteins in the knowledge base. It is sufficient to store at least one of genes, DNA, or proteins in the knowledge base. In other words, if at least one fact about genes, DNA, or proteins is stored in the knowledge base, it is possible to clarify the facts about other genes, DNA, or proteins that are not stored based on the stored genes, DNA, or proteins. This is because genes, DNA, and proteins are related to each other in that information about a specific protein is held by genes, and such genetic information is stored in a specific locus of DNA.

[0035] The knowledge base holds depression knowledge as multiple nodes and links that represent the relationship between two nodes, as shown in the simplified conceptual diagram of Figure 1. Figure 1 shows a simplified concept, and the genes, DNA, proteins, diseases, medicines, mental illness criteria, human identification symbols, questionnaires, diagnostic results, biometric information, and disease records specified in Figure 1 are each stored as concrete factual information as entities, as will be described later.

[0036] Next, in describing the depression diagnostic support system described below, the mechanism of onset of general depression will be briefly described. The onset, progression, and treatment of depression are believed to be related to the amount of monoamines transmitted between synapses in the brain, particularly serotonin and noradrenaline. However, it is also known that depression can also be caused by external factors such as changes in the living environment, such as the death of a family member, changing jobs, marriage, or divorce. Therefore, depression is broadly related to internal physiological factors and external factors such as environmental changes, and internal physiological factors are thought to be particularly related to stress resistance to environmental changes, the course of depression after onset, and treatment.

[0037] In addition, since the amount of serotonin is related to the onset of depression, the activity of serotonin transporters and serotonin receptors is related to depression. In this regard, there are drugs such as SSRIs (selective serotonin reuptake inhibitors) that attempt to treat depression by increasing the concentration of serotonin in the synaptic cleft and increasing the amount of serotonin transmitted between synapses by inhibiting the reuptake of serotonin present in the synaptic cleft by the serotonin transporter.

[0038] In addition, in recent years, there have been studies showing that in the HTTLPR (5-hydroxytryptamine transporter-linked polymorphic region) polymorphism in the SLC6A4 gene, SS-type individuals have weaker resistance to stress than LL-type individuals, i.e., they are at higher risk of developing depression (e.g., Non-Patent Document 2). This is because the promoter activity of HTTLPR, which is the promoter region of the SLC6A gene, is lower in the SS type compared to the LL type, resulting in lower gene expression levels. Also, there is a study that suggests a correlation between individuals with high DNA methylation levels in the CpG island present around the first exon of the SLC6A4 gene and individuals with a history of depression (also Non-Patent Document 2). MTHFR (methylenetetrahydrofolate reductase) is also related to depression. MTHFR is involved in folate metabolism in the blood and is related to the concentration of homocysteine ​​in folate metabolism. In addition, in the MTHFR single nucleotide polymorphism (SNP), C677T, people with the TT variant are more likely to be in a folate deficiency state than those with the wild type CC. In addition, people with folate deficiency have a higher risk of developing depression. Depression is also related to the function of the HPA axis (hypothalamus-pituitary-adrenal axis). The HPA axis has a feedback function that suppresses the secretion of adrenocorticotropic hormone via glucocorticoid receptors. In patients with depression, this feedback function is impaired, and excessive secretion of cortisol from the adrenal cortex is observed along with the excessive secretion of adrenocorticotropic hormone. When this leads to hypercortisolism, the glucocorticoid receptors continue to be stimulated, causing depressive symptoms.

[0039] (Hardware configuration) 2 is a hardware configuration diagram as an example of this embodiment. In this embodiment, an example is described in which a depression diagnostic support system is realized by a client-server system in which a diagnostician terminal 20, which is a client device, and a depression diagnostic support device 10, which is a server device, are configured to be able to communicate information with each other.

[0040] The diagnostician terminal 20 can be configured as a tablet terminal, a smartphone, a mobile phone terminal, a personal computer, etc., but in this embodiment, for example, a tablet terminal is used to configure the diagnostician terminal 20. The diagnostician terminal 20 is preferably capable of inputting diagnostician information and transmitting the diagnostician information to the depression diagnostic support device 10, and of receiving predetermined information from the depression diagnostic support device 10 and displaying it on a display means such as a screen.

[0041] The diagnostician terminal 20 preferably comprises a control unit constituted by a CPU (Central Processing Unit) or the like, a storage unit constituted by a RAM (Random Access Memory) or the like and working with the control unit to temporarily store information, an auxiliary storage unit constituted by a HDD (Hard Disk Drive) or the like and storing an OS (Operating System), various programs, and other data, a communication unit constituted by an antenna or the like and transmitting and receiving various types of information, a display unit constituted by a touch panel or the like and displaying various types of information, and an input unit constituted by the touch panel or a keyboard or the like and inputting various types of information.

[0042] The depression diagnostic support device 10 is, for example, a computer constituting a server. The depression diagnostic support device 10 includes a control unit 11, an input receiving unit 13, an output unit 14, an information search unit 15, and a storage unit 12.

[0043] The control unit 11 is configured with at least one of, for example, a CPU (Central Processing Unit), an MPU (Micro Processing Unit), or a GPU (Graphics Processing Unit), or a combination of these. The control unit 11 reads information and programs from the storage unit 12, performs predetermined processing, and controls other components that make up the depression diagnosis support device 10. In this embodiment, in cooperation with the other components, the control unit 11 performs processes such as receiving diagnosed person information, extracting attributes from the diagnosed person information, searching for nodes, searching for related nodes related to the searched nodes, and outputting the search results.

[0044] The input receiving unit 13 is configured with at least one of, for example, an antenna, a network card, a keyboard, a touch panel, etc., or a combination of these. The input receiving unit 13 is preferably capable of receiving information from a communication unit of the diagnostician terminal 20. The input receiving unit 13 cooperates with the control unit 11 to receive the diagnostician information transmitted by the communication unit of the diagnostician terminal 20. It is preferable that the depression diagnostic support device 10 and the diagnostician terminal 20 are directly or indirectly connected to each other via an antenna or a network card, a LAN (Local Area Network) or a WAN (Wide Area Network), or the like.

[0045] The output unit 14 is composed of at least one of, for example, an antenna, a network card, a display, or a projector, or a combination of these. The output unit 14 is preferably capable of transmitting information to a communication unit of the diagnostician terminal 20. The output unit 14 outputs information searched by the information search unit 15. In particular, in this embodiment, the output unit 14 transmits the information searched by the information search unit 15 to the diagnostician terminal 20. The output unit 14 may also display a result of a predetermined information processing in the depression diagnostic support device 10 on the display unit as information such as an image or text.

[0046] The information search unit 15 is composed of, for example, the control unit 11, the storage unit 12, and an information search program stored in the storage unit 12. The information search unit 15 searches for predetermined information from a knowledge base stored in the storage unit 12. The information search unit 15 searches for at least one of a biomedical node corresponding to a fact related to depression that is a gene, DNA (deoxyribonucleic acid), protein, disease, medicine, or mental illness diagnostic criterion included in the diagnosed person information, other biomedical nodes related to this biomedical node, and the link indicating the relationship between these biomedical nodes, a human identification node corresponding to a person having personal data information corresponding to information on a current or past record of a disease other than depression included in the diagnosed person information, or information on the living body of the diagnosed person, and a personal data node linked to this human identification node.

[0047] The storage unit 12 is composed of, for example, an auxiliary storage device that stores information, and a main storage device that temporarily stores information read from the auxiliary storage device. The auxiliary storage device can be composed of, for example, a magnetic disk such as a hard disk drive (HDD), a solid state drive (SSD), a flash memory such as a USB memory, or an optical disk. The storage unit 12 stores a knowledge base. The main storage device can be configured with a RAM (Random Access Memory), a register, a cache memory, or the like, which temporarily stores programs and data related to the execution of a given process.

[0048] The depression diagnosis support device 10 is configured with a control unit 11, an input receiving unit 13, an output unit 14, an information search unit 15, and a storage unit 12 all housed in a housing.

[0049] (Knowledge base configuration) Next, with reference to FIG. 3, an example of the configuration of the knowledge base stored in the storage unit 12 of the depression diagnostic support device 10 will be described. The knowledge base is formed by expressing depression knowledge, which is knowledge related to depression, in a semantic network so that it can be used in a systematic manner.

[0050] FIG. 3 is an exemplary visualization of some of the facts representing depression knowledge stored by nodes and links. In Figure 3, a node represented by a horizontally long circular shape with parallel lines above and below represents a biomedical node, a node represented by a wide rectangle above and below represents a human identification node, and a node represented by a narrow rectangle above and below represents a personal data node. A line connecting two nodes is a link indicating the relationship between the two nodes. A line with circles on both ends is not a link, but is a visualization of the fact that for convenience, nodes connected by a line with circles on both ends represent the same fact. A line with a circle on only one end represents a link. It should be noted that the nodes and links stored in the knowledge base are not limited to these.

[0051] Each biomedical node represents an entity of a gene, DNA (deoxyribonucleic acid), or protein, disease, medicine, and mental illness diagnostic criteria related to depression. A link connecting two of the multiple biomedical nodes represents a relationship between the linked biomedical nodes. A slightly elongated ellipse shown superimposed on a link represented by a straight line illustrates the content of the relationship in the link. Links that omit a circular shape omit explicit illustration in the embodiment, and such relationships will be explained later. Additionally, in one example of this embodiment in FIG. 3, the biomedical node further represents entities relating to DNA (deoxyribonucleic acid) methylation and single nucleotide polymorphisms.

[0052] More specifically, the biomedical nodes include a node representing facts about SSRIs (selective serotonin reuptake inhibitors), a node representing facts about SLC6A4 (solute carrier family 6 member 4, serotonin transporter gene), a node representing facts about the serotonin transporter, a node representing facts about the SS type of HTTLPR (one of the genotypes of a repeat polymorphism in the promoter region of the serotonin transporter), a node representing facts about 17q11.2 (the DNA locus encoding the serotonin transporter gene, located in subband 2 of band 1, region 1 of the q arm of chromosome 17), a node representing facts about DNA methylation of the SLC6A4 gene, a node representing facts about depression, and a node representing facts about the DSM-5 (Diagnostic and Statistical Manual of Mental Disorders). A node representing facts about the psychiatric diagnostic criteria for depression in the American Neuropsychiatric Disorders, 5th edition, a node representing facts about vascular dementia, a node representing facts about 5q12.3 (the DNA locus that encodes the HTR1A gene, located in subband 3 of band 2, region 1 of the q arm of chromosome 5), a node representing facts about the HTR1A gene (the gene that encodes 5HT1A, a type of serotonin receptor), a node representing facts about 5HT1A (the 5HT1A serotonin receptor, a subtype of serotonin receptor), and a node representing facts about insomnia. The map includes a node representing facts about the MTHFR gene (a gene that holds information about methylenetetrahydrofolate reductase), a node representing facts about MTHFR (methylenetetrahydrofolate reductase), a node representing facts about 1p36.22 (a DNA locus encoding the MTHFR gene located in subband 22 of band 6, region 3 of the p arm of chromosome 1), and a node representing facts about the TT type of C677T (the TT type is one of the genotypes of C677T, which is the position where a single nucleotide polymorphism (SNP) mutation occurs in the MTHFR gene). Note that biomedical nodes are not limited to these, and may include, for example, nodes of the SL type and LL type in HTTLPR, nodes of the CC type which is the wild type of C677T, or nodes of the CT type which is a heterozygote of the C677T mutant type, and nodes representing various other facts of genes, DNA, proteins, diseases, medicines, or mental disorder diagnostic criteria.

[0053] Furthermore, the plurality of biomedical nodes are linked by a link that indicates a relationship between two related biomedical nodes. In particular, the depression node and the vascular dementia node are both linked to indicate that there is a high possibility that the respective diseases co-occur. Furthermore, the depression node and the insomnia node are both linked to indicate that there is a high possibility that the respective diseases co-occur.

[0054] SSRIs are known as antidepressants prescribed to patients suffering from depression. SSRIs increase the concentration of serotonin in the brain by inhibiting the reuptake of serotonin by the serotonin transporter in the synaptic cleft, thereby exerting an antidepressant effect. A link between the SSRI node and the depression node represents the relationship between the therapeutic drug and the target disease. A link between the SSRI node and the serotonin transporter node represents the relationship between the drug and the target cell (or target organ).

[0055] SLC6A4 is a gene that codes for serotonin transporter information. The link between the SLC6A4 node and the serotonin transporter node represents the relationship between a gene, which is information on a specific base sequence, and a protein expressed by the gene. The link between the SLC6A4 node and the SS-type node of HTTLPR represents the relationship between a gene and one of the polymorphisms in the promoter region of the gene. The link between the SLC6A4 node and the 17q11.2 node represents the relationship between a gene and a locus in DNA where the gene is encoded. The link between the SLC6A4 node and the SLC6A4 methylation node represents the relationship between a gene and DNA methylation in the gene.

[0056] The serotonin transporter adjusts the concentration of serotonin in the synaptic cleft by taking up serotonin, a neurotransmitter present in the synaptic cleft in the brain, and adjusts the amount of serotonin received by the serotonin receptors present in the postsynaptic cells. The serotonin transporter node and the depression node are linked to show the relationship between depression and the protein that controls the amount of serotonin transmitted, which is one of the causes of depression. The serotonin transporter node and the SSRI node, and the serotonin transporter node and the SLC6A4 node are linked as described above.

[0057] HTTLPR is the promoter region of the SLC6A4 gene that codes for the serotonin transporter. HTTLPR has repeat polymorphisms of SS type, SL type, and LL type. The SS type has lower stress resistance and is more susceptible to depression than the SL type. Here, a node representing the SS type of HTTLPR is provided. The HTTLPR SS type node and the depression node are linked to represent the relationship between a genotype with a high probability of developing a disease and the disease in question. The HTTLPR SS type node and the SLC6A4 gene node are linked as described above.

[0058] 17q11.12 is the DNA locus that contains the genetic code for SLC6A4. The 17q11.12 node and the SLC6A4 node are linked as described above.

[0059] DNA methylation is considered to be the central mechanism of action of epigenetics, and methylation occurs for cytosine and adenine. It is known that when DNA methylation occurs at a high level in the promoter region of a certain gene, the level of protein expression in that gene is reduced. In other words, when DNA methylation occurs in the HTTLPR promoter region of SLC6A, the expression of the protein serotonin transporter is suppressed. The link between the DNA methylation node of the SLC6A4 gene and the depression node represents a relationship between one of the causes of a disease and the disease itself. The DNA methylation node of the SLC6A4 gene and the SLC6A4 node are linked as described above.

[0060] The depression node is linked to many other biomedical nodes, namely, the depression node is linked to the SSRI node, the serotonin transporter node, the SS type of HTTLPR node, the DNA methylation node of the SLC6A4 gene, the DSM-5 node, the vascular dementia node, the HTR1A node, the 5HT1A node, the insomnia node, and the TT type of C677T node.

[0061] DSM-5 is the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders. A DSM-5 node represents a fact of the diagnostic criteria for depression in DSM-5. More specifically, it is preferable that a DSM-5 node holds information on the threshold score of a depression diagnosis as a criterion for diagnosing depression. A link between the DSM-5 node and the depression node holds the relationship between the diagnostic criteria for a disease and the disease in question.

[0062] Vascular dementia is dementia caused by cerebral infarction or cerebral hemorrhage due to vascular disorders in the brain. Patients with vascular dementia often also suffer from depression. For this reason, a link between a vascular dementia node and a depression node represents a relationship in which the two diseases are highly likely to co-occur. In addition, it is preferable that the link indicating the co-occurrence relationship holds, as depression knowledge, the fact of the probability of the two diseases co-occurring.

[0063] 5q12.3 is the DNA locus that encodes the HTR1A gene. A link between the 5q12.3 node and the HTR1A node represents the relationship between the location of DNA genetic information and the gene retained at that location.

[0064] HTR1A is a gene that holds information about 5HT1A, a type of serotonin receptor. The link between the HTR1A node and the 5HT1A node represents the relationship between a gene, which is information about a specific base sequence, and the protein expressed by that gene. The link between the HTR1A node and the depression node represents the relationship between a gene and a disease related to that gene. The HTR1A node and the 5q12.3 node are linked as described above.

[0065] 5HT1A is a type of serotonin receptor. The 5HT1A node and the depression node are linked to each other, showing the relationship between the protein and the disease associated with that protein. The 5HT1A node and the HTR1A node are linked as described above.

[0066] Insomnia is a symptom seen in the majority of depression patients. A link between an insomnia node and a depression node indicates a high probability of the co-occurrence of the respective diseases. In addition, it is preferable that a node connecting an insomnia node and a depression node holds, as depression knowledge, a fact about the probability of the co-occurrence of insomnia and depression, for example, 80%.

[0067] The MTHFR gene is a gene that codes for MTHFR information. Links between the MTHFR gene node and the MTHFR node represent the relationship between the gene and the protein encoded by the gene. Links between the MTHFR gene node and the 1p36.22 node represent the relationship between the gene and the locus in DNA that holds the genetic information of the gene. Links between the MTHFR gene node and the TT type node of C677T represent the relationship between the gene and one of the genotypes at the position where a single nucleotide polymorphism occurs in the gene.

[0068] MTHFR is methylenetetrahydrofolate reductase. MTHFR is involved in folate metabolism and also in the production of neurotransmitters. The MTHFR node and the MTHFR gene node are linked as described above.

[0069] 1p36.22 is the DNA locus that encodes the genetic information for MTHFR. The 1p36.22 node and the MTHFR gene node are linked as described above.

[0070] C677T is one of the single nucleotide polymorphisms (SNPs) in the MTHFR gene. The TT type of C677T is one of the genotypes of the single nucleotide polymorphism C677T, and is the TT type, which is a homozygous mutant type. The link between the TT type node of C677T and the depression node represents the relationship between the genotype of the single nucleotide polymorphism that increases the risk of developing a certain disease and the disease. The TT type node of C677T and the MTHFR gene node are linked as described above.

[0071] In addition, the biomedical nodes are not limited to those described above. For example, the knowledge base preferably holds nodes representing facts such as the HPA axis, glucocorticoid receptors, adrenocorticotropic hormone, or hypercortisolism. For example, the knowledge base preferably holds a node of a DEX / CRH test as a test capable of measuring a feedback function related to the secretion of adrenocorticotropic hormone via glucocorticoid. For example, the node of the DEX / CRH test preferably holds information in which the degree of decline in the feedback function is associated with the possibility of depression and used as a certain index.

[0072] Next, the knowledge base described in this embodiment holds two human identification nodes. The human identification nodes identify each of a plurality of people who have been diagnosed with depression, and are provided with information on the depression diagnosis questionnaire and diagnosis results corresponding to the person. In this embodiment, two human identification nodes are illustrated, but this is for the sake of convenience of explanation, and it is preferable that the knowledge base further holds a plurality of human identification nodes.

[0073] Linked to each human identification node are a number of personal data nodes each representing personal data information including information regarding each human's current or past record of illnesses other than depression, or information regarding each human's biology.

[0074] In this embodiment, the human identification node for identifying human A holds identification information for identifying human A, as well as an answer to a Patient Health Questionnaire (PHQ), which is a questionnaire for diagnosing depression, and information indicating depression as a result of the PHQ, as property information. Note that it is preferable to use the PHQ-9 or PHQ-8, which are related to mental illness, as the PHQ.

[0075] As the identification information for identifying person A, for example, an identification code uniquely associated with each person may be provided, or multiple people may be identified by the name of the person. Furthermore, such an identification method is not limited to primary key information that is uniquely identified by only one key, and may be identified by a composite key that combines multiple pieces of information such as name, date of birth, and gender.

[0076] The questionnaire for diagnosing depression may be any questionnaire used for diagnosing depression in the person A, and is not limited to the PHQ. Examples of the questionnaire for diagnosing depression include the Hamilton Depression Rating Scale (HAM-D, SIGH-D, or GRID-HAMD), the Hamilton Anxiety Rating Scale (HAM-A), the Brief Psychiatric Rating Scale (BPRS), the Clinician-Administered Dissociative States Scale (CADSS), the Scale for Suicide Ideation (SSI), the Global Assessment of Functioning Scale (GAF), the Montgomery-Asberg Depression Rating Scale (MADRS), the Mood Disorder Questionnaire (MDQ), and the Structured Clinical Interview for DSM (SCID).

[0077] The result of the depression diagnosis of a human identified by the human identification node may be binary information indicating whether or not the person is depressed, or may be a numerical representation of the severity of the symptoms of depression.

[0078] Similarly, the human identification node that identifies human B holds identification information for identifying human B, as well as property information such as the answer to a PHQ (Patient Health Questionnaire), which is a questionnaire related to the diagnosis of depression, and the information that the PHQ result indicates depression.

[0079] Furthermore, a link between the human identification node that identifies human A and the human identification node that identifies human B represents the parent-child relationship between human A and human B. This shows that the genetic and genomic information of human A and human B is similar. Furthermore, by expressing the relationship between humans identified by the human identification node with a link, it is also possible to express the possibility of sympathetic psychosis (foliadou).

[0080] Next, the knowledge base described in this embodiment holds a number of personal data nodes, which are linked to the person identification node and each represent personal data information including information about a current or past record of a disease other than depression in each person, or information about the biology of each person. In this embodiment, the knowledge base holds personal data nodes that represent personal data information on two people identified by two person identification nodes. However, this is due to the fact that only two person identification nodes are shown for the sake of convenience, and it is preferable that the knowledge base further holds personal data nodes linked to multiple person identification nodes.

[0081] In the knowledge base according to this embodiment, personal data nodes representing the facts of SSRI, depression, SS type of HTTLPR, age 20, female, and asthma as a medical history are linked to the human identification node of human A. In addition, in the knowledge base according to this embodiment, personal data nodes representing the facts of SS type of HTTLPR, methylation of SLC6A4, depression, TT type of C677T, insomnia as a medical history, age 40, and male are linked to the human identification node of human B. The facts represented by each node will be described in more detail below.

[0082] The SSRI node, which is a personal data node linked to the human identification node of human A, represents the fact that an SSRI has been prescribed as a treatment for depression for human A. In addition, a node representing the fact of information about such a treatment may also hold facts such as the dosage and administration method determined at the time of prescription.

[0083] The depression node, which is a personal data node linked to Person A's human identity node, represents the fact that Person A suffers from depression.

[0084] The SS type node of HTTLPR, which is a personal data node linked to the human identification node of human A, represents the fact that the polymorphism of HTTLPR, which is the promoter region of the SLC6A4 gene of human A, is SS type.

[0085] The 20-year-old node, which is a personal data node linked to the human identification node of human A, represents the fact that human A is 20 years old. Note that the node representing the age of a human identified by the human identification node may represent the age in increments of one year, or may represent age classes such as teens and twenties.

[0086] The female node, which is a personal data node linked to the human identity node of Person A, represents the fact that Person A is female, preferably biological sex.

[0087] The SS type node of HTTLPR, which is a personal data node linked to the human identification node of human B, represents the fact that the polymorphism of HTTLPR, which is the promoter region of the SLC6A4 gene of human B, is SS type.

[0088] The methylation node of the SLC6A4 gene, which is a personal data node linked to the human identification node of human B, indicates the fact that high DNA methylation has occurred in the HTTLPR, which is the promoter region of the SLC6A4 gene in human B. In this case, the expression of the SLC6A4 gene in human B is suppressed.

[0089] The depression node, which is a personal data node linked to Person B's human identity node, represents the fact that Person B suffers from depression.

[0090] The TT type node of C677T, which is a personal data node linked to the human identification node of human B, represents the fact that C677T, a single nucleotide polymorphism in the MTHFR gene of human B, is a TT type genotype.

[0091] The insomnia node, which is a personal data node linked to Person B's human identity node, represents the fact that Person B has a history of insomnia.

[0092] The 40 years old node, which is a personal data node linked to the human identification node of person B, represents the fact that person B is 40 years old. In this case, too, age information may be represented as knowledge by a hierarchy, as explained for person A.

[0093] The Male node, which is a personal data node linked to Person B's human identity node, represents the fact that Person B is male.

[0094] In addition, in FIG. 3, among the personal data nodes linked to the human identification node of human A, the SSRI node, the depression node, and the SS type node of HTTLPR are connected to the SSRI node, the depression node, and the SS type node of HTTLPR in the biomedical node by lines with circles on both ends. Similarly, among the personal data nodes linked to the human identification node of human B, the SS type node of HTTLPR, the methylation node of SLC6A4 gene, the depression node, the TT type node of C677T, and the insomnia node are connected to the SS type node of HTTLPR, the methylation node of SLC6A4 gene, the depression node, the TT type node of C677T gene, and the insomnia node in the biomedical node by lines with circles on both ends. The nodes connected by lines with circles are merely a visual representation of the same facts, and are not intended to indicate that they are linked to each other. For such identical facts, it is preferable to unify the definition of the fact and express it in an ontology. For this purpose, it is preferable that, for example, personal data nodes and biomedical nodes are controlled by an ontology consisting of the same symbols. An example of the use of a particularly suitable ontology in this embodiment will be described later.

[0095] Additionally, the personal data information represented by a personal data node linked to a person identifying node may include marital status, standard of living, income, education level, occupation, physical examination results, or diet. For example, marital status, standard of living, income, level of education, and occupation are external factors such as stress related to the onset of depression. Also, physical examination results and diet are related to internal factors such as neurotransmitters, transporters, receptors, enzymes, amino acids, or components of body fluids such as blood, which are related to the onset of depression. Therefore, when these facts are represented by human identification nodes, the relationship between these facts and depression can be represented by a semantic network.

[0096] (Explanation of the processing flow) Next, the flow of processing of the depression diagnostic support system according to this embodiment will be described with reference to FIG.

[0097] First, in a diagnosed person information receiving step S1, the input receiving unit 13 of the depression diagnostic support device 10 receives input of diagnosed person information about a diagnosed person who is to be diagnosed with depression. In this embodiment, a doctor or the like inputs diagnosed person information to the diagnoser terminal 20, and this information is communicated between a communication unit of the diagnoser terminal 20 and the input receiving unit 13 of the depression diagnostic support device 10, thereby enabling the depression diagnostic support device 10 to receive the diagnosed person information.

[0098] Here, the diagnosed person information includes information corresponding to the facts related to depression, such as genes, DNA, or proteins, diseases, medicines, or mental illness criteria, or information related to the diagnosed person's current or past records of diseases other than depression, or information related to the diagnosed person's living body. Such information can be obtained by doctors or sequence operators, for example, through electronic medical records, questionnaires related to depression diagnosis, NGS (next generation sequencing), blood tests, and other diagnoses, tests, interviews, etc. More specifically, examples of the diagnosed person information that can be envisioned include genetic, DNA, or protein information that can be obtained by sequencing technology such as NGS, cell testing, or protein testing, vital signs, urine, blood, body temperature, blood pressure, weight, CT (Computed Tomography), MRI (Magnetic Resonance Imaging), or electroencephalogram information obtained by various tests, or information on mental illness criteria thresholds related to psychological tests, etc. Furthermore, the diagnosed person information can include information such as records of current or past illnesses, treatment history, current or past prescribed medicines, changes in the environment such as family, work, worries, and lifestyle, or the appearance and facial expression of the diagnosed person, which can be obtained from the diagnosed person's medical records or interviews.

[0099] In this embodiment, the diagnosed person information for explaining the present embodiment includes the following attribute information as an example. Age 30 ·woman -HTTLPR polymorphism is SS type · The single nucleotide polymorphism of C677T is TT type For example, a psychiatrist can use the patient terminal 20 to send the patient's medical record information and NGS information to the depression diagnostic device, and the patient information can be received by the input receiving unit 13 of the depression diagnostic device.

[0100] Next, in attribute extraction step S2, the depression diagnostic support device 10 extracts attribute information relating to facts related to depression, such as genes, DNA, proteins, diseases, medicines, or mental illness diagnostic criteria, or information regarding current or past records of diseases other than depression, or information regarding the body of the diagnosed person, from the medical record information and NGS information of the diagnosed person examined by the depression diagnostic support device 10 in diagnosed person information receiving step S1. Here, the depression diagnostic support device 10 extracts attribute information of the subject, such as "HTTLPR SS type" and "C677T TT type," as DNA facts related to depression, and attribute information of the subject, such as "30 years old" and "female," as information about the subject's biological body.

[0101] As a method for extracting such attributes, for example, the diagnosed person information, which is information such as electronic medical records or NGS transmitted from the diagnostician terminal 20 to the depression diagnostic support device 10, may be tagged for each predetermined attribute and structured, or the depression diagnostic support device 10 that receives the diagnosed person information may extract a part of the information that meets predetermined conditions as attribute information. For example, when structuring the diagnosed person information, a data structure such as RDF (Resource Description Framework) may be used.

[0102] Next, in node search step S3, the depression diagnostic support device 10 searches for a biomedical node or personal data node corresponding to the attribute information of the diagnosed person extracted in attribute extraction step S2 from the knowledge base stored in the memory unit 12 of the depression diagnostic device.

[0103] In this embodiment, as shown in FIG. 6, in node search step S3, the information search unit 15 of the depression diagnostic support device 10 searches the knowledge base for biomedical nodes and personal data nodes corresponding to the attribute information of the diagnosed person. 6, the nodes indicated by solid lines are nodes that are hit by the search in node search step S3. Note that the use of solid and dashed lines is for the sake of convenience in explaining the depression diagnosis support system in this embodiment. Here, the attribute information of the diagnosed person contained in the diagnosed person information is DNA, and "HTTLPE SS type" and "C677T TT type" have been extracted as facts related to depression, and "30 years old" and "female" have been extracted as information about the diagnosed person's biological body. Therefore, among the biomedical nodes held in the knowledge base, the corresponding "HTTLPR SS type node" and "C677T TT type node" are hits, and also, among the personal data nodes held in the knowledge base, the corresponding "female node" linked to the human identification node of Person A is hit.

[0104] Fig. 7 is a diagram for explaining the related node search step S4. In Fig. 7, nodes marked with a star represent biomedical nodes that were found in the search in node search step S3 as corresponding to the attribute information of the diagnosed person (hereinafter, for convenience, these nodes are referred to as corresponding nodes). In Fig. 7, the biomedical nodes and links that are additionally represented by solid lines compared to Fig. 6 are respectively nodes that were found in the related node search step S4 as related nodes related to the corresponding nodes marked with a star, and links that represent the relationships between the corresponding nodes and the related nodes. Here, the MTHFR gene node, SLC6A4 node, and depression node are newly represented by solid lines, and the link connecting the C677T TT type node and the MTHFR gene node, the link connecting the C677T TT type node and the depression node, the link connecting the HTTLPR SS type node and the SLC6A4 node, and the link connecting the HTTLPR SS type node and the depression node are also represented by solid lines.

[0105] As shown in Figure 7, in related node search step S4, other biomedical nodes related to the biomedical node (corresponding node) that was found to correspond to the attribute information of the diagnosed person in node search step S3, and links representing the relationship between the corresponding node and other biomedical nodes related to the corresponding node are searched for. As an example in this embodiment, the SLC6A4 node and the depression node connected by a direct link to the HTTLPR SS type node corresponding to the attribute information of the diagnosed person are hit by a search of related nodes as other biomedical nodes related to the corresponding node, and the link connecting the HTTLPR SS type node and the SLC6A4 node and representing the relationship between one of the polymorphisms in the promoter region and the gene in which the promoter region is located, and the link connecting the HTTLPR SS type node and the depression node and representing the relationship between a genotype with a high probability of developing a disease and the disease itself are also hit by the search. Furthermore, in this embodiment, the MTHFR gene node and depression node connected by a direct link to the C677T TT type node corresponding to the attribute information of the diagnosed person are hit by a search for related nodes as other biomedical nodes related to the corresponding node, and the link connecting the C677T TT type node and the MTHFR gene node, representing the relationship between a single nucleotide polymorphism in a gene and the gene in which the single nucleotide polymorphism occurs, and the link connecting the C677T TT type node and the depression node, representing the relationship between the genotype of a single nucleotide polymorphism that increases the risk of developing a specific disease and the disease, are also hit by the search.

[0106] In addition, by searching for corresponding nodes and related nodes related to the corresponding nodes, it is possible to systematically search for data at all levels, from higher-level knowledge to lower-level knowledge in related fields such as the diagnosis and treatment of depression, biomedicine, or personal health checkups, such as HTTLPR polymorphisms and SLC6A4, or C677T and MTHFR genes. In the present embodiment, in the step S4 of searching for related nodes, related nodes directly linked to the corresponding node are searched for. That is, in the present embodiment, related nodes at a depth of 1 from the corresponding node are searched for. However, in the step S4 of searching for related nodes, the search depth may be adjusted according to the strength of the relationship of the information to be obtained.

[0107] Fig. 8 is also a diagram for explaining the related node search step S4. In Fig. 8, the nodes marked with a star represent personal data nodes that are found in the search in the node search step S3 as corresponding to the attribute information of the person to be diagnosed (hereinafter, for convenience, these nodes are referred to as corresponding nodes). In Fig. 8, the human identification nodes, personal data nodes, and links that are additionally represented by solid lines compared to Fig. 6 represent nodes that are found in the related node search step S4 as related nodes related to the corresponding nodes marked with a star, and links that represent the relationships between the corresponding nodes and the related nodes. Here, the human identification node of human A and the 20-year-old node, SSRI node, depression node, and asthma node, which are personal data nodes linked to the human identification node of human A, are represented by solid lines, and links connecting the human identification node of human A and the 20-year-old node, SSRI node, depression node, HTTLPR SS type node, asthma node, and female node are represented by solid lines. Furthermore, the human identification node of human B and the 40-year-old node, SLC6A4 methylation node, depression node, insomnia node, and male node, which are personal data nodes linked to the human identification node of human B, are represented by solid lines, and links connecting the human identification node of human B and the 40-year-old node, HTTLPR SS type node, SLC6A4 methylation node, depression node, C677T TT type node, insomnia node, and male node are represented by solid lines. In addition, in this embodiment, a link that connects the human identification nodes of human A and human B searched for in relationship node search step S4 and represents the parent-child relationship between human A and human B is also searched for in relationship node search step S4, and the link connecting human A's human identification node and human B's human identification node is also represented by a solid line.

[0108] In the related node search step S4 shown in Fig. 8, a person identification node linked to the personal data node (corresponding node) hit in the node search step S3 as expressing the personal data information corresponding to the attribute information of the diagnosed person, and other personal data nodes linked to this person identification node are searched for. Note that in this embodiment, the link connecting the person identification node and the personal data node is also searched for, and if a predetermined relationship is indicated in this link, this relationship is also acquired. As an example in this embodiment, the search returns a human identification node for human A and a human identification node for human B that are linked to a personal data node representing at least one of the personal data information of the subject, which is 30 years old, female, SS type of HTTLPR, or TT type of C677T. Furthermore, in this embodiment, as an example, as personal data nodes linked to the human identification nodes hit by the search, a 20-year-old node, an SSRI node, a depression node, and an asthma node linked to the human identification node of human A are hit by the search of related nodes. Furthermore, as personal data nodes linked to the human identification nodes hit by the search, a 40-year-old node, an SLC6A4 methylation node, a depression node, an insomnia node, and a male node linked to the human identification node of human B are hit by the search of related nodes.

[0109] Here, by searching for attribute information of the person diagnosed, more specifically, information on current or past records of illnesses other than depression contained in the diagnosed person information, or personal data information corresponding to information on the living body of the diagnosed person, a human identification node corresponding to a person having personal data information corresponding to the diagnosed person information, and a personal data node linked to this human identification node, it is possible to search for information on the living body of a person who has been diagnosed with depression, such as current or past illnesses, genes, DNA, proteins, age, and sex, for each person who has been diagnosed with depression. This makes it easy to search for information based on past cases of depression diagnoses, and makes it possible to obtain information useful for diagnosing depression.

[0110] Next, in search result output step S5, the depression diagnostic support system outputs the nodes and links found in node search step S3 and related node search step S4 as search results. In this embodiment, the depression diagnostic support device 10 outputs the nodes and links found in node search step S3 and related node search step S4 as search results from the output unit 14. Then, the communication unit of the diagnostician terminal 20 receives information on the search results and displays it on the display unit of the diagnostician terminal 20. This allows the diagnostician to easily and reliably diagnose depression based on the facts about depression retrieved from the knowledge base.

[0111] (Explanation of the Actions and Effects of the Present Embodiment) The depression diagnosis support system according to this embodiment can search for a biomedical node corresponding to a fact related to depression that is a gene, DNA, protein, disease, medicine, or mental illness diagnostic criterion included in the diagnosed person information, other biomedical nodes related to this biomedical node, and links indicating the relationships between these biomedical nodes, and output them to the diagnoser terminal 20. Furthermore, based on information about current or past diseases other than depression included in the diagnosed person information, or information about the diagnosed person's living body, it can search for a human identification node of a human having information corresponding to current or past diseases other than depression or information about the diagnosed person's living body, and a personal data node linked to this human identification node, and output them to the diagnoser terminal 20. This makes it possible to search for and confirm depression knowledge represented by a plurality of nodes and links connecting the plurality of nodes, which is useful for diagnosing depression and related to the diagnostician's information according to the attributes of the diagnostician.

[0112] In this embodiment, the knowledge base includes the probability of depression occurring together with other diseases. Since the depression knowledge held in the knowledge base includes the probability of depression occurring together with other diseases, it is possible to diagnose depression taking into account the probability of depression occurring together with a current or past disease of the diagnosed person, based on the probability of depression and other diseases occurring simultaneously.

[0113] In this embodiment, the output unit 14 outputs the questionnaire or diagnostic result information when the questionnaire or diagnostic result information is assigned to the person identification node searched by the information search unit 15. This makes it possible to confirm the questionnaire answers or diagnostic result information at the time of depression diagnosis of a specific person held in the knowledge base, and to confirm the information represented by the personal data node of the specific person. This makes it possible to perform depression diagnosis more easily and reliably.

[0114] In this embodiment, the personal data information represented by each of the multiple personal data nodes may include at least one of information on gene expression, DNA methylation, or single nucleotide polymorphism of the human identified by the linked human identification node. Here, the gene expression information may represent, for example, a gene expression level indirectly estimated from gene polymorphism information, or may be represented as a personal data node by directly measuring the gene expression level. In this case, it is possible to confirm the relationship between the genetic information related to the onset of depression and the human identification node linked to the personal data node representing that genetic information. This makes it possible to diagnose depression by referencing information from past depression diagnoses in comparison with the genetic information of the person being diagnosed.

[0115] Furthermore, the personal data information held in the knowledge base preferably includes at least one of medicines prescribed for depression or treatment methods for depression. In this case, the information retrieval unit 15 can retrieve medicines or treatment methods prescribed for people diagnosed with depression who share common information about the body of the person based on information about current or past records of illnesses other than depression, or information about each organism including information about gene expression, DNA methylation, or single nucleotide polymorphism. This makes it easier to propose treatments for depression that are tailored to each individual, taking into account commonalities in biological information.

[0116] (Other embodiments) Next, as another embodiment, an example is shown in which the semantic network of the depression diagnostic support system is such that nodes that indicate the same fact among biomedical nodes, human identification nodes, and personal data nodes are expressed by an ontology with the definition of the fact unified. In addition to the example in which the definition of a certain fact is unified by an ontology, an example is also described in which a synonym corresponding to the fact represented by a certain node is set in the semantic network. It should be noted that the description of the same matters as those in the embodiments already described will be omitted.

[0117] FIG. 9 is a conceptual diagram showing an example of a knowledge base of a depression diagnostic support system according to another embodiment. The knowledge base in this embodiment holds, as biomedical nodes, a depression node, an asthma node, an SS type node of HTTLPR, a serotonin transporter node, a DSM5 node, a PHQ8 node, and an SSRI node. Also, the knowledge base in this embodiment holds, as human identification nodes, a human identification node for human A. Also, the knowledge base in this embodiment holds, as personal data nodes, a female node, a 20s node, an asthma node as medical history, an SS type node of HTTLPR, and a depression node as medical history.

[0118] The nodes that compose each semantic network, i.e., biomedical nodes, human identification nodes, and personal data nodes, that represent the same fact are represented by an ontology that unifies the definition of that fact. A description of each of these ontologies is shown in the table in Figure 9.

[0119] An ontology can be constructed by unifying the definition of certain facts through a combination of symbols consisting of certain alphanumeric characters, hiragana, katakana, kanji, etc., and representing the certain facts with the same symbols.

[0120] In Figure 9, the fact of depression is listed as "dis_mdd", the fact of asthma is listed as "dis_asthma", the fact of no disease is listed as "dis_null", the DSM5 definition of depression is listed as "dif_dsm5_mdd", the fact of PHQ9 is listed as "tes_phq9", the fact of the serotonin transporter gene is listed as "gen_slc6a4", the fact of HTTLPR SS type is listed as "gen_httlprss", the fact of HTTLPR SL type is listed as "gen_httlprsl", and the fact of HTTLPR LL type is listed as "gen_httlp The fact that it is a serotonin transporter is "pro_5htt", the fact that it is a female is "sex_f", the fact that it is a male is "sex_m", the fact that it is under 19 is "age_0019", the fact that it is in its 20s is "age_2029", the fact that it is in its 30s is "age_3039", the fact that a person A is "hum_123a", the fact that a person B is "hum_123b", and the fact that it is a selective serotonin reuptake inhibitor is "med_ssri".These facts are unified and represented by the ontology. It should be noted that the symbols of the ontology expressed here can be changed as appropriate, and any symbols may be used as long as each fact can be uniquely distinguished.

[0121] Here, for example, the "dis_mdd node", which is a depression node in the biomedical node, and the "dis_mdd node", which is a depression node as a medical history among the personal data nodes linked to the human identification node of person A, have a unified definition of the fact of depression through the ontology, which is the same symbol.

[0122] In this way, when the semantic network is such that the nodes showing the same fact among the biomedical node, the human identification node, and the personal data node are represented by an ontology with a unified definition of the fact, the same fact is linked between the biomedical node, the human identification node, and the personal data node using the ontology symbol as a key. This unifies the concepts between the nodes in the semantic network, making it easy to search multiple facts across the board. That is, for example, when searching for facts related to depression, the information search unit 15 of the depression diagnosis support device 10 searches for "dis_mdd," making it easy to search across the biomedical node, the human identification node, and the personal data node for "dis_mdd nodes" showing the same fact.

[0123] In addition, in this embodiment, synonyms corresponding to facts represented by specific nodes are set in the semantic network, and when the input receiving unit 13 receives input of diagnosed person information represented by synonyms, the information search unit 15 can search for nodes representing facts corresponding to the synonyms.

[0124] That is, for example, in this embodiment, when the information "Clinical depression" is included in the diagnostician information, the information retrieval unit 15 of the depression diagnosis support device 10 can retrieve a node including "Clinical depression" as a synonym. In this embodiment, since the "dis_mdd node" indicating the fact of depression has "Clinical depression" as a synonym, it is possible to retrieve the "dis_mdd node" corresponding to the diagnostician information from the semantic network.

[0125] Furthermore, the ontology may unify various relationships, such as the relationship between a disease and a gene, or the relationship between a disease and a single nucleotide polymorphism, by using a unified notation.

[0126] (Modification) In addition, technologies such as depression diagnosis support systems and graph convolutional networks (GCNs) may be used to predict the existence of information that is not held in the knowledge base, i.e., information for which nodes and relationships are not held. For example, by collapsing a graph consisting of a human identification node held in a semantic network and a personal data node linked to the human identification node, a learning model can be used to explore the relationship between the human represented by the collapsed human identification node and personal data information not held in the knowledge base. In this case, it is possible to predict personal data information and relationships not present in the knowledge base using a large number of human identification nodes and personal data stored in the knowledge base and links representing the relationships between them. This allows the knowledge base to be used more flexibly, making it possible to better support depression diagnosis.

[0127] The technical scope of the present invention is not limited to the matters described in the embodiment, and the matters described in the embodiment can be appropriately combined and implemented by those skilled in the art. Furthermore, all such modifications fall within the technical scope of the present invention. [Explanation of symbols]

[0128] 10. Depression diagnosis support device 11 Control section 12 Storage section 13 Input reception section 14 Output section 15 Information Search Department 20 Diagnostician terminal

Claims

1. A depression diagnosis support system that supports the diagnosis of depression using a computer, comprising: The system includes a storage unit in which a knowledge base is stored, an input receiving unit that receives input of subject information relating to a subject who is to be diagnosed with depression, an information retrieval unit that retrieves predetermined information from the knowledge base, and an output unit that outputs the information retrieved by the information retrieval unit, The knowledge base comprises: Depression knowledge, which is knowledge about depression, is represented by a semantic network so that it can be used in a systematic manner; maintaining a plurality of biomedical nodes representing respective entities of genes, DNA (deoxyribonucleic acid) or proteins, diseases, medicines, and mental illness diagnostic criteria related to depression, and a link connecting two biomedical nodes to each other and representing a relationship between the two biomedical nodes; A system for identifying each of a plurality of people who have been diagnosed with depression, and holding a person identification node to which information on a questionnaire related to the depression diagnosis and information on the diagnosis result corresponding to each person is attached, and a plurality of personal data nodes linked to the person identification node and each representing personal data information including information on a record of a current or past illness other than depression for each person, or information on the biological body of each person, The information search unit is searching for the biomedical node corresponding to a fact related to depression among genes, DNA (deoxyribonucleic acid), proteins, diseases, medicines, or mental illness diagnostic criteria included in the diagnosed person information; Searching for other biomedical nodes connected to the searched biomedical node by links and relationships represented by the links connecting these biomedical nodes; searching for the human identification node corresponding to a human having the personal data information corresponding to information on a current or past record of a disease other than depression included in the diagnosed person information, or information on the living body of the diagnosed person; A search means for searching the personal data node linked to the human identification node is provided. A depression diagnosis support system.

2. The semantic network includes: Nodes indicating the same fact among the biomedical node, the human identification node, or the personal data node are represented by an ontology with a unified definition of the fact.

2. The depression diagnostic support system according to claim 1 .

3. In the semantic network, synonyms corresponding to facts represented by specific nodes are set, The information search unit searches for a node representing the fact corresponding to the synonym when the input receiving unit receives the input of the diagnostician information represented by the synonym.

3. The depression diagnosis support system according to claim 1 or 2.

4. The depression knowledge includes the probability of depression occurring together with other diseases.

4. The depression diagnostic support system according to claim 1,

5. The output unit is When the questionnaire or the diagnostic result information is given to the human identification node searched for by the information search unit, the questionnaire or the diagnostic result information is output.

5. The depression diagnostic support system according to claim 1,

6. The personal data information represented by each of the plurality of personal data nodes includes: The linked human identification node includes at least one of information on gene expression, DNA (deoxyribonucleic acid) methylation, or single nucleotide polymorphism of the human identified by the linked human identification node.

6. The depression diagnostic support system according to claim 1,

7. The personal data information represented by each of the plurality of personal data nodes includes: At least one of a medication prescribed for depression or a method for treating depression is included, The information search unit is Based on information on current or past records of illnesses other than depression contained in the diagnosed person information, or information on each person's biological body including information on gene expression, DNA (deoxyribonucleic acid) methylation, or single nucleotide polymorphism, a search is performed for medicines or treatment methods prescribed for people who have been diagnosed with depression and who contain personal data information identical to the predetermined information.

7. The depression diagnostic support system according to claim 1,

Images