System for diagnosing patient's disease through symptom reconstruction

Abstract

To provide a system for diagnosing a patient's disease which can improve the accuracy of measurement of the similarity between a symptom of a patient and a symptom in a disease.SOLUTION: A system for diagnosing a patient's disease by a symptom reconstruction includes: a disease-symptom collection unit for collecting a group of known symptoms of a disease; a symptom reconstruction unit for generating a symptom reconstruction group by classifying the group of the symptoms of the disease in the same group set in advance; and a symptom similarity measuring unit for measuring a symptom similarity by comparing a symptom of the patient and the symptom reconstruction groups for all diseases.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a patient disease diagnosis system capable of diagnosing a patient disease using symptom similarity between the patient's symptoms and known diseases, and more particularly to a patient disease diagnosis system capable of diagnosing a patient disease through symptom reconstruction.

[0002] The present invention was made under the support of the Ministry of Science and ICT (MSIT) of the Republic of Korea under Project No. 2022-0-00333, in which the research management specialist agency for the above project is the Institute of Information and Communications Technology Planning (IITP), the research project name is "SW Computing Industry Source Technology Development (R&D, Informatization)", the research project name is "Integrated SW Solution Development for Multifaceted Analysis of Pediatric Rare Diseases", the lead agency is Suribillion Co., Ltd., and the research period is from April 1, 2022 to December 31, 2024.

[0003] This application claims priority to Korean Patent Application No. "10-2023-0139926" filed with the Korean Intellectual Property Office on October 18, 2023, the disclosure of which is incorporated herein by reference. [Background technology]

[0004] The symptom similarity between the patient's symptoms and known diseases is the most important information in diagnosing the patient's disease. In the case of a patient who has not yet been diagnosed, the general procedure is to measure the degree of similarity between the known symptoms of all known diseases and the patient's symptoms, and then to check the diseases with the highest symptom similarity.

[0005] To measure the similarity of a disease, one measures the similarity between the symptoms of the patient and the symptoms of the disease, which may vary between patients with the same disease.

[0006] In addition, in the case of genetic diseases, symptoms may be similar among family members who share the same genetic defect. However, even if other independent family members have the same genetic disease, symptoms may differ between the two families. Also, symptoms may differ depending on the population used in the literature report that is the basis for compiling disease symptom information. This can be seen from the fact that more than 300 diseases registered in OMIM (Online Mendelian Inheritance in Man) explicitly state that there are clinical differences between individuals with the same disease.

[0007] FIG. 1 shows symptom information for OMIM disease ID 620158.

[0008] Referring to FIG. 1, the symptom information does not distinguish whether the symptoms are all from one individual or from one family member. As an example, in the OMIM information for the disease, among the listed symptoms, dysphagia (a disorder that makes it difficult to swallow food) is a symptom reported only by a 47-year-old male among patients with the disease. In addition, memory impairment is a symptom not reported by a 58-year-old female among patients diagnosed with the disease. Meanwhile, a 6-year-old girl diagnosed with the disease reported gait disturbance, loss of speech ability, and epilepsy, which are symptoms not reported by patients diagnosed with the disease.

[0009] As mentioned above, even if an individual has the same disease, the symptoms expressed may vary between individuals.

[0010] Conventional methods for measuring the similarity between a patient's symptoms and symptoms contained in a disease have the problem of deriving an incorrect symptom similarity, which can lead to a misdiagnosis of the patient's disease. [Prior art documents] [Patent documents]

[0011] [Patent Document 1] Republic of Korea Patent Publication No. 10-2022-0138327 (Publication date: October 12, 2022) Summary of the Invention [Problem to be solved by the invention]

[0012] The technical problem to be solved by the present invention is to provide a disease diagnosis system for a patient that can improve the accuracy of measuring the symptom similarity between the symptoms of a patient and the symptoms included in a disease. [Means for solving the problem]

[0013] In order to solve the above problem, a system for diagnosing a patient's disease through symptom reconstruction according to an embodiment of the present invention includes a disease-symptom collection unit that collects a set of symptoms for which a disease is known, a symptom reconstruction unit that generates a symptom reconstructed set by classifying the set of symptoms of the disease into a pre-defined uniform group, and a symptom similarity measurement unit that measures symptom similarity by comparing the patient's symptoms with the symptom reconstructed sets for all diseases.

[0014] The symptom reconstruction unit can generate the symptom reconstruction set so as to satisfy the following formula 1.

[0015] <Expression 1> JPEG2025069942000002.jpg7170

[0016] where S is the set of all known symptoms for the disease, and s i is the symptom reconstruction set obtained by reconstructing S.

[0017] The symptom reconstruction set (s i ) should belong to a single group (G) diagnosed with the disease, and all symptom reconstruction sets (s i ) can consist of three or more symptoms.

[0018] The symptom reconstruction set (s i ) may not be subordinate to one another in the HPO hierarchy.

[0019] The reference population can be any one of an individual, identical twins, siblings, a single family, a single population reported in the literature, or an ethnic group with a similar genetic makeup.

[0020] The symptom reconstructed set may further include a disease discrimination unit for discriminating a disease of a patient using the symptom similarity for the symptom reconstructed set.

[0021] The disease discrimination unit can select the symptom reconstructed set for a disease that has the highest symptom similarity as a representative symptom similarity for the disease, and assign a ranking according to the representative symptom similarity for each disease.

[0022] The disease determination unit can select the symptom similarity having the highest symptom similarity, and determine the disease to which the symptom reconstructed set having the highest symptom similarity belongs as the patient's disease.

[0023] In addition to the above-mentioned technical problems of the present invention, other features and advantages of the present invention will be described below or will be clearly understood by those having ordinary skill in the art to which the present invention pertains from such techniques and descriptions. Effect of the Invention

[0024] The present invention as described above has the following advantages.

[0025] The present invention classifies a set of disease symptoms into the same group to generate a symptom reconstructed set, and then measures the symptom similarity between the symptom reconstructed set and the patient's symptoms, thereby improving the accuracy of symptom similarity compared to conventional techniques that measure the symptom similarity between symptom information of a disease and the patient's symptoms.

[0026] In addition, other features and advantages of the present invention can be understood through the embodiments of the present invention. [Brief description of the drawings]

[0027] [Figure 1] FIG. 1 shows symptom information for OMIM disease ID 620158. [Diagram 2] 1 is a schematic configuration diagram of a patient disease diagnosis system through symptom reconstruction according to an embodiment of the present invention. [Diagram 3] FIG. 2 is a schematic exemplary diagram for explaining what a patient disease diagnosis system through symptom reconstruction according to an embodiment of the present invention performs. [Figure 4] 4 is an exemplary diagram for explaining how a disease determination unit according to the present invention determines a disease; FIG. [Diagram 5] 11 is another exemplary view illustrating how the disease determination unit performs disease determination according to the present invention; FIG. [Figure 6] FIG. 13 is a diagram showing the ratio of population forms that yielded the highest similarity according to the symptom similarity measurement according to the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0028] It should be noted that in this specification, when assigning reference numbers to components in each drawing, the same components will have the same numbers as much as possible, even if they appear on different drawings.

[0029] Meanwhile, the meanings of the terms used in this specification should be understood as follows.

[0030] The singular expression should be understood to include the plural expression, unless the context clearly defines otherwise.

[0031] It should be understood that terms such as "comprise" or "have" do not preclude the presence or additional possibility of one or more other features, numbers, steps, operations, components, parts, or combinations thereof.

[0032] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, preferred embodiments of the present invention devised to solve the above problems will be described in detail with reference to the accompanying drawings.

[0033] FIG. 2 is a schematic configuration diagram of a system for diagnosing a patient's disease through symptom reconstruction according to one embodiment of the present invention, and FIG. 3 is a schematic exemplary diagram for explaining what the system for diagnosing a patient's disease through symptom reconstruction according to one embodiment of the present invention performs.

[0034] Referring to FIG. 2, a system 1000 for diagnosing a patient's disease through symptom reconstruction according to one embodiment of the present invention includes a disease-symptom collection unit 100, a symptom reconstruction unit 200, a symptom similarity measurement unit 300, and a disease discrimination unit 400.

[0035] The disease-symptoms collector 100 may collect a set of symptoms for which a disease is known.

[0036] The disease-symptom collection unit 100 can collect a set of symptoms for a disease from a database 2000 that organizes and stores known symptoms for diseases, such as OMIM (Online Mendelian Inheritance in Man), D0 (Disease Ontology), and HPO (Human Phenotype Ontology).

[0037] OMIM is a continuously updated database of human genetic diseases and genes. It contains information on the genes linked to specific diseases, clinical characteristics of the diseases, and related research references.

[0038] D0 provides a hierarchical and standardized ontology of human diseases that can be used to organize and integrate information linked to complex disease classification systems with other databases.

[0039] HPO provides a standardized ontology for human phenomenological abnormalities. It provides a standard language for describing the clinical characteristics of genetic disorders and can be used in the diagnosis and research of genetic disorders. HPO uses a hierarchical structure to structure specific symptoms or phenomena for diseases and to show the relationships between symptoms that are related to each other.

[0040] The symptom reconstructing unit 200 can classify a set of symptoms of a disease into a previously set identical group to generate a symptom reconstructed set.

[0041] The symptom reconstructor 200 generates a symptom reconstruction set so as to satisfy the following formula 1, where S is the set of all known symptoms for the disease, and s i is the symptom reconstruction set obtained by reconstructing S.

[0042] <Expression 1> JPEG2025069942000003.jpg6170

[0043] In this case, the symptom reconstruction set (s i ) should belong to a single group (G) diagnosed with the disease, and all symptom reconstruction sets (s i ) can consist of three or more symptoms. Here, the symptom reconstruction set (s i ) may not be subordinate to one another in the HPO hierarchy.

[0044] The symptom reconstructing unit 200, in constructing symptom information of a disease, can reconstruct a known symptom of the disease by a set of symptoms observed simultaneously in the same population.

[0045] Actual examples are shown in Table 1 below. In this case, the criteria for the "same group" may be an individual, identical twins, siblings, a single family, a single group reported in the literature, a racial group with a similar genetic makeup, etc. In the present invention, as an embodiment, only an individual, a family, and a single group reported in the literature are exemplified as the "same group", but the criteria for the "same group" are not limited to these three.

[0046] [Table 1]

[0047] The symptoms are displayed as HPO and the disease identifier is the OMIM ID. The cluster types used are INDIVIDUAL, FAMILY, and GROUP, each of which has a GROUP identifier. Since more individual, family, and group information may be added to the database in the future, even if there is only one cluster type, it always has a GROUP identifier (similar to the example of GROUP in 620481). It can also be seen that there is an "ALL" which combines all these symptoms and constructs them as in the existing method.

[0048] The symptom similarity measuring unit 300 can measure the symptom similarity by comparing the symptoms of a patient with the symptom reconstruction set for all diseases.

[0049] The symptom similarity measuring unit 300 can use various known methods using a symptom reconstruction set for the patient's symptoms and the disease.

[0050] Unlike the conventional technology in which symptom similarity is calculated using a disease ID and a set of symptom IDs corresponding to the disease ID, the symptom similarity measurement unit 300 according to the present invention can calculate symptom similarity by specifying both a disease ID and the ID of a reconstructed symptom reconstructed set (symptom reconstructed set).

[0051] That is, the symptom similarity measuring unit 300 according to the present invention can configure information from <disease ID, type of the same group, group ID, {symptom ID}>.

[0052] Table 2 below is disease-symptom information of the prior art, and Table 3 below is an example of disease-symptom information according to the present invention.

[0053] [Table 2]

[0054] [Table 3]

[0055] The symptom similarity measuring unit 300 can measure symptom similarities between a patient and all of the same group forms of the disease using disease-symptom information obtained by dividing the same group, and the highest symptom similarity among them can be used as the symptom similarity between the patient and the disease.

[0056] The disease discrimination unit 400 can discriminate the disease of the patient by using the symptom similarity between the symptom reconstructed set of the disease and the symptoms of the patient.

[0057] FIG. 4 is an exemplary diagram for explaining how the disease determination unit according to the present invention determines a disease.

[0058] Referring to FIG. 4, the disease discrimination unit 400 according to one embodiment of the present invention can determine the disease of a patient by selecting the symptom reconstructed set for a disease with the highest symptom similarity as the representative symptom similarity for the disease and assigning a ranking according to the representative symptom similarity for each disease.

[0059] FIG. 5 is another exemplary view illustrating how the disease determination unit according to the present invention determines a disease.

[0060] Referring to FIG. 5, a disease discrimination unit 400 according to another embodiment of the present invention can select the maximum symptom similarity, and discriminate the disease to which the symptom reconstructed set having the maximum symptom similarity belongs as the patient's disease.

[0061] To confirm the effectiveness of the present invention, the symptoms diagnosed by 1,633 patients who were diagnosed with a total of 309 diseases were reconstructed by the present invention, and the symptom similarity was confirmed. The symptom similarity was calculated according to the method described in the literature "Clinical Diagnostics in Human Genetics with Semantic Similarity Searches in Ontologies." The paper used the Mutual Information of each symptom node, but in the present invention, the depth of the symptom node (the shortest distance at the top node) was used.

[0062] As a result, when using the conventional method, i.e., when symptom information is used without classification based on the morphology of the same group, the highest similarity was only 34% of the total.

[0063] FIG. 6 is a diagram showing the ratio of population forms that yielded the highest similarity according to the symptom similarity measurement of the present invention.

[0064] Referring to FIG. 6, when the symptoms of a disease for which a patient has been diagnosed are classified into individual, family, and group and symptom similarity is measured according to the present invention, the highest symptom similarity was measured from the group, which was 43% of the total, the highest.

[0065] The conventional method accounted for 34% of cases, individuals for 15%, and the remainder, families for 8%.

[0066] When the symptoms of a group and a patient were most similar, the average score difference was 4.61 points when using the method of the prior art. However, when the similarity was calculated by dividing the symptoms of the group according to the present invention, the score difference was 5.18 points, with an average difference of 0.57 points. This is found to be a very significant difference with a one-sided paired t-test p-value of <0.001.

[0067] When the individual symptoms were classified and calculated to be most similar to the patient's symptoms, the average symptom similarity was 4.62 points when using the method of the conventional technology, while it was 5.06 points when using the method of the present invention, with an average difference of 0.44 points. These are found to be very significant differences with a one-sided paired t-test p-value of <0.001.

[0068] When the symptoms were classified by family lineage and calculated to be most similar to the patient's symptoms, the average symptom similarity was 3.90 points when using the method of the conventional technology, while it was 4.28 points when using the method of the present invention, with an average difference of 0.38 points, and these are found to be very significant differences with a one-sided paired t-test p-value of <0.001.

[0069] The largest score difference between the conventional method and the method according to the present invention is shown in Table 4 below.

[0070] [Table 4]

[0071] When comparing the conventional method of calculating similarity using the overall symptoms of a disease without dividing the population, with the method of the present invention in which symptoms are divided by population and symptom similarity with patients is calculated, it was found that a particularly large difference in scores occurs when symptoms are divided by individual or family separately for calculation.

[0072] In other words, even if people have the same disease, symptoms can vary between individuals, families, and groups, and it has been found that symptoms appearing in individuals or families, which are biologically more homogeneous groups, tend to be more similar to those of patients than symptoms divided into groups.

[0073] Through this, it was confirmed that when the disease diagnosis system for a patient through symptom reconstruction according to the present invention is applied, the symptom similarity between the symptoms of a diagnosed patient and the disease for which the patient has been diagnosed is measured to be even higher.

[0074] The above objects, other objects, features and advantages of the present invention can be easily understood through the above-described preferred embodiment in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiment described herein, and may be embodied in other forms. Rather, the embodiments described herein are provided so that the disclosed content will be thorough and complete, and will fully convey the concept of the present invention to those skilled in the art. In this specification, when a component is referred to as being on another component, it means that it can be formed directly on the other component, or that a third component can be placed between them.

[0075] Additionally, when an element (or component) is referred to as operating or executing ON another element (or component), it should be understood that the element (or component) operates or executes in the environment in which the other element (or component) operates or executes, or operates or executes through direct or indirect interaction with the other element (or component).

[0076] When an element, component, device, or system is referred to as including a component consisting of a program or software, the element, component, device, or system should be understood to include the hardware (e.g., memory, CPU, etc.) necessary for the program or software to execute or operate, as well as other programs or software (e.g., operating systems, drivers necessary to run the hardware, etc.), even if not explicitly stated.

[0077] Furthermore, unless specifically stated that an element (or component) is embodied, it should be understood that the element (or component) can be embodied in either software, hardware, or software and hardware form. [Explanation of symbols]

[0078] 100. Disease-symptom collection section 200...Symptom Reconstruction Department 300...Symptom similarity measurement part 400...Disease Identification Department 1000...Disease diagnosis system for patients through symptom reconstruction

Claims

1. a disease-symptom collection unit for collecting a set of symptoms for which a disease is known; A symptom reconstructing unit that generates a symptom reconstructed set by classifying a set of symptoms of the disease into a previously set identical group; a symptom similarity measurement unit for measuring symptom similarity by comparing the symptoms of a patient with a symptom reconstruction set for all diseases; A system for diagnosing a patient's disease through symptom reconstruction, including:

2. The system for diagnosing a patient's disease through symptom reconstruction according to claim 1 , wherein the symptom reconstruction unit generates the symptom reconstruction set so as to satisfy the following formula 1: <Formula 1> where S is the set of all known symptoms for the disease, and s ｉ is a symptom reconstruction set obtained by reconstructing S.

3. The symptom reconstruction set (s ｉ ) should belong to the single population (G) diagnosed with the disease, and all symptom reconstruction sets (s ｉ 3. The system for diagnosing a patient's disease through symptom reconstruction according to claim 2, wherein the symptom reconstruction should consist of three or more symptoms.

4. The symptom reconstruction set (s ｉ The system for diagnosing a patient's disease through symptom reconstruction according to claim 3, wherein the three or more symptoms included in the first symptom (1) are not in a hierarchical relationship with each other in the HPO.

5. The system for diagnosing a patient's disease through symptom reconstruction according to claim 1, wherein the standard of the same group is any one of an individual, identical twins, siblings, a single family, a single group reported in the literature, and an ethnic group with a similar genetic makeup.

6. The system for diagnosing a patient's disease through symptom reconstruction according to claim 1 , further comprising a disease discrimination unit for discriminating a patient's disease using the symptom similarity for the symptom reconstruction set.

7. The disease diagnosis system for patients through symptom reconstruction according to claim 6, characterized in that the disease discrimination unit selects the symptom reconstructed set for a disease with the highest symptom similarity as a representative symptom similarity for the disease, and assigns a ranking according to the representative symptom similarity for each disease.

8. The disease diagnosis system of claim 6, wherein the disease discrimination unit selects the highest symptom similarity and discriminates a disease to which the symptom reconstruction set having the highest symptom similarity belongs as the patient's disease.

Images