JP2019207634A - Diagnosis assist device, diagnosis assist method, and diagnosis assist program - Google Patents

Abstract

To enable a disease candidate to be extracted by using information existing between symptoms and diseases.SOLUTION: A diagnosis assist device 1 according to an embodiment of the present invention comprises: a receiving unit 141 for receiving input having at least one word indicating a symptom, an opinion, a disease state, and a disease; a graph acquiring unit 142 for acquiring a graph indicating correlation among the symptom, the opinion, the disease state, and the disease for each case stored in a storage unit; a score calculating unit 143 for specifying a path to an identified disease which is at least one of diseases included in the graph based on the symptom, the opinion, the disease state and the disease corresponding to a word in one or a plurality of graphs and for calculating points associated with the identified disease based on the specified path; and an output unit 144 for outputting, with respect to the identified disease, information corresponding to the point calculated by the score calculating unit 143.SELECTED DRAWING: Figure 2

Description

  The present invention relates to a diagnosis support apparatus, a diagnosis support method, and a diagnosis support program for supporting diagnosis in medicine.

  In medical practice, a doctor diagnoses a disease based on various information obtained by examining a patient. Patent Document 1 describes a system that extracts and displays a disease corresponding to an input symptom from a database in which the symptom and the disease are associated with each other.

Japanese Patent Laid-Open No. 2004-29927

  Even if it is the same symptom, it may be a different disease, and conversely, even if it is a different symptom, it may be the same disease, so doctors comprehensively judge the information obtained for each patient and determine the disease Need to diagnose. However, since the system described in Patent Document 1 only directly associates a symptom with a disease, it is not possible to use information between the disease, disease state, finding, and symptom when deriving the disease. . In addition, there is no existing technology that can learn by confirming the case that is the basis of the evaluation while combining various cases.

  The present invention has been made in view of these points, and provides a diagnosis support apparatus, a diagnosis support method, and a diagnosis support program that can extract disease candidates using information between disease, pathology, findings, and symptoms. For the purpose.

  The diagnosis support apparatus according to the first aspect of the present invention includes a reception unit that receives an input of at least one word indicating a symptom, a finding, a pathological condition, or a disease, and a symptom, a finding, or a pathological condition for each case that is stored in the storage unit. And an acquisition unit that acquires a graph showing a relationship between diseases, and a differential disease that is at least one disease included in the graph from symptoms, findings, pathologies, or diseases corresponding to the word in one or more of the graphs And a calculation unit that calculates a score associated with the differential disease based on the specified route, and outputs information corresponding to the score calculated by the calculation unit for each differential disease And an output unit.

  The storage unit stores the graph indicating a symptom, a finding, a pathological condition, and a direction of a chain between diseases for each case, and the calculating unit includes a symptom, a finding, and a pathological condition corresponding to the word in the graph. Alternatively, the route from the disease to the differential disease upstream of the graph may be specified.

  The calculation unit may identify at least one route as the route in ascending order of distance in the graph.

  The output unit may output a portion constituting the path in the graph in addition to information corresponding to the score.

  The output unit may output information indicating the case corresponding to the route in addition to the information corresponding to the score.

  The calculation unit may calculate the score associated with the differential disease based on a weight set for each of the associations on the identified route.

  The calculation unit includes the weight of the relationship between the symptom or finding and the symptom or finding, the weight of the relationship between the symptom or finding and the disease state or disease, and the disease state or disease and the disease state or disease. The points may be calculated by setting the weights related to each other to be different from each other.

  The calculation unit may specify the path constituted by a plurality of the cases from a symptom, a finding, a disease state, or a disease corresponding to the word to the differential disease in the graph.

  The calculation unit may calculate the score based on the number of cases constituting the path.

  The output unit outputs related information associated with the differential disease, and the reception unit receives the selected related information in addition to the word when any of the related information is selected. May be.

  In addition to the word received by the reception unit, the calculation unit may specify the path from a symptom, a finding, a disease state, or a disease corresponding to a related word associated with the word to the differential disease.

  The accepting unit accepts an input of a word indicating a symptom, a finding, a disease state, or a disease to be excluded in specifying the route in addition to the input of the word, and the calculating unit corresponds to the word in the graph The path from the symptom, finding, disease state or disease to the differential disease to the symptom, finding, disease state or disease corresponding to the exclusion subject may be specified.

  The reception unit may receive an input of a disease classification in addition to the input of the word, and the output unit may output information corresponding to the score of the differential disease included in the classification.

  The graph stored in the storage unit may indicate the association between symptoms, findings, disease states, and diseases by at least one of a directed graph and an undirected graph.

  In the diagnosis support method according to the second aspect of the present invention, the processor accepts an input of at least one word indicating a symptom, a finding, a disease state or a disease, and a symptom or a finding for each case stored in the storage unit. Obtaining a graph showing an association between a pathological condition and a disease, and distinguishing one or more symptoms included in the graph from symptoms, findings, pathological conditions or diseases corresponding to the word in the one or more of the graphs A path to a disease is identified, and a score associated with the differential disease is calculated based on the identified path, and information corresponding to the score calculated by the calculating step is determined for each differential disease. And a step of outputting.

  The diagnosis support program according to the third aspect of the present invention includes a step of receiving at least one word indicating a symptom, a finding, a disease state, or a disease in a computer, and a symptom and a finding for each case stored in the storage unit. Obtaining a graph showing an association between a pathological condition and a disease, and distinguishing one or more symptoms included in the graph from symptoms, findings, pathological conditions or diseases corresponding to the word in the one or more of the graphs A path to a disease is identified, and a score associated with the differential disease is calculated based on the identified path, and information corresponding to the score calculated by the calculating step is determined for each differential disease. And a step of outputting.

  The diagnosis support apparatus according to the fourth aspect of the present invention includes a reception unit that receives an input of at least one word indicating a symptom, a finding, a disease state, or a disease, and a symptom, a finding, or a disease state for each case that is stored in the storage unit. And an acquisition unit that acquires a graph showing a relationship between diseases, and a differential disease that is at least one disease included in the graph from symptoms, findings, pathologies, or diseases corresponding to the word in one or more of the graphs An output unit that identifies a path to the path and displays a portion of the graph that constitutes the path on the display unit for each of the differential diseases.

  According to the present invention, it is possible to extract a disease candidate using information between a symptom and a disease.

It is a mimetic diagram of a diagnosis support system concerning this embodiment. It is a block diagram of a diagnosis support device according to the present embodiment. It is a figure showing the disease recall chain graph which shows the direction of the linkage between a symptom, a finding, a disease state, and a disease. It is a schematic diagram of the search screen which the display part which concerns on this embodiment displays. It is a schematic diagram explaining the cost (weight) set to the disease recall chain graph which a graph memory | storage part memorize | stores. It is a schematic diagram explaining the score calculation part the calculation method of the score (score) with respect to a disease recall chain graph. It is a schematic diagram of the detailed display screen which the display part which concerns on this embodiment displays. It is a schematic diagram of the modification of the search screen which the display part which concerns on this embodiment displays. It is a schematic diagram of the modification of the search screen which the display part which concerns on this embodiment displays. It is a figure which shows the flowchart of the diagnostic assistance method which the diagnostic assistance apparatus which concerns on this embodiment performs.

[Outline of diagnosis support system]
FIG. 1 is a schematic diagram of a diagnosis support system according to the present embodiment. The diagnosis support system includes a diagnosis support apparatus 1. The diagnosis support apparatus 1 includes a storage unit 11, an input unit 12, and a display unit 13. The storage unit 11, the input unit 12, and the display unit 13 may be directly connected to the diagnosis support apparatus 1, or may be connected via a network such as the Internet or a local area network. The diagnosis support system may include other devices such as a server and a terminal.

  The diagnosis support apparatus 1 is a computer that supports diagnosis of a disease by extracting one or a plurality of diseases based on information input by a user and outputting them as differential diseases. The user who uses the diagnosis support apparatus 1 is a doctor, for example. A user who is a doctor examines a patient and inputs information related to the patient to the diagnosis support apparatus 1 using the input unit 12. And the user who is a doctor refers to the information used as the candidate of the differential disease displayed on the display part 13 by the diagnosis assistance apparatus 1, and diagnoses a patient's disease. Hereinafter, a candidate for a differential disease is simply referred to as a differential disease. The user who uses the diagnosis support apparatus 1 is not limited to a doctor, and may be a medical worker such as a nurse or a general person.

[Configuration of Diagnosis Support Device 1]
FIG. 2 is a block diagram of the diagnosis support apparatus 1 according to the present embodiment. In FIG. 2, arrows indicate main data flows, and there may be data flows not shown in FIG. In FIG. 2, each block represents a functional unit configuration, not a hardware (device) unit configuration. Therefore, the blocks shown in FIG. 2 may be implemented in a single device, or may be separately implemented in a plurality of devices. Data exchange between the blocks may be performed via any means such as a data bus, a network, a portable storage medium, or the like.

  The diagnosis support apparatus 1 includes a control unit 14 in addition to the storage unit 11, the input unit 12, and the display unit 13 described above. The storage unit 11 is a storage medium including a ROM (Read Only Memory), a RAM (Random Access Memory), a hard disk drive, and the like. The storage unit 11 stores a program executed by the control unit 14 in advance. The storage unit 11 may be provided outside the diagnosis support apparatus 1. In this case, the storage unit 11 transmits and receives signals to and from the control unit 14 via the interface of the diagnosis support apparatus 1.

  The storage unit 11 includes a graph storage unit 111 and a related word storage unit 112. Each of the graph storage unit 111 and the related word storage unit 112 may be a storage area on the storage unit 11 or may be a database configured on the storage unit 11. The graph storage unit 111 stores in advance a disease recall chain graph indicating a directional linkage and a non-directional linkage between symptoms, findings, pathological conditions, and diseases for each case. The related word storage unit 112 stores in advance related words of words that can be input from the user and related information related to the differential disease.

  The input unit 12 includes an input device such as a keyboard and a mouse for receiving user input. The input unit 12 inputs a signal indicating a user operation to the control unit 14. The display unit 13 includes a display device such as a liquid crystal display or a CRT (Cathode Ray Tube) display for displaying various types of information. The display unit 13 displays information according to the signal received from the control unit 14. The input unit 12 and the display unit 13 may be integrally configured by using a touch screen capable of detecting the position of contact by the user as the display unit 13.

  The control unit 14 includes a reception unit 141, a graph acquisition unit 142, a score calculation unit 143, and an output unit 144. The control unit 14 is a processor such as a CPU (Central Processing Unit), for example, and by executing a program stored in the storage unit 11, as the reception unit 141, the graph acquisition unit 142, the score calculation unit 143, and the output unit 144. Function. The functions of the reception unit 141, the graph acquisition unit 142, the score calculation unit 143, and the output unit 144 will be described later. At least a part of the function of the control unit 14 may be executed by an electric circuit. Further, at least a part of the functions of the control unit 14 may be executed by a program executed via a network.

  The diagnosis support apparatus 1 according to the present embodiment is not limited to the specific configuration shown in FIG. The diagnosis support apparatus 1 is not limited to one apparatus, and may be configured by connecting two or more physically separated apparatuses in a wired or wireless manner.

[Explanation of disease recall chain graph]
FIG. 3 is a diagram illustrating a disease recall chain graph showing a relationship among symptoms, findings, pathological conditions, and diseases stored in the graph storage unit 111. A disease is a disease (eg, appendicitis). A disease state is a general term for changes in the structure or function of the human body that is the background of a disease (for example, inflammation). The findings are items observed as a result of visual inspection, palpation, percussion, auscultation, examination, etc. on the patient or the like (for example, calcification of the pleura found by CT (Computed Tomography) examination). Symptoms are changes in individual conditions that appear in the patient (eg, nausea). One or a plurality of diseases recalled by the diagnosis support apparatus 1 based on words indicating symptoms, findings, pathological conditions, or diseases received from the user are referred to as differential diseases.

  The disease recall chain graph is a directed graph that shows a directional chain and an undirected chain, with specialists extracting the associations of symptoms, findings, pathologies, and diseases that may have occurred in actual patients in cases reported in the past. Is described as an undirected graph indicating In other words, the disease recall chain graph shows, for each case, the causal logic of the linkage from the underlying cause disease to the disease and pathology, and the symptoms or findings observed in the disease (i.e. Observation). In FIG. 3, a directional chain is represented by an arrow, and a non-directional chain is represented by a broken line. The disease recall chain graph may include a test name effective for disease evaluation and a prescription example for the disease.

  The disease recall chain graph is composed of nodes (nodes) representing each symptom, finding, disease state and disease, and edges (branches) connecting the nodes while maintaining the direction from the cause to the result. And shows an association between diseases. In the disease recall chain graph, a symptom, a finding, a disease state, and a disease that are relatively on the upstream side are upstream, and a symptom, a finding, a disease state, and a disease that are relatively on the downstream side are downstream. For example, in case 1 shown in FIG. 3, appendicitis is upstream (ie, cause), and iliopsoas abscess is downstream (ie, result). In addition, the disease recall chain graph may include a disease and symptoms and findings observed in the disease. For example, in case 1, appendicitis is a disease and fever is a symptom observed with appendicitis.

  In addition, diseases that have a common pathology and are known to combine (temporal arteritis and polymyalgia rheumatica, etc.), diseases that have the same pathology and that do not have the same pathology (polymyalgia rheumatica) And RS3PE syndrome), pathologies whose causality is reversed depending on severity (heart failure and atrial fibrillation, etc.), and symptoms that are usually known to occur at the same time (nausea, vomiting, etc.) are undirected chains (undirected graphs) Combined. For example, in case 2 shown in FIG. 3, nausea and vomiting are linked by a non-directional chain.

  The disease recall chain graph illustrated in FIG. 3 is a tree format graph. The disease recall chain graph is not limited to this, and may be other directed graphs or undirected graphs. The disease recall chain graph may be a combination of a directed graph and an undirected graph. The disease recall chain graph is stored in the graph storage unit 111 in any format readable by the control unit 14, such as a database table, a list of character strings, binary data, or the like.

  The diagnosis support apparatus 1 uses the disease recall chain graph stored in the graph storage unit 111 in this way to identify and output a differential disease based on information input by the user.

[Description of disease identification method]
FIG. 4 is a schematic diagram of a search screen displayed by the display unit 13 according to the present embodiment. The search screen includes an input field A and a result field B. The display unit 13 may display only the input field A when the differential disease has not been extracted yet.

  The input field A includes a search word field A1 for inputting words indicating symptoms, findings, pathological conditions, or diseases used by the user for disease extraction, and a virtual transmission button A2 for transmitting the input content. .

  When the reception unit 141 detects that the user has pressed the transmission button A2 using the input unit 12, the reception unit 141 receives the content input by the user in the search term column A1 as input information. The graph acquisition unit 142 acquires the disease recall chain graph stored in the graph storage unit 111. And the score calculation part 143 calculates the score (score) for every differential disease based on the input information which the reception part 141 received using the disease recall chain graph which the graph acquisition part 142 acquired.

  The score calculated by the score calculation unit 143 is based on the cost (weight) set in the disease recall chain graph, and the causal logic on the path specified in the disease recall chain graph is close (that is, between the cause and the result). The smaller the logical gap of (), the smaller. The farther the causal logic on the path is, the larger it is. That is, the score corresponds to the distance in the disease recall chain graph. Therefore, the route with the smallest score among the routes for which the scores are calculated is referred to as the shortest route.

  The score calculated by the score calculation unit 143 is not limited to the definition shown here, and other definitions may be used. For example, the score may be defined such that the larger the value, the closer the causal logic is, and the smaller the value, the farther the causal logic is.

  A method for calculating a score by the score calculation unit 143 will be described below. FIG. 5 is a schematic diagram for explaining the cost (weight) set in the disease recall chain graph stored in the graph storage unit 111. FIG. 5 represents a first node S corresponding to a symptom or finding as a circle, a second node D corresponding to a disease state or a disease as a rectangle, and an edge L between them as an arrow.

  The score calculation unit 143 obtains one whole graph in which the graphs of the plurality of cases are integrated by identifying the places where the same disease appears in the graphs of the plurality of cases as described in FIG. Each node of the overall graph is associated with one or more cases that are the original. The score calculation unit 143 may acquire an entire graph stored in advance in the graph storage unit 111, or may generate an entire graph by integrating graphs of a plurality of cases when calculating a score. The example of FIG. 5 represents an overall graph in which case α and case β are integrated. As a result, the score calculation unit 143 configures a route by combining a plurality of cases even when not all of the symptoms, findings, pathological conditions, or diseases input by the user are included in one case, and the user inputs A differential disease corresponding to a symptom, finding, disease state or disease can be extracted.

  The score calculation unit 143 uses at least one of the node cost Wn set for each node and the edge cost We set between the nodes as the cost (weight) set for each chain included in the disease recall chain graph. To get. The node cost Wn and the edge cost We are weights used to calculate the score of the path on the disease recall chain graph. That is, the score of the route passing through the node and the edge in the disease recall chain graph is determined based on the node cost Wn and the edge cost We.

  The score calculation unit 143 may acquire the node cost Wn and the edge cost We stored in advance in the graph storage unit 111. Alternatively, the score calculation unit 143 may determine the node cost Wn and the edge cost We for each node and edge of the disease recall chain graph when calculating the score.

  The node cost Wn and the edge cost We may be set differently for all nodes and edges, or may be set uniformly for all nodes and edges.

  The node cost Wn may be set to a different value depending on whether the node corresponds to a symptom, a finding, a disease state, or a disease. For example, the node cost Wn of symptoms or findings and the node cost Wn of disease state or disease are set to be different from each other. The edge cost We may be set to a different value depending on whether the nodes at both ends of the edge correspond to a symptom, a finding, a disease state, or a disease. For example, the edge cost We of an edge between a symptom or finding and a symptom or finding, an edge cost We of an edge between a symptom or finding and a disease state or disease, and an edge between a disease state or disease and a disease state or disease Are set to be different from each other.

  Specifically, the edge cost We is reduced because the logical gap between the symptom and the pathological condition tends to be small, and the edge cost We is decreased because the logical gap between the pathological condition and the pathological condition tends to be large. Enlarge. In this way, by changing the cost of the node or edge for each symptom, finding, disease state, or disease, a differential disease can be extracted in a form close to the actual causal logic.

  The node cost Wn and the edge cost We may be set to different values depending on the symptom, finding, disease state, or disease appearance probability corresponding to the node and the edge. For example, a symptom, a finding, a disease state, or a disease having a higher appearance probability in the entire graph has a lower node cost Wn and an edge cost We, and a symptom, a finding, a disease state, or a disease that has a lower appearance probability in the entire graph. The node cost Wn and the edge cost We are set so that the node cost Wn and the edge cost We are increased. Thereby, the differential disease can be extracted by reflecting the symptom, the findings, the pathological condition, the ease of appearance of the disease, and the difficulty of appearance.

  FIG. 6 is a schematic diagram illustrating how the score calculation unit 143 calculates a score (score) for the disease recall chain graph. First, the score calculation unit 143 selects one or a plurality of nodes corresponding to a symptom, a finding, a disease state, or a word indicating a disease input by the user in the disease recall chain graph. Here, the score calculation unit 143 may select a node that completely matches a word indicating a symptom, finding, disease state, or disease input by the user, or may select a node that partially matches the word. Nodes that match the word associated with can be selected.

  Next, the score calculation unit 143 selects a path from one or a plurality of nodes corresponding to a word indicating a symptom, a finding, a disease state, or a disease input by the user to a differential disease node upstream in common with those nodes. And the score of the identified route is calculated. FIG. 6 shows two first nodes S corresponding to words indicating a symptom, a finding, a disease state, or a disease input by the user, and a second node D of a differential disease located upstream thereof, by hatching.

  Specifically, the score calculation unit 143 is configured to select one from one or a plurality of nodes corresponding to a word indicating a symptom, a finding, a disease state, or a disease input by the user to a differential disease node upstream of those nodes. One or more path scores are calculated.

The score calculation unit 143 calculates the score by, for example, an equation of (ΣWn + ΣWe) × k. ΣWn is the total value of the node costs Wn on the route. ΣWe is the total value of the edge costs We on the route. k is a coefficient corresponding to the number of cases constituting the path, and is represented by, for example, an expression of log ₂ (number of cases + 1). Thereby, k takes a larger value as the number of cases is larger, and takes a smaller value as the number of cases is smaller. This is because when the number of cases constituting the route is large, the logical separation tends to be large. In addition, since k is defined as a logarithm of the number of cases, the smaller the number of cases (for example, when the number of cases is changed from 1 to 2), the greater the influence on the score is. In the example of FIG. 6, since two cases are combined to form a route from the first node S to the upstream second node D, the number of cases constituting the route is two.

  And the score calculation part 143 specifies the path | route with the lowest score (namely, shortest path | route) among the 1 or several path | routes from the node corresponding to the word which the user input to the node of the differential disease in the upstream. At the same time, the score calculated for the identified route is output. Further, the score calculation unit 143 may specify a plurality of routes (for example, a predetermined number of routes in ascending order of score) whose score satisfies a predetermined condition.

  The score calculation unit 143 can use a known route search algorithm such as the Dijkstra method that can specify a route based on the node cost Wn, the edge cost We, and the number of cases constituting the route. The score calculation unit 143 may calculate a score using either the node cost Wn or the edge cost We. Further, the edge cost We of the edge across the disease recall chain graph of a plurality of cases may be set to be high so as to reflect the number of cases constituting the route.

  With such a score calculation method, the score calculation unit 143 has close causal logic on the route based on the node cost Wn, the edge cost We, and the number of cases constituting the route (that is, the logical relationship between the cause and the result). The score is calculated so that it is smaller as the gap is smaller and larger as the causal logic on the path is farther away.

  Specifically, the greater the number of nodes and edges included in the route from the first node S to the upstream second node D, the greater the score of the route, so the causal logic is farther away (ie, the second node D Is likely to be the cause of the first node S). On the other hand, the smaller the number of nodes and edges included in the route from the first node S to the upstream second node D, the smaller the score of the route, so the causal logic is closer (that is, the second node D is the first one). It is estimated that there is a high probability of causing the node S).

  Moreover, since the score of a path | route becomes so large that there are many cases which comprise the path | route from the 1st node S to the 2nd node D upstream, it cannot explain by the cause and effect which occurred in the known case. It means that the probability of causal logic is low. On the other hand, the smaller the number of cases constituting the route from the first node S to the second node D upstream, the smaller the score of the route, and thus it is estimated that the probability of causal logic is high.

  As described above, the score calculation unit 143 calculates a score for each differential disease based on information on a route from the symptom or the like input by the user to the differential disease upstream thereof. Therefore, the diagnosis support apparatus 1 can extract a differential disease more accurately by using information between a symptom and the like that could not be considered in the past and the disease. It is also possible to create a graph that collects information on all cases in relation to disease and symptom, and to evaluate each case separately in relation to disease and illness. It is also possible to express the relationship between the relationship and the strength of the relationship between the disease and the disease.

  In the present embodiment, the score calculation unit 143 searches for a differential disease in the upstream direction of the disease recall chain graph from the symptoms and the like input by the user, but the score calculation unit 143 proceeds in the downstream direction of the disease recall chain graph or the disease You may search for a differential disease in both the upstream direction and the downstream direction of the recall chain graph.

  In the example of FIGS. 5 and 6, the score calculation unit 143 calculates the score using one overall graph in which the disease recall chain graphs of a plurality of cases are integrated. As another method, the score calculation unit 143 may calculate the score using a plurality of disease recall chain graphs defined for each case.

  The output unit 144 outputs information corresponding to the score calculated by the score calculation unit 143 by a predetermined method. Returning to FIG. 4, a method of outputting information by the output unit 144 will be described. The output unit 144 causes the display unit 13 to display the result column B as information corresponding to the score calculated by the score calculation unit 143. The result column B includes a differential disease B1, a score B2, case information B3, and a route B4.

  Specifically, the output unit 144 ranks the symptom, the finding, the pathological condition, or the differential disease that is upstream of the disease corresponding to the word input by the user in the overall graph in descending order of the score calculated by the score calculation unit 143. To do. Then, the output unit 144 causes the display unit 13 to display a predetermined number (or all) of diseases as the differential disease B1 in the order from the lowest score together with the ranking of the scores. The output unit 144 also causes the display unit 13 to display the score calculated by the score calculation unit 143 as the score B2 for each differential disease B1. Thereby, the user can recognize the closeness of the causal logic, that is, the small logical separation between the cause and the result for each differential disease B1.

  In addition, the output unit 144 causes the display unit 13 to display case information B3 (for example, an ID for identifying a case) indicating cases constituting the route specified by the score calculation unit 143. When the user selects any of the cases indicated by the case information B3 using the input unit 12, the output unit 144 acquires the disease recall chain graph of the selected case from the graph storage unit 111 and displays it on the display unit 13. You may let them. Thereby, the user can obtain information on cases related to the differential disease B1.

  Further, the output unit 144 causes the display unit 13 to display a part constituting the route B4 specified by the score calculation unit 143 in the disease recall chain graph. Thus, the user can know what causal logic, that is, what symptoms, findings, pathological conditions, or disease chains have extracted the differential disease B1. In the example of FIG. 4, the output unit 144 displays one route B4 for one differential disease B1 on the display unit 13, but displays a plurality of routes B4 for one differential disease B1 on the display unit 13. Also good.

  The information output method illustrated in FIG. 4 is an example, and the output unit 144 may output information by other methods. For example, the output unit 144 may print information using a printer. The output unit 144 may store data indicating information in the storage unit 11.

  When the user selects one of the routes B4 using the input unit 12 on the search screen of FIG. 4, the selected route B4 is changed to the detailed display screen of FIG. FIG. 7 is a schematic diagram of a detail display screen displayed by the display unit 13 according to the present embodiment. The detail display screen includes a detail display field C. The detailed display column C includes a differential disease C1, a score C2, case information C3, and an extended route C4. The differential disease C1, the score C2, and the case information C3 are the differential disease B1, the score B2, and the case information B3 corresponding to the route B4 selected by the user on the search screen of FIG.

  The output unit 144 further includes at least a part of other nodes and edges connected to the route B4 in addition to the portion constituting the route B4 selected by the user on the search screen of FIG. 4 in the disease recall chain graph. Is displayed on the display unit 13. For example, in the example of FIG. 7, “tachycardia” and “weight loss” downstream of “Graves' disease” are not included in the route B4, but are displayed in the extended route C4 as a portion connected to the route B4. ing. Thereby, the user can obtain information such as other symptoms related to the differential disease C1.

  When the user selects any of the cases indicated by the case information C3 using the input unit 12, the output unit 144 displays the display mode (that is, the color or pattern) of the node corresponding to the case selected in the extended path C4. , Shape, etc.) to be different from other nodes. Thereby, when the extended route C4 is configured from a plurality of cases, the user can distinguish which case each node is included in.

[First Modification]
The diagnosis support apparatus 1 may accept priority target, exclusion target, and category designation as input information, in addition to words indicating symptoms, findings, disease states, or diseases used by the user for disease extraction. FIG. 8 is a schematic diagram of a modified example of the search screen displayed by the display unit 13 according to the present embodiment. The search screen includes an input field A, a result field B, and a category designation field E. The contents of the result column B are the same as the search screen of FIG.

  In addition to the search word field A1 and the send button A2 similar to the search screen of FIG. 4, the input field A is a priority object field A3 for inputting a word indicating a symptom, a finding, a disease state, or a disease to be prioritized by the user. And an exclusion target column A4 for inputting words indicating symptoms, findings, pathological conditions, or diseases to be excluded by the user. When the user selects any node or edge displayed in the result column B using the input unit 12, the output unit 144 selects the selected node or edge in the priority target column A3 or the exclusion target column A4. You may add or delete from priority object column A3 or exclusion object column A4.

  The category designation field E includes a category E1 and a virtual designation button E2 for switching presence / absence of designation for each category E1. The disease recall chain graph of each case stored in the graph storage unit 111 is classified in advance into at least one of a plurality of categories E1. When the user presses the designation button E2 for each category E1 using the input unit 12, the user can switch the designation of each category E1. The category E1 can be displayed with the number of differential diseases obtained as a result, for example.

  The category designation field E can be switched between a display state and a non-display state according to a user operation. The category designation column E in the display state is superimposed on the result column B, and the category designation column E in the non-display state is removed from the result column B.

  When the reception unit 141 detects that the user has pressed the transmission button A2 using the input unit 12, the contents input by the user in the search term column A1, the priority target column A3, and the exclusion target column A4, and the category E1 The presence / absence of designation is accepted as input information.

  The score calculation unit 143 calculates a score for each disease using the score calculation method described with reference to FIG. However, the score calculation unit 143 sets a node cost Wn of a node corresponding to a word indicating a symptom, a finding, a disease state, or a disease that is a priority object, as compared with a case where it is not a priority object. Thereby, the diagnosis support apparatus 1 can lower the score of the route including the symptom, finding, pathological condition, or disease that the user wants to prioritize, and can adjust the rank of the differential disease.

  Further, the score calculation unit 143 calculates a score by excluding a route including a node or an edge corresponding to a word indicating a symptom, a finding, a disease state, or a disease to be excluded. Thereby, the diagnosis support apparatus 1 can extract a differential disease by narrowing down to a path from which a symptom, a finding, a disease state, or a disease that the user wants to exclude is excluded.

  Further, the output unit 144 causes the display unit 13 to display only the differential diseases classified in the designated category. Thereby, the diagnosis support apparatus 1 can extract the differential diseases by narrowing down to cases classified into the category designated by the user.

  Moreover, the reception part 141 may receive the information which concerns on patients, such as age and sex, as input information. In this case, the score calculation unit 143 limits the path to the path including the node and the edge of the disease recall chain graph for the patient similar to the information related to the patient such as the input age and sex (or the node and the edge). Set the cost to low) and calculate the score. Thereby, the diagnosis support apparatus 1 can extract a differential disease by preferentially using a patient case similar to information related to the patient specified by the user.

[Second Modification]
The diagnosis support apparatus 1 may display information related to the input word or the differential disease in addition to the information indicating the differential disease. FIG. 9 is a schematic diagram of a modified example of the search screen displayed by the display unit 13 according to the present embodiment. The search screen includes an input field A, a result field B, a related word field F, and a related information field G. The contents of the input field A are the same as the search screen of FIG.

  The result column B includes the differential disease B1, the score B2, the case information B3, and the route B4 similar to those in the search screen of FIG. However, the case information B3 displays only one case as a representative case among one or a plurality of cases constituting the route B4. In the path B4, the portion constituting the path B4 specified by the score calculation unit 143 in the disease recall chain graph is displayed as a character string for each of words indicating symptoms, findings, pathologies, or diseases input by the user. ing.

  The related word field F includes a related word F1 related to the word input in the search word field A1, and a selection field F2 for switching presence / absence of selection for each related word F1. The related word F1 is at least a word that partially matches (for example, a forward match, an intermediate match, or a backward match) with the word input in the search word column A1 and a synonym associated with the word input in the search word column A1. On the other hand. The synonym may be a word having the same or similar meaning as the word input in the search word column A1, and is a word corresponding to a higher concept or a lower concept of the word input in the search word column A1. Also good.

  In the related word storage unit 112, various words related to medical care and synonym relationships between them are stored in advance. When calculating the score, the score calculation unit 143 stores, as related words, a word partially matching the word input in the search word field A1 and a synonym of the word input in the search word field A1. Obtained from the unit 112. The score calculation unit 143 calculates a score using the related words acquired from the related word storage unit 112 in addition to the words indicating the symptoms, findings, pathological conditions, or diseases input by the user.

  And when the output part 144 displays the information corresponding to the score which the score calculation part 143 calculated on the result column B for every differential disease, the related word F1 which shows the information acquired from the related word memory | storage part 112, and a selection column F2 is displayed on the display unit 13. At this time, the output unit 144 displays the selection field F2 on the display unit 13 in a state where the related word F1 used for calculating the score is selected.

  Furthermore, when the user switches the presence / absence of selection in the selection field F2 for any related word F1, the score calculation unit 143 includes related words in addition to the words indicating the symptoms, findings, disease states, or diseases input by the user. Of the related words acquired from the storage unit 112, the score is recalculated using the related words selected by the user. The output unit 144 displays information corresponding to the score recalculated by the score calculation unit 143 in the result column B. As described above, since the diagnosis support apparatus 1 calculates a score by searching for a route using not only the word input by the user but also its related word, there is a difference in expression in the disease recall chain graph for each case. Even so, it is possible to extract the differential disease corresponding to the word input by the user.

  The related information column G includes related information G1. For example, the related information G1 is information associated with the differential disease displayed in the result column B. In this case, the related word storage unit 112 stores related information associated with each disease in advance. The related information is, for example, a test name effective for additional evaluation of a differential disease or a prescription example for a differential disease.

  When the output unit 144 displays information corresponding to the score calculated by the score calculation unit 143 in the result column B for each differential disease, the output unit 144 acquires related information from the related word storage unit 112, and displays the related information G1 as the display unit 13 To display. When the user selects any related information G1, the output unit 144 adds the selected related information G1 to the search term column A1. Thus, the diagnosis support apparatus 1 can propose additional information to the user.

[Diagnosis support method flowchart]
FIG. 10 is a diagram illustrating a flowchart of a diagnosis support method executed by the diagnosis support apparatus 1 according to the present embodiment. The reception unit 141 receives words indicating symptoms, findings, disease states, or diseases input by the user as input information (S11). The graph acquisition unit 142 acquires the disease recall chain graph stored in the graph storage unit 111 (S12).

  The score calculation unit 143 acquires the cost (weight) set in the disease recall chain graph (that is, the node and the edge) acquired by the graph acquisition unit 142 in step S12 (S13). In addition, the score calculation unit 143 selects one or a plurality of nodes corresponding to a word indicating a symptom, a finding, a disease state, or a disease input by the user in the disease recall chain graph (S14).

  The score calculation unit 143 uses the cost acquired in step S13 to identify the route from the node selected in step S14 to the upstream differential disease node in common with those nodes, and the identified route A score is calculated (S15).

  Then, the output unit 144 outputs information corresponding to the score calculated by the score calculation unit 143 in step S15 by a predetermined method (S16).

[Effect of this embodiment]
The diagnosis support apparatus 1 according to the present embodiment includes a symptom for each case, a finding, a disease state, and a disease recall chain graph indicating a chain direction between the diseases, from the symptom input by the user to a differential disease upstream thereof. A score for each differential disease is calculated based on the route information. Therefore, the diagnosis support apparatus 1 can more accurately extract a candidate for a differential disease by using information between a symptom and the like that could not be considered in the past and the disease.

  Moreover, the diagnosis support apparatus 1 displays the part which comprises the specified path | route in the disease recall chain graph. This allows the user to know what causal logic (reason for derivation) or recall chain, that is, what symptom, finding, pathology or disease chain the candidate for the differential disease was extracted from. Can do.

  In addition, the diagnosis support apparatus 1 specifies a path from a symptom or the like input by the user to a differential disease upstream thereof by combining disease recall chain graphs of a plurality of cases. Therefore, even if not all of the symptoms etc. input by the user are included in one case, a path can be constructed by combining a plurality of cases, and differential disease candidates corresponding to the symptoms etc. input by the user can be extracted. .

  As mentioned above, although this invention was demonstrated using embodiment, the technical scope of this invention is not limited to the range as described in the said embodiment, A various deformation | transformation and change are possible within the range of the summary. is there. For example, the specific embodiments of device distribution / integration are not limited to the above-described embodiments, and all or a part of them may be configured to be functionally or physically distributed / integrated in arbitrary units. Can do. In addition, new embodiments generated by any combination of a plurality of embodiments are also included in the embodiments of the present invention. The effect of the new embodiment produced by the combination has the effect of the original embodiment.

  The processor of the diagnosis support apparatus 1 becomes a main body of each step (process) included in the diagnosis support method shown in FIG. That is, the processor of the diagnosis support apparatus 1 reads out a program for executing the diagnosis support method shown in FIG. 10 from the storage unit, executes the program, and controls each part of the diagnosis support apparatus 1, so that FIG. Execute the indicated diagnostic support method. Part of the steps included in the diagnosis support method shown in FIG. 10 may be omitted, the order between steps may be changed, and a plurality of steps may be performed in parallel.

DESCRIPTION OF SYMBOLS 1 Diagnosis support apparatus 11 Memory | storage part 12 Input part 13 Display part 14 Control part 141 Reception part 142 Graph acquisition part 143 Score calculation part 144 Output part

Claims (17)

A reception unit that accepts input of at least one word indicating a symptom, a finding, a disease state, or a disease;
An acquisition unit for acquiring a graph indicating a relationship between symptoms, findings, pathological conditions, and diseases for each case, stored in the storage unit;
Identifying a path from a symptom, a finding, a disease state or a disease corresponding to the word in one or more of the graphs to a differential disease which is at least one disease included in the graph, and based on the identified path, A calculation unit for calculating a score associated with the differential disease;
An output unit that outputs information corresponding to the score calculated by the calculation unit for each of the differential diseases;
A diagnosis support apparatus having The storage unit stores the graph showing the symptom, findings, pathophysiology, and direction of linkage between diseases for each case,
The diagnosis support apparatus according to claim 1, wherein the calculation unit identifies the path from a symptom, a finding, a disease state, or a disease corresponding to the word in the graph to the differential disease that is upstream of the graph.   The diagnosis support apparatus according to claim 1, wherein the calculation unit identifies at least one route as the route in ascending order of distance in the graph.   The diagnosis support apparatus according to any one of claims 1 to 3, wherein the output unit outputs, in addition to information corresponding to the score, a portion constituting the path in the graph.   The diagnosis support apparatus according to any one of claims 1 to 4, wherein the output unit outputs information indicating the case corresponding to the route in addition to the information corresponding to the score.   The diagnosis according to any one of claims 1 to 5, wherein the calculation unit calculates the score associated with the differential disease based on a weight set for each of the associations on the identified route. Support device.   The calculation unit includes the weight of the relationship between the symptom or finding and the symptom or finding, the weight of the relationship between the symptom or finding and the disease state or disease, and the disease state or disease and the disease state or disease. The diagnosis support apparatus according to claim 6, wherein the weights related to each other are set to be different from each other, and the score is calculated.   The calculation unit according to any one of claims 1 to 7, wherein the calculation unit specifies the path constituted by a plurality of the cases from a symptom, a finding, a disease state, or a disease corresponding to the word to the differential disease in the graph. The diagnosis support apparatus according to one item.   The diagnosis support apparatus according to claim 8, wherein the calculation unit calculates the score based on the number of cases constituting the path. The output unit outputs related information associated with the differential disease,
The diagnosis support apparatus according to any one of claims 1 to 9, wherein the reception unit receives the selected related information in addition to the word when any of the related information is selected.   The said calculation part specifies the said path | route from the symptom, finding, pathology, or disease corresponding to the related word linked | related with the said word to the said differential disease in addition to the said word which the said reception part received. To 10. The diagnosis support apparatus according to any one of 1 to 10. In addition to the input of the word, the reception unit receives an input of a word indicating a symptom, a finding, a disease state, or a disease to be excluded in specifying the route,
The calculation unit identifies the pathway from which symptoms, findings, pathologies or diseases corresponding to the exclusion target are excluded from symptoms, findings, pathologies or diseases corresponding to the words in the graph to the differential diseases. The diagnosis support apparatus according to any one of claims 1 to 11. The reception unit receives an input of a disease classification in addition to the input of the word,
The diagnosis support apparatus according to any one of claims 1 to 12, wherein the output unit outputs information corresponding to the score of the differential disease included in the classification.   The graph stored in the storage unit indicates the association between symptoms, findings, pathological conditions, and diseases by at least one of a directed graph and an undirected graph. Diagnosis support device. Processor
Receiving an input of at least one word indicating a symptom, finding, condition or disease;
Acquiring a graph showing a relationship between symptoms, findings, pathological conditions, and diseases for each case stored in the storage unit;
Identifying a path from a symptom, a finding, a disease state or a disease corresponding to the word in one or more of the graphs to a differential disease which is at least one disease included in the graph, and based on the identified path, Calculating a score associated with the differential disease;
Outputting the information corresponding to the score calculated by the calculating step for each of the differential diseases;
Diagnosis support method to execute. On the computer,
Receiving an input of at least one word indicating a symptom, finding, condition or disease;
Acquiring a graph showing a relationship between symptoms, findings, pathological conditions, and diseases for each case stored in the storage unit;
Identifying a path from a symptom, a finding, a disease state or a disease corresponding to the word in one or more of the graphs to a differential disease which is at least one disease included in the graph, and based on the identified path, Calculating a score associated with the differential disease;
Outputting the information corresponding to the score calculated by the calculating step for each of the differential diseases;
Diagnosis support program to execute A reception unit that accepts input of at least one word indicating a symptom, a finding, a disease state, or a disease;
An acquisition unit for acquiring a graph indicating a relationship between symptoms, findings, pathological conditions, and diseases for each case, stored in the storage unit;
Identifying a path from a symptom, a finding, a disease state or a disease corresponding to the word to a differential disease which is at least one disease included in the graph in one or more of the graphs, and for each differential disease, An output unit for displaying a portion constituting the route in the display unit,
A diagnosis support apparatus having

Images