Information processing device, information processing method, and control program

Abstract

The problem addressed by the present invention is to recognize a change of the condition of a subject. An information processing device (1) comprises: a data acquisition unit (21) that acquires medical examination data about the subject; a classification unit (22) that inputs the acquired medical examination data about the subject to a classification model that can classify the subject into any one of a plurality of expression types on the basis of the medical examination data, and classifies the subject into any one of the plurality of expression types; and a subject transition information output unit (23) that outputs subject transition information indicating a transition of the subject from a first classification result to a second classification result.

Description

Information processing device, information processing method, and control program

[0001] The present invention relates to an information processing device that outputs a prediction result regarding the prognosis of a subject.

[0002] For example, Patent Document 1 describes a prediction method for predicting the occurrence of fatal symptoms such as sepsis and cardiac arrest from vital signs.

[0003] Japanese Patent Application Publication No. 2021-177429

[0004] In order for doctors and other medical professionals to determine early on whether a patient may develop life-threatening symptoms, it is important to understand changes in the patient's condition, including potential risks.

[0005] An object of one aspect of the present invention is to provide an information processing device and an information processing method that are capable of understanding changes in the condition of a subject.

[0006] In order to solve the above-mentioned problems, an information processing device according to one aspect of the present invention includes a data acquisition unit that acquires medical data related to a subject; a classification unit that inputs the acquired medical data related to the subject into a classification model that can classify the subject into one of a plurality of phenotypes based on the medical data, and classifies the subject into one of the plurality of phenotypes; and a subject transition information output unit that outputs subject transition information that indicates a transition from a first classification result obtained based on the medical data related to the subject at a first time point to a second classification result obtained based on the medical data related to the subject at a second time point after the first time point, wherein the medical data includes at least attribute information, vital sign information, and symptom information, and the symptom information includes at least the location of a disease and the severity of the disease, or the location of an injury and the severity of symptoms in the location.

[0007] In order to solve the above-mentioned problems, an information processing method according to one aspect of the present invention includes: a data acquisition step in which a computer acquires medical data related to a subject; a classification step in which the computer inputs the acquired medical data related to the subject into a classification model capable of classifying the subject into one of a plurality of phenotypes based on the medical data, and classifies the subject into one of the plurality of phenotypes; and a subject transition information output step in which the computer outputs subject transition information indicating a transition from a first classification result obtained based on the medical data related to the subject at a first time point to a second classification result obtained based on the medical data related to the subject at a second time point after the first time point, wherein the medical data includes at least attribute information, vital sign information, and symptom information, and the symptom information includes at least the site of a disease and the extent of the disease, or the site of an injury and the extent of symptoms at the site.

[0008] The information processing device according to each aspect of the present invention may be realized by a computer. In this case, the control program for the information processing device that causes the computer to operate as each unit (software element) of the information processing device to realize the information processing device on the computer, and a computer-readable recording medium on which the control program is recorded, also fall within the scope of the present invention.

[0009] According to one aspect of the present invention, it is possible to grasp changes in the condition of a subject.

[0010] 1 is a block diagram showing an example of the configuration of an information processing system including an information processing device according to a first embodiment of the present invention. FIG. 1 is a graph in which the first principal component score and the second principal component score of a subject are plotted on a coordinate system in which the first axis represents the first principal component score of a classification model and the second axis represents the second principal component score. FIG. 2 is a flowchart showing an example of the flow of processing performed by the information processing device. FIG. 3 is a diagram showing an example of a display screen according to the first embodiment of the present invention. FIG. 4 is a block diagram showing an example of the configuration of an information processing system including an information processing device according to a second embodiment of the present invention. FIG. 5 is a diagram showing an overview of the classification process. FIG. 6 is a table showing the classification results of an analysis cohort. FIG. 7 is a table showing the classification results of a validation cohort. FIG. 8 is a heat map showing the survival rate and the distribution of variables for each clinical phenotype. FIG. 9 is a diagram showing the results of a hierarchical cluster analysis of the principal component scores of each phenotype. FIG. 10 is a plot showing the results of a principal component analysis of the centroids of each phenotype. FIG. 11 is a graph showing serum proteins that are differentially expressed between high-mortality phenotypes and other phenotypes. FIG. 12 is a diagram showing the results of a gene ontology (GO) enrichment analysis of proteins. FIG. 13 is a diagram plotting each variable of serum proteomics analysis data against the standardized value of each variable. 1 is a diagram showing a transition of the classification results of a subject classified using the information processing device according to embodiment 1 of the present invention, and FIG. 2 is a diagram showing a transition of the classification results of another subject classified using the information processing device according to embodiment 1 of the present invention.

[0011] First Embodiment An information processing system according to one embodiment of the present invention will be described in detail below.

[0012] (Configuration of Information Processing System 100) The configuration of the information processing system 100 will be described with reference to Fig. 1. Fig. 1 is a block diagram showing an example of the configuration of the information processing system 100 including an information processing device 1 according to one aspect of the present invention.

[0013] 1, the information processing system 100 includes an information processing device 1, an external device 5 that transmits medical data to the information processing device 1, and a presentation device 4 that acquires information output from the information processing device 1 and presents the information. Note that FIG. 1 shows an example in which the information processing system 100 is introduced into a medical institution H1.

[0014] The external device 5 may be, for example, a server device that stores and manages medical data, or a terminal device into which a system user such as a medical professional inputs medical data. As will be described later, an electronic medical record system may be used as the external device 5.

[0015] 1 illustrates an example in which the information processing device 1 acquires medical data from an external device 5 that is separate from the information processing device 1, but the present invention is not limited to this. For example, the information processing device 1 may be configured to be built into the external device 5.

[0016] The presentation device 4 may be a display capable of presenting information output from the information processing device 1. In one example, the presentation device 4 may be a display included in the information processing device 1 or the external device 5. Alternatively, the presentation device 4 may be a computer, tablet terminal, or the like used by a medical professional belonging to the medical institution H1.

[0017] The information processing device 1 and the external device 5, and the information processing device 1 and the presentation device 4 may be connected by wireless communication or by wired communication.

[0018] (Configuration of Information Processing Device 1) As shown in Fig. 1, the information processing device 1 includes a control unit 2 and a storage unit 3. The storage unit 3 may store a classification model 31. The classification model 31 will be described later.

[0019] The storage unit 3 may store, in addition to the classification model 31, control programs for each unit executed by the control unit 2, an OS program, application programs, etc. The storage unit 3 may also store various data that is read when the control unit 2 executes these programs. The storage unit 3 is configured by a non-volatile storage device such as a hard disk or flash memory. In addition to the storage unit 3, the storage unit may also include a volatile storage device such as a RAM (Random Access Memory) that is used as a working area for temporarily storing data in the process of executing the various programs described above.

[0020] <Configuration of control unit 2> The control unit 2 may be configured by a control device such as a CPU (Central Processing Unit) or a dedicated processor. Each unit of the control unit 2 shown in Fig. 1 can be realized by a control device such as a CPU reading out a program stored in a storage unit 3 realized by a ROM (Read Only Memory) or the like into a RAM or the like and executing the program. The control unit 2 includes a data acquisition unit 21, a classification unit 22, and a target transition information output unit 23.

[0021] The data acquisition unit 21 acquires medical data related to the subject from the external device 5. The data acquisition unit 21 may store the acquired medical data in the storage unit 3.

[0022] The medical data includes the following items related to the subject: ・Attribute information ・Vital sign information ・Symptom information

[0023] The attribute information may include, for example, the age and gender of the subject, and the Charlson Polypharmacy Scale (CPS).

[0024] The vital sign information may include, for example, at least one of respiratory rate, pulse rate, systolic blood pressure, level of consciousness (Glasgow Coma Scale (GCS) or Japan Coma Scale (JCS)), and body temperature.

[0025] If the subject is suffering from a disease, the symptom information includes the location of the disease and the severity of the disease. Diseases may be distinguished from trauma, which will be described later. The specified disease may include at least one of myocardial infarction, cerebral infarction, heat stroke, cerebral hemorrhage, cardiac arrest, poisoning, disseminated intravascular coagulation (DIC), acute respiratory distress syndrome (ARDS), burns, hemorrhagic shock (including, for example, gastrointestinal bleeding, postpartum hemorrhage, etc.), and infectious diseases. Examples of infectious diseases include sepsis and COVID-19. The diseased organ may specifically be the myocardium, lungs, cerebrospinal cord, peripheral nerves, bones, liver, kidneys, spleen, gastrointestinal tract, bladder, uterus, ovaries, testes, skin, etc., and the severity of the disease may be the degree of progression of the disease. Furthermore, the severity of the disease may be indicated, for example, by one of a plurality of preset graded scales.

[0026] If the subject has an injury, the symptom information includes the location of the injury and the severity of the symptoms at the location. The location of the injury may include, for example, the head and neck, face, chest, abdomen, limbs, and body surface. The severity of the symptoms at the location of the injury may be a severity. The severity of the symptoms may be indicated, for example, by one of a plurality of preset graded scales.

[0027] The symptom information may include at least one of information indicating the type of injury to the subject's body, the degree of injury, and a biological response caused by the injury. For example, the symptom information may be represented by an Abbreviated Injury Scale (AIS) code and / or an Injury Severity Score (ISS).

[0028] The type of injury is not particularly limited, and may be, for example, a type classified by a known score for assessing the severity of the injury. An example of a known score for assessing the severity of the injury is the AIS coding (https: / / www.aaam.org / abbreviated-injury-scale-ais / ). Specific examples of the AIS coding include AIS 2015, AIS 2005 Update 2008, and AIS90 Update 98. The latest AIS coding updates may be included at https: / / www.aaam.org / abbreviated-injury-scale-ais / . More specifically, the type of injury may be an abrasion, a contusion, a crush injury, a contusion, an incision, a split wound, a laceration, a puncture wound, an avulsion, a burn, a fracture, or the like.

[0029] If the subject has a disease and an injury, the symptom information may include the location of the disease and the severity of the disease, and the location of the injury and the severity of the symptoms at that location. The symptom information may include at least the location of the disease and the severity of the disease, or the location of the injury and the severity of the symptoms at that location.

[0030] The data acquisition unit 21 may have a preprocessing function for performing preprocessing on the acquired medical data. For example, the data acquisition unit 21 may select and acquire only necessary items from among items included in the acquired medical data.

[0031] The data acquisition unit 21 may acquire at least a portion of the medical data related to the subject from the subject's electronic medical record. For example, the data acquisition unit 21 may acquire vital sign information from the electronic medical record stored in the external device 5. This allows at least a portion of the medical data to be acquired from the diagnosis results entered into the electronic medical record by the medical professional treating the subject, thereby simplifying the task of acquiring medical data. Furthermore, since vital sign information is measured periodically and entered into the electronic medical record, acquiring the vital sign information from the electronic medical record enables regular updates of the medical data.

[0032] When AIS codes are included in medical data, entering them requires advanced knowledge and a certain level of proficiency. Moreover, because AIS codes are highly categorized and include approximately 2,000 different codes, even medical professionals often have to refer to text while entering them. This results in a time-consuming process. Therefore, the data acquisition unit 21 may acquire AIS codes using a language generation model (e.g., ChatGPT). Specifically, the data acquisition unit 21 may acquire AIS codes by using a language generation model that has previously learned AIS codes to ask medical professionals questions to identify the AIS code and then identifying the AIS code based on the medical professional's answers to the questions. This eliminates the need for medical professionals to refer to text, thereby reducing the time it takes for the data acquisition unit 21 to acquire AIS codes. Alternatively, the data acquisition unit 21 may acquire AIS codes by inputting images of the target area of ​​the subject into a model generated by machine learning to output AIS codes based on input images (e.g., X-ray images, CT images, etc.).

[0033] The classification unit 22 inputs the medical data acquired by the data acquisition unit 21 into the classification model 31 read from the storage unit 3, and classifies the subject into one of multiple phenotypes. In this specification, the term "phenotype" refers to a group when a patient's disease and / or injury is classified into multiple groups based on clinical characteristics. The phenotype may also be referred to as a phenotype.

[0034] The classification model 31 will now be described. The classification model 31 is a clustering model generated using medical data on each of a plurality of sample individuals. The medical data may be medical data having the same items as the medical data on the subject acquired by the data acquisition unit 21. The classification model 31 may be generated using medical data on each of a plurality of sample individuals who have suffered from a disease and / or an injury. The medical data may be medical data created from the time when the sample individual is transported by ambulance to a medical institution for the disease and / or the injury of the sample individual, until the time when the sample individual has finished receiving medical treatment for the disease and / or the injury at the medical institution. Since it is expected that the sample individual will be transported by ambulance to the medical institution again and become a subject, the sample individuals may include the subject.

[0035] The classification model 31 may be generated by machine learning using medical data on each of a plurality of sample individuals. For example, k-means, latent class analysis (LCA), silhouette analysis, consensus clustering, random forest, neural network, principal component analysis, hierarchical clustering, etc. may be used as the machine learning method.

[0036] The classification unit 22 inputs medical data about the subject at each of a plurality of time points, including a first time point and a second time point after the first time point, into the classification model 31 and outputs a classification result for each of the plurality of time points. For example, the time when the subject is transported to the hospital by ambulance and medical data is first acquired and input into the classification model 31 can be defined as the first time point, and a predetermined time (12 hours, 24 hours, etc.) can be defined as the second time point. Furthermore, the second time point can be defined as a new first time point, and a predetermined time can be defined as a new second time point. In the following description, the classification result of the subject at the first time point may be referred to as the first classification result, and the classification result of the subject at the second time point may be referred to as the second classification result. The classification unit 22 outputs classification results at a plurality of time points, including the first classification result and the second classification result, to the subject transition information output unit 23. The classification unit 22 may also store the classified classification results in the storage unit 3.

[0037] The subject transition information output unit 23 outputs subject transition information indicating a transition of the classification result of the subject output from the classification unit 22. The subject transition information output unit 23 includes information indicating at least a transition from the first classification result to the second classification result.

[0038] The subject transition information output unit 23 may output subject transition information further including the subject's classification results at time points other than the first and second time points. The subject transition information output unit 23 may output subject transition information including the subject's classification results at predetermined time intervals (e.g., every 12 hours, every 24 hours, etc.). The subject transition information output unit 23 may output text showing, in chronological order, information indicating the phenotype into which the subject was classified at each time point as the transition of the subject's classification results. The subject transition information output unit 23 outputs the subject transition information, allowing medical professionals such as doctors to understand changes in the subject's condition. As a result, medical professionals can identify cases with a high risk of death early and recognize the need to provide necessary treatment to the subject early.

[0039] The subject transition information output unit 23 may display the subject transition information on a graph plotting the first classification result and the second classification result on a coordinate system with the first axis representing the first principal component score calculated by principal component analysis of the multiple phenotypes classified by the classification model 31 and the second axis representing the second principal component score. Principal component analysis is a statistical analysis method that aggregates data having multiple variables to create principal components. In this case, specifically, the subject transition information output unit 23 first calculates the center of gravity of each item of the medical data of the sample individuals included in each phenotype for each of the multiple phenotypes. The subject transition information output unit 23 calculates the first and second principal components of the classification model 31 by performing principal component analysis on the calculated centers of gravity of each item. The subject transition information output unit 23 calculates the first and second principal component scores of the subject at each of the multiple time points from the medical data of the subject at the multiple time points. The subject transition information output unit 23 creates a graph in which the calculated first principal component score and second principal component score at each time point are plotted on a coordinate system with the first principal component score as the first axis and the second principal component score as the second axis. The subject transition information output unit 23 may cause the created graph to be presented on the presentation device 4. This allows a medical professional to visually recognize the transition of the phenotype into which the subject has been classified.

[0040] FIG. 2 is a diagram showing an example of the graph. As shown in FIG. 2, the subject transition information output unit 23 may cause the presentation device 4 to display a graph in which positions indicating the classification results of the subject at each time point are plotted on a coordinate system. The subject transition information output unit 23 may also cause the graph to display arrows indicating the transition of the positions indicating the classification results of the subject at each time point, plotted on the graph. This makes it easier for medical professionals to understand the transition of the phenotype into which the subject has been classified. The subject transition information output unit 23 may also display the date and time when the subject was classified near the plot on the graph. This makes it easier for medical professionals to understand the transition of the phenotype into which the subject has been classified.

[0041] The graph may display a curve indicating the degree of approximation from the center of gravity of each phenotype in the graph, as shown in Figure 2. The graph may also display an area in the graph where the proportion of patients who will die within a predetermined period after the acquisition of medical data is high (the area displayed as area C in the example shown in Figure 2). This makes it easier for medical professionals to recognize that the subject is at potential risk of developing a fatal symptom when the position indicating the subject's classification result is moving closer to that area.

[0042] (Processing Executed by Information Processing Device 1) Next, the processing (information processing method) executed by the information processing device 1 will be described with reference to Fig. 3. Fig. 3 is a flowchart showing an example of the flow of processing executed by the information processing device 1.

[0043] In the process executed by the information processing device 1, first, the data acquisition unit 21 acquires medical data related to the subject from the external device 5 (step S1: data acquisition step). The data acquisition unit 21 may select and discard multiple items included in the medical data as necessary. In step S1, the data acquisition unit 21 acquires medical data related to the subject at each of multiple time points including a first time point and a second time point that is later than the first time point.

[0044] Next, the classification unit 22 inputs the medical data acquired by the data acquisition unit 21 into the classification model 31 read from the storage unit 3, and classifies the subject into one of multiple phenotypes (step S2: classification step). In step S2, each piece of medical data related to the subject at each of multiple time points including a first time point and a second time point after the first time point is input into the classification model 31, and the classification result at each time point is output.

[0045] As described above, the information processing device 1 according to one aspect of the present invention includes a classification unit that inputs medical data about a subject into the classification model 31 and classifies the subject into one of a plurality of phenotypes, and a subject transition information output unit that outputs subject transition information indicating a transition from a first classification result obtained based on the medical data about the subject at a first time point to a second classification result obtained based on the medical data about the subject at a second time point after the first time point. This allows a user to understand changes in the subject's condition. As a result, if a potential risk of a fatal symptom occurring in the subject is detected, a medical professional can recognize the need to provide the subject with necessary treatment early.

[0046] (Application example of information processing system 100) Next, an application example of the information processing system 100 will be described. In this application example, a case will be described in which the information processing device 1 is configured to be built into the external device 5. In this case, a user inputs medical data related to a subject to the external device 5. FIG. 4 is a diagram showing an example of a display screen 50 displayed on a display device provided in the external device 5. As shown in FIG. 4, the display screen 50 may include a first area A1, a second area A2, and a third area A3.

[0047] The first area A1 is an area for supporting the user in inputting medical data related to the subject. The user can input medical data by selecting or operating selection buttons, etc. displayed in the first area A1, using a mouse, etc., provided on the external device 5. The second area A2 is an area for displaying a graph in which the centroids of each phenotype are plotted on a coordinate system with the first axis representing the first principal component score and the second axis representing the second principal component score. Each centroid is displayed with a size corresponding to the number of subjects classified into each phenotype. The third area A3 is an area for displaying a graph corresponding to the graph shown in FIG. 2.

[0048] In this application example, the data acquisition unit 21 acquires medical data input by the user by operating selection buttons or the like displayed in the first area A1. When an instruction to execute classification processing is received from the user after the medical data has been input, the classification unit 22 performs classification using the medical data. Then, the subject transition information output unit 23 displays, in the third area A3, subject transition information indicating the transition of the classification results of the subject output from the classification unit 22. With the above configuration, the user can check the input medical data and the subject transition information on one screen.

[0049] In another application example of the information processing system 100, a graph equivalent to the graph shown in Fig. 2 may be displayed on the electronic medical record screen, which makes it easier for medical personnel to understand the progress of the subject's condition.

[0050] [Embodiment 2] Another embodiment of the present invention will be described below. For ease of explanation, the same reference numerals will be used to designate components having the same functions as those described in the above embodiment, and the description thereof will not be repeated.

[0051] FIG. 5 is a block diagram showing an example of the configuration of an information processing system 100A including an information processing device 1A according to this embodiment.

[0052] 5, the information processing device 1A includes a control unit 2A instead of the control unit 2 in the first embodiment. The control unit 2A includes an estimation unit 24 in addition to the configuration of the first embodiment.

[0053] 5 , in the information processing device 1A, transition information of past classification results of a specimen (hereinafter referred to as specimen transition information 32) is stored in the storage unit 3. Specifically, the specimen transition information 32 includes specimen transition information indicating a transition from a third classification result in which the specimen is classified into one of the plurality of phenotypes based on medical data about the specimen at a certain point in time (hereinafter referred to as the third point in time), to a fourth classification result in which the specimen is classified into one of the plurality of phenotypes based on medical data about the specimen at a fourth point in time that is after the third point in time and before the second point in time.

[0054] The estimation unit 24 compares the sample transition information 32 read from the storage unit 3 with the transition information of the subject's classification results (i.e., the subject transition information) output from the subject transition information output unit 23, and outputs an estimation result regarding the subject's prognosis. Specifically, the estimation unit 24 compares the subject transition information with the sample transition information 32, and extracts transition information of the sample's classification results including a transition of the classification results similar to the transition of the subject's classification results. The estimation unit 24 extracts from a database, based on the extracted transition information of the sample's classification results, the sample's subsequent disease and / or disease transition after the similar time point, and estimates the subject's prognosis based on the extracted disease transition. The database may be stored in the storage unit 3 or the external device 5.

[0055] The sample transition information 32 may be stored in the storage unit 3 with a flag set for transition information of classification results for a sample subject whose symptoms have rapidly worsened. In this case, the estimation unit 24 may compare the subject transition information with the sample transition information 32, and if there is transition information of classification results for multiple sample subjects that is similar to the subject transition information, estimate the subject's prognosis based on the flagged transition information of classification results. This reduces the likelihood that a medical professional will assume that the subject's symptoms are unlikely to rapidly worsen when there is a possibility that the subject's symptoms will rapidly worsen. Furthermore, medical professionals can share cases of sample subjects whose symptoms have rapidly worsened, enabling early therapeutic intervention for the subject.

[0056] The estimation result estimated by the estimation unit 24 may be displayed by the presentation device 4. In this case, if there is a possibility that the subject's symptoms will rapidly worsen, warning information may also be displayed.

[0057] [Example of implementation by software] The functions of the information processing device 1, 1A can be realized by a program for causing a computer to function as the information processing device, and a program for causing a computer to function as each control block of the information processing device (particularly each part included in the control unit 2, 2A).

[0058] In this case, the information processing device includes a computer having at least one control device (e.g., a processor) and at least one storage device (e.g., a memory) as hardware for executing the program, and the functions described in each of the above embodiments are realized by executing the program using the control device and the storage device.

[0059] The program may be stored non-transitory on one or more computer-readable storage media. The storage media may or may not be included in the information processing device. In the latter case, the program may be supplied to the information processing device via any wired or wireless transmission medium.

[0060] Furthermore, some or all of the functions of the control blocks can be realized by logic circuits. For example, an integrated circuit in which a logic circuit that functions as each of the control blocks is formed is also included in the scope of the present invention. In addition, the functions of the control blocks can also be realized by, for example, a quantum computer.

[0061] (Summary) An information processing device according to aspect 1 of the present disclosure includes a data acquisition unit that acquires medical data related to a subject, a classification unit that inputs the acquired medical data related to the subject into a classification model that can classify the subject into one of a plurality of phenotypes based on the medical data, and classifies the subject into one of the plurality of phenotypes, and a subject transition information output unit that outputs subject transition information that indicates a transition from a first classification result obtained based on the medical data related to the subject at a first time point to a second classification result obtained based on the medical data related to the subject at a second time point after the first time point, wherein the medical data includes at least attribute information, vital sign information, and symptom information, and the symptom information includes at least the site of a disease and the severity of the disease, or the site of an injury and the severity of symptoms in the site.

[0062] The information processing device according to aspect 2 of the present disclosure may be configured in the above-mentioned aspect 1 to further include an estimation unit that outputs an estimation result regarding the prognosis of the subject based on a comparison result between specimen transition information indicating a transition from a third classification result in which the specimen is classified into one of the plurality of phenotypes based on the medical data regarding the specimen at a third time point, to a fourth classification result in which the specimen is classified into one of the plurality of phenotypes based on the medical data regarding the specimen at a fourth time point that is after the third time point and before the second time point, and the subject transition information.

[0063] An information processing device according to aspect 3 of the present disclosure may be configured in the above-mentioned aspect 1 or 2 such that the subject transition information output unit causes the presentation device to present the subject transition information as a graph in which the first classification result and the second classification result are plotted on a coordinate system in which a first axis represents a first principal component score calculated by principal component analysis of the plurality of phenotypes and a second axis represents a second principal component score.

[0064] In an information processing device according to aspect 4 of the present disclosure, in any of aspects 1 to 3 above, the classification model may be a clustering model generated based on medical data relating to each of a plurality of sample subjects who have suffered from a disease and / or an injury, which medical data was created from the time the sample subject was rushed to a medical institution for the disease and / or the injury, until the time the sample subject finished receiving medical treatment for the disease and / or the injury at the medical institution.

[0065] In an information processing device according to aspect 5 of the present disclosure, in any one of aspects 1 to 4 above, the data acquisition unit may acquire at least a portion of the medical data regarding the subject from the subject's electronic medical record.

[0066] In an information processing device according to a sixth aspect of the present disclosure, in any one of the first to fifth aspects, the symptom information may include an AIS code.

[0067] In an information processing device according to a seventh aspect of the present disclosure, in any one of the first to sixth aspects, the attribute information may include age and gender.

[0068] In an information processing device according to aspect 8 of the present disclosure, in any of aspects 1 to 7 above, the vital sign information may include at least one of respiratory rate, pulse rate, systolic blood pressure, level of consciousness (Glasgow Coma Scale (GCS) or Japan Coma Scale (JCS)), and body temperature.

[0069] An information processing method according to a ninth aspect of the present disclosure includes: a data acquisition step in which a computer acquires medical data related to a subject; a classification step in which the computer inputs the acquired medical data related to the subject into a classification model capable of classifying the subject into one of a plurality of phenotypes based on the medical data, and classifies the subject into one of the plurality of phenotypes; and a subject transition information output step in which the computer outputs subject transition information indicating a transition from a first classification result obtained based on the medical data related to the subject at a first time point to a second classification result obtained based on the medical data related to the subject at a second time point after the first time point, wherein the medical data includes at least attribute information, vital sign information, and symptom information, and the symptom information includes at least the site of a disease and the extent of the disease, or the site of an injury and the extent of symptoms at the site.

[0070] A control program according to aspect 10 of the present disclosure is a control program for causing a computer to function as an information processing device described in any one of aspects 1 to 8 above, and is a control program for causing a computer to function as the data acquisition unit, the classification unit, and the target transition information output unit.

[0071] The present invention is not limited to the above-described embodiments, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present invention.

[0072] An embodiment of the present invention will be described below. Fig. 6 is a diagram showing an outline of the classification process carried out in this embodiment.

[0073] We obtained medical data for 158,918 patients registered in the Japan Trauma Data Bank (2013-2017). All data were collected from the time a patient was transported to a medical institution by ambulance until the time they completed treatment for their illness or injury at that institution. Of the acquired data, data for 12,565 patients were excluded from classification due to non-blunt trauma. Of the remaining 146,353 blunt trauma patients, 75,315 patients were excluded because they did not meet the inclusion criteria. Specifically, patients who did not meet the inclusion criteria were not transported (n = 36,267), had an ISS of 75 (n = 13,667), were in cardiopulmonary arrest at the time of admission (n = 6,555), were pregnant (n = 69), and had abnormal or missing data (n = 18,757).

[0074] <Classification of Analysis Cohort (1)> Data of 71,038 patients were used as the analysis cohort for classification. The analysis cohort used for this classification is hereinafter also referred to as analysis cohort D.

[0075] The variables used for classification were age, sex, CPS, respiratory rate (RR), pulse rate (HR), systolic blood pressure (SBR), Grass-Coma Scale (GCS), head and neck AIS, face AIS, chest AIS, abdominal AIS, limb AIS, body surface AIS, injury severity score (ISS), RTS (Revised Trauma Score), TRISS (Trauma and Injury Severity Score), and number of survivors.

[0076] RTS is a severity assessment based on physiological indicators. Specifically, RTS is calculated from the level of consciousness (GCS), systolic blood pressure, and respiratory rate. The most severe score is 0 points, and the best score is 7.84 points.

[0077] TRISS is a predicted survival rate calculated by adding physiological severity, anatomical severity, and age factors.

[0078] Each data set was averaged and then clustered using the k-means method. The optimal number of clusters was evaluated using silhouette analysis, and data with negative silhouette coefficients were excluded from classification.

[0079] As a result of the classification, the analysis cohort was classified into eight phenotypes (D-1 to D-8). The classification results are shown in FIG. 7. In FIG. 7, [IQR] indicates the interquartile range. When comparing each phenotype with the survival rate of patients of each phenotype in FIG. 7, it was found that the survival rate of one phenotype (D-8) out of the eight phenotypes was extremely low. In other words, it was found that the mortality rate of one of the classified phenotypes was high.

[0080] Next, using only the data for the D-8 phenotype, which had a high mortality rate, further classification was performed using latent class analysis (LCA) (VarSelLCM package). As a result of classification, the D-8 data were divided into four clusters (D-8α, D-8β, D-8γ, and D-8δ). Evaluation of the four phenotypes classified by LCA revealed that D-8α represented multiple trauma in young people (n = 464), D-8β represented head trauma with hypothermia (n = 178), D-8γ represented severe head injury in elderly people (n = 957), and D-8δ represented multiple trauma with a predicted mortality rate higher than the actual mortality rate (n = 579).

[0081] <Classification of the validation cohort> Data from 42,780 patients (January 2013 to June 2015) registered in the Japan Trauma Data Bank were classified using the same method as for analysis cohort D. Data showing negative silhouette coefficients were excluded from the cluster, and 38,097 patients were finally classified. The classification results are shown in Figure 8. In Figure 8, the validation cohort was also classified into eight phenotypes, V-1 to V-8, as in the analysis cohort. It was also found that patients classified into one of the phenotypes (V-8) out of V-1 to V-8 had a high mortality rate.

[0082] Next, we further classified the data for the V-8 phenotype, which had a high mortality rate, using the LCA method. As a result, the V-8 data were classified into four clusters (V-8α, V-8β, V-8γ, and V-8δ), similar to the D-8 data.

[0083] <Comparison between analysis cohort and validation cohort> Figure 9 is a heat map showing the survival rate for each classified phenotype in the analysis cohort and validation cohort, and the distribution of variables for each phenotype. The survival rate for each clinical phenotype is shown in a bar graph at the top, and the heat map in the lower panel lists each variable, with each cell listing the median value of each variable for each phenotype.

[0084] Figure 10 shows the results of hierarchical clustering analysis of the principal component scores of the centroids of each phenotype. Principal component analysis was performed on each phenotype in the analysis cohort and validation cohort. Subsequently, the principal component scores of the centroids of each phenotype were calculated, and hierarchical cluster analysis was performed. Figure 10 confirms that clusters of high mortality phenotypes are paired together.

[0085] Figure 11 is a plot of the coordinates of the centers of gravity calculated in Figure 10. The horizontal axis is the first principal component axis, and the vertical axis is the second principal component axis. The size of each plot indicates the number of patients. Figure 11 visualizes pairs with similar phenotypes in the analysis cohort and the validation cohort.

[0086] <Classification of Analysis Cohorts (2) Evaluation of Molecular Pathology> After biological invasion in a given disease or trauma, damage-associated molecular patterns (DMAPs) associated with tissue damage bind as ligands to pattern recognition receptors present on immune cells. Activated intracellular transcription factors bind to DNA in the nucleus, and proteins are translated from the transcribed messenger RNA (mRNA), leading to the progression of a systemic inflammatory response. Therefore, comprehensive evaluation of blood proteins using proteomics analysis makes it possible to understand the molecular pathology of trauma.

[0087] In this example, an analysis cohort different from the above-mentioned analysis cohort D was used, and after classification, each cluster was evaluated using serum proteomics analysis data.

[0088] Clinical data from 90 patients admitted to the Advanced Emergency and Critical Care Center at Osaka University Hospital between February 2017 and March 2021 (Osaka University cohort in Figure 6) were collected and classified using the same method as in <Classification of Analysis Cohorts (1)>. Additionally, general research data from patients admitted to the hospital and serum proteomic analysis data analyzed within 72 hours of injury were used to evaluate the molecular pathology of each classified phenotype. This study was conducted in accordance with the Declaration of Helsinki and approved by the Osaka University Ethics Committee (IRB approval no. 16260, 885). The analysis cohort used for this classification is hereafter referred to as Analysis Cohort B.

[0089] As a result of the classification, analysis cohort B was classified into eight phenotypes (B-1 to B-8), similar to analysis cohort D. Each classified phenotype was evaluated using serum proteomics analysis data. In the serum proteomics analysis data, the contents of the components shown below were used as variables. Note that the concentrations of the components shown below may also be used as variables in serum proteomics analysis.・Sodium ・Potassium ・Chloride ・Total bilirubin ・AST (aspartate aminotransferase) ・ALT (alanine aminotransferase) ・LDH (serum lactate dehydrogenase) ・BUN (urea nitrogen) ・Creatinine ・C-Reactive Protein ・WBC count ・Hemoglobin ・Platelet count ・PT-INR (prothrombin) time ・APTT (activated partial thromboplastin time) ・FDP (fibrinogen-fibrin degradation products) - D-dimer - Lactate Evaluation revealed that the B-8 phenotype exhibits excessive inflammation compared to other phenotypes, including an acute inflammatory response, dysregulated complement activation pathways, impaired coagulation, and platelet degranulation pathways.

[0090] In analysis cohort B, differences in serum protein expression between the high mortality phenotype and other phenotypes were evaluated. Specifically, differential expression analysis was performed using Welch's two-group t-test. False discovery rates (FDRs) were calculated using the Benjamin-Hochberg method (see Benjamini, Y. & Hochberg, Y. Journal of the Royal Statistical Society. Series B (Methodological) 57, 289-300 (1995)). Proteins with an FDR of less than 0.2 were considered to be significantly expressed.

[0091] Figure 12 is a volcano plot showing serum proteins whose expression was differentially observed between the high mortality phenotype and other phenotypes. Plots shown in dark colors indicate proteins whose expression was upregulated, and plots shown in light colors indicate proteins whose expression was downregulated. There were 11 upregulated proteins and 26 downregulated proteins.

[0092] Figure 13 shows the results of a gene ontology (GO) enrichment analysis of proteins. The top 12 significant GO terms are shown in Figure 13 along with their FDR-corrected p-values ​​calculated by the Benjamin-Hochberg method. The size of the plot indicates the number of differentially expressed proteins in the list of significant proteins associated with the GO term.

[0093] GO enrichment analysis revealed that phenotype B-8 (equivalent to phenotype D-8 / V-8) exhibited excessive inflammation, including enhanced acute inflammatory response, dysregulation of the complement activation pathway, and coagulation disorders, including reduced coagulation and platelet granule pathways.

[0094] 14 is a plot of each variable in the serum proteomics analysis data for phenotypes B-1 to B-7 and phenotype B-8 (high mortality rate) against the standardized values ​​of each variable. The variables were standardized so that the mean value of the variables was set to 0 and the standard deviation (SD) was set to 1. Phenotype B-8 was found to have high laboratory values ​​related to coagulation, fibrinolysis, and inflammation, and lower fibrinogen and platelet values ​​than the other phenotypes.

[0095] In this example, multiple patients were classified into multiple phenotype groups based on medical data collected from the time the patient was transported by ambulance due to their illness or injury until their first medical examination after their arrival at the medical institution. The analysis cohort was classified into 11 phenotype groups. The validation cohort was also classified into 11 phenotype groups, confirming the reproducibility of this classification method. From the above, a classification model (hereinafter referred to as this classification model) used in the information processing device 1 of the present invention was obtained. Note that the classification model that can be used in the present invention is not limited to this classification model.

[0096] FIG. 15 shows the transition of the classification results for subject O1, classified using the information processing device 1 of the present invention incorporating this classification model. FIG. 15 shows the transition of the classification results at the time of admission, 12 hours, 24 hours, and 48 hours after admission. Subject O1 is a 14-year-old patient who visited the hospital with an unstable pelvic fracture (ISS: 48, TRISS Ps: 0.59). Hemorrhagic shock was present at the time of admission and craniotomy was performed after admission. Subject O1's intracranial pressure gradually increased after admission, increasing the likelihood of death within a predetermined period. As shown in FIG. 15 , as time passes, the mark indicating the position of the subject's classification result gradually approaches the region with a high percentage of patients who will die within a predetermined period (region C shown in FIG. 15 ). This is consistent with the clinical course in which subject O1 became more likely to die within a predetermined period, demonstrating the effectiveness of outputting subject transition information indicating the transition of the subject's classification results using the information processing device 1 of the present disclosure.

[0097] FIG. 16 shows the transition of the classification results of subject O2, classified using the information processing device 1 of the present invention incorporating this classification model. FIG. 16 shows the transition of the classification results at the time of admission, and 12 hours, 24 hours, and 48 hours after admission. Subject O2 is a 20-year-old patient who visited the hospital with a severe head injury (ISS: 17, TRISS Ps: 0.81) and underwent craniotomy and hematoma removal after admission. Subject O2 gradually recovered through treatment after admission. As shown in FIG. 16 , the mark indicating the position of the subject's classification result moves from within the region with a high rate of patient deaths (region C shown in FIG. 16 ) to outside of this region within a predetermined period. This coincides with the clinical course in which subject O2 recovered through treatment after admission, demonstrating the effectiveness of outputting subject transition information indicating the transition of the subject's classification results using the information processing device 1 of the present disclosure.

[0098] 1, 1A Information processing device 4 Presentation device 21 Data acquisition unit 22 Classification unit 23 Object transition information output unit 24 Estimation unit 31 Classification model 32 Sample transition information

Claims

1. An information processing device comprising: a data acquisition unit that acquires medical data regarding a subject; a classification unit that inputs the acquired medical data regarding the subject into a classification model that can classify the subject into one of a plurality of phenotypes based on the medical data, and classifies the subject into one of the plurality of phenotypes; and a subject transition information output unit that outputs subject transition information indicating a transition from a first classification result, obtained based on the medical data regarding the subject at a first time point, to a second classification result, obtained based on the medical data regarding the subject at a second time point after the first time point, wherein the medical data includes at least attribute information, vital sign information, and symptom information, and the symptom information includes at least the site of a disease and the extent of the disease, or the site of an injury and the extent of symptoms in the site.

2. The information processing device of claim 1, further comprising an estimation unit that outputs an estimation result regarding the prognosis of the subject based on a comparison result between specimen transition information indicating a transition from a third classification result in which the specimen is classified into one of the multiple phenotypes based on the medical data regarding the specimen at a third time point to a fourth classification result in which the specimen is classified into one of the multiple phenotypes based on the medical data regarding the specimen at a fourth time point that is after the third time point and before the second time point, and the subject transition information.

3. The information processing device according to claim 1, wherein the subject transition information output unit causes a presentation device to present the subject transition information as a graph in which the first classification result and the second classification result are plotted on coordinates having a first axis representing a first principal component score calculated by principal component analysis of the multiple phenotypes and a second axis representing a second principal component score.

4. The information processing device according to claim 1, wherein the classification model is a clustering model generated based on medical data relating to each of a plurality of subjects suffering from a disease and / or an injury, the medical data being created from the time the subject is rushed to a medical institution for the disease and / or injury and the time the subject finishes receiving treatment for the disease and / or injury at the medical institution.

5. The information processing device according to claim 1, wherein the data acquisition unit acquires at least a portion of the medical data relating to the subject from an electronic medical record of the subject.

6. The information processing device according to claim 1, wherein the symptom information includes an AIS (Abbreviated Injury Scale) code.

7. The information processing device according to claim 1, wherein the attribute information includes age and gender.

8. The information processing device according to claim 1, wherein the vital sign information includes at least one of respiratory rate, pulse rate, systolic blood pressure, level of consciousness (Glasgow Coma Scale (GCS) or Japan Coma Scale (JCS)), and body temperature.

9. An information processing method comprising: a data acquisition step in which a computer acquires medical data on a subject; a classification step in which the computer inputs the acquired medical data on the subject into a classification model capable of classifying the subject into one of a plurality of phenotypes based on the medical data, and classifies the subject into one of the plurality of phenotypes; and a subject transition information output step in which the computer outputs subject transition information indicating a transition from a first classification result obtained based on the medical data on the subject at a first time point to a second classification result obtained based on the medical data on the subject at a second time point after the first time point, wherein the medical data includes at least attribute information, vital sign information, and symptom information, and the symptom information includes at least the site of a disease and the extent of the disease, or the site of an injury and the extent of symptoms at the site.

10. A control program for causing a computer to function as the information processing device according to claim 1, the control program causing a computer to function as the data acquisition unit, the classification unit, and the object transition information output unit.

Images