JP2020144623A - Information processor, information processing method, and program - Google Patents

Abstract

To provide information for performing clinical inference concerning initial diagnosis of a patient.SOLUTION: A patient information acquisition section 101 acquires information including information on a patient as patient information. A disease information generation section 104 generates and acquires first information on a disease that may be related to the patient on the basis of the patient information. A presentation section 105 presents the first information or information on the basis of the first information to a medical staff. An additional information generation section 107 generates or acquires second information on the disease that may be related to the patient on the basis of the first information. The presentation section 105 can also present the second information as information on the basis of the first information.SELECTED DRAWING: Figure 3

Description

The present invention relates to an information processing device, an information processing method, and a program.

Until now, medical professionals who make initial diagnoses need to infer which disease is most likely from interviews with patients and clinical findings, and refer them to appropriate specialists as necessary.
For example, Patent Document 1 proposes a technique for extracting factors for affirming or denying a specific disease from sentences such as medical records and creating an explanation of the basis for a definitive diagnosis.

Japanese Unexamined Patent Publication No. 2015-138402

However, in the prior art including the above-mentioned Patent Document 1, it is the medical professional who infers and specifies which disease is most likely from the medical interview and the medical record. Therefore, individual differences will increase depending on the amount of experience and body of knowledge of medical staff.
The present invention has been made in view of such a situation, and an object of the present invention is to provide information for inferring which disease is likely to occur.

An object of the present invention is to provide information for making clinical reasoning related to the initial diagnosis of a patient.

In order to achieve the above object, the information processing device of one aspect of the present invention is
Patient information acquisition means for acquiring information including information about patients as patient information,
A first generation acquisition means for generating or acquiring a first information about a disease that may be related to the patient based on the patient information.
A presenting means for presenting the first information or information based on the first information to a medical professional, and
To be equipped.
Each of the information processing methods and programs of one aspect of the present invention is an information processing method and program corresponding to the above-mentioned information processing device of one aspect of the present invention.

According to the present invention, it is possible to provide information for making clinical reasoning related to the initial diagnosis of a patient.

It is a figure which shows the example of the flow of information processing using the inference support terminal which concerns on one Embodiment of the information processing apparatus of this invention. It is a block diagram which shows an example of the hardware configuration of the inference support terminal of FIG. It is a functional block diagram which shows an example of the functional structure of the inference support terminal of FIG. This is an example of the patient information and the original problem list display screen displayed on the inference support terminal of FIG. This is an example of the original problem list and the medical terminology problem list display screen displayed on the inference support terminal of FIG. This is an example of the red flag additional information screen displayed on the inference support terminal of FIG. This is an example of the disease narrowing additional information screen displayed on the inference support terminal of FIG. This is an example of a criterion additional information screen displayed on the inference support terminal of FIG. This is an example of a clinical inference support screen displayed on the inference support terminal of FIG. This is an example of a diagnostic hypothesis screen displayed on the inference support terminal of FIG. This is an example of the hypothesis deduction screen displayed on the inference support terminal of FIG. This is an example of the initial diagnosis screen displayed on the inference support terminal of FIG. This is an example of the OPQRST additional information screen displayed on the inference support terminal of FIG. This is an example of the algorithm approach additional information screen displayed on the inference support terminal of FIG.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

First, in explaining the present embodiment, the process of initial diagnosis by a doctor, which is a premise thereof, will be briefly described.
In the process of initial diagnosis, doctor D, who is a medical professional, asks patient P, who is a patient, and makes clinical inferences based on the contents, so that the medical condition and disease of patient P are not clear. , Perform initial diagnosis.
This clinical reasoning is a reasoning that doctor D explains the basic idea (thinking process) and decision-making process in the initial diagnosis, and is systematically to some extent especially among doctors and researchers who specialize in the field. It is understood as a converted theory.
The inference support terminal according to the embodiment of the information processing apparatus of the present invention typically improves the accuracy of the initial diagnosis and the skill related to the initial diagnosis for the doctor D who is not accustomed to the initial diagnosis ( Used for educational purposes).

FIG. 1 is a diagram illustrating an example of a flow of a service (hereinafter, referred to as “the service”) to which the inference support terminal according to the embodiment of the information processing apparatus of the present invention is applied.
FIG. 1 shows a patient P who wishes to be examined by a doctor D, a doctor D who examines the patient P, and an inference support terminal 1 used by the doctor D. In the example of FIG. 1, the person who provides this service to the doctor D will be referred to as the service provider M, and the following description will be given.

The inference support terminal 1 is used by the doctor D as described above. The reasoning support terminal 1 is mainly used to assist the process of clinical reasoning by the doctor D in the process of clinical reasoning in the initial diagnosis.
Further, in the example of FIG. 1, the chief complaint (Chief Complaint), which is the symptom mainly complained by the patient P, is chest pain (Chest Pain). The inference support terminal 1 is used by the doctor D who performs the initial diagnosis for such a patient P.
Hereafter, under such a premise, the overall flow of each step will be explained with respect to an example of the flow of this service.

In step ST1, doctor D conducts an interview with patient P. Here, the interview means that the doctor D asks the current subjective symptoms, age and gender, medical history of patient P and his / her family, etc. at the time of diagnosis, but the interview referred to here is not necessarily an interview as a medical practice. This includes not only the above but also a wide range of general questions.

In step ST2, the doctor D inputs the content of the interview to the inference support terminal 1. Here, the information input by the doctor D will be referred to as patient information in the following description.
Here, the content of the interview is, for example, text data (hereinafter, referred to as "interview result") regarding the result of the interview with the patient P by the doctor D. However, the interview result is not limited to the result of the patient P responding to the question of the doctor D.
That is, specifically, for example, the contents of the questionnaire filled out by patient P before the examination, the vital sign of patient P, the medical record of patient P, various test results of patient P, the voice data recording the response of patient P, and the patient. An explanatory text explaining the state of P, a text describing matters of concern to doctor D, and the like may be included.
In other words, the doctor D inputs various information including the interview result and contributing to the clinical reasoning related to the initial diagnosis into the reasoning support terminal 1 as patient information.

In step ST3, the inference support terminal 1 generates disease information which is information indicating a potential disease candidate possessed by the patient P based on the patient information input in step ST2.
That is, the reasoning support terminal 1 determines a potential disease candidate possessed by the primary patient P based on a predetermined algorithm from the patient information input by the doctor D, that is, various information contributing to clinical reasoning. .. The algorithm used here can be arbitrarily determined by the service provider M, the doctor D, and the like, and an example of a specific algorithm will be described later with reference to FIG. 3 and the like.
In the example of FIG. 1, the reasoning support terminal 1 is, for example, a potential possession of the patient P as a result of performing an calculation by a predetermined algorithm from the information included in the patient information that the chief complaint of the patient P is chest pain. "Acute myocardial infarction" and "reflux esophagitis" are determined as candidates for various diseases.

In step ST4, the inference support terminal 1 presents the disease information generated in step ST3 to the doctor D.
By confirming the presented disease information, the doctor D first first confirms the candidate for the disease that the patient P is suffering from. However, for example, when doctor D is not accustomed to the initial diagnosis, doctor D forgets to ask patient P about the matters that must be asked in the interview, or patient P's response is insufficient. It is quite possible that the interview will be completed regardless. In such cases, Doctor D cannot identify the disease that Patient P is suffering from.
Specifically, for example, it is assumed that doctor D cannot identify the disease that patient P is truly affected by from a plurality of disease candidates that patient P may be affected by.
Further, for example, it is assumed that the doctor D thinks that the patient P is extremely likely to be affected by a predetermined disease, but cannot be sure.
Also, for example, even if doctor D cannot identify the disease that patient P is suffering from, he would like to prioritize the possibility that patient P is suffering from a fatal disease. It is also assumed that this is the case.

Therefore, in step ST5, the doctor D inputs to the inference support terminal 1 a request necessary for himself / herself to accurately carry out the initial diagnosis.
The request referred to here may be a request that contributes to the accurate initial diagnosis by Doctor D. Specifically, in the present embodiment, the inference support terminal 1 can be provided with the following three functions as related functions.
That is, the inference support terminal 1 has a function of identifying whether or not the patient P is suffering from a serious disease, a function of narrowing down candidates for the disease in which the patient P is suffering, in response to a request from the doctor D. It has the function of identifying whether or not patient P has a specific disease.

In step ST6, the inference support terminal 1 generates additional information indicating matters to be additionally asked by the doctor D to the patient P based on the request information input in step ST5.
Here, among the functions of the above-mentioned inference support terminal 1, an example in which additional information related to the function of identifying whether or not a person is suffering from a serious disease is generated will be described.
Specifically, for example, additional information that the reasoning support terminal 1 should examine signs that the occurrence of chest pain was "sudden" to identify whether it is a serious illness such as acute myocardial infarction. To generate.

In step ST7, the inference support terminal 1 presents the additional information generated in step ST6 to the doctor D.

In step ST8, doctor D asks patient P an additional interview based on the additional information presented in step ST7.

In step ST9, the doctor D inputs the content of the additional interview to the inference support terminal 1. Here, the content of the additional interview input to the doctor D is hereinafter referred to as an additional inquiry result. The patient information described above may also include the results of this additional interview.

In step ST10, the inference support terminal 1 generates new disease information (hereinafter referred to as “updated disease information”) based on the patient information including the additional interview result input in step ST9.

In step ST11, the inference support terminal 1 presents the updated disease information generated in step ST10 to the doctor D.

In step ST12, doctor D identifies the disease that patient P is suffering from as an initial diagnosis based on the updated disease information presented in step ST11.
In this updated disease information, as compared with the above-mentioned disease information, candidates for the disease that patient P is suffering from are shown based on more detailed information of patient P and the like. It is expected that post-disease information can be used as more accurate information.

The inference support terminal 1 can generate disease information including information for distinguishing between a serious disease and a normal disease. Here, a serious illness is a illness that requires a large-scale hospital or an emergency response. In addition, a normal disease is a disease that can be treated by follow-up observation or a normal prescription drug.
In addition, the inference support terminal 1 can include information on the degree of the disease-likeness (hereinafter, referred to as “disease-likeness”) in each of the potential disease candidates possessed by the patient P as the disease information. In addition, the inference support terminal 1 can include, as the disease information, information on which content of the patient information improves or lowers the disease-likeness of each of the disease candidates.

This service is generally implemented including such a series of flows. The doctor D who receives the provision of this service can obtain the following merits by using this service, for example.

Doctor D can practice the process of thinking of clinical reasoning by performing clinical reasoning related to the initial diagnosis using the reasoning support terminal 1. As a result, it is more likely that the doctor D can improve the accuracy of the initial diagnosis and the skills related to the initial diagnosis, rather than simply performing the initial diagnosis as a normal task.

Further, the doctor D can obtain a candidate for the disease affecting the patient P as the disease information from the patient information by performing the clinical reasoning related to the initial diagnosis using the reasoning support terminal 1. This makes it possible to learn the relationship between patient information and disease candidates.

In addition, the doctor D can use various additional information in the process of performing clinical inference related to the initial diagnosis by using the inference support terminal 1. That is, the doctor D has an item for specifying whether or not the patient P has a serious disease, an item for narrowing down candidates for the disease that the patient P has, and a patient P who has a specific disease. It is possible to obtain information on items that specify whether or not the patient has a disease. This allows Doctor D to learn what additional interviews can be done to identify the disease and lead to an initial diagnosis.

Further, even if the doctor D is inferior in the clinical reasoning skill related to the initial medical treatment, the doctor D executes the clinical reasoning related to the initial diagnosis with a predetermined quality by using the reasoning support terminal 1. can do.

Here, the algorithm used in the generation of the above-mentioned disease information will be briefly described. However, the algorithm adopted, including the post-odds described later, is merely an example, and can be arbitrarily determined by the service provider M, doctor D, or the like.

Regarding step ST3 described above, the inference support terminal 1 calculates "post-hoc odds" based on the patient information.
Here, the odds are an index showing the probability that the patient has a predetermined disease, and take a numerical value between 0 and positive infinity. The closer the odds are to 0, the lower the "disease-likeness". The higher the odds, the higher the "disease-likeness".
In addition, the concepts of "pre-odds" and "post-odds" will be described.
"Preliminary odds" are predicted indicators (odds) of the existence probability of a certain disease. The "posterior odds" are indexes (odds) indicating the existence probability of a certain disease, which is predicted after the occurrence of risks, symptoms, etc. included in the patient information.
Specifically, the "post-odds" is the sum of the likelihood ratios (likelihood ratio, hereinafter appropriately referred to as "LR") corresponding to each of the plurality of items with respect to the "pre-odds". It can be calculated by

Specifically, for example, in the emergency outpatient department, the prior odds for a patient with sudden headache to have subarachnoid hemorrhage are 1.4 (60% when converted to prior probability). If the actual patient P is "female", the likelihood ratio is 1.1. If patient P has a "history of hypertension", the likelihood ratio is 2.5. In this case, the posterior odds are approximately 4.1 (80% when converted to posterior probabilities), which is the result of multiplying the likelihood ratio. As a result, this patient is initially diagnosed with a high probability of subarachnoid hemorrhage.

The calculation algorithm of the posterior odds of the inference support terminal 1 in the example in which the posterior odds are adopted as the disease-likeness has been described above.

FIG. 1 is a block diagram showing an example of a hardware configuration of an inference support terminal according to an embodiment of the information processing apparatus of the present invention.

The reasoning support terminal 1 includes a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, a RAM (Random Access Memory) 13, a bus 14, an input / output interface 15, an output unit 16, and an input unit. A 17, a storage unit 18, a communication unit 19, and a drive 20 are provided.

The CPU 11 executes various processes according to the program recorded in the ROM 12 or the program loaded from the storage unit 18 into the RAM 13.
Data and the like necessary for the CPU 11 to execute various processes are also appropriately stored in the RAM 13.

The CPU 11, ROM 12 and RAM 13 are connected to each other via the bus 14. An input / output interface 15 is also connected to the bus 14. An output unit 16, an input unit 17, a storage unit 18, a communication unit 19, and a drive 20 are connected to the input / output interface 15.

The output unit 16 is composed of a display, a speaker, and the like, and outputs various information as images and sounds.
The input unit 17 is composed of a keyboard, a mouse, and the like, and inputs various information.

The storage unit 18 is composed of a hard disk, a DRAM (Dynamic Random Access Memory), or the like, and stores various data.
The communication unit 19 communicates with other devices (server 2 and process control terminal 3 in the example of FIG. 5) via a network N including the Internet.

A removable medium 31 made of a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is appropriately mounted on the drive 20. The program read from the removable media 31 by the drive 20 is installed in the storage unit 18 as needed.
In addition, the removable media 31 can also store various data stored in the storage unit 18 in the same manner as the storage unit 18.

FIG. 3 is a functional block diagram showing an example of the functional configuration of the inference support terminal of FIG.

As shown in FIG. 3, in the CPU 11 of the inference support terminal 1, the patient information acquisition unit 101, the medical term conversion unit 102, the posterior odds calculation unit 103, the disease information generation unit 104, the presentation unit 105, and the request The reception unit 106, the additional information generation unit 107, the differential diagnosis acquisition unit 108, and the learning unit 109 function.
Further, a disease likelihood ratio DB 300, a red flag DB 400, a criterion DB 500, and a learning result DB 600 are provided in one area of the storage unit 18 of the inference support terminal 1.

The patient information acquisition unit 101 acquires information including information about the patient P as patient information.
As described above, the patient information includes not only the interview results, but also, for example, the contents of the questionnaire filled out by the patient P before the examination, the vital sign of the patient P, the medical record of the patient P, the various test results of the patient P, and the patient. It may include audio data recording P's response, explanatory text explaining the condition of patient P, text describing matters of concern to doctor D, and the like.

The medical term conversion unit 102 converts the patient information acquired by the patient information acquisition unit 101 into medical terms.
This will be specifically described with an example. For example, the medical term conversion unit 102 extracts a keyword necessary for identifying the disease in which the patient P is suffering from the interview results of the patient information acquired by the patient information acquisition unit 101, and first Generate the next problem list (hereinafter referred to as "original problem list").
It should be noted that such keyword extraction may be performed by using an existing technique such as morphological analysis.
Then, with respect to the original problem list generated in this way, the medical term conversion unit 102 converts each keyword included in the original problem list into a medical term generally available in medicine based on a predetermined method. It is converted to generate a new problem list (hereinafter referred to as "medical term problem list").
The medical terms listed in the medical term problem list are terms in which the keywords included in the specific answers of patient P are medically classified and replaced as higher-level concepts. Further, the medical terms referred to here are, for example, "hypertension", "chest pain" and the like. The universalized medical term is called a Semantic Qualifier (hereinafter referred to as "SQ").

The ex post facto odds calculation unit 103 calculates each index of the disease based on the medical terms described in the medical term problem list acquired by the medical term conversion unit 102.
That is, the ex-post odds calculation unit 103 calculates the ex-post odds of each disease for the patient P based on the medical terms described in the medical term problem list acquired by the medical term conversion unit 102.

The disease information generation unit 104 generates disease information related to a disease that may be related to the patient P based on the patient information.
That is, the disease information generation unit 104 generates information indicating potential disease candidates possessed by the target patient P based on the index (for example, the posterior odds) calculated by the posterior odds calculation unit 103.

The presentation unit 105 presents the disease information generated by the disease information generation unit 104 or the information based on the disease information to the medical staff. That is, it is generated as disease information and the information is presented to the doctor D.
Doctor D attempts an initial diagnosis based on the presented disease information. Here, the doctor D will explain as an example in which the information on the potential disease candidate possessed by the patient P is presented, but the disease cannot be identified as the initial diagnosis.

The request reception unit 106 receives a request from the doctor D who is presented with the disease information. The doctor D inputs to the inference support terminal 1 a request for what kind of additional information he / she wants to be presented as a support for determining whether the patient P has a disease.

The additional information generation unit 107 generates additional information or questions about a disease that may be associated with patient P, based on the disease information.
Further, the additional information generation unit 107 is provided with a generation information determination unit 151, a red flag question generation unit 152, a disease narrowing question generation unit 153, and a criterion question generation unit 154.

The generation information determination unit 151 is among the red flag question generation unit 152, the disease narrowing question generation unit 153, and the criterion question generation unit 154, which will be described later, based on the request from the doctor D received by the request reception unit 106. Which of the above will generate additional information or questions based on the request from Doctor D. The details of the request from the doctor D received by the request reception unit 106 will be described later with reference to FIG. 6 and the like.

When the generation information determination unit 151 determines that the red flag question generation unit 152 generates additional information or a question, the red flag question generation unit 152 determines whether the patient P has a serious illness. Generate additional information or questions that are useful for Doctor D to make a decision.
Specifically, for example, the red flag question generation unit 152 obtains information from the red flag DB 400 from the chief complaint of the patient P obtained from the patient information or the like, including the contents to be confirmed regarding the red flag associated with the chief complaint. Extract additional information or questions to be further confirmed with patient P and generate that information as additional information or questions.

When the generation information determination unit 151 determines that the disease narrowing question generation unit 153 generates additional information or a question, the disease narrowing question generation unit 153 is determined by the disease candidate doctor D who is affected by the patient P. Generate additional information or questions that will be useful for narrowing down.
Specifically, for example, the disease narrowing question generation unit 153 confirms the contents to be confirmed regarding the terms associated with the disease candidate from the disease candidates affected by the patient P obtained from the patient information and the like. From the disease likelihood ratio DB, additional information or a question to be further confirmed by the patient P is extracted, and the information is generated as the additional information or the question.

When the generation information determination unit 151 determines that the criterion question generation unit 154 generates additional information or a question, the criterion question generation unit 154 determines whether the patient P has a specific disease or not by the doctor D. Generate additional information or questions that will be useful to identify.
Specifically, for example, the criteria question generation unit 154 obtains information obtained from patient information and the like, including information to be confirmed regarding the disease that doctor D intends to identify, among the disease candidates that patient P is suffering from. From the DB 500, additional information or a question to be further confirmed by the patient P is extracted, and the information is generated as the additional information or a question.

The presentation unit 105 presents the additional information generated by the additional information generation unit 107 to the doctor D.

Thereby, the doctor D can be presented with additional information based on the request of the doctor D. Based on the additional information, the doctor D can identify the disease and make an initial diagnosis by asking the patient P a question or a test as a question.
Further, when the doctor D does not reach the initial diagnosis, the reasoning support terminal 1 can input further patient information to the doctor D. The inference support terminal 1 in which further patient information is input can update the disease information. Doctor D can identify the disease and make an initial diagnosis by repeating the interview.

The differential diagnosis acquisition unit 108 acquires the result of the differential diagnosis of the patient P who has undergone the initial diagnosis. That is, the patient P who has undergone the initial diagnosis obtains the result of identifying the disease as a result of the differential diagnosis by performing a large-scale hospital, emergency response, follow-up observation, usual prescription drug, and the like.
The result of the differential diagnosis can be utilized as teacher data in the learning department described later.

The learning unit 109 can acquire and learn various inputs and outputs of each of the above-mentioned functional blocks. Specifically, for example, the learning unit 109 can acquire and learn patient information, an original problem list, a medical language problem list, post-hoc odds, and the result of a differential diagnosis. The learned result is stored in the learning result DB 600. The inference support terminal 1 can support the inference of the doctor D based on the learning result stored in the learning result DB 600.

Thereby, for example, the learning unit 109 can learn about the analysis related to the extraction of keywords from the patient information acquired from the patient information acquisition unit 101 and the original problem list acquired from the medical term conversion unit 102. As a result, the medical term conversion unit 102 using the learning result can generate the original problem list more accurately.

In addition, for example, the learning unit 109 is medically classified into general keywords from the original problem list acquired from the medical term conversion unit 102 and the medical term problem list acquired from the medical term conversion unit 102. You can learn how to associate with terms that have been replaced as higher-level concepts. As a result, the medical term conversion unit 102 using the learning result can generate the medical term problem list more accurately.

In addition, for example, the learning unit 109 narrows down the doctor D to the candidate for the disease affected by the patient P from the disease information acquired from the disease information generation unit 104 and the request of the doctor D acquired from the request reception unit 106. You can learn about the relationship between your needs. As a result, the additional information generation unit 107 using the learning result can generate additional information in consideration of the request received from a typical doctor D.

Hereinafter, the service provided by the service will be described based on an example of a screen related to the provision of the service displayed on the inference support terminal 1 of FIG. 2 shown in FIGS. 4 to 14.

FIG. 4 is an example of the patient information and the original problem list display screen displayed on the inference support terminal of FIG.
In the example of the screen of FIG. 4, the interview result input field W1 is displayed in the upper left part of the screen of FIG. Further, in the upper and lower parts of the interview result input field W1 in FIG. 4, fields for inputting the sex, vital signs, and pre-existing illness of patient P are displayed. In these fields, Doctor D inputs patient information, which is various information that contributes to clinical reasoning related to the initial diagnosis, including the result of the interview.
The original problem list column W2 is displayed in the center of the screen of FIG. In the original problem list column W2, an original problem list, which is a list of words that can be a plurality of problems extracted from patient information, is displayed, which is generated by the medical term conversion unit 102 of FIG.
The chief complaint column W3 is displayed in the upper right corner of the screen of FIG. In the screen example of FIG. 4, the chief complaint (Chief Complaint) is chest pain (Chest Pain).

Here, on the screen of FIG. 4, the interview result input field W1 and the original problem list field W2 are displayed side by side. As a result, the doctor D who confirms the screen of FIG. 4 can learn which keyword should be extracted as the original problem list from the patient information. Further, the doctor D can add or delete a keyword extracted from the patient information by operating the original problem list column W2 on the screen of FIG.

Next, the doctor D operates the SQ button B1. As a result, the screen of the example of FIG. 5 is displayed.

FIG. 5 is an example of the original problem list and the medical terminology problem list display screen displayed on the inference support terminal of FIG.
In the example of the screen of FIG. 5, the original problem list column W2 is displayed on the left side of the screen of FIG. Further, in the example of the screen of FIG. 5, the medical term problem list column W4 is displayed on the right side of the screen of FIG.

Here, on the screen of FIG. 5, the original problem list column W2 and the medical term problem list column W4 are displayed side by side. That is, in FIG. 5, the original problem list converted by the medical term conversion unit 102 of FIG. 3 and the medical term problem list are displayed side by side.
Further, the doctor D can change the SQ corresponding to the original problem list by operating the medical term problem list column W4 on the screen of FIG. As a result, the doctor D who confirms the screen of FIG. 5 can learn how the original problem list and the medical term SQ correspond to each other.

The interview result included in the patient information is text data describing the contents of the patient P's raw answer. Therefore, it is highly possible that the described symptoms and terms are expressions that are difficult to interpret as medical terms. Therefore, the inference support terminal 1 converts such an expression that is difficult to interpret into a medical term, and processes the interview result so that the subsequent processing can be easily performed.
As a result, the inference support terminal 1 can improve the accuracy of calculating the posterior odds for each disease by converting the patient information including the so-called natural language into the control word SQ.
Next, the inference support terminal 1 calculates an index of each disease likeness based on the medical term problem list. That is, each SQ in the medical terminology problem list is associated with the likelihood ratio of each disease, and the post-disease odds, which is an index of the probability of each disease, is calculated. The details of the calculation of the ex post facto odds are as described in FIG.

FIG. 6 is an example of a clinical inference support screen displayed on the inference support terminal of FIG.
In the example of the screen of FIG. 6, there is a description of "Diagnostic Reasoning clinical reasoning" at the upper part of the screen of FIG. Further, in the lower left part of the screen of FIG. 6, there is a description of "Problem Representation problem representation", and in the lower right part of the screen of FIG. 6, there is a description of "Hypothetico-deducive method hypothesis deduction method". On the other hand, in the right center portion of the screen of FIG. 6, a circular figure and a reciprocating arrow are described. Based on the problem representation, through hypothesis deduction, we show the process of clinical reasoning in which various judgments are repeatedly used to narrow down the disease and make an initial diagnosis.

The medical terminology problem list is displayed on the left side of the screen of FIG. That is, it is shown that doctor D should make clinical reasoning based on the medical terminology problem list which is a representation of the problem. The inference support terminal 1 can present the disease information, which is the information of the potential disease candidate possessed by the patient P, to the doctor D based on the medical term problem list.

Doctor D, who was presented with the disease information, initially operated the "Red Flags (Prognotic approach)" button B2, the "Questions" button B3, and the "Criteria" button B4 in FIG. 6 as a request for clinical reasoning. Diagnosis can be made. That is, the doctor D can make the request receiving unit 106 of the inference support terminal 1 of FIG. 3 accept the request by operating the buttons B2 to B4.

FIG. 7 is an example of a red flag additional information screen displayed on the inference support terminal of FIG.
The screen of FIG. 7 is displayed when the doctor D operates the "Red Flags (Prognotic approach)" button B2 as one of the approaches based on Bayons'theorem among the screens of FIG.

On the left side of the screen of FIG. 7, a medical term problem list column W4 is displayed. Further, a red flag column W9 is displayed in the center of the screen of FIG. 7.
In the red flag column W9, the contents to be confirmed regarding the red flag associated with the chief complaint shown in the medical terminology problem list column W4 are listed. The inference support terminal 1 can provide additional information to the doctor D via the red flag column W9.
The doctor D selects the red flag from the red flag column W5, and selects whether the patient P applies from YES and NO. As a result, the inference support terminal 1 can acquire information on whether the patient P corresponds to the red flag as further patient information.

FIG. 8 is an example of a disease narrowing additional information screen displayed on the inference support terminal of FIG.
The screen of FIG. 8 is displayed when the doctor D operates the "Questions" button B6 as one of the approaches based on Bayons'theorem among the screens of FIG.

On the left side of the screen of FIG. 8, the medical term problem list column W4 is displayed. Further, a question column W6 is displayed on the right side of the screen of FIG.
In the question column W6, the contents to be confirmed regarding the disease information based on the medical term problem list shown in the medical term problem list column W4 are listed. The inference support terminal 1 can provide additional information to the doctor D via the question column W6.
The doctor D selects a question from the question column W6, and selects whether the patient P applies from YES and NO. As a result, the inference support terminal 1 can acquire information on whether the patient P corresponds to the question as further patient information.

FIG. 9 is an example of a criterion additional information screen displayed on the inference support terminal of FIG.
The screen of FIG. 9 is displayed when the doctor D operates the "Criteria" button B7 as one of the approaches based on Bayons' theorem among the screens of FIG.

The diagnostic hypothesis column W8 is displayed on the left side of the screen of FIG. Further, the criterion column W7 is displayed on the right side of the screen of FIG.
In the criterion column W7, items for specifying whether or not the patient P is affected by the disease based on the disease information shown in the diagnostic hypothesis column W8 are listed. The inference support terminal 1 can provide additional information to the doctor D via the criterion column W7.
The doctor D selects an item for specifying whether or not the patient has a certain disease from the question column W6, and selects whether or not the patient P applies from YES and NO. As a result, the inference support terminal 1 can acquire information on whether the patient P corresponds to the item as further patient information.

Although one embodiment of the present invention has been described above, the present invention is not limited to the above-described embodiment, and modifications, improvements, etc. within the range in which the object of the present invention can be achieved are included in the present invention. is there.

Here, although the description is omitted in the above-described embodiment, further functions that can be provided in the inference support terminal 1 will be described with reference to FIGS. 6 and 10 to 14.

Doctor D, who was presented with the disease information, requested the clinical reasoning to include the "Diagnotic Hypothesis" button B5, the "Abduction (hypothetico-Deducive method)" button B6, the "Initial Diagnosis" button B6, and the "Initial Diagnosis" button B7 in FIG. The initial diagnosis can be performed by operating B8, the "Abduction Approach" button B9. That is, the doctor D can have the request reception unit of the inference support terminal accept the request by operating the buttons B2 to B4.

FIG. 10 is an example of a diagnostic hypothesis screen displayed on the inference support terminal of FIG.
The screen of FIG. 10 is displayed when the doctor D operates the "Diagnostic Hypothesis" button B5 indicating the first step of the arrow coming out of the medical term problem list in the screen of FIG.

On the left side of the screen of FIG. 10, a medical term problem list column W4 is displayed. Further, a diagnostic hypothesis column W8 is displayed in the center of the screen of FIG. 7. Further, on the right side of the screen of FIG. 7, the VINDICATE + P column W9 is displayed.
Here, VINDICATE refers to Vascular (vascular), Infection (infection), Neoplasm (neoplasm), Degenerative (degenerative), Idiopathic (addiction), Congenital (congenital), Autoimune (allergy / autoimmunity), Traum. Trauma), an acronym for Endocrine (metabolism / endocrine system). In addition, P is an acronym for Psychogenic (psychopathic).
That is, in making clinical reasoning, doctor D first shows that the pathological condition can be considered from the viewpoint of VINDICATE + P as a diagnostic hypothesis, which is the first stage of the arrow coming out of the medical term problem list, which is a representation of the problem. ing.
Doctor D selects the item applicable to the patient from the VINDICATE + P column W9. As a result, the inference support terminal 1 can acquire information on whether the patient P corresponds to the item as further patient information.

FIG. 11 is an example of a hypothesis deduction screen displayed on the inference support terminal of FIG.
The screen of FIG. 11 shows the case where the doctor D operates the “Abduction (hypothetico-Deductive method)” button B6 indicating the first stage of the arrow coming out of the medical term problem list in the screen of FIG. ,Is displayed.

On the left side of the screen of FIG. 11, the medical term problem list column W4 is displayed. Further, a diagnostic hypothesis column W8 is displayed in the center of the screen of FIG. Further, on the right side of the screen of FIG. 11, a hypothesis deduction column W10 is displayed.
The doctor D selects from YES and NO whether the patient P applies to the item from the hypothesis deduction column W10. As a result, the inference support terminal 1 can acquire information on whether the patient P corresponds to the item as further patient information.

FIG. 12 is an example of the initial diagnosis screen displayed on the inference support terminal of FIG.
The screen of FIG. 12 is displayed when the doctor D operates the “Initial Diagnosis” button B7 indicating the third stage of the arrow coming out of the medical term problem list in the screen of FIG.

The diagnostic hypothesis column W8 is displayed on the left side of the screen of FIG. Further, the initial diagnosis column W11 is displayed on the right side of the screen of FIG.
In the initial diagnosis column W11, candidates for the disease affecting the patient P are listed based on the disease information.
The doctor D can make an initial diagnosis based on the initial diagnosis column W11 by referring to the post-hoc odds of each disease.

FIG. 13 is an example of the OPQRST additional information screen displayed on the inference support terminal of FIG.
The screen of FIG. 13 is displayed when the doctor D operates the "OPQRST" button B8 as one of the approaches based on Bayons' theorem among the screens of FIG.

On the left side of the screen of FIG. 13, the medical term problem list column W4 is displayed. Further, the OPQRST column W12 is displayed on the right side of the screen of FIG.
Here, OPQRST refers to Onset (onset mode), Palliative / Palliative factor (exacerbation / revelation factor), Quality (property), Region / Radiation / relatated symptom (site / radiation / related symptom), Severity (strength). It is an acronym for Temporal symptoms.
That is, it is shown that the doctor D can consider the pathological condition of the pain complained by the patient from the viewpoint of OPQRST when making clinical reasoning.
The doctor D selects the item applicable to the patient from the OPQRST column W12. As a result, the inference support terminal 1 can acquire information on whether the patient P corresponds to the item as further patient information.

FIG. 14 is an example of the algorithm approach additional information screen displayed on the inference support terminal of FIG.
The screen of FIG. 14 is displayed when the doctor D operates the "Algorithmic Approach" button B9 as one of the approaches based on Bayons' theorem in the screen of FIG.

On the left side of the screen of FIG. 14, the medical term problem list column W4 is displayed. Further, the algorithm approach column W13 is displayed on the right side of the screen of FIG.
In the algorithm approach column W13, a question is displayed based on a predetermined algorithm. In response to the displayed question, doctor D selects from YES and NO whether patient P applies to the item. As a result, the inference support terminal 1 can acquire information on whether the patient P corresponds to the item as further patient information. The inference support terminal 1 that has acquired further patient information displays further questions in the algorithm approach column W13 based on a predetermined algorithm. Similarly, the doctor D selects from YES and NO whether the patient P applies to the item in response to the displayed question. By repeating this, the candidates for the disease affecting the patient P can be narrowed down.
The additional functions that can be provided in the inference support terminal 1 have been described above with reference to FIGS. 6 and 10 to 14.

The inference support terminal 1 can further take the following aspects.

For example, in the above-described embodiment, as a process related to the initial diagnosis, the doctor D asks the patient P to make a clinical reasoning based on the contents, but the present invention is not particularly limited to this.
That is, for example, the patient information does not need to be input to the inference support terminal 1 by the doctor D. Furthermore, the patient information may be input to the inference support terminal 1 as an example of clinical inference related to a case or initial diagnosis.
This allows Doctor D to train in clinical reasoning and initial diagnosis thinking and processes in a way that is more tailored to his or her skills and circumstances.

Further, for example, in the above-described embodiment, the inference support terminal 1 generates an original problem list based on the patient information, converts the generated original problem list into a medical term problem list, and sets the post-odds based on the medical term problem list. Although it has been described as being calculated, it is not particularly limited to this.
That is, for example, from the result of the interview, the doctor D may create a list or the like similar to the medical term problem list by himself / herself and input it to the inference support terminal 1.
Furthermore, the doctor D does not necessarily have to ask the patient P a written or oral question as an interview, and it is sufficient if the doctor D can obtain information for identifying the disease affecting the patient P by any method.
Thereby, for example, the doctor D can develop the ability to think of a medical term problem list for inferring a candidate for a disease in which the patient P is suffering from the patient information in the clinical reasoning related to the initial medical care.

Further, for example, the flow of this service in the above-described embodiment shown in FIG. 1 and the like is merely an example, and is not particularly limited thereto.
That is, the presentation of the inference result by the inference support terminal 1 does not necessarily have to be performed in the order in the above-described embodiment, and is arbitrary. Similarly, the requests from the doctor D to the inference support terminal 1 do not necessarily have to be made in the order in the above-described embodiment.
Specifically, for example, doctor D may make a request to the inference support terminal multiple times as necessary, or confirm in advance a question related to a predetermined disease before the interview. You may.
In other words, the doctor D may use various functions provided in the inference support terminal at any time according to his / her skill and situation.

Further, in the above-described embodiment, the doctor D inputs the patient information to the inference support terminal 1, and then requests the inference support terminal 1 for additional information in order to identify the candidate for the disease that the patient P is suffering from. However, it is not particularly limited to this. That is, for example, the doctor D may repeatedly input the patient information any number of times without requesting additional information. Furthermore, the doctor D can make a request for additional information, but the type and number of times of the additional information is not limited, and the request can be made any number of times in any order.
As a result, doctor D can go through the process of practical reasoning by advancing the reasoning by repeating trial and error in the process of clinical reasoning. That is, the doctor D can improve the skills related to the initial diagnosis.

In the description of the above-described embodiment, the disease information generation unit 104 is said to generate disease information regarding a disease that may be related to the patient P based on the patient information, but the present invention is not particularly limited thereto. That is, for example, it is sufficient to generate or acquire disease information. Specifically, for example, the disease information generation unit 104 may be able to acquire information on a disease that may be related to the patient P from a predetermined database or the like based on the patient information.

In the above description of the embodiment, the differential diagnosis acquisition unit 108 acquires the result of the differential diagnosis of the patient P who has undergone the initial diagnosis, and the result of the differential diagnosis is utilized as teacher data in the learning unit 109. However, it is not limited to this. That is, for example, the patient information and the result of the differential diagnosis may be transmitted to a server (not shown). Thereby, for example, in the server, it can be accumulated as big data including the patient information and the result of the differential diagnosis and further used. That is, various secondary uses are possible, such as utilizing patient information including information on patient P and information on the result of differential diagnosis using the result of identifying a disease as information contributing to case research.

Further, for example, the series of processes described above can be executed by hardware or software.
In other words, the functional configuration of FIG. 3 is merely an example and is not particularly limited.
That is, it suffices if the information processing system is provided with a function capable of executing the above-mentioned series of processes as a whole, and what kind of functional block is used to realize this function is not particularly limited to the example of FIG. Further, the location of the functional block is not particularly limited to FIG. 3, and may be arbitrary. For example, the functional block of the inference support terminal 1 may be transferred to another server or the like (not shown).
Further, one functional block may be configured by a single piece of hardware, a single piece of software, or a combination thereof.

When a series of processes are executed by software, the programs constituting the software are installed on a computer or the like from a network or a recording medium.
The computer may be a computer embedded in dedicated hardware.
Further, the computer may be a computer capable of executing various functions by installing various programs, for example, a general-purpose smartphone or a personal computer in addition to a server.

The recording medium containing such a program is not only composed of a removable medium (not shown) distributed separately from the device main body in order to provide the program to doctor D and the like, but also in a state of being preliminarily incorporated in the device main body. It is composed of a recording medium or the like provided to doctor D or the like.

In the present specification, the steps for describing a program recorded on a recording medium are not necessarily processed in chronological order, but also in parallel or individually, even if they are not necessarily processed in chronological order. It also includes the processing to be executed.
Further, in the present specification, the term of the system shall mean an overall device composed of a plurality of devices, a plurality of means, and the like.

In other words, the information processing apparatus to which the present invention is applied only needs to have the following configuration, and can take various embodiments.
That is, the information processing device to which the present invention is applied (for example, the inference support terminal 1 in FIG. 2) is
A patient information acquisition means (for example, the patient information acquisition unit 101 in FIG. 3) for acquiring information including information about the patient as patient information, and
Based on the patient information, a first generation acquisition means (for example, a disease information generation unit 104 in FIG. 3) for generating or acquiring a first information (for example, disease information) regarding a disease that may be related to the patient, and
A presentation means for presenting the first information or information based on the first information to a medical professional (for example, a presentation unit 105 in FIG. 3) and
To be equipped.
This makes it possible to provide information for inferring which disease is most likely.

Further, a second generation acquisition means (for example, the additional information generation unit 107 in FIG. 3) for generating or acquiring a second information (additional information) regarding a disease that may be related to the patient based on the first information is further provided.
The presenting means may also present the second information as information based on the first information.

The second information may also include a question for the patient.
As a result, the inference support terminal 1 can generate a question or the like to be asked to the patient P as additional information.

Further, a reception means (for example, the request reception unit 106 in FIG. 3) for receiving a request from the medical worker who is presented with the disease information is further provided.
The second generation acquisition means can also generate the second information based on the request in addition to the first information.

In addition, the first information includes information indicating a serious disease.
The second information is information (eg, a red flag) that assists in determining whether the patient has the serious illness.

In addition, the first information includes information indicating a plurality of diseases.
The second information is information (for example, additional information) that assists in determining whether the disease of the patient includes at least one of the diseases specified in the first information.

1 ... Reasoning support terminal, 11 ... CPU, 16 ... Output unit, 17 ... Input unit, 101 ... Patient information acquisition unit, 102 ... Medical term conversion unit, 103 ... Ex-post odds calculation unit, 104 ... disease information generation unit, 105 ... presentation unit, 106 ... request reception unit, 107 ... additional information generation unit, 108 ... differential diagnosis acquisition unit, 109 ... -Learning unit, 151 ... Generation information determination unit, 152 ... Red flag question generation unit, 153 ... Disease narrowing question generation unit, 154 ... Criteria question generation unit, 300 ... Disease likelihood Ratio DB, 400 ... Red flag DB, 500 ... Criteria DB, 600 ... Learning result DB

Claims (8)

Patient information acquisition means for acquiring information including information about patients as patient information,
A first generation acquisition means for generating or acquiring a first information about a disease that may be related to the patient based on the patient information.
A presenting means for presenting the first information or information based on the first information to a medical professional, and
Information processing device equipped with. A second generation acquisition means for generating or acquiring a second information regarding a disease that may be related to the patient based on the first information is further provided.
The presenting means presents the second information as information based on the first information.
The information processing apparatus according to claim 1. The second information includes questions to the patient.
The information processing device according to claim 2. Further provided with a reception means for receiving a request from the medical worker who is presented with the second information,
The second generation acquisition means generates the second information based on the request in addition to the first information.
The information processing device according to claim 2 or 3. The first information includes information indicating a serious disease.
The second information is information that assists in determining whether the patient has the serious disease.
The information processing device according to any one of claims 2 to 4. The first information includes information indicating a plurality of diseases.
The second information is information that assists in determining whether the disease of the patient includes at least one of the diseases specified in the first information.
The information processing device according to any one of claims 2 to 4. In the information processing method executed by the information processing device
Patient information acquisition step to acquire information including information about the patient as patient information,
A generation acquisition step of generating or acquiring disease information regarding a disease that may be related to the patient based on the patient information.
A presentation step of presenting the disease information or information based on the disease information to a medical professional, and
Information processing methods including. On the computer
Patient information acquisition step to acquire information including information about the patient as patient information,
A generation acquisition step of generating or acquiring disease information regarding a disease that may be related to the patient based on the patient information.
A presentation step of presenting the disease information or information based on the disease information to a medical professional, and
A program that executes control processing including.

Images