JP7479168B2 - Interview device - Google Patents

Description

One embodiment disclosed in the specification relates to a medical interview device.

Conventionally, when patients visited a hospital for a medical examination, they received a pre-prepared medical questionnaire at the hospital reception and filled out the questionnaire with their symptoms.

Because these pre-prepared medical questionnaires are created for an unspecified number of patients, there is the problem that it is difficult to accurately record the patient's actual symptoms. For this reason, a system has been proposed that generates new medical questionnaire items based on the results of the interview and asks the questions again.

However, in this type of conventional system, the new interview questions only asked more detailed questions about the symptoms that the patient had already answered, and did not efficiently examine the patient's symptoms.

JP 2012-113629 A

The medical interview device according to the embodiment includes a first reception unit, a presentation unit, a second reception unit, and a first determination unit. The first reception unit receives an initial response to an initial questionnaire asking about symptoms. The presentation unit presents an additional questionnaire asking about symptoms related to the initial response based on the initial response. The second reception unit receives an additional response to the additional questionnaire. The first determination unit determines at least one first disease name based on the initial response and the additional response.

FIG. 1 is a system block diagram illustrating an example of a configuration of an inquiry system according to an embodiment. FIG. 2 is a functional block diagram illustrating an example of the functional configuration of the user terminal and the doctor terminal. FIG. 3 is a functional block diagram illustrating an example of a functional configuration of the server device according to the embodiment. FIG. 4 is a diagram showing an example of the initial questionnaire. FIG. 5 is a diagram showing an example of the correspondence between symptoms and disease names stored in the disease name/symptom database. FIG. 6 is a diagram showing an example of selecting a symptom for which an additional question is desirable. FIG. 7 is a diagram showing an example of the additional medical questionnaire. FIG. 8 is a diagram showing an example of a display when the server device causes the doctor terminal to display symptoms obtained by a medical interview. FIG. 9 is a diagram showing an example of medical interview information that the server device causes the doctor terminal to display. FIG. 10 is a flowchart showing an example of the flow of processing performed by the interview system. FIG. 11 is a diagram for explaining an inquiry system 5 according to the second embodiment.

Hereinafter, an embodiment of an interview device and an interview system will be described with reference to the drawings.
First Embodiment
FIG. 1 is a diagram illustrating an example of a system configuration of a medical interview system according to an embodiment.

(Explanation of the overall configuration of the medical interview system)
As shown in FIG. 1, the medical interview system 1 according to the present embodiment includes a server device 2, a doctor terminal 4, and a user terminal 10. The server device 2 and the user terminal 10 are connected to each other via a communication network N so that they can communicate with each other. The communication network N is, for example, a dedicated line such as the Internet or a VPN (Virtual Private Network). The server device 2 and the doctor terminal 4 are directly or indirectly communicable with each other via, for example, an in-hospital LAN (Local Area Network) 6 installed in a hospital. Since each patient possesses a user terminal 10, there may be multiple user terminals 10. Since the doctor terminal 4 is installed in each examination room, there will be multiple doctor terminals 4 when there are multiple examination rooms.

The server device 2 accepts an initial response to an initial questionnaire asking about symptoms from the user terminal 10, and causes the user terminal 10 to present an additional questionnaire asking about symptoms related to the initial response based on the initial response. The server device 2 accepts additional responses to the additional questionnaire from the user terminal 10, and determines at least one first disease name based on the initial response and the additional responses. Note that the term "determine" in this embodiment includes estimation using an AI model and identification by calculation using a table.

The server device 2 also transmits an initial questionnaire to the user terminal 10, asking about symptoms. The server device 2 also receives a response to the initial questionnaire from the user terminal 10, and infers at least one disease name based on the symptoms answered to the initial questionnaire. The server device 2 then infers other symptoms from the inferred disease name, and generates an additional questionnaire asking about symptoms to be reconfirmed according to the other inferred symptoms. The server device 2 may obtain an additional questionnaire according to the response to the initial questionnaire from among the additional questionnaires already created. The server device 2 also transmits an additional questionnaire to the user terminal 10, and receives a response to the additional questionnaire from the user terminal 10. The server device 2 then further narrows down disease name candidates based on the response to the additional questionnaire. The server device 2 is an example of a medical interview device. The user terminal 10 is an example of a terminal device.

The doctor terminal 4 is, for example, a terminal such as a personal computer that is installed in each examination room. The doctor terminal 4 receives from the server device 2 medical interview information (medical interview results) including responses to the initial interview form and additional interview form and the disease name estimated by the server device 2, and displays the received medical interview information. The doctor then examines the patient by referring to the medical interview information, and inputs the medical record information from the electronic medical record, i.e., the patient's symptoms and the doctor's findings, from the doctor terminal 4.

The user terminal 10 receives the initial questionnaire and the additional questionnaire from the server device 2. The user terminal 10 also accepts responses to the initial questionnaire and the additional questionnaire from, for example, a patient. The user terminal 10 also transmits the responses to the initial questionnaire and the additional questionnaire to the server device 2. The user terminal 10 is a mobile terminal such as a smartphone or tablet PC (Personal Computer) that is carried by the patient or a patient's family member.

(Description of the functional configuration of the user terminal and the doctor terminal)
2 is a functional block diagram showing an example of the functional configuration of the user terminal 10 and the doctor terminal 4. As shown in FIG. 2, the user terminal 10 and the doctor terminal 4 have a configuration in which a control unit 101, a storage unit 102, a communication I/F 103, and an I/O device control unit 104 are connected via a bus 105. The user terminal 10 may have at least a function of displaying a medical questionnaire, a function of filling out the medical questionnaire, and a function of transmitting and receiving the medical questionnaire via a network or a communication line, and may be, for example, a smartphone. The doctor terminal 4 may have at least a function of viewing a medical questionnaire, and a function of transmitting and receiving the medical questionnaire via a network or a communication line, and may be, for example, a medical record terminal.

The control unit 101 has a function as a processor and is composed of a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), etc. The control unit 101 controls each part of the doctor terminal 4 and the user terminal 10 by the CPU reading out the control programs stored in the ROM and the storage unit 102, expanding them into the RAM, and executing them sequentially. The storage unit 102 is a hard disk drive device or the like, and stores the control programs executed by the CPU of the control unit 101 and various setting data. For example, the control programs stored in the storage unit 102 include a browser that communicates with the server device 2, and a file communication app that transmits and receives questionnaire files to and from the server device 2.

The communication I/F 103 controls communication between the user terminal 10 and the server device 2, and between the doctor terminal 4 and the server device 2. Specifically, the communication I/F 103 of the user terminal 10 performs data communication with the server device 2 via the communication network N. In addition, the communication I/F 103 of the doctor terminal 4 performs data communication with the server device 2 via the hospital LAN 6.

The I/O device control unit 104 controls various I/O devices under the control of the control unit 101. The I/O devices controlled by the I/O device control unit 104 are, for example, an operation unit 106 such as a keyboard, mouse, or touch panel, and a display unit 107 such as a liquid crystal display.

In particular, in the user terminal 10, the communication I/F 103 described above receives an initial questionnaire and an additional questionnaire from the server device 2. In addition, the communication I/F 103 transmits answers to the initial questionnaire (initial answers) and answers to the additional questionnaire (additional answers) to the server device 2. Note that the communication I/F 103 is an example of a receiving unit and a transmitting unit.

In addition, in the user terminal 10, the display unit 107 displays the initial questionnaire and the additional questionnaire. The control unit 101 receives responses to the initial questionnaire and the additional questionnaire from the patient. The control unit 101 is an example of a reception unit.

(Description of the Functional Configuration of the Server Device)
3 is a functional block diagram showing an example of a functional configuration of the server device 2. As shown in FIG.

The control unit 201 has the function of a processor and is composed of a CPU, ROM, RAM, etc., and the CPU controls each part of the server device 2 by expanding the control program P1 stored in the ROM and the memory unit 220 into the RAM and executing it sequentially.

The control unit 201 realizes, as functional units, a first reception unit 202, a disease name estimation unit 203, a symptom estimation unit 204, a medical questionnaire regeneration unit 205, a second reception unit 206, an estimated disease name identification unit 207, a medical questionnaire information generation unit 208, a display control unit 209, and a communication control unit 210, all of which are shown in FIG. 3, as a result of the CPU sequentially executing the control program P1.

The first reception unit 202 receives a response to the initial questionnaire C1 that asks about symptoms. Specifically, the server device 2 receives the completed initial questionnaire C1 from the user terminal 10. The first reception unit 202 may receive the initial questionnaire C1 each time a user who is filling out the initial questionnaire C1 at the user terminal 10 operates a "Next" button or the like to indicate the next questionnaire item on the initial questionnaire C1. The server device 2 determines that there is a request for a questionnaire when it detects that an application for inputting the initial questionnaire C1 has been started at the user terminal 10, or when it detects that a hospital name has been input at the user terminal 10, and causes the communication control unit 210 to transmit the initial questionnaire C1 to the user terminal 10.

The disease name estimation unit 203 estimates (or judges) at least one estimated disease name based on the answer received by the first reception unit 202. The disease name estimation unit 203 estimates at least one estimated disease name as a first or second disease name based on the patient's symptoms (initial answer) entered in the initial questionnaire C1 received by the first reception unit 202. More specifically, the disease name estimation unit 203 estimates at least one estimated disease name using a table that associates symptoms with disease names that cause the symptoms. In addition, the disease name estimation unit 203 estimates at least one estimated disease name using a trained model that receives at least one symptom as input and outputs at least one second disease name. The disease name estimation unit 203 is an example of a first or second judgment unit.

The symptom estimation unit 204 estimates symptoms of the estimated disease based on the estimated disease name estimated by the disease name estimation unit 203. When the disease name estimation unit 203 estimates only one estimated disease name, the symptom estimation unit 204 estimates other symptoms that may develop due to the one estimated disease name.

The questionnaire regeneration unit 205 generates an additional questionnaire C2 that asks about symptoms corresponding to the symptoms estimated by the symptom estimation unit 204. The questionnaire regeneration unit 205 also highlights the symptoms estimated by the symptom estimation unit 204 in the generated additional questionnaire C2. If the disease name estimation unit 203 estimates only one estimated disease name, the questionnaire regeneration unit 205 generates an additional questionnaire C2 that asks about other symptoms that may be caused by the one estimated disease name. The questionnaire regeneration unit 205 and the symptom estimation unit 204 are an example of a first generation unit. That is, the questionnaire regeneration unit 205 and the symptom estimation unit 204 generate an additional questionnaire using a trained model that takes at least one disease name as input and outputs a questionnaire. The questionnaire regeneration unit 205 and the symptom estimation unit 204 also generate an additional questionnaire by calculation using a table that associates multiple disease names with multiple questions.

Furthermore, when the disease name estimation unit 203 estimates a plurality of estimated disease names, the questionnaire regeneration unit 205 generates an additional questionnaire C2 including questions for symptoms estimated to be most likely to have developed from among the symptoms estimated by the symptom estimation unit 204 from each of the plurality of estimated disease names. Note that the questionnaire regeneration unit 205 may distinguish the possibility of development into levels, for example, from 1, the lowest, to 5, the highest, and include questions related to symptoms whose possibility of development is higher than a predetermined level in the additional questionnaire C2.

In addition, when the disease name estimation unit 203 estimates multiple estimated disease names, the questionnaire regeneration unit 205 generates an additional questionnaire C2 including questions for symptoms that overlap with the multiple estimated disease names among the symptoms estimated by each of the multiple estimated disease names by the symptom estimation unit 204.

In addition, when the disease name estimation unit 203 estimates multiple estimated disease names, the questionnaire regeneration unit 205 generates an additional questionnaire C2 including questions for symptoms estimated from each of the multiple estimated disease names, which have little overlap among the multiple estimated disease names, from among the symptoms estimated from each of the multiple estimated disease names by the symptom estimation unit 204.

The medical questionnaire regeneration unit 205 may obtain an additional medical questionnaire C2 corresponding to the answer to the initial medical questionnaire C1 from among the additional medical questionnaires C2 that have already been created. A specific example of the additional medical questionnaire C2 will be described later (Figure 7).

The second reception unit 206 accepts responses to the additional questionnaire C2 generated by the questionnaire regeneration unit 205.

The estimated disease name identification unit 207 identifies (or determines) the estimated disease name estimated by the disease name estimation unit 203 as the first or second disease name based on the initial answer to the initial questionnaire and the symptoms received by the second reception unit 206 as additional answers. That is, the estimated disease name identification unit 207 identifies (or determines) the estimated disease name estimated by the disease name estimation unit 203 as the first or second disease name using a table that associates symptoms with disease names that cause the symptoms. In addition, the estimated disease name identification unit 207 identifies (or determines) the estimated disease name estimated by the disease name estimation unit 203 as the first or second disease name using a trained model that inputs at least one symptom and outputs at least one disease name. Furthermore, when the disease name estimation unit 203 estimates only one estimated disease name, the estimated disease name identification unit 207 determines whether or not there is a possibility that the one estimated disease name has developed based on the answer to the additional questionnaire C2. The presumed disease name identification unit 207 is an example of a first or second determination unit.

The medical interview information generating unit 208 generates medical interview information 35 including the answers (initial answers) to the initial medical interview form C1 and the answers (additional answers) to the additional medical interview form C2, and the disease name identified by the estimated disease name identifying unit 207. The medical interview information generating unit 208 is an example of a second generating unit. The medical interview information 35 highlights symptoms that characterize the disease name identified by the estimated disease name identifying unit 207.

The display control unit 209 causes the initial questionnaire C1 and the additional questionnaire C2 to be displayed on the display unit 107 of the user terminal 10. The display control unit 209 also causes the medical interview information 35 to be displayed on the display unit 107 of the doctor terminal 4 where the doctor inputs the medical record information.

The communication control unit 210 controls the communication between the control unit 201 and the memory unit 220. The communication control unit 210 also controls the communication between the server device 2 and the user terminal 10, and the communication between the server device 2 and the doctor terminal 4.

The storage unit 220 is a hard disk drive device or the like, and stores the control program P1 executed by the CPU of the control unit 201, various data, and the like. The storage unit 220 also stores a disease name and symptom database DB, a disease name estimation model M1, a symptom estimation model M2, and an initial questionnaire C1.

The control program P1 executed by the control unit 201 may be provided in an installable or executable format recorded on a computer-readable recording medium such as a CD-ROM, a flexible disk (FD), a CD-R, or a DVD (Digital Versatile Disk).

The control program P1 may also be configured to be stored on a computer connected to a network such as the Internet and provided by being downloaded via the network. Furthermore, the control program P1 may also be configured to be provided or distributed via a network such as the Internet.

The disease name and symptoms database DB is a database that manages the names and symptoms of various diseases. The disease name and symptoms database DB is generated based on the medical records of doctors at each hospital and guidelines from academic societies, etc.

The disease names and symptoms stored in the disease name and symptoms database DB are associated with each other. For example, in the disease name and symptoms database DB, each uniquely assigned disease name ID is associated with the symptoms that are the main complaint of that disease. Furthermore, each symptom is assigned a unique ID, and is associated with the disease name that is the source of that symptom.

The disease name estimation model M1 is a model that outputs one or more disease names for at least one symptom. Such a disease name estimation model M1 can be realized by a first model (AI model) that stores a symptom, a disease name that causes the symptom, and information that indicates the likelihood of the association between the symptom and the disease name in association with each other, and a second model that performs calculations using a table that stores a symptom and a disease name that causes the symptom in association with each other.

The disease name estimation model M1 as the first model is a program that realizes an AI model such as a neural network. That is, the disease name estimation model M1 is, for example, a trained model (such as a deep neural network) that takes at least one symptom as input and outputs one or more disease names. A program that realizes such a disease name estimation model M1 can be constructed by machine learning using multiple training data consisting of a combination of at least one symptom that is generally the main complaint of patients with the disease and the disease name diagnosed by a doctor.

The second model, disease name estimation model M1, is a calculation program that determines at least one disease name based on at least one symptom, based on a table that has been created in advance and that defines multiple correspondences between disease names and symptoms.

Note that whether the disease name estimation model M1 is the first model or the second model, it can be formed using, for example, data stored in the disease name and symptom database DB. Furthermore, information indicating the likelihood of the association between symptoms and disease names may be calculated, for example, based on the number of samples used to train the disease name estimation model M1. Specifically, if M patients are diagnosed with pneumonia out of N patients who have the symptom "coughing," the likelihood that pneumonia will be inferred from the cough is registered as M/N. Note that this likelihood changes as the disease name estimation model M1 is updated.

The symptom estimation model M2 is a model (information) that stores an association between a disease name, symptoms caused by the disease name, and information that indicates the likelihood of the association between the disease name and the symptoms. The symptom estimation model M2 is a model that outputs one or more symptoms for at least one disease name. Such a symptom estimation model M2 can be realized by a third model (AI model) that stores a disease name, at least one symptom caused by a disease with the disease name, and information that indicates the likelihood of the association between the disease name and the symptoms in association with each other, a fourth model that performs calculations using a table that stores a disease name and symptoms caused by a disease with the disease name in association with each other, etc.

The symptom estimation model M2 as a third model is a program that realizes an AI model such as a neural network. That is, the symptom estimation model M2 may be, for example, a trained model (such as a deep neural network) that inputs at least one disease name and outputs one or more symptoms. A program that realizes such a symptom estimation model M2 can be constructed by machine learning using multiple training data consisting of a combination of a disease name diagnosed by a doctor and at least one symptom that is generally the main complaint of patients with that disease.

The fourth model, symptom estimation model M2, is a calculation program that determines at least one symptom based on at least one disease name, based on a table that defines multiple correspondences between symptoms and disease names that have been created in advance.

Note that whether the symptom estimation model M2 is the third model or the fourth model, it can be formed using data stored in the disease name and symptom database DB, for example. Furthermore, information indicating the likelihood of the association between the disease name and the symptom may be calculated, for example, based on the number of samples used to train the symptom estimation model M2. Specifically, if there are M patients who have the symptom "coughing" out of N patients who have the symptom "pneumonia," the likelihood that coughing will be inferred from pneumonia is registered as M/N. Note that this likelihood changes as the symptom estimation model M2 is updated.

The disease name estimation model M1 and the symptom estimation model M2 may be created and updated using existing clinical data, or may be created and updated using data collected by operating the medical interview system 1 of this embodiment, i.e., the correspondence between the contents of the medical interview form and the actual diagnosis results.

The disease name estimation model M1 and the symptom estimation model M2 have a learning function. That is, when the disease name estimated by the server device 2 differs from the disease name diagnosed by the doctor, the server device 2 updates the disease name estimation model M1 and the symptom estimation model M2 by associating the patient's symptoms with the disease name diagnosed by the doctor through the operation of a model update unit not shown in FIG. 3.

The initial questionnaire C1 is a questionnaire that is first presented to a patient using the user terminal 10 and asks about the patient's symptoms. Details will be described later (Figure 4).

(Explanation of the initial medical questionnaire)
4 is a diagram showing an example of the initial questionnaire C1. The initial questionnaire C1 includes symptom information 20, a check box 21, and an input completion button 22. The patient checks the initial questionnaire C1 displayed on the display unit 107 of the user terminal 10. The patient then operates the operation unit 106 (e.g., a touch panel) to input a check mark 23 in the check box 21 that corresponds to the symptom information 20 related to the patient's own symptoms from among the symptom information 20 displayed on the initial questionnaire C1. Specifically, the control unit 101 of the user terminal 10 displays a check mark 23 in the check box 21 that has been touched.

After inputting all of the patient's symptoms, the patient presses the input completion button 22. The control unit 101 of the user terminal 10 detects that the input completion button 22 has been pressed, and transmits the initial questionnaire C1 to the server device 2.

The initial medical questionnaire C1 is created in advance for the purpose of checking the patient's symptoms comprehensively and is stored in the storage unit 220 of the server device 2. The initial medical questionnaire C1 includes items for various physical symptoms so that the patient's symptoms can be checked comprehensively.

For patients who wish to visit a specific specialty, an initial questionnaire C1 specialized for that specialty (e.g., otolaryngology, ophthalmology, orthopedics, etc.) may be prepared in the memory unit 220.

In addition, the initial questionnaire C1 shown in FIG. 4 is answered by entering a check mark 23 in a check box 21, but the answering format is not limited to this method. That is, the patient may enter their symptoms in words (e.g., headache) or in sentences (e.g., my head hurts).

Furthermore, the patient may input the severity of the symptoms as a numerical value (e.g., on a scale of 1 to 10). The first reception unit 202 of the server device 2, which receives the results of filling out the initial questionnaire C1, is configured to be able to receive information entered according to the input format of the initial questionnaire C1. Specifically, the server device 2 is equipped with a text recognition function that can receive responses in the form of, for example, words or sentences.

(Explanation of the method for predicting disease candidates)
5 is a diagram showing an example of the correspondence between symptoms and disease names stored in the disease name/symptom database DB. The disease name/symptom database DB stores symptoms 40 and suspected disease names 41 in association with each other.

The disease name estimation unit 203 estimates a disease name using the disease name estimation model M1. For example, as shown in FIG. 4, assume that the patient's main complaint in the initial medical questionnaire C1 is "chest pain." In this case, the disease name estimation unit 203 estimates, from the symptom 40 column, the disease name that presents symptoms related to "chest pain" as the most likely disease name, using the disease name estimation model M1 created based on the disease name and symptoms database DB.

Specifically, when the disease name estimation model M1 is described in a table format, the disease name estimation unit 203 reads out the estimated disease name associated with the symptom "chest pain" from the table. When the disease name estimation model M1 is described in a mathematical model such as a neural network, the disease name estimation unit 203 may predict the estimated disease name based on an evaluation value corresponding to the possibility of developing the symptom, which is output to the output layer of the neural network when the symptom "chest pain" is input. At this time, the disease name estimation unit 203 may output only one estimated disease name that outputs the highest evaluation value as the most likely estimated disease name, or may output multiple estimated disease names that output evaluation values higher than a predetermined threshold as likely estimated disease names. In addition, when the disease name estimation model M1 has a certainty that the symptom and the estimated disease name are associated with each other, the certainty may be used to determine the estimated disease name to be output.

In the example of Figure 5, the symptom of "chest pain" hatched in the figure is suspected to be spontaneous pneumothorax, lung cancer, reflux esophagitis, angina pectoris, acute myocardial infarction, acute heart failure, acute myocarditis, bacterial pneumonia, pleurisy, sudden esophageal rupture, acute pancreatitis, etc.

In this embodiment, of the 11 diseases listed above, five diseases, spontaneous pneumothorax, lung cancer, angina pectoris, acute myocardial infarction, and bacterial pneumonia, are estimated to be highly likely diseases.

(Explanation of the method for estimating symptom candidates)
The symptom estimation unit 204 estimates symptoms using the symptom estimation model M2. More specifically, the symptom estimation unit 204 estimates whether there is another symptom that may have developed, based on the disease name estimated by the disease name estimation unit 203 in response to the answer to the initial questionnaire C1.

The symptom estimation unit 204 may estimate one symptom that is most likely to have developed, or may estimate multiple symptoms that are likely to have developed. In this case, there is a possibility that the symptom that is the main complaint in the initial questionnaire C1 may be estimated, and there is also a possibility that a symptom that is not the main complaint in the initial questionnaire C1 may be estimated. The medical interview system 1 judges that it is desirable to reconfirm with the patient the presence or absence of a symptom that is not the main complaint in the initial questionnaire C1 from among the symptoms estimated by the symptom estimation unit 204.

Note that symptoms that are not the main complaint in the initial medical questionnaire C1 include symptoms that are not included as questions in the initial medical questionnaire C1, and symptoms that were included as questions in the initial medical questionnaire C1 but were not answered.

The medical questionnaire regeneration unit 205 generates an additional medical questionnaire C2 including symptoms that should be reconfirmed based on the symptoms estimated by the symptom estimation unit 204. A specific example of the additional medical questionnaire C2 will be described later (Figure 7).

Figure 6 shows an example of selecting symptoms for which additional questions are desirable. In particular, Figure 6 shows an example of selecting additional questions based on the suspected illnesses that are suspected when a patient complains of cough, fever, and chest pain in the initial questionnaire C1.

The symptom estimation unit 204 estimates symptoms caused by the estimated disease name from the five estimated disease names estimated by the disease name estimation unit 203. More specifically, the symptom estimation unit 204 estimates, among the symptoms estimated from the estimated disease names estimated by the disease name estimation unit 203, symptoms that were not included as questions in the initial medical questionnaire C1 and symptoms that were included as questions in the initial medical questionnaire C1 but were not answered.

Figure 6 shows an example in which the disease name estimation unit 203 estimates five suspected disease names, as described in Figure 5: spontaneous pneumothorax 41a, lung cancer 41b, angina pectoris 41c, acute myocardial infarction 41d, and bacterial pneumonia 41e.

The symptom estimation unit 204 uses the symptom estimation model M2 to estimate symptoms that are likely to occur for each disease. Then, the symptom estimation unit 204 selects one or more symptoms that are likely to develop from the estimated symptoms.

Specifically, the symptom estimation unit 204 may select, for each disease name, the symptom that is most likely to develop. For example, the symptom estimation unit 204 may output only one symptom that is most likely to develop, using information that indicates the likelihood of the association between the disease name and the symptom stored in the symptom estimation model M2.

In addition, when the symptom estimation model M2 is configured as a neural network, the symptom estimation unit 204 may predict symptoms based on the evaluation value output to the output layer of the neural network when a specific disease name is input. In this case, the symptom estimation unit 204 may output only one symptom that outputs the highest evaluation value as the symptom most likely to develop.

The symptom estimation unit 204 may also use information indicating the likelihood of the association between the disease name and the symptom stored in the symptom estimation model M2 to output multiple symptoms having a likelihood higher than a predetermined threshold as symptoms that are likely to develop. The symptom estimation unit 204 may also output multiple symptoms having an evaluation value higher than a predetermined threshold as symptoms that are likely to develop, based on the evaluation value output to the output layer of the neural network.

Furthermore, the symptom estimation unit 204 may output symptoms that appear in multiple disease name candidates as symptoms that are likely to develop.

In addition, the symptom estimation unit 204 may output a symptom that the same patient has frequently entered in the initial questionnaire C1 in the past as a symptom that is likely to develop.

Furthermore, the symptom estimation unit 204 may output symptoms estimated from multiple estimated disease names that make it easier to identify the disease name.

For example, the symptom estimation unit 204 may output symptoms that do not overlap among multiple disease names. In addition, the symptom estimation unit 204 may output symptoms that the same patient has entered in the initial questionnaire C1 a small number of times in the past.

The symptom estimation unit 204 may also output symptoms that are useful for distinguishing between the incorrect disease name and the correct disease name when a misdiagnosis has occurred in the past. Furthermore, the symptom estimation unit 204 may output symptoms estimated by taking into account the patient's medical history for each of the above-mentioned estimations by referring to the patient's medical history.

The additional questions shown in FIG. 6 are symptoms extracted as a result of each of the above-mentioned estimations. That is, shortness of breath and palpitations 40a are examples of symptoms that overlap widely between multiple disease names. Also, weight loss 40b, bone and joint pain 40c, and bloody sputum 40d are examples of symptoms that each have a one-to-one correspondence with lung cancer 41b, that is, symptoms that overlap little between multiple disease names.

Regarding sputum, it is advisable to ask about the condition of the sputum, for example, whether there is bloody sputum 40d or purulent sputum 40e, in order to identify the name of the disease.

The supplemental questions about chest pain, i.e., 40f, which subsides within 30 minutes, 40g, which continues for more than 30 minutes, and 40h, pain in the upper abdomen, are examples of questions that ask in more detail about the state of the chest pain in order to identify the name of the illness. Such supplemental questions are also applicable, for example, to situations in which a patient complaining of a headache is asked about the state of the headache (does it hurt all over the head, does it hurt in a specific part, what kind of pain it is, when did it start hurting, etc.).

The questionnaire regeneration unit 205 generates an additional questionnaire C2 (FIG. 7) described later based on the additional questionnaire items shown in FIG. 6.

(Explanation of the additional medical questionnaire)
Fig. 7 is a diagram showing an example of the additional questionnaire C2. In particular, the left side of Fig. 7 shows an example of the additional questionnaire C2a, and the right side of Fig. 7 shows an example of the additional questionnaire C2b.

The additional medical questionnaire C2 follows the display format of the initial medical questionnaire C1, taking into consideration ease of understanding for the patient. It is desirable to highlight symptoms that are estimated by the symptom estimation unit 204 and require reconfirmation. There is no restriction on the method of highlighting, but it is possible to add hatching, underline, change the text color, change the font, make the text bold, etc.

The left side of Figure 7 shows an example of a supplemental questionnaire C2a in which the symptoms 24 that should be highlighted are hatched.

As another form of highlighting, symptoms that should be reconfirmed may be rearranged and displayed at the top of the symptom information 20. Also, predicted symptoms may be additionally displayed at a specific position on the screen, such as, "Are you suffering from a headache?".

The questionnaire regeneration unit 205 may also generate an additional questionnaire C2b by adding additional questions to the initial questionnaire C1. The right side of FIG. 7 shows an example of the additional questionnaire C2b. The additional questionnaire C2b includes an additional question area 25 in addition to the initial questionnaire C1.

In the additional question area 25, an additional question 26 and a detailed question 27 are displayed. The additional question 26 is a question that is determined to require reconfirmation based on the estimation result of the symptom estimation unit 204. In the example on the right side of FIG. 7, the additional question 26 includes items to which no answer was given in the initial questionnaire C1 and items that were not included in the initial questionnaire C1. Therefore, the additional question 26 may overlap with items displayed in the initial questionnaire C1. The patient may check a check box in the additional question area 25, or may check a check box 21 of an item corresponding to the initial questionnaire C1. In addition, when a check box is checked in either the additional question area 25 or an item corresponding to the initial questionnaire C1, the other check box may also be checked.

Detailed question 27 is, for example, a question for obtaining more detailed information about a specific symptom. One example is the question described in FIG. 6 that asks more details about the state of chest pain.

When the patient has completed inputting information into the additional medical questionnaire C2 (C2a, C2b), the user terminal 10 determines that inputting information into the medical questionnaire is complete by pressing the input completion button 22, in the same manner as when inputting information into the initial medical questionnaire C1 described above.

In addition, when the number of symptoms entered into the additional questionnaire C2 (or the initial questionnaire C1) reaches a certain number, the input of the questionnaire may be forcibly terminated. Also, the symptoms entered from the user terminal 10 may be transmitted to the server device 2 each time, and the server device 2 may estimate the disease name, and when the number of estimated disease name candidates reaches a certain number, the input of the questionnaire may be forcibly terminated. Conversely, when the estimated disease name candidates are narrowed down to a certain number or less, the input of the questionnaire may be forcibly terminated.

(Explanation of how to display symptoms on the doctor's terminal)
FIG. 8 is a diagram showing an example of a display when the server device 2 causes the doctor terminal 4 to display symptoms obtained by a medical interview.

Based on the answers to the initial questionnaire C1 and the additional questionnaire C2, the medical interview information generation unit 208 organizes the patient's symptoms into a form that is easy for a medical professional (e.g., a doctor) who actually examines the patient to confirm.

The medical interview information generating unit 208 then generates the patient's subjective symptom information 28. The subjective symptom information 28 is displayed so that medical professionals can easily confirm symptoms that characterize the disease when examining the patient. The left side of Figure 8 shows subjective symptom information 28a, which is an example of subjective symptom information 28. Subjective symptom information 28a is an example in which symptoms that tend to characterize the disease are highlighted by being displayed in order of the likelihood of developing.

The left side of Figure 8 specifically shows subjective symptom information 28a that is likely to characterize "lung cancer" as a result of the estimated disease name identification unit 207 re-estimating (or determining) the disease name based on the answers to the additional questionnaire C2. That is, the left side of Figure 8 is an example in which symptoms are sorted and displayed in order of likelihood of occurring in the case of "lung cancer" based on the answers to the initial questionnaire C1 and the additional questionnaire C2. When "lung cancer" develops, the symptoms reported by the patient are least likely to characterize lung cancer in the following order: "bloody sputum," "weight loss," "chest pain," "shortness of breath, palpitations," and "cough."

In addition, if the estimated disease name identification unit 207 identifies a disease name based on the answers to the additional questionnaire C2 and multiple disease names are selected, the subjective symptom information 28a may display the symptoms sorted in order of the most overlapping or least overlapping among the disease name candidates.

Furthermore, the subjective symptom information 28a may be displayed by sorting the symptoms in descending order of the number of times they have been entered into the initial questionnaire C1 and the additional questionnaire C2.

The right side of Figure 8 shows subjective symptom information 28b, which is an example of subjective symptom information 28. Subjective symptom information 28b is an example in which, when the estimated disease name identification unit 207 identifies a disease name based on the answers to the additional questionnaire C2 and finds that the disease is likely to be "lung cancer," symptoms that are likely to characterize "lung cancer" are highlighted by applying hatching. Note that the highlighting method is not limited to hatching, and may also include underlining, changing the text color, changing the font, making the text bold, etc.

(Explanation of medical interview information to be displayed on doctor's terminal)
9 is a diagram showing an example of medical interview information 35 that the server device 2 causes the doctor terminal 4 to display. The medical interview information 35 includes subjective symptom information 28, medical history 29, disease name candidates 30, past medical examination records 31, vital signs 32, and diagnosis records 33.

The subjective symptom information 28 is as described above. In FIG. 9, symptoms that characterize the estimated disease candidate 30 are highlighted and arranged in order of likelihood of onset. Medical history 29 is the patient's past medical history. Note that the medical history can be acquired from an identification number that identifies the individual patient. The disease candidate 30 is a disease candidate estimated from the initial questionnaire C1 and the additional questionnaire C2. FIG. 9 shows that both lung cancer and pneumonia are possible. Note that the disease candidate may display one most likely disease name, or multiple likely disease names.

Past medical records 31 are information recorded in the electronic medical record at the time of past medical examinations (medical history, past medication information, test information, etc.). Vital signs 32 are, for example, medical information (blood test results, etc.) related to a disease name selected from the disease name candidates 30 (lung cancer and pneumonia in the example of FIG. 9). Diagnosis records 33 are the patient's past diagnosis records related to the disease name candidates 30. When an information acquisition button provided near the disease name candidates 30 is pressed, the diagnosis records 33 are acquired from the electronic medical record as information related to the checked disease name of the disease name candidates 30, and are displayed in the diagnosis record 33 column in FIG. 9. The diagnosis record 33 in FIG. 9 shows an example of a medical image of the patient taken in the past.

The doctor examines the patient with reference to the medical interview information 35 and diagnoses the patient's illness. If the name of the illness estimated as a candidate disease name 30 in the medical interview information 35 differs from the actual name of the illness determined as a result of the diagnosis, the medical interview system 1 determines that it has made an incorrect estimation.

At this time, the server device 2 associates the symptom information entered in the initial questionnaire C1 and the additional questionnaire C2 with the disease name diagnosed by the doctor. Then, the associated data is used as new learning data to update the disease name estimation model M1 and the symptom estimation model M2. At this time, it is desirable to train the disease name estimation model M1 and the symptom estimation model M2 by giving a higher contribution to the learning data indicating the relationship between the patient's symptom information obtained in the current diagnosis and the disease name diagnosed by the doctor than other learning data. Furthermore, learning may be performed after making the contribution of the learning data of the disease name group in which the medical interview system 1 has made an incorrect estimation to the disease name estimation model M1 and the symptom estimation model M2 larger than the learning data of the other disease name groups.

(Explanation of the process flow performed by the medical interview system)
10 is a flowchart showing an example of the flow of processing performed by the medical interview system 1. The flow of processing performed by the user terminal 10, the flow of processing performed by the server device 2, and the flow of processing performed by the doctor terminal 4 will be described below in that order.

(Explanation of the process flow performed by the user terminal)
The patient starts an application for executing a medical interview on his/her user terminal 10, and requests a medical interview sheet from the server device 2 (step S10). When requesting the medical interview sheet, patient information such as the patient's name may be input.

The control unit 101 of the user terminal 10 determines whether the communication I/F 103 has received the initial questionnaire C1 from the server device 2 (step S11). If it is determined that the initial questionnaire C1 has been received (step S11: Yes), the process proceeds to step S12. On the other hand, if it is not determined that the initial questionnaire C1 has been received (step S11: No), step S11 is repeated.

The control unit 101 accepts input from the patient into the initial questionnaire C1 (step S12).

The control unit 101 cooperates with the operation unit 106 to determine whether the input completion button 22 has been pressed (step S13). If it is determined that the input completion button 22 has been pressed (step S13: Yes), the process proceeds to step S14. On the other hand, if it is not determined that the input completion button 22 has been pressed (step S13: No), the process returns to step S12.

If it is determined in step S13 that the input completion button 22 has been pressed, then in step S14, the control unit 101 causes the communication I/F 103 to send the initial questionnaire C1 to the server device 2.

Then, the control unit 101 determines whether the communication I/F 103 has received the additional questionnaire C2 from the server device 2 (step S15). If it is determined that the additional questionnaire C2 has been received (step S15: Yes), the control unit 101 proceeds to step S16. On the other hand, if it is determined that the additional questionnaire C2 has not been received (step S15: No), the control unit 101 repeats step S15.

The control unit 101 accepts input from the patient to the additional medical questionnaire C2 (step S16). The control unit 101 cooperates with the operation unit 106 to determine whether the input completion button 22 has been pressed (step S17). If it is determined that the input completion button 22 has been pressed (step S17: Yes), the process proceeds to step S18. On the other hand, if it is not determined that the input completion button 22 has been pressed (step S17: No), the process returns to step S16.

If it is determined in step S17 that the input completion button 22 has been pressed, then in step S18, the control unit 101 causes the communication I/F 103 to send the additional questionnaire C2 to the server device 2. Thereafter, the user terminal 10 ends the process of FIG. 10.

(Explanation of the flow of processing performed by the server device)
The first reception unit 202 determines whether or not there is a request for a medical questionnaire from the user terminal 10 (step S20). Note that, for example, in step S20, the server device 2 can also acquire patient information from the user terminal 10 together with the request for the medical questionnaire. If it is determined that there is a request for a medical questionnaire from the user terminal 10 (step S20: Yes), the process proceeds to step S21. On the other hand, if it is not determined that there is a request for a medical questionnaire from the user terminal 10 (step S20: No), step S20 is repeated.

If it is determined in step S20 that a medical questionnaire has been requested from the user terminal 10, then in step S21, the first reception unit 202 causes the communication control unit 210 to send an initial medical questionnaire C1 to the user terminal 10.

Next, the first reception unit 202 determines whether the initial questionnaire C1 has been received from the user terminal 10 (step S22). If it is determined that the initial questionnaire C1 has been received from the user terminal 10 (step S22: Yes), the process proceeds to step S23. On the other hand, if it is not determined that the initial questionnaire C1 has been received from the user terminal 10 (step S22: No), the process repeats step S22.

The disease name estimation unit 203 estimates the disease name based on the contents entered in the initial questionnaire C1 (step S23).

The symptom estimation unit 204 estimates symptoms based on the disease name estimated in step S23 (step S24).

The questionnaire regeneration unit 205 generates an additional questionnaire C2 based on the symptoms estimated in step S24 (step S25).

The questionnaire regeneration unit 205 causes the communication control unit 210 to send the additional questionnaire C2 to the user terminal 10 (step S26).

The estimated disease name identification unit 207 determines whether the additional questionnaire C2 has been received from the user terminal 10 (step S27). If it is determined that the additional questionnaire C2 has been received from the user terminal 10 (step S27: Yes), the process proceeds to step S28. On the other hand, if it is not determined that the additional questionnaire C2 has been received from the user terminal 10 (step S27: No), step S27 is repeated.

If it is determined in step S27 that an additional medical questionnaire C2 has been received from the user terminal 10, then in step S28, the estimated disease name determination unit 207 re-estimates the patient's disease name.

Next, the medical interview information generation unit 208 generates medical interview information 35 (step S29).

Next, the medical interview information generating unit 208 causes the communication control unit 210 to transmit the medical interview information 35 to the doctor terminal 4 (step S30). After that, the server device 2 ends the process of FIG. 10.

(Explanation of the processing flow performed by the doctor terminal)
The control unit 101 of the doctor terminal 4 determines whether the medical interview information 35 has been received (step S40). If it is determined that the medical interview information 35 has been received (step S40: Yes), the process proceeds to step S41. On the other hand, if it is determined that the medical interview information 35 has been received (step S40: No), step S40 is repeated.

When it is determined in step S40 that the medical interview information 35 has been received, the control unit 101 of the doctor terminal 4 displays the medical interview information 35 on the display unit 107 (step S41). Then, the doctor terminal 4 ends the process of FIG. 10.

As described above, in the server device 2 (medical interview device) of the embodiment, the first reception unit 202 receives a response to an initial questionnaire C1 asking about symptoms, and the disease name estimation unit 203 estimates at least one disease name based on the symptoms received by the first reception unit 202. The symptom estimation unit 204 estimates the symptoms of the disease based on the disease name estimated by the disease name estimation unit 203, and the questionnaire regeneration unit 205 (generation unit) generates an additional questionnaire C2 asking about symptoms to be reconfirmed according to the symptoms estimated by the symptom estimation unit 204. Then, the second reception unit 206 receives a response to the additional questionnaire C2 generated by the questionnaire regeneration unit 205, and the estimated disease name identification unit 207 identifies the estimated disease name estimated by the disease name estimation unit 203 based on the symptoms received by the second reception unit 206. Therefore, the patient's condition can be efficiently collected by conducting another interview using the additional interview form C2 generated based on the patient's response to the initial interview form C1. Furthermore, the patient can fill out the initial interview form C1 and the additional interview form C2 before arriving at the hospital, which shortens the waiting time at the hospital. Furthermore, the effort required for medical staff to collect symptom information from patients at the hospital can be reduced.

The medical interview information generating unit 208 may also guide the user to a medical department based on the results of the additional medical interview form C2. This can be achieved by registering the medical department associated with the disease name in the disease name and symptoms database DB described above. Furthermore, if the patient's symptoms are deemed highly urgent based on the results of the additional medical interview form C2, the medical interview information generating unit 208 may instruct the user terminal 10 to call an ambulance.

In addition, in the server device 2 (medical interview device) of the embodiment, the disease name estimation unit 203 estimates the estimated disease name using a disease name estimation model M1 that stores associations between symptoms and disease names that cause the symptoms. Therefore, the disease name can be estimated easily and efficiently. In addition, by updating the disease name estimation model M1, it is possible to easily respond when a new disease name is registered.

In addition, in the server device 2 (interview device) of the embodiment, the symptom estimation unit 204 estimates symptoms using a symptom estimation model M2 that stores associations between disease names and symptoms caused by the disease names. Therefore, it is possible to easily and efficiently estimate symptoms. In addition, by updating the symptom estimation model M2, it is possible to easily respond when new symptoms are registered.

In addition, in the server device 2 (medical interview device) of the embodiment, the symptom estimation model M2 is information that stores an association between a disease name, symptoms caused by the disease name, and information that indicates the likelihood of the association between the disease name and the symptoms, and when the disease name estimation unit 203 estimates multiple estimated disease names, the questionnaire regeneration unit 205 (generation unit) generates an additional questionnaire that includes a question for the symptom that is estimated to be most likely to have occurred based on the symptom estimation model M2 from among the symptoms estimated by the symptom estimation unit 204 from each of the multiple estimated disease names. Therefore, it is possible to narrow down disease name candidates from the answer to the question in the additional questionnaire C2.

In addition, in the server device 2 (medical interview device) of the embodiment, when the disease name estimation unit 203 estimates multiple estimated disease names, the questionnaire regeneration unit 205 (generation unit) generates an additional questionnaire C2 including questions for symptoms that overlap with the multiple estimated disease names among the symptoms estimated by the symptom estimation unit 204 from each of the multiple estimated disease names. Therefore, it is possible to select a likely disease name candidate from the answer to the question in the additional questionnaire C2.

In addition, in the server device 2 (medical interview device) of the embodiment, when the disease name estimation unit 203 estimates multiple estimated disease names, the questionnaire regeneration unit 205 (generation unit) generates an additional questionnaire C2 including questions for symptoms that have little overlap among the multiple estimated disease names, among the symptoms estimated by the symptom estimation unit 204 from each of the multiple estimated disease names. Therefore, it is possible to narrow down disease name candidates from the answers to the questions in the additional questionnaire C2.

In addition, in the server device 2 (medical interview device) of the embodiment, the medical interview sheet regeneration unit 205 (generation unit) generates an additional medical interview sheet C2 in which the symptoms estimated by the symptom estimation unit 204 are highlighted. This makes it easier for the patient to pay attention when filling out the additional medical interview sheet C2.

In addition, in the server device 2 (medical interview device) of the embodiment, the medical interview information generating unit 208 generates medical interview information 35 including the answers to the initial medical interview form C1 and the additional medical interview form C2, and the disease name identified by the estimated disease name identifying unit 207. Then, the display control unit 209 causes the medical interview information 35 to be displayed on the doctor terminal 4 to which the doctor inputs the medical record information. Therefore, the doctor who examines the patient can easily check the results of the medical interview.

In addition, in the server device 2 (medical interview device) of the embodiment, the medical interview information 35 highlights symptoms that characterize the disease name identified by the estimated disease name identification unit 207. Therefore, the doctor who performs the examination can more easily check the results of the patient's medical interview, thereby improving the efficiency of the examination.

In addition, the medical interview system 1 of the embodiment collects the symptoms of the patient by cooperation between the server device 2 and the user terminal 10. Specifically, in the server device 2, the first reception unit 202 receives a response to an initial medical interview sheet C1 asking about symptoms, and the disease name estimation unit 203 estimates at least one estimated disease name based on the symptoms received by the first reception unit 202. The symptom estimation unit 204 estimates the symptoms of the disease name based on the estimated disease name estimated by the disease name estimation unit 203, and the medical interview sheet regeneration unit 205 (generation unit) generates an additional medical interview sheet C2 that asks about symptoms to be reconfirmed according to the symptoms estimated by the symptom estimation unit 204. Then, the second reception unit 206 receives a response to the additional medical interview sheet C2 generated by the medical interview sheet regeneration unit 205, and the estimated disease name identification unit 207 identifies the estimated disease name estimated by the disease name estimation unit 203 based on the symptoms received by the second reception unit 206. At this time, in the user terminal 10, the communication I/F 103 (receiving unit) receives the initial questionnaire C1 and the additional questionnaire C2 from the server device 2. Then, the display unit 107 displays the initial questionnaire C1 and the additional questionnaire C2, and the control unit 101 (reception unit) accepts the responses to the initial questionnaire C1 and the additional questionnaire C2. The communication I/F 103 (transmitting unit) transmits the responses to the initial questionnaire C1 and the additional questionnaire C2 to the server device 2. Therefore, the patient's medical condition can be collected efficiently.

Second Embodiment
Next, a medical interview device and a medical interview system according to a second embodiment will be described. Generally, an electronic medical record contains information such as a patient's medical history, current medical history, and medications that the patient regularly takes. The medical interview device and the medical interview system according to the second embodiment generate a more accurate additional medical interview form based on the information in the electronic medical record and the name of a disease estimated from the answers to the initial medical interview form.

Figure 11 is a diagram for explaining the medical interview system 5 according to the second embodiment. Compared to the medical interview system 1 according to the first embodiment shown in Figure 1, the medical interview system 5 differs in that it further includes a clinical information database 3.

As shown in FIG. 11, the clinical information database 3 includes an electronic medical record database 3a and an image database 3b. Here, clinical information refers to information recorded in the electronic medical record, information associated with the electronic medical record (e.g., nursing records, meal information, medical accounting information, etc.), image data, etc., and is managed for each patient.

The electronic medical record database 3a is a database that stores and manages electronic medical records and information associated with the electronic medical records for each patient. The electronic medical record database 3a in this embodiment includes a database serving as an HIS.

The image database 3b is a database that stores and manages image data acquired by various modalities and information associated with the image data for each patient. The image database 3b in this embodiment includes databases such as a PACS and a RIS.

The server device 2 acquires information recorded in the electronic medical record of the patient from the electronic medical record database 3a via the network based on the patient information received from the user terminal 10. Specifically, the symptom estimation unit 204 of the server device 2 acquires information recorded in the electronic medical record of the patient from the electronic medical record database 3a via the network based on the patient information received from the user terminal 10.

Here, the information recorded in the electronic medical record of the patient acquired by the symptom estimation unit 204 of the server device 2 is the patient's medical history, current medical history, and regularly taken medications recorded in the electronic medical record, as well as information regarding medical interviews conducted by a doctor in the past about the patient (contents of the medical interview and the patient's responses thereto), etc.

The symptom estimation unit 204 of the server device 2 uses the symptom estimation model M2 to estimate at least one symptom based on the disease name estimated by the disease name estimation unit 203 and the information recorded in the electronic medical record of the patient obtained from the electronic medical record database 3a.

Such a symptom estimation model M2 is a program that realizes an AI model such as a neural network. That is, the symptom estimation model M2 is, for example, a trained model (such as a deep neural network) that inputs at least one disease name and information recorded in the patient's electronic medical record, and outputs one or more symptoms. A program that realizes such a symptom estimation model M2 can be constructed by machine learning using multiple training data consisting of a combination of the disease name diagnosed by a doctor, information recorded in the electronic medical record, and at least one symptom that is generally the main complaint of patients with that disease. Note that the symptom estimation unit 204 of the server device 2 may further estimate at least one symptom based on the initial questionnaire, as necessary.

The questionnaire regeneration unit 205 of the server device 2 generates an additional questionnaire C2 based on at least one symptom estimated by the symptom estimation unit 204. At this time, the questionnaire regeneration unit 205 of the server device 2 can also generate an additional questionnaire C2 in which information included in the additional questionnaire C2 that overlaps with the initial questionnaire C1 is separated by referring to the initial questionnaire C1.

In this embodiment, the symptom estimation unit 204 and the questionnaire regeneration unit 205 have been described as being separate components. In contrast, the symptom estimation unit 204 and the questionnaire regeneration unit 205 may be realized as a single trained model (such as a deep neural network). Such a trained model can be constructed by machine learning using multiple training data in which the name of the disease diagnosed by the doctor, the information recorded in the electronic medical record, and the initial questionnaire C1 are used as input data, and at least one symptom that does not overlap with the initial questionnaire C1 is used as output data (teacher data).

As described above, the server device 2 as the interview device according to this embodiment estimates at least one symptom in the symptom estimation unit 204 based on the disease name estimated by the disease name estimation unit 203 and the information recorded in the patient's electronic medical record obtained from the electronic medical record database 3a. Furthermore, the server device 2 generates an additional interview sheet C2 in the interview sheet regeneration unit 205 based on the at least one symptom estimated by the symptom estimation unit 204. Therefore, it is possible to generate an even more accurate additional interview sheet C2 that also includes the information recorded in the patient's electronic medical record. Furthermore, since a high-quality additional interview sheet C2 based on the patient's electronic medical record is used, it is possible to realize even more accurate disease name estimation.

(Variation 1)
In the above-described second embodiment, when the symptom estimation unit 204 of the server device 2 estimates at least one symptom, an example is shown in which information recorded in the electronic medical record of the patient obtained from the electronic medical record database 3a is used. In contrast, for example, when the disease name estimation unit 203 estimates the name of a disease or when the estimated disease name specification unit 207 specifies the name of a disease, the information recorded in the electronic medical record of the patient obtained from the electronic medical record database 3a can be used as necessary.

(Variation 2)
In the above-described embodiments, the server device 2 is installed in a hospital. However, the server device 2 may be installed anywhere, not limited to a hospital, as long as the server device 2 is in an environment where it can communicate with the user terminal 10 and the doctor terminal 4 via a network. For example, the server device 2 may be installed as a server on a cloud.

(Variation 3)
In the above-described embodiments, the server device 2 of the medical interview system 1 functions as the medical interview device. However, the medical interview system 1 may be realized by installing an application that executes the process executed in the server device 2 on the user terminal 10.

The term "processor" used in the above description means, for example, a circuit such as a CPU (central processing unit), a GPU (Graphics Processing Unit), or an application specific integrated circuit (ASIC), a programmable logic device (for example, a simple programmable logic device (SPLD), a complex programmable logic device (CPLD), and a field programmable gate array (FPGA)). When the processor is, for example, a CPU, the processor realizes a function by reading and executing a program stored in a memory circuit. On the other hand, when the processor is, for example, an ASIC, instead of storing a program in a memory circuit, the function is directly incorporated as a logic circuit in the circuit of the processor. Note that each processor in this embodiment is not limited to being configured as a single circuit for each processor, but may be configured as a single processor by combining multiple independent circuits to realize its function. Furthermore, the multiple components in FIG. 3 may be integrated into a single processor to realize its function.

Although several embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the gist of the invention. These embodiments and their modifications are within the scope of the invention and its equivalents as set forth in the claims, as well as the scope and gist of the invention.

1. Medical interview system 2. Server device (medical interview device)
4 Doctor terminal 10 User terminal (terminal device)
101 Control unit (reception unit)
103 Communication I/F (receiving unit, transmitting unit)
107 Display unit 202 First reception unit 203 Disease name estimation unit 204 Symptom estimation unit 205 Questionnaire regeneration unit (generation unit)
206 Second reception unit 207 Estimated disease name specification unit 208 Interview information generation unit C1 Initial interview sheet C2 (C2a, C2b) Additional interview sheet DB Disease name and symptom database M1 Disease name estimation model M2 Symptom estimation model

Claims (13)

a first reception unit that receives an initial response to an initial questionnaire asking about symptoms;
a presentation unit that presents an additional questionnaire inquiring about symptoms related to the initial answer based on the initial answer;
a second reception unit that receives additional responses to the additional questionnaire;
a first determination unit that determines at least one first disease name based on the initial answer and the additional answer;
a second determination unit that determines at least one second disease name based on the initial response;
a first generation unit that generates the additional questionnaire based on the determined at least one second disease name;
Equipped with
The first generation unit generates the additional medical questionnaire using a trained model that receives the determined at least one second disease name as an input and outputs a medical questionnaire .
Interview device. a first reception unit that receives an initial response to an initial questionnaire asking about symptoms;
a presentation unit that presents an additional questionnaire inquiring about symptoms related to the initial answer based on the initial answer;
a second reception unit that receives additional responses to the additional questionnaire;
a first determination unit that determines at least one first disease name based on the initial answer and the additional answer;
a second determination unit that determines at least one second disease name based on the initial response;
a first generation unit that generates the additional questionnaire based on the determined at least one second disease name;
Equipped with
The first generation unit generates the additional medical questionnaire using a trained model that receives the determined at least one second disease name and information recorded in the patient's electronic medical record as input and outputs a medical questionnaire.
Interview device. a first reception unit that receives an initial response to an initial questionnaire asking about symptoms;
a presentation unit that presents an additional questionnaire inquiring about symptoms related to the initial answer based on the initial answer;
a second reception unit that receives additional responses to the additional questionnaire;
a first determination unit that determines at least one first disease name based on the initial answer and the additional answer;
a second determination unit that determines at least one second disease name based on the initial response;
a first generation unit that generates the additional questionnaire based on the determined at least one second disease name;
Equipped with
the first generation unit generates the additional questionnaire including questions for symptoms that have little overlap among the plurality of second disease names among symptoms related to each of the plurality of second disease names that have been determined ;
Interview device. a first reception unit that receives an initial response to an initial questionnaire asking about symptoms;
a presentation unit that presents an additional questionnaire inquiring about symptoms related to the initial answer based on the initial answer;
a second reception unit that receives additional responses to the additional questionnaire;
a first determination unit that determines at least one first disease name based on the initial answer and the additional answer;
a second determination unit that determines at least one second disease name based on the initial response;
a first generation unit that generates the additional questionnaire based on the determined at least one second disease name;
Equipped with
The first generation unit generates an additional questionnaire including questions for symptoms that overlap with the plurality of second disease names among symptoms related to each of the plurality of second disease names that have been determined.
Interview device. The medical interview device according to claim 1 , wherein the first determination unit determines at least one of the first disease names by using a table in which symptoms are associated with disease names that cause the symptoms. The medical interview device according to claim 1 , wherein the first determination unit determines at least one of the first disease names using a trained model that receives at least one symptom as an input and outputs at least one disease name. The first determination unit determines at least one of the first disease names using a trained model that receives at least one symptom and information recorded in a patient 's electronic medical record as input and outputs at least one disease name.
The interview device according to claim 1 . The medical interview device according to claim 1 , wherein the second determination unit determines at least one of the second disease names by using a table in which symptoms are associated with disease names that cause the symptoms. The interview device according to claim 1 , wherein the second determination unit determines at least one second disease name using a trained model that receives at least one symptom as input and outputs at least one second disease name. The medical interview device according to any one of claims 1 to 6, wherein the second determination unit determines at least one second disease name using a trained model that receives as input at least one symptom and information recorded in the patient 's electronic medical record and outputs at least one second disease name. The medical interview device according to claim 1 , wherein the first generating unit generates the additional medical interview sheet by performing a calculation using a table in which a plurality of disease names are associated with a plurality of medical interviews. The interview device according to claim 2 , wherein the first generation unit generates the additional interview sheet in which the estimated symptoms are emphasized. a second generating unit configured to generate medical interview information including answers to the initial medical interview sheet and the additional medical interview sheet and the first disease name;
a communication control unit that transmits the interview information to a doctor terminal where a doctor inputs medical record information;
The interview device according to claim 1 , further comprising:

Images