Medical information processing method, medical information processing device, and program

Abstract

The goal is to reduce the burden on patients and medical staff during diagnosis. [Solution] The medical information processing method of the embodiment includes: generating an event log, which is a log of events that occurred in the patient, arranged in chronological order, based on the patient's personal data; estimating a turning event, which is an event that was a turning point for the patient, based on the event log; obtaining the patient's answers to a plurality of items related to the medical interview; and outputting the answers from the patient's answers that correspond to the turning event as attention information via an output interface.

Description

Technical Field

[0001] The embodiments disclosed in this specification and the drawings relate to a medical information processing method, a medical information processing apparatus, and a program.

Background Art

[0002] Mental illnesses such as schizophrenia, depression, and bipolar disorder are difficult to diagnose solely based on objective data and images. Although electroencephalogram tests and CT (Computed Tomography) tests may be performed, these are carried out for the purpose of differentiating from other diseases (so-called differential diagnosis). Therefore, the interview by having the patient describe their experiences in words is the most important method for diagnosing mental illnesses.

[0003] A general interview focuses on the patient's physical symptoms and health status. In contrast, an interview for mental disorders focuses on the patient's emotions, thoughts, and behavior patterns. It is important to approach from a broader perspective, such as the patient's daily life, human relationships, stress level, etc., and find biases in the patient's self-awareness and self-evaluation.

[0004] As a prior art, a WEB (World Wide Web) interview service using the Internet is known. In such an interview service, for interview items, the patient himself / herself can answer (input or select) to fill in the interview items (that is, the patient's answers to the interview items can be digitized). Also, in an interview using AI (Artificial Intelligence), it is possible to推测 the suspected symptoms and diseases according to the input content.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Summary of the Invention

[0006] The problem that the embodiments disclosed in this specification and drawings aim to solve is to reduce the burden on patients and medical staff during diagnosis. However, the problems that the embodiments disclosed in this specification and drawings aim to solve are not limited to the above problem. Problems corresponding to the effects of each configuration shown in the embodiments described later can also be positioned as other problems. [Means for solving the problem]

[0007] This is a medical information processing method using a computer according to an embodiment. The medical information processing method includes: generating an event log, which is a log of events that occurred in the patient in chronological order, based on the patient's personal data; estimating a turning event, which is an event that was a turning point for the patient, based on the event log; obtaining the patient's answers to a plurality of items related to the medical interview; and outputting the answers from the patient that correspond to the turning event as attention information via an output interface. [Brief explanation of the drawing]

[0008] [Figure 1] A diagram showing an example configuration of the medical information processing system 1 in the first embodiment. [Figure 2] A diagram showing an example configuration of the user interface 10 in the first embodiment. [Figure 3] A diagram showing an example configuration of the medical information processing device 100 in the first embodiment. [Figure 4] A flowchart showing the sequence of processes of the processing circuit 120 in the first embodiment. [Figure 5] A diagram illustrating an example of questionnaire items extracted from a questionnaire form. [Figure 6] A diagram illustrating an example of an event extracted from a patient's personal data. [Figure 7] A diagram illustrating an example of an event managed as a record. [Figure 8] A diagram illustrating an example of a primary event log. [Figure 9] A diagram illustrating an example of a method for estimating turning events. [Figure 10] A diagram illustrating an example of how estimated answers are displayed. [Figure 11] A diagram illustrating other examples of how estimated answers are displayed. [Figure 12] A diagram illustrating an example of how estimated responses and turning events are displayed. [Figure 13] A diagram illustrating other examples of events managed as records. [Figure 14] A diagram illustrating a method for estimating turning events from outward appearances. [Figure 15] A diagram showing an example configuration of the medical information processing device 100 in the third embodiment. [Figure 16] A flowchart illustrating the sequence of processes in the processing circuit 120 in the third embodiment. [Figure 17] A schematic diagram showing the sequence of processing steps of the processing circuit 120 in the third embodiment. [Modes for carrying out the invention]

[0009] The following describes the embodiment of the medical information processing method, medical information processing device, and program with reference to the drawings.

[0010] (First Embodiment) [Configuration of the medical information processing system] Figure 1 is a diagram showing an example configuration of the medical information processing system 1 in the first embodiment. The medical information processing system 1 includes, for example, a user interface 10 and a medical information processing device 100. The user interface 10 and the medical information processing device 100 are connected to each other via a communication network NW.

[0011] The communication network NW may mean the entire information communication network using telecommunication technologies. For example, the communication network NW includes wireless / wired LANs such as hospital backbone LAN (Local Area Network), the Internet, as well as telephone communication line networks, optical fiber communication networks, cable communication networks, satellite communication networks, and the like.

[0012] The user interface 10 is used by patients and medical staff. In this embodiment, the patient is typically a patient suffering from a mental illness, but is not limited thereto, and may be a patient suffering from other diseases or injuries.

[0013] For example, the user interface 10 is a touch interface or a voice user interface, and more specifically, a terminal device such as a personal computer, a tablet terminal, or a mobile phone. The medical staff is typically a doctor, but may also be a nurse or other person involved in medical care. For example, the patient may touch-input or voice-input his or her answers to the medical interview into the user interface 10. Also, the medical staff may verbally conduct a medical interview with the patient, hear the answers from the patient, and input the hearing results into the user interface 10.

[0014] In this embodiment, "medical treatment" may include not only treatments such as surgeries and medication, but also examinations and any other medical acts before, during, or after treatment.

[0015] The user interface 10 transmits information input by patients and medical staff to the medical information processing device 100 via the communication network NW, or receives information from the medical information processing device 100.

[0016] The medical information processing device 100 receives information from the user interface 10 via the communication network NW and processes the received information. The medical information processing device 100 then transmits the processed information to the user interface 10 via the communication network NW. In addition to transmitting the processed information to the user interface 100, the medical information processing device 100 may also transmit it to a dedicated terminal for medical staff installed within the hospital.

[0017] The medical information processing device 100 may be a single device, or it may be a system in which multiple devices connected via a communication network NW work together. In other words, the medical information processing device 100 may be implemented by multiple computers (processors) included in a distributed computing system or a cloud computing system. Furthermore, the medical information processing device 100 does not necessarily have to be a separate device from the user interface 10, and may be an integrated device with the user interface 10.

[0018] [User Interface Configuration] Figure 2 is a diagram showing an example configuration of the user interface 10 in the first embodiment. The user interface 10 includes, for example, a communication interface 11, an input interface 12, an output interface 13, a memory 14, and a processing circuit 20.

[0019] The communication interface 11 communicates with the medical information processing device 100 and other devices via the communication network NW. The communication interface 11 includes, for example, a NIC (Network Interface Card) and an antenna for wireless communication.

[0020] The input interface 12 receives various input operations from the operator (e.g., a patient), converts the received input operations into electrical signals, and outputs them to the processing circuit 20. For example, the input interface 12 includes a mouse, keyboard, trackball, switch, button, joystick, touch panel, etc. The input interface 12 may also be a user interface that accepts audio input, such as from a microphone. If the input interface 12 is a touch panel, the input interface 12 may also incorporate the display function of the display 13a included in the output interface 13, which will be described later.

[0021] In this specification, the input interface 12 is not limited to those equipped with physical operating components such as a mouse or keyboard. For example, an electrical signal processing circuit that receives an electrical signal corresponding to an input operation from an external input device provided separately from the device and outputs this electrical signal to a control circuit is also included as an example of the input interface 12.

[0022] The output interface 13 includes, for example, a display 13a and a speaker 13b.

[0023] The display 13a displays various types of information. For example, the display 13a displays images generated by the processing circuit 20, or a GUI (Graphical User Interface) for receiving various input operations from the operator. For example, the display 13a may be an LCD (Liquid Crystal Display), a CRT (Cathode Ray Tube) display, or an organic EL (Electro Luminescence) display.

[0024] Speaker 13b outputs the information input from processing circuit 20 as sound.

[0025] Memory 14 can be implemented by, for example, semiconductor memory elements such as RAM (Random Access Memory) or flash memory, a hard disk, or an optical disc. These non-transient storage media may also be implemented by other storage devices connected via a communication network NW, such as NAS (Network Attached Storage) or external storage server devices. Memory 14 may also include non-transient storage media such as ROM (Read Only Memory) or registers.

[0026] The processing circuit 20 includes, for example, an acquisition function 21, an output control function 22, and a communication control function 23. The processing circuit 20 realizes these functions, for example, by a hardware processor (computer) executing a program stored in the memory 14 (storage circuit).

[0027] In the processing circuit 20, the hardware processor refers to circuits such as a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), an Application Specific Integrated Circuit (ASIC), or a programmable logic device (e.g., a Simple Programmable Logic Device (SPLD) or a Complex Programmable Logic Device (CPLD), or a Field Programmable Gate Array (FPGA)). Instead of storing the program in memory 14, the program may be directly embedded within the hardware processor's circuitry. In this case, the hardware processor functions by reading and executing the program embedded within the circuitry. The program may be stored in memory 14 beforehand, or it may be stored on a non-temporary storage medium such as a DVD or CD-ROM, and installed from the non-temporary storage medium to memory 14 when the non-temporary storage medium is inserted into the drive device (not shown) of the user interface 10. A hardware processor is not limited to being a single circuit; it may also be composed of multiple independent circuits combined to perform various functions. Alternatively, multiple components may be integrated into a single hardware processor to perform various functions.

[0028] The acquisition function 21 acquires input information via the input interface 12 or acquires information from the medical information processing device 100 via the communication interface 11.

[0029] The output control function 22 displays the information acquired by the acquisition function 21 as an image or video on the display 13a, or outputs it as sound from the speaker 13b.

[0030] The communication control function 23 transmits the information input to the input interface 12 to the medical information processing device 100 via the communication interface 11.

[0031] [Configuration of medical information processing device] Figure 3 is a diagram showing an example configuration of the medical information processing device 100 in the first embodiment. The medical information processing device 100 includes, for example, a communication interface 111, an input interface 112, an output interface 113, a memory 114, and a processing circuit 120.

[0032] The communication interface 111 communicates with the user interface 10 and other interfaces via the communication network NW. The communication interface 111 includes, for example, a NIC.

[0033] The input interface 112 receives various input operations from the operator, converts the received input operations into electrical signals, and outputs them to the processing circuit 120. For example, the input interface 112 includes a mouse, keyboard, trackball, switch, button, joystick, touch panel, etc. The input interface 112 may also be a user interface that accepts audio input, such as from a microphone. If the input interface 112 is a touch panel, the input interface 112 may also incorporate the display function of the display 113a included in the output interface 113, which will be described later.

[0034] In this specification, the input interface 112 is not limited to those equipped with physical operating components such as a mouse or keyboard. For example, an electrical signal processing circuit that receives an electrical signal corresponding to an input operation from an external input device provided separately from the device and outputs this electrical signal to a control circuit is also included as an example of the input interface 112.

[0035] The output interface 113 includes, for example, a display 113a and a speaker 113b.

[0036] The display 113a displays various types of information. For example, the display 113a displays images generated by the processing circuit 120, or a GUI for receiving various input operations from the operator. For example, the display 113a may be an LCD, a CRT display, an organic EL display, etc.

[0037] Speaker 113b outputs the information input from processing circuit 120 as sound.

[0038] Memory 114 can be implemented by, for example, semiconductor memory elements such as RAM or flash memory, a hard disk, or an optical disc. These non-transient storage media may also be implemented by other storage devices connected via a communication network NW, such as a NAS or an external storage server device. Memory 114 may also include non-transient storage media such as ROM or registers.

[0039] Memory 114 stores programs executed by the hardware processor, as well as patient personal data and model information.

[0040] Patient personal data may include, for example, images, videos, text, location information, and medical and health information. These images, videos, and text may include, for example, data posted on social networking services (SNS). Medical and health information may include the medical questionnaire form described later.

[0041] Model information refers to information (a program or data structure) that defines the pre-trained model MDL. The pre-trained model MDL may be implemented by large-scale language models (LLMs) such as GPT (Generative Pre-trained Transformer), PaLM (Pathways Language Model), LLaMA (Large Language Model Meta AI), or Claude. Alternatively, the pre-trained model MDL may be implemented by other generative AIs such as VAE (Variational Auto-Encoder) or GAN (Generative Adversarial Networks). Details about the pre-trained model MDL will be discussed later.

[0042] The processing circuit 120 includes, for example, an acquisition function 121, a medical interview item extraction function 122, an event extraction function 123, an event log generation function 124, a turning event estimation function 125, a medical interview answer estimation function 126, an output control function 127, and a communication control function 128. The event log generation function 124 is an example of a "generation unit," the turning event estimation function 125 is an example of a "first estimation unit," and the medical interview answer estimation function 126 is an example of an "acquisition unit." The output control function 127 is an example of an "output control unit," and the communication control function 128 is another example of an "output control unit."

[0043] The processing circuit 120 realizes these functions, for example, by having a hardware processor (computer) execute a program stored in the memory 114 (storage circuit).

[0044] In the processing circuit 120, the hardware processor refers to, for example, a CPU, GPU, application-specific integrated circuit, or programmable logic device (e.g., a simple programmable logic device or a complex programmable logic device, a field-programmable gate array, etc.). Instead of storing the program in memory 114, the program may be directly incorporated into the hardware processor's circuit. In this case, the hardware processor realizes its function by reading and executing the program incorporated into the circuit. The program may be stored in memory 114 beforehand, or it may be stored in a non-temporary storage medium such as a DVD or CD-ROM, and installed from the non-temporary storage medium to memory 114 when the non-temporary storage medium is mounted in the drive device (not shown) of the medical information processing device 100. The hardware processor is not limited to being configured as a single circuit; it may be configured as a single hardware processor by combining multiple independent circuits to realize each function. Alternatively, multiple components may be integrated into a single hardware processor to realize each function.

[0045] [Processing flow of medical information processing equipment] The following describes a series of processes performed by the processing circuit 120 of the medical information processing device 100, following the flowchart. Figure 4 is a flowchart showing the flow of a series of processes performed by the processing circuit 120 in the first embodiment.

[0046] First, the acquisition function 121 acquires the medical questionnaire forms of one or more patients (step S100).

[0047] A medical questionnaire is an electronic document (i.e., a medical history form) that medical staff (e.g., doctors) use to confirm the patient's perceived symptoms and health status, especially during initial consultations. For example, a medical questionnaire form contains several items (hereinafter referred to as "questionnaire items") that the patient is asked as part of the questionnaire, and each questionnaire item is associated with the symptoms and health status that the patient has answered. Among the multiple questionnaire items, there may be some that the patient does not answer (so-called blanks).

[0048] The medical questionnaire form may be displayed on the display 13a of the user interface 10, or it may be printed on paper and distributed to the patient. If the patient fills out the paper questionnaire, the medical staff of the medical institution may input the entered information into the user interface 10. In this case, OCR (Optical Character Recognition / Reader) may be used. The contents of the questionnaire form may be output as audio from the speaker 113b of the user interface 10, or the medical staff of the medical institution may read it aloud. If the patient speaks their answers to the user interface 10, the user interface 10 may acquire the answers spoken by the patient via the microphone. Alternatively, or in addition to this, the medical staff may interview the patient to obtain the answers spoken by them. After interviewing the patient, the medical staff may input the results of the interview into the user interface 10. Furthermore, the questionnaire items included in the questionnaire form do not need to be predetermined and can be freely determined by the medical staff at the time of the examination. The answers to the medical interview are also known as the chief complaint in medical terms.

[0049] For example, if personal data including a medical questionnaire form is stored in memory 114, the acquisition function 121 may acquire the medical questionnaire form from memory 114.

[0050] Alternatively, for example, the acquisition function 121 may access the user interface 10 via the communication interface 111 and acquire the medical questionnaire form from the user interface 10.

[0051] Furthermore, for example, if a medical questionnaire form is entered into the input interface 112, the acquisition function 121 may acquire the medical questionnaire form from the input interface 112.

[0052] Furthermore, for example, if a storage medium (such as flash memory) on which a medical questionnaire form is recorded is mounted in the drive device of the medical information processing device 100, the acquisition function 121 may acquire the medical questionnaire form from the storage medium.

[0053] Next, the questionnaire item extraction function 122 extracts questionnaire items from the acquired patient questionnaire form (step S102).

[0054] Figure 5 shows an example of questionnaire items extracted from a questionnaire form. As shown in the figure, for example, the questionnaire item extraction function 122 may extract information (patient responses) associated with questionnaire items such as present illness history, lifestyle history, family history, and preferences history.

[0055] For extracting questionnaire items, for example, a pre-trained model MDL of a large-scale language model may be used. In this case, the questionnaire item extraction function 122 inputs the questionnaire form into the pre-trained model MDL. In response, the pre-trained model MDL extracts information (patient responses) that corresponds to questionnaire items such as present illness history, lifestyle history, family history, and preferences history from the input questionnaire form.

[0056] The processes S100 and S102 described above may be performed before the processes in this flowchart begin. In other words, the processes S100 and S102 may be pre-processing.

[0057] Returning to the flowchart explanation, the event extraction function 123 then extracts events from the patient's personal data that occurred in the patient or in the patient's life (step S104).

[0058] The events can be any events that a patient may experience at some point in their life, such as starting school, changing schools, taking entrance exams, advancing to higher education, living alone, finding a job, getting married, giving birth, raising children, retiring, buying a house or car, or traveling.

[0059] Figure 6 shows an example of an event extracted from a patient's personal data. As shown in the figure, for example, if image data is stored in memory 114 as personal data, the event extraction function 123 may estimate the event represented by the image data from the metadata attached to the image data (date, location information, title, etc.) or from the image data itself.

[0060] For event extraction, for example, a pre-trained model MDL of a large-scale language model may be used. In this case, the event extraction function 123 inputs image data and its metadata to the pre-trained model MDL. More specifically, the event extraction function 123 inputs a dataset labeled as the correct answer for the event already known to the pre-trained model MDL as an example prompt. Then, the event extraction function 123 inputs image data and its metadata for which the event is unknown to the pre-trained model MDL. In response, the pre-trained model MDL outputs the event represented by the input image data. In the illustrated example, the event "transferring schools" is extracted from the image data, and the location where the event occurred is further identified as "a junior high school in Hokkaido" from the location information. The above example explains how to extract events from image data, but events can be extracted from video data and text data in the same way using the pre-trained model MDL.

[0061] The event extraction function 123 may extract events from various types of data included in personal data, such as images, videos, and text. Each event extracted by the event extraction function 123 is managed as a single record.

[0062] Figure 7 shows an example of an event managed as a record. As shown in the diagram, each event may be managed in the format of "When," "Where," "Who," "What," "Why," and "How" (i.e., the 5W1H format).

[0063] Returning to the flowchart explanation, the event log generation function 124 then generates a primary event log in which the events extracted from the personal data by the event extraction function 123 are arranged in chronological order (step S106). The primary event log is an example of a "primary log".

[0064] Next, the event extraction function 123 extracts events from the primary event log that correspond to each of the multiple questionnaire items extracted from the questionnaire form (step S108).

[0065] Figure 8 shows an example of a primary event log. As shown in the figure, multiple events (records) are arranged chronologically in the primary event log. The event extraction function 123 extracts events from the primary event log that match the questionnaire items (present illness history, lifestyle history, family history, hobbies history, etc.) extracted from the patient's questionnaire form by the questionnaire item extraction function 122. The "Where" column (Item A in the figure) represents event logs related to "educational history" and "work history," which are part of "lifestyle history." Therefore, the event extraction function 123 may extract the "Where" column from the primary event log as lifestyle history.

[0066] Returning to the flowchart explanation, the event log generation function 124 then generates a secondary event log (step S110) in which the events extracted from the primary event log by the event extraction function 123 (i.e., events corresponding to the questionnaire items) are arranged in chronological order. The secondary event log is an example of a "secondary log".

[0067] Returning to the flowchart explanation, the turning event estimation function 125 then estimates the turning event based on the secondary event log (step S112).

[0068] A turning event is one of several events included in a secondary event log that represents a turning point for the patient. More specifically, a turning event is an event that resulted in a change in the patient's emotions, thoughts, or behavioral patterns.

[0069] For example, the turning event estimation function 125 calculates the similarity between each of the multiple events included in the secondary event log and other events that are adjacent to each other in time (hereinafter referred to as adjacent events).

[0070] More specifically, the turning event estimation function 125 may convert each of the multiple events included in the secondary event log into a vector and calculate the cosine similarity between the vectors.

[0071] Figure 9 is a diagram illustrating an example of a turning event estimation method. As shown in the diagram, the turning event estimation function 125 extracts words (e.g., "Tokyo", "Hokkaido", "Junior High School", "High School") from the strings contained in each event and converts these words into vectors.

[0072] For splitting strings into words, methods such as morphological analysis may be used. For vectorizing words, pre-trained models of large-scale language models such as MDL may be used, or other machine learning models such as Bag of Words models, Word2vec models, and Transformer models may be used.

[0073] The turning event estimation function 125 may convert phrases or sentences into vectors instead of words. Vectors may also be interpreted as embeddings or distributed representations.

[0074] The turning event estimation function 125 selects one event of interest (hereinafter referred to as the "interest event") from among multiple events included in the secondary event log. The turning event estimation function 125 calculates the cosine similarity between the word vector extracted from the interest event and the word vector extracted from adjacent events that are time-series adjacent to the interest event.

[0075] Suppose the string in the event of interest is "a junior high school in Tokyo" and the string in the adjacent event is "a junior high school in Hokkaido". In this case, the turning event estimation function 125 calculates the cosine similarity between the vector of the word "Tokyo" originating from the event of interest and the vector of the word "Hokkaido" originating from the adjacent event, and also calculates the cosine similarity between the vector of the word "junior high school" originating from the event of interest and the vector of the word "junior high school" originating from the adjacent event.

[0076] The turning event estimation function 125 assigns a flag indicating a change to a focus event whose cosine similarity with an adjacent event exceeds a threshold. The turning event estimation function 125 then selects other events from among the multiple events included in the secondary event log that were not previously selected as focus events, calculates the similarity (e.g., cosine similarity) between the new focus events and adjacent events, and assigns a flag to the new focus events if the similarity exceeds a threshold. The turning event estimation function 125 repeats this process to determine whether or not to assign a flag to each of the multiple events included in the secondary event log.

[0077] The turning event estimation function 125 estimates events that have been flagged (events whose similarity exceeds a threshold) from among multiple events included in the secondary event log as turning events.

[0078] Returning to the flowchart explanation, the next step is to estimate the patient's response to each of the multiple questionnaire items extracted from the questionnaire form, based on the secondary event log (step S114).

[0079] For example, the questionnaire response estimation function 126 uses a trained model MDL of a large-scale language model to generate sentences corresponding to questionnaire items (present illness, lifestyle, family history, hobbies, etc.) from the strings contained in each of the multiple events in the secondary event log, and estimates these sentences as the patient's answers. Among the multiple questionnaire items, there may be questionnaire items for which answers could not be estimated (so-called blanks). For example, if the secondary event log does not contain an event related to "family history," "family history" will be blank in the estimated answer.

[0080] Next, the output control function 127 outputs the patient's answers estimated by the medical interview response estimation function 126 (hereinafter referred to as estimated answers) via the output interface 113 (step S116). For example, the output control function 127 may display the estimated answers on the display 113a included in the output interface 113. Alternatively, the communication control function 128 may transmit the estimated answers to the user interface 10 or the like via the communication interface 111. This completes the processing of this flowchart.

[0081] Figure 10 shows an example of how estimated responses are displayed. As shown in the figure, patient responses are estimated for multiple questionnaire items, such as present illness history, lifestyle history, family history, and preferences history. If there are questionnaire items for which no response was estimated, those items may be emphasized compared to those for which responses were estimated. For example, if no response was estimated for the questionnaire item "infancy," the questionnaire item "infancy" will be emphasized, as shown in Item B in the figure.

[0082] Figure 11 shows another example of the display of estimated answers. For example, the output control function 127 may emphasize estimated answers corresponding to turning events (hereinafter referred to as "first estimated answers") compared to estimated answers that do not correspond to turning events (hereinafter referred to as "second estimated answers"). For example, suppose events containing strings such as "leave of absence," "declining grades," and "transfer to junior high school" are estimated as turning events. In such cases, as shown in Item C in the figure, estimated answers estimated from strings included in turning events are emphasized compared to estimated answers estimated from strings included in other events. This makes it easy for medical staff to understand which events in particular were turning points for the patient from among the estimated answers. The first estimated answer is an example of a "first answer," and the second estimated answer is an example of a "second answer."

[0083] The output control function 127 may output turning events in addition to the estimated answer via the output interface 113.

[0084] Figure 12 shows an example of the display of estimated answers and turning events. As mentioned above, the estimated answers may include the first estimated answer. In this case, the output control function 127 may, as shown in ItemD in the figure, display a pop-up of the turning event used to estimate the first estimated answer when the mouse hovers over the string representing the first estimated answer, or transition to another page where the turning event is displayed via a hyperlink.

[0085] According to the first embodiment described above, the processing circuit 120 of the medical information processing device 100 generates a secondary event log in which events that occurred in the patient's life are arranged chronologically based on the patient's personal data. Based on the secondary event log, the processing circuit 120 estimates turning events that were significant for the patient, and also estimates the patient's answers to each of several questionnaire items. The processing circuit 120 outputs the estimated answers via the output interface 113 or transmits the estimated answers to the user interface 10 via the communication interface 111. In this case, the processing circuit 120 outputs the first estimated answer corresponding to the turning event as noteworthy information. Specifically, the processing circuit 120 emphasizes the first estimated answer corresponding to the turning event compared to the second estimated answer that does not correspond to the turning event. This makes it easy for medical staff to understand which events were particularly turning points for the patient from among the estimated answers. In addition, the patient does not have to recall their memories to answer the questionnaire items. In other words, the burden on the patient and medical staff during diagnosis can be reduced.

[0086] (Second Embodiment) The second embodiment will now be described. In the first embodiment, each event was described as being in the format of "When," "Where," "Who," "What," "Why," and "How" (i.e., the 5W1H format). In contrast, the second embodiment differs from the first embodiment in that, in addition to the 5W1H, appearance is also included in the event. The following description will focus on the differences from the first embodiment, and the points common to both embodiments will be omitted. In the description of the second embodiment, parts that are the same as in the first embodiment will be denoted by the same reference numerals.

[0087] In the second embodiment, the event extraction function 123 extracts changes in the patient's physical characteristics as events from the patient's personal data.

[0088] Figure 13 illustrates another example of an event managed as a record. As shown in the figure, each event is managed not only by the 5W1H format but also by external characteristics. Patient external characteristics may include, for example, height, weight, hair color, clothing, and facial expression (see Item E in the figure).

[0089] In the second embodiment, the turning event estimation function 125 estimates a turning event based on the patient's appearance included in multiple events in the secondary event log.

[0090] Figure 14 illustrates a method for estimating turning events from outward appearance characteristics. For example, outward appearance characteristics include features related to the patient's emotions, such as hair color, clothing, and facial expression (hereinafter referred to as first outward appearance characteristics), as shown by Item F in the figure, and features not related to the patient's emotions, such as height (hereinafter referred to as second outward appearance characteristics), as shown by Item G in the figure.

[0091] The turning event estimation function 125 may extract words from a string representing one of the patient's physical characteristics, the first physical characteristic, and estimate the turning event based on the cosine similarity of the vectors of those words.

[0092] Furthermore, in the second embodiment, the turning event estimation function 125 may estimate as a turning event a predetermined event in which there was no change in the patient's emotions, thoughts, or behavioral patterns, even though such a change should have occurred in those events.

[0093] A designated event is one that is likely to cause a significant change in the patient's emotions, thoughts, or behavioral patterns, such as the death of a family member or the birth of a child. In such designated events, messages posted by the patient on social media tend to be pessimistic or euphoric. The patient's physical characteristics also tend to change. Therefore, designated events tend to have a high cosine similarity to adjacent events, and the probability of them being estimated as turning events increases.

[0094] However, in patients with mental illness, even in the event of a predetermined event, there may be no change in emotions, thoughts, or behavioral patterns (i.e., the cosine similarity with adjacent events is below the threshold), and the predetermined event that should be estimated as a turning event may be overlooked. Therefore, the turning event estimation function 125 may estimate the event of interest as a turning event if it is a predetermined event, regardless of the cosine similarity with adjacent events.

[0095] According to the second embodiment described above, the processing circuit 120 of the medical information processing device 100 estimates a turning event based on one of the first external characteristics, which is one of the patient's external characteristics, the first external characteristic. This allows medical staff to easily understand what kind of change in external characteristics constitutes a turning point for the patient.

[0096] (Third embodiment) The third embodiment will now be described. The third embodiment differs from the first or second embodiment in that it compares the estimated patient's response (i.e., estimated response) for each of the multiple questionnaire items with the patient's actual response (hereinafter referred to as the actual response) for each of the multiple questionnaire items. The following description will focus on the differences from the first or second embodiment, and will omit explanations of points common to the first or second embodiment. In the description of the third embodiment, the same reference numerals will be used for parts that are the same as those in the first or second embodiment.

[0097] Figure 15 is a diagram showing an example configuration of the medical information processing device 100 in the third embodiment. The processing circuit 120 of the medical information processing device 100 includes, for example, the acquisition function 121, the medical interview item extraction function 122, the event extraction function 123, the event log generation function 124, the turning event estimation function 125, the medical interview answer estimation function 126, the output control function 127, and the communication control function 128 described above, as well as a speech recognition function 129 and an answer judgment function 130.

[0098] The speech recognition function 129 performs speech recognition processing on audio data obtained when a patient verbally answers questionnaire items or when a doctor interviews a patient, and converts the audio data into text data. For example, suppose a patient verbally answers questionnaire items to the user interface 10. In this case, the user interface 10 functions as a voice user interface and acquires the patient's audio data. The speech recognition function 129 performs speech recognition processing on the audio data acquired by the user interface 10. This obtains the patient's actual answers.

[0099] The answer determination function 130 determines whether the estimated answer and the actual answer match. For example, the answer determination function 130 converts both the estimated answer and the actual answer into vectors and calculates the cosine similarity between these vectors. Then, if the cosine similarity is above a threshold, the answer determination function 130 may determine that the estimated answer and the actual answer match.

[0100] Figure 16 is a flowchart showing a series of processing steps in the processing circuit 120 in the third embodiment. Figure 17 is a schematic diagram showing a series of processing steps in the processing circuit 120 in the third embodiment.

[0101] First, the acquisition function 121 acquires audio data of the doctor's interview with the patient (step S200). For example, the acquisition function 121 may acquire the doctor's voice data via the microphone of the user interface 10.

[0102] Next, the speech recognition function 129 performs speech recognition processing on the doctor's voice data and converts the voice data into text data (step S202).

[0103] Next, the event log generation function 124 generates a secondary event log based on the text data (step S204).

[0104] Specifically, the questionnaire item extraction function 122 extracts questionnaire items from text data derived from the doctor's voice data. The event extraction function 123 extracts events corresponding to each of the multiple questionnaire items extracted from the text data from the primary event log. The event log generation function 124 generates a secondary event log in which the events extracted from the primary event log (i.e., events corresponding to the questionnaire items) are arranged in chronological order.

[0105] Next, the questionnaire response estimation function 126 estimates the patient's response to each of the multiple questionnaire items extracted from the text data, based on the secondary event log (step S206). This obtains the estimated responses.

[0106] Next, the acquisition function 121 acquires audio data of the patient's responses to the medical interview (step S208). For example, the acquisition function 121 may acquire the patient's audio data via the microphone of the user interface 10.

[0107] Next, the speech recognition function 129 performs speech recognition processing on the patient's voice data and converts the voice data into text data (step S210). This obtains the patient's actual answers to the medical interview.

[0108] Next, the response determination function 130 compares the estimated response with the actual response and determines whether the responses match (step S212).

[0109] If it is determined that the estimated answer and the actual answer do not match, the output control function 127 outputs predetermined information via the output interface 113 (step S214).

[0110] The prescribed information is information used to inform a doctor when the estimated response does not match the actual response. For example, a secondary event log of a patient with alcohol dependence may reveal that they drink regularly or frequently, but the patient may give a false actual response such as, "I only drink alcohol occasionally." In such cases, the estimated response and the actual response do not match, so the prescribed information is output.

[0111] Furthermore, the communication control function 128 may transmit predetermined information to the user interface 10 or the like via the communication interface 111. This completes the processing of this flowchart.

[0112] According to the third embodiment described above, the processing circuit 120 of the medical information processing device 100 determines whether the estimated answer and the actual answer match, and if they do not match, it outputs predetermined information via the output interface 113. This makes it easy for medical staff to know that the patient gave an answer that differs from the answer to the medical interview estimated from the secondary event log.

[0113] (Other embodiments) Other embodiments will be described below. In the embodiments described above, the user interface 10 and the medical information processing device 100 were described as separate devices, but the invention is not limited to this. For example, the user interface 10 and the medical information processing device 100 may be a single integrated device. For example, the processing circuit 20 of the user interface 10 may, in addition to the acquisition function 21, output control function 22, and communication control function 23, further include the acquisition function 121, questionnaire item extraction function 122, event extraction function 123, event log generation function 124, turning event estimation function 125, questionnaire answer estimation function 126, output control function 127, and communication control function 128, speech recognition function 129, and answer judgment function 130, which are also provided by the processing circuit 120 of the medical information processing device 100. In this case, the user interface 10 can perform the various flowchart processing described above in a standalone (offline) manner.

[0114] Furthermore, in the embodiments described above, the questionnaire response estimation function 126 was described as estimating the patient's response to each of the multiple questionnaire items extracted from the questionnaire form based on the secondary event log, but it is not limited to this. The patient's response to each of the multiple questionnaire items does not necessarily need to be estimated; for example, it may be obtained from the patient themselves or a doctor using the user interface 10. In other words, actual responses may be used instead of estimated responses.

[0115] According to at least one embodiment described above, the processing circuit 120 of the medical information processing device 100 generates a secondary event log in which events that occurred in the patient's life are arranged chronologically based on the patient's personal data. Based on the secondary event log, the processing circuit 120 estimates turning events that were significant for the patient, and also estimates the patient's answers to each of the multiple questionnaire items. As mentioned above, the patient's answers do not necessarily need to be estimated and may be obtained from the patient themselves or from a doctor, etc.

[0116] The processing circuit 120 outputs estimated answers via the output interface 113 and transmits estimated answers to the user interface 10 via the communication interface 111. In this process, the processing circuit 120 outputs the first estimated answer corresponding to the turning event as noteworthy information. Specifically, the processing circuit 120 emphasizes the first estimated answer corresponding to the turning event compared to the second estimated answer that does not correspond to the turning event. This allows medical staff to easily identify which events were particularly turning points for the patient from among the estimated answers. Furthermore, patients do not have to recall their memories to answer the questionnaire items. In other words, the burden on both patients and medical staff during diagnosis can be reduced.

[0117] While several embodiments have been described, these embodiments are presented as examples only and are not intended to limit the scope of the invention. These embodiments can be carried out in a variety of other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents. [Explanation of symbols]

[0118] 1…Medical information processing system, 10…User interface, 11…Communication interface, 12…Input interface, 13…Output interface, 14…Memory, 20…Processing circuit, 21…Acquisition function, 22…Output control function, 23…Communication control function, 100…Medical information processing device, 111…Communication interface, 112…Input interface, 113…Output interface, 114…Memory, 120…Processing circuit, 121…Acquisition function, 122…Medical questionnaire item extraction function, 123…Event extraction function, 124…Event log generation function, 125…Turning event estimation function, 126…Medical questionnaire answer estimation function, 127…Output control function, 128…Communication control function

Claims

1. A computer-based method for processing medical information, Based on the patient's personal data, an event log is generated, which is a log of events that occurred in the patient arranged in chronological order. Based on the event log, estimate the turning event, which was a pivotal event for the patient. To obtain the patient's responses to multiple items related to the medical interview, Of the patient's responses, the responses corresponding to the turning event are output as attention information via the output interface. A medical information processing method including [the specified term].

2. This further includes emphasizing the first response, which corresponds to the aforementioned turning event, compared to the second response, which does not correspond to the aforementioned turning event. The medical information processing method according to claim 1.

3. In addition to the first response which is the response corresponding to the turning event, the turning event is further output via the output interface, The medical information processing method according to claim 1 or 2.

4. The events included in the event log include a first physical characteristic of the patient that is related to the patient's emotions, and a second physical characteristic of the patient that is not related to the patient's emotions. Further comprising estimating the turning event based on the first outward appearance characteristic, The medical information processing method according to claim 1 or 2.

5. Based on the event log, estimate the patient's responses to the multiple items. To determine whether the actual response obtained from the patient matches the estimated response obtained from the patient, and If the actual answer and the estimated answer do not match, the system further includes outputting predetermined information. The medical information processing method according to claim 1 or 2.

6. For each of the multiple events included in the event log, the similarity to adjacent events, which are events adjacent to each other in chronological order, is calculated, and The further includes estimating an event whose similarity to the adjacent event is greater than or equal to a threshold as the turning event, The medical information processing method according to claim 1 or 2.

7. Convert each of the multiple events included in the event log into a vector, and This further includes calculating the cosine similarity between the aforementioned vectors as the similarity, The medical information processing method according to claim 6.

8. Further comprising estimating the event in which a change occurred in the patient's emotional, thought, or behavioral patterns as the turning event, The medical information processing method according to claim 1 or 2.

9. This further includes estimating a predetermined event in which there was no change in the patient's emotions, thoughts, or behavioral patterns, in a predetermined event where such a change should occur, as the turning event. The medical information processing method according to claim 1 or 2.

10. This further includes extracting the aforementioned items from one or more questionnaire forms. The medical information processing method according to claim 1 or 2.

11. Extracting the aforementioned events from the aforementioned personal data, To generate a primary log by arranging the primary events, which are the events extracted from the personal data, in chronological order. From the primary log, extract the events corresponding to each of the multiple items that the patient is asked about as part of the medical interview, and The process further includes generating a secondary log as the event log, which is a chronological arrangement of the secondary events extracted from the primary log. The medical information processing method according to claim 1 or 2.

12. A generation unit generates an event log, which is a log of events that occurred in the patient, arranged in chronological order, based on the patient's personal data. Based on the event log, a first estimation unit estimates the turning event, which is the event that marked a turning point for the patient. An acquisition unit that acquires the patient's responses to multiple items related to the medical interview, An output control unit outputs, via an output interface, the responses from the patient that correspond to the turning event as attention information, A medical information processing device equipped with [a specific feature].

13. A program to be executed by a computer, Based on the patient's personal data, an event log is generated, which is a log of events that occurred in the patient arranged in chronological order. Based on the event log, estimate the turning event, which was a pivotal event for the patient. To obtain the patient's responses to multiple items related to the medical interview, Of the patient's responses, the responses corresponding to the turning event are output as attention information via the output interface. A program that includes this.

Images