Program, Information Processing Method, and Information Processing Apparatus

Abstract

Provide a program or the like that can present an appropriate medical institution for a patient to visit. 【Solution means】The program acquires patient information regarding a patient who requires surgery, generates surgical plan information representing the surgery to be performed on the patient based on the patient information, and causes a computer to execute a process of generating a list of medical institutions or doctors with experience in the surgery represented by the surgical plan information. Preferably, the program generates the list by searching a database that stores the surgical performance of each medical institution or each doctor for medical institutions or doctors with experience in the surgery represented by the surgical plan information.

Description

[Technical field]

[0001] The present invention relates to a program, an information processing method, and an information processing device. [Background technology]

[0002] There is a system that supports the diagnosis of patients who require surgery. For example, Patent Document 1 discloses a program that inputs a prompt including a medical image of the affected area of the patient into a large language model (hereinafter referred to as "LLM"), and acquires and displays surgical plan information representing a surgical plan from the LLM. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 7553165 Summary of the Invention [Problem to be solved by the invention]

[0004] One aspect is to provide a program or the like that can suggest appropriate medical facilities. [Means for solving the problem]

[0005] The program is for patients who need surgery. including a medical image of the affected part Acquire patient information; By inputting a prompt including generating surgical planning information representative of a surgical procedure to be performed on the patient; generating a query for instructing a listing of medical institutions with a record of the surgical operation represented by the surgical plan information, and by inputting a prompt including the query and information of each medical institution retrieved from the Internet based on the query into a language model The surgical plan information indicates a surgical procedure performed by a medical device having a proven track record. related to A process for generating the list is performed by a computer. Effect of the Invention

[0006] In one aspect, a program or the like is provided that can suggest appropriate medical facilities. [Brief description of the drawings]

[0007]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

[0008] The present invention will be described in detail below with reference to the drawings showing embodiments thereof. (Embodiment 1) Figure 1 is a diagram showing an example of the configuration of a medical support system. In this embodiment, a medical support system is described that uses an LLM 50 to generate surgical plan information representing a surgical operation to be performed on a patient, and generates and presents a list of medical institutions or doctors with experience in performing the surgical operation represented by the generated surgical plan information. The medical support system includes an information processing device 1, terminals 2, 2, 2..., and a generation server 3. Each device is communicatively connected via a network N such as the Internet.

[0009] The information processing device 1 is an information processing device capable of various information processing and information transmission and reception, such as a server computer, a personal computer, etc. In this embodiment, it is assumed that the information processing device 1 is a server computer, and hereinafter it will be read as server 1 for simplicity. The server 1 provides an application for comprehensively managing surgical cases and schedules for a medical institution performing surgical operations. Specifically, the server 1 databases and manages surgical cases (records) and schedules, and shares information with each user (medical staff), who is a member of the surgical team.

[0010] Note that as the "user" of this system, medical staff (such as doctors) who diagnose patients is assumed, but the "user" may also be the patient himself / herself.

[0011] In this embodiment, the server 1 has a function of generating a surgical plan to be performed on a patient and presenting it to the user as one function of the above application. Specifically, the server 1 inputs a prompt including patient information (such as a medical image obtained by imaging the affected part of the patient) regarding a patient who requires a surgical operation to the LLM 50, generates surgical plan information representing the surgical operation to be performed on the patient, and presents it to the user.

[0012] The terminal 2 is a terminal device used by a user who uses this system, such as a personal computer, a smartphone, a tablet terminal, etc. The terminal 2 accepts the input of patient information on a surgical plan creation screen (see Fig. 4) described later. The server 1 obtains patient information from the terminal 2, inputs it to the LLM 50, generates surgical plan information, and outputs it to the terminal 2.

[0013] The generation server 3 is a server computer that generates a response sentence using the LLM 50 from input data such as images and texts. The LLM 50 is a large language model such as GPT (Generative Pre-trained Transformer) or BERT (Bidirectional Encoder Representations from Transformer), but the model is not particularly limited. When the server 1 acquires patient information from the terminal 2, it creates a prompt including the patient information and sends it to the generation server 3 to cause the generation server 3 to generate surgical plan information.

[0014] As described above, the server 1 generates surgical plan information necessary for the patient using the LLM 50 and presents it to the user. On the other hand, when the user (medical staff) who has received the presentation of the surgical plan is a non-specialist doctor, etc., it is necessary to judge the medical institutions and doctors capable of performing the surgery and introduce them to the patient. However, the work of searching for an appropriate medical institution to visit and introducing it to the patient is not an easy task and is a burden on the user.

[0015] Therefore, in the present embodiment, medical institutions or doctors capable of performing the surgical operation represented by the surgical plan information generated by the LLM 50 are listed and presented to the user. This reduces the burden on the user associated with judging the appropriateness of the medical institution to visit.

[0016] FIG. 2 is a block diagram showing a configuration example of the server 1. The server 1 includes a control unit 11, a main memory unit 12, a communication unit 13, and an auxiliary storage unit 14. The control unit 11 is a processor such as one or more CPUs (Central Processing Units), MPUs (Micro-Processing Units), GPUs (Graphics Processing Units), etc. By reading and executing the program P stored in the auxiliary storage unit 14, it performs various information processing. The main storage unit 12 is a temporary storage area such as SRAM (Static Random Access Memory), DRAM (Dynamic Random Access Memory), etc., and temporarily stores the data necessary for the control unit 11 to execute arithmetic processing. The communication unit 13 is a communication module for performing communication-related processing, and transmits and receives information with the outside.

[0017] The auxiliary storage unit 14 is a non-volatile storage area such as a hard disk, a large-capacity memory, etc., and stores the program P (program product) and other data necessary for the control unit 11 to execute processing. In addition, the auxiliary storage unit 14 stores a medical fee DB 141, a medical institution DB 142, a doctor DB 143, and a surgery schedule / history DB 144. The medical fee DB 141 is a database that stores medical fee information (information related to the surgical procedure) for each surgical procedure of a surgical operation. The medical institution DB 142 is a database that stores information on medical institutions that use this system (application). The doctor DB 143 is a database that stores information on each doctor working at a medical institution that uses this system. The surgery schedule / history DB 144 is a database that stores the schedule of surgeries scheduled at each medical institution that uses this system and the history (performance) of surgeries that have been performed.

[0018] Note that the auxiliary storage unit 14 may be an external storage device connected to the server 1. Also, the server 1 may be a multi-computer composed of a plurality of computers, or may be a virtual machine virtually constructed by software.

[0019] Also, in this embodiment, the server 1 is not limited to the above configuration and may include, for example, an input unit that receives operation inputs, a display unit that displays images, etc. Further, the server 1 may be provided with a reading unit that reads a portable storage medium 1a such as a CD (Compact Disk)-ROM or a DVD (Digital Versatile Disc)-ROM, and reads and executes the program P from the portable storage medium 1a.

[0020] FIG. 3 is a diagram showing an example of the record layouts of the medical fee DB 141, the medical institution DB 142, the doctor DB 143, and the surgery schedule / history DB 144.

[0021] The medical fee DB 141 includes a medical fee code column, a surgical procedure name column, a body part name column, a vector value column, and a medical fee point number column. The medical fee code column stores a medical fee code (for example, K code). The surgical procedure name column, the body part name column, the vector value column, and the medical fee point number column store a surgical procedure name, a body part (affected part) name, a vector value, and a medical fee point number in association with the medical fee code, respectively. The vector value will be described later.

[0022] The medical institution DB 142 includes a medical institution ID column, a medical institution name column, a department name column, and an address column. The medical institution ID column stores a medical institution ID, which is an identifier of each medical institution. The medical institution name column, the department name column, and the address column store a medical institution name, a department name, and an address in association with the medical institution ID, respectively.

[0023] The doctor DB 143 includes a doctor ID column, a surname column, an affiliated medical institution column, and a specialty column. The doctor ID column stores a doctor ID, which is an identifier of each doctor. The surname column, the affiliated medical institution column, and the specialty column store a doctor's surname, a medical institution ID of the affiliated medical institution, and a specialty in association with the doctor ID, respectively.

[0024] The surgical schedule / history DB 144 includes a surgical date / time column, a patient name column, a performing medical institution column, an in-charge doctor column, a surgical procedure name column, and a surgical site column. The surgical date / time column stores the surgical date / time (scheduled surgical date / time or actual surgical date / time). The patient name column, the performing medical institution column, the in-charge doctor column, the surgical procedure name column, and the surgical site column store, respectively, in association with the surgical date / time, the name of the patient undergoing the surgery, the medical institution ID of the medical institution performing the surgery, the doctor ID of the doctor in charge of the surgery, the surgical procedure, and the name of the site where the surgery is to be performed.

[0025] Figure 4 is a diagram showing an example of a surgical plan creation screen. In Figure 4, an example of an input screen for patient information displayed by the terminal 2 when creating a surgical plan is illustrated. Hereinafter, the outline of the present embodiment will be described.

[0026] The surgical plan creation screen includes a chief complaint / medical history input field 41, an image input field 42, an image finding input field 43, and a generation button 44. The chief complaint / medical history input field 41 is an input field for receiving the input of the patient's chief complaint (symptoms) and medical history. The terminal 2 receives the input of text representing the patient's chief complaint and medical history via the chief complaint / medical history input field 41.

[0027] The image input field 42 is an input field for receiving the input of a medical image (e.g., an X-ray image) of the patient's affected part. The terminal 2 receives the input of an image of the patient's affected part via the image input field 42.

[0028] The image finding input field 43 is an input field for receiving the input of findings for the medical image. The terminal 2 receives the input of text representing the findings for the medical image via the image finding input field 43.

[0029] The generation button 44 is a button for instructing the generation of surgical plan information based on the patient information (medical image and text) input in the chief complaint / medical history input field 41, the image input field 42, and the image finding input field 43. When an operation input to the generation button 44 is received, the server 1 generates the surgical plan information.

[0030] FIG. 5 is an explanatory diagram regarding the generation process of surgical planning information. In FIG. 5, it illustrates how surgical planning information is generated by inputting a prompt including a medical image (for example, an X-ray image) and other text data into the LLM50.

[0031] In this embodiment, an X-ray image is treated as the medical image, but the medical image may be an image of other modalities (for example, an ultrasonic image, a CT (Computed Tomography) image, an MRI (Magnetic Resonance Imaging) image, etc.) other than the X-ray image.

[0032] The LLM50 is a machine learning model that has learned a large amount of training data and is a model that generates a response sentence when receiving input of input data (images, texts, etc.). As described above, the LLM50 is a large language model such as GPT, BERT, etc., but the model is not particularly limited.

[0033] In this embodiment, the server 1 uses the LLM50 for formulating a surgical plan for a surgical operation. Specifically, the server 1 generates surgical planning information representing the surgical operation to be performed on the patient by inputting a prompt including a medical image of the affected part of the patient and text data such as the image findings described in the medical record into the LLM50.

[0034] For example, the server 1 creates a prompt including a query (instruction sentence) for instructing the LLM50 to generate a plurality of surgical planning information and patient information (medical image and text data) input on the surgical planning screen. By outputting the prompt to the generation server 3, the server 1 causes the LLM50 to generate a plurality of surgical planning information.

[0035] In this embodiment, the server 1 uses the medical fee DB 141 to generate surgical plan information by means of the RAG (Retrieval Augmented Generation) technology. The medical fee DB 141 is a database that stores information (medical fee information) related to each surgical procedure of surgical operations, such as the name of the surgical procedure, the name of the site, and the medical fee points. In the medical fee DB 141, information related to each surgical procedure is stored in association with the vector value obtained by converting the patient information.

[0036] The server 1 converts the patient information into a vector value (feature quantity) by inputting a prompt including the patient information to the LLM 50. Then, the server 1 uses the converted vector value as a search query to search for information related to the surgical procedure from the medical fee DB 141. For example, the server 1 pre-converts the patient information of representative cases corresponding to each surgical procedure by a predetermined converter (model for embedding), and stores it in the medical fee DB 141 in association with the name of the surgical procedure, the name of the site, etc. For example, the server 1 acquires information related to the top several surgical procedures that are approximated to the vector value obtained by converting the patient information of the patient currently being processed from the medical fee DB 141.

[0037] In this embodiment, although the medical image and the text data are collectively described as being converted into one vector value, the medical image and the text data may be respectively converted into vector values, and information related to the surgical procedure may be acquired from the medical fee DB 141 using at least one of the vector values.

[0038] The server 1 inputs a prompt including the acquired information and the patient information used as the search query to the LLM 50. Thereby, the LLM 50 generates surgical plan information (response text) with reference to the information related to the surgical procedure acquired from the medical fee DB 141. The LLM 50 generates a plurality of surgical plan information with reference to the information related to each of the top several surgical procedures.

[0039] Server 1 may output the surgical plan information (response text) generated by LLM50 as it is to terminal 2 for display. However, in this embodiment, a word representing the surgical procedure name is extracted from the surgical plan information generated by LLM50, and the extracted word is displayed.

[0040] FIG. 6 is a diagram showing an example of a list display screen. In FIG. 6, a screen example is illustrated in which the surgical plan information (surgical procedure name) generated using LLM50 is displayed, and a list of medical institutions or doctors with experience in the surgical procedures represented by the surgical plan information is displayed.

[0041] The list display screen includes a patient information display column 61, a surgical plan display column 62, a search condition setting column 63, and a list display column 64. The patient information display column 61 is a display column for displaying patient information (text data) other than medical images input on the surgical plan creation screen (FIG. 4). Note that terminal 2 may also display medical images together.

[0042] The surgical plan display column 62 is a display column for displaying the name of the surgical procedure (surgical procedure name) to be performed on the patient. Terminal 2 displays the surgical procedure name represented by the surgical plan information generated using LLM50 in the surgical plan display column 62. Note that a plurality of surgical procedure names are displayed in the surgical plan display column 62 in a pull-down manner, and the user can select an appropriate surgical procedure.

[0043] The search condition setting column 63 is an input column for setting other search conditions that can be set when searching for medical institutions or doctors where the patient should receive medical treatment. For example, terminal 2 accepts input for setting a geographical range including a reference position ("current location" in FIG. 6) and the distance from the reference position as the first search condition. In addition, terminal 2 accepts input for setting the surgical available period (within what period the surgery is possible) as the second search condition.

[0044] The list display column 64 is a display column that displays a list of medical institutions or doctors with experience in the surgical operations displayed in the surgical plan display column 62. For example, the terminal 2 switches to display a list of medical institutions or a list of doctors by accepting an operation input to the tab 641. FIG. 6 illustrates the case of displaying a list of medical institutions.

[0045] The server 1 searches for medical institutions or doctors with experience in the surgical operations displayed in the surgical plan display column 62 from a database that stores the surgical operation results of each medical institution or each doctor, that is, the surgical schedule / history DB 144. That is, the server 1 refers to the surgical procedure name stored in the surgical procedure name column of the surgical schedule / history DB 144, the surgical site name stored in the surgical site column, etc. (see FIG. 3), and extracts medical institutions or doctors with experience in the surgical operations displayed in the surgical plan display column 62. For example, the server 1 extracts medical institutions or doctors whose number of times (number of people) of performing the operation is equal to or greater than a threshold value. Note that the threshold value may be changeable by the user.

[0046] Furthermore, the server 1 narrows down to medical institutions or doctors that meet the search conditions set in the search condition setting column 63. Specifically, when a setting is made for the geographical range, the server 1 acquires the position information of the user as the reference position (for example, if it is the "current location", the position information of the current terminal 2), and narrows down to medical institutions located within the set distance from the user's position, or doctors working at the medical institutions. Also, when a setting is made for the surgical available period, the server 1 reads schedule information representing the surgical schedules of each medical institution or each doctor from the surgical schedule / history DB 144, and narrows down to medical institutions or doctors that can perform the operation within the set period.

[0047] In this way, the server 1 searches for medical institutions or doctors that have experience in the surgical operations represented by the surgical plan information generated using the LLM 50 and that satisfy other geographical and time conditions. The server 1 generates a list of the searched medical institutions or doctors and displays it in the list display column 64.

[0048] For example, when displaying a list of medical institutions, as shown in FIG. 6, the terminal 2 displays, in addition to the name of the medical institution and the name of the department, the surgical record (the number of times (number of people) the corresponding surgery has been performed), the location, and the date on which surgery can be performed soon. For example, the terminal 2 arranges and displays each medical institution in descending order of the surgical record. Note that the terminal 2 may display each medical institution in ascending order of the distance from the reference position, or may display each medical institution in ascending order of the date on which surgery can be performed soon.

[0049] As described above, in the present embodiment, not only can a surgical plan be formulated, but also a medical institution or doctor with a surgical record in accordance with the surgical plan can be presented. Thereby, the user (medical staff or patient) can be comprehensively supported.

[0050] FIG. 7 is a flowchart showing an example of a processing procedure executed by the server 1. Based on FIG. 7, the processing content executed by the server 1 will be described. The control unit 11 of the server 1 receives an input of patient information regarding a patient who requires a surgical operation via the terminal 2 (step S11). The patient information includes a medical image obtained by imaging the affected part of the patient and text representing the chief complaint, medical history, findings on the medical image, etc. of the patient.

[0051] The control unit 11 converts the patient information into vector values by inputting a prompt including the patient information to the LLM 50 (step S12). The control unit 11 acquires (searches) information regarding the surgical procedure based on the converted vector values from the medical fee DB 141 that stores information (medical fee information) regarding each surgical procedure in association with the vector values (step S13).

[0052] The control unit 11 generates surgical plan information representing the surgical operation to be performed on the patient by inputting a prompt including the acquired information regarding the surgical procedure and the patient information used as a search query to the LLM 50 (step S14). Specifically, the control unit 11 inputs a prompt instructing the generation of a plurality of pieces of surgical plan information to the LLM 50 to generate a plurality of pieces of surgical plan information (response sentences). The control unit 11 outputs a list display screen showing the generated surgical plan information to the terminal 2 for display.

[0053] The control unit 11 receives input for setting other search conditions when searching for a medical institution or a doctor via the terminal 2 (step S15). Specifically, the control unit 11 receives input for setting the geographical range (reference position and distance from the reference position) and the operable period when searching for a medical institution or a doctor.

[0054] The control unit 11 searches a database (surgery schedule / history DB144) that stores the surgery performance of each medical institution or each doctor for a medical institution or a doctor with a track record of the surgical operation represented by the surgery plan information generated in step S14 (step S16). Specifically, the control unit 11 searches for a medical institution or a doctor that has a track record of the corresponding surgical operation, is located within the geographical range set in step S15, and is operable within the period set in step S15. The control unit 11 generates a list indicating the searched medical institutions or doctors (step S17). The control unit 11 causes the generated list to be displayed on the list display screen (step S18) and ends a series of processes.

[0055] Note that in the above, the surgery plan information is generated from patient information using the LLM50 (large language model), but the present embodiment is not limited to this. For example, the server 1 may generate the surgery plan information using a machine learning model other than the LLM50 (for example, a convolutional neural network) or a rule-based algorithm. That is, it is sufficient that the server 1 can generate the surgery plan information based on the patient information, and the generation algorithm is not limited to the LLM50.

[0056] From the above, according to the first embodiment, an appropriate medical institution to visit can be presented.

[0057] (Second Embodiment) In Embodiment 1, a form was described in which a medical institution or doctor with experience in the corresponding surgery is searched from a pre-prepared database to generate a list. In this embodiment, a form will be described in which a medical institution with surgical experience is searched from the Internet to generate a list. Note that the same reference numerals are used for the content overlapping with Embodiment 1, and the description thereof is omitted.

[0058] FIG. 8 is an explanatory diagram regarding the list generation process according to Embodiment 2. In FIG. 8, a state of generating a list of medical institutions is schematically illustrated by inputting a prompt including a query instructing the listing of medical institutions and information on medical institutions searched from the Internet based on the query into the LLM50. The outline of this embodiment will be described below.

[0059] The server 1 generates surgical plan information representing the surgery to be performed on the patient by inputting a prompt including patient information into the LLM50, as in Embodiment 1. In this embodiment, the server 1 generates a list of medical institutions by inputting a prompt instructing the listing of medical institutions with experience in the surgery represented by the surgical plan information into the LLM50.

[0060] First, the server 1 generates a query instructing the listing of medical institutions. For example, in addition to the surgical plan information, the server 1 refers to search conditions (e.g., geographical range) arbitrarily set by the user, and generates a query by applying the surgical procedure name and place name to a predetermined template sentence. In the example of FIG. 8, the server 1 generates a query by applying the surgical procedure name represented by the surgical plan information and the place name of the reference location set as the search condition to the template "The current location is [place name]. Please list medical institutions with experience in [surgical procedure name] surgery."

[0061] Next, server 1 searches for information on each medical institution from the Internet based on the generated query. Note that the search process may be executed by generation server 3 instead of server 1. Server 1 extracts a predetermined word from the query generated above, and performs a search using the extracted word as a search query. In the example of FIG. 8, server 1 extracts "total hip arthroplasty" (surgical procedure name), "surgical results", "medical institution", "Tokyo" (place name), etc. as search queries and performs a search.

[0062] For example, server 1 searches for the home pages, article sites, review sites, etc. on which the surgical results of each medical institution are described based on the above search queries. In Japan, medical institutions of a certain scale or more (medical institutions that receive a certain amount or more of medical fees) are obliged to publish the number of patients by major surgical department by medical department (the number of patients by medical department up to the top 5). That is, the surgical results of each medical institution are publicly available on the home page of each medical institution, etc. Therefore, server 1 searches for information on medical institutions with experience in the corresponding surgical procedure by searching these websites based on search queries including the surgical procedure name and place name.

[0063] Server 1 creates a prompt including a query instructing to list medical institutions and the information on the medical institutions searched based on the query. Then, server 1 generates a list of medical institutions (response text) by inputting the created prompt into LLM50. As a result, a list of medical institutions with experience in the corresponding surgical procedure can be generated without creating a database of the surgical results of each medical institution.

[0064] In this embodiment, the server 1 combines the method of searching for medical institutions with surgical performance from the Internet as described above and generating a list, and the method of searching for medical institutions with surgical performance from the database (Surgical Schedule / History DB 144) and generating a list as in Embodiment 1. That is, the server 1 performs each method simultaneously in parallel and generates a list of medical institutions with the performance of the corresponding surgical operation respectively. Then, the server 1 integrates the lists generated by each method and displays them in the list display column 64 of the list display screen as the final list.

[0065] FIG. 9 is a diagram showing an example of a list display screen according to Embodiment 2. In FIG. 9, similar to FIG. 6, an example of a screen for displaying a list of medical institutions is illustrated.

[0066] The difference from the list of medical institutions shown in FIG. 6 is that a star mark is attached to the medical institutions whose surgical performance is stored in the surgical schedule / history DB 144, that is, the medical institutions registered in this system, and they are displayed at the top of the list. In the example of FIG. 9, star marks are attached to "AA Hospital" and "BB Hospital", indicating that these medical institutions are registered in this system. On the other hand, no star mark is attached to "CC Hospital", indicating that this medical institution is a medical institution searched from the Internet.

[0067] In the case of medical institutions registered in this system, information on doctors working in the medical institutions and the surgical schedules in the medical institutions can be grasped. On the other hand, it is difficult to grasp the surgical schedules of medical institutions searched from the Internet. Therefore, the "CC Hospital" in FIG. 9 shows that the surgical available date is "unknown". For such reasons, the server 1 displays the medical institutions registered in this system at the top of the list.

[0068] As described above, the server 1 integrates the list of medical institutions retrieved from the surgical schedule / history DB 144 and the list of medical institutions retrieved from the Internet, and generates a list that prioritizes the medical institutions listed in the former list and presents it to the user. As a result, while providing the user with various options for the medical institution to visit, it is possible to preferentially present medical institutions that can grasp the surgical schedule and the like.

[0069] FIG. 10 is a flowchart showing an example of a processing procedure executed by the server 1 according to the second embodiment. After retrieving from the database (surgical schedule / history DB 144) a medical institution or doctor with a record of the surgical operation represented by the surgical plan information (step S16), the server 1 executes the following processing. The control unit 11 of the server 1 generates a query (instruction sentence) for instructing the listing of medical institutions with a record of the surgical operation represented by the surgical plan information generated in step S14 (step S201). Specifically, in addition to the surgical plan information, the control unit 11 refers to the search condition regarding the geographical range set in step S15, and generates a query specifying a geographical reference position and a surgical procedure name with a surgical record.

[0070] The control unit 11 extracts a predetermined word (for example, a place name representing a reference position, a surgical procedure name, etc.) from the generated query, and searches the Internet for information on each medical institution using the extracted word as a search query (step S202). Specifically, the control unit 11 searches a homepage, an article site, a review site, etc. where the surgical record of the medical institution is described.

[0071] The control unit 11 generates a list of medical institutions with surgical performance represented by the surgical plan information by inputting a prompt including the query generated in step S201 and the information retrieved in step S202 into the LLM 50 (step S203). The control unit 11 integrates the list of medical institutions retrieved from the database in step S16 and the list of medical institutions generated by the LLM 50 in step S203 (step S204). Specifically, the control unit 11 generates a list in which the medical institutions retrieved from the database are listed higher than the medical institutions listed by the LLM 50. The control unit 11 causes the list of medical institutions (or doctors) to be displayed on the list display screen (step S205), and ends a series of processes.

[0072] As described above, according to the second embodiment, a list can be suitably generated without creating a database of the surgical performance of each medical institution.

[0073] (Modification example) In the first embodiment, the form of presenting a user with a list of medical institutions (or doctors) with surgical performance represented by the surgical plan information has been described. In this modification example, the form of displaying each medical institution included in the list on a map image will be described.

[0074] FIG. 11 is a diagram showing an example of a map image. In this modification example, instead of or in addition to displaying a list of medical institutions in the list display column 64 (see FIG. 6) of the list display screen, a map image showing the positions of each medical institution with surgical performance is displayed.

[0075] Specifically, as shown in FIG. 11, the server 1 generates a map image indicating the positions of each medical institution included in the list by balloons and causes it to be displayed on the terminal 2. For example, in the balloon, in addition to the medical institution name and address, the surgical performance (the number of times the corresponding surgery has been performed) and the nearest available surgery date are displayed. In this way, the server 1 generates a map image in which the number of surgeries performed and the available surgery date are superimposed on the positions corresponding to each medical institution included in the list and causes it to be displayed on the terminal 2.

[0076] According to this modification example, it is possible to easily grasp the location of a medical institution with a record of surgical results for which a surgical plan has been formulated.

[0077] The embodiments disclosed this time should be considered as illustrative in all respects and not restrictive. The scope of the present invention is indicated by the claims, rather than the above description, and is intended to include all modifications within the meaning and scope equivalent to the claims.

[0078] The matters described in each embodiment can be combined with each other. In addition, the independent claims and dependent claims described in the claims can be combined with each other in all possible combinations regardless of the citation form. Furthermore, although the claims use a form (multi-claim form) of describing claims that cite two or more other claims, it is not limited to this. A form of describing a multi-claim (multi-multi-claim) that cites at least one multi-claim may be used.

Explanation of Reference Numerals

[0079] 1 Server (information processing device) 11 Control unit 12 Main memory unit 13 Communication unit 14 Auxiliary storage unit 141 Medical fee DB 142 Medical institution DB 143 Physician DB 144 Surgery schedule / history DB P Program 2 Terminal 3 Generation server 50 LLM (language model)

Claims

1. Obtain patient information including a medical image of the affected part of a patient in need of surgery, By inputting a prompt including the patient information into a language model, generate surgical plan information representing the surgery to be performed on the patient, Generate a query instructing to list medical institutions with experience in the surgery represented by the surgical plan information, By inputting a prompt including the query and information of each medical institution retrieved from the Internet based on the query into a language model, generate a list of medical institutions with experience in the surgery represented by the surgical plan information A program for causing a computer to execute the process.

2. Search a database storing the surgical performance of each medical institution or each doctor for a medical institution or doctor with experience in the surgery represented by the surgical plan information, Generate a list integrating the list of medical institutions generated using a language model and the list of medical institutions or doctors retrieved from the database The program according to claim 1.

3. Obtain the location information of the user, Based on the location information, generate the list indicating medical institutions within a predetermined distance from the user's location The program according to claim 1.

4. Obtain schedule information representing the surgical schedule of each medical institution, Based on the schedule information, generate the list indicating medical institutions capable of performing surgery within a predetermined period The program according to claim 1.

5. Display on a display unit a map image indicating the location of each medical institution included in the generated list The program according to claim 1.

6. Obtain patient information including a medical image of the affected part of a patient in need of surgery, By inputting a prompt including the patient information into a language model, generate surgical plan information representing the surgery to be performed on the patient, Generate a query instructing to list medical institutions with experience in the surgery represented by the surgical plan information, By inputting a prompt including the query and information of each medical institution retrieved from the Internet based on the query into a language model, generate a list of medical institutions with experience in the surgery represented by the surgical plan information An information processing method executed by a computer.

7. An information processing apparatus including a control unit, The control unit, Obtain patient information including a medical image of the affected part of a patient in need of surgery, By inputting a prompt including the patient information into a language model, surgical plan information representing a surgical operation to be performed on the patient is generated. A query is generated to instruct listing of medical institutions with experience in the surgical operation represented by the surgical plan information. By inputting a prompt including the query and information of each medical institution retrieved from the Internet based on the query into a language model, a list of medical institutions with experience in the surgical operation represented by the surgical plan information is generated. An information processing apparatus.

Images