Information processing method, non-transitory storage medium, and information processing device

Abstract

To provide an information processing method, a storage medium, and an information processing device that allow medical data to be efficiently utilized.SOLUTION: A method generates data for medical AI / drug discovery AI by causing a computer to execute: acquiring a plurality of pieces of image information in each of which personal information included in the image information has been subjected to anonymization processing; acquiring a plurality of pieces of clinical information that are obtained by anonymizing associated personal information for medical practice for a patient; selecting target image information suitable as the data for medical AI / drug discovery AI from among the plurality of pieces of image information; selecting target clinical information to be used as the data for medical AI / drug discovery AI from among the plurality of pieces of clinical information; associating disease information related to a disease and organ information related to an organ with the target image information and the target clinical information; and executing standardization on the target image information and the target clinical information.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The present disclosure relates to an information processing method, a non-transitory storage medium, and an information processing apparatus.

Background Art

[0002] Conventionally, techniques for processing medical-related data are known. For example, Patent Document 1 describes a system for processing medical information that detects the region where a medical institution that has acquired clinical data exists and sets confidential items based on the region. However, with the technique described in Patent Document 1, the utilization of medical-related data cannot be sufficiently streamlined.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] An object of the present disclosure is to improve the efficiency of utilization of medical-related data.

Means for Solving the Problems

[0005] An information processing method according to one aspect of the present disclosure causes a computer to acquire a plurality of pieces of image information, each of the plurality of pieces of image information being generated by a medical device having an imaging function and the personal information included in the image information being anonymized, acquire a plurality of pieces of clinical information regarding a patient's medical treatment, the plurality of pieces of clinical information being anonymized with respect to the associated personal information, select target image information appropriate as data for medical AI / drug discovery AI from the plurality of pieces of image information, select target clinical information to be used as data for medical AI / drug discovery AI from the plurality of pieces of clinical information, associate disease information and organ information regarding a disease or an organ with the target image information and the target clinical information as necessary, and perform standardization on the target image information and the target clinical information, thereby generating data for medical AI / drug discovery AI.

[0006] A non-transitory storage medium according to another aspect of the present disclosure stores a program that causes a computer to acquire a plurality of pieces of image information, each of the plurality of pieces of image information being generated by a medical device having an imaging function and the personal information included in the image information being anonymized, acquire a plurality of pieces of clinical information regarding a patient's medical treatment, the plurality of pieces of clinical information being anonymized with respect to the associated personal information, select target image information appropriate as data for medical AI / drug discovery AI from the plurality of pieces of image information, select target clinical information to be used as data for medical AI / drug discovery AI from the plurality of pieces of clinical information, associate disease information and organ information regarding a disease or an organ with the target image information and the target clinical information as necessary, and perform standardization on the target image information and the target clinical information, thereby generating data for medical AI / drug discovery AI.

[0007] An information processing apparatus according to another aspect of the present disclosure acquires a plurality of pieces of image information, each of the plurality of pieces of image information being generated by a medical device having an imaging function and the personal information included in the image information being anonymized, acquires a plurality of pieces of clinical information regarding the medical treatment of a patient, the plurality of pieces of clinical information being anonymized with respect to the associated personal information, selects appropriate target image information as data for medical AI / drug discovery AI from the plurality of pieces of image information, selects target clinical information to be used as data for medical AI / drug discovery AI from the plurality of pieces of clinical information, associates disease information and organ information regarding a disease or an organ with the target image information and the target clinical information as necessary, and performs standardization on the target image information and the target clinical information, and includes a computer that generates data for medical AI / drug discovery AI by performing these operations.

Advantages of the Invention

[0008] According to the present disclosure, the utilization of data in medical AI / drug discovery AI can be made more efficient.

Brief Description of the Drawings

[0009]

Figure 1

Figure 2

Figure 3

Figure 4

Embodiments for Carrying Out the Invention

[0010] <1. Overview> The system 1 according to this embodiment (hereinafter simply referred to as "system 1") is an all-in-one platform for research and development in the fields of medical and drug discovery image processing. The system 1 can generate high-quality learning / inference data for medical AI and drug discovery AI by performing (1) data collection, (2) data selection, (3) annotation as needed, and (4) standardization.

[0011] Referring to FIG. 1, an example of the operation of the system 1 will be described. The system 1 includes a data provider device 3, an information processing device 2, and a data user device 4. In one embodiment, the data provider device 3 is a server device managed by a data provider. The data provider is, for example, a medical institution including a hospital. In one embodiment, the data user device 4 is a server device managed by a data user. The data user is, for example, an AI development company, a medical device development company, a pharmaceutical company, and a research institution.

[0012] The information processing device 2 acquires a plurality of image information and a plurality of clinical information from the data provider device 3 (S1). The plurality of image information is generated by a medical device having an imaging function (hereinafter referred to as "medical imaging device"). The data provider device 3 may receive the image information from the medical imaging device and transmit the image information to the information processing device 2. The plurality of clinical information includes, for example, information related to the diagnosis and treatment of patients in a medical institution.

[0013] In one embodiment, the medical imaging device is a device that photographs the body of a patient or a subject (hereinafter referred to as "subject"). The medical imaging device may be, for example, a CT (Computed Tomography) scanner, an MRI (Magnetic Resonance Imaging) scanner, an X-ray scanner, an ultrasonic scanner, an endoscope scanner, a PET (Positron Emission Tomography) scanner, and a WSI (Whole Slide Imaging) scanner.

[0014] The information processing apparatus 2 generates data for medical AI / drug discovery AI by performing data selection, annotation, and standardization on a plurality of pieces of image information and a plurality of pieces of clinical information acquired from the data provider apparatus 3. The information processing apparatus 2 transmits the data for medical AI / drug discovery AI to the data user apparatus 4 (S2). Further, the information processing apparatus 2 can also provide the data user apparatus 4 with an AI model trained with the data for medical AI / drug discovery AI (S3).

[0015] According to the system 1, a data user can easily access the data for medical AI / drug discovery AI. In the system 1, the information processing apparatus 2 can also be said to be a data hub.

[0016] Hereinafter, the detailed configuration and operation of the system 1 will be described.

[0017] <2. Functional Configuration> With reference to FIG. 2, the functional configuration of the system 1 will be described. In the system 1, the information processing apparatus 2 is configured to be communicable with the data provider apparatus 3 and the data user apparatus 4 via the communication network 6.

[0018] [Information Processing Apparatus 2] The information processing apparatus 2 is a server apparatus that generates data for medical AI / drug discovery AI. The information processing apparatus 2 includes a control unit 10, a storage unit 12, a network interface unit 14, and a bus 16. The control unit 10, the storage unit 12, and the network interface unit 14 are electrically connected via the bus 16.

[0019] In the present disclosure, "an apparatus generates data or information" may be either that the control unit of the apparatus makes the data or information, which is a result of a predetermined operation, processable, or that the apparatus stores the data or information obtained as a result of a predetermined operation in the storage unit of the apparatus.

[0020] (Control Unit 10) The control unit 10 functions as an acquisition unit 100, an image processing unit 102, a learning unit 104, an output unit 106, and a determination unit 108 by executing various programs stored in a storage unit 12 described later.

[0021] - Acquisition Unit 100 - The acquisition unit 100 acquires a plurality of pieces of image information generated by a medical imaging device. The image information is, for example, an image file in DICOM (Digital Imaging and Communications in Medicine) format, NIfTI (Neuroimaging Informatics Technology Initiative) format, TIFF (Tagged Image File Format) format, SVS format, NDPI format, or the like.

[0022] The acquisition unit 100 may directly receive a plurality of pieces of image information from the medical imaging device, or may receive a plurality of pieces of image information stored in a server device communicable with the medical imaging device.

[0023] The acquisition unit 100 further acquires a plurality of pieces of clinical information related to the patient's medical treatment. In one embodiment, the acquisition unit 100 acquires clinical information directly or indirectly from a server managed by a medical institution including a hospital. The clinical information includes, for example, information related to age, gender, height, weight, past history / family history, surgical history, chief complaint, main symptoms / course, nursing observation records, referring department, diagnosis name, findings, and molecular diagnosis results. The clinical information is, for example, a file including text such as an EXCEL (registered trademark) file and a CSV file.

[0024] In the present disclosure, "acquiring information" includes making the information processable in the control unit 10. "Acquiring information" may be any of receiving the information from another device, reading the information from the storage unit 12, and obtaining the information as a result of a predetermined process.

[0025] - Image Processing Unit 102 - The image processing unit 102 generates data for medical AI / drug discovery AI from a plurality of pieces of image information. The image processing unit 102 includes an anonymization processing unit 102a, a selection unit 102b, an association unit 102c, and a normalization unit 102d.

[0026] ~Anonymization Processing Unit 102a~ The anonymization processing unit 102a applies anonymization processing to the personal information included in each of the plurality of pieces of image information acquired by the acquisition unit 100. Further, the anonymization processing unit 102a anonymizes the personal information associated with the plurality of pieces of clinical information acquired by the acquisition unit 100. Applying anonymization processing to the personal information included in the image information and anonymizing the personal information associated with the clinical information include making the personal information incomprehensible or difficult to understand for humans and computers. The anonymization processing unit 102a applying anonymization processing to the personal information included in the image information includes, for example, deleting, blurring, or changing the face information of an image including the face of the person who is the subject of the photograph to the face information of another person's face. The anonymization processing unit 102a anonymizing the personal information associated with the clinical information includes, for example, deleting the personal information, encrypting the personal information, and replacing the personal information with alternative text.

[0027] ~Selection Unit 102b~ The selection unit 102b selects target image information appropriate as data for medical AI / drug discovery AI from the plurality of pieces of image information acquired by the acquisition unit 100.

[0028] In one embodiment, the selection unit 102b selects target image information from a plurality of pieces of image information based on an operation on a terminal device used by a user (e.g., a doctor, etc.). In one embodiment, the terminal device displays the plurality of pieces of image information and receives an operation regarding the selection of image information appropriate as data for medical AI / drug discovery AI from the user. Thereafter, the terminal device transmits information regarding the selected image information to the information processing device 2. The selection unit 102b of the information processing device 2 receives the information and selects the image information selected by the user from among the plurality of pieces of image information as the target image information.

[0029] In one embodiment, the selection unit 102b selecting the target image information from a plurality of pieces of image information includes determining whether a lesion included in each of the plurality of pieces of image information includes a complication. The selection unit 102b may determine, for example, based on an input of a user who has visually inspected the image information, whether the lesion included in the image information includes a complication. That is, the selection unit 102b may select, as the target image information, the image information determined by the user not to include a complication among the plurality of pieces of image information. Further, the selection unit 102b may select, as the target image information, the image information in which a single disease is specified by a predetermined classification model among the plurality of pieces of image information. The classification model is, for example, a classification model based on SVM (Support Vector Machine), a classification model based on deep learning, a classification model based on a decision tree, or the like.

[0030] In addition, the selection unit 102b selects target clinical information to be used as data for medical AI / drug discovery AI from the plurality of pieces of clinical information acquired by the acquisition unit 100. In one example, when the clinical information includes a plurality of items and corresponding values, the selection unit 102b may select, as the target clinical information, the clinical information in which the value corresponding to a predetermined item is set from the plurality of pieces of clinical information. Here, when the clinical information includes information such as "age: 30 years old" and "diagnosis name: lung cancer", "age" and "diagnosis name" correspond to items, and "30 years old" and "lung cancer" correspond to the values corresponding to those items. The selection unit 102b selects, for example, when the predetermined items are "age" and "diagnosis name", from the plurality of pieces of clinical information, the clinical information in which the values corresponding to "age" and "diagnosis name" are set (that is, the values are not blank, empty, or default values, etc.) as the target clinical information. The predetermined items may be, for example, age, gender, height, weight, past history / family history, surgical history, chief complaint, main symptoms / course, nursing observation record, referring department, diagnosis name, findings, and molecular diagnosis results, etc.

[0031] ~Association unit 102c~ The association unit 102c associates disease information and organ information related to diseases and organs with the target image information and the target clinical information. The association of disease information by the association unit 102c with the target image information and the target clinical information may be to individually associate disease information with each of the target image information and the target clinical information, or may be to associate disease information with the set of the target image information and the target clinical information.

[0032] The association unit 102c may, for example, analyze a specific pattern or feature included in the target image information to associate disease information or organ information with the target image information. The analysis may be performed using a predetermined classification model. The association unit 102c may, for example, associate disease information or organ information with the target image information based on the input of a user who has visually inspected the target image information. Associating disease information or organ information with the target image information can also be referred to as annotation.

[0033] In one embodiment, the association unit 102c associating disease information or organ information with the target image information includes specifying the position of a lesion or organ structure included in the target image information. Specifying the position of a lesion or organ structure included in the target image information includes, in the target image information, drawing the contour of the organ structure where the lesion or disease is located, and coloring the location of the organ structure where the lesion or disease is located, and coloring the locations of each organ structure regardless of the lesion or disease, etc. Specifying the position of a lesion or organ structure included in the target image information includes filling in the position (or the region including the position) in the target image information. Specifying the position of a lesion or organ structure included in the target image information can also be referred to as segmentation.

[0034] In addition, in one embodiment, the associating unit 102c associating disease information or organ information with the target image information includes specifying the classification of the lesion or organ structure included in the target image information. Specifying the classification of the lesion includes, for example, specifying the name of the disease corresponding to the lesion and specifying the degree of progression of the lesion (e.g., cancer stage, etc.).

[0035] The associating unit 102c may associate disease information with the target clinical information based on at least a part of the target clinical information. For example, when the target clinical information includes findings, the associating unit 102c may specify disease information based on the findings and associate the specified disease information with the target clinical information. Also, when the target clinical information includes a diagnosis name, the associating unit 102c may associate the diagnosis name as disease information with the target clinical information.

[0036] In one embodiment, at least one of the association between the target image information and the disease information or organ information and the association between the target clinical information and the disease information or organ information can be double-checked by an expert. Specifically, the system 1 may be configured such that a first user associates disease information with the target image information and the target clinical information, and then a second user (e.g., an expert) having specialized knowledge compared to the first user verifies the validity of the association. In this case, the information processing apparatus 2 may be configured to generate data for medical AI / drug discovery AI based on the target image information and the target clinical information for which the association has been certified as valid by the second user.

[0037] ~Standardization unit 102d~ The standardization unit 102d performs standardization on the target image information and the target clinical information. The standardization for the target image information may include at least any one of noise removal, contrast adjustment, edge enhancement, and color tone correction. Further, the standardization for the target clinical information may include extracting a part of the data included in the clinical information (for example, extracting age, gender, height, weight, past history / family history, surgical history, chief complaint, main symptoms / course, nursing observation records, referring department, diagnosis name, radiology report, and molecular diagnosis results, etc. from the clinical information), and unifying its format and unit, etc. The standardization by the standardization unit 102d for the target image information and the target clinical information (especially the standardization for the target clinical information) is particularly important for achieving the purpose of collecting information from a plurality of medical institutions, etc. and promoting the utilization of that information. The standardization can also be called preprocessing or normalization.

[0038] The standardization unit 102d performing standardization on the target image information includes aligning a plurality of pieces of target image information generated by a plurality of different medical imaging devices photographing a common subject (which can also be called paired scanning). More specifically, when the first target image information generated by the first medical device and the second target image information generated by the second medical device are selected by the selection unit 102b from a plurality of pieces of image information, the standardization unit 102d performing standardization on the target image information includes aligning the first target image information and the second target image information. For example, when an MR image (corresponding to the first target image information) generated by an MRI scanner (corresponding to the first medical device) photographing a predetermined subject and a CT image (corresponding to the second target image information) generated by a CT scanner (corresponding to the second medical device) photographing the predetermined subject are selected by the selection unit 102b, the standardization unit 102d performing standardization on the target image information includes aligning the MR image and the CT image.

[0039] In the present disclosure, "aligning the first target image information and the second target image information" may be any of associating the second target image information with the first target image information, synthesizing the first target image information and the second target image information, and comparing the first target image information and the second target image information.

[0040] In addition, the standardization unit 102d performing standardization on the target image information includes aligning a plurality of pieces of target image information generated by a common medical imaging device photographing a common subject at different times. More specifically, when the first target image information generated based on the photographing of a predetermined medical device at a first time point and the second target image information generated based on the photographing of the predetermined medical device at a second time point different from the first time point are selected by the selection unit 102b from a plurality of pieces of image information, the standardization unit 102d performing standardization on the target image information includes aligning the first target image information and the second target image information.

[0041] In this way, by aligning a plurality of pieces of target image information, more useful data for medical AI / drug discovery AI can be generated. Specifically, a plurality of pieces of target image information generated by a plurality of different medical imaging devices photographing a common subject may have complementary properties to each other. Therefore, by aligning a plurality of pieces of target image information (for example, indicating that the plurality of pieces of target image information are a set), more suggestive data for medical AI / drug discovery AI can be generated.

[0042] Regarding aligning a plurality of pieces of target image information generated by a common medical imaging device photographing a common subject at different times, the plurality of pieces of target image information may show changes along a time series. Therefore, similar to the above, by aligning a plurality of pieces of target image information, more suggestive data for medical AI / drug discovery AI can be generated.

[0043] The image processing unit 102 generates data for medical AI / drug discovery AI by applying an anonymization unit 102a, a selection unit 102b, an association unit 102c, and a normalization unit 102d to a plurality of pieces of image information. That is, the data for medical AI / drug discovery AI includes information in which disease information and organ information are associated with target image information selected from a plurality of pieces of image information in which personal information has been anonymously processed, and which has been further normalized.

[0044] ―Learning unit 104― The learning unit 104 causes a learning model to learn the data for medical AI / drug discovery AI generated by the image processing unit 102. As a result, the learning unit 104 generates a medical AI / drug discovery AI model that has learned the data for medical AI / drug discovery AI. The learning model may be any of a neural network (including a transformer, a generative adversarial network, a convolutional neural network, etc.), an SVM, a decision tree, and the like.

[0045] ―Output unit 106― The output unit 106 outputs at least one of the data for medical AI / drug discovery AI obtained by the image processing unit 102 and the medical AI / drug discovery AI model obtained by the learning unit 104.

[0046] In the present disclosure, "outputting information" may be any of displaying the information, outputting the information as audio, and transmitting the information to another device. Also, in the present disclosure, a predetermined device transmitting information to another device is not limited to the predetermined device directly transmitting the information to the other device, and may be either transmitting the information to the other device via the communication network 6 or transmitting the information to the other device via a device that relays the information.

[0047] ―Decision unit 108― The determination unit 108 determines the data provision fee to be paid to the data provider. Specifically, when the output unit 106 transmits medical AI / drug discovery AI data or a medical AI / drug discovery AI model to the user, the determination unit 108 determines at least a part of the usage fee for the medical AI / drug discovery AI data or the medical AI / drug discovery AI model received from (or to be received from) the data user as the data provision fee to be paid to the data provider of the target image information.

[0048] When determining the data provision fee, the determination unit 108 can consider a plurality of factors. For example, the quantity and quality of the target image information provided by the data provider, the rarity of the target image information provided by the data provider (for example, the rarity of diseases and drugs included in the data), the presence, type, and quality of clinical information / annotations associated with the target image information, the number of times the medical AI / drug discovery AI data based on the target image information and clinical information provided by the data provider is used, the functionality of the medical AI / drug discovery AI model based on the medical AI / drug discovery AI data, the scale and / or commerciality of the data user (for example, whether the data user is a large enterprise, a small and medium-sized enterprise, a research facility, etc.), etc. Based on these, the data provision fee can be calculated. These factors may be appropriately adjusted to ensure the fairness and reasonableness of the data provision fee.

[0049] (Memory unit 12) The memory unit 12 stores various information that the information processing device 2 should store for operation. The memory unit 12 stores various programs executed by the control unit 10.

[0050] (Network interface unit 14) The network interface unit 14 realizes communication with other devices or systems via the communication network 6.

[0051] [Data provider device 3] In one embodiment, the data provider device 3 is a server device managed by a data provider. The data provider device 3 is configured to be communicable with a plurality of medical imaging devices. The data provider device 3 can transmit the image information received from the plurality of medical imaging devices to the information processing device 2 via the communication network 6.

[0052] [Data user device 4] In one embodiment, the data user device 4 is a server device managed by a data user. The data user device 4 can receive at least one of medical AI / drug discovery AI data and a medical AI / drug discovery AI model from the information processing device 2 via the communication network 6.

[0053] [Communication network 6] The communication network 6 realizes communication among the information processing device 2, the data provider device 3, and the data user device 4. The communication network 6 realizes communication based on, for example, the TCP / IP protocol. Although omitted in FIG. 2, the system 1 may include a plurality of data provider devices 3 and a plurality of data user devices 4. In this case, the information processing device 2 may acquire a plurality of pieces of image information from each of the plurality of data provider devices 3.

[0054] <3. Operations> With reference to FIG. 3, an example of the operation of the information processing device 2 will be described.

[0055] First, the information processing device 2 acquires a plurality of pieces of image information from the data provider device 3 (S100a). Next, the information processing device 2 performs anonymization processing on each of the plurality of pieces of image information (S102a). Next, the information processing device 2 selects target image information appropriate as medical AI / drug discovery AI data from the plurality of pieces of image information (S104a). Similarly, the information processing device 2 acquires a plurality of pieces of clinical information from the data provider device 3 (S100b), performs anonymization (S102b), and selects target clinical information (S104b). Next, the information processing device 2 associates disease information and organ information with the target image information and the target clinical information (S106). Next, the information processing device 2 performs normalization on the target image information (S108). Next, the information processing device 2 stores the medical AI / drug discovery AI data generated as a result of S108 in the storage unit 12 (S110). The information processing device 2 may transmit the stored medical AI / drug discovery AI data to the data user device 4 (S112a), or may use it to generate a medical AI / drug discovery AI model (S112b-1). When a medical AI / drug discovery AI model is generated, the medical AI / drug discovery AI model is transmitted to the data user device 4 (S112b-2).

[0056] <4. Hardware Configuration> Referring to FIG. 4, an example of a hardware configuration when the devices included in the above-described system 1 are realized by a computer 70 will be described. Note that the functions of each device can also be realized by dividing them among a plurality of devices.

[0057] As shown in FIG. 4, the computer 70 includes a processor 700, a storage device 702, an input I / F 704, a data I / F 706, a communication I / F 708, and a display device 710.

[0058] The processor 700 controls various processes in the computer 70 by executing programs stored in the storage device 702. For example, each functional unit included in the control unit 10 of the information processing device 2 can be realized by the processor 700 executing a program stored in the storage device 702.

[0059] The memory device 702 is a memory medium such as a RAM (Random Access Memory). The RAM temporarily stores the program code of the program executed by the processor 700 and the data required during the execution of the program.

[0060] The memory device 702 is also a non-volatile memory medium such as a hard disk drive (HDD) or a flash memory. The memory device 702 stores an operating system and various programs for implementing the above-described components. The memory medium storing the various programs may be a non-transitory computer readable medium. In addition, the memory device 702 can store a table for registering various information and a DB for managing the table. Such programs and data are referred to by the processor 700 by being loaded into the memory device 702 as needed.

[0061] The input I / F 704 is a device for receiving an input from a user. Specific examples of the input I / F 704 include a camera, buttons, a microphone, a keyboard, a mouse, a touch panel, various sensors, wearable devices, and the like. The input I / F 704 may be connected to the computer 70 via an interface such as USB (Universal Serial Bus).

[0062] The data I / F 706 is a device for inputting data from the outside of the computer 70. Specific examples of the data I / F 706 include a drive device for reading data stored in various memory media. The data I / F 706 may be provided outside the computer 70. In that case, the data I / F 706 is connected to the computer 70 via an interface such as USB.

[0063] The communication I / F 708 is a device for performing data communication via the communication network 6, either wired or wirelessly, with a device external to the computer 70. It is conceivable that the communication I / F 708 is provided outside the computer 70. In that case, the communication I / F 708 is connected to the computer 70 via an interface such as USB.

[0064] The display device 710 is a device for displaying various kinds of information. Specific examples of the display device 710 include, for example, a liquid crystal display, an organic EL (Electro-Luminescence) display, a display of a wearable device, and the like. The display device 710 may be provided outside the computer 70. In that case, the display device 710 is connected to the computer 70 via, for example, a display cable. Also, when a touch panel is adopted as the input I / F 704, the display device 710 can be configured integrally with the input I / F 704.

[0065] Moreover, the components included in the device of the system 1 described in the above embodiment are assumed to be such that, when a program stored in the storage device 702 is executed by the processor 700, the defined processing is realized in cooperation with other hardware. In other words, these components are assumed as software or firmware as well as the corresponding hardware, and in both concepts, they are also described as "function", "means", "section", "processing circuit", "unit", or "module", etc., and can be read as each other.

[0066] <5. Modification Example> The matters described in the above embodiment can be appropriately changed within a non - contradictory range.

[0067] The information processing apparatus 2 may further have a function of recording information related to the transmission and reception of medical AI / drug discovery AI data on a blockchain. The information related to the transmission and reception of medical AI / drug discovery AI data may be the medical AI / drug discovery AI data itself, the transmission and reception history of the medical AI / drug discovery AI data, or a non-fungible token (NFT) for accessing the medical AI / drug discovery AI data. The information processing apparatus 2 may, for example, store medical AI / drug discovery AI data in a blockchain so that the data user apparatus 4 can access the medical AI / drug discovery AI data on the blockchain. Further, the information processing apparatus 2 may, for example, transfer a non-fungible token to the data user apparatus 4 on the blockchain so that the data user apparatus can access the medical AI / drug discovery AI data corresponding to the non-fungible token. According to this configuration, the security regarding the use of medical AI / drug discovery AI data can be enhanced. In another embodiment, the information processing apparatus 2 may record medical AI / drug discovery AI data or information related to the transmission and reception thereof in IPFS (InterPlanetary File System).

[0068] <6. Embodiments of the Present Disclosure> The above embodiments are merely examples of the present disclosure and do not limit the scope of application of the present disclosure. The apparatuses included in the system 1 may be physically existing apparatuses or so-called virtual machines. The apparatuses included in the system 1 may be configured by a plurality of physically or logically separated apparatuses. For example, a part of the functions of the information processing apparatus 2 may be implemented in a first information processing apparatus, and another part may be implemented in a second information processing apparatus operable in cooperation with the first information processing apparatus. Further, the apparatuses included in the system 1 may be apparatuses on a cloud (i.e., computing resources available via a network).

[0069] The present disclosure includes, for example, the following embodiments. The corresponding configurations in the above embodiments are shown in parentheses.

[0070] [Appendix 1] An information processing method according to an aspect of the present disclosure includes causing a computer 70 to acquire a plurality of pieces of image information (acquisition unit 100), wherein each of the plurality of pieces of image information is generated by a medical device having an imaging function and the personal information included in the image information is anonymized; acquire a plurality of pieces of clinical information regarding a patient's medical treatment, wherein the plurality of pieces of clinical information have the associated personal information anonymized (anonymization processing unit 102a); select target image information appropriate as data for medical AI / drug discovery AI from the plurality of pieces of image information; select target clinical information to be used as data for medical AI / drug discovery AI from the plurality of pieces of clinical information (selection unit 102b); associate disease information and organ information regarding a disease or an organ with the target image information and the target clinical information (association unit 102c); and perform normalization on the target image information and the target clinical information (normalization unit 102d), thereby generating data for medical AI / drug discovery AI.

[0071] [Appendix 2] In the information processing method described in Appendix 1, associating disease information and organ information regarding a disease or an organ with the target image information (association unit 102c) may include specifying the position and / or classification of a lesion and / or organ structure included in the target image information.

[0072] [Appendix 3] In the information processing method described in any one of Appendices 1 to 2, selecting target image information from the plurality of pieces of image information (selection unit 102b) may be performed based on an operation on a terminal device used by a user.

[0073] [Appendix 4] In the information processing method described in any one of Appendices 1 to 3, selecting target image information from the plurality of pieces of image information (selection unit 102b) may include determining whether a lesion included in each of the plurality of pieces of image information includes a complication.

[0074] [Appendix 5] In the information processing method according to any one of Supplementary Notes 1 to 4, selecting target image information from a plurality of pieces of image information (selection unit 102b) may include selecting, from the plurality of pieces of image information, first target image information generated by a first medical device and second target image information generated by a second medical device (selection unit 102b), and performing normalization on the target image information (normalization unit 102d) may include aligning the first target image information and the second target image information.

[0075] [Supplementary Note 6] In the information processing method according to any one of Supplementary Notes 1 to 5, selecting target image information from a plurality of pieces of image information (selection unit 102b) may include selecting, from the plurality of pieces of image information, first target image information generated based on imaging at a first time point of a predetermined medical device and second target image information generated based on imaging at a second time point different from the first time point of the predetermined medical device (selection unit 102b), and performing normalization on the target image information (normalization unit 102d) may include aligning the first target image information and the second target image information.

[0076] [Supplementary Note 7] The information processing method according to any one of Supplementary Notes 1 to 6 may further cause the computer 70 to record information regarding transmission and reception of medical AI / drug discovery AI data on a blockchain.

[0077] [Supplementary Note 8] The information processing method according to any one of Supplementary Notes 1 to 7 may further cause the computer 70 to generate a learning model obtained by training medical AI / drug discovery AI data (learning unit 104).

[0078] [Supplementary Note 9] The information processing method according to any one of Supplementary Notes 1 to 8 causes the computer 70 to transmit medical AI / drug discovery AI data or a medical AI / drug discovery AI model to a user (output unit 106), and determines that at least a part of the usage fee for the medical AI / drug discovery AI data or the medical AI / drug discovery AI model received from the user is to be paid to the provider of the target image information as a provision fee (determination unit 108), and may execute the above.

[0079] [Supplementary Note 10] A non-transitory storage medium according to another aspect of the present disclosure causes the computer 70 to acquire a plurality of pieces of image information (acquisition unit 100), wherein each of the plurality of pieces of image information is generated by a medical device having an imaging function, and personal information included in the image information is anonymized, and to acquire a plurality of pieces of clinical information regarding a patient's medical treatment, wherein the plurality of pieces of clinical information are anonymized with respect to the associated personal information (anonymization processing unit 102a), to select target image information appropriate as medical AI / drug discovery AI data from the plurality of pieces of image information, to select target clinical information to be used as medical AI / drug discovery AI data from the plurality of pieces of clinical information (selection unit 102b), to associate disease information and organ information regarding a disease or an organ with the target image information and the target clinical information (association unit 102c), and to perform standardization on the target image information and the target clinical information (standardization unit 102d), thereby generating a program for generating medical AI / drug discovery AI data and storing the program.

[0080] [Supplementary Note 11] An information processing apparatus 2 according to another aspect of the present disclosure acquires a plurality of pieces of image information (acquisition unit 100), wherein each of the plurality of pieces of image information is generated by a medical device having an imaging function, and personal information included in the image information is anonymized (anonymization processing unit 102a). The apparatus also acquires a plurality of pieces of clinical information regarding patient treatment, wherein the plurality of pieces of clinical information have associated personal information anonymized. The apparatus selects appropriate target image information as data for medical AI / drug discovery AI from the plurality of pieces of image information, and selects target clinical information to be used as data for medical AI / drug discovery AI from the plurality of pieces of clinical information (selection unit 102b). The apparatus associates disease information and organ information regarding diseases and organs with the target image information and the target clinical information (association unit 102c), and performs standardization on the target image information and the target clinical information (standardization unit 102d). By performing these operations, the apparatus includes a computer 70 that generates data for medical AI / drug discovery AI.

Explanation of Signs

[0081] 1…System, 2…Information processing apparatus, 3…Data provider apparatus, 4…Data user apparatus, 6…Communication network, 10…Control unit, 12…Storage unit, 70…Computer, 100…Acquisition unit, 102…Image processing unit, 102a…Anonymization processing unit, 102b…Selection unit, 102c…Association unit, 102d…Standardization unit, 104…Learning unit, 106…Output unit, 108…Determination unit

Claims

1. On the computer, Acquiring a plurality of pieces of image information, each of the plurality of pieces of image information being generated by a medical device having an imaging function, and personal information included in the image information being anonymized; acquiring a plurality of clinical information related to a medical treatment of a patient, the plurality of clinical information having associated personal information anonymized; Selecting target image information appropriate as data for medical AI / drug discovery AI from the plurality of image information; Selecting target clinical information to be used as data for medical AI / drug discovery AI from the plurality of clinical information; Associating disease information and organ information related to a disease or an organ with the target image information and the target clinical information; performing standardization on the subject image information and the subject clinical information; An information processing method for generating data for medical AI / drug discovery AI by executing the above.

2. The information processing method according to claim 1 , wherein associating the disease information and organ information regarding a disease or an organ with the target image information includes identifying a position and / or a classification of a lesion and / or an organ structure included in the target image information.

3. The information processing method according to claim 1 , wherein the selection of the target image information from the plurality of pieces of image information is performed based on an operation on a terminal device used by a user.

4. The information processing method according to claim 1 , wherein selecting the target image information from the plurality of pieces of image information includes determining whether or not a lesion included in each of the plurality of pieces of image information includes a complication.

5. Selecting the target image information from the plurality of image information includes selecting, from the plurality of image information, first target image information generated by a first medical device and second target image information generated by a second medical device; The method of claim 1 , wherein performing standardization on the target image information comprises aligning the first target image information and the second target image information.

6. Selecting the target image information from the plurality of image information includes selecting, from the plurality of image information, first target image information generated based on imaging of a predetermined medical device at a first time point, and second target image information generated based on imaging of the predetermined medical device at a second time point different from the first time point; The method of claim 1 , wherein performing standardization on the target image information comprises aligning the first target image information and the second target image information.

7. The computer includes: The information processing method according to claim 1 , further comprising recording information regarding the transmission and reception of the medical AI / drug discovery AI data on a blockchain.

8. The computer includes: The information processing method according to claim 1 , further comprising: generating a learning model by learning the data for medical AI / drug discovery AI.

9. The computer includes: Transmitting the medical AI / drug discovery AI data to a user; determining at least a portion of the usage fee for the medical AI / drug discovery AI data received from the user as a provision fee to be paid to the provider of the target image information; The information processing method according to claim 1 , further comprising the step of:

10. On the computer, Acquiring a plurality of pieces of image information, each of the plurality of pieces of image information being generated by a medical device having an imaging function, and personal information included in the image information being anonymized; acquiring a plurality of clinical information related to a medical treatment of a patient, the plurality of clinical information having associated personal information anonymized; Selecting target image information appropriate as data for medical AI / drug discovery AI from the plurality of image information; Selecting target clinical information to be used as data for medical AI / drug discovery AI from the plurality of clinical information; Associating disease information and organ information related to a disease or an organ with the target image information and the target clinical information; performing standardization on the subject image information and the subject clinical information; A non-transitory storage medium that stores a program for generating data for medical AI / drug discovery AI by executing the program.

11. Acquiring a plurality of pieces of image information, each of the plurality of pieces of image information being generated by a medical device having an imaging function, and personal information included in the image information being anonymized; acquiring a plurality of clinical information related to a medical treatment of a patient, the plurality of clinical information having associated personal information anonymized; Selecting target image information appropriate as data for medical AI / drug discovery AI from the plurality of image information; Selecting target clinical information to be used as data for medical AI / drug discovery AI from the plurality of clinical information; Associating disease information and organ information related to a disease or an organ with the target image information and the target clinical information; performing standardization on the subject image information and the subject clinical information; An information processing device comprising a computer that generates data for medical AI / drug discovery AI by executing the above.

Images