Information processing system, information processing method, and program

Abstract

We provide information processing systems, information processing methods, and programs. [Solution] The information processing system 90 includes a metadata acquisition unit 910 that acquires metadata including information representing the image type of a medical image, an evaluation unit 930 that evaluates the information representing the image type based on the metadata, an identification unit 940 that identifies the image type of a medical image according to the evaluation result by the evaluation unit, and if the evaluation result does not satisfy predetermined conditions, an identification unit that identifies the image type based on specific related information including at least one of the information other than the information representing the image type included in the metadata and the medical image, and an association unit 950 that associates label information including the identified image type with the medical image.

Description

[Technical Field]

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

[0002] Patent Document 1 describes predicting metadata corresponding to a medical image by inputting the medical image and using a prediction model learned by a neural network. [Prior Art Document] [Patent Document] [Patent Document 1] Japanese Unexamined Patent Publication No. 2020-191063 [Summary of the Invention] [Means for Solving the Problems]

[0003] According to one embodiment of the present invention, an information processing system is provided. The information processing system may include a metadata acquisition unit that acquires metadata related to a medical image, the metadata including information representing an image type of the medical image. The information processing system may include an evaluation unit that evaluates information representing the image type based on the metadata. The information processing system may include a specifying unit that specifies the image type of the medical image according to an evaluation result by the evaluation unit, wherein when the evaluation result does not satisfy a predetermined condition, the specifying unit specifies the image type based on specification-related information including at least one of information other than the information representing the image type included in the metadata and the medical image. The information processing system may include an association unit that associates label information including the specified image type with the medical image.

[0004] In the information processing system, the evaluation unit may evaluate the appropriateness of the information representing the image type based on the metadata. In the information processing system, the specifying unit may specify the image type based on the specification-related information when the appropriateness does not satisfy a predetermined criterion.

[0005] In any of the aforementioned information processing systems, the evaluation unit may evaluate the appropriateness based on the metadata, based on at least one of the degree of certainty of the information representing the image type and the degree of conformity to the notation format of the information representing the image type.

[0006] In any of the aforementioned information processing systems, the identification unit may identify the image type based on the metadata and the identification-related information including the medical image if the degree of appropriateness does not meet a predetermined standard.

[0007] In any of the above-mentioned information processing systems, the identification unit may identify the image type based on the information representing the image type if the degree of appropriateness satisfies the criteria.

[0008] In any of the above-mentioned information processing systems, if the evaluation result does not satisfy the above-mentioned conditions, the identification unit may determine the scope of the identification related information used to identify the image type from among the identification related information that includes at least one of the information other than the information representing the image type included in the metadata and the medical image, based on the content of the deficiency identified in the evaluation by the evaluation unit.

[0009] In any of the aforementioned information processing systems, the evaluation unit may evaluate the information representing the image type by evaluating the consistency between the information representing the image type included in the metadata and at least one of the imaging parameters and the characteristics of the medical image included in the metadata.

[0010] In any of the above-mentioned information processing systems, if the inspection results obtained by a single inspection include multiple series, which are sets of one or more medical images obtained by a single measurement of a single type, the identification unit may identify the image type of each of the multiple series. In any of the above-mentioned information processing systems, if there are multiple series among the multiple series in which the identified image type is the same, the association unit may associate the label information with one of the multiple series according to the results of evaluating each of the multiple series, and may not associate the label information with the remaining series.

[0011] In any of the above information processing systems, the identification unit may identify the image type of a medical image and obtain the confidence level of the identified image type by inputting the medical image into a learning model that takes a medical image as input and outputs the image type of the medical image and the confidence level of the image type, which is generated by machine learning using a plurality of training data that associate medical images with corresponding image types. In any of the above information processing systems, the association unit may associate the label information, including the identified image type, with the medical image if the confidence level is greater than or equal to a predetermined threshold.

[0012] In any of the aforementioned information processing systems, the association unit may generate the label information in accordance with a naming rule that generates the label information such that the notation of the portion indicating the specified image type includes a common notation component among multiple medical images that share the specified image type.

[0013] In any of the aforementioned information processing systems, the association unit may associate the label information with the medical image in such a way as to anonymize the subject related to the medical image.

[0014] In any of the above information processing systems, the metadata acquisition unit may acquire metadata that includes multiple pieces of information other than direct identification information that can directly identify the subject, but which indicate the attributes of the subject. In any of the above information processing systems, the association unit may acquire identification information based on a combination of information indicating the attributes of the subject, in order to identify the same subject among multiple medical images, and include it in the label information.

[0015] Any of the above information processing systems may include a receiving unit that receives a user's modification instruction or confirmation operation for the specified image type. In any of the above information processing systems, the association unit may, when the receiving unit receives the modification instruction, associate the label information including the image type modified in accordance with the modification instruction with the medical image, and when the receiving unit receives the confirmation operation, associate the label information including the image type confirmed in accordance with the confirmation operation with the medical image.

[0016] In any of the above-mentioned information processing systems, the association unit may associate the label information with the medical image, in addition to the image type identified by the identification unit, the label information which includes information identified based on the metadata indicating at least one of the following: the area where the medical image was taken, whether or not a contrast agent was used in the measurement of the medical image, and the data format of the medical image.

[0017] In any of the above-mentioned information processing systems, the association unit may associate the label information with the medical image by replacing the information representing the image type included in the metadata with label information including the identified image type.

[0018] According to one embodiment of the present invention, there is provided an information processing method executed by a computer. The information processing method may comprise an evaluation step of acquiring metadata relating to a medical image including information representing an image type of the medical image, and evaluating the information representing the image type based on the metadata. The information processing method may comprise an identification step of identifying the image type of the medical image according to an evaluation result in the evaluation step, wherein when the evaluation result does not satisfy a predetermined condition, the image type is identified based on identification-related information including at least one of information other than the information representing the image type included in the metadata and the medical image. The information processing method may comprise an association step of associating label information including the identified image type with the medical image.

[0019] According to one embodiment of the present invention, there is provided a program for causing a computer to execute any one of the information processing methods described above.

[0020] It should be noted that the above summary of the invention does not list all of the necessary features of the present invention. Subcombinations of these feature groups may also constitute inventions. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] [Figure 1] An example of a related art is schematically shown. [Figure 2] An example of a related art is schematically shown. [Figure 3] An example of an information processing system 90 is schematically shown. [Figure 4] An example of metadata 200 is schematically shown. [Figure 5] An example of metadata 200 is schematically shown. [Figure 6] An example of an information processing system 90 and an example of a functional configuration of an information processing apparatus 900 are schematically shown. [Figure 7] An example of a processing flow by the information processing system 90 is schematically shown. [Figure 8]An example of a processing flow by the information processing system 90 is schematically shown. [Figure 9] An example of a processing flow by the information processing system 90 is schematically shown. [Figure 10] An example of a hardware configuration of a computer 1200 that functions as the information processing apparatus 900 or the server 909 is schematically shown. Mode for Carrying Out the Invention

[0022] Hereinafter, the present invention will be described through embodiments of the invention, but the following embodiments do not limit the invention according to the claims. In addition, not all combinations of features described in the embodiments are essential to the solution of the invention.

[0023] FIG. 1 schematically shows an example of a conventional technique. In the example shown in FIG. 1, a medical institution 80 includes a server 800 that stores and manages measurement results obtained from modalities such as MRI (Magnetic Resonance Imaging), X-ray CT (Computed Tomography), and plain X-ray imaging. The server 800 may be a medical image management apparatus. The server 800 is, for example, PACS (Picture Archiving and Communication System).

[0024] In the example shown in Figure 1, the server 800 stores the study results 10. The study results 10 may include a series, and the series may include medical images. In this example, one study result 10 includes a series 11, which is the result of an MRI T1-weighted sequence; a series 12, which is the result of an MRI T2-weighted sequence; a series 13, ..., which is the result of an X-ray CT arterial layer sequence. In this example, series 11 includes multiple medical images 101, medical image 102, ..., obtained by measuring an MRI T1-weighted sequence once. In this example, metadata 200 is associated with series 11. Metadata 200 may be associated with each of the medical images 101, medical image 102, ... Metadata 200 may be associated with the study results 10.

[0025] In this embodiment, "associating" may include logically linking multiple pieces of information such that when one piece of information is retrieved, referenced, etc., the other piece of information becomes callable. For example, "associating" may include embedding one piece of information into another, assigning a common index to both pieces of information, and maintaining mutual pointers to both pieces of information in a relational database.

[0026] In this embodiment, "associating" may include making it possible to directly or indirectly reference one piece of information from another piece of information. For example, "associating" may include uniquely linking multiple pieces of information in a database using information that can identify each of the pieces of information as a key. For example, associating a medical image 100 with metadata 200 may include linking the two by associating information that can identify the medical image 100 with information that can identify the metadata 200. Associating a medical image 100 with metadata 200 may also include associating a common key information with both the medical image 100 and the metadata 200.

[0027] In this embodiment, if no particular distinction is made between series 11, series 12, series 13, etc., they may simply be referred to as "series." In this embodiment, if no particular distinction is made between medical image 101, medical image 102, etc., they may simply be referred to as "medical image 100."

[0028] In the example shown in Figure 1, the test result 10 may be the result of tests performed per visit to the patient. The test result 10 may include one or more series.

[0029] In this embodiment, a series may be a set of one or more medical images included in a single imaging sequence in a measurement of a single modality. In the example shown in Figure 1, the multiple medical images included in a series may be a set of images including multiple two-dimensional images that are continuous in the thickness direction and acquired by tomography. For example, if a set of images including 15 two-dimensional images is acquired by performing tomography once in a T1-weighted sequence of head MRI with a slice thickness of 5 mm and a slice interval of 2 mm, then the series of the T1-weighted sequence of head MRI includes a set of images including 15 medical images.

[0030] In the example shown in Figure 1, the metadata 200 includes information 201 representing the image type, imaging parameters 202, ... Each of these may be metadata. Each of these may be associated with a series 11. Each of these may be associated with a medical image 100 included in the series 11. The following explanation will use the case where metadata 200 is associated with a medical image 100 as an example.

[0031] In this embodiment, metadata 200 may be a so-called DICOM tag defined in accordance with the DICOM (Digital Imaging and Communications in Medicine) standard, which is an international standard for medical images. A DICOM tag is an identifier for identifying individual information attributes associated with a medical image, and is represented by a combination of two hexadecimal numbers, such as "(0008,103E)". Specific examples of DICOM tags will be described later, but information 201 representing the image type is, for example, the Series Description (0008,103E) DICOM tag. Since text can be freely written in the Series Description, each medical institution, or each device or person in charge, may write text explaining the contents of the series in their own unique format.

[0032] In this embodiment, the image type may represent an imaging sequence. The image type may represent the specific imaging sequence, imaging conditions, and contrast agent usage for each modality. The image type may also include imaging sequences specialized for a particular area or analysis method. The following are examples of image types for MRI, X-ray CT, and plain X-ray radiography, but these are merely examples, and the type of modality and imaging sequence are not limited to these. In this embodiment, plain X-ray radiography refers to an imaging method in which radiation such as X-rays is irradiated onto a subject and the difference in transmitted dose is imaged, and may include the imaging style generally called radiography.

[0033] The image type may represent the MRI imaging sequence. For example, image types include T1-weighted images, T2-weighted images, diffusion-weighted imaging (DWI), FLAIR (Fluid Attenuated Inversion Recovery), T2*-weighted images, magnetic resonance angiography (MRA), and proton density-weighted images.

[0034] The image type may represent the imaging sequence of X-ray CT. For example, the image type may be a plain CT image (non-contrast image), arterial phase image, portal venous phase image, equilibrium phase (delayed phase) image, angiography image (CTA: Computed Tomography Angiography), bone-enhanced reconstruction image, lung field conditional reconstruction image, low-dose imaging image, etc.

[0035] The image type may represent a sequence of plain X-ray imaging. For example, the image type may be a frontal-anterior (PA) or anterior-posterior (AP) image, a lateral image, an oblique image, an inspiratory image, an expiratory image, a supine image, an upright image, etc.

[0036] In the example shown in Figure 1, the accuracy of the information 201 representing the image type included in the metadata 200 was not always guaranteed. For example, differences in modality equipment manufacturers, differences in models from the same manufacturer, differences in the physicians treating the patients, and differences in the technicians performing the imaging could cause inconsistencies in the text information recorded as the information 201 representing the image type, even for the same image type. For example, if the image type is a T1-weighted MRI image, one manufacturer's equipment might record "T1 SE" as the information 201 representing the image type, while another manufacturer's equipment might record "T1 RSSE".

[0037] Another case where the accuracy of the image type information 201 included in metadata 200 cannot be guaranteed is when, due to human error or other reasons, inappropriate information that differs from the actual imaging sequence is recorded as the image type information 201. For example, even if the actual imaging sequence is a T1-weighted MRI image, information indicating that it is a T2-weighted image may be mistakenly recorded in the free-text field of metadata 200.

[0038] If the information 201 representing the image type contains inconsistencies or errors, and its accuracy (degree of conformity to the notation format, or reliability of the information) is low, problems such as those listed below may occur. For example, when extracting a specific image type from server 800 later for use in clinical research, statistical analysis, and clinical diagnosis, it may become difficult to narrow down the search using keywords, or medical images of a different image type than the target image type may be included, reducing the accuracy of the analysis. To resolve this situation, it becomes necessary to manually check and sort each image from a vast amount of data.

[0039] Figure 2 schematically shows an example of the conventional technology. In the example shown in Figure 2, multiple medical institutions (medical institution 80, medical institution 81, ...) manage test results using their respective servers (server 800, server 801, ...). In this example, the servers of the multiple medical institutions are connected to each other via network 99 so that they can communicate with each other.

[0040] In the example shown in Figure 2, server 800 of medical institution 80 and server 801 of medical institution 81 store series 11, which is an MRI T1-weighted image. In this example, the series 11 on server 800 of medical institution 80 is associated with the text information "T1 SE" as information 201 representing the image type, while the series 11 on server 801 of medical institution 81 is associated with the text information "T1 RSSE," which is different from that of medical institution 80.

[0041] Thus, in conventional data management systems, even for substantially the same image type (imaging sequence), the content of the information 201 representing the image type (e.g., Series Description) could differ from one medical institution to another. This was due, for example, to differences in the equipment manufacturers and models used by each medical institution, or to differences in the documentation rules independently operated by each medical institution.

[0042] Such inconsistencies (variations in terminology) in information representing image types across different medical institutions become an obstacle when attempting to integrate and database a vast collection of medical images gathered from multiple institutions. For example, even when trying to extract "T1-weighted images" from all data for a specific pathological analysis, it becomes necessary to search using different keywords for each medical institution, and if unknown terminology is included, the search may fail. Therefore, in order to perform highly accurate statistical analysis or build datasets for machine learning using data from multiple medical institutions, it was necessary for someone with specialized knowledge to visually check the data from each institution and manually relabel it, among other post-processing steps.

[0043] Figure 3 schematically shows an example of an information processing system 90. In the example shown in Figure 3, the information processing system 90 includes an information processing device 900. The information processing system 90 may consist only of the information processing device 900, or it may consist of the information processing device 900 and other devices. For example, if the processing of the information processing device 900 is distributed among multiple devices, the information processing system 90 may be implemented by these multiple devices. The information processing system 90 may include a server 909. The information processing system 90 may include a user terminal 700. The information processing system 90 may include servers for each of multiple medical institutions (server 800, server 801, ...).

[0044] In the example shown in Figure 3, the information processing device 900 is a device for managing medical images. In this example, the information processing device 900 is connected to the server 800 of medical institution 80, the server 801 of medical institution 81, etc. via the network 99. The information processing system 90 may manage the medical images on these servers in an integrated manner.

[0045] In the example shown in Figure 3, the server 909 is connected to the information processing device 900 in a communicative manner. In this example, the server 909 is directly connected to the information processing device 900, but the server 909 may be located on the network 99 and connected to the information processing device 900 via the network 99. In this example, the server 909 may function as a data warehouse for the integrated management of medical images from multiple medical institutions. Furthermore, the information processing device 900 and the server 909 may be implemented using cloud computing.

[0046] In the example shown in Figure 3, network 99 may include a WAN (Wide Area Network). Network 99 may include the Internet. Network 99 may include a connection to a cloud computing environment. Network 99 may include a LAN (Local Area Network). Network 99 may include a dedicated network for medical institutions. Network 99 may include a mobile communication network. Network 99 may comply with 5G (5th Generation). Network 99 may comply with LTE (Long Term Evolution). Network 99 may comply with 6G (6th Generation) or later communication systems.

[0047] In the example shown in Figure 3, the user terminal 700 is connected to the information processing device 900, server 909, server 800, server 801, etc. via the network 99, enabling communication. The user 70 may use the user terminal 700 to send search requests, etc., to these devices and view the corresponding examination results 10, series, medical images 100, etc.

[0048] Examples of user terminals 700 include, but are not limited to, PCs (Personal Computers), tablet devices, smartphones, wearable devices such as smartwatches and smart glasses, and electronic devices with wireless capabilities. In the example shown in Figure 3, the user terminal 700 is a PC.

[0049] In the example shown in Figure 3, the information processing device 900 evaluates information 201 representing the image type associated with the medical image 101 based on the metadata 200 associated with the medical image 101, identifies the image type according to the evaluation result, and processes the identification of label information 30 including the identified image type 301 to newly associate with the medical image 101.

[0050] In the example shown in Figure 3, the information processing device 900 acquires metadata 200, which includes information 201 representing the image type of the medical image 101. In this example, "T1 SE" is acquired as the information 201 representing the image type. Based on the acquired metadata 200, the information processing device 900 evaluates the information 201 representing the image type. For example, the information processing device 900 evaluates whether the information 201 representing the image type is appropriate or not.

[0051] For example, the information processing device 900 evaluates the information 201 representing the image type based on the imaging parameters 202 included in the metadata 200. For example, the information processing device 900 evaluates the appropriateness based on the degree of certainty of the information 201 representing the image type. For example, the information processing device 900 evaluates the degree of certainty of the information 201 representing the image type based on the consistency between the information 201 representing the image type and the imaging parameters 202, and evaluates the appropriateness based on that degree.

[0052] In the example shown in Figure 3, the information processing device 900 may identify the image type of the medical image 101 according to the evaluation result of the information 201 representing the image type. For example, if the evaluation result does not satisfy predetermined conditions, the information processing device 900 may identify the image type based on specific related information that includes at least one of the information other than the information 201 representing the image type contained in the metadata 200, and the medical image 101. If the evaluation result satisfies the conditions, the information processing device 900 may identify the image type based on the information 201 representing the image type.

[0053] For example, if the evaluation result of the information representing the image type 201 does not indicate that the information representing the image type 201 is appropriate, the information processing device 900 identifies the image type based on the specific related information, and if the evaluation result indicates that the information representing the image type 201 is appropriate, it identifies the image type based on the information representing the image type 201.

[0054] For example, if the appropriateness of the information 201 representing the image type does not meet a predetermined standard, the information processing device 900 identifies the image type based on specific related information that includes at least one of the information other than the information 201 representing the image type contained in the metadata 200, and the medical image 101. If the appropriateness meets the standard, the information processing device 900 may identify the image type based on the information 201 representing the image type.

[0055] In the example shown in Figure 3, the information processing device 900 compares the information 201 representing the image type, "T1 SE," with the setting value of the MRI device included in the imaging parameter 202. The information processing device 900 may evaluate the appropriateness so that the more closely the setting value matches the setting for acquiring a T1-weighted image, the higher the appropriateness, and the less closely it matches, the lower the appropriateness. If the appropriateness is less than a predetermined threshold, the information processing device 900 may identify the image type based on specific related information. If the appropriateness is greater than a predetermined threshold, the information processing device 900 may identify the image type based on the information 201 representing the image type.

[0056] The information processing device 900 evaluates the appropriateness to be lower, for example, based on whether the setting value of the MRI device included in the imaging parameter 202 matches the setting value for acquiring a T1-weighted image to a low degree and matches the setting value for a T2-weighted image to a high degree. If the appropriateness is lower than a predetermined threshold, the information processing device 900 may identify the image type of the medical image 101 as a T2-weighted image according to the evaluation result of the appropriateness. The information processing device 900 may associate label information including the identified image type 301 (T2) with the medical image 101.

[0057] The information processing device 900 may evaluate the information representing the image type 201 using the condition that the text contained in the information representing the image type 201 matches the master data in which standard names are registered. For example, the information processing device 900 evaluates the information representing the image type 201 as appropriate if the text contained in the information representing the image type 201 matches the master data, and evaluates the information representing the image type 201 as inappropriate if they do not match.

[0058] In the example shown in Figure 3, the information processing device 900 may evaluate the appropriateness based on the degree of conformity of the notation format of the information 201 representing the image type. For example, the information processing device 900 evaluates the appropriateness to be greater when the notation format of the information 201 representing the image type conforms to a predetermined notation format to a greater extent. For example, if it is predetermined that the information 201 representing the image type for a T1-weighted image is to be represented as "T1", the information processing device 900 may evaluate the appropriateness to be high if the information 201 representing the image type is "T1", to be moderate if it contains "T1", as in "T1 SE" in this example, and to be low if it does not contain "T1" at all.

[0059] In the example shown in Figure 3, the information processing device 900 may evaluate the appropriateness based on the metadata 200, considering both the degree of certainty of the information 201 representing the image type and the degree of conformity of the information 201 representing the image type to the notation format. For example, the information processing device 900 evaluates the appropriateness to be higher when the information 201 representing the image type matches a predetermined notation format (e.g., "T1"), and imaging parameters 202 such as repetition time (TR) and echo time (TE) are within the standard setting range for the image type (T1-weighted image) indicated by the notation format.

[0060] In the example shown in Figure 3, if the information 201 representing the image type is "T1 SE", the information processing device 900 first evaluates the notation format by determining the degree of agreement with a predetermined regular expression or master data (e.g., "T1"), and calculates a moderate fit score because it is a partial match. Next, the information processing device 900 obtains the values ​​of the imaging parameters 202 as an evaluation of the likelihood and determines whether these match the numerical range of a typical T1-weighted image. For example, the information processing device 900 determines whether TR is less than or equal to several hundred ms and TE is less than or equal to several tens of ms. For example, the information processing device 900 may evaluate the final appropriateness as "high (meets the criteria)" if the score of the notation format is greater than or equal to a predetermined value and the imaging parameters match the numerical range of the corresponding image type, and as "low (does not meet the criteria)" otherwise.

[0061] In the example shown in Figure 3, the information processing device 900 may identify the image type of the medical image 101 based on the medical image 101. For example, the information processing device 900 analyzes the pixel data of the medical image 101 to extract brightness values ​​of the cerebrospinal fluid region and contrast between tissues, and identifies the image type based on these features. For example, the information processing device 900 identifies whether the medical image 101 is a T1-weighted image or a T2-weighted image. This makes it possible to accurately identify the image based on its actual characteristics even if the metadata 200 is missing or there are errors in the information of the imaging parameters 202.

[0062] In the example shown in Figure 3, if the information processing device 900 determines that the appropriateness does not meet the criteria and that the information 201 representing the image type is unreliable, it may acquire the pixel data of the medical image 101. The information processing device 900 may perform statistical analysis of the brightness values ​​and inference using a machine learning model on the acquired pixel data to extract the contrast trends of tissues in the image, such as cerebrospinal fluid and fat. Based on the fact that the extracted contrast trends match the characteristics of a T1-weighted image, such as "cerebrospinal fluid is low signal (black)," the information processing device 900 may identify the image type as "T1." This makes it possible to identify the image type based on the actual image, even if there is an error in the information 201 representing the image type and it is incorrectly labeled as T2 when it is actually T1.

[0063] In the example shown in Figure 3, the information processing device 900 may identify the image type of the medical image 101 based on both the metadata 200 and the medical image 101. For example, the information processing device 900 identifies the image type by comprehensively evaluating imaging conditions such as imaging parameters 202 obtained from the metadata 200 and image features obtained from the analysis of the pixel data of the medical image 101.

[0064] As a specific example, in MRI, the information processing device 900 may narrow down the candidate image types to "T1-weighted image" or "T2-weighted image" based on imaging parameters 202 such as TR and TE included in the metadata 200. The information processing device 900 may make the final determination by analyzing the medical image 101. The information processing device 900 may extract the signal intensity of the cerebrospinal fluid region from the pixel data and identify it as a "T1-weighted image" if the region has a low signal (black), and as a "T2-weighted image" if it has a high signal (white).

[0065] As another specific example, consider a case in X-ray CT where only "CT Abdomen (abdominal CT)" is recorded in the information 201 representing the image type, and the contrast-enhanced phase (plain, arterial phase, portal venous phase, etc.) is unknown. In this case, the information processing device 900 may first obtain a tag indicating the elapsed time after contrast agent injection and information on the injection protocol from the imaging parameters 202 included in the metadata 200. The information processing device 900 may analyze the pixel data of the medical image 101 and calculate the brightness values ​​of specific anatomical structures in the image (e.g., the aorta or portal vein). The information processing system 90 may compare the "elapsed time after injection" obtained from the metadata 200 with the "brightness values ​​of major blood vessels" obtained from the image analysis. For example, based on the fact that the elapsed time is within the range of the arterial phase, and the brightness value of the aorta is high above a threshold, indicating a significant contrast effect, the information processing device 900 may identify the image type of the medical image 101 as an "arterial phase image". In this way, by complementaryly utilizing objective imaging conditions from metadata 200 and visual real-world information from medical images 101, it becomes possible to identify image types with high reliability, even when identification would be difficult using only one of the pieces of information.

[0066] When the information processing device 900 identifies an image type based on information 201 representing the image type, the information processing device 900 identifies the image type included in the information 201 representing the image type as the image type 301 of the medical image 101. For example, if the information 201 representing the image type is "T1 SE", the information processing device 900 may identify the image type 301 as "T1 SE". The information processing device 900 may also identify the image type 301 as "T1".

[0067] In the example shown in Figure 3, the information processing device 900 associates the medical image 101 with the label information 30 by having the server 909 store information indicating that the medical image 101 and the label information 30 are in a corresponding relationship. For example, the information processing device 900 generates label information 30 including a specified image type 301, and link information indicating that the medical image 101 and the label information 30 are in a corresponding relationship, separately from the actual data of the medical image 101, and stores these in the server 909. For example, the information processing device 900 generates indexes or pointers that can identify the medical image 101 and the label information 30 as link information and stores these in the server 909. This relational data management structure allows the medical image 101 and the label information 30 to be logically linked in the database without directly modifying the file itself of the medical image 101.

[0068] The information processing device 900 may associate the medical image 101 with the label information 30 in other ways. For example, the information processing device 900 may associate the medical image 101 with the label information 30 by adding the label information 30 to the metadata 200 of the medical image 101, or by adding the label information 30 to a free-text field such as a private tag included in the metadata 200 of the medical image 101.

[0069] The information processing device 900 may associate the identified image type 301 with the medical image 101. For example, the information processing device 900 may associate the identified image type 301 with the medical image 101 by: linking the identified image type 301 to the medical image 101 and storing it in the server 909; adding the identified image type 301 to the metadata 200 of the medical image 101; adding the identified image type 301 to the free description field of the metadata 200 of the medical image 101; or replacing the information 201 representing the image type in the metadata 200 of the medical image 101 with the identified image type 301.

[0070] The information processing device 900 may store multiple medical images and label information 30 in the server 909. For example, the information processing device 900 may acquire multiple medical images stored in the servers of multiple medical institutions, associate label information 30 with each of the multiple medical images, and store the multiple medical images and label information 30 in the server 909. The information processing device 900 may store the multiple medical images themselves on the servers of the medical institutions and not store them in the server 909. For example, the information processing device 900 may store information representing the relationship between multiple medical images and the corresponding label information 30, along with the label information 30, in the server 909, but not the multiple medical images themselves. The information processing device 900 may store some of the multiple medical images in the server 909 and leave others out. The information processing device 900 may decide whether or not to store medical images in the server 909 on a case-by-case basis, such as for each medical institution or each image type. The information processing device 900 may decide whether or not to store the medical image 101 in the server 909, based on the evaluation result of the information 201 representing the image type corresponding to the medical image 101.

[0071] Figure 4 schematically shows an example of metadata 200. In the example shown in Figure 4, the types of metadata 200 and the content of the information contained in metadata 200 are explained. In this example, metadata 200 includes patient information, examination information, device information, imaging parameters, and image information. In this example, patient information includes name, individual identification information, date of birth, sex, age, and medical history. Examination information may include the date and time of the examination and information indicating the modality (MRI, X-ray CT, plain X-ray, etc.). Device information may include the manufacturer name and model name of the device.

[0072] In the example shown in Figure 4, the imaging parameters include items common to all modalities and items specific to each modality. In this example, items common to all modalities that acquire tomographic images include slice thickness, slice interval, and slice position. Among the items specific to each modality, MRI items may include imaging sequence (T1-weighted, T2-weighted, FLAIR, etc.), TR value, TE value, FOV, matrix size, gain, brightness level, filter settings, and information on injected contrast agent. X-ray CT items may include X-ray exposure time, X-ray tube voltage, X-ray tube current, imaging range, imaging angle, representative values ​​of exposure dose such as CTDIvol (Computed Tomography Dose Index volume), gain, brightness level, filter settings, and information on injected contrast agent.

[0073] In the example shown in Figure 4, the image information includes the size and resolution of the image. Although the pixel information of the medical image 101 may sometimes be treated as image information in the metadata 200, for the sake of explanation in this embodiment, the pixel information of the medical image 101 will be treated and explained as the medical image 101 itself, rather than as part of the metadata 200.

[0074] The example shown in Figure 4 is merely an example, and the metadata 200 is not limited to the example shown in Figure 4. The metadata 200 may include other types of information different from those shown in Figure 4, and each type may include other information different from those shown in Figure 4. The information processing device 900 may use each of these diverse pieces of metadata 200 for evaluation of the information 201 representing the image type, and for identifying the image type 301, etc.

[0075] Figure 5 schematically shows an example of metadata 200. In the example shown in Figure 5, the classification of metadata 200 and specific examples of DICOM tags are explained.

[0076] In the example shown in Figure 5, the DICOM tag classified as series identification information includes Series Description (0008,103E), which is text information describing the contents of the series; Series Instance UID (0020,000E), which uniquely identifies the series; Study Instance UID (0020,000D), which uniquely identifies the study; and Series Number (0020,0011), which indicates the series number. The series identification information may correspond to the information 201 representing the image type in this embodiment. The Series Description may correspond to the information 201 representing the image type in this embodiment.

[0077] In the example shown in Figure 5, the DICOM tags classified as imaging conditions include Repetition Time TR (0018,0080) indicating the repetition time, Echo Time TE (0018,0081) indicating the echo time, Inversion Time TI (0018,0082) indicating the inversion time, Flip Angle (0018,1314) indicating the flip angle, and Diffusion b-value (0018,9087) indicating the diffusion b-value. The imaging conditions may correspond to the imaging parameters 202 in this embodiment.

[0078] In the example shown in Figure 5, the DICOM tags classified as imaging protocol information include Protocol Name (0018, 1030) indicating the protocol name and Sequence Name (0018, 0024) indicating the sequence name. In this example, the DICOM tags classified as imaging sequence and device information include Acquisition Number (0020, 0012) indicating the acquisition number, Instance Number (0020, 0013) indicating the instance number, Manufacturer (0008, 0070) indicating the device manufacturer's name, and Scanning Sequence (0018, 0020) indicating the format of the imaging sequence.

[0079] The example shown in Figure 5 is merely one example, and the classification of metadata 200 and the DICOM tags included in each classification are not limited to the example shown in Figure 5. The information processing device 900 may use each of these diverse pieces of metadata 200 for evaluation of information 201 representing the image type, and for identification of the image type 301, etc.

[0080] Figure 6 schematically shows an example of an information processing system 90 and an example of the functional configuration of an information processing device 900. The differences between the example shown in Figure 6 and the example shown in Figure 3 will be explained in detail. Figure 6 schematically shows an example of the functional configuration of an information processing device 900.

[0081] In the example shown in Figure 6, the information processing device 900 includes a metadata acquisition unit 910, an image acquisition unit 920, an evaluation unit 930, a specification unit 940, an association unit 950, a reception unit 960, an output control unit 970, a model generation unit 980, and a storage unit 990. It is not essential that the information processing device 900 includes all of these. For example, the information processing device 900 may not include the image acquisition unit 920, the reception unit 960, and the model generation unit 980.

[0082] The metadata acquisition unit 910 acquires metadata 200 related to the medical image 101. The metadata 200 includes information 201 representing the image type of the medical image 101.

[0083] The image acquisition unit 920 acquires a medical image 101.

[0084] The evaluation unit 930 evaluates the information 201 representing the image type based on the metadata 200. The evaluation unit 930 may evaluate whether the information 201 representing the image type based on the metadata is appropriate. The evaluation unit 930 may evaluate the information 201 representing the image type by evaluating the consistency between the information 201 representing the image type included in the metadata 200 and at least one of the imaging parameters 202 and the characteristics of the medical image 101 included in the metadata 200.

[0085] The evaluation unit 930 may evaluate the appropriateness of the information 201 representing the image type based on the metadata. The evaluation unit 930 may evaluate the appropriateness based on the metadata 200, based on at least one of the degree of certainty of the information 201 representing the image type and the degree of conformity of the information 201 representing the image type to the notation format. The evaluation unit 930 may evaluate the appropriateness by evaluating the consistency between the information 201 representing the image type included in the metadata 200 and at least one of the imaging parameters 202 and the features of the medical image 101 included in the metadata 200.

[0086] The evaluation unit 930 may evaluate the image type information 201 using the condition that the text contained in the image type information 201 matches the master data in which standard names are registered. For example, the evaluation unit 930 evaluates the image type information 201 as appropriate if the text contained in the image type information 201 matches the master data, and evaluates the image type information 201 as inappropriate if they do not match.

[0087] The evaluation unit 930 may evaluate the information representing the image type 201 as appropriate if the similarity between the text contained in the information representing the image type 201 and the master data is equal to or greater than a predetermined threshold, and may evaluate the information representing the image type 201 as inappropriate if it is less than the threshold.

[0088] The evaluation unit 930 may evaluate the image type information 201 by comprehensively evaluating its consistency with other information. For example, the evaluation unit 930 evaluates the consistency between the image type information 201 and the imaging parameters 202. For example, if the image type information 201 indicates a "T1-weighted image," and the echo time (TE) included in the imaging parameters 202 is set to a long time (e.g., 80ms or more) commonly used in T2-weighted images, the evaluation unit 930 will determine that there is a contradiction between the two and evaluate the image type information 201 as inappropriate. The evaluation unit 930 may also evaluate the appropriateness of the image type information 201 as low.

[0089] The evaluation unit 930 may evaluate the consistency between the image type information 201 and the medical image 101. For example, if the image type information 201 indicates "contrast-enhanced CT," and the analysis of the pixel data of the medical image 101 reveals that the brightness values ​​of anatomical structures such as the heart and major blood vessels are within the standard range for non-contrast imaging, the evaluation unit 930 will determine that the consistency is low because no contrast effect is observed, and will evaluate the image type information 201 as inappropriate. The evaluation unit 930 may also evaluate the appropriateness of the image type information 201 as low.

[0090] The evaluation unit 930 may evaluate the consistency between the three elements: the information 201 representing the image type, the imaging parameters 202, and the medical image 101. For example, if the information 201 representing the image type indicates "diffusion-weighted image (DWI)," and the diffusion B value (b-value) of the imaging parameters 202 is set to "0" or an extremely low value, the evaluation unit 930 will evaluate the image as appropriate as a DWI if the medical image 101 is part of a group of images composed of multiple images with different b values ​​(e.g., high b-value images) within the same imaging series. On the other hand, if (1) the medical image 101 exists alone in the absence of other images with different b values, or (2) despite the imaging parameter 202 being set to a sufficiently high b value, no signal suppression of the water component is observed in the pixel data of the medical image 101, or the pixel data does not show signal characteristics corresponding to the set b value, the evaluation unit 930 may determine that these are mutually contradictory and evaluate the appropriateness of the information 201 representing the image type as low.

[0091] In this way, the evaluation unit 930 does not evaluate the information 201 representing the image type in isolation, but rather compares it with other items in the metadata 200 and the actual image, thereby enabling accurate detection of information deficiencies caused by, for example, human input errors or device setting errors.

[0092] The identification unit 940 identifies the image type of the medical image 101 according to the evaluation result by the evaluation unit 930. If the evaluation result by the evaluation unit 930 does not meet predetermined conditions, the identification unit 940 identifies the image type 301 of the medical image 101 based on the identification-related information. The identification-related information includes at least one of the information other than the information representing the image type included in the metadata 200 and the medical image 101.

[0093] The identification unit 940 may identify the image type of the medical image 101 based on the identified related information if the evaluation unit 930 evaluates that the information 201 representing the image type is not appropriate. The identification unit 940 may identify the image type based on the identified related information including the metadata 200 and the medical image 101 if the evaluation result by the evaluation unit 930 does not meet predetermined conditions. The identification unit 940 may determine the scope of the identified related information used to identify the image type based on the content of the deficiencies identified in the appropriateness evaluation if the evaluation result by the evaluation unit 930 does not meet predetermined conditions.

[0094] The identification unit 940 may identify the image type based on the identified related information if the degree of appropriateness does not meet predetermined criteria. The identification unit 940 may identify the image type based on the identified related information including metadata 200 and medical image 101 if the degree of appropriateness does not meet predetermined criteria. The identification unit 940 may identify the image type based on the information 201 representing the image type if the degree of appropriateness meets the criteria. The identification unit 940 may determine the scope of the identified related information used to identify the image type based on the content of the deficiencies identified in the evaluation of the degree of appropriateness if the degree of appropriateness does not meet the criteria.

[0095] The identification unit 940 may determine the range of identification-related information used to identify the image type 301 depending on the nature of the deficiency in the information 201 representing the image type. The nature of the deficiency in the information 201 representing the image type may be the result of an evaluation of appropriateness. For example, if the information 201 representing the image type partially matches a known expression such as "T1 SE" and it is determined that only the notation format is deficient, the identification unit 940 will identify the image type by referring only to specific imaging parameters 202 from the metadata 200, such as TR value and TE value, without using the medical image 101. This allows for the rapid identification of the correct label while avoiding image analysis, which has a high processing load. If the information 201 representing the image type is blank, or if there is a suspected error that contradicts the actual imaging content (i.e., the appropriateness is extremely low), the identification unit 940 may identify the image type by combining a part of the metadata 200 (imaging conditions and device information) with the pixel data of the medical image 101.

[0096] In cases of specific deficiencies in CT images, such as when only the presence or absence of contrast agent use is unknown, the identification unit 940 may identify the deficiency by referring only to information contained in specific DICOM tags related to contrast agent injection, rather than referring to all items in the metadata 200. For example, the identification unit 940 may refer to values ​​contained in DICOM tags representing the name of the contrast agent, the total amount of contrast agent injected, the time when the injection of the contrast agent started, the time when the injection of the contrast agent ended, and the name of the active ingredient of the contrast agent. By limiting the referenced information according to the nature of the deficiency in this way, the computational resources of the entire information processing device 900 can be optimized, enabling high-speed batch processing of a large number of medical images.

[0097] The association unit 950 associates the label information 30, which includes the image type 301 identified by the identification unit 940, with the medical image 101. The association unit 950 may generate the label information 30 according to a naming rule that generates the label information 30 so that the notation of the part indicating the identified image type 301 includes common notation components among multiple medical images that share the identified image type. The association unit 950 may generate the label information 30 according to a naming rule that generates the label information 30 so that the label information 30 includes common notation components corresponding to the identified image type among multiple medical images that share the identified image type. For example, the association unit 950 may generate label information 30 that not only has common particles and symbols such as hyphens, but also includes common notation components corresponding to the identified image type. Assume that the identified image type 301 is a "T1-weighted image". For example, even if the information 201 representing the original image type uses different notations such as "T1_SE", "T1-weighted", and "T1WI", the association unit 950 may consistently generate label information 30 (e.g., "Standard_T1") that includes "T1", a common notation component, for these, in accordance with the naming convention.

[0098] If the image type 301 includes whether or not a contrast agent was used, the association unit 950 may generate label information 30 according to a naming convention in which images with contrast are commonly assigned the notation component "CE (Contrast Enhanced)" and images without contrast are assigned the notation component "Plain". This allows, for example, medical images collected from different medical institutions or devices to be searched collectively using the same keyword if they are of the same image type.

[0099] The association unit 950 may follow a naming convention that generates structured text containing, for example, modality, imaging site, image type, and presence or absence of contrast using delimiters such as underscores. For example, according to such a naming convention, structured text containing the text MRI_HEAD_T1_CE is generated. In this way, by using hierarchical common notation components, it becomes possible to assign a consistent, cross-sectional index based on rules to image groups that were previously managed with different names at each facility.

[0100] The association unit 950 may associate the label information 30 with the medical image 101 in such a way as to anonymize the subject related to the medical image 101. For example, the metadata acquisition unit 910 may acquire metadata 200 which includes multiple pieces of information other than direct identification information that can directly identify the subject, but which indicates the attributes of the subject, and the association unit 950 may acquire identification information based on a combination of the information indicating the attributes of the subject, in order to identify the same subject among multiple medical images, and include it in the label information 30. For example, the association unit 950 may acquire identification information based on a combination of information such as date of birth, sex, blood type, place of birth, birth weight, height, and anatomical features such as organs and skeleton, and include it in the label information 30.

[0101] This improves the possibility that, for example, the association unit 950 can accurately link a group of medical images of the same subject, scattered across different medical institutions, as a series of time-series data, without acquiring direct identification information such as names or My Number, and without being affected by life events such as changes of address. This makes it possible, for example, to perform prognosis analysis and disease progression prediction while protecting the privacy of the subject.

[0102] The association unit 950 may associate label information 30 with the medical image 101, in addition to the image type 301 identified by the identification unit 940, which includes information identified based on the metadata 200 indicating at least one of the imaging site of the medical image 101, whether or not a contrast agent was used in the measurement of the medical image 101, and the data format of the medical image 101. This improves the searchability and extraction accuracy of medical image data, for example, because the label information 30 includes at least one of the imaging site, whether or not a contrast agent was used, and the data format. For example, in the creation of training data for specific clinical research or machine learning, even inexperienced individuals can easily and accurately collect the optimal image set that matches the purpose from a vast amount of data. Furthermore, because supplementary information that was not standardized across different medical institutions or devices is labeled in a standardized format, the efficiency of data integration is improved, for example.

[0103] The association unit 950 may associate the label information 30 with the medical image 101 by replacing the information 201 representing the image type contained in the metadata 200 with the label information 30 which includes the identified image type 301. The association unit 950 may replace the information 201 representing the original image type contained in the metadata 200 with the label information 30 which includes the appropriate image type 301 identified by the identification unit 940. As a result, for example, information 201 representing the image type, such as a Series Description which may contain defects or inconsistencies in notation, may be replaced with the identified image type 301.

[0104] This means that, for example, even if the original metadata 200 contains errors or inaccurate representations, it will be overwritten with standardized, accurate label information 30. As a result, subsequent systems and users can handle the data based on the correct image type without being confused by information inconsistencies. Furthermore, by optimizing information within the framework of existing DICOM tags while suppressing increases in data volume, it is possible to improve data quality while maintaining compatibility with existing PACS and other systems.

[0105] The association unit 950 may associate the label information 30, which includes the identified image type 301, with the medical image 101 by recording the label information 30 in an area of ​​a different item from the information 201 representing the image type included in the metadata 200. For example, the association unit 950 controls the free-text field included in the metadata 200 of the medical image 101 to describe the identified image type 301 and some or all of the rest of the label information 30. For example, the association unit 950 controls the private tag in the metadata 200 to describe the identified image type 301 and some or all of the rest of the label information 30.

[0106] This allows, for example, the recording of standardized label information 30 in a separate free-text field while retaining the image type information 201 contained in the original metadata 200 without overwriting it. This ensures information traceability while improving data quality. As a result, compatibility with existing systems is maintained, and the original input information remains accessible, allowing for stable and highly accurate filtering and data extraction optimized for specific research or analysis purposes without interfering with existing metadata items.

[0107] The association unit 950 may associate the label information 30, including the identified image type 301, with the medical image 101 by storing the label information 30 independently of the metadata 200. For example, the association unit 950 may associate the label information 30 with the medical image 101 by storing the label information 30 and relational information that associates the label information 30 with the medical image 101 in a storage area separate from the server, such as the server 800, where the medical image 101 is stored. For example, the association unit 950 may associate the label information 30 with the medical image 101 by having the medical image 101 stored in the server 800 and having the label information 30 and relational information that associates the label information 30 with the medical image 101 stored in the server 909.

[0108] This allows label information 30 to be managed independently in an external database without rewriting the medical image 101 itself or its metadata 200. Therefore, even in existing system environments that require read-only storage or strict change management, information can be aggregated without affecting the original data. This enables centralized cross-searching and advanced filtering using relational information for image sets distributed across multiple different servers, allowing for the construction of a large-scale and secure data warehouse.

[0109] When the test results obtained in a single test include multiple series, which are sets of one or more medical images obtained in a single measurement of one type, the identification unit 940 may identify the image type of each of the multiple series. If there are multiple series among the multiple series in which the identified image type is the same, the association unit 950 may associate label information 30 with one of the multiple series, depending on the results of evaluating each of the multiple series, and may not associate label information 30 with the remaining series.

[0110] A single test result 10 may include multiple series. A series may be a set of one or more medical images 101 obtained by a single measurement of one type. The identification unit 940 may identify the image type of each of these multiple series.

[0111] In some cases, multiple series may exist within a single series that share the same identified image type 301. For example, if imaging is repeated, multiple "T1-weighted image" series may be mixed together in a single inspection result 10. In such cases, the association unit 950 may, based on the evaluation results of each of the multiple series with the same image type, associate label information 30 with one of these series (representative series) and not associate label information 30 with the remaining series.

[0112] The evaluation unit 930 may evaluate each series identified as belonging to the same type, assessing the appropriateness of the consistency between metadata and imaging parameters, as well as the quality of the image itself (low noise, presence or absence of artifacts, number of slices, etc.), and determine that the series with the highest evaluation is the appropriate series for analysis, etc. The evaluation unit 930 may also compare the acquisition times of multiple series and determine that the most recently acquired series is the appropriate series.

[0113] The evaluation unit 930 may quantitatively evaluate multiple series of the same image type based on the number of slices included in the image information and the amount of noise such as artifacts obtained from the analysis of the pixel data. The evaluation unit 930 may preferentially select a series that shows a higher degree of consistency between the imaging conditions included in the metadata 200 and the anatomical features extracted from the pixel data of the actual medical image 101 as a valid series suitable for analysis.

[0114] The association unit 950 may associate label information 30 with a series that has been evaluated as appropriate by the evaluation unit 930. On the other hand, for other series with low evaluation results, the association unit 950 may consider them as incidental data or unnecessary duplicate data, and may not associate label information 30 with them, excluding them from registration in the database. When the association unit 950 replaces the information 201 representing the image type with label information 30, the association unit 950 may process the information 201 representing the image type for other series with low evaluation results in a manner that restricts the use of the medical image 101 in subsequent processing. For example, the association unit 950 may delete the information 201 representing the image type and replace it with label information 30 indicating that the quality of the medical image 101 is unsuitable for use in subsequent processing. This makes it possible to build a database that has cleansed low-quality and unnecessary data, for example. Furthermore, it is possible to suppress bias due to duplicate data in secondary uses such as statistical analysis and machine learning, and improve learning efficiency.

[0115] The reception unit 960 may accept various requests. For example, the reception unit 960 may accept requests from users. For example, the reception unit 960 may accept display requests, search requests, and data update requests from users via a user terminal 700 or the like. The reception unit 960 may also accept correction instructions and confirmation operations from users regarding the identification results of image type 301. When the reception unit 960 accepts a correction instruction from a user, the association unit 950 may associate the label information 30, which includes the image type corrected in accordance with the correction instruction, with the medical image. When the reception unit 960 accepts a confirmation operation, the association unit 950 may associate the label information 30, which includes the image type confirmed in accordance with the confirmation operation, with the medical image.

[0116] The output control unit 970 controls the output of various types of information. For example, the output control unit 970 may control the output of requested information based on a user request received by the reception unit 960. For example, the output control unit 970 may control the output of requested medical images 101, metadata 200, and label information 30, etc. The output of information by the output control unit 970 may include display output to a display or the like provided by the information processing device 900, transmission output to a user terminal 700, and print output to paper or the like.

[0117] The output control unit 970 may be controlled to output information prompting the user to confirm the likelihood of the identified image type based on the confidence level of the image type identified by the identification unit 940. For example, if the confidence level is lower than a predetermined threshold, the output control unit 970 may output a message prompting the user to confirm the likelihood of the image type identified by the identification unit 940. The identification unit 940 may output the confidence level value to the user. The reception unit 960 may receive a response from the user to the information prompting confirmation. For example, the reception unit 960 may receive the aforementioned correction instructions or confirmation operations from the user.

[0118] The output control unit 970 may output the medical image 101, to which the label information 30 has been associated by the association unit 950, to an external device such as a server 909. The output control unit 970 may also control the output of this information along with a message prompting the user to confirm the appropriateness calculated by the evaluation unit 930, the candidate image type 301 identified by the identification unit 940, etc.

[0119] The identification unit 940 may use a learning model that takes a medical image as input and outputs the image type of the medical image and the confidence level of the image type, which is generated by machine learning using multiple training data that associate medical images with corresponding image types. The identification unit 940 may input the medical image 101 into the learning model to identify the image type of the medical image 101 and obtain the confidence level of the identified image type. The association unit 950 may associate label information 30, which includes the identified image type, with the medical image 101 if the confidence level is above a predetermined threshold. This effectively reduces the risk of incorrect label information 30 being associated when identifying the image type of the medical image 101 by image analysis using machine learning, by performing threshold processing based on the output confidence level. This makes it possible to extract and store only data with a certain level of accuracy, even for medical images 100 with significant metadata deficiencies or case images that are difficult to distinguish, thereby reducing the cost of manual verification and enabling the construction of a highly reliable data warehouse.

[0120] If the confidence level of the image analysis using machine learning is below a threshold, the association unit 950 does not need to associate label information 30 with the medical image 101. In such cases, the association unit 950 may associate label information 30 containing information indicating that it should be ignored with the medical image 101. The association unit 950 may also associate label information 30 indicating that the confidence level is below a threshold, or label information 30 indicating the confidence level value, with the medical image 101.

[0121] If the confidence level is below a threshold, the association unit 950 may associate the medical image with specific flags or label information 30 indicating that the medical image is "of unknown type" or "excluded from analysis." This makes it possible to filter out unreliable data in a subsequent data warehouse, and maintain a cleansed, high-quality dataset without requiring manual verification of all items.

[0122] The model generation unit 980 may generate various types of learning models. For example, the model generation unit 980 may take a medical image as input and generate a learning model that outputs the image type of the medical image and the confidence level of that image type. The model generation unit 980 may generate the learning model by machine learning using multiple learning data sets that associate medical images with corresponding image types. The model generation unit 980 may update the learning model by machine learning using learning data that includes information on the image type identified by the identification unit 940. The model generation unit 980 may update the learning model by machine learning using learning data that includes information on the image type modified by the user from the information on the image type identified by the identification unit 940. The model generation unit 980 may update the learning model by machine learning using learning data that includes information on the image type confirmed by the user from the information on the image type identified by the identification unit 940.

[0123] The storage unit 990 stores various types of information. For example, the storage unit 990 stores label information 30. The storage unit 990 may also retrieve and store information stored on the medical institution's server. For example, the storage unit 990 may further store at least one of the following: test results, series, medical images, and metadata. The storage unit 990 may control the server 909 to store this information. The server 909 may be the storage unit 990.

[0124] Figure 7 schematically shows an example of the processing flow by the information processing system 90. In the example shown in Figure 7, the process is described from when the information processing system 90 acquires metadata 200 including the image type of the medical image 100, to when it evaluates the appropriateness of the information 201 representing the image type, and when it associates label information 30 with the medical image 100 and outputs it.

[0125] In step 102 (sometimes abbreviated as S), the metadata acquisition unit 910 acquires metadata 200, which includes information 201 representing the image type of the medical image 100.

[0126] In S104, the image acquisition unit 920 acquires a medical image 100. The medical image 100 may be pixel data. S104 is not mandatory. The order in which S104 is executed is not limited to the order shown in the example in Figure 7. For example, S104 may be executed only if it is determined that the medical image 100 is necessary based on the evaluation result of S108 described later, and may be executed between S108 and S112.

[0127] In S106, the evaluation unit 930 evaluates the appropriateness of the information 201 representing the image type based on the metadata 200 acquired by the metadata acquisition unit 910 in S102. The evaluation unit 930 may evaluate the appropriateness based on at least one of the degree of certainty of the information 201 representing the image type and the degree of conformity of the information 201 representing the image type to the notation format.

[0128] In S108, based on the evaluation result by the evaluation unit 930 in S106, it is determined whether the information 201 representing the image type satisfies predetermined criteria. The predetermined criteria may include, for example, the degree of agreement between the text contained in the information 201 representing the image type and a known regular expression. The predetermined criteria may also include the consistency between the content of the information 201 representing the image type and other information such as imaging parameters 202. If the determination result is YES, the process proceeds to S110; otherwise, it proceeds to S112. This determination may be performed by the evaluation unit 930.

[0129] In S110, the identification unit 940 identifies the image type based on the information 201 representing the image type. This allows for quick identification with low processing load, provided the metadata is reliable.

[0130] In S112, the identification unit 940 identifies the image type 301 of the medical image 100 based on identification-related information that includes at least one of the information 201 representing the image type included in the metadata 200 and the medical image 101. Depending on the nature of the deficiency, the identification unit 940 may perform the identification with optimal accuracy and computational cost by limiting the range of metadata items to be referenced or by combining it with the analysis of pixel information.

[0131] In S114, the association unit 950 associates the label information 30, which includes the image type 301 identified in S110 or S112, with the medical image 100. The association unit 950 may generate the label information 30 and attach it to the medical image 100. The association unit 950 may perform the association by updating the existing metadata 200, adding missing information, or registering the medical image 100 and the label information 30 in a relational relationship.

[0132] In S116, the output control unit 970 controls the output of the medical image 100 based on the label information 30 associated in S114. For example, the output control unit 970 performs a search by comparing the search word received from the user with the information representing the image type 301 included in the label information 30, and controls the output of the medical image 100 if a match is found in the search. The output control unit 970 may also control the output to output the result of the search, without outputting the medical image 100. The output control unit 970 may also control the output to output metadata 200 associated with the medical image 100 that was found in the search.

[0133] Figure 8 schematically shows an example of the processing flow by the information processing system 90. In the example shown in Figure 8, the information processing system 90 acquires metadata 200 including the image type of the medical image 100, evaluates the appropriateness of the information 201 representing the image type, and associates label information 30 with the medical image 100 only if it does not meet predetermined criteria.

[0134] In S202, the metadata acquisition unit 910 acquires metadata 200, which includes information 201 representing the image type of the medical image 101. S202 may be the same as S102 in the example shown in Figure 7.

[0135] In S204, the image acquisition unit 920 acquires a medical image 100. S204 may be the same as S104 in the example shown in Figure 7.

[0136] In S206, the evaluation unit 930 evaluates the appropriateness of the information 201 representing the image type based on the metadata 200 acquired by the metadata acquisition unit 910 in S202. S206 may be the same as S106 in the example shown in Figure 7.

[0137] In S208, it is determined whether the information 201 representing the image type meets predetermined criteria, based on the evaluation result by the evaluation unit 930 in S206. S208 may be the same as S108 in the example shown in Figure 7. However, if the determination result is YES, the process proceeds to S216; if NO, it proceeds to S212.

[0138] In S212, the identification unit 940 identifies the image type 301 of the medical image 100 based on identification related information which includes at least one of the information other than the information 201 representing the image type included in the metadata 200, and the medical image 100. S212 may be the same as S112 in the example shown in Figure 7.

[0139] In S214, the association unit 950 associates the label information 30, which includes the image type 301 identified in S212, with the medical image 100. Here, the association unit 950 may generate the label information 30 and attach it to the medical image 100. The association unit 950 associates the label information 30 with the medical image 100 by replacing the information 201 representing the image type with the label information 30 which includes the identified image type 301. The association unit 950 may replace the information 201 representing the image type with the image type 301 included in the label information 30.

[0140] In S216, the output control unit 970 controls the output to output the medical image 100. If the determination result in S208 is YES, the output control unit 970 controls the output to output the medical image 100 based on the original image type 301. If the determination result in step S208 is NO, the output control unit 970 controls the output to output the medical image 100 based on the image type 301 included in the label information 30 obtained in S214 when the association unit 950 replaces the information 201 representing the image type.

[0141] Figure 9 schematically shows an example of the processing flow by the information processing system 90. In the example shown in Figure 9, the information processing system 90 acquires metadata 200 including the image type of the medical image 100, and then identifies the image type 301 based solely on the metadata 200 without pre-evaluating the information 201 representing the image type, and associates the label information 30 with it.

[0142] In S302, the metadata acquisition unit 910 acquires metadata 200, which includes information 201 representing the image type of the medical image 100. S302 may be the same as S102 in the example shown in Figure 7.

[0143] In S304, the identification unit 940 identifies the image type of the medical image 100 based on the metadata 200 acquired by the metadata acquisition unit 910 in S302. The identification unit 940 identifies the image type of the medical image 100 based, for example, on imaging parameters 202 included in the metadata 200. The identification unit 940 may also identify the image type of the medical image 100 based solely on the metadata 200 without using the medical image 100.

[0144] In S306, the association unit 950 associates the label information 30, which includes the image type 301 identified in S304, with the medical image 100. S306 may be the same as S114 in the example shown in Figure 7.

[0145] In S308, the output control unit 970 controls the output of the medical image 100 based on the label information 30 associated in S306. S308 may be the same as S116 in the example shown in Figure 7.

[0146] Figure 10 schematically shows an example of the hardware configuration of a computer 1200 that functions as an information processing device 900 or a server 909. A program installed on the computer 1200 can cause the computer 1200 to function as one or more "parts" of the apparatus according to this embodiment, or to cause the computer 1200 to execute operations associated with the apparatus according to this embodiment or such one or more "parts", and / or to cause the computer 1200 to execute a process or a stage of such process according to this embodiment. Such a program may be executed by the CPU 1212 to cause the computer 1200 to execute specific operations associated with some or all of the blocks in the flowcharts and block diagrams described herein.

[0147] The computer 1200 according to this embodiment includes a CPU 1212, a GPU 1213, RAM 1214, and a graphics controller 1216, which are interconnected by a host controller 1210. The computer 1200 also includes input / output units such as a communication interface 1222, a storage device 1224, a DVD drive, and an IC card drive, which are connected to the host controller 1210 via an input / output controller 1220. The DVD drive may be a DVD-ROM drive and a DVD-RAM drive, etc. The storage device 1224 may be a hard disk drive and a solid-state drive, etc. The computer 1200 also includes legacy input / output units such as a ROM 1230 and a keyboard, which are connected to the input / output controller 1220 via an input / output chip 1240.

[0148] The CPU 1212 operates according to the programs stored in the ROM 1230 and RAM 1214, thereby controlling each unit. The graphics controller 1216 acquires the image data generated by the CPU 1212 and stores it in the frame buffer provided in RAM 1214 or within itself, so that the image data is displayed on the display device 1218.

[0149] The communication interface 1222 communicates with other electronic devices via a network. The storage device 1224 stores programs and data used by the CPU 1212 in the computer 1200. The DVD drive reads programs or data from a DVD-ROM or the like and provides them to the storage device 1224. The IC card drive reads programs and data from an IC card and / or writes programs and data to an IC card.

[0150] The ROM 1230 stores boot programs and / or hardware-dependent programs of the computer 1200, which are executed by the computer 1200 upon activation. The input / output chip 1240 may also connect various input / output units to the input / output controller 1220 via USB ports, parallel ports, serial ports, keyboard ports, mouse ports, etc.

[0151] The program is provided on a computer-readable storage medium such as a DVD-ROM or IC card. The program is read from the computer-readable storage medium and installed on a storage device 1224, RAM 1214, or ROM 1230, which are examples of computer-readable storage media, and executed by the CPU 1212. The information processing described within these programs is read by the computer 1200, resulting in coordination between the program and the various types of hardware resources described above. The apparatus or method may be configured to realize the operation or processing of information in accordance with the use of the computer 1200.

[0152] For example, when communication is performed between a computer 1200 and an external device, the CPU 1212 may execute a communication program loaded into RAM 1214 and, based on the processing described in the communication program, instruct the communication interface 1222 to perform communication processing. Under the control of the CPU 1212, the communication interface 1222 reads transmission data stored in a transmission buffer area provided in a recording medium such as RAM 1214, storage device 1224, DVD-ROM, or IC card, transmits the read transmission data to the network, or writes received data received from the network to a reception buffer area provided on the recording medium.

[0153] Furthermore, the CPU 1212 may read all or necessary parts of a file or database stored on an external recording medium such as the storage device 1224, a DVD drive (DVD-ROM), or an IC card into the RAM 1214, and perform various types of processing on the data in the RAM 1214. The CPU 1212 may then write the processed data back to the external recording medium.

[0154] Various types of information, such as various types of programs, data, tables, and databases, may be stored on the recording medium and subjected to information processing. The CPU 1212 may perform various types of processing on the data read from RAM 1214, including various types of operations, information processing, conditional judgments, conditional branching, unconditional branching, information retrieval / replacement, etc., as described throughout this disclosure and specified by the program instruction sequence, and write the results back to RAM 1214. The CPU 1212 may also retrieve information in files, databases, etc., within the recording medium. For example, if multiple entries are stored in the recording medium, each having an attribute value of a first attribute associated with an attribute value of a second attribute, the CPU 1212 may search among the multiple entries for an entry that matches the specified condition for the attribute value of the first attribute, read the attribute value of the second attribute stored in that entry, and thereby obtain the attribute value of the second attribute associated with the first attribute that satisfies the predetermined condition.

[0155] The program or software module described above may be stored on or near the computer 1200 in a computer-readable storage medium. Alternatively, a recording medium such as a hard disk or RAM provided within a server system connected to a dedicated communication network or the Internet can be used as a computer-readable storage medium, thereby providing the program to the computer 1200 via the network.

[0156] In this embodiment, blocks in the flowchart and block diagram may represent a stage in a process in which an operation is performed or a "part" of a device that has the role of performing an operation. A particular stage and "part" may be implemented by a dedicated circuit, a programmable circuit supplied with computer-readable instructions stored on a computer-readable storage medium, and / or a processor supplied with computer-readable instructions stored on a computer-readable storage medium. The dedicated circuit may include digital and / or analog hardware circuits, and may include integrated circuits (ICs) and / or discrete circuits. The programmable circuit may include reconfigurable hardware circuits, such as field-programmable gate arrays (FPGAs) and programmable logic arrays (PLAs), which include logical AND, logical OR, exclusive OR, negated AND, negated OR, and other logical operations, flip-flops, registers, and memory elements.

[0157] A computer-readable storage medium may include any tangible device capable of storing instructions that can be executed by a suitable device, and as a result, a computer-readable storage medium having instructions stored therein will comprise a product that includes instructions that can be executed to create means for performing operations specified in a flowchart or block diagram. Examples of computer-readable storage media may include electronic storage media, magnetic storage media, optical storage media, electromagnetic storage media, semiconductor storage media, etc. More specific examples of computer-readable storage media may include floppy disks, diskettes, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), electrically erasable programmable read-only memory (EEPROM), static random access memory (SRAM), compact disk read-only memory (CD-ROM), digital multipurpose disc (DVD), Blu-ray® disc, memory stick, integrated circuit card, etc.

[0158] Computer-readable instructions may include assembler instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk®, Java®, C++, and traditional procedural programming languages ​​such as the C programming language or similar programming languages.

[0159] Computer-readable instructions may be provided locally or via a wide area network (WAN) such as a local area network (LAN) or the internet to a processor or programmable circuit of a general-purpose computer, a special-purpose computer, or another programmable data processing device, so that the processor or programmable circuit of the programmable data processing device, such as a computer, may execute the instructions to generate means for performing operations specified in a flowchart or block diagram. Here, the computer may be a PC (personal computer), a tablet computer, a smartphone, a workstation, a server computer, a general-purpose computer, or a special-purpose computer, and may also be a computer system in which multiple computers are connected. Such a computer system in which multiple computers are connected is also called a distributed computing system and is a computer in a broad sense. In a distributed computing system, multiple computers execute a program collectively by each computer executing a part of the program and passing data during program execution between computers as needed.

[0160] Examples of processors include computer processors, central processing units (CPUs), processing units, microprocessors, digital signal processors, controllers, and microcontrollers. A computer may have one or more processors. In a multiprocessor system with multiple processors, each processor executes a portion of the program, and the processors collectively execute the program by passing program execution data between them as needed. For example, in the execution of multitasking, each of the multiple processors may execute a portion of each task in small chunks by switching tasks at each time slice. In this case, which part of a program each processor executes changes dynamically. Which part of a program each of the multiple processors executes may also be statically determined by multiprocessor-aware programming.

[0161] Although the present invention has been described above using embodiments, the technical scope of the present invention is not limited to the scope described in the above embodiments. It will be apparent to those skilled in the art that various modifications or improvements can be made to the above embodiments. It will be clear from the claims that such modified or improved forms may also be included in the technical scope of the present invention.

[0162] It should be noted that the execution order of operations, procedures, steps, and stages in the devices, systems, programs, and methods shown in the claims, specifications, and drawings is not explicitly stated as "before," "prior to," etc., and that these can be implemented in any order unless the output of a previous process is used in a later process. Even if the operation flow in the claims, specifications, and drawings is described using phrases such as "first," "next," etc. for convenience, it does not mean that it is essential to perform the operations in that order. [Explanation of Symbols]

[0163] 10 Inspection results, 11 Series, 12 Series, 13 Series, 30 Label information, 70 User, 80 Medical institution, 81 Medical institution, 90 Information processing system, 99 Network, 100 Medical image, 101 Medical image, 102 Medical image, 200 Metadata, 201 Information representing image type, 202 Imaging parameters, 301 Image type, 700 User terminal, 800 Server, 801 Server, 900 Information processing device, 909 Server, 910 Metadata acquisition unit, 920 Image acquisition unit, 930 Evaluation unit, 940 Identification unit, 950 Association unit, 960 Reception unit, 970 Output control unit, 980 Model generation unit, 990 Storage unit, 1200 Computer, 1210 Host controller, 1212 CPU, 1213 GPU, 1214 RAM, 1216 Graphics controller, 1218 Display device, 1220 I / O controller, 1222 communication interface, 1224 storage device, 1230 ROM, 1240 I / O chip

Claims

1. A metadata acquisition unit acquires metadata relating to a medical image, which includes information representing the image type of the medical image and parameters of the medical image. An evaluation unit that evaluates information representing the image type based on the metadata, A specification unit for identifying the image type of the medical image according to the evaluation result of the evaluation unit, wherein if the evaluation result does not satisfy predetermined conditions, the specification unit identifies the image type based on specification-related information including at least one of the information other than the information representing the image type included in the metadata and the medical image, An association unit that associates label information including the identified image type with the medical image. Equipped with, The metadata acquisition unit further acquires parameters of other medical images included in the same image group as the medical image, The evaluation unit is an information processing system that evaluates information representing the image type of the medical image based on whether the parameters of the medical image and the parameters of the other medical image are of the same type and have different values ​​from each other.

2. The information representing the image type of the medical image is information indicating that the image type of the medical image is a diffusion-weighted image, The parameters of the aforementioned medical image and the parameters of the other medical image each include the diffusion b value. The information processing system according to claim 1, wherein the evaluation unit evaluates information representing the image type of the medical image based on the fact that the medical image is included in the same group of images as the other medical images having a diffusion b value different from the diffusion b value of the medical image.

3. The diffusion b value of the medical image is 0 or low, The diffusion b value of the other medical image is higher than the diffusion b value of the aforementioned medical image. The information processing system according to claim 2, wherein the evaluation unit evaluates the information representing the image type of the medical image as appropriate, based on the fact that the medical image is included in the same image group as the other medical images, even if the diffusion b value of the medical image is 0 or low.

4. The system further comprises an output control unit that receives a search word, searches for the medical image associated with the label information corresponding to the search word, and controls the system to output at least one of the medical image that matches the search and metadata related to the medical image as a search result, The information processing system according to claim 2, wherein when the output control unit outputs the medical image associated with the label information as the search result, it controls the output control unit to further output other medical images included in the same group of images as the medical image as the search result.

5. The identification unit identifies the image type of the medical image by inputting the medical image into a learning model, which is generated by machine learning using multiple training data that associate medical images with corresponding image types, and which takes a medical image as input and outputs the image type of the medical image and the confidence level of the image type. The identification unit also obtains the confidence level of the identified image type. The information processing system according to any one of claims 1 to 4, wherein the association unit associates the label information, including the image type identified when the confidence level is above a predetermined threshold, with the medical image.

6. The association unit generates the label information according to a naming rule that generates the label information such that the notation of the portion indicating the specified image type includes a common notation component among multiple medical images that share the specified image type, as described in any one of claims 1 to 4.

7. The information processing system according to any one of claims 1 to 4, wherein the association unit associates the label information with the medical image in such a way as to anonymize the subject related to the medical image.

8. The metadata acquisition unit acquires metadata that includes multiple pieces of information other than direct identification information that can directly identify the subject, but which indicate the attributes of the subject. The information processing system according to claim 7, wherein the association unit acquires identification information that enables the identification of the same subject among a plurality of medical images based on a combination of information indicating the attributes of the subject, and includes it in the label information.

9. The system further includes a reception unit that receives user requests for modification or confirmation operations for the specified image type, The information processing system according to any one of claims 1 to 4, wherein the association unit associates the label information, including the image type modified in accordance with the modification instruction, with the medical image when the reception unit receives the modification instruction, and associates the label information, including the image type confirmed in accordance with the confirmation operation, with the medical image when the reception unit receives the confirmation operation.

10. The information processing system according to any one of claims 1 to 4, wherein the association unit associates the label information with the medical image by replacing the information representing the image type contained in the metadata with label information including the identified image type.

11. A metadata acquisition unit that acquires metadata relating to a medical image, which includes information representing the image type of the medical image, An evaluation unit that evaluates information representing the image type based on the metadata, A specification unit for identifying the image type of the medical image according to the evaluation result of the evaluation unit, wherein if the evaluation result does not satisfy predetermined conditions, the specification unit identifies the image type based on specification-related information including at least one of the information other than the information representing the image type included in the metadata and the medical image, An association unit that associates label information including the identified image type with the medical image. Equipped with, The identification unit, when the evaluation result does not satisfy the conditions, determines the range of the specific related information used to identify the image type from the information other than the information representing the image type included in the metadata and the medical image, based on the content of the deficiencies identified in the evaluation by the evaluation unit.

12. A metadata acquisition unit that acquires metadata relating to a medical image, which includes information representing the image type of the medical image, An evaluation unit that evaluates information representing the image type based on the metadata, A specification unit for identifying the image type of the medical image according to the evaluation result of the evaluation unit, wherein if the evaluation result does not satisfy predetermined conditions, the specification unit identifies the image type based on specification-related information including at least one of the information other than the information representing the image type included in the metadata and the medical image, An association unit that associates label information including the identified image type with the medical image. Equipped with, When the test results obtained from a single test include multiple series, which are sets of one or more medical images obtained from a single measurement of a single type, The specified unit identifies the image type of each of the multiple series, In the case where, among the aforementioned multiple series, there are multiple series in which the identified image type is the same, The association unit is an information processing system that, in accordance with the results of evaluating each of the multiple series of identical identified image types, associates the label information with one of the multiple series of identical identified image types, and does not associate the label information with the remaining series.

13. A method of information processing performed by a computer, An evaluation step involves obtaining metadata related to the medical image, including information representing the image type of the medical image and parameters of the medical image, and evaluating the information representing the image type based on the metadata. A selection step in which the image type of the medical image is identified according to the evaluation results in the evaluation step, wherein if the evaluation results do not satisfy predetermined conditions, the image type is identified based on specific related information which includes at least one of the information other than the information representing the image type included in the metadata and the medical image; A correlation step in which label information including the identified image type is associated with the medical image. Equipped with, The evaluation step includes a step of further acquiring parameters of other medical images included in the same image group as the medical image, and evaluating information representing the image type of the medical image based on whether the parameters of the medical image and the parameters of the other medical images are of the same type and have different values.

14. An information processing method performed by a computer, An evaluation step involves obtaining metadata related to the medical image, which includes information representing the image type of the medical image, and evaluating the information representing the image type based on the metadata. A selection step in which the image type of the medical image is identified according to the evaluation results in the evaluation step, wherein if the evaluation results do not satisfy predetermined conditions, the image type is identified based on specific related information which includes at least one of the information other than the information representing the image type included in the metadata and the medical image; A correlation step in which label information including the identified image type is associated with the medical image. Equipped with, The identification step includes, if the evaluation result does not satisfy the conditions, a step of determining the range of the specific related information used to identify the image type from the specific related information which includes at least one of the information other than the information representing the image type included in the metadata and the medical image, based on the content of the deficiency identified in the evaluation step.

15. An information processing method performed by a computer, An evaluation step involves obtaining metadata related to the medical image, which includes information representing the image type of the medical image, and evaluating the information representing the image type based on the metadata. A selection step in which the image type of the medical image is identified according to the evaluation results in the evaluation step, wherein if the evaluation results do not satisfy predetermined conditions, the image type is identified based on specific related information which includes at least one of the information other than the information representing the image type included in the metadata and the medical image; A correlation step in which label information including the identified image type is associated with the medical image. Equipped with, When the test results obtained from a single test include multiple series, which are sets of one or more medical images obtained from a single measurement of a single type, The aforementioned identification step involves identifying the image type of each of the multiple series, In the case where, among the aforementioned multiple series, there are multiple series in which the identified image type is the same, The association step is an information processing method in which, based on the results of evaluating each of the multiple series of identical identified image types, the label information is associated with one of the multiple series of identical identified image types, and the label information is not associated with the remaining series.

16. A program for causing a computer to perform the information processing method described in any one of claims 13 to 15.

Images