Analysis device, analysis method, and program

Abstract

To provide an analyzer, a method for analysis, and a program that can efficiently and properly show the reliability of each result when there are more than one result of analysis or diagnosis of medical information.SOLUTION: An analyzer includes: an analysis unit 311 for acquiring "first medical information" obtained by performing computer processing on medical information; a data acquisition unit 33 for acquiring "second medical information" created by a user on the basis of the medical information; and a comparison processing unit 312 for comparing the "first medical information" acquired by the analysis unit 311 and the "second medical information" acquired by the data acquisition unit 33. The comparison processing unit 312 has an output unit for outputting statistical information on the basis of the "first medical information" and "second medical information".SELECTED DRAWING: Figure 4

Description

[Technical Field]

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

[0002] In recent years, with the development of AI (Artificial Intelligence) technology, AI analysis has been introduced into the medical field, and attempts are being made to use AI to assist in the analysis and diagnosis of medical information, such as image diagnosis, which was previously performed by doctors. For example, Patent Document 1 discloses a diagnostic support device that detects differences between first medical information based on information created by a user and second medical information obtained by computer processing, and displays the differences between the two on a display unit in a display format that corresponds to the combination of the lesion names contained in the first medical information and the lesion names contained in the second medical information.

[0003] In clinical settings, there is a need to perform tests and diagnoses appropriately and quickly, and to streamline and optimize diagnostics to reduce the burden on doctors. The introduction of AI analysis is expected to contribute to the efficiency and optimization of such diagnoses. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent No. 5501491 Summary of the Invention [Problem to be solved by the invention]

[0005] However, the technology disclosed in Patent Document 1 does not evaluate or analyze the analysis results of the AI ​​or the diagnosis results of the radiologist. Therefore, if there is a difference between the two, it is unclear which is more reliable and to what extent. As a result, it was necessary to verify both the AI ​​analysis results and the diagnostic results of the radiologist, which created the problem that diagnostic efficiency and optimization were not necessarily achieved.

[0006] The present invention has been made in consideration of the problems in the prior art described above, and aims to provide an analysis device, analysis method, and program that can efficiently and appropriately indicate the degree of reliability of each result when there are multiple analysis results and diagnostic results of medical information. [Means for solving the problem]

[0007] In order to solve the above problem, the invention described in claim 1 is an analysis device, comprising: an analysis unit that performs computer processing on medical information to obtain first medical information related to the medical information; an acquisition unit that acquires second medical information created by a user based on the medical information; a comparison processing unit that compares the first medical information acquired by the analysis unit with the second medical information acquired by the acquisition unit, The comparison processing unit performs an output process of setting one of the compared first medical information and second medical information as correct data and outputting statistical information for evaluating the other information. and the first medical information is set as first correct answer data; Based on the first correct answer data, the correct answer rate of the second medical information is output as the statistical information. It is characterized by:

[0008] Also, claims 11 The invention described in is an analysis method, an analyzing step of performing computer processing on the medical information to obtain first medical information related to the medical information; an acquisition step of acquiring second medical information created by a user based on the medical information; a comparison process step of comparing the first medical information acquired in the analysis step with the second medical information acquired in the acquisition step; an output step of setting one of the first medical information and the second medical information compared as correct data and outputting statistical information for evaluating the other information; stomach, In the comparison process, the first medical information is set as first correct answer data, In the output step, a correct answer rate of the second medical information is output as the statistical information based on the first correct answer data. It is characterized by:

[0009] Also, claims 13 The invention described in is a program, On the computer, an analysis function for performing computer processing on the medical information to obtain first medical information related to the medical information; an acquisition function for acquiring second medical information created by a user based on the medical information; a comparison processing function that compares the first medical information acquired by the analysis function with the second medical information acquired by the acquisition function; Realize this, The comparison processing function performs an output process of setting one of the compared first medical information and second medical information as correct data and outputting statistical information for evaluating the other information. and the first medical information is set as first correct answer data; Based on the first correct answer data, the correct answer rate of the second medical information is output as the statistical information. It is characterized by: [Effects of the Invention]

[0010] According to the present invention, when there are multiple analysis results and diagnosis results of medical information, it is possible to efficiently and appropriately indicate the degree of reliability of each result. [Brief explanation of the drawings]

[0011] [Figure 1] 1 is a diagram illustrating the overall configuration of a medical image system according to an embodiment of the present invention. [Figure 2] 1 is a block diagram showing a functional configuration of an embodiment of an analysis device according to the present invention; [Figure 3](a) is an example of statistical information that lists the analysis results of AI and the diagnostic results of radiologists for a single medical image diagnosis. (b) is an example of statistical information that lists the cumulative results of analysis results of AI and diagnostic results of radiologists over a certain period of time. [Figure 4] 10 is a flowchart showing an analysis process in a first QA pattern. [Figure 5] FIG. 5 is an explanatory diagram schematically showing the flow of the analysis process shown in FIG. 4. [Figure 6] FIG. 10 is an explanatory diagram illustrating a flow of a modified example of the analysis process. [Figure 7] 10 is a flowchart showing an analysis process in a second QA pattern. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, an embodiment of an analysis device, an analysis method, and a program according to the present invention will be described, although the scope of the invention is not limited to the illustrated examples.

[0013] [Configuration of medical imaging system] The analysis device in this embodiment performs analysis of medical images, which are medical information, in a medical image system, for example. FIG. 1 shows the system configuration of a medical image system 100.

[0014] As shown in Fig. 1, the medical image system 100 includes a modality 1, a console 2, an analysis device 3, an interpretation terminal 4, an image server 5, etc., which are connected via a communication network N such as a LAN (Local Area Network), a WAN (Wide Area Network), or the Internet. Each device constituting the medical image system 100 conforms to the HL7 (Health Level Seven) or DICOM (Digital Image and Communications in Medicine) standards, and communication between the devices is performed in accordance with HL7 or DICOM. Note that the number of modalities 1, consoles 2, interpretation terminals 4, etc. is not particularly limited.

[0015] The modality 1 is an image generating device such as an X-ray device (DR, CR), an ultrasound diagnostic device (US), a CT, or an MRI, and generates medical images as medical information by capturing an image of a patient's examination target area as a subject based on examination order information transmitted from a Radiology Information System (RIS) (not shown) or the like. In accordance with the DICOM standard, additional information (patient information, examination information, image ID, etc.) is written to the header of the image file of the medical image generated by the modality 1. The medical image thus annotated with additional information is transmitted to an analysis device 3 or an interpretation terminal 4 via a console 2 or the like.

[0016] The console 2 is an imaging control device that controls imaging in the modality 1. The console 2 outputs imaging conditions and image reading conditions to the modality 1 and acquires image data of medical images captured in the modality 1. The console 2 is configured with a control unit, display unit, operation unit, communication unit, memory unit, etc. (not shown), and each unit is connected by a bus.

[0017] The analysis device 3 is a device that performs various analyses on medical images, which are medical information. The analysis device 3 is configured as a PC, a mobile terminal, or a dedicated device. In this embodiment, the analysis device 3 includes a medical image management device such as a PACS (Picture Archiving and Communication System).

[0018] FIG. 2 is a block diagram showing the functional configuration of the analysis device 3. As shown in FIG. As shown in Figure 2, the analysis device 3 is configured with a control unit 31, a memory unit 32, a data acquisition unit 33, a data output unit 34, an operation unit 35, a display unit 36, etc., and each unit is connected by a bus 37.

[0019] The data acquisition unit 33 is an acquisition unit that acquires various data from an external device (for example, the console 2 or an interpretation terminal 4 described later). The data acquisition unit 33 is configured, for example, by a network interface or the like, and is configured to receive data from an external device connected by wire or wirelessly via the communication network N. In this embodiment, the data acquisition unit 33 is configured by a network interface or the like, but it can also be configured by a port into which a USB memory, an SD card, or the like can be inserted. In this embodiment, the data acquisition unit 33 acquires image data of medical images, for example, from the console 2. The data acquisition unit 33 also acquires, as "second medical information," from the interpretation terminal 4, diagnosis results (detection result information of lesions that can be read from medical images) related to the medical images created by a user (for example, a doctor, etc.) based on the medical images, which are medical information, and interpretation reports, which are interpretation results by radiologists (for example, radiologists who perform primary interpretation and secondary interpretation, etc.).

[0020] The data output unit 34 is used to externally output information processed by the analysis device 3. As the data output unit 34, for example, a network interface for communicating with the image interpretation terminal 4 or the image server 5, a connector for connecting to an external device (for example, a display device, a printer, etc., not shown), a port for various media such as a USB memory, etc. can be applied.

[0021] The operation unit 35 is composed of a keyboard with various keys, a pointing device such as a mouse, or a touch panel attached to the display unit 36. The operation unit 35 allows a user to input data, and specifically outputs operation signals input by key operations on the keyboard, mouse operations, or touch operations on the touch panel to the control unit 31.

[0022] The display unit 36 ​​is configured to include a monitor such as an LCD (Liquid Crystal Display), and displays various screens according to instructions of a display signal input from the control unit 31. The number of monitors is not limited to one, and multiple monitors may be provided. As will be described later, the display unit 36 ​​appropriately displays statistical information and the like output from the control unit 31 (comparison processing unit 312 of the control unit 31).

[0023] The control unit 31 is configured with a CPU (Central Processing Unit), RAM (Random Access Memory), etc., and comprehensively controls the operations of each unit of the analysis device 3. Specifically, the CPU reads out various processing programs stored in a program storage unit 321 of the storage unit 32, loads them into the RAM, and executes various processes in accordance with the programs. In this embodiment, the control unit 31 functions as an analysis unit 311, a comparison processing unit 312, etc. in cooperation with the programs.

[0024] The analysis unit 311 acquires "first medical information" by computer processing of the medical information. Specifically, it detects and analyzes lesions in the medical images acquired by the data acquisition unit 33, and outputs the results of the detection and analysis of one or more types of lesions as "first medical information." The computer processing used here may be, for example, AI analysis using AI (Artificial Intelligence) that performs image diagnosis and image analysis, including lesion detection using CAD (Computer Aided Diagnosis).

[0025] The control unit 31 also functions as a learning unit (not shown) that learns the correspondence between medical information (medical images in this embodiment) and medical information (such as the name of a lesion), and the analysis unit 311 obtains "first medical information" by computer processing the medical information (medical images) based on the correspondence between the medical information (medical images) and medical information learned by the learning unit. That is, for example, a machine learning model created by deep learning or other methods using a large amount of training data (pairs of medical images showing lesions and correct labels (lesion areas in the medical images and the diagnosis of the lesion (type of lesion), etc.)) is used to detect and analyze lesions from input medical images. The "first medical information" thus acquired is information such as the name of the lesion and the location of the lesion, and is attached to the image data of the medical image as additional information.

[0026] The comparison processing unit 312 compares the "first medical information" acquired by the analysis unit 311 with the "second medical information" acquired by the data acquisition unit 33. That is, it compares the two and outputs the comparison result. Specifically, it clarifies matches and mismatches (differences) between the "first medical information" and the "second medical information." As a prerequisite for comparing the two, the comparison processing unit 312 structures the radiological report, etc. created by the user (radiological interpreter), and performs processing to extract character strings, etc. that can be compared with the "first medical information," which is the result of the AI ​​analysis. Although not shown in the figure, the storage unit 32 stores dictionary data, etc., that specifies the correspondence between character strings, etc., used to generate structured data.

[0027] The comparison processing unit 312 also includes an output unit that outputs statistical information based on the "first medical information" and the "second medical information." That is, the comparison processing unit 312 calculates statistical information based on the "first medical information" and the "second medical information." The statistical information calculated by the comparison processing unit 312 is output to, for example, the display unit 36 ​​of the analysis device 3. The display unit 36 ​​is capable of displaying the output statistical information. The output destination of the statistical information is not limited to this, and the statistical information may be output to the data output unit 34 so that the statistical information can be displayed on an external display device or various terminal devices.

[0028] Here, statistical information includes numerical and chart information obtained by examining a group under certain conditions such as time and region, and aggregating and processing the results. Statistical information also includes information that quantitatively expresses the attributes of a group in numerical and chart form based on the distribution of the individual components of the group. Statistical information includes the match and mismatch rates between medical information, and examples of match and mismatch rates between medical information include the accuracy rate of other medical information when certain medical information is used as correct data.

[0029] In this embodiment, the match rate and mismatch rate of the result of comparing the "first medical information" and the "second medical information" are calculated as statistical information. In other words, the comparison processing unit 312 determines either the "first medical information" or the "second medical information" as the correct data, and calculates the match rate or mismatch rate of the other data relative to the correct data, thereby obtaining the reliability of the other data relative to the correct data. For example, the comparison processing unit 312 sets the "first medical information" that is the result of the AI ​​analysis as the "first correct answer data," and calculates (outputs) the accuracy rate of the "second medical information" that is the diagnostic result of the radiologist based on this "first correct answer data," thereby making it possible to know the reliability of the diagnostic result of the radiologist relative to the AI ​​analysis. Conversely, the "second medical information," which is the diagnosis result of the radiologist, can be used as the "second correct answer data," and the accuracy rate of the "first medical information," which is the result of the AI ​​analysis, can be calculated (output) as statistical information based on this "second correct answer data." This allows the reliability of the AI ​​analysis results relative to the diagnosis of the radiologist to be known.

[0030] The statistical information calculated and output by the comparison processing unit 312 is not limited to this. For example, character strings (keywords) that are commonly found in the "first medical information" and the "second medical information" may be extracted as statistical information. Also, various percentages and statistics, such as the distribution of character strings that are not common to both (such as the percentage of what character strings are found), may be extracted as statistical information. The comparison processing unit 312 may output a table (matrix, heatmap information, etc.) that summarizes such information in a list of findings as statistical information.

[0031] Figure 3(a) is an example of statistical information that lists the "first medical information," which is the analysis result of AI, and the "second medical information," which is the diagnosis result of the user (specialist in primary and secondary reading) in the diagnosis of a certain medical image. In Figure 3(a), the vertical axis shows the analysis result of AI, and the horizontal axis shows the diagnosis result of the doctor (radiography doctor). A single medical image can reveal a variety of findings, including not only a single lesion but also the possibility of multiple lesions. For example, Figure 3(a) shows an example in which the medical image being analyzed and diagnosed is a chest X-ray image. In this case, possible findings from the AI ​​and radiologist include "nodular shadow," "infiltrate," "reticular shadow," and "pneumothorax." The diagonal lines shown with light shading in the example shown indicate items where the AI's analysis results and the radiologist's diagnosis agree. In the example of Figure 3(a), both the AI ​​and the radiologist diagnosed the medical image as containing a "nodular shadow."

[0032] In the example shown, the AI's analysis results indicated that in addition to "nodule shadows," "infiltration shadows" and "reticular shadows" were also recognized, but the radiologist's diagnosis only recognized "nodule shadows" and nothing else. By outputting a list of statistical information like this, it becomes clear which judgments are consistent and which are inconsistent, as well as which diagnoses are consistent between the AI ​​and the radiologist, and what kind of pathology is likely to be present in the image.

[0033] Figure 3(b) is an example of statistical information that lists the "first medical information," which is the result of AI analysis, and the "second medical information," which is the diagnosis result of the user (a specialist in primary and secondary reading) for medical image diagnoses over a certain period of time, such as one week or one month. In Figure 3(b), as in Figure 3(a), the vertical axis shows the AI ​​analysis results and the horizontal axis shows the doctor's (radiography doctor's) diagnosis results for a case where the medical image to be analyzed and diagnosed is a chest X-ray image. By collecting statistical information over a certain period of time, it is possible to see trends such as the areas where the AI ​​analysis results and the radiology doctor's diagnosis results agree and the areas where they tend to differ. For example, the rate at which non-pneumothorax is judged to be "pneumothorax" is relatively low, but the rates at which "infiltration" and "reticular shadows" are recognized by both the AI ​​and the radiology doctor are high, suggesting that these are areas where both the AI ​​and the radiology doctor are prone to misidentification.

[0034] In Figures 3(a) and 3(b), for the sake of simplicity, only four types of findings are shown: "nodule," "infiltrate," "reticular shadow," and "pneumothorax," and only a limited number of findings are shown for each type. However, in reality, analysis and judgment are performed on, for example, 10 or so items from a single medical image. In this regard, by presenting statistical information compiled in a table, as in Figures 3(a) and 3(b), it is possible to clearly present the agreement / disagreement status and trends between the AI ​​analysis results and the diagnostic results of the radiologist, even when there are a large number of findings or items.

[0035] The display unit 36, to which the statistical information output from the comparison processing unit 312 is output, displays the statistical information appropriately according to its contents. For example, the output mode (display mode) of the statistical information may be changed depending on the statistical information (match rate, mismatch rate, etc.). For example, only the mismatched items may be displayed in a conspicuous color or in large letters. Furthermore, the method of changing the output mode (display mode) of statistical information according to the statistical information (match rate, mismatch rate, etc.) is not limited to changing the color, text, etc. when displayed. For example, statistical information may be output by displaying a design such as a mark. In this case, a threshold may be set for the statistical information (match rate, mismatch rate, etc.), and if the value is above a certain value, an "O" may be displayed, and if the value is below the certain value, an "X" may be displayed. Using marks or the like rather than numbers makes it easier for doctors and other medical professionals to intuitively grasp and has excellent visibility. Therefore, when the number of cases (the number of cases for which statistical information is output in the comparison processing unit 312) is particularly large, using marks or the like can speed up and streamline processing.

[0036] Furthermore, for example, the way the AI ​​analysis results are displayed on the screen may differ depending on whether the accuracy rate of the AI ​​analysis results for the correct data is high or low. For example, if the accuracy rate of the AI ​​analysis results is low and not very reliable, both the AI ​​analysis results and the radiologist's diagnosis results may be displayed prominently for items where the AI ​​analysis results and the radiologist's diagnosis results do not match. On the other hand, if the accuracy rate of the AI ​​analysis results is high and reliable, only the AI ​​analysis results may be displayed prominently for items where the two do not match.

[0037] As mentioned above, the statistical information indicates the rate of agreement / disagreement between the AI ​​analysis results and the diagnostic results of the radiologist. In this case, the diagnostic results of the radiologist may be those of the hospital as a whole, or those of an individual radiologist. When various statistical information can be calculated by changing the perspective, such as for the entire hospital or for individual radiologists, the display method may be switched depending on the case. When calculating statistical information from various angles like these, for example, when the reliability of the AI ​​is high, information showing the agreement / disagreement rate between the diagnosis results of individual radiologists and the analysis results of the AI ​​can be used as a tool for supervisors to check the accuracy rate of new radiologists (i.e., the extent to which the new radiologist was able to arrive at the same diagnosis result as the AI). Furthermore, by narrowing down the statistical information to tests by modality 1 or by department, it will also be possible to verify the accuracy rate (diagnostic accuracy) by modality 1 or by department. If such verification becomes possible, it can be used as a reference value for improvement measures, such as holding radiology study sessions for departments with low accuracy rates. Furthermore, by dividing the calculation range of statistical information by the date of reading, etc., it will be possible to use this statistical information as an objective reference for determining the degree of improvement, for example, to see how much the accuracy rate in 2020 has improved in 2021.

[0038] The memory unit 32 is composed of an HDD (Hard Disk Drive), semiconductor memory, etc., and stores programs for executing various processes including the analysis of medical information such as medical images, which will be described later, as well as parameters, files, etc. required for executing the programs. The storage unit 32 of this embodiment stores, for example, a program storage unit 321 that stores various programs, a learning data storage unit 322, a statistical information storage unit 323, and the like. As described above, the learning data storage unit 322 stores, for example, a large amount of learning data and a machine learning model created by learning using this data through deep learning or the like. Although not shown in the figure, dictionary data, etc. used to structure the "second medical information," such as the radiological report created by the user (radiological interpreter) as described above, and generate structured data is also stored in the memory unit 32.

[0039] The image interpretation terminal 4 is a computer device that includes, for example, a control unit, an operation unit, a display unit, a storage unit, a communication unit, etc., and that reads out medical images, which are medical information, from the image server 45, etc., and displays them for image interpretation. The users (primary and secondary radiologists) interpret the medical images on the radiology terminal 4 and create a radiology report or the like as the "second medical information" that is the radiologist's diagnosis regarding the medical images.

[0040] The image server 5 is, for example, a server that constitutes a PACS (Picture Archiving and Communication Systems), and stores in a database the medical images output from the modality 1 in association with patient information (patient ID, patient name, date of birth, age, sex, height, weight, etc.), examination information (examination ID, examination date and time, type of modality, examination area, requesting department, purpose of examination, etc.), image ID of the medical image, information on the AI ​​analysis results output from the analysis unit 311 of the analysis device 3 (i.e., "first medical information") and the interpretation report created by the user (radiography doctor) on the interpretation terminal 4, "second medical information" which is the diagnosis result of the radiology doctor, matching results (comparison results) and statistical information (e.g., information such as that shown in Figures 3(a) and 3(b)) output from the comparison processing unit 312 of the analysis device 3, etc. Furthermore, the image server 5 reads out the medical image requested by the image interpretation terminal 4 and various additional information attached to the medical image from the database, and displays them on the image interpretation terminal 4.

[0041] [Analysis Method in This Embodiment] In this embodiment, the analysis method includes an analysis step of acquiring "first medical information" obtained by computer processing of medical information such as medical images, an acquisition step of acquiring "second medical information" created by a user based on the medical information, and a comparison processing step of comparing the "first medical information" acquired in the analysis step with the "second medical information" acquired in the acquisition step, and the comparison processing step includes an output step of outputting statistical information based on the "first medical information" and the "second medical information."

[0042] According to the analysis method of this embodiment, by taking into consideration both the analysis results of the AI ​​and the diagnostic results of the radiologist, quality assurance (hereinafter referred to as "QA") can be performed on the diagnostic accuracy of medical information (medical images). As for QA patterns, the flow of the analysis method differs depending on how reliable the AI ​​analysis results are. The first QA pattern is a flow that is adopted when the AI ​​analysis results are not very reliable and the AI ​​analysis cannot be fully trusted.

[0043] FIG. 4 is a flowchart showing the analysis process in the first QA pattern, and FIG. 5 is an explanatory diagram showing a schematic flow of the process. As shown in Figures 4 and 5, in the first QA pattern, first, the data acquisition unit 33 acquires "second medical information (in the primary interpretation)," which is the diagnosis result of the primary interpretation physician, such as an interpretation report created by the primary interpretation physician by interpreting the target medical information (here, a medical image) (step S1; acquisition process). In addition, the target medical information (medical image) is analyzed by AI (AI application) in the analysis unit 311 of the analysis device 3, and the "first medical information" which is the analysis result (diagnosis result) is obtained (step S2; analysis process).

[0044] Then, the comparison processing unit 312 matches (compares) the "first medical information" with the "second medical information" (step S3; comparison processing step). In this embodiment, the comparison processing unit 312 calculates and outputs statistical information as appropriate (step S4; output process). The type of statistical information to be output may be set appropriately depending on the type of medical image (such as which part the image depicts) and the reliability of the AI ​​analysis results. The user (primary radiologist, secondary radiologist, etc.) may select the type of statistical information to be calculated and output.

[0045] The comparison processing unit 312 compares the "first medical information" with the "second medical information" and determines whether they match (step S5). If they match (step S5; YES), the diagnosis result and analysis result are used as information on a definitive diagnosis (hereinafter referred to as "definitive diagnosis information") (step S6). On the other hand, if they do not match (if they differ, step S5; NO), the image is sent to a second image interpreter (final image interpreter) for secondary interpretation (step S7). The result of this secondary interpretation is then regarded as "definitive diagnosis information" (step S8). When the image is sent to secondary interpretation, as shown in FIG. 5, the target medical information (medical image) and statistical information calculated and output based on the "first medical information," "second medical information," and "first medical information" and "second medical information" are sent to the image interpretation terminal 4 operated by the secondary image interpreter, and the secondary image interpreter performs secondary interpretation by referring to this information and creates a diagnostic result such as an image interpretation report as "second medical information (in secondary interpretation)."

[0046] In this way, in the process shown in Figures 4 and 5, the "second medical information (in the primary reading)," which is the diagnosis result of the primary radiologist, is matched (compared) with the "first medical information," which is the analysis result (diagnosis result) of the AI, and the results (matching results and statistical information) are sent to the secondary radiologist. This allows the secondary interpretation physician to refer to this information when interpreting the images, enabling them to, for example, focus on checking only the areas where the two parties' judgments differ, thereby enabling efficient and appropriate interpretation. In particular, if the statistical information is displayed in a way that highlights areas where the judgments between the "first medical information" and the "second medical information" differ by changing the color, or if, when an item where the judgments between the two differ is touched, the area in the medical image where an abnormality related to that item has been determined is marked or highlighted by changing the color, the efficiency and accuracy of interpretation (secondary interpretation) can be further improved.

[0047] When "definitive diagnosis information" is issued for a medical image as medical information, the medical image is associated with the "definitive diagnosis information" determined by the analysis device 3 and stored in the image server 5. There are no particular limitations on where medical images and various accompanying information (such as "first medical information," "second medical information," statistical information, etc.), "definitive diagnosis information," etc. are stored. For example, if the analysis device 3, interpretation terminal 4, and image server 5 constitute a PACS, this information may be collectively stored and managed in an information management server on the PACS. Furthermore, it is not necessary to store all of the above information; for example, only the medical image and the "definitive diagnosis information" may be stored as a set.

[0048] The term "definitive diagnosis information" as used herein refers to a diagnosis determined by a radiologist based on medical images and the findings and analysis results obtained based on the images. The final diagnosis for a patient is determined by a clinician, taking into consideration the "definitive diagnostic information" from the medical images as well as test data and examination data obtained from various tests and examinations.

[0049] In Figures 4 and 5, an example is shown in which either the "second medical information" in the primary interpretation or the "first medical information" in the AI ​​analysis is set as correct data, and the reliability of the other information relative to the correct data is calculated as statistical information. However, the comparison when calculating the reliability is not limited to the "second medical information (in the primary interpretation)" and the "first medical information."

[0050] For example, as shown in FIG. 6, the comparison processing unit 312 may consider either the "first medical information" obtained by AI analysis or the "second medical information (in the secondary interpretation)" ("definitive diagnosis information"), which is the diagnosis result of the secondary interpretation physician, as correct data, and calculate (output) the match rate or mismatch rate of the other data relative to the correct data as statistical information. That is, for example, the comparison processing unit 312 may treat the "second medical information (in the secondary interpretation)" ("definitive diagnosis information"), which is the diagnosis result of the secondary radiologist, as "second correct answer data," and calculate (output) the accuracy rate of the "first medical information" by the AI ​​analysis based on this "definitive diagnosis information" as statistical information. This allows the reliability of the AI ​​analysis of the diagnosis result of the secondary radiologist ("definitive diagnosis information") to be known. Conversely, the "first medical information" that is the result of the AI ​​analysis can be used as the "second correct answer data," and the accuracy rate of the second medical information (in the second reading) that is the diagnosis result of the secondary radiologist ("definitive diagnosis information") can be calculated (output) as statistical information based on this "second correct answer data." This makes it possible to know the reliability of the diagnosis of the secondary radiologist relative to the AI ​​analysis results.

[0051] In this way, when calculating the match / mismatch rate between the "first medical information" and the "second medical information (in the secondary interpretation)" ("definitive diagnosis information"), using either one of them as the correct data, as shown in Figure 6, the "second medical information (in the primary interpretation)" resulting from the primary interpretation is only compared with the "first medical information" (i.e., statistical information such as the match rate is not calculated), and the target medical information (medical image) and the "first medical information," "second medical information," and the match information (match / mismatch information) between the "first medical information" and the "second medical information" are sent to the interpretation terminal 4 operated by the secondary interpretation physician. The secondary interpretation physician then performs secondary interpretation by referring to this information and prepares a diagnostic result such as an interpretation report as the "second medical information (in the secondary interpretation)," which becomes the "definitive diagnosis information." As shown in Figure 6, if the results of the AI ​​analysis of the "first medical information" and the "second medical information (in the primary reading)" are not processed separately, and both cases of agreement and disagreement are sent to the secondary reading, the calculation of statistical information (e.g., agreement rate, disagreement rate, etc.) may be performed only between the "first medical information" and the "second medical information (in the secondary reading)" ("definitive diagnosis information"), or may be performed both between the "first medical information" and the "second medical information (in the primary reading)" and between the "first medical information" and the "second medical information (in the secondary reading)" ("definitive diagnosis information"). Furthermore, when the comparison processing unit 312 compares the "first medical information" obtained by AI analysis with the "second medical information (in the secondary interpretation)" (i.e., "definitive diagnosis information"), which is the diagnosis result of the secondary interpretation physician, it is possible to simply compare the two without calculating statistical information. This allows the image diagnosis to be completed more quickly. In this case, too, the information on the comparison results, indicating whether the judgment of the secondary radiologist (final radiologist) matches the analysis results of the AI, is attached to the "definitive diagnosis information," providing useful information for clinicians who make the final diagnosis by comprehensively considering the results of the image diagnosis and various test information, etc.

[0052] In contrast to the above, if the reliability of the AI ​​analysis results is extremely high and the AI ​​analysis can be fully trusted, the second QA pattern flow is adopted. FIG. 7 is a flowchart showing the analysis process in the second QA pattern. As shown in Figure 7, in the second QA pattern, first, the target medical information (medical image) is analyzed by AI in the analysis unit 311 of the analysis device 3, and the "first medical information" which is the analysis result (diagnosis result) is obtained (step S11). In this case, it is not necessary to obtain the "second medical information (in the primary reading)," which is the diagnosis result of the primary reading physician.

[0053] Then, the control unit 31 determines whether the "first medical information" determines that the medical image is normal (step S12). That is, the control unit 31 determines whether the "first medical information" includes any abnormal findings in the medical image. If the "first medical information" determines that the medical image is normal (step S12; YES), the "first medical information", which is the analysis result by the AI, is set as "definitive diagnosis information" (step S13).

[0054] In this case, the work of the radiologist can be reduced for images that the AI ​​judges to be normal, reducing the burden on the radiologist and enabling efficient image diagnosis. Furthermore, even if the "first medical information" judges the medical image to be normal, this does not immediately become "definitive diagnosis information." Instead, a "normal label" indicating that the medical image is normal and does not contain abnormal findings can be attached to the "first medical information," which is the result of analysis by AI, and the medical image to be judged and the "first medical information" can be sent to a secondary radiologist (or final radiologist), and the diagnosis result of the secondary radiologist can be considered "definitive diagnosis information." In this case, the burden on the radiologist can be reduced by assigning a "normal label" while still seeking a diagnosis from the secondary radiologist (final radiologist).

[0055] On the other hand, if the "first medical information" does not determine that the medical image is normal (if abnormal findings are included, step S12; NO), the medical image to be judged and the "first medical information" are sent to a secondary image interpreter (or final image interpreter) for secondary interpretation (step S14). The diagnosis result of the secondary image interpreter is then set as "definitive diagnosis information" (step S15). In this case, the comparison processing unit 312 may also compare the "first medical information" that is the analysis result of the AI ​​with the "definitive diagnosis information" that is the diagnosis result of the secondary radiologist, and calculate (output) statistical information such as the match rate and mismatch rate. When the statistical information is output, the statistical information is also stored in the image server 5 or the like together with the medical images and the "definitive diagnosis information." By linking and storing the statistical information with the medical images and the "definitive diagnosis information," it becomes possible to refer to it later when a clinician makes a final diagnosis, etc.

[0056] 〔effect〕 As described above, the analysis device 3 of this embodiment includes an analysis unit 311 that acquires "first medical information" obtained by computer processing of medical information (i.e., AI analysis), a data acquisition unit 33 that acquires "second medical information" created by a user (i.e., a radiologist who performs primary and secondary interpretation) based on the medical information, and a comparison processing unit 312 that compares the "first medical information" acquired by the analysis unit 311 with the "second medical information" acquired by the data acquisition unit 33. Comparing the judgment of the radiologist with the analysis results of AI allows for more careful diagnosis and appropriate image diagnosis. In particular, when the subject of comparison with the analysis results of AI is the judgment (definitive diagnosis) made by a secondary radiologist, the judgment of the human doctor can be calmly verified, leading to more appropriate image diagnosis.

[0057] The comparison processing unit 312 of the analysis device 3, which is equipped with an analysis unit 311 that acquires "first medical information" obtained by computer processing of medical information (i.e., AI analysis), a data acquisition unit 33 that acquires "second medical information" created by a user (i.e., a radiologist who performs primary and secondary interpretation) based on the medical information, and a comparison processing unit 312 that compares the "first medical information" acquired by the analysis unit 311 with the "second medical information" acquired by the data acquisition unit 33, may be equipped with an output unit that outputs statistical information based on the "first medical information" and the "second medical information". When AI analysis is introduced into image diagnosis, multiple analysis and diagnosis results for medical images (medical information) are obtained, including medical information obtained by AI analysis and medical information created by radiologists who perform primary and secondary interpretation. In this case, if the reliability of each result cannot be verified, it is ultimately impossible to determine which judgment to base the final definitive diagnosis on, making it impossible to achieve efficient interpretation. In this regard, by outputting statistical information such as the match rate and mismatch rate for each result, the degree of reliability can be effectively indicated, enabling efficient and appropriate definitive diagnoses to be made for medical images, which are medical information.

[0058] Furthermore, if the statistical information is the rate of agreement or disagreement between the "first medical information" and the "second medical information," it will be possible to know to what extent the analysis results by the AI ​​match the diagnosis results of the radiologist. If there is a large discrepancy between the analysis results by AI and the diagnosis made by a radiologist who is a specialist in radiological interpretation, the reliability of each result may not be very high. By obtaining an index of reliability in this way, it is possible to make an appropriate diagnosis of medical images, which are medical information.

[0059] Furthermore, the comparison processing unit 312 regards the "first medical information" as the "first correct answer data" and outputs the correct answer rate of the "second medical information" based on the "first correct answer data" as statistical information. Whether the AI ​​analysis results or the radiologist's diagnosis are correct depends on which judgment result is considered correct. In this regard, by using the "first medical information" obtained through AI analysis as correct data, it is possible to calculate the reliability of the diagnostic results of the radiologist based on this information.

[0060] Furthermore, the comparison processing unit 312 regards the "second medical information" as the "second correct answer data" and outputs the correct answer rate of the "first medical information" as statistical information based on the "second correct answer data." In this way, by using the diagnostic results of the radiologist as the "second medical information," it is possible to calculate the reliability of the analysis results by AI.

[0061] Furthermore, the "second medical information" may be "definitive diagnosis information" which is the diagnosis result of a secondary radiologist (or final radiologist). In this case, by outputting statistical information based on the "first medical information" and the "second medical information," it is possible to obtain the reliability of the "definitive diagnosis information" as the "second medical information."

[0062] Furthermore, when the comparison processing unit 312 treats the "second medical information," which is the "definitive diagnosis information," as the "second correct answer data" and outputs the accuracy rate of the "first medical information" as statistical information based on this "second correct answer data," it is possible to calculate the reliability of the analysis results by AI for the "definitive diagnosis information."

[0063] In addition, the analysis unit 311 or the control unit 31 including the analysis unit 311 is provided with a learning unit that learns the correspondence between medical information (e.g., medical images) and medical information (e.g., names of lesions, etc.), and the analysis unit 311 obtains ``first medical information'' by computer processing the medical information based on the correspondence between the medical information and medical information learned by the learning unit. When the analysis unit 311 uses a machine learning model in this way, the accuracy of the analysis can be improved through repeated learning, making it possible to obtain more reliable "first medical information."

[0064] [Modification] Although the embodiment of the present invention has been described above, it goes without saying that the present invention is not limited to such an embodiment and that various modifications are possible without departing from the spirit of the present invention.

[0065] For example, in the above embodiment, the medical information to be analyzed by the analysis device is a medical image, but the medical information is not limited to a medical "image." Information obtained through various tests on patients may be broadly included in medical information, and for example, results obtained from various tests, such as electrocardiogram waveform data, heart sound data, and data related to blood flow, may also be included in medical information.

[0066] Furthermore, in the above embodiment, the analysis unit 311 acquires "first medical information" obtained by computer processing of medical information (i.e., AI analysis), and the comparison processing unit 312 compares this "first medical information" with "second medical information" created by a user (i.e., a radiologist who performs primary and secondary interpretation) based on the medical information, and calculates (outputs) statistical information such as the match rate and mismatch rate between the two, but the objects that the comparison processing unit 312 compares, etc. are not limited to this.

[0067] For example, the analysis unit 311 may perform analysis (computer processing) on ​​the medical information using an AI different from the AI ​​that acquired the “first medical information” to acquire the “third medical information.” In this case, the comparison processing unit 312 may compare the “first medical information” that is the analysis result by one AI with the “third medical information” that is the analysis result by another AI. The comparison processing unit 312 may then treat the "third medical information" as the "third correct answer data" and calculate (output) the accuracy rate of the "first medical information" based on the "third correct answer data" as statistical information. In this case, the reliability of the AIs can be compared. For example, by looking at the extent to which a newly introduced AI at a certain facility can produce the same analysis results as the analysis results of another AI that is already highly reliable (the "third correct data," or "third medical information") (i.e., what the accuracy rate is), it is possible to calculate the reliability of the newly introduced AI.

[0068] Furthermore, in the above embodiment, as mentioned above, an example was given of the case where the objects that the comparison processing unit 312 compares are the "second medical information" which is the diagnosis result of the user (i.e., the radiologist) and the "first medical information" which is the analysis result of the AI, but the objects that the comparison processing unit 312 compares are not limited to this. For example, the data acquisition unit 33 may acquire the "fourth medical information," which is the diagnosis result of the primary radiologist, and the "fifth medical information," which is the diagnosis result of the secondary radiologist, based on the medical information, and the comparison processing unit 312 may compare the "fourth medical information" with the "fifth medical information," or calculate (output) statistical information such as the agreement rate and disagreement rate between the two. In this case as well, by determining either the "fourth medical information" or the "fifth medical information" as correct data, the reliability of the other can be calculated.

[0069] In this embodiment, in FIG. 1, the analysis device 3, the interpretation terminal 4, and the image server 5 are illustrated as separate and independent devices, but the analysis device 3 and the image server 5, or the analysis device 3, the interpretation terminal 4, and the image server 5 may be configured as a single device or a single system.

[0070] It goes without saying that the present invention is not limited to the above-described embodiments and modifications, and can be modified as appropriate without departing from the spirit of the present invention. [Explanation of symbols]

[0071] 1. Modality 2 Console 3 Analysis device 4. Image reading terminal 5. Image Server 31 Control Unit 32 Storage section 33 Data Acquisition Section 36 Display section 100 Medical Imaging Systems 311 Analysis Department 312 Comparison processing section

Claims

1. an analysis unit that performs computer processing on the medical information to obtain first medical information related to the medical information; an acquisition unit that acquires second medical information created by a user based on the medical information; a comparison processing unit that compares the first medical information acquired by the analysis unit with the second medical information acquired by the acquisition unit, the comparison processing unit sets one of the compared first medical information and the second medical information as correct answer data, performs output processing to output statistical information evaluating the other medical information, and sets the first medical information as first correct answer data; An analysis device that outputs the accuracy rate of the second medical information as the statistical information based on the first correct answer data.

2. The analysis device according to claim 1 , wherein the statistical information is a match rate or a mismatch rate between the first medical information and the second medical information.

3. the comparison processing unit determines the second medical information as second correct answer data, 3. The analysis device according to claim 1, wherein the accuracy rate of the first medical information is output as the statistical information based on the second correct answer data.

4. An analysis unit that acquires first medical information related to the medical information by performing computer processing on the medical information; an acquisition unit that acquires second medical information created by a user based on the medical information; a comparison processing unit that compares the first medical information acquired by the analysis unit with the second medical information acquired by the acquisition unit, the analysis unit acquires third medical information related to the medical information obtained by a second computer processing different from the computer processing of the medical information; the comparison processing unit sets one of the compared first medical information and second medical information as correct answer data, performs output processing to output statistical information evaluating the other medical information, and sets the third medical information as third correct answer data; An analysis device that outputs a correct answer rate for the first medical information based on the third correct answer data.

5. An analysis device as described in claim 4, wherein the statistical information is the agreement rate or disagreement rate between the first medical information and the second medical information.

6. The analysis device according to claim 1 , wherein the second medical information is definitive diagnosis information.

7. the second medical information is definitive diagnosis information; the comparison processing unit determines the second medical information as second correct answer data, 7. The analysis device according to claim 1, wherein the accuracy rate of the first medical information is output as the statistical information based on the second correct answer data.

8. a learning unit for learning correspondence between medical information and clinical information; The analysis unit, based on the correspondence between the medical information and medical information learned by the learning unit, The analysis device according to claim 1 , wherein the first medical information is obtained by computer processing of the medical information.

9. an analysis unit that performs computer processing on the medical information to obtain first medical information related to the medical information; an acquisition unit that acquires definitive diagnosis information created by a user based on the medical information; comparing the first medical information acquired by the analysis unit with the definitive diagnosis information acquired by the acquisition unit, and setting the first medical information as first correct answer data; a comparison processing unit that outputs a correct answer rate of the definitive diagnosis information as statistical information based on the first correct answer data; An analysis device comprising:

10. An analysis unit that acquires first medical information related to the medical information by performing computer processing on the medical information; an acquisition unit that acquires definitive diagnosis information created by a user based on the medical information; a comparison processing unit that compares the first medical information acquired by the analysis unit with the definitive diagnosis information acquired by the acquisition unit; Equipped with the analysis unit acquires third medical information related to the medical information obtained by a second computer processing different from the computer processing of the medical information; The comparison processing unit further sets the third medical information as third correct answer data, An analysis device that outputs a correct answer rate for the first medical information based on the third correct answer data.

11. an analyzing step of performing computer processing on the medical information to obtain first medical information related to the medical information; an acquiring step of acquiring second medical information created by a user based on the medical information; a comparison process step of comparing the first medical information acquired in the analysis step with the second medical information acquired in the acquisition step; an output step of setting one of the first medical information and the second medical information compared as correct data and outputting statistical information for evaluating the other medical information; In the comparison process, the first medical information is set as first correct answer data, In the output step, a correct answer rate of the second medical information is output as the statistical information based on the first correct answer data.

12. An analysis step of acquiring first medical information related to the medical information by performing computer processing on the medical information; an acquiring step of acquiring second medical information created by a user based on the medical information; a comparison process step of comparing the first medical information acquired in the analysis step with the second medical information acquired in the acquisition step; an output step of setting one of the first medical information and the second medical information compared as correct data and outputting statistical information for evaluating the other medical information; the analyzing step includes acquiring third medical information related to the medical information obtained by a second computer processing different from the computer processing of the medical information, The comparison processing step further comprises determining the third medical information as third correct answer data, A computer-based analysis method that outputs the accuracy rate of the first medical information based on the third correct answer data.

13. On the computer, an analysis function for performing computer processing on the medical information to obtain first medical information related to the medical information; an acquisition function for acquiring second medical information created by a user based on the medical information; a comparison processing function that compares the first medical information acquired by the analysis function with the second medical information acquired by the acquisition function; Realize this, the comparison processing function sets one of the compared first medical information and second medical information as correct answer data, performs output processing to output statistical information for evaluating the other medical information, and sets the first medical information as first correct answer data; A program that outputs the accuracy rate of the second medical information as the statistical information based on the first correct answer data.

14. On the computer, an analysis function for performing computer processing on the medical information to obtain first medical information related to the medical information; an acquisition function for acquiring second medical information created by a user based on the medical information; a comparison processing function that compares the first medical information acquired by the analysis function with the second medical information acquired by the acquisition function; Realize this, the analysis function acquires third medical information related to the medical information obtained by a second computer processing different from the computer processing of the medical information; the comparison processing function sets one of the compared first medical information and second medical information as correct answer data, performs output processing to output statistical information for evaluating the other medical information, and further sets the third medical information as third correct answer data; A program that outputs a rate of accuracy of the first medical information based on the third correct answer data.

Images