Image processing device, control method, and program

Abstract

It outputs information to support diagnosis, which can reduce the burden on both doctors and patients. [Solution] The image processing device includes a first acquisition means for acquiring biological information relating to a biological body to be diagnosed, a second acquisition means for acquiring a medical image of the area being examined, a derivation means for deriving first information relating to the diagnosis of the area being examined based on the medical image, and an output means for generating and outputting second information relating to the diagnosis of the area being examined based on the first information, biological information, and change information relating to changes in the area being examined due to aging.

Description

Technical Field

[0001] The present invention relates to an image processing apparatus, a control method, and a program, and particularly to a diagnostic technique based on a medical image obtained by medical imaging.

Background Art

[0002] In recent years, in various medical fields, diagnosis using imaging images has been performed. In Patent Document 1, in the field of dental orthodontics, an imaging image of a patient's dentition is used to predict the future state of the patient's dentition and provide a treatment recommended for correcting the dentition to a suitable state.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] By the way, among human body parts, there are some whose characteristics change with aging. For example, in the field of dental medicine as in Patent Document 1, the teeth in a person's oral cavity show characteristic changes such as growing from deciduous teeth to permanent teeth with aging, and the thickness of enamel and dentin increasing and the strength increasing. For a human body part whose characteristics change with aging, the content of treatment or treatment may differ depending on the state showing which characteristics.

[0005] Therefore, even if body parts are located in the same area, their characteristics can change with age, so it is necessary to correctly determine the state of that body part for diagnosis. Ideally, such determination should be made by a specialist observing the patient's body part. However, the practice of requiring a doctor to assess the condition of all teeth each time, even for some procedures where a doctor's assistant primarily performs the work, was cumbersome. Furthermore, it could be a physical burden for the patient, in terms of the waiting time until the doctor arrives, the time spent on examinations for assessment, and the overall treatment time.

[0006] This invention has been made in view of the above-mentioned problems, and aims to provide an image processing device, a control method, and a program that output information for diagnostic support that can reduce the burden on both doctors and patients. [Means for solving the problem]

[0007] To achieve the aforementioned objectives, the image processing apparatus of the present invention is characterized by comprising: a first acquisition means for acquiring biological information relating to a biological body to be diagnosed; a second acquisition means for acquiring a medical image of a part of the biological body being examined; a derivation means for deriving first information relating to the diagnosis of the part of the body being examined based on the medical image; and an output means for generating and outputting second information relating to the diagnosis of the part of the body being examined based on the first information, biological information, and change information relating to changes in the part of the body being examined due to aging. [Effects of the Invention]

[0008] With this configuration, the present invention makes it possible to output information for diagnostic support that can reduce the burden on both doctors and patients. [Brief explanation of the drawing]

[0009] [Figure 1] Figure showing an example of the configuration of a support system relating to embodiments and modifications of the present invention. [Figure 2]Block diagram illustrating the hardware configuration of the image processing apparatus 101 according to embodiments and modified examples of the present invention. [Figure 3] Block diagram illustrating the hardware configuration of the electronic medical record terminal 106 according to embodiments and modified examples of the present invention. [Figure 4] A diagram illustrating the outline of the diagnostic imaging according to Embodiment 1 of the present invention. [Figure 5] Figure illustrating an existence time table for embodiments and modified examples of the present invention. [Figure 6] Another figure illustrating an existence time table relating to embodiments and modifications of the present invention. [Figure 7] Figure illustrating the correction of evaluation values ​​according to Embodiment 1 of the present invention. [Figure 8] Flowchart illustrating output processing performed in the image processing apparatus 101 according to embodiments and modified versions of the present invention. [Figure 9] A flowchart illustrating the correction process according to Embodiment 1 of the present invention. [Figure 10] A diagram illustrating the correction process according to Modification 1 of the present invention. [Figure 11] A flowchart illustrating the correction process according to Modification 1 of the present invention. [Figure 12] Another diagram illustrating the correction process according to Modification 1 of the present invention. [Figure 13] Another flowchart illustrating the correction process according to Modification 1 of the present invention. [Figure 14] This figure shows an example of the system configuration according to Modification 3 of the present invention. [Figure 15] Another figure showing an example of the system configuration according to Modification 3 of the present invention. [Figure 16] Another figure showing an example of the system configuration according to Modification 3 of the present invention. [Modes for carrying out the invention]

[0010] [Embodiment 1] Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. Note that the following embodiments do not limit the invention according to the claims. Although a plurality of features are described in the embodiments, not all of these plurality of features are essential for the invention, and the plurality of features may be arbitrarily combined. Further, in the accompanying drawings, the same or similar configurations are denoted by the same reference numerals, and redundant descriptions are omitted.

[0011] One embodiment described below is an example of an image processing apparatus used in a support system capable of assisting diagnosis (hereinafter, image diagnosis) using a medical image obtained by photographing an examination site of a living body. An example of applying the present invention to the image processing apparatus will be described. However, the present invention is applicable to any device capable of outputting information related to diagnosis for assisting image diagnosis based on a medical image.

[0012] 《Configuration of Support System》 The system configuration of the support system according to the present embodiment is illustrated in FIG. 1. The support system records an image of a patient's affected part and provides a function of assisting diagnosis based on the image of the affected part to medical staff such as doctors. In the present embodiment, the support system is used in dentistry, and the image of the affected part is an intraoral image obtained by photographing the inside of the patient's oral cavity.

[0013] As illustrated, the support system includes an in-hospital system 109 and a patient information DB 110 that provide various functions related to hospital operation, and other configurations that provide functions related to dental treatment.

[0014] The hospital system 109 issues unique patient identification information (patient ID), such as a patient registration number, to each patient using the hospital (dental clinic), and manages information about each patient (patient information). Patient information is managed using the patient information DB 110. In addition to the patient ID, patient information may include information such as the patient's name, age, and purpose of visit. The hospital system 109 may also be configured to provide functions such as registering first-time patients and issuing medical examination slips. Various types of information to be included in patient information may be entered by hospital staff based on, for example, a medical questionnaire filled out when a first-time patient registers.

[0015] On the other hand, the configuration that provides functions related to dental treatment includes an electronic medical record terminal 106 used by physicians to view and edit electronic medical records, and an electronic medical record system 107 that manages the information (medical record information) of said electronic medical records. The management of medical record information is performed using a dental information database 108. The medical record information is configured to manage medical records for each patient, and includes, for example, information on diagnosis results and symptoms for medical treatment performed on patients within the most recent predetermined period. Therefore, the patient ID of each patient is associated with the medical record information for that patient. In addition, the dental information database 108 also records other information that can be referenced in relation to dental treatment. In the following explanation, this information managed by the dental information database 108 may be collectively referred to as dental information.

[0016] In this embodiment, the system is configured to use image diagnosis for dental treatment, and the support system includes an imaging device 103 for photographing the patient's oral cavity 104. The oral cavity images acquired by imaging using the imaging device 103 are, for example, visible light image data, and are recorded with a patient ID associated with them. For this reason, the imaging device 103 is configured to acquire the patient ID from a code image printed on, for example, a medical examination form, prior to the patient's treatment, and the photographer (dental staff or dentist) performs an operation to set the patient ID of the patient to be photographed before taking the oral cavity images. The oral cavity images are then transmitted from the imaging device 103 to the image processing device 101, and are registered and managed in the image management DB 102 by the image processing device 101. The oral cavity images registered in the image management DB 102 are configured to be displayed together with the electronic medical record on the electronic medical record terminal 106. The image processing device 101 is also capable of performing various image processing, such as analysis processing, on the oral cavity images, and can output the information obtained by image processing to the electronic medical record terminal 106 as support information to assist in diagnosis as needed.

[0017] Therefore, when viewing or editing an electronic medical record for a single patient on the electronic medical record terminal 106, patient information and medical record information associated with the patient's patient ID are acquired via the electronic medical record system 107 and the in-hospital system 109, respectively. In addition, an intraoral image associated with the patient's patient ID and supporting information obtained regarding that intraoral image are transmitted to the electronic medical record terminal 106 by the image processing device 101. The transmission and reception of this information is performed via the network 105.

[0018] <Hardware configuration of the image processing device> The hardware configuration of the image processing device 101 according to this embodiment will be described with reference to the block diagram in Figure 2.

[0019] The CPU (Central Processing Unit) 201 is a control device that controls the operation of each block in the image processing device 101. The CPU 201 reads the operation programs of each block stored in, for example, the ROM (Read Only Memory) 202 or the recording device 204, and loads them into the RAM (Random Access Memory) 203. The CPU 201 then controls the operation of each block by executing the programs loaded into the RAM 203.

[0020] ROM 202 and recording device 204 are devices configured to enable permanent information storage. Recording device 204 may be a hard disk drive (HDD) or a solid state drive (SSD). In addition to the operating program, ROM 202 and recording device 204 can store information such as parameters and setting information used for the operation of each block, and the results of various processes. In contrast, RAM 203 is a device configured to enable temporary information storage.

[0021] Input I / F205 is a user interface such as a mouse or keyboard. When Input I / F205 detects that input has been received from any of the user interfaces, it outputs the corresponding control signal to CPU201.

[0022] Communication I / F206 is an interface for sending and receiving information with external devices. Communication I / F206 may convert the data (information) to be transmitted into a format corresponding to the communication method with the external device as needed.

[0023] The display device 207 is, for example, an LCD. The display device 207 is used, for example, to display information generated by the operation of each block of the image processing device 101.

[0024] The graphics board 208 is hardware for performing various image processing tasks. The graphics board 208 has a GPU (Graphics Processing Unit) 211 and GPU memory 212. As will be described in detail later, in the support system of this embodiment, the image processing device 101 provides a function to infer symptoms occurring in the patient's oral cavity 104, for example, based on oral cavity images, and output this information to the electronic medical record terminal 106. Generally, GPUs are superior to CPUs in parallel processing capabilities, so they can efficiently construct inference models using deep learning (machine learning) and perform inference using these inference models. For this reason, in this embodiment, the GPU 211, or the CPU 201 and GPU 211 working together, is configured to perform inference processing using oral cavity images as input. When the GPU 211 is operating, the GPU memory 212 is used as a workspace.

[0025] Each piece of hardware in the image processing device 101 is configured to communicate information via the bus 209. That is, the bus 209 enables bidirectional communication of data or control signals between blocks.

[0026] <Hardware configuration of the electronic medical record terminal> Furthermore, the hardware configuration of the electronic medical record terminal 106 according to this embodiment will be described with reference to the block diagram in Figure 3.

[0027] The CPU 301 is a control device that controls the operation of each block in the electronic medical record terminal 106. The CPU 301 reads the operation programs of each block stored in, for example, the ROM 302 or the recording device 304, and loads them into the RAM 303. The CPU 301 then controls the operation of each block by executing the programs loaded into the RAM 303.

[0028] The ROM 302 and the recording device 304 are devices configured to enable permanent information storage. The recording device 304 may be an HDD or an SSD. In addition to the operating program, the ROM 302 and the recording device 304 can store information such as parameters and setting information used for the operation of each block, and the results of various processes. In contrast, the RAM 303 is a device configured to enable temporary information storage.

[0029] The input interface 305 is a user interface such as a mouse, keyboard, or touch panel. When the input interface 305 detects that input has been received from any of these user interfaces, it outputs the corresponding control signal to the CPU 301.

[0030] Communication I / F306 is an interface for sending and receiving information with external devices. Communication I / F306 may convert the data (information) to be transmitted into a format corresponding to the communication method with the external device as needed.

[0031] The display device 307 is, for example, an LCD. The display device 307 is used, for example, to display information generated by the operation of each block of the electronic medical record terminal 106. In the support system of this embodiment, the display device 307 of the electronic medical record terminal 106 is used to display the electronic medical record during patient treatment, etc.

[0032] Each piece of hardware in the electronic medical record terminal 106 is configured to communicate information via bus 308. That is, bus 308 enables bidirectional communication of data or control signals between blocks.

[0033] Overview of Diagnostic Imaging First, an overview of a dental examination, including image diagnosis, using the support system of this embodiment will be described with reference to the drawings.

[0034] Figure 4 shows an overview of the flow from patient registration to examination in a dental clinic. In sequence 400, once registration for treatment is completed at the hospital reception, a patient's consultation slip 402 is issued. This consultation slip 402 includes patient ID information, for example, in the form of a barcode. This patient ID information can be read by the imaging device 103 or the electronic medical record terminal 106 and used to identify the patient.

[0035] Subsequently, in sequence 410, the patient is guided to examination room 411 for intraoral imaging, where the imaging operator (dentist or dental staff) 412 uses imaging device 103 to photograph the inside of the oral cavity 104. At this time, for capturing intraoral images, a method such as the five-image method, which is one of the common imaging methods in dental practice, is employed.

[0036] An electronic medical record terminal 106 is provided in the examination room 411, and the electronic medical record 413 may be displayed on the display device 307 of the electronic medical record terminal 106 in a display mode for sequence 410. In displaying the electronic medical record 413, the electronic medical record terminal 106 reads (acquires) the patient ID information attached to the consultation slip 402. The electronic medical record terminal 106 then acquires corresponding information from the electronic medical record system 107 and the in-hospital system 109 based on the acquired patient ID. More specifically, the electronic medical record system 107 acquires the medical record information associated with the received patient ID from the dental information DB 108 and sends it back to the electronic medical record terminal 106. Similarly, the in-hospital system 109 acquires the patient information associated with the received patient ID from the patient information DB 110 and sends it back to the electronic medical record terminal 106. Based on this information, the electronic medical record terminal 106 can configure the information of the electronic medical record 413 to be displayed on the electronic medical record terminal 106 and display it on the display device 307.

[0037] Furthermore, the captured intraoral images are uploaded to the image processing device 101, for example, directly from the imaging device 103, or from the electronic medical record terminal 106 by connecting the imaging device 103 to the electronic medical record terminal 106 via a wired / wireless connection, and registered in the image management DB 102.

[0038] After the acquisition of intraoral images is complete, the dentist 421 examines the patient 401 in sequence 420. In the examination in sequence 420, the dentist 421 examines the patient 401 while referring to and using the electronic medical record 422 on the electronic medical record terminal 106, similar to sequence 410. At this time, the electronic medical record 422 includes intraoral images 423a to e taken of the patient 401, as shown in the figure. In this embodiment, intraoral imaging is performed using the five-image method, so the electronic medical record 422 contains intraoral images 423a relating to the maxilla, 423e relating to the mandible, 423b relating to the left lateral view, 423d relating to the right lateral view, and 423c relating to the frontal view.

[0039] As described above, each intraoral image is associated with a patient ID. Therefore, when displaying the image in the electronic medical record 422, the electronic medical record terminal 106 transmits the patient ID to the image processing device 101 and retrieves the intraoral image 423 related to patient 401. More specifically, the image processing device 101 retrieves the intraoral image associated with the received patient ID from the image management DB 102 and sends it back to the electronic medical record terminal 106.

[0040] The image processing device 101 is also configured to output support information for diagnostic assistance for the intraoral image. In this embodiment, the support information includes, for each tooth in the intraoral image, the dental formula of the tooth and the inference result of the tooth's condition (including symptoms appearing in the tooth). Here, the dental formula is a symbol used to indicate whether a human tooth is a deciduous tooth or a permanent tooth. The dental formula is common to both the upper and lower dental arches in the oral cavity, and symbols A to E are assigned to deciduous teeth and 1 to 8 to permanent teeth, in order from the frontmost tooth to the back of the dental arch on both the left and right sides. Specifically, for permanent teeth 1 to 8, 1 is the central incisor, 2 is the lateral incisor, 3 is the canine, 4 is the first premolar, 5 is the second premolar, 6 is the first molar, 7 is the second molar, and 8 is the third molar (wisdom tooth). Furthermore, A through E in the deciduous teeth indicate the following teeth: A is the central incisor, B is the lateral incisor, C is the canine, D is the first molar, and E is the second molar.

[0041] Therefore, at least some of the intraoral images 423 displayed in the electronic medical record 422 are overlaid with supporting information obtained for those images. This supporting information can, for example, inform the dentist 421 of teeth that should be carefully diagnosed and treated prior to the diagnosis. The dentist 421 can then visually inspect the patient 401's oral cavity and make a diagnosis while referring to the supporting information displayed in this way. The final diagnosis made by the dentist 421 is entered, for example, via a predetermined user interface of the electronic medical record 422 and transmitted to the electronic medical record system 107, where it is reflected in the patient 401's medical record information. Furthermore, the intraoral images of patient 401 obtained in this way, associated with dental formulas and condition information, can be stored, for example, in an external shared database and utilized for identification during disasters and for regional medical cooperation.

[0042] Furthermore, if there is an error in the support information provided by the image processing device 101, the dentist 421 may provide feedback to the image processing device 101 by, for example, making correction inputs in the electronic medical record terminal 106. In a configuration in which machine learning for generating support information is performed in the image processing device 101, as in the support system of this embodiment, the inference accuracy of the inference model can be improved by feeding back the correction information made by the dentist 421. Specifically, the inference accuracy of the inference model can be improved by providing the dental formula and tooth condition information corrected by the dentist 421 in the annotation process to the intraoral images taken in connection with the treatment of the patient 401 in this case, and then performing new machine learning.

[0043] 《Support information for determining dental formulas》 Incidentally, during the growth process, human teeth are replaced from deciduous teeth to permanent teeth. Therefore, even at the same position in the dentition, the type of tooth erupting at that position may differ due to differences in the growth process. Consequently, when the image processing device 101 takes an intraoral image as input and has the inference model infer the dental formula for each tooth, it is possible that both deciduous and permanent dental formulas may be obtained as inference results. In other words, the inference model may not be able to distinguish whether the tooth in question is a deciduous or permanent tooth, and may output multiple types of dental formulas as possible candidates (hereinafter referred to as dental formula candidates). In particular, for intraoral images of patients aged 4 to 15 years, when deciduous and permanent teeth are mixed in the dentition due to the eruption of permanent teeth, the inference model may infer multiple dental formula candidates for some teeth.

[0044] Therefore, the inference model that infers the dental formula of a tooth based on an intraoral image is configured to output, as an inference result for the target tooth (hereinafter referred to as the target tooth), a candidate dental formula that is thought to be the target tooth, and an evaluation value indicating the probability (likelihood) that the target tooth is the candidate dental formula. The evaluation value for the candidate dental formula is derived, for example, as a value between 0 and 1, with a value closer to 1 indicating a higher probability that the dental formula of the target tooth is the candidate dental formula.

[0045] Furthermore, since the distribution of teeth in the dentition can vary from patient to patient, different symbols (e.g., 2 and 3 for the target tooth) may be output as candidate dental formulas even for the same type of tooth (deciduous or permanent) at a specific location. For this reason, among the candidate dental formulas inferred by the inference model for the image of the target tooth included in the intraoral image, only those dental formula candidates whose evaluation value is above a predetermined threshold (e.g., 0.5) will be included in the support information as candidate dental formulas for the target tooth.

[0046] On the other hand, the evaluation values ​​for each dental formula candidate obtained through inference represent the likelihood output by the inference model constructed using image-based machine learning. Therefore, even if evaluation values ​​are obtained for multiple dental formula candidates, the dental formula candidate with the highest evaluation value does not necessarily represent the correct dental formula of the target tooth. Since teeth in the process of eruption vary in size, it is possible that a dental formula candidate related to a deciduous tooth may have a higher evaluation value than a permanent tooth.

[0047] Therefore, the image processing device 101 is configured to correct the evaluation values ​​of multiple dental formula candidates for the target tooth, which are inferred by the inference model, to values ​​that reflect the trend of the condition of the target tooth at the patient's age. As a result, the evaluation value of the dental formula candidate that is more likely for the target tooth at the patient's age will be relatively higher than the evaluation values ​​of the other dental formula candidates. Consequently, a more accurate (more age-appropriate) estimation result of the dental formula of the target tooth can be provided to the dentist as supporting information.

[0048] <Correction process based on patient age> In executing the correction process for the evaluation values ​​of such dental formula candidates, the CPU 201 of the image processing device 101 acquires information on the age of the target patient. Here, the target patient is the patient who is the subject of the intraoral image, which is input to the inference model and from which dental formula candidates and evaluation values ​​are derived. In this embodiment, the information on the age of the target patient is acquired from the patient information relating to the target patient. For this reason, when the CPU 201 inputs an intraoral image to the inference model and performs inference regarding dental formula candidates, it acquires patient information managed in the patient information DB 110, which is associated with the patient ID associated with the intraoral image. The patient information may be provided by the in-hospital system 109 in response to an acquisition request including the patient ID being sent from the image processing device 101 to the in-hospital system 109. Alternatively, the patient information may be provided to the image processing device 101 from the electronic medical record terminal 106 displaying the electronic medical record of the target patient.

[0049] CPU201 uses an inference model to infer candidate dental formulas and evaluation values ​​for each tooth that appears in the intraoral image. It then extracts dental formula candidates whose evaluation values ​​are above a threshold for correction processing. Based on the patient's age information, CPU201 performs a correction process to adjust the evaluation values ​​of the extracted dental formula candidates for each tooth.

[0050] The evaluation values ​​are corrected based on the information in the table shown in Figure 5. The table shows the average time of eruption (indicated by △ in the figure) and, if they are deciduous teeth, the average time of shedding (indicated by ▼ in the figure) for each tooth in each dental formula. Each row in the table corresponds to a dental formula, with records for deciduous teeth A-E and permanent teeth 1-8 arranged vertically. Each column in the table indicates age (or months).

[0051] At least some of the permanent teeth are indicated in the record with a triangle symbol (△) in two rows. The upper row indicates the time when the upper jaw teeth begin to erupt, and the lower row indicates the time when the lower jaw teeth begin to erupt. For example, the tooth in formula 1 can begin to erupt in both the upper and lower rows at age 7.

[0052] Thus, the table in Figure 5 defines the periods when deciduous teeth and permanent teeth exist in the human oral cavity, and will be referred to as the presence period table below. In other words, the presence period table is a table that defines information (change information) regarding age-related changes in teeth that erupt at each position in the human dentition. The presence period table may be stored, for example, in the dental information DB 108 and configured to be accessible to the image processing device 101 at any time.

[0053] According to the presence timing table, on average, the replacement of deciduous teeth with permanent teeth begins around the age of 6, and permanent teeth are fully erupted by around the age of 13. Therefore, in the support system of this embodiment, for example, the period from age 4 to 15 is defined as the period in which deciduous teeth and permanent teeth coexist in the oral cavity (hereinafter referred to as the mixed period), and if the patient's age falls within this period, a correction process is performed on the evaluation value output by the inference model. Information on the mixed period may be predetermined and stored in the recording device 204, for example, or it may be generated by the CPU 201 by referring to the presence timing table and recorded in the recording device 204, etc.

[0054] Furthermore, as shown in Figure 6, the existence timing table also includes information on correction coefficients that are multiplied by the evaluation value of each candidate tooth formula during the correction process. The correction coefficients are set based on the probability that each tooth formula is present at a given age. For example, looking at the correction coefficients in column 601 for the age of 9, the tooth formulas "D" and "4," which represent teeth erupting in the fourth position from the front, have correction coefficients of "1.0" and "0.7," respectively. This corresponds to the tendency for teeth erupting in the fourth position from the front in the oral cavity of a 9-year-old child to be more likely to be deciduous teeth than permanent teeth.

[0055] For example, let's consider a case where, as shown in Figure 7(a), the estimation results for the fourth tooth from the front on the upper left side are 0.54 for the deciduous tooth formula candidate "upper left D" and 0.56 for the permanent tooth formula candidate "upper left 4". In this case, if the patient is 9 years old, a correction coefficient of "1.0" is multiplied by the former evaluation value and "0.7" is multiplied by the latter evaluation value. As a result, the evaluation values ​​obtained after correction are 0.54 for the deciduous tooth formula candidate "upper left D" and 0.39 for the permanent tooth formula candidate "upper left 4", as shown in Figure 7(b), and supporting information is output that indicates a high likelihood that the target tooth is the tooth with formula D on the upper left side. In other words, for the target tooth, before correction, the evaluation value indicated a slightly higher likelihood that it was the tooth with formula 4 on the upper left side, but after correction, the evaluation value can be adjusted to reflect the trend in the patient's age, indicating a high likelihood that it is the tooth with formula D. In other words, the support information provided to the electronic medical record terminal 106 after correction processing is adjusted to increase the likelihood of the candidate dental formulas that would likely be present for the patient's age. This allows the dentist to be presented with support information that prioritizes age-related trends, even when the inference model is unable to distinguish between deciduous and permanent teeth.

[0056] In this embodiment, the correction coefficient is defined as a value between 0 and 1 in the existence time table, and a method of correction is described in which the evaluation value of dental formula candidates with a low probability based on age trends is lowered. However, the implementation of the present invention is not limited to this. The method of setting the correction coefficient may be changed to increase the evaluation value of dental formula candidates with a high probability based on age trends, or to increase the evaluation value of dental formula candidates with a high probability based on age trends while lowering the evaluation value of dental formula candidates with a low probability. In addition, the correction coefficient may be a value set by a dentist or the like in correspondence with the patient's age. Furthermore, the correction coefficient does not need to be defined for each dental formula at each age as shown in the existence time table, but may be configured to be derivable by, for example, a function with age as a variable.

[0057] As described above, in the support system of this embodiment, the image processing device 101 first derives the inference results of an inference model based on intraoral images taken of the patient's teeth, which are the examination site, as first information. The inference results include one or more candidate tooth formulas and an evaluation value indicating the likelihood of each candidate tooth formula for each tooth that is the examination site. Furthermore, the image processing device 101 applies a correction process to the evaluation values ​​obtained as first information, based on the patient's age and an existence time table that shows the age-related changes in the teeth erupting in the oral cavity. As a result, the image processing device 101 can generate evaluation values ​​for each candidate tooth formula that reflect the age-related change trend as second information for each tooth, and output them as support information.

[0058] The support information output in this manner is displayed in the electronic medical record terminal 106, for example, as shown in Figure 7(b), superimposed on the image of the tooth specified by selection or other operation in the intraoral image. This display shows candidate dental formulas for the specified tooth with evaluation values, allowing the dentist to refer to more accurate information before visual confirmation. Alternatively, the dentist can determine which dental formula is for the target tooth based on the evaluation values ​​in this display.

[0059] Output Processing Hereinafter, the output processing performed in the image processing device 101 of this embodiment for outputting support information will be described in detail using the flowchart in Figure 8. The processing corresponding to the flowchart can be realized by the CPU 201 reading the corresponding processing program stored in, for example, the ROM 202 or the recording device 204, loading it into the RAM 203, and executing it. This output processing will be described as starting, for example, when the electronic medical record terminal 106 receives a request from the electronic medical record terminal 106 to output support information for a specified intraoral image.

[0060] In S801, CPU201 acquires the intraoral image to be processed (hereinafter referred to as the target image). More specifically, CPU201 acquires the corresponding intraoral image (target image) from the image management DB102 based on information that uniquely identifies the intraoral image included in the output request.

[0061] In S802, CPU201 retrieves patient information corresponding to the patient ID associated with the target image. More specifically, CPU201 sends a request to the hospital system 109 to transmit patient information including the patient ID, and in response receives the relevant patient information managed in the patient information DB 110.

[0062] In this embodiment, the image processing device 101 is configured to estimate (1) a candidate dental formula including evaluation values ​​for each tooth in which an image appears in the target image, and (2) the state of each tooth (including symptoms, treatment marks, etc.) in order to output support information related to the target image. For this reason, the image processing device 101 has both an inference model for estimating the former and an inference model for estimating the latter. Accordingly, in this output process, the following processes S803 to S807 and the process S808 are performed in parallel in order to output each estimation result to be included in the support information.

[0063] In S803, CPU201 inputs the target image to the inference model and infers a candidate dental formula and its evaluation value for each tooth whose image appears in the target image. At this time, CPU201 excludes dental formula candidates whose evaluation value does not meet the threshold from among the dental formula candidates inferred by the inference model.

[0064] In S804, CPU201 determines whether the patient's age falls within the mixed period. Specifically, CPU201 refers to the patient's age information included in the patient information obtained in S802 and determines whether it falls within the age range defined as the mixed period (for example, 4 to 15 years old). If CPU201 determines that the patient's age falls within the mixed period, it moves the process to S805; otherwise, it moves the process to S809.

[0065] In S805, CPU201 selects teeth that appear in the target image but have not yet undergone correction processing as target teeth.

[0066] In S806, CPU201 performs a correction process to correct the evaluation value of the dental formula candidate obtained as an inference result for the target tooth.

[0067] <Correction process> Here, the correction process performed in this step will be explained in detail using the flowchart in Figure 9.

[0068] In S901, the CPU201 determines whether the dental formula candidates obtained as inference results for the target tooth include both deciduous tooth and permanent tooth dental formula candidates. In the correction process of this embodiment, if the dental formula candidates obtained as inference results are limited to either deciduous teeth or permanent teeth, it is determined that there is no error in the estimation of the tooth type (whether it is a deciduous tooth or a permanent tooth), and the correction is not applied. In other words, the situation in which the evaluation value is corrected in this correction process is when there are multiple dental formula candidates obtained as inference results for the target tooth, and when deciduous tooth and permanent tooth dental formula candidates are mixed together. Therefore, if the dental formula candidates obtained as inference results include both deciduous tooth and permanent tooth dental formula candidates, the CPU201 moves the process to S902, and if only one of them is included, the correction process is completed.

[0069] In S902, CPU201 obtains correction coefficients from the existence time table that correspond to the candidate dental formulas obtained as inference results for the target tooth. Specifically, CPU201 refers to the patient's age column in the record of the existence time table corresponding to each of the candidate dental formulas obtained as inference results, and obtains the correction coefficient for each candidate dental formula.

[0070] In S903, the CPU201 updates the evaluation value of the dental formula candidate obtained as an inference result for the target tooth by multiplying it by the correction coefficient obtained in S901, and completes this correction process. As a result of this step, for example, the evaluation value of each dental formula candidate in the inference results for the target tooth stored in RAM203 will be the corrected value.

[0071] Once the correction process is complete, CPU201 transfers the output processing to S807.

[0072] In S807, CPU201 determines whether correction processing has been performed on all teeth in the target image. If CPU201 determines that correction processing has been performed on all teeth, it moves the process to S809; if it determines that there are teeth that have not yet been processed, it returns the process to S805.

[0073] On the other hand, in S808, CPU201 inputs the target image to the inference model and infers the state of each tooth in which an image appears in the target image.

[0074] Once the inference of candidate dental formulas, evaluation values, and conditions is complete, the CPU 201 generates and outputs support information that integrates this information in S809. In this embodiment, the output destination of the support information is the electronic medical record terminal 106, and the CPU 201 transmits the generated support information to the communication I / F 206, causing the electronic medical record terminal 106 to receive it.

[0075] As described above, the image processing device of this embodiment can output information for diagnostic support that can reduce the burden on both doctors and patients. More specifically, it can provide evaluation values ​​as support information that have been corrected to reflect the tendency of teeth that have erupted at the patient's age for multiple candidate dental formulas inferred based on intraoral images, thereby enabling the presentation of more accurate estimation results to dentists and other medical professionals.

[0076] In this embodiment, the existence period table has been described as being of one type, but the implementation of the present invention is not limited to this. For example, since physical changes in the human body can differ depending on sex, an existence period table may be prepared for each sex. In this case, the correction coefficient for the evaluation value of the dental formula candidate is obtained from the corresponding existence period table, using not only the patient's age information included in the patient information, but also the patient's sex information. Also, since physical changes in the human body can vary depending on dietary culture and living environment, an existence period table may be prepared for each nationality or country of birth. In this case, the correction coefficient is obtained from the existence period table corresponding to the nationality or country of birth information included in the patient information. Furthermore, when multiple types of existence period tables are provided in this way, the coexistence period may also be set differently for each corresponding biological information (sex, nationality, country of birth, etc.).

[0077] [Example 1] In the embodiments described above, a method was described in which the evaluation value is corrected based on information on the tendency of teeth erupting at the patient's age in order to increase the accuracy of the estimation results regarding the patient's teeth output as support information. However, the method for increasing the accuracy of the estimation results is not limited to this. The present invention can also increase the accuracy of the estimation results by employing other methods in addition to, or instead of, the method of correcting based on information on the tendency of teeth erupting at the patient's age as in Embodiment 1.

[0078] Correction processing based on past diagnostic results One method involves referring to past diagnostic results for the target tooth. For example, if a past diagnosis has already determined that the tooth in the target tooth's position is a permanent tooth, then the deciduous tooth formula candidates included in the estimation results for the target tooth can be determined to be unnecessary (not suitable as a formula). Also, for example, if the size of the tooth image in the target position has increased between the intraoral image taken in the most recent diagnosis of the patient and the current image, it would support the conclusion that the tooth is a permanent tooth during the mixed-age period. That is, since the size of deciduous teeth does not increase by the time the mixed-age period arrives, if the size of the tooth image is larger than in the most recent intraoral image, it means either that the deciduous tooth has been replaced by a permanent tooth, or that the newly erupted permanent tooth has grown. Conversely, if it is immediately after the deciduous tooth has been replaced by a permanent tooth, the size of the tooth image may be smaller than in the most recent intraoral image, as shown in Figure 10, for example, so it is not possible to uniformly exclude deciduous tooth formula candidates. Here, the growth of the patient's teeth (increase in image size) can be identified, for example, by including a fixed-size object that serves as a reference point within the field of view when taking each intraoral image.

[0079] Therefore, by referring to past diagnostic results, it is possible to exclude inapplicable dental formula candidates from among those inferred by the inference model. Alternatively, similar to Embodiment 1, the evaluation value of the likely dental formula candidate can be corrected to be relatively higher than the evaluation value of the deciduous tooth dental formula candidate.

[0080] <Correction process> The correction process using a method that references past diagnostic results of the target tooth will be explained below using the flowchart in Figure 11. The correction process shown in the flowchart in Figure 11 may be performed in place of the correction process S806 of the correction process in Embodiment 1, or it may be performed in addition to the said correction process. Prior to the execution of the correction process, if information related to the patient's most recent diagnostic results (including intraoral images taken at that time) exists in the dental information DB 108, such information will be stored in the RAM 203 of the image processing device 101. The information related to the patient's most recent diagnostic results may be, for example, part of the medical record information used to display the electronic medical record in the electronic medical record terminal 106. This information is transmitted to the image processing device 101 when the electronic medical record terminal 106 sends it to the image processing device 101, or when the electronic medical record system 107 retrieves it from the dental information DB 108 and sends it back in response to the image processing device 101 sending an acquisition request.

[0081] In S1101, the CPU 201 determines whether the candidate dental formulas obtained as inference results for the target tooth include candidate dental formulas for deciduous teeth. If the candidate dental formulas obtained as inference results include candidate dental formulas for deciduous teeth, the CPU 201 moves the process to S1102; otherwise, it completes this correction process.

[0082] In S1102, the CPU 201 determines whether information relating to the patient's most recent diagnosis is stored in the RAM 203. If the CPU 201 determines that information relating to the patient's most recent diagnosis is stored in the RAM 203, it moves the process to S1103. If the CPU 201 determines that information relating to the patient's most recent diagnosis is not stored in the RAM 203, i.e., if there are no past diagnosis results for the patient, it completes this correction process.

[0083] In S1103, CPU201 determines whether the target tooth has been determined to be a permanent tooth in the most recent diagnostic results. If CPU201 determines that the target tooth has already been determined to be a permanent tooth, it moves the process to S1105; otherwise, it moves the process to S1104.

[0084] In S1104, the CPU201 determines whether the size of the image of the target tooth has increased compared to the most recent diagnosis. More specifically, the CPU201 makes this determination based on the difference in the size of the tooth images at the target tooth's position between the intraoral image to be output and the intraoral image taken in the most recent diagnosis. If the CPU201 determines that the size of the image of the target tooth has increased compared to the most recent diagnosis, it moves the process to S1105; otherwise, it completes this interpolation process.

[0085] In S1105, the CPU201 completes the correction process by excluding deciduous tooth candidate dental formulas from the candidate dental formulas obtained as inference results for the target tooth. In other words, when this step is performed, it is confirmed that the target tooth is at least a permanent tooth.

[0086] In this case, if there is only one candidate dental formula after exclusion, the CPU 201 may determine that the remaining candidate dental formula is the dental formula of the target tooth. In this case, since the dental formula determination for the target tooth has been completed in the support information provided to the electronic medical record terminal 106, instead of displaying an evaluation value and requesting the dentist to determine the dental formula, the system may display the dental formula of the target tooth as if the determination had been completed.

[0087] By using this correction method, which references past diagnostic results of the target tooth, the candidate dental formulas inferred by the inference model can be narrowed down to those with higher accuracy.

[0088] Corrective processing based on other teeth (order of tooth replacement) Another method involves referring to the estimation, assessment, or diagnosis results of other teeth. As shown in the existence timing table, tooth replacement generally occurs in the order from the front teeth to the back teeth. That is, tooth replacement generally occurs in the order that teeth with dental formula A are replaced by teeth with dental formula 1, teeth with dental formula B are replaced by teeth with dental formula 2, teeth with dental formula C are replaced by teeth with dental formula 3, teeth with dental formula D are replaced by teeth with dental formula 4, and teeth with dental formula E are replaced by teeth with dental formula 5.

[0089] Therefore, for example, if, with respect to other teeth located in front of the target tooth, the determination that they are deciduous teeth has already been confirmed at the time of the correction process for the target tooth, or if the estimated results only consist of deciduous tooth formula candidates and there is a high probability that they are deciduous teeth, then the probability that the target tooth is a permanent tooth is low. In addition, if, in the estimation results for other teeth, the evaluation value of the deciduous tooth formula candidate is higher than, for example, the evaluation value of the permanent tooth formula candidate by a predetermined value, then it may be treated as a high probability that the other teeth are deciduous teeth. In other words, based on the premise that the order of tooth replacement is not reversed, if there is a high probability that other teeth are deciduous teeth, then even if deciduous tooth and permanent tooth formula candidates are inferred for the target tooth located behind those other teeth, the probability that it is the latter is low.

[0090] For example, as shown in Figure 12, suppose that for tooth 1201, the fourth tooth from the left of the maxilla, the inference results yield a candidate deciduous tooth formula (maxilla left formula D) and a candidate permanent tooth formula (maxilla left formula 4). In this case, if the estimation result for tooth 1202, the second tooth from the right of the maxilla, suggests that tooth 1202 is likely to be a deciduous tooth with maxilla right formula B, then tooth 1201 is likely to be a deciduous tooth. In other words, the inference result for the candidate permanent tooth formula (maxilla left formula 4) for tooth 1201 is likely to be incorrect.

[0091] Therefore, by referring to the estimation results of other teeth located anterior to the target tooth, the evaluation values ​​of the less likely dental formula candidates inferred by the inference model can be relatively reduced. More specifically, if it can be determined from the estimation results of the other teeth that there is a high probability that they are deciduous teeth, the evaluation value of the deciduous tooth dental formula candidate in the inference results for the target tooth may be corrected to be relatively higher than the evaluation value of the permanent tooth dental formula candidate. Alternatively, the evaluation value of the permanent tooth dental formula candidate in the inference results for the target tooth may be corrected to be relatively lower than the evaluation value of the deciduous tooth dental formula candidate, or corrections that achieve both may be applied to the evaluation values ​​of both teeth.

[0092] <Correction process> The correction process using a method that references past diagnostic results of the target tooth will be explained below using the flowchart in Figure 13. The correction process shown in the flowchart in Figure 13 may be performed in place of the correction process in S806 of the correction process in Embodiment 1, or it may be performed in addition to the said correction process or the correction process shown in the flowchart in Figure 11. Alternatively, at least one of the correction processes shown in Figures 9, 11, and 13 may be performed in output process S806.

[0093] In S1301, the CPU 201 determines whether the candidate dental formula obtained as an inference result for the target tooth includes both a candidate dental formula for a deciduous tooth and a candidate dental formula for a permanent tooth. If the candidate dental formula obtained as an inference result includes both a candidate dental formula for a deciduous tooth and a candidate dental formula for a permanent tooth, the CPU 201 moves the process to S1302; if it includes only one of them, the correction process is completed.

[0094] In S1302, the CPU201 determines whether there are other teeth growing in an anterior position to the target tooth. If the CPU201 determines that there are other teeth in an anterior position to the target tooth, it moves the process to S1303; otherwise, it completes this correction process.

[0095] In S1303, CPU201 determines whether there are any teeth among the other teeth that are presumed to be deciduous teeth (hereinafter referred to as presumed deciduous teeth). Here, a presumed deciduous tooth refers to a tooth that has been determined to be a deciduous tooth as described above, whose inference result consists only of deciduous tooth formula candidates, and whose evaluation value of the deciduous tooth formula candidate is higher than the evaluation value of the permanent tooth formula candidate, exceeding a threshold. If CPU201 determines that a presumed deciduous tooth exists among the other teeth, it moves the process to S1304; otherwise, it completes this correction process.

[0096] In S1304, the CPU201 adjusts and updates the evaluation value of the dental formula candidate obtained as an inference result for the target tooth, and completes this correction process. Here, the adjustment of the evaluation value can be done, for example, by multiplying the evaluation value of the permanent tooth dental formula candidate by a value less than 1.

[0097] By using this correction method, which references the estimation results of other teeth, the evaluation value of the dental formula candidate inferred by the inference model can be adjusted to a more accurate state.

[0098] In the correction process shown in Figure 13, an adjustment was made to the evaluation value of the candidate dental formula for the target tooth based on the estimation results of other teeth anterior to the target tooth, among the estimation results obtained for the same intraoral image. However, the adjustment of evaluation values ​​based on other teeth is not limited to this. For example, if it has been confirmed in a previous diagnosis that other teeth posterior to the target tooth are permanent teeth, the estimation result for the target tooth may be adjusted to relatively lower the evaluation value of the candidate dental formula for a deciduous tooth. Alternatively, if the estimation result in this case determines that the tooth is a permanent tooth, the estimation result for the target tooth may be adjusted to relatively lower the evaluation value of the candidate dental formula for a deciduous tooth.

[0099] [Differentiation 2] In the modified example described above, the image processing device 101 provides support information to the electronic medical record terminal 106, the electronic medical record terminal 106 displays the estimated results of each tooth, namely the candidate dental formula and evaluation value, superimposed on the intraoral image, and finally the dentist determines the dental formula of the tooth. However, the implementation of the present invention is not limited to this, and if the dental formula of a tooth is clear from the estimation results, the determination of the dental formula of the tooth may be finalized on the image processing device 101 side without relying on the dentist.

[0100] [Difference 3] In the embodiments described above, the present invention was implemented in an image processing device 101 that generates support information in a support system used in dentistry. However, the implementation of the present invention is not limited to this. The function of estimating the dental formula of each tooth in the oral cavity based on an intraoral image is not limited to use in dental treatment at a medical institution, but may also be provided, for example, for checkups that do not require a doctor's consultation or for simple home examinations.

[0101] Accordingly, the present invention is also applicable to systems that do not have an image management DB 102, an electronic medical record terminal 106, an electronic medical record system 107, an in-hospital system 109, and a patient information DB 110, as shown in Figure 14. In the system of Figure 14, intraoral images captured by the imaging device 103 are transmitted to the examination terminal 1401 and sent to the image processing device 101 via the network 105. The image processing device 101 has an inference model infer candidate teeth and evaluation values ​​for the received intraoral images, performs a correction process, and returns the corrected evaluation values ​​obtained as estimation results for the teeth in the oral cavity 104 to the examination terminal 1401. This allows the estimation results to be displayed together with the intraoral images on the examination terminal 1401. In this embodiment, biometric information (age, gender, nationality, etc.) about the person being examined is input and stored on the examination terminal 1401. Furthermore, the intraoral images and estimation results may be uploaded to and registered in the dental information DB 108 for management, or they may be stored and managed in the examination terminal 1401. The intraoral images and estimation results can be used as past estimation results in subsequent examinations.

[0102] Furthermore, the present invention is also applicable to a system that does not have an image processing device 101, as shown in Figure 15. In this embodiment, the examination terminal 1401 can be any information and communication terminal with imaging capabilities, such as a smartphone. In this embodiment, the various functions that were performed by the image processing device 101 are performed by the examination terminal 1401. At this time, the inference model is stored in the dental information DB 108 and should be downloaded to the examination terminal 1401 prior to the execution of the examination.

[0103] Furthermore, as shown in Figure 16, the present invention can also be implemented on the inspection terminal 1401 alone. In this embodiment, all the information necessary for the inspection terminal 1401 (biological information, existence time table, inference model, past estimation results) is aggregated.

[0104] Furthermore, the intraoral images associated with the estimation results output by the system shown in Figures 14 to 16 may be stored in an external shared database as described above, or they may be provided to a dental clinic when the examined person visits for dental treatment. By providing such information, doctors can easily grasp the condition of the patient's oral cavity and the status of tooth eruption, and as a result, a reduction in consultation time in dental treatment can be expected.

[0105] [Differentiation Example 4] The embodiments and modifications described above describe the application of the present invention to the estimation of a dental formula for human teeth, but the implementation of the present invention is not limited thereto. For example, in an embodiment in which a time-of-existence table is provided for animals such as dogs and cats, whose teeth are replaced as they age, the present invention can also be applied to the estimation of tooth types for animals other than humans.

[0106] [Difference 5] In the embodiments and modifications described above, a method was explained in which, for teeth whose type (dental formula) changes with age, support information is output that corrects the estimated type based on biological information including age and a table of existence time. However, the implementation of the present invention is not limited to this, and the present invention is applicable in any manner in which an estimated result is output for diagnostic purposes, based on a medical image taken of a biological part that is a subject to diagnosis and whose condition changes with age. In this case, the estimated result generated as first information includes a candidate state for the biological part and its evaluation value, and second information of the biological part is generated and output based on the evaluation value, biological information relating to the organism, and change information relating to changes in the biological part due to aging.

[0107] [Embodiment 2] In the embodiments and modifications described above, a method was explained in which, based on medical images of a biological site, a candidate state, which is an estimated result of the biological site, and an evaluation value indicating the likelihood of the candidate state are derived as first information, and then corrected to generate and output second information. On the other hand, from the viewpoint of outputting information for diagnostic support, the first and second information are not limited to information that estimates the state of the biological site, but can be other forms of information. Other forms of such first and second information may include, for example, display parameters (window parameters) in the image diagnosis of breast diseases.

[0108] In diagnosing breast diseases, specialists interpret medical images such as mammograms and X-ray CT scans of the breast. When interpreting mammograms, physicians identify potential lesions in the breast (such as tumor shadows and microcalcification clusters). While medical images such as mammograms and X-ray CT scans have 10-bit to 16-bit color depth, the display devices used for interpretation may have lower color depth capabilities, such as 8-bit. In such cases, physicians interpret the images while performing various display processing techniques tailored to the region of interest. These processing techniques include, for example, zooming in on the region of interest, black and white inversion, and windowing.

[0109] Here, windowing is the process of converting (mapping) the grayscale levels of a medical image in the region of interest to the maximum number of grayscale levels that the display device can show, thereby creating a brightness and contrast state that makes the region of interest easier to observe. For grayscale conversion, it is necessary to set the window width, which is the range of pixel values ​​to be mapped to the maximum number of grayscale levels, and the window level, which is the pixel value to be at the center of that window width. Since it is cumbersome for doctors to set these parameters manually, in recent years, an auto-windowing function that automates the setting of these parameters based on, for example, the pixel value distribution of mammography has been adopted.

[0110] In mammography, pixel values ​​depend on the density of the breast tissue. High density of the breast tissue results in a low-dose area and therefore high pixel values, while low density results in a high-dose area and therefore low pixel values. Generally, the density of the breast tissue is very high and uniform in young women and nulliparous women. On the other hand, as the breast tissue regresses with age and fat replacement progresses, the density of the breast tissue tends to decrease and become non-uniform with age. Therefore, in young patients (subjects), areas with high pixel values ​​are distributed over a wide area within the breast, making it difficult to detect potential lesions in those areas.

[0111] Therefore, the image processing device 101 of this embodiment adjusts the window parameters derived based on the pixel value distribution of the mammography to values ​​that take into account the density trend of the breast tissue at the patient's age, and then outputs them to a device that controls the display of the mammography. That is, the CPU 201 of the image processing device 101 processes the window parameters (first information) derived based on the mammography to change them based on the patient's age and information on the change in density of the breast tissue with aging. For example, if the patient is young, the CPU 201 sets the window level of the window parameters of the first information higher and suppresses the window width to the higher frequency side, making it easier to see subtle changes in pixel values ​​in areas with high pixel values. That is, the CPU 201 further changes the window parameters of the first information to a state where lesion candidates are easier to see even in areas with high density of breast tissue, generates window parameters for the second information, and outputs them to a device that controls the display.

[0112] As described above, the image processing apparatus of this embodiment can output display parameters for diagnostic support that can reduce the burden on both doctors and patients. Specifically, from the doctor's perspective, potential lesions become easier to see even in areas of high-density breast tissue in medical images, especially when the patient is young, making it possible to perform accurate diagnoses and appropriate treatments. From the patient's perspective, the number of re-examinations decreases, reducing the physical burden.

[0113] [Other embodiments] The present invention can also be realized by supplying a program that implements one or more of the functions of the above-described embodiments to a system or device via a network or storage medium, and by having one or more processors in the computer of that system or device read and execute the program. It can also be realized by a circuit (e.g., an ASIC) that implements one or more functions.

[0114] The invention is not limited to the embodiments described above, and various modifications and variations are possible without departing from the spirit and scope of the invention. Accordingly, claims are attached to disclose the scope of the invention.

[0115] [Summary of Embodiments and Modifications] The disclosures herein include the following image processing apparatus, control method, and program. (Item 1) A first acquisition means for acquiring biological information about the organism to be diagnosed, A second acquisition means for acquiring medical images of the area being examined in the living body, A derivation means for deriving first information relating to the diagnosis of the examination site based on the medical image, An output means that generates and outputs second information relating to the diagnosis of the examination site based on the first information, the biological information, and change information relating to changes in the examination site due to aging, has An image processing apparatus characterized by the following: (Item 2) The output means generates the second information by correcting the first information based on the biological information and the change information. The image processing apparatus according to item 1, characterized in that it is a picture processing apparatus. (Item 3) The aforementioned examination site is a biological site whose condition changes with age. The first piece of information includes an estimated result of which of a plurality of candidate states the state of the inspected area is. The image processing apparatus according to item 2, characterized in that (Item 4) The estimation result includes, for each of the plurality of candidate states, an evaluation value indicating the likelihood that the state of the inspection site is that candidate state. The output means corrects the evaluation value based on the biological information and the change information. The image processing apparatus according to item 3, characterized in that (Item 5) The aforementioned examination site is a tooth. The first piece of information includes an estimated result of which of several candidate dental formulas the dental formula of the examined site is. The image processing apparatus according to item 2, characterized in that (Item 6) The estimation results include, for each of the multiple candidate dental formulas, an evaluation value indicating the likelihood that the dental formula of the examination site is that candidate dental formula. The output means corrects the evaluation value based on the biological information and the change information. The image processing apparatus according to item 5, characterized in that... (Item 7) The aforementioned list of candidate dental formulas includes dental formulas for deciduous teeth and dental formulas for permanent teeth. The image processing apparatus according to item 5 or 6, characterized by the features described herein. (Item 8) The derivation means includes the inference results inferred by the inference model using the medical image as input in the first information as the estimation result. An image processing apparatus according to any one of items 3 to 7, characterized in that it is an image processing apparatus. (Item 9) The aforementioned biological information further includes information on the diagnostic results of other examination sites of the biological body, The output means corrects the first information based on the diagnostic results information of the other inspection site and the change information. An image processing apparatus according to any one of items 3 to 8, characterized in that it is an image processing apparatus. (Item 10) The aforementioned biological information further includes information on past diagnostic results of the examination site of the biological body, The output means corrects the first information based on the information of past diagnostic results. An image processing apparatus according to any one of items 3 to 9, characterized in that (Item 11) The aforementioned biological information further includes past medical images of the examination site of the biological body, The output means corrects the first information based on the difference between the image of the examination site in the medical image and the past medical image. An image processing apparatus according to any one of items 3 to 10, characterized in that (Item 12) The system further includes a display means for displaying the second information output by the output means and the medical image. An image processing apparatus according to any one of items 3 to 11, characterized in that (Item 13) The system further comprises a determination means for determining the diagnostic result of the examination site of the living organism based on the second information output by the output means. An image processing apparatus according to any one of items 3 to 12, characterized in that (Item 14) The first piece of information is a display parameter set in relation to the display of the medical image, The output means outputs the display parameters, which have been modified based on the biological information and the change information, as the second information. The image processing apparatus according to item 1, characterized in that it is a picture processing apparatus. (Item 15) The aforementioned examination site is the breast, The aforementioned medical images are mammograms or X-ray CT images. The aforementioned display parameters include the window width and window level in windowing. The image processing apparatus according to item 14, characterized in that (Item 16) The aforementioned biological information includes information on the age of the organism, The aforementioned change information defines the trend in the state of the examination site corresponding to age. The output means corrects the first information based on the trend of the state of the examination site corresponding to the age of the living organism. An image processing apparatus according to any one of items 1 to 15, characterized in that (Item 17) The aforementioned change information is provided for each gender, The aforementioned biological information further includes information on the sex of the organism, The output means corrects the first information based on the change information corresponding to the sex of the living organism. The image processing apparatus according to item 16, characterized in that (Item 18) The aforementioned change information is provided for each nationality. The aforementioned biological information further includes information on the nationality of the organism, The output means corrects the first information based on the change information corresponding to the nationality of the living organism. The image processing apparatus according to item 16 or 17, characterized in that it is an image processing apparatus. (Item 19) The first acquisition step involves obtaining biological information about the organism to be diagnosed, A second acquisition step involves acquiring a medical image of the area being examined in the biological body, A derivation step of deriving first information relating to the diagnosis of the examination site based on the medical image, An output step that generates and outputs second information relating to the diagnosis of the examination site based on the first information, the biological information, and change information relating to changes in the examination site due to aging, has A method for controlling an image processing apparatus, characterized by the features described above. (Item 20) A program for causing a computer to function as one of the means of an image processing apparatus described in any one of items 1 through 18. [Explanation of Symbols]

[0116] 101: Image processing device, 201: CPU, 202: ROM, 203: RAM, 204: Recording device, 208: Graphics card, 211: GPU, 212: GPU memory, 1401: Inspection terminal

Claims

1. A first acquisition means for acquiring biological information about the organism to be diagnosed, A second acquisition means for acquiring medical images of the area being examined in the living body, A derivation means for deriving first information relating to the diagnosis of the examination site based on the medical image, An output means that generates and outputs second information relating to the diagnosis of the examination site based on the first information, the biological information, and change information relating to changes in the examination site due to aging, has An image processing apparatus characterized by the following:

2. The output means generates the second information by correcting the first information based on the biological information and the change information. The image processing apparatus according to claim 1.

3. The aforementioned examination site is a biological site whose condition changes with age. The first piece of information includes an estimated result of which of a plurality of candidate states the state of the inspected area is. The image processing apparatus according to claim 2.

4. The estimation result includes, for each of the plurality of candidate states, an evaluation value indicating the likelihood that the state of the inspection site is that candidate state. The output means corrects the evaluation value based on the biological information and the change information. The image processing apparatus according to claim 3.

5. The aforementioned examination site is a tooth. The first piece of information includes an estimated result of which of several candidate dental formulas the dental formula of the examined site is. The image processing apparatus according to claim 2.

6. The estimation results include, for each of the multiple candidate dental formulas, an evaluation value indicating the likelihood that the dental formula of the examination site is that candidate dental formula. The output means corrects the evaluation value based on the biological information and the change information. The image processing apparatus according to feature 5.

7. The aforementioned list of candidate dental formulas includes dental formulas for deciduous teeth and dental formulas for permanent teeth. The image processing apparatus according to feature 5.

8. The derivation means includes the inference results inferred by the inference model using the medical image as input in the first information as the estimation result. The image processing apparatus according to claim 3.

9. The aforementioned biological information further includes information on the diagnostic results of other examination sites of the biological body, The output means corrects the first information based on the diagnostic results information of the other inspection site and the change information. The image processing apparatus according to claim 3.

10. The aforementioned biological information further includes information on past diagnostic results of the examination site of the biological body, The output means corrects the first information based on the information of past diagnostic results. The image processing apparatus according to claim 3.

11. The aforementioned biological information further includes past medical images of the examination site of the biological body, The output means corrects the first information based on the difference between the image of the examination site in the medical image and the past medical image. The image processing apparatus according to claim 3.

12. The system further includes a display means for displaying the second information output by the output means and the medical image. The image processing apparatus according to claim 3.

13. The system further comprises a determination means for determining the diagnostic result of the examination site of the living organism based on the second information output by the output means. The image processing apparatus according to claim 3.

14. The first piece of information is a display parameter set in relation to the display of the medical image, The output means outputs the display parameters, which have been modified based on the biological information and the change information, as the second information. The image processing apparatus according to claim 1.

15. The aforementioned examination site is the breast, The aforementioned medical images are mammograms or X-ray CT images. The aforementioned display parameters include the window width and window level in windowing. The image processing apparatus according to feature 14.

16. The aforementioned biological information includes information on the age of the organism, The aforementioned change information defines the trend in the state of the examination site corresponding to age. The output means corrects the first information based on the trend of the condition of the examination site corresponding to the age of the living organism. The image processing apparatus according to claim 1.

17. The aforementioned change information is provided for each gender, The aforementioned biological information further includes information on the sex of the organism, The output means corrects the first information based on the change information corresponding to the sex of the living organism. The image processing apparatus according to claim 16.

18. The aforementioned change information is provided for each nationality. The aforementioned biological information further includes information on the nationality of the organism, The output means corrects the first information based on the change information corresponding to the nationality of the living organism. The image processing apparatus according to feature 16.

19. The first acquisition step involves obtaining biological information about the organism to be diagnosed, A second acquisition step involves acquiring a medical image of the area being examined in the aforementioned living organism, A derivation step of deriving first information relating to the diagnosis of the examination site based on the medical image, An output step that generates and outputs second information relating to the diagnosis of the examination site based on the first information, the biological information, and change information relating to changes in the examination site due to aging, has A method for controlling an image processing apparatus, characterized by the features described above.

20. A program for causing a computer to function as one of the means of an image processing apparatus according to any one of claims 1 to 18.

Images