Prediction system, prediction device, prediction method, control program, and recording medium

Abstract

The present invention makes it possible to highly accurately predict the effect of a drug when the drug is administered to a subject. A prediction system according to the present invention comprises: a storage unit that stores subject information pertaining to a bone of a subject; and a reception unit that receives drug efficacy prediction information pertaining to the effect of a drug when the drug is administered to the subject. The subject information includes at least a medical image showing the bone of the subject. The prediction system generates the drug efficacy prediction information on the basis of a drug efficacy prediction model that predicts the effect of a drug having an action on bones on the basis of the subject information and drug information pertaining to the drug.

Description

[Technical field]

[0001] The present disclosure relates to a prediction system, a prediction device, a control method, and a control program for predicting the effect of a drug that acts on bones. [Background technology]

[0002] Patent Document 1 discloses a technique for predicting the effect of an anticancer drug. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-013583 Summary of the Invention

[0004] A prediction system according to one embodiment of the present disclosure includes a memory unit that stores object information relating to the bones of a subject, and a receiving unit that receives efficacy prediction information relating to the effect of a drug when administered to the subject, wherein the object information includes at least a medical image depicting the bones of the subject, and generates the efficacy prediction information based on an efficacy prediction model that predicts the effect of the drug based on the object information and drug information relating to a drug having an action on bones.

[0005] A prediction device according to one embodiment of the present disclosure includes an acquisition unit that acquires object information related to the bones of a subject, and a prediction unit that outputs efficacy prediction information related to the effect of a drug when administered to the subject, wherein the object information includes at least a medical image showing the bones of the subject, and the prediction unit has an efficacy prediction model that predicts the effect of the drug based on the object information and drug information related to a drug having an action on bones.

[0006] In addition, a control method according to one aspect of the present disclosure includes a storage step of storing object information regarding the bones of a subject, a generation step of generating efficacy prediction information regarding the effect of a drug when administered to the subject based on an efficacy prediction model that predicts the effect of the drug based on the object information and drug information regarding a drug having an action on bones, and a receiving step of receiving the efficacy prediction information, wherein the object information includes at least a medical image showing the bones of the subject.

[0007] Furthermore, the prediction system according to each aspect of the present disclosure may be realized by one or more computers. In this case, the control program for the prediction system that causes a computer to operate as each unit (software element) of the prediction system to realize the prediction system, and the computer-readable recording medium on which the control program is recorded, also fall within the scope of the present disclosure. [Brief description of the drawings]

[0008] [Figure 1] FIG. 1 is a diagram illustrating a configuration example of a prediction system according to an embodiment of the present disclosure. [Diagram 2] FIG. 13 is a diagram illustrating a configuration example of a prediction system according to another aspect of the present disclosure. [Diagram 3] 1 is a block diagram illustrating an example of a configuration of a prediction device according to an aspect of the present disclosure. [Figure 4] FIG. 2 is a diagram showing an example of the configuration of a drug efficacy prediction model executed by a prediction unit. [Diagram 5] 13 is a flowchart showing an example of the flow of a learning process by a learning unit. [Figure 6] 13 is a flowchart showing an example of a flow of a process performed by the prediction device. [Figure 7] FIG. 13 is a diagram illustrating an example of a medical service realized by the prediction device. [Figure 8] FIG. 13 is a block diagram illustrating an example of a configuration of a prediction device according to another aspect of the present disclosure. [Figure 9] 13 is an example of a processed image including information on the basis of which an image change due to another reason is derived. [Figure 10] 13 is an example of a browser image showing analysis results both including and excluding image changes due to other reasons. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] Drugs for improving the bone condition of a subject include drugs that act on bone formation, drugs that act on bone resorption, and drugs that balance bone remodeling. These drugs may be administered to a subject alone or in combination.

[0010] According to one aspect of the present disclosure, it is possible to accurately predict the effect of a drug when the drug is administered to a subject.

[0011] The present disclosure will be described in detail below.

[0012] The condition of the bones of a subject may be influenced by the balance between bone formation and bone resorption, lifestyle habits such as diet and exercise, and genetics. For example, osteoporosis is a disease characterized by low bone mass and abnormalities in the microstructure of bone tissue, which increases bone fragility and increases the risk of fracture. Osteoporosis is often seen in elderly people and postmenopausal women. Osteoporosis may also be caused by, but is not limited to, local circulatory disorders, calcium metabolic disorders, and the like.

[0013] If the deterioration of the subject's bone condition leads to fractures in areas important for motor function, surgery and hospitalization may be required. This can result in a significant decline in the subject's quality of life (QOL). In order to reduce the decline in the subject's QOL, it is important to detect the deterioration of the subject's bone condition as early as possible and begin appropriate intervention, including the administration of appropriate medication.

[0014] Each embodiment of the present disclosure will be described below. In the following description, the subject is a human (i.e., a "subject"), but the subject is not limited to a human. The subject may be, for example, a non-human mammal, such as an equine, feline, canine, bovine, or porcine animal. The present disclosure also includes, among the following embodiments, embodiments in which the "subject" is replaced with "animal" if the embodiment is applicable to these animals.

[0015] [Embodiment 1] (Outline of Prediction Device 1) A prediction device 1 according to one embodiment of the present disclosure outputs efficacy prediction information regarding the effect of a drug when administered to a subject, based on subject information, which is information regarding a subject (hereinafter sometimes referred to as a "subject"), and drug information regarding the drug.

[0016] The subject information includes at least one of fracture risk factor information related to the subject's fracture risk factors, bone strength information related to the strength of the subject's bones, information related to the subject's health condition, bone density information indicating the bone density of the subject's bones, bone metabolism information indicating the bone metabolism state of the subject's bones, and medical images showing the subject's bones. In the present disclosure, the subject's bones may be multiple bones included in each part of the subject's body, such as the chest, lower back, neck, hip joint, head, legs, arms, fingers, and feet. Alternatively, in the present disclosure, the subject's bones may be specific bones, such as the spine, thoracic vertebrae, lumbar vertebrae, pelvis, cervical vertebrae, skull, femur, fibula, tibia, radius, metacarpals, phalanges, calcaneus, metatarsals, and phalanges of the subject.

[0017] The subject information may include fracture risk factor information regarding the subject's fracture risk factors. For example, the fracture risk factor information may be information used by a fracture risk assessment tool (FRAX (registered trademark); or the results thereof). The fracture risk factor information may include at least one of the following information group A.

[0018] (Information group A): Type (e.g., race), age, sex, weight, height, presence or absence of fracture, fracture location, fracture history, family history of fracture (e.g., parents), glucocorticoids, rheumatoid arthritis, secondary osteoporosis, underlying diseases (e.g., food and / or drug allergies, diseases related to the development of osteoporosis and diseases unrelated to it, etc.), smoking history, drinking habits (e.g., drinking frequency and amount, etc.), occupational history, exercise history, medical history (e.g., history of bone disease), menstruation (e.g., cycle and presence or absence, etc.), menopause (e.g., possibility and presence or absence, etc.), estimated time of menopause (e.g., estimated year or month or date, etc.), artificial joint (e.g., type, presence or absence of spinal implant or knee joint, and timing of replacement surgery), blood test results, urine test results, medications being taken, gene sequence.

[0019] The information group A is information that is useful for judging the subject's bone density, bone metabolism, possibility of fracture, and balance between bone formation and bone resorption, etc. Therefore, by using subject information including at least one piece of information in the information group A, the prediction device 1 can more accurately predict the effect of a drug to be administered to the subject.

[0020] The information on the presence or absence of a fracture may be detected based on a medical image of the subject. The fracture location may be information on the fracture location of the subject that is estimated based on a medical image of the subject. Alternatively, the information on the presence or absence of a fracture may be information on a location of the subject that is predicted to have a high probability of fracture in the future.

[0021] The information on the results of a blood test may be, for example, information on the results of at least one of a biochemical test, a glucose metabolism test, and an endocrine system test.

[0022] The biochemical test may include, for example, information regarding the test results of at least one of the blood test group B below.

[0023] (Blood test group B): Total protein (TP), albumin (ALB), aspartate aminotransferase (AST), alanine aminotransferase (ALT), lactate dehydrogenase (LDH), bone-type ALP, creatinine (Cre), calcium (Ca), inorganic phosphorus (IP), total cholesterol (T-Cho), high-density lipoprotein-cholesterol (HDL-C), triglycerides (TG), low-density lipoprotein-cholesterol (calculated LDL-C), C-reactive protein (CRP).

[0024] The glucose metabolism test may include, for example, information on the test results of at least one of the glucose metabolism test group C described below.

[0025] (Glucose metabolism system test group C): Glucose (Glu), hemoglobin A1c (HbA1c).

[0026] The endocrine system test may include, for example, information on the test results of at least one of the endocrine system test group D below.

[0027] (Endocrine system test group D): intact parathyroid hormone (intact PTH), 25 hydroxyvitamin D (25(OH)D), 1.25 hydroxydivitamin D (1.25(OH)2D), thyroid-stimulating hormone (YSH), CS, estradiol (E2), and free testosterone (free-TST).

[0028] The information regarding the blood test results may include, for example, information regarding the test results of at least one of the following urine test group E.

[0029] (Urine test group E): Calcium (Ca), phosphorus (P), creatinine (Cre).

[0030] The information on the medication being taken may include, for example, the name of the medication, the amount being taken, the period of taking the medication, etc. The information on the medication being taken may include information on the steroid drug being used.

[0031] The bone density information may be, for example, a measurement value obtained by measuring the bone density of a bone at a predetermined site of a subject. The bone density may be measured by a dual-energy X-ray absorptiometry (DXA) method, an ultrasound method, a micro densitometry (MD) method, or the like. The bone density may also be estimated from a plain X-ray image, a magnetic resonance imaging (MRI) image, a computed tomography (CT) image, an ultrasound image, or a PET image showing the bones of the subject. The DXA method is a method for measuring bone density using X-rays of two different energies. The DXA method is generally performed on the lumbar vertebrae, the proximal femur, and the distal radius. In a DXA device that measures bone density using the DXA method, when the bone density of the lumbar vertebrae is measured, X-rays are irradiated from the front of the lumbar vertebrae of the subject. In a DXA device, when the bone density of the proximal femur is measured, X-rays are irradiated from the front of the proximal femur of the subject. The ultrasound method is a method of measuring bone density by applying ultrasound to bones such as the lumbar vertebrae, femur, heel, and shin. Here, "front of the lumbar vertebrae" and "front of the proximal femur" refer to the direction of correctly facing the imaging site such as the lumbar vertebrae and the proximal femur, and may be the ventral side of the subject's body or the back side of the subject. The proximal femur includes at least one of the neck, trochanter, shaft, and the entire proximal femur (neck, trochanter, and shaft). The MD method is a method of measuring bone density by simultaneously imaging the bones of the hand and an aluminum plate using X-rays and comparing the shading of the images. When measuring bone density by the MD method, X-rays are irradiated from the back side of both hands. The location of the bone measured by the MD method is, for example, the second metacarpal bone of the hand. By using bone density information indicating the bone density measured by the DXA method, which has high measurement accuracy, the accuracy of the drug efficacy prediction information output from the prediction device 1 can be improved. The bone density information may be information on bone quality, which will be described later, instead of bone density, or may include information on bone quality in addition to bone density.

[0032] The bone density information may be the bone density (estimated value) of the bone of the subject estimated by the prediction device 1 from a medical image showing the bone of the subject. Alternatively, the bone density information may be the bone density (predicted value) of the future bone of the subject predicted by the prediction device 1 from a medical image showing the bone of the subject. In this way, the prediction device 1 may be configured to estimate and / or predict the bone density of the bone of the subject from a medical image showing the bone of the subject. In this case, the prediction device 1 may be equipped with a bone density estimation / future prediction model that has been trained using a medical image showing the bone of the subject as an explanatory variable and a measurement value of the bone density of the bone of the subject whose medical image was taken as an objective variable. The prediction device 1 may be equipped with a bone density future prediction model that has been trained using a medical image showing the bone of the subject as an explanatory variable and a measurement value of the bone density of the bone of the subject whose bone has been taken for a predetermined time since the time when the medical image was taken as an objective variable.

[0033] The prediction device 1 may calculate at least one of the estimated value and the predicted value of the bone density of the bone from the whole bone or a part of the bone in the medical image. When the prediction device 1 calculates the estimated value and the predicted value of the bone density of the bone from a part of the medical image, the prediction device 1 may calculate at least one of the estimated value and the predicted value of the bone density of the bone from a plurality of parts. When the prediction device 1 partially calculates the estimated value and the predicted value of the bone density of the bone, the prediction device 1 partially calculates the area mainly including cancellous bone, or partially calculates the area mainly including cortical bone. By being able to partially calculate at least one of the estimated value and the predicted value of the bone density from the medical image in this way, it becomes easier to predict fractures of the vertebrae or the proximal femur. Not limited to bone density, the bone quality described later can also be calculated from the whole bone or a part of the bone in the medical image. As a method of focusing on the bone density or the bone quality from a part of the medical image, for example, a learned segmentation method or the like can be used.

[0034] The bone density may be, for example, a value used in the osteoporosis guidelines (such as, but not limited to, the 2015 Prevention and Treatment Guidelines of the Japan Osteoporosis Society, the same below). In addition, the bone density may be any value related to the density of the bone, and a unique index may be used. Specifically, the bone density may be, for example, bone mineral density per unit area (g / cm 2 ), bone mineral density per unit volume (g / cm 3 ), YAM(%), T-score, and / or Z-score. YAM is an abbreviation for "Young Adult Mean" and may be referred to as young adult average percent and young average percent. The bone density information may include at least one of a measured value, an estimated value, and a predicted value of bone density, as well as information regarding its location (e.g., at least one of the proximal femur and the vertebrae).

[0035] Here, the prediction device 1 may be configured to use the bone density of the subject's bones measured by a DXA device, MD method, ultrasound method, or the like as bone density information instead of a medical image. In this case, the bone density measured by the subject within one year, for example, by a DXA device may be used as the bone density information. Alternatively, the bone density information may be the future bone density of the subject's bones predicted by the prediction device 1 from at least one of the bone density measured by the DXA device, the fracture risk factor information, and the bone metabolism information.

[0036] The bone metabolism information may include at least one of bone formation marker information on osteoblasts, bone resorption marker information on osteoclasts, and bone matrix-related marker information on bone quality of the subject. The bone metabolism information may be analyzed, for example, from a blood or urine sample of the subject. For example, data analyzed before drug treatment may be used as the bone metabolism information. For example, the bone metabolism information may be data analyzed once, or multiple data obtained by analyzing samples collected multiple times. When there are multiple data, the bone metabolism information may be used as a rate of change, an average value, or a basal value. In addition to the bone metabolism information analyzed before drug treatment, the bone metabolism information may be analyzed after drug treatment (for example, 3 to 6 months after the start of drug administration).

[0037] Here, the bone formation marker information may be information regarding at least one of alkaline phosphatase and type I procollagen-N-propeptide, the bone resorption marker information may be information regarding at least one of deoxypyridinoline, type I collagen cross-linked N-telopeptide and type I collagen cross-linked C-telopeptide, and tartrate-resistant acid phosphatase-5b, and the bone matrix-related marker information may be information regarding at least one of undercarboxylated osteocalcin, pentosidine, and homocysteine.

[0038] The bone quality can be based on at least one of the following: statistical properties of bone, geometric properties of bone, mechanical properties of bone, and chemical properties of bone. The bone quality can be based on at least one of the following: bone metabolism marker, sex, race, menopause, age, cortical bone condition, cancellous bone condition, cancellous bone trabecular condition, disease information, bone evaluation information, drug information, and fracture. More specifically, the bone quality can be based on at least one of the following: bone formation marker, bone resorption marker, bone quality marker (e.g., vitamin K value), cortical bone thickness, trabecular density, trabecular direction, and cancellous bone structure index (trabecular bone score), but is not limited thereto. The cancellous bone structure index is an index obtained from raw data (density grayscale image) of a lumbar DXA image from the front (the subject's anterior-posterior direction).

[0039] The disease information may include at least one of osteoporosis, rheumatism, bone necrosis (e.g., femoral head necrosis, etc.), systemic sclerosis, kidney disease, osteopetrosis, etc. The bone evaluation information may include information evaluated by a fracture risk assessment tool. The drug information may include at least one of trade name, generic name, dosage, administration period, and administration method (e.g., oral, intravenous injection, intramuscular injection, subcutaneous injection, etc.) for drugs including at least one of drugs that suppress bone resorption, drugs that promote bone formation, and other drugs (e.g., calcium preparations, vitamin preparations, female hormone preparations, etc.).

[0040] The bone quality may also include, for example, the type of medullary cavity shape. For example, the Dorr classification can be used for the medullary cavity shape. For example, the medullary cavity shape can be classified as follows using at least one of the thickness of the cortical bone and the shape of the medullary cavity. Type A: The cortical bone is thick and the medullary cavity is narrow and thin. Type B: Between Type A and Type C, the medullary cavity is neither narrow nor wide. Type C: The cortical bone is thin and the medullary cavity is wide.

[0041] The prediction device 1 can use any medical image that can analyze the bone density of the bone of the subject to be predicted, the shape of the bone, and the state of the tissues (e.g., tendons, etc.) surrounding the bone. Therefore, the medical image may include at least one of a plain X-ray image, a computed tomography (CT) image, a magnetic resonance imaging (MRI) image, a PET image, and an ultrasound image. When a plain X-ray image is used as a medical image, the part of the body captured by the plain X-ray image is not particularly limited. The plain X-ray image used as a medical image may include at least one of the head, neck, chest, lower back, hip joint, knee joint, ankle joint, foot, toe, shoulder joint, elbow joint, wrist joint, hand, finger, or jaw joint of the subject. In addition, the plain X-ray image for estimation used to estimate the bone density of the bone of the subject and the fracture site may be a front image in which the target part is viewed from the front, or a side image in which the target part is viewed from the side. When using CT images as medical images, for example, at least one of a three-dimensional image, a cross-sectional image perpendicular to the body axis connecting the head and legs (e.g., horizontal section), and a cross-sectional image parallel to the body axis (e.g., sagittal section or coronal section) can be used.

[0042] When the area to be analyzed is the subject's hip joint (such as the femoral neck and head), the prediction device 1 may use plain X-ray images taken from directions other than the front and side, such as axial images and Lauenstein images, as medical images.

[0043] The subject information may include time information corresponding to each piece of information included in the subject information. The time information may be the time or point in time when each piece of information included in the subject information was acquired, the examination date when the subject was examined, the diagnosis date when the subject was diagnosed, and the date when each piece of information was entered into an electronic medical record or the like.

[0044] The target information may be information including time information as described above, or may be information about the subject at substantially the present time. "Substantially the present time" means that a predetermined period is regarded as the same as the present time. Here, the predetermined period may be arbitrarily set. For example, within the past three months, within the past six months, or within the past year may be set as the predetermined period. More specifically, even if the time when one piece of target information is acquired and the time when the other piece of target information is acquired are different within a predetermined period, they can be regarded as the same time. The target information may be, for example, information about the subject at the time when the bone density information and bone metabolism information of the subject are acquired, or at the time when a medical image of the subject's bones is captured.

[0045] The subject information may be information about a subject not at the present time, but in the past or in the future. For example, the subject information may be information about a subject at the time when the subject's bone density information and bone metabolism information were previously acquired, or at the time when a medical image of the subject's bones was previously captured.

[0046] The drug information may include information on drugs having at least one of an effect on bone formation, an effect on bone resorption, and an effect of balancing bone remodeling. The drug information may include at least information on the name and / or mechanism of action of the drug. By using such drug information, the prediction device 1 can more accurately predict the effect of the drug administered to the subject. The drug information may include information on the chemical structure and physical properties of each drug.

[0047] The name of the drug may be, for example, at least one of the chemical name, generic name, trade name, and common name of the active ingredient. The name of the drug may be a development code of the drug manufacturer, clinical trial institution, and / or medical institution. The name of these drugs is not limited to one, and may hold multiple names, in which case the usability of the data can be improved. The information on the mechanism of action may be, for example, biochemical information on the mechanism by which the active ingredient contained in the drug acts on the target molecule, physiological and pharmacological information on important conditions for the active ingredient to act on the target molecule, and the like. Alternatively, the information on the mechanism of action may include, for example, at least one of the information on the action on bone formation, the action on bone resorption, and the action of balancing bone remodeling.

[0048] The drug information may also include auxiliary information on the dosage form, manufacturer, lot number, serial number, administration method, administration conditions, administration period, administration interval, contraindication information, side effect information, drug price, price, and country of approval for each drug. The administration method may include, for example, oral administration or injection. Furthermore, the injection in the administration method may include information on the injection site, such as intravenous, intramuscular, intrathecal, or subcutaneous. When using drug information including these auxiliary information, the prediction device 1 may output drug efficacy prediction information together with the auxiliary information on each drug, and present the drug efficacy prediction information and the auxiliary information to medical personnel. Alternatively, the prediction device 1 may output drug efficacy prediction information on the effect of the drug when the drug is administered for each auxiliary information. In this way, the prediction device 1 can support medical personnel in determining a treatment plan including drug administration to the subject. Here, the effect of the drug refers to the effect on the subject when the drug is administered to the subject, and includes the degree and / or presence of the effect.

[0049] The efficacy prediction information may be fracture risk information indicating how the risk of bone fracture of the subject changes due to administration of a drug to the subject. The fracture risk information in the efficacy prediction information includes, for example, information indicating that the fracture risk is increased, the same, or decreased. Alternatively, the efficacy prediction information may be a prediction result of the future bone density of the bone of the subject. Alternatively, the efficacy prediction information may be a prediction result of the current bone density of the bone of the subject. The efficacy prediction information may be information regarding whether or not a drug considered for administration to the subject can improve the bone condition of the subject. Alternatively, the efficacy prediction information may be information regarding to what extent a drug considered for administration to the subject can improve the bone condition of the subject. The efficacy prediction information may be information indicating the bone condition of the subject after administration of a drug to the subject for a predetermined period of time, or information indicating changes occurring in the bone of the subject before and after a predetermined period of time. Some drugs that act on the bone condition do not have an immediate effect and their effects appear after administration for a predetermined period of time. According to the above configuration, the prediction device 1 can output efficacy prediction information indicating the effect of a drug even if the drug does not have an immediate effect.

[0050] The efficacy prediction information may be information on a drug that may change the bone condition of a subject when the drug is administered to the subject for a predetermined period of time, or prediction information on a drug that should be considered for administration to the subject. In this case, the efficacy prediction information may include at least one of information on a drug that acts on bone formation, information on a drug that acts on bone resorption, and information on a drug that balances bone remodeling. When the efficacy prediction information includes information on both of the above drugs, each drug included in the efficacy prediction information may be associated with action information indicating whether the drug acts on bone formation or on bone formation.

[0051] Examples of drugs that act on bone formation include, but are not limited to, teriparatide acetate and teriparatide (genetic recombination). Examples of drugs that act on bone resorption include, but are not limited to, calcitonin preparations, bisphosphonate preparations, and anti-RANKL monoclonal antibodies. Examples of drugs that act to balance bone remodeling include active vitamin D3 preparations (e.g., calcitriol, eldecalcitol, or alphacalcidol).

[0052] More specifically, the efficacy prediction information may be information indicating the degree of improvement of a given drug in a subject as a percentage, or information indicating the improvement rate of bone density as a percentage.

[0053] Furthermore, the efficacy prediction information may include a plurality of pieces of prediction information on efficacy according to, for example, an administration period for administering a drug to a subject. Specifically, the efficacy prediction information may include two or more pieces of prediction information on efficacy for each administration period for administering a drug to a subject, for example, 3 months, 6 months, 1 year, and 3 years from now. Alternatively, when the efficacy prediction information includes prediction information for a plurality of drugs, the appropriate administration period may differ for each drug. Alternatively, the efficacy prediction information may be prediction information on efficacy at a time when the classification defined in a guideline or the like changes (for example, a time when the classification changes from the class "osteoporosis" to the class "bone mass reduction").

[0054] The drug efficacy prediction information may include a prediction image showing a prediction result regarding how the bone of the subject will change due to the administered drug. Such drug efficacy prediction information utilizes target information including a medical image showing the bone of the subject. The prediction image may show a prediction result showing an effect on the tissues surrounding the bone in addition to a prediction result showing a change in the bone of the subject. Furthermore, the prediction image may be displayed in multiple predictive images as the administration period progresses, or the predicted image may be changed like a video in one image area. The display of multiple predictive images may be, for example, a parallel display of predicted images corresponding to multiple administration periods.

[0055] The efficacy prediction information may also include side effect prediction information regarding side effects predicted when a drug is administered to a subject. The side effect prediction information may be, for example, information regarding at least one side effect from the following side effects:

[0056] (Side effects) Fatigue, allergic reactions, anaphylactic shock, headache, vomiting, gastrointestinal disorders, constipation, vascular disorders, heartburn, kidney disorders, liver disorders, jaundice, hypercalcemia, hypocalcemia, osteonecrosis of the jaw, osteomyelitis of the jaw, osteonecrosis of the external auditory canal, urinary tract stones, convulsions, palpitations, loss of appetite, and hyperuricemia.

[0057] The side effect prediction information may include prediction information regarding specific side effects that may occur in a subject administered with a drug, or may include information indicating the possibility that one or more side effects will occur in the subject (such as the occurrence probability). The side effect prediction information may include a plurality of pieces of prediction information regarding the occurrence of side effects according to the administration period of the drug administered to the subject. Furthermore, the side effect prediction information may include prediction information predicting how the probability of a side effect occurring in the subject will change due to the intake of a drug other than the drug to be predicted, diet, or daily behavior. For example, the side effect prediction information may include prediction information indicating how the probability of a side effect occurring in the subject will change depending on the amount (e.g., intake period and intake amount) of a drug other than the drug to be predicted that is taken.

[0058] When a drug is administered to a subject, the prediction device 1 outputs drug efficacy prediction information that accurately predicts the effect of the drug. This allows the prediction device 1 to present to the subject and medical personnel in charge of the subject how the subject's bone condition may change due to the administration of the drug before the drug is administered.

[0059] (Configuration of prediction system 100a) First, a configuration of a prediction system 100a according to an embodiment of the present disclosure will be described with reference to Fig. 1. Fig. 1 is a diagram showing a configuration example of the prediction system 100a in a medical facility 8 in which a prediction device 1 is installed. In the following, an example of the prediction device 1 that acquires medical images of bones of a subject as target information will be described.

[0060] The prediction system 100a includes a prediction device 1 and one or more terminal devices 7 communicably connected to the prediction device 1. The prediction device 1 is a computer that outputs drug efficacy prediction information from a medical image showing a subject's bones and transmits the drug efficacy prediction information to the terminal device 7. The prediction device 1 may be a cloud-based device, or an on-premise device provided in a medical facility 8 or a company that provides analysis services.

[0061] The terminal device 7 (receiving unit) functions as a receiving unit that receives efficacy prediction information in the prediction system 100a. That is, the terminal device 7 receives efficacy prediction information from the prediction device 1 and presents the efficacy prediction information. The terminal device 7 is a computer used by a medical professional such as a doctor belonging to the medical facility 8. The terminal device 7 is, for example, a personal computer, a tablet terminal, a smartphone, etc. The terminal device 7 has a communication unit that transmits and receives data to and from other devices, an input unit such as a keyboard and a microphone, a display unit that can display information included in the efficacy prediction information, an output unit such as a speaker, etc. In the medical facility 8 shown in FIG. 1, a LAN (local area network) is provided, and an example is shown in which the prediction device 1 and the terminal device 7 are connected to the LAN, but this is not limited to this. For example, the network in the medical facility 8 may be the Internet, a telephone communication line network, an optical fiber communication network, a cable communication network, a satellite communication network, etc.

[0062] In addition to the prediction device 1 and the terminal device 7, a medical image management device 6 (storage unit) that can function as a storage unit that stores target information in the prediction system 100a may be communicatively connected to a LAN in the medical facility 8. The medical image management device 6 is a computer that functions as a server for managing medical images captured in the medical facility 8. In this case, the prediction device 1 may obtain a medical image showing the bones of the subject from the medical image management device 6.

[0063] An electronic medical record management device 5 (storage unit) that can function as a storage unit that stores target information in the prediction system 100a may be communicatively connected to the LAN in the medical facility 8. The electronic medical record management device 5 is a computer that functions as a server for managing various target information related to the subject. In this case, the prediction device 1 may acquire various target information related to the subject from the electronic medical record management device 5.

[0064] The LAN in the medical facility 8 may be communicatively connected to an external communication network. In the medical facility 8, the prediction device 1 and the terminal device 7 may be directly connected without going through a LAN.

[0065] (Configuration of Prediction Device 1) Next, the configuration of the prediction device 1 applied to the prediction system 100a shown in Fig. 1 will be described with reference to Fig. 3. Fig. 3 is a block diagram showing an example of the configuration of the prediction device 1.

[0066] The prediction device 1 includes a control unit 2 that comprehensively controls each unit of the prediction device 1, and a storage unit 3 that stores various data used by the control unit 2. The control unit 2 includes an acquisition unit 21, a drug information application unit 22, a prediction unit 23, an output control unit 24, and a learning unit 25. The storage unit 3 stores a control program 31, which is a program for performing various controls of the prediction device 1, as well as learning data 32, a drug database 33, and a learned drug efficacy prediction model 34 (drug efficacy prediction model). In one example, the storage unit 3 may store target information.

[0067] <Acquisition part 21> The acquisition unit 21 acquires target information including at least one of risk factor information on the subject's fracture risk factors, bone density information indicating the subject's bone density, bone metabolism information indicating the bone metabolism state of the subject, and medical images showing the bones of the subject. For example, the acquisition unit 21 shown in FIG. 3 may be capable of acquiring risk factor information on the subject's fracture risk factors, bone density information indicating the subject's bone density, and bone metabolism information indicating the bone metabolism state from the electronic medical record management device 5. The acquisition unit 21 shown in FIG. 3 may be capable of acquiring medical images that are images showing the bones of the subject from the medical image management device 6. The target information is input data input to the prediction unit 23.

[0068] <Drug information application unit 22> The drug information application unit 22 acquires drug information on drugs having at least one of information on drugs that have an effect on bone formation, an effect on bone resorption, and information on drugs that balance bone remodeling. For example, as shown in FIG. 3, the drug information application unit 22 may acquire drug information by reading out the drug database 33 in the storage unit 3. Alternatively, the drug information application unit 22 may acquire drug information from an external drug information management device via a network such as the Internet. The drug information can be input data to the prediction unit 23.

[0069] <Prediction Section 23> The prediction unit 23 generates and outputs efficacy prediction information regarding the effect of a drug when the drug is administered to a subject, by inputting subject information into a trained efficacy prediction model 34 that predicts the effect of a drug having an action on bones based on drug information regarding the drug. The prediction unit 23 may output at least one of a prediction result indicating a difference between a future bone density and a current bone density of the subject's bones and a prediction result indicating a change from the current bone density of the subject's bones as efficacy prediction information.

[0070] For example, when a medical image and bone metabolism information of a subject are input as the target information, the prediction unit 23 may be configured to output a list of drugs to be considered for administration to the subject and efficacy information (described later) on the effect of each drug included in the list, even if drug information on a drug to be administered to the subject is not input. That is, input of drug information is not essential for the prediction unit 23 to output efficacy information.

[0071] The prediction unit 23 may output a prediction result regarding the fracture site of the subject. Alternatively, the prediction unit 23 may output a prediction result indicating the possibility that the subject will have a fracture in the future. A specific example of information output from the prediction unit 23 will be described later (see FIG. 7).

[0072] The trained efficacy prediction model 34 may be trained using training data 32 in which subject information on each of a plurality of subjects to which a drug has been administered and drug information on the drug administered to each of the subjects are used as explanatory variables, and the effect of the drug on each of the subjects is used as a response variable. The trained efficacy prediction model 34 is, for example, a model obtained as a result of a training process performed by a training unit 25 described below on an untrained neural network. The subject may be of the same biological species as the target, and the following description will be given taking the case where the subject is a human as an example.

[0073] The prediction unit 23 may output a prediction result when the drug is administered and a prediction result when the drug is not administered. In this case, the trained efficacy prediction model 34 may be trained using the training data 32 in which subject information on each of a plurality of subjects to which the drug is administered and subject information on each of a plurality of subjects to which the drug is not administered are used as explanatory variables, and the effect of the drug on each of the subjects is used as a response variable.

[0074] The prediction unit 23 may output a prediction result when different types of drugs are administered. In this case, the trained efficacy prediction model 34 may be trained using training data 32 in which subject information on each of a plurality of subjects administered with drug A and subject information on each of a plurality of subjects administered with drug B, which is a different type from drug A, are used as explanatory variables, and the effect of the drug on each of the subjects is used as a response variable.

[0075] The prediction unit 23 may output a prediction result when different combinations of drugs are administered. In this case, the trained efficacy prediction model 34 may be trained using the training data 32 in which subject information on each of a plurality of subjects receiving different combinations of drugs is used as an explanatory variable, and the effect of the drug on each of the subjects is used as a response variable.

[0076] In response to target information (e.g., medical images) and efficacy information being input to the input layer 231 (see FIG. 4), the prediction unit 23 performs a calculation based on the learned efficacy prediction model and outputs efficacy prediction information from the output layer 232 (see FIG. 4). As an example, the prediction unit 23 may be configured to extract features from the target information and efficacy information and use them as input data. The following known algorithms may be applied to extract the features. · Convolutional neural network (CNN). Autoencoder. Recurrent neural network (RNN). ·LSTM (Long Short-Term Memory). ·ConvLSTM (Convolutional Long Short-Term Memory).

[0077] The trained efficacy prediction model 34 is a calculation model used when the prediction unit 23 performs calculations based on input data. The trained efficacy prediction model 34 is generated by executing machine learning using the training data 32 described later on an untrained neural network. Specific examples of the training data 32, the configuration of the neural network, and the training process will be described later.

[0078] <Output control unit 24> The output control unit 24 transmits the efficacy prediction information output from the prediction unit 23 to the terminal device 7. The prediction device 1 may be configured to include a display unit (not shown). In that case, the output control unit 24 causes the display unit to display the efficacy prediction information.

[0079] <Study Section 25> The learning unit 25 controls the learning process for an untrained neural network. The learning unit 25 executes the learning process for an untrained neural network to create a neural network that functions as the prediction unit 23 from the untrained neural network. This learning uses learning data 32 (described later). A specific example of the learning performed by the learning unit 25 will be described later.

[0080] (Training data 32) The training data 32 is data including explanatory variables and objective variables used in machine learning for generating a trained efficacy prediction model 34 from an untrained neural network. In the training data 32, subject information on each of a plurality of subjects to which a drug is administered and drug information on the drug administered to each of the subjects are the explanatory variables, and the effect of the drug on each of the subjects is the objective variable.

[0081] Here, the subject information includes at least one of risk factor information on the subject's fracture risk factors, bone density information indicating the subject's bone density, bone metabolism information indicating the subject's bone metabolism state, and a medical image showing the subject's bones. The subject information may include risk factor information on the subject's fracture risk factors. The fracture risk factor information may include at least one of the information group A described below. Here, the bone density information may be a measurement value obtained by measuring the bone density of a specific part of the subject. The bone density may be measured by a dual-energy X-ray absorptiometry (DXA) method, an ultrasound method, a micro densitometry (MD) method, or the like. Alternatively, the bone density may be estimated from a plain X-ray image, a magnetic resonance imaging (MRI) image, a positron emission tomography (PET) image, and a computed tomography (CT) image showing the subject's bones.

[0082] The effect of the drug on each subject may be the degree to which the drug administered to each subject improved the bone condition of each subject, or may be information indicating the bone condition of each subject after the drug has been administered continuously for a predetermined period of time, or the change in the bone condition before and after the drug has been administered continuously for a predetermined period of time.

[0083] (Configuration of prediction unit 23) Hereinafter, the configuration of the prediction unit 23 will be described with reference to Fig. 4. The configuration shown in Fig. 4 is just an example, and the configuration of the prediction unit 23 is not limited to this.

[0084] As shown in FIG. 4, the prediction unit 23 performs calculations on input data input to an input layer 231 based on a trained efficacy prediction model 34, and outputs efficacy prediction information from an output layer 232.

[0085] The prediction unit 23 in FIG. 4 includes a neural network having an input layer 231 and an output layer 232. The neural network may be any neural network suitable for handling time-series information. For example, it may be LSTM. The neural network may be any neural network suitable for handling time-series information and position information in combination. For example, the neural network may be a ConvLSTM network combining CNN and LSTM. The input layer 231 can extract a feature of the time change of the input data. The output layer 232 can calculate a new feature based on the feature extracted by the input layer 231, the time change of the input data, and the initial value. The input layer 231 and the output layer 232 have a plurality of LSTM layers. Each of the input layer 231 and the output layer 232 may have three or more LSTM layers.

[0086] (Learning process by learning unit 25) The learning process for generating the trained efficacy prediction model 34 will be described below with reference to Fig. 5. Fig. 5 is a flowchart showing an example of the flow of the learning process performed by the learning unit 25.

[0087] The learning unit 25 acquires the learning data 32 from the storage unit 3 (step S1). The learning data 32 includes explanatory variables to be input to the input layer 231. Here, the explanatory variables are subject information on each of a plurality of subjects to which a drug has been administered, and drug information on the drug administered to each of the subjects.

[0088] Next, the learning unit 25 inputs subject information about a certain subject A and drug information about a drug administered to the subject A to the input layer 231 (step S2).

[0089] Next, the learning unit 25 acquires output data related to the effect of the drug administered to the subject A from the output layer 232 (step S3). This output data contains the same content as the objective variable of the learning data 32. In FIG. 5, the order of steps S2 and S3 may be reversed. Alternatively, in FIG. 5, steps S2 and S3 may be configured to be executed simultaneously.

[0090] Next, the learning unit 25 acquires a response variable for the subject A included in the learning data 32. Then, the learning unit 25 compares the output data acquired in step S3 with the response variable for the subject A to calculate an error (step S4), and adjusts the efficacy prediction model being trained so as to reduce the error (step S5).

[0091] Any known method can be applied to adjust the efficacy prediction model during learning. For example, the error backpropagation method may be adopted as a method for adjusting the efficacy prediction model. The adjusted efficacy prediction model becomes a new efficacy prediction model, and the prediction unit 23 uses the new efficacy prediction model in subsequent calculations. In the stage of adjusting the efficacy prediction model, parameters (e.g., filter coefficients, weighting coefficients, etc.) used by the prediction unit 23 may be adjusted.

[0092] If the error is not within a predetermined range and explanatory variables for all subjects included in the learning data 32 have not been input (NO in step S6), the learning unit 25 returns to step S2 and repeats the learning process. If the error is within a predetermined range and explanatory variables for all subjects included in the learning data 32 have been input (YES in step S6), the learning unit 25 ends the learning process.

[0093] When the above-mentioned learning process is adopted, the prediction unit 23 can output efficacy prediction information on the effect of a drug to be administered to the subject from the subject information and drug information. Also, when the above-mentioned learning process is adopted, the prediction unit 23 can output efficacy prediction information on one or more drugs that are candidates for the drug to be administered to the subject from the subject information.

[0094] [Another embodiment] The learning data 32 may include information on each combination of drugs administered to a subject as an explanatory variable, and information on the effect of the drugs on each subject as a response variable. When a learning process using such learning data 32 is adopted, the prediction unit 23 can output efficacy prediction information for each combination of drugs administered to a subject.

[0095] (Processing performed by prediction device 1) Hereinafter, the flow of processing performed by the prediction device 1 will be described with reference to Fig. 6. Fig. 6 is a flowchart showing an example of the flow of processing performed by the prediction device 1. Fig. 6 shows an example of processing for outputting efficacy prediction information from target information.

[0096] The target information on the target bone is stored in the medical image management device 6 and the electronic medical record management device 5 (storage step). First, the acquisition unit 21 acquires the target information (step S11: acquisition step) and inputs it to the prediction unit 23.

[0097] The prediction unit 23 inputs the subject information into the learned efficacy prediction model 34, thereby generating and outputting efficacy prediction information regarding the effect of the drug when the drug is administered to the subject (step S12: generation step and prediction step). The efficacy prediction information output from the prediction unit 23 may be transmitted to the terminal device 7, in which case the terminal device 7 receives the efficacy prediction information (reception step).

[0098] In step 12, drug information regarding drugs that have at least one of an effect on bone formation and an effect on bone resorption may be input into the trained efficacy prediction model 34 together with the subject information.

[0099] (Medical services realized by prediction device 1) FIG. 7 is a diagram showing an example of a medical service realized by the prediction device 1. As shown in FIG.

[0100] When a subject is examined at a medical facility 8 based on the results of a regular health check or subjective symptoms such as a suspected fracture, subject information on the subject may be collected by a user (e.g., a medical professional) during the examination. The subject information includes at least any of bone density information indicating the bone density of an acquired bone, bone metabolism information indicating the bone metabolism state of the subject's bone, and a medical image showing the bone of the subject.

[0101] When a user inputs the acquired subject information to the prediction device 1, the prediction device 1 outputs efficacy prediction information regarding a drug (and / or a candidate thereof) to be administered to a subject. Alternatively, a configuration may be adopted in which a user inputs the acquired subject information and drug information regarding a drug (and / or a candidate thereof) to be administered to a subject to the prediction device 1.

[0102] The prediction device 1 may output future bone density when the subject does not receive drug administration and future bone density when the subject receives drug administration as drug efficacy prediction information. Here, the output bone density is bone mineral density per unit area (g / cm 2 ), bone mineral density per unit volume (g / cm 3), YAM(%), T-score, and / or Z-score. YAM is an abbreviation for “Young Adult Mean” and may be referred to as young adult average percent or young average percent. For example, the output layer 230 may display bone mineral density per unit area (g / cm 2 ) and bone density expressed by YAM may be output, or bone density expressed by YAM, bone density expressed by T score, and bone density expressed by Z score may be output. In addition, the bone density output may be categorized as follows: T score of -1.0 or more is displayed as normal bone density, T score of -1.0 to -2.5 is displayed as low bone density or osteopenia, and T score of -2.5 or less is displayed as osteoporosis. Alternatively, it may be the YAM ratio. The YAM ratio is a value indicating the diagnostic criterion, and may be categorized as "normal" in the range of 100% to 80% of the YAM ratio, "bone loss" in the range of 80% to 70% of the YAM ratio, and "osteoporosis" in the range of less than 70% of the YAM ratio.

[0103] FIG. 7 shows efficacy prediction information such as "If left as is, bone density will be 〇% after X years", "If drug A is used, bone density will be 〇% after X years", and "If drug B is used, bone density will be 〇% after X years". Based on the efficacy prediction information, a user who has obtained such efficacy prediction information can appropriately consider and determine whether or not to administer a drug to a subject, and which of drug A and drug B is more preferable to administer. The prediction device 1 may present the probability of future fracture as the efficacy prediction information in FIG. 7 instead of bone density, or may present the probability of future fracture in addition to bone density.

[0104] FIG. 7 shows only an example. For example, the drug prediction information may include drug efficacy prediction information such as "When drug A and drug B are used in combination, bone density will be 〇% after X years," or "When drug A, drug B, drug C, and... are used in combination, bone density will be 〇% after X years." The drug prediction information may also include drug efficacy prediction information such as "When drug A is administered in P mg / day, bone density will be 〇% after X years," or "When drug A is administered in Q mg / day, bone density will be 〇% after X years." Since some drugs have a limited length of administration, the drug prediction information may include drug efficacy prediction information such as "When drug A is administered for the first K years and drug B thereafter, bone density will be 〇% after X years." The drug prediction information may also include information on fracture risk, such as "The probability of developing a fracture in M ​​years is 〇%," instead of bone density.

[0105] The prediction device 1 may be capable of outputting efficacy prediction information shown in (1) to (6) in Fig. 7. In Fig. 7, "target drug" refers to one or more drugs to be administered to a subject and / or candidate drugs.

[0106] (1) The degree of effectiveness of the drug when administered to the subject.

[0107] (2) Information regarding whether the subject is likely to be effectively treated with the drug.

[0108] (3) The species on which the target drug may be effective.

[0109] (4) The extent of the drug effect when the target drugs are combined.

[0110] (5) The cause of the subject's osteoporosis.

[0111] (6) Relationship with concomitant medications. For example, the efficacy prediction information shown in (5) above may be information indicating whether the cause is bone formation alone, bone resorption alone, or both bone formation and bone resorption.

[0112] When various additional information regarding a subject is input, the prediction device 1 may output the type of medicine suitable for the subject.

[0113] For subjects with extremely low bone mass, subjects who have already experienced a fracture, and subjects at high risk of fracture, the prediction device 1 may output the type of drug that is highly effective in increasing bone density. Examples of drugs that are highly effective in increasing bone density include parathyroid hormone drugs. Here, the information on bone mass may be information measured by a bone density measuring device such as DXA, or information obtained by estimating bone density from an X-ray image using learned parameters. For the fracture risk, information on fracture risk factors by FRAX (registered trademark) may be used, or information predicted from an X-ray image using learned parameters may be used, or a combination of these may be used. Bone mass is the sum of bone salts and bone matrix proteins. In the present disclosure, bone mass is an index related to bone density, and the bone mass is the amount of bone tissue in the skeleton.

[0114] The prediction device 1 may output efficacy prediction information according to the type of drug effective in suppressing fracture. Such efficacy prediction is useful, for example, for subjects with a high risk of fracture. Examples of drugs effective in suppressing fracture include bisphosphonates and denosumab. The prediction device 1 may extract a type of drug effective for fractures at a specific site for a subject with a high risk of fracture at a specific site, and output efficacy prediction information according to the drug. More specifically, for example, when the specific site is the proximal femur, the prediction device 1 may extract a type of drug effective for fracture healing at the proximal femur, and output efficacy prediction information according to the drug. The specific site may include at least one of the vertebral body and the proximal femur, but is not limited thereto. The fracture prediction at the specific site may be based on partially estimating or predicting at least one of the bone density and bone quality in an area mainly including at least one of cancellous bone and cortical bone.

[0115] More specifically, when predicting at least one of bone mineral density and fracture of vertebrae as efficacy prediction information, the prediction device 1 may partially focus on a region mainly including cancellous bone of vertebrae. Also, when predicting at least one of bone mineral density and fracture of proximal femur as efficacy prediction information, the prediction device 1 may partially focus on a region mainly including cortical bone of femur.

[0116] For a subject who is considered to have increased bone resorption based on the measurement of the bone metabolism marker, the prediction device 1 may output the type of drug that corresponds to a bone resorption inhibitor. Examples of drugs that correspond to a bone resorption inhibitor include calcitonin drugs, bisphosphonate preparations, anti-RANKL antibodies, selective estrogen receptor modulators (SERM (Selective Estrogen Receptor Modulator) preparations), and female hormone preparations.

[0117] For a subject who is thought to have decreased bone formation, the prediction device 1 may output the type of drug that corresponds to an osteogenesis promoting drug. Examples of drugs that correspond to an osteogenesis promoting drug include active vitamin D3, parathyroid hormone drugs, and vitamin K2 preparations.

[0118] The information on the drug contained in the efficacy prediction information output by the prediction device 1 may be selected based on information on the administration period, swallowing, and cost. Here, the administration period may include at least one of information on the subject's age and menopause. For subjects who need to take the drug for a long period of time, the prediction device 1 may output a type of drug that is suitable for long-term administration. Examples of drugs that are suitable for long-term administration include SERM preparations.

[0119] For subjects who are thought to have no problem taking medicines orally, the prediction device 1 may output types of medicines that correspond to oral agents. On the other hand, for subjects who are thought to have difficulty taking medicines orally, the prediction device 1 may output types of medicines that correspond to intravenous formulations and subcutaneous injection formulations. For subjects at risk of fracture and with progressing symptoms of osteoporosis, the prediction device 1 may output types of medicines that correspond to high-cost medicines. Examples of high-cost medicines include parathyroid hormone medicines.

[0120] The prediction device 1 may output a type of drug suitable for a subject based on information on a disease, symptoms, etc., for which a drug is contraindicated or inappropriate for the subject. For example, for a subject suffering from renal failure, the prediction device 1 may be configured not to output information on drugs such as bisphosphonates and SERMs. For example, for a subject with reduced renal function, the prediction device 1 may be configured not to output information on active vitamin D3 preparations. Without being limited thereto, the prediction device 1 may present efficacy prediction information even if a drug is contraindicated or inappropriate for the subject. In this case, it may be clearly indicated that the drug is contraindicated or inappropriate, and this may be displayed in a manner that is easy for the user to understand. The prediction device 1 may also display the reason why the drug is contraindicated or inappropriate for the subject so that the user can understand.

[0121] The prediction device 1 may be configured to output a virtual image corresponding to the future bone density predicted for the bones of the subject, and the virtual image. For example, when a plain chest X-ray image of the subject is used as the target information, the image of the bones shown in the plain chest X-ray image may be an image that reflects the appearance corresponding to the future bone density. The prediction device 1 may be capable of outputting an image and / or a video showing the difference between the plain chest X-ray image of the target information and the virtual image (i.e., a change in bone density). For example, by having the subject visually recognize such an image and / or video, the risk of fractures occurring in the subject's bones and the risk of developing osteoporosis can be clearly shown to the subject. Therefore, even for subjects who have no subjective symptoms, the need for measures against the risk of fractures occurring and the risk of developing osteoporosis can be effectively conveyed, leading to early intervention.

[0122] By outputting at least any one of the above information (1) to (6) as efficacy prediction information, the prediction device 1 can inform the user whether or not a drug is expected to be effective if administered to a subject. Having obtained such efficacy prediction information, the user can appropriately consider and select a drug to be administered to a subject.

[0123] [Embodiment 2] (Configuration of prediction system 100b) The prediction device 1 may not be a computer installed in a specific medical facility 8, but may be communicatively connected to a LAN installed in each of a plurality of medical facilities 8 via a communication network 9. Fig. 2 is a diagram showing a configuration example of a prediction system 100b according to another embodiment of the present disclosure.

[0124] The LAN in the medical facility 8a may be communicably connected to one or more terminal devices 7a (receiving units), as well as an electronic medical record management device 5a (storage unit), and a medical image management device 6a (storage unit). The LAN in the medical facility 8b may be communicably connected to one or more terminal devices 7b (receiving units), as well as an electronic medical record management device 5b (storage unit), and a medical image management device 6b (storage unit). In the following, when there is no particular distinction between the medical facilities 8a and 8b, they will be referred to as the "medical facility 8." Furthermore, when there is no particular distinction between the terminal devices 7a and 7b, and the medical image management devices 6a and 6b, they will be referred to as the "terminal devices 7" and the "medical image management devices 6," respectively.

[0125] 2 shows an example in which LANs of a medical facility 8a and a medical facility 8b are connected to a communication network 9. The prediction device 1 is only required to be communicably connected to devices in each medical facility via the communication network 9, and is not limited to the configuration shown in FIG 2. For example, the prediction device 1 may be installed in the medical facility 8a or the medical facility 8b.

[0126] In the prediction system 100b having such a configuration, the prediction device 1 can acquire medical images of a subject Pa who has been examined at a medical facility 8a from a medical image management device 6a in the medical facility 8a. The prediction device 1 then transmits efficacy prediction information on the effects of a drug that is being considered for administration to the subject Pa to a terminal device 7a installed in the medical facility 8a. Similarly, the prediction device 1 can acquire medical images of a subject Pb who has been examined at a medical facility 8b, and transmit efficacy prediction information on the effects of a drug that is being considered for administration to the subject P to a terminal device 7b installed in the medical facility 8b.

[0127] In this case, the medical image of each subject may include identification information (e.g., facility ID) unique to each medical facility 8 that examines each subject, and identification information (e.g., subject ID, patient ID) unique to each subject that is assigned to each subject. Based on this identification information, the prediction device 1 can correctly transmit drug efficacy prediction information output from the medical image of the subject to the terminal device 7 of each medical facility 8 where the subject was examined.

[0128] [Embodiment 3] (Outline of prediction device 1a) Other embodiments of the present disclosure will be described below. For convenience of explanation, the same reference numerals will be given to components having the same functions as those described in the above embodiment, and the description thereof will not be repeated.

[0129] Fig. 8 is a block diagram showing the configuration of a prediction device 1a according to this embodiment. The prediction device 1a differs from the prediction device 1 shown in Fig. 3 in that it further includes an image analysis function and a function for generating processed image data 351 in which a medical image is modified so as to include region of interest information 352 relating to a region of interest focused on during the image analysis process.

[0130] The prediction device 1a includes a control unit 2a that controls each unit of the prediction device 1a in an integrated manner, and a storage unit 3a that stores various data used by the control unit 2a. The control unit 2a includes an acquisition unit 21, a drug information application unit 22, a prediction unit 23, an output control unit 24a, a learning unit 25, and an image analysis unit 26. The storage unit 3a stores a control program 31, which is a program for performing various controls of the prediction device 1a, as well as learning data 32, a drug database 33, a learned drug efficacy prediction model 34 (drug efficacy prediction model), and analysis result data 35. The analysis result data 35 may store attention region information 352 output from the image analysis unit 26. The analysis result data 35 also stores processed image data 351 output from the image analysis unit 26. The processed image data 351 is image data obtained by adding the analysis result and attention region information to the medical image to be analyzed by the image analysis unit 26.

[0131] The processed image data 351 may be image data in which information indicating the region that is the main basis is added to the input image data. The information indicating the region is, for example, information that indicates the range of the region, such as coloring or framing. In image analysis, since a certain region of an image is often the main basis for estimation, the user can confirm the region that is the basis for estimation by using information indicating such a region. In this case, the display image displayed on the terminal device 7 used by the user includes a processed image in which the attention region information is added to the medical image input to the first analysis model 261. The output control unit 24a may generate a display image in which the medical image and the processed image are displayed side by side. Such a display image makes it easy for the user to compare the analyzed original image and the processed image. Alternatively, the terminal device 7 may be capable of switching between the medical image and the processed image and displaying them by a user operation. For example, the terminal device 7 may alternately display the medical image and the processed image by the user clicking a switching button displayed on the display screen.

[0132] <Image Analysis Unit 26> The image analysis unit 26 includes a first analysis model 261. The first analysis model 261 is, for example, a trained machine model that analyzes medical images, and outputs an estimation result related to a change in the state of bones from the input medical image. In addition to the estimation result, the first analysis model 261 may output area of ​​interest information 352 related to an area of ​​interest that is an area in the medical image and that is focused on in the process of outputting the estimation result. The image analysis unit 26 may store the area of ​​interest information 352 in the storage unit 3a. Here, the area in the medical image may be, for example, an area inside the outer periphery of the medical image. The area of ​​interest information 352 may be located in the entire area in the medical image, or may overlap only a part of the area. Alternatively, if the area of ​​interest is not in the medical image (or, for example, the entire medical image), it may not overlap the area in the medical image. In this case, for example, a message indicating that there is no region of interest (or, for example, that the entire medical image is the region of interest) may be displayed outside the medical image.

[0133] Also, the attention area information 352 may be a heat map showing the area on which the estimation result is based. In this case, for example, the outer edge of the heat map indicates the segmentation area. The heat map is a method of expressing the magnitude of bone density with the density of any color. For example, the attention area information 352 may be a heat map showing the degree of attention, or a heat map showing the bone density value. Also, the attention area information 352 may be a heat map showing the possibility (probability) of a fracture. In this case, the processed image may be an image in which a heat map showing the attention area information is superimposed on a medical image. The image used for the heat map may be a still image or a video. By showing it as a video, for example, by fading various heat maps in order, it becomes easy to visually recognize the relationship between each heat map. Also, in the case of a bone density heat map including areas other than the segmentation area, a part of the segmentation area may be surrounded by a frame.

[0134] The medical image input to the first analysis model 261 is an image including an image of the subject's bone, and outputs an estimation result regarding the state of the bone. For example, the first analysis model 261 may be a trained model trained to output an estimation result or a calculation result regarding the state of the bone, such as bone density, a relative comparison of bone density, or a possibility of fracture, from an X-ray image of the bone. The bone density may be a calculated bone density of a bone part included in the medical image, or may be an average bone density of the whole body estimated from the image data. Alternatively, the bone density may be the bone density of the lowest part. A known method can be used to calculate the bone density from the medical image. The relative comparison of the bone density is the ratio of the estimated bone density to the YAM. The possibility of fracture is the possibility of a bone at a specific part (for example, the femoral neck, etc.) being fractured. The estimation result regarding the state of the bone may be an estimation result at the time when the medical image is taken, or may be a prediction at a time when a predetermined period has passed since that time.

[0135] The first analytical model 261 may output at least one of the following pieces of information as an estimation result. -Bone density estimated at the time a medical image is taken. -Bone density predicted at a specified time after a medical image is taken. -The estimated location and likelihood of a fracture at the time the medical image is taken. - Bone density at a given time prior to when the medical image was taken. -Prediction of fracture location and likelihood at a specified time after the medical image is taken.

[0136] If the bone density of the subject's bone has not been measured, the acquiring unit 21 cannot acquire the bone density information of the subject from the electronic medical record management device 6. In such a case, the acquiring unit 21 may acquire, from the image analyzing unit 26, an estimation result including the bone density of the bone estimated by the first analytical model 261 from a medical image including an image of the subject's bone. In this case, the predicting unit 23 may input the estimation result regarding the bone density acquired from the first analytical model 261 to the trained efficacy prediction model 34 as subject information.

[0137] Next, a learning method of the first analysis model 261 will be described. Learning of the first analysis model 261, for example, learning to estimate bone density, can be performed using learning data including an X-ray image of a bone with a specified bone density as an explanatory variable and the bone density of the subject shown in the X-ray image as an objective variable. Learning to estimate the possibility of fracture can be performed using learning data including an X-ray image of the subject's bone as an explanatory variable and data on whether the subject has had a fracture within a predetermined period thereafter as an objective variable. Relative comparison of bone density does not need to be learned, and is obtained by dividing the estimated bone density by YAM.

[0138] Furthermore, future prediction learning may be performed using, as training data, an X-ray image of a bone with a specified bone density and data on the degree of bone density of the subject after a predetermined period of time thereafter, or data on whether or not a fracture has occurred. For example, the data on whether or not a fracture has occurred may be information linking information on a fracture to an X-ray image showing a fractured bone, and information on a non-fracture to an X-ray image showing a non-fracture. By including information on the subject from which the training data is obtained, such as age, sex, and weight, and information on lifestyle habits, such as the amount of exercise, diet, smoking, and drinking, in the training data, a first analysis model 261 capable of more accurate estimation or prediction can be constructed.

[0139] The first analytical model 261 may be a cytopathological analysis model. In this case, the input medical image may be a microscopic image of a subject's cells, and the estimation result may be information on a pathological mutation of the cells.

[0140] In medical images, not only image changes due to changes in bone density and the like estimated by the first analytical model, but also changes due to other reasons may appear. Other reasons include some kind of disease, treatment scars, various types of information intentionally inserted, objects worn by the subject, etc. Examples of some kind of disease include calcification of arteries, hardening of bones, fractures, and tumors in organs. Examples of treatment scars include implants and bone cement. Examples of intentionally inserted information include letters such as "L" and "R" that indicate directions. Examples of objects worn by the subject include necklaces.

[0141] Due to such causes, pixel values ​​(e.g., brightness) of medical images often change compared to cases without other causes. In addition, implants, necklaces, and the like have peculiar shapes, which may hide information contained in the medical images used by the first analysis model 261 when estimating bone density. Changes due to such reasons may affect the evaluation of bone density based on the estimation results of the image analysis unit 26. Usually, a doctor who sees a medical image notices such changes and interprets the evaluation results of bone density after taking into account the information that caused such changes. However, in some cases, a doctor may not notice such changes in the medical image. Or, a doctor may not be able to judge to what extent the changes in the medical image due to other reasons are taken into account in the estimation results output from the image analysis unit 26. In such cases, a doctor may be confused about how to interpret the output estimation results.

[0142] Therefore, the image analysis unit 26 may further include a second analysis model 262. The second analysis model 262 is a trained machine model that detects changes in the medical image due to other reasons, such as changes in the medical image due to some disease, treatment scars, various information intentionally inserted, things worn by the subject, etc. The image analysis unit 26 may correct the medical image according to the detection result. Here, the correction may be at least one of image position adjustment (left / right, up / down directions or rotation with respect to the original position), brightness adjustment, and mask processing. Alternatively, the image analysis unit 26 may analyze the medical image that has been subjected to these corrections using the first analysis model 261.

[0143] The second analysis model 262 can be trained using images on which a doctor has previously annotated the positions of images due to some disease, treatment scars, various pieces of information intentionally inserted, items worn by the subject, etc. The second analysis model 262 can be constructed by a known learning method for constructing an image analysis model.

[0144] The image analysis unit 26 may generate the processed image data 351 using the medical image input to the first analysis model 261, the analysis result output from the first analysis model 261, and the region of interest information. Alternatively, the image analysis unit 26 may correct the image using the result output from the second analysis model 262, and input the image to the first analysis model 261 to create the processed image data. Alternatively, the image analysis unit 26 may generate the processed image data 351 by adding information on the basis for deriving a change in the image due to other reasons (derivation basis data) to the image data generated using the medical image input to the first analysis model 261, the analysis result output from the first analysis model 261, and the region of interest information.

[0145] For example, when the second analysis model 262 detects a change in the medical image due to another reason, the image analysis unit 26 may generate processed image data 351 by excluding a region corresponding to the detected change from the original medical image. Then, the image analysis unit 26 may input the generated processed image data 351 to the first analysis model 261 to output an estimation result (including bone density) regarding the state of the bones shown in the processed image data 251. According to this configuration, the image analysis unit 26 can output an estimation result from the medical image including a region corresponding to the change detected by the second analysis model 262, and an estimation result from the medical image excluding the region corresponding to the change detected by the second analysis model 262. By presenting both of these estimation results to the user, the user can evaluate the influence of the region corresponding to the change detected by the second analysis model 262 on the estimation result.

[0146] The output control unit 24a transmits the efficacy prediction information output from the prediction unit 23 and the processed image data 351 generated by the image analysis unit 26 to the terminal device 7. The output control unit 24a may generate a web page using the efficacy prediction information output from the prediction unit 23 and the processed image data 351 generated by the image analysis unit 26. In this case, the output control unit 24a may transmit access information for accessing the generated web page to the terminal device 7.

[0147] FIG. 9 is an example of a processed image including information (derivation basis data) on the basis of the derivation basis of the image change due to other reasons, generated by the image analysis unit 26 based on the analysis results of the first analysis model 261 and the second analysis model 262. The area indicated by the frame 1502 in FIG. 9 is an area in which the implant 1502A for fixing the spine is shown. The area indicated by the frame 1503 is an area in which the calcified abdominal aorta 1503A is shown. In the figure, each area is shown by a dotted line, but may be shown by different colors, such as a red frame and a yellow frame. The area indicated by the frame 1501 indicates an area that is the basis for derivation of the analysis result analyzed by the first analysis model 261. In this way, the range in which the image change due to other reasons is detected may be outside the area (frame 1501) that is the basis for derivation when the analysis of the state of the object is performed.

[0148] 10 is an example of a browser image 1600 that displays analysis results both in the case where image changes due to other reasons are included and in the case where they are excluded. In this browser image 1600, for example, an analysis image 1602 and an analysis result 1601 including efficacy prediction information are displayed. The analysis image 1602 may be the same as the image shown in FIG.

[0149] The analysis result 1601 says, "When the dotted area is included, your bone density is XX g / cm 2 "If you exclude the dotted area, your bone density is YYg / cm 2 In addition, the analysis result 1601 shown in FIG. 10 displays drug efficacy prediction information such as "Probability that drug XX will be effective XX %."

[0150] It may be possible to allow a medical professional such as a doctor to select one or both of areas 1604 and 1605 indicated by dotted lines in analysis image 1602. For example, when an operation on selection button 1606 is accepted after area 1604 is selected, a detailed analysis result for area 1604 may be displayed in analysis result 1601. Back button 1607 and end button 1608 are as before.

[0151] [Embodiment 4] Other embodiments of the present disclosure will be described below. For convenience of explanation, the same reference numerals will be given to components having the same functions as those described in the above embodiment, and the description thereof will not be repeated.

[0152] Each function of the prediction devices 1 and 1a, including the efficacy prediction model that generates efficacy prediction information based on target information and drug information on drugs having an action on bones, may be realized by cloud computing. The prediction devices 1 and 1a may be, for example, a cloud server (not shown). In this case, in the prediction systems 100 and 100a, the cloud server may receive target information from the terminal devices 7, 7a, and 7b and generate efficacy prediction information based on the target information and drug information on drugs having an action on bones. The cloud server may then transmit the generated efficacy prediction information to the terminal device 7, which is the source of the target information. In this embodiment, the terminal devices 7, 7a, and 7b function as a storage unit that stores target information to be transmitted to the cloud server, and also function as a receiving unit that receives efficacy prediction information from the cloud server.

[0153] [Software implementation example] The control blocks of the prediction device 1 (particularly the acquisition unit 21, the drug information application unit 22, the prediction unit 23, the output control unit 24, and the learning unit 25) may be realized by a logic circuit (hardware) formed in an integrated circuit (IC chip) or the like, or may be realized by software.

[0154] In the latter case, the prediction device 1 includes a computer that executes instructions of a program, which is software that realizes each function. The computer includes, for example, one or more processors, and a computer-readable recording medium that stores the program. The processor in the computer reads the program from the recording medium and executes it, thereby achieving the object of the present disclosure. The processor may be, for example, a CPU (Central Processing Unit). The recording medium may be a "non-transient tangible medium," such as a ROM (Read Only Memory), as well as a tape, a disk, a card, a semiconductor memory, a programmable logic circuit, or the like. The device may further include a RAM (Random Access Memory) that expands the program. The program may be supplied to the computer via any transmission medium (such as a communication network or a broadcast wave) that can transmit the program. One aspect of the present disclosure may also be realized in the form of a data signal embedded in a carrier wave, in which the program is embodied by electronic transmission.

[0155] The invention according to the present disclosure has been described above based on the drawings and examples. However, the invention according to the present disclosure is not limited to the above-mentioned embodiments. In other words, the invention according to the present disclosure can be modified in various ways within the scope of the present disclosure, and the embodiments obtained by appropriately combining the technical means disclosed in the different embodiments are also included in the technical scope of the invention according to the present disclosure. In other words, it should be noted that a person skilled in the art can easily make various modifications or corrections based on the present disclosure. It should also be noted that these modifications or corrections are included in the scope of the present disclosure.

[0156] [Summary 1] A prediction device according to aspect 1A of the present disclosure includes an acquisition unit that acquires subject information including at least one of fracture risk factor information related to a subject's fracture risk factors, bone density information indicating the bone density of the subject's bone, bone metabolism information indicating the bone metabolism state of the subject's bone, and a medical image showing the subject's bone, and a prediction unit that outputs efficacy prediction information related to the effect of a drug when administered to the subject by inputting the subject information into a drug efficacy prediction model that predicts the effect of a drug based on drug information related to the drug that has an effect on bone.

[0157] The prediction device according to aspect 2A of the present disclosure, in accordance with aspect 1A above, may be arranged such that the drug information includes information on a drug having at least one of an effect on bone formation and an effect on bone resorption.

[0158] A prediction device according to aspect 3A of the present disclosure, in aspect 1A or 2A above, may be configured such that the efficacy prediction model is trained using training data in which subject information on each of a plurality of subjects to which the drug has been administered and drug information on the drug administered to each of the subjects are used as explanatory variables, and the effect of the drug in each of the subjects is used as a response variable.

[0159] A prediction device according to aspect 4A of the present disclosure, in any one of aspects 1A to 3A above, may be such that the drug information includes at least the name of the drug and information relating to a mechanism of action.

[0160] A prediction device according to aspect 5A of the present disclosure, in any of aspects 1A to 4A above, may be configured such that the drug information includes information regarding each of a combination of multiple drugs that may change the bone condition of the subject, and the prediction unit outputs the efficacy prediction information for each of the combinations of multiple drugs.

[0161] In the prediction device according to aspect 6A of the present disclosure, in any of aspects 1A to 5A above, the prediction unit may output, as the efficacy prediction information, information indicating the state of the subject's bones after administration of the drug to the subject continues for a predetermined period of time, or information indicating changes occurring in the subject's bones before and after the predetermined period of time.

[0162] In the prediction device of aspect 7A of the present disclosure, in any of aspects 1A to 6A above, the prediction unit may output, as the efficacy prediction information, information regarding a drug that may change the bone condition of the subject after administration to the subject for a predetermined period of time.

[0163] The prediction device according to an aspect 8A of the present disclosure is any one of the above aspects 1A to 7A, wherein the prediction unit outputs a prediction result of future bone mineral density of the subject as the efficacy prediction information.

[0164] A prediction device according to an aspect 9A of the present disclosure is any one of the above aspects 1A to 8A, wherein the prediction unit outputs a prediction result of a current bone mineral density of the subject as the efficacy prediction information.

[0165] In the prediction device according to aspect 10A of the present disclosure, in any one of aspects 1A to 9A above, the prediction unit may output, as the efficacy prediction information, at least one of a prediction result indicating a difference between a future bone density and a current bone density of the subject's bone, and a prediction estimation result indicating a change from the current in the bone density of the subject's bone.

[0166] The prediction device according to an aspect 11A of the present disclosure is any one of the above aspects 1A to 10A, wherein the prediction unit outputs a prediction result regarding a fracture site of the subject as the efficacy prediction information.

[0167] The prediction device according to aspect 12A of the present disclosure is any one of aspects 1A to 11A above, wherein the prediction unit outputs a prediction result indicating a possibility that the subject will have a bone fracture in the future as the efficacy prediction information.

[0168] In a prediction device according to aspect 13A of the present disclosure, in any of aspects 1A to 12A above, the subject information may include at least one of the following information regarding the subject's type, age, sex, weight, height, presence or absence of fracture, fracture location, fracture history, family history of fracture, glucocorticoids, rheumatoid arthritis, secondary osteoporosis, underlying disease, smoking history, drinking habits, occupational history, exercise history, medical history, blood test results, urine test results, medications being taken, and gene sequence.

[0169] The prediction device according to Aspect 14A of the present disclosure is in any one of Aspects 1A to 13A above, wherein the bone density information is a measurement value obtained by measuring the bone density of a predetermined part of the subject.

[0170] In a prediction device according to aspect 15A of the present disclosure, in any one of aspects 1A to 14A above, the bone metabolism information may include at least any one of bone formation marker information regarding osteoblasts, bone resorption marker information regarding osteoclasts, and bone matrix-related marker information regarding bone quality of the subject.

[0171] In the prediction device of aspect 16A of the present disclosure, in the above aspect 15A, the bone formation marker information may be information regarding at least any one of alkaline phosphatase and type I procollagen N-propeptide, the bone resorption marker information may be information regarding at least any one of deoxypyridinoline, type I collagen cross-linked N-telopeptide and type I collagen cross-linked C-telopeptide, and tartrate-resistant acid phosphatase-5b fraction (TRACP-5b), and the bone matrix-related marker information may be information regarding at least any one of undercarboxylated osteocalcin, pentosidine, and homocysteine.

[0172] A prediction device according to aspect 17A of the present disclosure, in any of aspects 1A to 16A above, may include at least one of an X-ray image, a CT (Computed Tomography) image, an MRI (Magnetic Resonance Imaging) image, a PET (Positron Emission Tomography) image, and an ultrasound image of the subject's body.

[0173] A prediction system according to aspect 18A of the present disclosure includes a prediction device according to any one of aspects 1A to 17A above, and a terminal device communicatively connected to the prediction device, the terminal device presenting the efficacy prediction information.

[0174] The prediction method according to aspect 19A of the present disclosure includes an acquisition step of acquiring subject information including at least one of fracture risk factor information related to the subject's fracture risk factors, bone density information indicating the bone density of the subject's bone, bone metabolism information indicating the bone metabolism state of the subject's bone, and a medical image showing the subject's bone, and a prediction step of inputting the subject information into a drug efficacy prediction model that predicts the effect of a drug having an effect on bone based on drug information related to the drug, thereby outputting drug efficacy prediction information related to the effect of the drug when the drug is administered to the subject.

[0175] A control program according to aspect 20A of the present disclosure is a control program for causing a computer to function as a prediction device described in any one of aspects 1A to 17A above, and is a control program for causing a computer to function as the acquisition unit and the prediction unit.

[0176] A recording medium according to aspect 21A of the present disclosure is a computer-readable recording medium having recorded thereon the control program according to aspect 20A above.

[0177] [Summary 2] A prediction system according to aspect 1B of the present disclosure comprises a memory unit that stores subject information relating to the bones of a subject, and a receiving unit that receives efficacy prediction information relating to the effect of a drug when administered to the subject, wherein the subject information includes at least a medical image depicting the bones of the subject, and generates the efficacy prediction information based on an efficacy prediction model that predicts the effect of the drug based on the subject information and drug information relating to a drug having an action on bones.

[0178] A prediction system according to aspect 2B of the present disclosure, in aspect 1B above, may be configured such that the efficacy prediction model is trained using training data in which subject information on each of a plurality of subjects to which the drug has been administered and drug information on the drug administered to each of the subjects are used as explanatory variables, and the effect of the drug in each of the subjects is used as a response variable.

[0179] In the prediction system according to aspect 3B of the present disclosure, in aspect 1B or 2B above, the efficacy prediction information may include at least any of a prediction result indicating a difference between a future bone density and a current bone density of the subject's bone, a prediction result indicating a change from the current in the bone density of the subject's bone, a prediction result in the case where a medication is administered and a prediction result in the case where a medication is not administered, and a prediction result in the case where a different type of medication is administered.

[0180] A prediction system according to aspect 4B of the present disclosure, in any of aspects 1B to 3B above, may include the subject information including at least one of fracture risk factor information relating to the subject's fracture risk factors, bone density information indicating the bone density of the subject's bone, and bone metabolism information indicating the bone metabolism state of the subject's bone.

[0181] A prediction system according to aspect 5B of the present disclosure, in any one of aspects 1B to 4B above, may be such that the drug information includes information regarding a drug having at least one of an effect on bone formation and an effect on bone resorption.

[0182] A prediction system according to aspect 6B of the present disclosure, in any one of aspects 1B to 5B above, may be such that the drug information includes at least the name of the drug and / or information regarding the mechanism of action.

[0183] A prediction system according to aspect 7B of the present disclosure, in any of aspects 1B to 6B above, may include the drug information including information on each of a combination of multiple drugs that may alter the bone condition of the subject, and the drug efficacy prediction information including information on each of the combinations of multiple drugs.

[0184] In the prediction system of aspect 8B of the present disclosure, in any of aspects 1B to 7B above, the efficacy prediction information may include information indicating the condition of the subject's bones after administration of the drug to the subject continues for a predetermined period of time, or information indicating changes occurring in the subject's bones before and after the predetermined period of time.

[0185] In the prediction system of aspect 9B of the present disclosure, in any of aspects 1B to 8B above, the efficacy prediction information may include information regarding a drug that may change the bone condition of the subject after administration to the subject for a specified period of time.

[0186] In the prediction system according to aspect 10B of the present disclosure, in any one of aspects 1B to 9B above, the efficacy prediction information may include a prediction result of future bone mineral density of the subject's bone.

[0187] A prediction system according to aspect 11B of the present disclosure, in any one of aspects 1B to 10B above, may be such that the efficacy prediction information includes a prediction result of a current bone mineral density of the subject's bone.

[0188] A prediction system according to aspect 12B of the present disclosure, in any one of aspects 1B to 11B above, may be arranged such that the efficacy prediction information includes a prediction result regarding a fracture site of the subject.

[0189] A prediction system according to aspect 13B of the present disclosure, in any one of aspects 1B to 12B above, may be arranged such that the efficacy prediction information includes a prediction result indicating a possibility that the subject will suffer a fracture in the future.

[0190] A prediction system according to aspect 14B of the present disclosure, in any of aspects 1B to 13B above, comprises the subject information including at least one of fracture risk factor information relating to the subject's fracture risk factors, bone density information indicating the bone density of the subject's bone, and bone metabolism information indicating the bone metabolism state of the subject's bone, and the bone density information may be a measured value obtained by measuring the bone density of a specified portion of the bone of the subject.

[0191] A prediction system according to aspect 15B of the present disclosure, in any of aspects 1B to 14B above, is characterized in that the subject information includes at least one of fracture risk factor information regarding the subject's fracture risk factors, bone density information indicating the bone density of the subject's bone, and bone metabolism information indicating the bone metabolism state of the subject's bone, the medical image is a simple X-ray image of the subject's body, and the bone density information may be an estimated value estimated based on a bone density prediction model that estimates the bone density of the bone from the simple X-ray image of the subject.

[0192] A prediction system according to aspect 16B of the present disclosure, in aspect 15B above, may be configured such that the bone density prediction model is trained using training data in which medical images showing a subject's bones are used as explanatory variables and a measured value of the bone density of the subject's bones from which the medical images were taken is used as a response variable.

[0193] A prediction device according to aspect 17B of the present disclosure includes an acquisition unit that acquires object information relating to the bones of a subject, and a prediction unit that outputs efficacy prediction information relating to the effect of a drug when administered to the subject, wherein the object information includes at least a medical image depicting the bones of the subject, and the prediction unit has an efficacy prediction model that predicts the effect of the drug based on the object information and drug information relating to a drug having an action on bones.

[0194] A prediction method according to aspect 18B of the present disclosure includes a storage step of storing object information relating to the bones of a subject, a generation step of generating efficacy prediction information relating to the effect of a drug when administered to the subject based on an efficacy prediction model that predicts the effect of the drug based on the object information and drug information relating to a drug having an action on bones, and a receiving step of receiving the efficacy prediction information, wherein the object information includes at least a medical image showing the bones of the subject.

[0195] The control program of aspect 19B of the present disclosure is a control program for causing a computer to function as a prediction system of any one of aspects 1B to 16B above, and is a control program for causing a computer to function as the memory unit and the receiving unit.

[0196] A recording medium according to embodiment 20B of the present disclosure is a computer-readable recording medium having the control program of embodiment 19B recorded therein. [Explanation of symbols]

[0197] 1, 1a Prediction device 5 Medical image management device (storage unit) 6 Electronic medical record management device (memory unit) 7 Terminal device (storage unit, receiving unit) 21 Acquisition Department 22 Drug Information Application Department 23 Prediction Department 24 Output control section 25 Learning Department 32 Teacher Data 34 Trained drug efficacy prediction model 100a, 100b Prediction System S11 Acquisition step S12 Generation Step

Claims

1. a storage unit that stores target information related to a bone of a target; a prediction unit that uses a drug efficacy prediction model that predicts an effect of the drug based on the subject information and drug information on the drug having an action on bones, and outputs drug efficacy prediction information on an effect of the drug when the drug is administered to the subject, based on the subject information and the drug information on the drug having an action on bones; a receiving unit for receiving the drug efficacy prediction information, The object information includes at least a medical image showing a bone of the object. Prediction system.

2. The drug efficacy prediction model is trained using training data in which subject information on each of a plurality of subjects to which the drug is administered and drug information on the drug administered to each of the subjects are used as explanatory variables, and the effect of the drug on each of the subjects is used as a response variable. The prediction system of claim 1 .

3. The drug efficacy prediction information includes at least one of a prediction result showing a difference between a future bone density and a current bone density of the subject's bone, a prediction result showing a change in the bone density of the subject's bone from the current, a prediction result when a drug is administered and a prediction result when a drug is not administered, and a prediction result when a different type of drug is administered. The prediction system of claim 1 .

4. The subject information includes at least one of fracture risk factor information regarding a fracture risk factor of the subject, bone density information indicating the bone density of the bone of the subject, and bone metabolism information indicating the bone metabolism state of the bone of the subject. The prediction system of claim 1 .

5. The drug information includes information on a drug having at least one of an effect on bone formation and an effect on bone resorption. The prediction system of claim 1 .

6. The drug information includes at least the name and / or mechanism of action of the drug; The prediction system of claim 1 .

7. the drug information includes information regarding each of a combination of a plurality of drugs that can change a bone condition of the subject; The drug efficacy prediction information includes information on each of the combinations of the multiple drugs. The prediction system of claim 1 .

8. The efficacy prediction information includes information showing a bone condition of the subject after administration of the drug to the subject is continued for a predetermined period of time, or information showing changes occurring in the bone of the subject before and after the predetermined period of time. The prediction system of claim 1 .

9. The drug efficacy prediction information includes information about a drug that can change the bone condition of the subject after administration to the subject for a predetermined period of time. The prediction system of claim 1 .

10. The drug efficacy prediction information includes a prediction result of future bone mineral density of the subject's bone. The prediction system of claim 1 .

11. The drug efficacy prediction information includes a prediction result of a current bone mineral density of the subject's bone. The prediction system of claim 1 .

12. The drug efficacy prediction information includes a prediction result regarding a fracture site of the subject. The prediction system of claim 1 .

13. The drug efficacy prediction information includes a prediction result indicating a possibility that the subject will have a bone fracture in the future. The prediction system of claim 1 .

14. The subject information includes at least one of fracture risk factor information regarding a fracture risk factor of the subject, bone density information indicating a bone density of the bone of the subject, and bone metabolism information indicating a bone metabolism state of the bone of the subject, The bone density information is a measurement value obtained by measuring the bone density of a bone at a predetermined site of the subject. The prediction system of claim 1 .

15. The subject information includes at least one of fracture risk factor information regarding a fracture risk factor of the subject, bone density information indicating a bone density of the bone of the subject, and bone metabolism information indicating a bone metabolism state of the bone of the subject, the medical image is a plain X-ray image of the subject's body; The bone density information is an estimated value estimated based on a bone density prediction model that estimates bone density of a bone from the plain X-ray image of the subject. The prediction system of claim 1 .

16. The bone mineral density prediction model is trained using training data in which a medical image showing a bone of a subject is used as an explanatory variable and a measurement value of the bone mineral density of the bone of the subject from which the medical image was taken is used as a response variable. The prediction system of claim 15.

17. an acquisition unit for acquiring object information relating to a bone of a target; a prediction unit that uses a drug efficacy prediction model that predicts an effect of the drug based on the subject information and drug information on the drug having an action on bones to output drug efficacy prediction information on the effect of the drug when the drug is administered to the subject, based on the subject information and drug information on the drug having an action on bones; The object information includes at least a medical image showing a bone of the object. Prediction device.

18. A method for determining whether a bone is a bone of a subject, comprising: a generation step in which a computer generates efficacy prediction information on an effect of a drug when the drug is administered to the subject from the subject information and the drug information on the drug having an action on bones, based on a drug efficacy prediction model that predicts an effect of the drug based on the subject information and drug information on the drug having an action on bones; A receiving step of receiving the drug efficacy prediction information by a computer, The object information includes at least a medical image showing a bone of the object. Forecasting methods.

19. A control program for causing a computer to function as the prediction system according to claim 1 , the control program causing a computer to function as the storage unit, the prediction unit, and the receiving unit.

20. A computer-readable recording medium having the control program according to claim 19 recorded thereon.

Images