Disease prediction system and program

Abstract

To predict the future development of a disease.SOLUTION: The present invention has a prediction unit for inputting input information of a subject into a neural network and predicting the future development of a disease of the subject. The input information has at least first information including an image and second information on a hormone-like acting substance of the subject. The second information includes a measured value of a hormone-like acting substance in the blood or in the urine of the subject. The neural network is generated by machine learning which uses first learning information, which is the same type as the first information, and second learning information, which is the same type as the second information, as learning data and using a diagnosis result of the disease as teacher data.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present disclosure relates to a disease prediction system.

Background Art

[0002] Patent Document 1 describes a diagnostic support device for osteoporosis, which is one of the diseases.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In order to grasp the risk of future diseases, etc., it is required to predict the future onset of these diseases.

Means for Solving the Problems

[0005] A disease prediction system according to an embodiment of the present invention includes a prediction unit that inputs input information regarding a subject to a neural network to predict the onset of a future disease of the subject. The input information has first information including at least an image of the subject and second information regarding a hormone-like substance of the subject. The second information includes measured values of hormone-like substances in the blood or urine of the subject. The neural network uses first learning information of the same type as the first information and second learning information of the same type as the second information as learning data, and is generated by machine learning using the diagnosis result of the disease as teacher data.

Brief Description of the Drawings

[0006]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Embodiments for Carrying Out the Invention

[0007] The disease prediction system 1 of the present disclosure can predict, for example, the onset of diseases in middle-aged and elderly people. Diseases in middle-aged and elderly people include, for example, osteoporosis, osteoarthritis, spondylosis, fractures, sarcopenia, frailty, menopausal disorders, ED (Erectile Dysfunction), periodontal disease, etc.

[0008] In the following, unless otherwise specified, osteoporosis will be described as an example of the disease.

[0009] FIG. 1 shows the concept of the configuration of the disease prediction system 1 of the present disclosure.

[0010] The disease prediction system 1 of the present disclosure has a terminal device 2 and a prediction device 3. The terminal device 2 acquires data of a subject used for disease diagnosis. The prediction device 3 can predict the onset of future diseases based on the data acquired by the terminal device 2.

[0011] As described above, the terminal device 2 can acquire data. The data acquired by the terminal device 2 is input to the prediction device 3 as input information I. Note that the input information I has information on the subject's hormone-like substances (second information I2) in addition to the data of the subject used for disease diagnosis (first information I1). And the terminal device 2 has a first terminal device 21 that acquires the first information I1 and a second terminal device 22 that acquires the second information.

[0012] When the first terminal device 21 is used, for example, in the diagnosis of osteoporosis by the disease prediction system 1, it may be, for example, a simple X-ray imaging device or a bone mass measurement device. Also, in this case, the first information I1 may have a medical image such as a simple X-ray image.

[0013] The second terminal device 22 can acquire second information I2 that affects a disease related to the first information I1 acquired by the first terminal device 21. The second information I2 may be, for example, information on the presence or absence or concentration of a hormone-like substance. The second information I2 may be, for example, a measured value of a hormone-like substance in the blood or urine of the subject. Specifically, when the first information I1 is related to osteoporosis, the second information I2 may be data indicating the presence or absence of non-steroidal estrogen. The second terminal device 22 may be, for example, an inspection device using a surface acoustic wave (SAW) sensor.

[0014] After acquiring the input information I, the terminal device 2 may transfer the input information I to the prediction device 3. For example, when acquiring a simple X-ray image of a subject, the terminal device 2 (the first terminal device 21) is installed, for example, in an X-ray room to take an X-ray photograph of the subject. Then, the image data is transferred from the terminal device 2 to the prediction device 3, and through the prediction device 3, the future onset of the disease can be predicted.

[0015] Note that the terminal device 2 does not necessarily transfer the input information I directly to the prediction device 3. In this case, for example, the input information I acquired by the terminal device 2 may be stored in a storage medium, and the input information I may be input to the prediction device 3 via the storage medium.

[0016] After acquiring the second input information I2, the second terminal device 22 may transfer the second input information I2 to the prediction device 3. For example, when an inspection device using a SAW sensor transfers information on the concentration of a hormone-like substance in the blood or urine of a subject, it may transfer information as a numerical value of the concentration, or it may transfer information on the primary data before conversion to the concentration. The primary data may be, for example, information regarding a phase change, amplitude change, frequency change, etc. of a detection signal by a sensor.

[0017] FIG. 2 shows a concept of the configuration of the prediction device 3 according to the present embodiment.

[0018] The prediction device 3 can predict the future onset of a disease of a subject from the input information I input to the prediction device 3. For example, when the disease prediction system 1 predicts osteoporosis, the prediction device 3 can predict the future onset of osteoporosis of the subject from the input information I including the medical image acquired by the terminal device 2 and output the predicted prediction result O.

[0019] The prediction device 3 includes an input unit 31, a control unit 34, an output unit 33, and a storage unit 35. The input unit 31, the output unit 33, the control unit 34, and the storage unit 35 are electrically connected to each other, for example, by a bus 60. The input unit 31 receives the input information I from the terminal device 2. The control unit 34 can predict the onset of a disease based on the input information I based on the prediction unit 32 described later by executing a control program. The output unit 33 can output the prediction result O predicted by the prediction unit 32. The storage unit 35 stores a control program and various data, parameters, etc. necessary for control.

[0020] The prediction device 3 of the present disclosure has a plurality of electronic components and circuits. In other words, the prediction device 3 can form each component constituting the prediction device 3 by a plurality of electronic components and circuits. The plurality of electronic components may be, for example, active elements such as transistors or diodes, or passive elements such as capacitors, and may be formed by a conventionally well-known method.

[0021] As described above, the input unit 31 receives the input information I. The input unit 31 may have a communication unit so that the input information I acquired by the terminal device 2 is directly input from the terminal device 2. Further, the input unit 31 may include an input device capable of inputting the input information I or other information. The input device may be, for example, a keyboard, a touch panel, or a mouse.

[0022] The control unit 34 can comprehensively manage the operation of the prediction device 3 by controlling other components of the prediction device 3. The control unit 34 can also be referred to as a control device or a control circuit. The control unit 34 includes at least one processor to provide control and processing capabilities for executing various functions, as described in more detail below.

[0023] In one embodiment, the processor includes, for example, one or more circuits or units configured to execute one or more data calculation procedures or processes by executing instructions stored in a related memory. In other embodiments, the processor may be firmware (e.g., discrete logic components) configured to execute one or more data calculation procedures or processes.

[0024] According to various embodiments, the processor includes one or more processors, controllers, microprocessors, microcontrollers, application-specific integrated circuits (ASICs), digital signal processing devices, programmable logic devices, field-programmable gate arrays, or any combination of these devices or configurations, or other known device and configuration combinations, and may execute the functions described below. In this example, the control unit 34 includes, for example, a CPU (Central Processing Unit).

[0025] The storage unit 35 includes a non-temporary recording medium that can be read by the CPU of the control unit 34, such as a ROM (Read Only Memory) and a RAM (Random Access Memory). A control program for controlling the prediction device 3 is stored in the storage unit 35. Various functions of the control unit 34 are realized by the CPU of the control unit 34 executing the control program in the storage unit 35. The control program can also be said to be a prediction program for causing the computer device 1 to function as the prediction device 3.

[0026] In this example, by the control unit 34 executing the control program in the storage unit 35, a prediction unit 32 capable of estimating the prediction result O is formed in the control unit 34. The prediction unit 32 includes, for example, a neural network. The control program can also be said to be a program for causing the computer device 1 to function as a neural network (prediction unit 32). The configuration example of the neural network will be described in detail later.

[0027] In addition to the control program, the storage unit 35 stores learned parameters related to the neural network, estimation data (hereinafter also referred to as "input information"), learning data, and teacher data. The learning data and the teacher data are data used when training the neural network. The learned parameters and the estimation data are data used when the learned neural network estimates the onset of a disease.

[0028] The prediction unit 32 can predict the future onset of a disease of the subject from the input information I input to the input unit 31. The prediction unit 32 has AI (Artificial Intelligence). The AI of the present disclosure may be, for example, a neural network.

[0029] In addition, the prediction unit 32 has been pre-trained. That is, by applying machine learning to the prediction unit 32 using learning data and teacher data, the prediction unit 32 can calculate the prediction result O from the input information I. Note that the learning data or the teacher data may be data corresponding to the input information I input to the prediction device 3 and the prediction result O output from the prediction device 3.

[0030] Figs. 3 and 4 show the concept of the configuration of the prediction unit 32 of the present disclosure.

[0031] The prediction unit 32 has a first neural network 321 and a second neural network 322. The first neural network 321 may be any neural network suitable for handling time-series information. For example, the first neural network 321 may be a ConvLSTM network that combines a CNN (Convolutional Neural Network) and an LSTM (Long short-term memory). The second neural network 322 may be, for example, a convolutional network composed of a CNN.

[0032] Fig. 3 shows the concept of the configuration of the first neural network 321 of the present disclosure. Fig. 4 shows the concept of the configuration of the second neural network 322.

[0033] The first neural network 321 has an encoding unit E and a decoding unit D. The encoding unit E can extract the feature amounts of the temporal change and the position information of the input information I. The decoding unit D can calculate new feature amounts based on the feature amounts extracted by the encoding unit E, the temporal change of the input information I, and the initial value.

[0034] The encoding unit E has a plurality of ConvLSTM layers (Convolutional Long short - term memory) E1. The decoding unit D has a plurality of ConvLSTM layers (Convolutional Long short - term memory) D1. Note that the contents learned by each of the plurality of ConvLSTM layers E1 may be different. The contents learned by each of the plurality of ConvLSTM layers D1 may be different. For example, a certain ConvLTSM layer may learn fine - grained contents such as changes in each pixel, while another ConvLSTM layer may learn rough contents such as changes in the overall image.

[0035] The second neural network 322 has a conversion unit C. The conversion unit C can convert the feature amount calculated by the decoding unit D into bone mass. The conversion unit C has a plurality of convolutional layers C1, a plurality of pooling layers C2, and a fully - connected layer C3. The fully - connected layer C3 is located in front of the output unit 33, and the plurality of convolutional layers C1 and the plurality of pooling layers C2 are alternately arranged.

[0036] During the learning of the prediction unit 32, the learning data is input to the encoding unit E of the prediction unit 32, and the teacher data is compared with the output data output from the conversion unit C of the prediction unit 32.

[0037] The output unit 33 can display the prediction result O. The output unit 33 is, for example, a liquid crystal display or an organic EL display. The output unit 33 can display various information such as characters, symbols, and graphics. The output unit 33 can display, for example, numbers or images.

[0038] <An example of input information, learning data, and teacher data> The input information I has the first information I1 of the subject used for diagnosing the disease. The first information I1 may be, for example, the test results or medical images of the subject. Specifically, for example, when the disease prediction system 1 predicts osteoporosis, it may be the bone mass, bone metabolism markers, X-ray images, CT images, etc. of the chest, lumbar spine, or proximal femur of the subject.

[0039] In the case of osteoporosis, specifically, the first information I1 has the image data of the plain X-ray image showing the bones of the subject. Also, the bones to be imaged are mainly cortical bone and cancellous bone of biological origin, but the target bones may include artificial bone mainly composed of calcium phosphate or regenerated bone artificially manufactured by regenerative medicine or the like.

[0040] In addition, the input information I may have the second information regarding the hormone-like substances that affect the disease related to the first information I1. The second information I2 may be, for example, information on the presence or absence or concentration of the hormone-like substances. The second information I2 may be, for example, information on the hormone-like substances that affect osteoporosis when the first information I1 is data related to osteoporosis. Specifically, it may be the concentration information of at least one type such as fibroblast growth factor 23 (FGF23), leptin, insulin, sclerostin, etc. Furthermore, if the subject is female, it may be the concentration information of at least one type such as steroid estrogens such as estrone, estradiol, estriol, non-steroid estrogens, progesterone, or estrogen receptors. If the subject is male, it may be the concentration information of at least one type such as androgens such as testosterone, dihydrotestosterone, or androgen receptors.

[0041] When the second information I2 is, for example, information regarding a hormone-like substance that affects sarcopenia in the case where the first information I1 is data related to sarcopenia, it may be information regarding at least one type of concentration such as ghrelin, leptin, adiponectin, etc. Additionally, if the subject is female, it may be information regarding at least one type of concentration such as steroid estrogens like estrone, estradiol, estriol, non-steroidal estrogens, progesterone, or estrogen receptors. If the subject is male, it may be information regarding at least one type of concentration such as androgens like testosterone, dihydrotestosterone, or androgen receptors.

[0042] In addition, when the disease prediction system 1 predicts osteoarthritis or spondylosis, the first information I1 may be an X-ray image of the affected part. Also, when predicting a fracture, the first information I1 may be bone mass, bone metabolism markers, fracture history, and X-ray images. When predicting sarcopenia, the first information I1 may be an X-ray image of the lower leg, muscle mass, grip strength, and walking speed. When predicting frailty, the first information I1 may be information regarding grip strength, walking speed, activity level, oral function, fatigue, and sociality. When predicting menopausal disorders, the first information I1 may be the presence or absence of flushing of the face, ease of sweating, coldness of the face and hands, shortness of breath / palpitations, ease of falling asleep / deepness of sleep, degree of irritability, degree of depression, headache / dizziness / nausea, ease of fatigue, and shoulder / back / hand pain. When predicting ED, the first information I1 may be information regarding the success and satisfaction of sexual intercourse. When predicting periodontal disease, the first information I1 may be information such as X-ray images, CT images, periodontal pocket depth, attachment level, and oral hygiene status.

[0043] The learning data has first learning information of the same type as the first input information I1. For example, if the first input information I1 is a plain X-ray image, the learning data may also have a plain X-ray image. Furthermore, if the first input information I1 is bone mass, the learning data may also have bone mass.

[0044] The learning data takes into account the change over time. That is, the learning data may be a series of data at different time axes for examining the same person. Also, the learning data may be a group of data of other people. Further, when the learning data is image data, the image data may be a series of data at different time axes for photographing the same person at the same body part. As a result, the prediction unit 32 for which the learning process has been performed can predict the onset of a future disease.

[0045] Also, the learning data has second learning information of the same type as the second input information I2. That is, the learning data is information regarding a hormone-like substance of the subject. Also, the second learning information may be information for a predetermined period at or around the time of acquisition of the first learning information. For example, if the subject is a female, the second input information I2 may be concentration information of at least one type such as steroid estrogens such as estrone, estradiol, and estriol, non-steroid estrogens, progesterone, or estrogen receptors.

[0046] Note that the learning data may be acquired using the terminal device 2 in the same manner as the input information I. For the first learning information, for example, it may be acquired using a first terminal device 21 such as a simple X-ray imaging device. Also, for the second learning information, for example, it may be acquired using a second terminal device 22 such as a SAW (Surface Acoustic Wave) sensor. If the learning data is, for example, the concentration of a non-steroid estrogen, it can be acquired using the second terminal device 22 as described above.

[0047] The teacher data includes a disease diagnosis result corresponding to the learning data. For example, when predicting osteoporosis, the teacher data is the actually measured bone mass or the diagnosis result of osteoporosis corresponding to each of a plurality of learning image data. The actually measured bone mass or the diagnosis result of osteoporosis only needs to be evaluated at approximately the same time as the time when the learning image data was taken.

[0048] Teacher data, for example in the case of bone mass, may be measured by, for example, the DEXA (Dual-Energy X-ray Absorptiometry) method or the ultrasonic method.

[0049] <Neural network learning example> The prediction unit 32 is optimized by machine learning using learning data and teacher data so as to be able to calculate a prediction result O from the input information I. That is, the control unit 34 calculates the prediction result O from the input information I based on the approximation formula (prediction unit 32) optimized by machine learning, but adjusts the parameters in the prediction unit 32 so that the difference between the pseudo-prediction result calculated from the learning data input to the input unit 31 and output from the output unit 33 and the teacher data becomes small, and thus machine learning is performed. As a result, the prediction unit 32 can perform an operation based on the learned parameters on the input information I and output the prediction result O.

[0050] As a method for adjusting the parameters, for example, the error backpropagation method is adopted. The parameters include, for example, the parameters used in the encoding unit E, the decoding unit D, and the conversion unit C. Specifically, the parameters include the weight coefficients used in the ConvLSTM layers of the encoding unit E and the decoding unit D, and the convolutional layer and the fully connected layer of the conversion unit C.

[0051] As described above, in the disease prediction system 1, the onset of a disease can be predicted by outputting the prediction result O from the input information I.

[0052] Here, for example, it is disclosed that a conventional osteoporosis diagnosis support device determines osteoporosis using feature amounts related to cortical bone. However, the conventional osteoporosis diagnosis support device diagnoses the current osteoporosis and does not show the possibility of future onset.

[0053] In contrast, the disease prediction system 1 according to the present invention can predict the onset of a future disease of the subject from the input information I. Therefore, it is possible to predict how the disease will progress after the data is acquired from the input information I at the time of data acquisition. And the disease prediction system 1 according to the present invention also has information (second information) regarding a hormone-like substance as the input information I. For example, since the way the disease progresses varies depending on the presence or absence of a hormone-like substance, etc., the accuracy of future prediction can be improved by inputting the second information.

[0054] The prediction result O of the disease prediction system 1 only needs to be a prediction for the future relative to the acquisition date of the input information I. For example, the disease prediction system 1 may perform predictions from 3 months to 50 years after the acquisition date of the input information I, more preferably from 6 months to 10 years after.

[0055] The prediction result O may be output as a future numerical value. For example, when predicting bone mass, the prediction result O may be a numerical value represented by at least one of, for example, YAM (Young Adult Mean), T-score, and Z-score.

[0056] In addition, when predicting osteoarthritis or spondylosis, the prediction result O may be, for example, the KL (Kellgren-Lawrence) grade classification. Also, when predicting fractures, sarcopenia, frailty, menopausal disorders, ED, periodontal disease, the prediction result O may be the presence or absence of onset. The presence or absence of onset may be determined by a predetermined threshold value based on the numerical value output by the prediction result.

[0057] The prediction result O may output the future onset probability of the disease. The onset probability may output, for example, the onset probability on a specific date. In this case, it may be output like "the probability of the disease after 1 year is 10%". Also, it may output when the onset date is, etc. In this case, it may be output like "there is a possibility of developing osteoporosis after 7 years", for example.

[0058] The prediction result O may output a secular change. For example, the transition of the prediction result O with respect to the time axis, such as the onset probability after 1 year, the onset probability after 5 years, the onset probability after 10 years, etc., may be output.

[0059] The disease prediction system 1 may output a prediction result O for predicting the onset of a plurality of diseases from one type of first information I1. In this case, for example, by inputting the concentration of non-steroidal estrogen in urine as the first information I1, the onset of diseases such as osteoporosis and menopausal disorder can be predicted simultaneously.

[0060] Also, the disease prediction system 1 may output a prediction result O for predicting the onset of a plurality of diseases from one type of first information I1 and one type of second information I2. In this case, for example, by inputting an X-ray image as one type of first information I1 and the concentration of non-steroidal estrogen in urine as one type of second information I2, the onset of a plurality of diseases such as osteoporosis and osteoarthritis can be predicted simultaneously.

[0061] The input information I may include third information I3 having the individual data of the subject. The third information I3 may be information regarding the health status of the subject. The individual data may include, for example, any one or more types of information regarding age information, gender information, height information, weight information, systolic blood pressure, total cholesterol, triglyceride, bad cholesterol (LDL-C, triglyceride), good cholesterol (HDL-cholesterol), insulin resistance index (HOMA-R index), blood glucose level, presence or absence of menopause, sperm count, etc.

[0062] In this case, as the learning data, third learning information of the same type as the third information I3 may be learned. Also, the third learning information may be information at the time of acquisition of the first learning information or within a predetermined period before and after the acquisition.

[0063] Also, the third piece of information I3 may include lifestyle information. The lifestyle information may include any one or more types of information related to eating habits, drinking habits, smoking habits, or exercise habits such as walking speed or number of steps.

[0064] The input information I may have fourth information I4 regarding the intervention for the subject. The fourth information I4 may be information regarding the planned (future) changes in the subject's lifestyle or the like. The fourth information I4 may include, for example, any one or more types of information related to changes in diet, changes in lifestyle, changes in weight, physical therapy, drug therapy, or supplements to be taken.

[0065] In this case, as learning data, fourth learning information of the same type as the fourth information I4 may be learned. Also, the fourth learning information may be information at the time of acquisition of the first learning information or within a predetermined period before or after the acquisition.

[0066] Also, when inputting information regarding supplements to be taken as the fourth information I4, the information regarding supplements may include any one or more types of information related to the intake of calcium, vitamin D, vitamin K, branched-chain amino acids, flavonoids, or probiotics. The probiotics may be, for example, Bacillus subtilis strain C-3102 or the like.

[0067] In the disease prediction system 1, the prediction result O may output a first result based on the input information I selected from the input information I excluding the fourth information I4, and a second result based on the input information I including the fourth information I4 and at least the first input information I1.

[0068] In the disease prediction system 1, not only the prediction result O which is a future prediction but also the current diagnosis result may be output. As a result, the change in the disease over time can be compared.

[0069] The input information I may have fifth input information I5 having the subject's bone metabolism marker information. The bone metabolism information may be, for example, bone resorption ability or bone formation ability. These can be measured, for example, by at least one of the bone resorption markers type I collagen cross-linked N-telopeptide (NTX), type I collagen cross-linked C-telopeptide (CTX), tartrate-resistant acid phosphatase (TRACP-5b), deoxypyridinoline (DPD), the bone formation marker bone alkaline phosphatase (BAP), type I collagen cross-linked N-propeptide (P1NP), and the bone-related matrix marker undercarboxylated osteocalcin (ucOC). The bone resorption marker may be measured using serum or urine as a specimen.

[0070] In this case, fifth learning information of the same type as the fifth information I5 may be learned as learning data. Also, the fifth learning information may be information at the time of acquisition of the first learning information or within a predetermined period before and after the acquisition.

[0071] The disease prediction system 1 may have a third terminal device that acquires the fifth input information I5. Also, using the second terminal device 22, the fifth input information I5 may be measured simultaneously with the second input information I2. For example, the DPD of the bone metabolism marker and the concentration of the non-steroidal estrogen, a hormone-like substance, may be measured simultaneously from a single urine specimen of the subject. Further, the second terminal device 22 may measure a plurality of the bone metabolism marker, which is one of the first input information I1, and the hormone-like substance of the second input information I2 simultaneously. Specifically, the DPD and CTX of the bone metabolism marker and the concentrations of the steroidal estrogen and non-steroidal estrogen of the hormone-like substance may be measured simultaneously from a single urine specimen of the subject.

[0072] <Other Embodiments> FIG. 5 shows the concept of the configuration of the prediction unit 32a using the linear regression model of the present disclosure.

[0073] In the above-described embodiment, the prediction unit 32 was described by using a neural network as the prediction unit 32. However, the prediction unit 32 may be a linear regression model that uses input information as an explanatory variable, prediction parameters as coefficients of the explanatory variable, and a prediction result as a target variable. The prediction parameters may be optimized by the least squares method using learning data and teacher data. That is, by using the learning data as an explanatory variable vector and the teacher data as a target variable vector, an explanatory variable coefficient vector that minimizes the sum of squared errors may be determined.

[0074] The prediction unit 32a using the linear regression model calculates the prediction result Y by inputting the input information into the explanatory variables X1, X2 ··· Xk. The coefficients θ1, θ2 ··· θk of the explanatory variables, which are the prediction parameters, may be optimized in advance by the least squares method using, for example, learning data and teacher data.

[0075] In the prediction unit 32a using the linear regression model, for example, diseases such as osteoporosis, fracture, sarcopenia, frailty, menopause, ED (Erectile Dysfunction), or periodontal disease can be easily predicted.

[0076] Note that the present invention is not limited to the examples of the above-described embodiments, and includes various modifications as long as there is no contradiction in the content. Also, the embodiments according to the present invention can be combined as appropriate.

[0077] For example, in the above example, when a neural network is used as the prediction unit 32, an example in which a ConvLSTM network is applied was described, but the present invention is not limited to this. For example, the prediction unit 32 may use an RNN (Recurrent Neural Network) or a GAN (Generative Adversarial Network). Also, the prediction unit 32 may combine a plurality of neural networks. Specifically, a composite neural network combining a ConvLSTM network and a convolutional neural network may be used.

[0078] In addition, in the above example, as the prediction unit 32, an example having a neural network and an example having a linear regression model were each described. However, in the present invention, the prediction device 3 of the disease prediction system 1 may have a plurality of different prediction units. For example, it may simultaneously have a prediction unit 32a using a linear regression model and a prediction 32b using a ConvLSTM neural network.

[0079] In this case, in the disease prediction system 1, the prediction unit 32a outputs a first prediction result O1. Further, the prediction unit 32b outputs a second prediction result as the second prediction result O2. As a result, as the prediction result O, the first prediction result and the second prediction result can be compared.

[0080] The disease prediction system 1 may output a third prediction result based on the first prediction result and the second prediction result as the prediction result O. As a result, for example, the result of correcting the first prediction result based on the second prediction result (the third prediction result) can be used as the prediction result O.

[0081] 〔Other aspects〕 A disease prediction system according to an aspect of the present disclosure includes an input unit to which input information having first information of a subject used for diagnosing a disease and second information regarding a hormone-like substance of the subject is input, and a control unit that predicts the onset of the future disease of the subject from the input information input to the input unit.

Explanation of reference numerals

[0082] 1 Disease prediction system 2 Terminal device 21 First terminal device 22 Second terminal device 3 Prediction device 31 Input unit 32 Prediction unit 32a Prediction unit 32b Prediction unit 33 Output unit 34 Control unit 35 Storage unit O Prediction result I Input information

Claims

1. A prediction system comprising a prediction unit that inputs input information regarding a subject into a neural network to predict the onset of a future disease of the subject, wherein the input information has first information including at least an image of the subject and second information regarding a hormone-like substance of the subject, the second information includes measured values of hormone-like substances in the blood or urine of the subject, the neural network is generated by machine learning using first learning information of the same type as the first information and second learning information of the same type as the second information as learning data, and using the diagnosis result of the disease as teacher data, and the prediction unit outputs a Kellgren-Lawrence (KL) grade classification as a prediction result.

2. The disease prediction system according to claim 1, wherein the prediction unit predicts the onset of two or more diseases among the diseases.

3. The disease prediction system according to claim 1 or 2, wherein the second information is measured using a Surface Acoustic Wave (SAW) sensor.

4. The disease prediction system according to any one of claims 1 to 3, wherein the second learning information is obtained using a Surface Acoustic Wave (SAW) sensor.

5. The disease prediction system according to any one of claims 1 to 4, wherein the image is a simple X-ray image.

6. The disease prediction system according to any one of claims 1 to 5, wherein the second information is information on the presence or concentration of a hormone-like substance.

7. The disease prediction system according to any one of claims 1 to 6, wherein the learning data is a series of data at different time points of examining the same person.

8. The disease prediction system according to any one of claims 1 to 7, wherein the first learning information is a series of data at different time points of photographing the same site.

9. The disease prediction system according to any one of claims 1 to 8, wherein the learning data is a data group of other people.

10. The disease prediction system according to any one of claims 1 to 9, wherein the prediction unit simultaneously predicts the onset of a plurality of the diseases.

11. The disease prediction system according to any one of claims 1 to 10, further comprising an output unit that outputs the prediction result predicted by the prediction unit.

12. The disease prediction system according to any one of claims 1 to 11, wherein the prediction unit outputs a change over time.

13. The disease prediction system according to any one of claims 1 to 12, wherein the input information includes third information regarding the health condition of the subject.

14. The disease prediction system according to claim 13, wherein the third information includes at least one of age information, gender information, height information, and weight information.

15. The disease prediction system according to any one of claims 1 to 14, wherein the input information includes fourth information regarding an intervention to the subject.

16. The disease prediction system according to claim 15, wherein the prediction unit predicts a first result based on input information excluding the fourth information and a second result based on input information including the fourth information.

17. The disease prediction system according to any one of claims 1 to 16, wherein the input information includes fifth information having bone metabolism marker information of the subject.

18. The disease prediction system according to any one of claims 1 to 17, which outputs in combination with the current diagnosis result of the subject.

19. A program for causing a computer to function as the disease prediction system according to claim 1, the program for causing a computer to function as the prediction unit.

Images