Information processing device, information processing method, and information processing program

Abstract

An information processing device comprising at least one processor, wherein the processor: acquires a plurality of items of first biometric information measured over time regarding a subject; accepts, with respect to second biometric information regarding the subject of which the type is different from the first biometric information and which is correlated with the first biometric information, designation of prediction timing indicating a desired timing for prediction; and performs prediction of the second biometric information at the predicted timing on the basis of the plurality of items of first biometric information.

Description

[Technical Field]

[0001] The present disclosure relates to an information processing device, an information processing method, and an information processing program. [Background technology]

[0002] Conventionally, a technique for predicting other biological information at a certain time point based on biological information at the same time point has been known. For example, Japanese Patent Application Laid-Open Publication No. 2009-136446 describes a technique for receiving an electrocardiogram signal from a subject and estimating an ultrasound image during a time period when ultrasound is not being transmitted or received using a regression model of the electrocardiogram signal and the ultrasound image. Furthermore, International Publication No. WO 2020 / 013230 describes a technique for estimating glucose levels for periods other than those during which blood glucose measurements were actually performed from measured glycoalbumin concentrations based on the correlation between glycoalbumin concentrations and glucose levels. Summary of the Invention [Problem to be solved by the invention]

[0003] Incidentally, the accuracy of a diagnosis based on biological information increases when the biological information at the time of measurement is in a state suitable for diagnosis. For example, when diagnosing a subject suspected of having arrhythmia by taking cardiac images, obtaining a cardiac image at the time of arrhythmia occurrence increases the accuracy of the diagnosis. However, in reality, since the probability of arrhythmia occurring exactly at the timing of imaging is low, it may be possible to obtain only a cardiac image in a better state than when the arrhythmia occurred, i.e., a cardiac image that is not suitable for diagnosis.

[0004] For example, when diagnosing a subject suspected of having high blood pressure by taking cardiac images, obtaining cardiac images when the subject's blood pressure is in a normal state increases the accuracy of the diagnosis. However, it is known that some subjects suspected of having high blood pressure experience particularly high blood pressure while in a medical institution due to tension and stress (so-called "white coat hypertension"). For such subjects, even if cardiac images are taken under normal conditions, cardiac images obtained under conditions different from normal conditions because the subject is in a medical institution may result in an overdiagnosis.

[0005] As described above, even with similar biological information (cardiac images), the state and timing suitable for diagnosis may differ depending on the type of disease being diagnosed. Therefore, there is a demand for a technology that supports appropriate diagnosis by predicting biological information in a state and timing suitable for diagnosis based on other biological information that is correlated with a certain biological information.

[0006] The present disclosure provides an information processing device, an information processing system, an information processing method, and an information processing program that can support appropriate diagnosis. [Means for solving the problem]

[0007] A first aspect of the present disclosure is an information processing device comprising at least one processor, which acquires a plurality of first biometric information measured over time on a subject, accepts a specification of a prediction timing indicating a desired timing for prediction of second biometric information on the subject, the second biometric information being of a different type from the first biometric information and correlated with the first biometric information, and predicts the second biometric information at the prediction timing based on the plurality of first biometric information.

[0008] A second aspect of the present disclosure is that, in the above-mentioned first aspect, each of the multiple first biometric information is assigned the date and time of measurement of the first biometric information, and the processor may predict the second biometric information by taking into account changes over time in the first biometric information indicated by the multiple first biometric information.

[0009] In a third aspect of the present disclosure, in the second aspect, the processor may predict the second biological information using a recurrent neural network (RNN) or a long short-term memory (LSTM).

[0010] A fourth aspect of the present disclosure is that in any one of the first to third aspects above, the processor may predict first biometric information at a predicted timing based on a plurality of first biometric information, and predict second biometric information based on the first biometric information at the predicted timing.

[0011] A fifth aspect of the present disclosure is that in any one of the first to fourth aspects above, the processor may acquire at least one second biometric information measured on the subject, and predict the second biometric information based on the acquired first biometric information and second biometric information.

[0012] According to a sixth aspect of the present disclosure, in any one of the first to fifth aspects, the processor may accept, as the prediction timing, a designation of a date and time at which prediction of the second biological information is desired.

[0013] A seventh aspect of the present disclosure is that, in any one of the first to fifth aspects above, the processor may accept a specification of a condition that the first biometric information must satisfy, and specify the date and time when the first biometric information satisfies the condition as the predicted timing.

[0014] An eighth aspect of the present disclosure is the seventh aspect, wherein each of the plurality of first biometric information is assigned the measurement date and time of the first biometric information, and the processor acquires a plurality of second biometric information measured on the subject, the second biometric information being assigned the measurement date and time of the second biometric information, interpolates the second biometric information at a point in time when the first biometric information is measured but the second biometric information is not measured based on the second biometric information before and after that point in time, determines a pattern according to the relationship between the first biometric information at that point in time and the interpolated second biometric information, and specifies predetermined conditions for each pattern as conditions that the first biometric information must satisfy.

[0015] A ninth aspect of the present disclosure is any one of the first to eighth aspects, wherein the predicted timing may be in the past than the present time.

[0016] A tenth aspect of the present disclosure is any one of the first to ninth aspects, wherein the first biological information may be measured more frequently than the second biological information.

[0017] An eleventh aspect of the present disclosure is any one of the first to tenth aspects above, wherein the first biometric information and the second biometric information may vary non-periodically in accordance with the behavior of the subject.

[0018] A twelfth aspect of the present disclosure is any one of the first to eleventh aspects, wherein the first biometric information indicates at least one of body temperature, heart rate, electrocardiogram, electromyogram, blood pressure, arterial oxygen saturation, blood glucose level, and lipid level, and the second biometric information may indicate at least one of electrocardiogram, electroencephalogram, medical image taken by a medical imaging device, and the results of at least one of hematological test, infectious disease test, biochemical test, and urinalysis.

[0019] A thirteenth aspect of the present disclosure is an information processing method in which a computer acquires a plurality of first biometric information measured over time regarding a subject, accepts a specification of a prediction timing indicating a desired timing for prediction of second biometric information regarding the subject, the second biometric information being of a different type from the first biometric information and correlated with the first biometric information, and executes a process to predict the second biometric information at the prediction timing based on the plurality of first biometric information.

[0020] A fourteenth aspect of the present disclosure is an information processing program that causes a computer to acquire a plurality of first biometric information measured over time regarding a subject, accept a specification of a prediction timing indicating a desired timing for prediction of second biometric information regarding the subject, the second biometric information being of a different type from the first biometric information and correlated with the first biometric information, and execute a process of predicting the second biometric information at the prediction timing based on the plurality of first biometric information. [Effects of the Invention]

[0021] According to the above aspects, the information processing device, the information processing method, and the information processing program of the present disclosure can support appropriate diagnosis. [Brief explanation of the drawings]

[0022] [Figure 1] FIG. 1 is a schematic configuration diagram of an information processing system. [Figure 2] 1 is an example of first biometric information and second biometric information. [Figure 3] FIG. 2 is a block diagram illustrating an example of a hardware configuration of an information processing device. [Figure 4] FIG. 2 is a block diagram showing an example of a functional configuration of an information processing device according to a first exemplary embodiment. [Figure 5] 10 is an example of time-series data of first biological information and second biological information. [Figure 6] 10 is an example of correlation data between the first biological information and the second biological information. [Figure 7]10 is an example of correlation data between the first biological information and the second biological information. [Figure 8] 10 is an example of a screen displayed on a display. [Figure 9] 10 is a flowchart illustrating an example of a determination process. [Figure 10] FIG. 10 is a block diagram showing an example of a functional configuration of an information processing device according to a second exemplary embodiment. [Figure 11] 10 is an example of a condition predetermined for each pattern. [Figure 12] FIG. 2 is a block diagram illustrating an example of a functional configuration of a prediction unit. [Figure 13] 10 is an example of a screen displayed on a display. [Figure 14] 10 is a flowchart illustrating an example of a prediction process. [Figure 15] 10 is an example of a screen displayed on a display. DETAILED DESCRIPTION OF THE INVENTION

[0023] Hereinafter, examples of embodiments for carrying out the technology of the present disclosure will be described in detail with reference to the drawings.

[0024] [First Exemplary Embodiment] An example of the configuration of an information processing system 1 according to this exemplary embodiment will be described with reference to Fig. 1. As shown in Fig. 1, the information processing system 1 includes an information processing device 10, at least one first measuring device 11, and at least one second measuring device 12. The information processing device 10 and the first measuring device 11, and the information processing device 10 and the second measuring device 12, are capable of communicating with each other via wired or wireless communication.

[0025] The first measuring device 11 has a function of measuring first biological information of the user over time. The first biological information may be, for example, information indicating at least one of body temperature, heart rate, electrocardiogram, electromyogram, blood pressure, arterial oxygen saturation (SpO2), blood glucose level, and lipid level. In these cases, the first measuring device 11 may be, for example, a thermometer, a heart rate monitor, a self-monitoring blood glucose monitor, or a wearable device such as a smartwatch equipped with sensors that measure biological information such as heart rate and arterial oxygen saturation.

[0026] The second measurement device 12 has a function of measuring second biological information of the user. The second biological information is a different type of biological information from the first biological information, and is measured less frequently than the first biological information (i.e., the first biological information is measured more frequently than the second biological information).

[0027] The second biological information may be, for example, information indicating at least one of electrocardiograms, electroencephalograms, medical images captured by a medical imaging device, and the results of at least one of hematological tests, infectious disease tests, biochemical tests, and urinalysis. Examples of medical imaging devices include devices that perform CR (Computed Radiography), CT (Computed Tomography), MRI (Magnetic Resonance Imaging), ultrasound imaging, fundus photography, PET (Positron Emission Tomography), and PAI (PhotoAcoustic Imaging). Using these medical imaging devices as the second measuring device 12 allows for obtaining medical images as the second biological information.

[0028] A hematological test is a test that obtains, for example, white blood cell count, red blood cell count, hemoglobin concentration, etc. as test results. A biochemical test is a test that obtains, for example, various indicators related to enzymes, proteins, sugars, lipids, electrolytes, etc. as test results. An infectious disease test is a test that obtains, for example, the presence or absence of various infectious diseases such as influenza infection and COVID-19 infection as test results. A urinalysis is a test that obtains, for example, urinary sugar, urinary protein, urinary occult blood, etc. as test results. When these various test results are used as the second biological information, a known analyzer that analyzes blood, urine, etc. as test specimens can be used as the second measuring device 12.

[0029] The first biological information and the second biological information may each vary non-periodically according to the subject's behavior. Examples of the subject's behavior include eating, exercise, and sleep. For example, blood glucose levels, which are an example of the first biological information, are known to rise after the subject eats. Furthermore, for example, abnormal shadows are known to be prominent in fundus images, which are an example of the second biological information, when the subject experiences a postprandial hyperglycemic spike after eating.

[0030] In this exemplary embodiment, the first biological information and the second biological information are biological information that are known to be correlated with each other. Figure 2 shows an example of a set of first biological information and second biological information that are correlated with each other. Figure 2 also shows the "disease name" diagnosed based on the second biological information.

[0031] Incidentally, the accuracy of a diagnosis based on the second biological information is increased when the second biological information at the time of measurement is in a state suitable for diagnosis. For example, when diagnosing a subject suspected of having arrhythmia by capturing a cardiac image as the second biological information, the accuracy of the diagnosis is increased if a cardiac image can be obtained at the time the arrhythmia occurs. However, since it is unlikely that arrhythmia actually occurs at the exact timing of the imaging, there is a possibility that only a cardiac image in a better state than when the arrhythmia occurs, i.e., a cardiac image that is not suitable for diagnosis, can be obtained.

[0032] Furthermore, for example, when diagnosing a subject suspected of having high blood pressure by capturing cardiac images as second biological information, obtaining cardiac images when the subject's blood pressure is in a normal state would increase the accuracy of the diagnosis. However, it is known that some subjects suspected of having high blood pressure experience particularly high blood pressure while in a medical institution due to tension and stress (so-called "white coat hypertension"). For such subjects, even if cardiac images are taken under normal conditions, cardiac images obtained under conditions different from normal conditions because the subject is in a medical institution may result in an overdiagnosis.

[0033] As described above, even with the same second biological information, the state suitable for diagnosis may differ depending on the type of disease being diagnosed. Therefore, the information processing device 10 according to this exemplary embodiment assists in appropriate diagnosis by determining the state in which the second biological information was measured based on the first biological information correlated with the second biological information. The detailed configuration of the information processing device 10 will be described below.

[0034] First, an example of the hardware configuration of an information processing device 10 according to this exemplary embodiment will be described with reference to FIG. 3. As shown in FIG. 3, the information processing device 10 includes a CPU (Central Processing Unit) 21, a non-volatile storage unit 22, and a memory 23 serving as a temporary storage area. The information processing device 10 also includes a display 24 such as a liquid crystal display, an input unit 25 such as a keyboard, a mouse, and buttons, and a network I / F (Interface) 26 for wired or wireless communication with the first measurement device 11, the second measurement device 12, and an external network (not shown). The CPU 21, the storage unit 22, the memory 23, the display 24, the input unit 25, and the network I / F 26 are connected via a bus 28 such as a system bus and a control bus so that various information can be exchanged between them. Examples of the information processing device 10 include a personal computer, a server computer, a tablet device, a smartphone, and a wearable device.

[0035] The storage unit 22 is realized by a storage medium such as a hard disk drive (HDD), a solid state drive (SSD), or a flash memory. The storage unit 22 stores an information processing program 27 for the information processing device 10. The CPU 21 reads the information processing program 27 from the storage unit 22, loads it into the memory 23, and executes the loaded information processing program 27. The CPU 21 is an example of a processor of the present disclosure.

[0036] Next, an example of the functional configuration of the information processing device 10 according to this exemplary embodiment will be described with reference to Fig. 4. As shown in Fig. 4, the information processing device 10 includes an acquisition unit 30, a determination unit 32, and a control unit 36. The CPU 21 executes an information processing program 27, thereby functioning as the acquisition unit 30, the determination unit 32, and the control unit 36.

[0037] The acquiring unit 30 acquires, from the first measuring device 11, a plurality of pieces of first biological information that are first biological information measured on the subject and that are assigned the measurement dates and times of the first biological information. The acquiring unit 30 also acquires, from the second measuring device 12, a plurality of pieces of second biological information that are measured on the subject and are of a different type from the first biological information and correlated with the first biological information and that are assigned the measurement dates and times of the second biological information.

[0038] Note that when the biological information acquired from the first measuring device 11 and the second measuring device 12 is expressed as a numerical value (for example, body temperature, blood glucose level, white blood cell count, etc.), the acquiring unit 30 may acquire the numerical value as the first biological information and the second biological information as is. On the other hand, when the biological information acquired from the first measuring device 11 and the second measuring device 12 is not expressed as a numerical value (for example, an electrocardiogram and medical images, etc.), the acquiring unit 30 may extract feature amounts from the biological information and acquire the extracted feature amounts as the first biological information and the second biological information. As a feature extraction method, for example, a known feature extraction technique can be used, such as a method using a trained model that receives input biological information such as an electrocardiogram and medical images and outputs feature amounts.

[0039] The determination unit 32 determines the state in which the measured second biological information was measured, based on the first biological information and the second biological information acquired by the acquisition unit 30. Specifically, the determination unit 32 determines which of three patterns the measured second biological information falls into: a state suitable for diagnosis, a state better than the state suitable for diagnosis, or a state worse than the state suitable for diagnosis. A specific method of determination by the determination unit 32 will be described below.

[0040] Fig. 5 shows examples of time-series data of the first biological information and the second biological information for three cases C1 to C3, each of which has a different tendency of change in the first biological information. The time-series data in Fig. 5 is created based on a plurality of pieces of first biological information and a plurality of pieces of second biological information acquired by the acquisition unit 30 and the measurement dates and times assigned thereto. In Fig. 5, adjacent pieces of data in the chronological order are connected by straight lines. In the following description, it is assumed that the larger the first biological information and the second biological information, the worse the condition, and the smaller the first biological information and the second biological information, the better the condition.

[0041] In case C1 of Fig. 5, the first biological information fluctuates more slowly than in cases C2 and C3. In case C2, the first biological information fluctuates more sharply than in case C1, making it difficult to acquire the second biological information at the time when the first biological information fluctuates, which is the case seen in, for example, a subject with arrhythmia. In case C3, the first biological information fluctuates sharply when acquiring the second biological information, which is the case seen in, for example, a subject with white coat hypertension.

[0042] There is a correlation between first biometric information and second biometric information measured at the same time point, such that when the first biometric information is large, the second biometric information is also large, and when the first biometric information is small, the second biometric information is also small. FIG. 6 shows an example of correlation data in which the correlation between the first biometric information and the second biometric information is predetermined. FIG. 6 is a diagram in which the horizontal axis represents the first biometric information and the vertical axis represents the second biometric information, and the regression line RL, the upper allowable limit UL, and the lower allowable limit LL are generated in advance by performing a regression analysis based on the actual performance of the combination of the first biometric information and the second biometric information at the same time point. The regression line RL, the upper allowable limit UL, and the lower allowable limit LL in FIG. 6 are correlation data in which the correlation between the first biometric information and the second biometric information is predetermined. The correlation data is stored in advance in, for example, the storage unit 22.

[0043] For example, when the standard deviation of the regression line RL is σ, the upper limit UL and the lower limit LL may be defined as the regression line RL±σ, respectively. If it is assumed that the probability distribution of the combination of the first biometric information and the second biometric information follows a normal distribution, the combination of the first biometric information and the second biometric information falls between the lower limit LL and the upper limit UL with a probability of 34% above and 34% below the regression line RL (a total of 68%). In other words, it is estimated that 68% of the combinations of the first biometric information and the second biometric information at the same time fall between the lower limit LL and the upper limit UL. Note that while the correlation data shown in FIG. 6 is represented by a regression line, this is not limiting and the correlation data may be represented by a regression curve.

[0044] By using the regression line RL, the upper allowable limit UL, and the lower allowable limit LL, it is possible to classify the patterns according to the relationship between the first biometric information and the second biometric information. Specifically, when the combination of the first biometric information and the second biometric information is between the lower allowable limit LL and the upper allowable limit UL, it is called pattern P1, when it is below the lower allowable limit LL, it is called pattern P2, and when it is above the upper allowable limit UL, it is called pattern P3.

[0045] 5, since there is a correlation between the first biometric information and the second biometric information measured at the same time, it is difficult to classify the patterns based on the actual measurement values. Therefore, the determination unit 32 interpolates the second biometric information at time ta, when the first biometric information is measured but the second biometric information is not, based on the second biometric information before and after time ta (hereinafter, the interpolated second biometric information at time ta is referred to as the "interpolated value"). Furthermore, the determination unit 32 determines patterns P1 to P3 according to the relationship between the first biometric information at time ta and the interpolated value of the second biometric information at time ta.

[0046] The determination unit 32 may derive the interpolated value of the second biometric information at time ta, for example, by linear interpolation based on the second biometric information before and after time ta. That is, the determination unit 32 may derive a value on a line connecting two pieces of second biometric information before and after time ta as the interpolated value of the second biometric information at time ta. Alternatively, for example, the determination unit 32 may derive the interpolated value of the second biometric information at time ta based on an approximation curve based on multiple pieces of second biometric information before and after time ta.

[0047] Specifically, the determination unit 32 derives second biometric information correlated with the first biometric information at time ta (hereinafter referred to as the "theoretical value" of the second biometric information) by comparing the first biometric information at time ta with correlation data (regression line RL in FIG. 6). Then, the determination unit 32 determines whether the pattern is P1, P2, or P3 depending on whether the degree of match between the theoretical value of the second biometric information at time ta and the interpolated value of the second biometric information is within an allowable range.

[0048] The case where the degree of match between the theoretical value of the second biological information at time ta and the interpolated value of the second biological information is within the allowable range means that the interpolated value of the second biological information is between the allowable lower limit LL and the allowable upper limit UL in Figure 6, and this case is determined to be pattern P1. Case C1 in Figure 5 corresponds to pattern P1. Pattern P1 indicates that the interpolated value of the second biological information is also correlated with the first biological information, which means that the measured second biological information is likely to be valid and that the condition is likely to be suitable for diagnosis.

[0049] On the other hand, when the degree of match between the theoretical value of the second biometric information at time ta derived based on the correlation data and the interpolated value of the second biometric information is not within the allowable range, that is, when the interpolated value of the second biometric information is not between the allowable lower limit LL and the allowable upper limit UL in Fig. 6, the pattern is determined to be P2 or P3. In this case, the determination unit 32 may determine whether the pattern is P2 or P3 depending on the magnitude relationship between the theoretical value of the second biometric information at time ta derived based on the correlation data and the interpolated value of the second biometric information.

[0050] For example, the determination unit 32 may determine that the pattern is P2 when the interpolated value of the second biological information is below the lower allowable limit LL in Fig. 6. Case C2 in Fig. 5 corresponds to pattern P2. In the case of pattern P2, the interpolated value of the second biological information is smaller than the theoretical value of the second biological information, which means that the actually measured value of the second biological information may be smaller than the original value, that is, the measured second biological information may be in a better state than is suitable for diagnosis.

[0051] Furthermore, for example, the determination unit 32 may determine that the pattern is P3 when the interpolated value of the second biological information exceeds the upper limit of tolerance UL in Fig. 6. Case C3 in Fig. 5 corresponds to pattern P3. In the case of pattern P3, the interpolated value of the second biological information is larger than the theoretical value of the second biological information, which means that the actually measured value of the second biological information may be larger than the original value, that is, the measured second biological information may be in a worse state than is suitable for diagnosis.

[0052] The time point ta used for the determination can be any time point when the first biological information is measured but the second biological information is not, but it is preferably a time point when the first biological information indicates an abnormality. Examples of the time point when the first biological information indicates an abnormality include a time point when the first biological information exceeds a predetermined threshold value and a time point when the first biological information reaches a maximum value in a predetermined period.

[0053] On the other hand, there may be a case where the first biological information indicates an abnormality only when the second biological information is measured, as shown in case C3 in Fig. 5. In such a case, the determination unit 32 may use the time ta used for determination as the time when the first biological information is measured, the second biological information is not measured, and the first biological information indicates normal.

[0054] In order to improve the reliability of the determination, it is preferable that the determination unit 32 determine patterns P1 to P3 according to the relationship between the first biometric information and the interpolated value of the second biometric information at multiple time points when the first biometric information was measured but the second biometric information was not. FIG. 7 is a diagram showing combinations of the first biometric information and the interpolated value of the second biometric information at multiple time points for each of cases C1 to C3 of FIG. 5 on the correlation data of FIG. 6. As shown in cases C2 and C3 of FIG. 7, when the combinations of the first biometric information and the interpolated value of the second biometric information at multiple time points straddle different patterns, the determination unit 32 may prioritize patterns P2 and P3 over pattern P1. For example, the determination unit 32 may prioritize a pattern that applies to a combination of the first biometric information and the interpolated value of the second biometric information at more time points.

[0055] The control unit 36 ​​controls the display 24 to display the pattern determined by the determination unit 32 and guidance corresponding to the pattern. For example, when the determination unit 32 determines pattern P2, the control unit 36 ​​may provide guidance that the measured second biological information is in a better state than is appropriate for diagnosis, i.e., that a diagnosis should be made taking into consideration the possibility that the actual state is worse than the measured second biological information. Alternatively, when the determination unit 32 determines pattern P3, the control unit 36 ​​may provide guidance that the measured second biological information is in a worse state than is appropriate for diagnosis, i.e., that a diagnosis should be made taking into consideration the possibility that the actual state is better than the measured second biological information.

[0056] Fig. 8 shows an example of a screen D1 displayed on the display 24 by the control unit 36. The screen D1 in Fig. 8 corresponds to case C2 in Fig. 5, and is a screen displayed when the determination unit 32 determines that the condition is pattern P2. In Fig. 8, the control unit 36 ​​advises that a diagnosis should be made taking into consideration the possibility that the actual condition is worse than the measured second biological information.

[0057] Next, the operation of the information processing device 10 according to this exemplary embodiment will be described with reference to Fig. 9. In the information processing device 10, the CPU 21 executes the information processing program 27, thereby executing the determination process shown in Fig. 9. The determination process is executed, for example, when a user issues an instruction to start execution via the input unit 25.

[0058] In step S10, the acquisition unit 30 acquires a plurality of pieces of first biological information from the first measurement device 11 and acquires a plurality of pieces of second biological information from the second measurement device 12. In step S12, the determination unit 32 interpolates the second biological information at time ta when the first biological information is measured but the second biological information is not measured, based on the second biological information before and after time ta acquired in step S10.

[0059] In step S14, the determination unit 32 determines one of patterns P1 to P3 according to the relationship between the first biometric information at time ta and the interpolated value of the second biometric information at time ta interpolated in step S12. In step S16, the control unit 36 ​​controls the display 24 to display a guide corresponding to the pattern determined in step S14, and then ends this determination process.

[0060] As described above, the information processing device 10 includes at least one processor. The processor acquires a plurality of first biological information measured on a subject, the first biological information being assigned with the measurement date and time of the first biological information, and a plurality of second biological information measured on the subject, the second biological information being of a different type from the first biological information and correlated with the first biological information, the second biological information being assigned with the measurement date and time of the second biological information. The processor also interpolates the second biological information at a time when the first biological information was measured but the second biological information was not measured, based on the second biological information measured before and after the time, and determines a pattern corresponding to the relationship between the first biological information at the time and the interpolated second biological information. In other words, the information processing device 10 can assist in appropriate diagnosis by determining the conditions under which the second biological information was measured.

[0061] In the first exemplary embodiment, the pattern is determined based on two types of biometric information, the first biometric information and the second biometric information, but the present invention is not limited to this. For example, the acquisition unit 30 may acquire three or more types of biometric information, and the determination unit 32 may determine the pattern based on the three or more types of biometric information.

[0062] Further, for example, the determination unit 32 may have the user select any two types of biometric information from the three or more types of biometric information, and determine the pattern based on the selected two types of biometric information. Further, for example, the determination unit 32 may have the user select the type of disease to be diagnosed, and determine the pattern based on two types of biometric information predetermined for each type of disease from the three or more types of biometric information.

[0063] Second Exemplary Embodiment In the first exemplary embodiment, the state under which the second biological information was measured is determined based on the first biological information. When the measured second biological information is not suitable for diagnosis, it is desirable to support the diagnosis by predicting and presenting second biological information suitable for diagnosis. Therefore, the information processing device 10 according to this exemplary embodiment has a function to predict second biological information suitable for diagnosis in addition to the functions of the first exemplary embodiment. Hereinafter, an example of the functional configuration of the information processing device 10 according to this exemplary embodiment will be described, but some of the configuration similar to that of the first exemplary embodiment will not be described.

[0064] The acquiring unit 30 acquires, from the first measuring device 11, a plurality of pieces of first biological information measured over time on the subject, each piece having the measurement date and time of the first biological information attached thereto. The acquiring unit 30 also acquires, from the second measuring device 12, at least one piece of second biological information measured on the subject, which is of a different type from the first biological information and correlated with the first biological information. The first biological information and the second biological information are the same as those in the first exemplary embodiment, and therefore description thereof will be omitted.

[0065] As described in the first exemplary embodiment above, the determination unit 32 determines whether the measured second biological information corresponds to a state suitable for diagnosis (pattern P1), a state better than a state suitable for diagnosis (pattern P2), or a state worse than a state suitable for diagnosis (pattern P3).

[0066] The prediction unit 34 receives a designation of a prediction timing indicating a timing at which prediction is desired for the second biological information. Specifically, the prediction unit 34 designates a predetermined condition for each pattern determined by the determination unit 32 as a condition to be satisfied by the first biological information, and designates a date and time at which the first biological information satisfies the condition as a prediction timing. FIG. 11 shows an example of the predetermined conditions to be satisfied by the first biological information for each of patterns P1 to P3. Note that, since the prediction of the second biological information is performed based on first biological information that has already been measured (details will be described later), the prediction timing is a time in the past.

[0067] Furthermore, the prediction unit 34 predicts the second biological information at a specified prediction timing based on the plurality of pieces of first biological information and at least one piece of second biological information acquired by the acquisition unit 30. Hereinafter, a method for predicting the second biological information will be described with reference to Fig. 12. Fig. 12 is a block diagram showing an example of the functional configuration of the prediction unit 34. In Fig. 12, circled numbers indicating the order of explanation are given for ease of understanding, but these circled numbers do not necessarily indicate the order of processing.

[0068] First, the prediction unit 34 uses an encoder 40A that extracts features from the first biometric information to extract features from each of the plurality of pieces of first biometric information acquired by the acquisition unit 30 (1 in FIG. 12). As the encoder 40A, for example, a trained model that has been trained in advance to use the first biometric information as input and the features as output can be applied. As such a training model, a convolutional neural network (CNN) and a residual network (ResNet) can be applied, or ensemble training that integrates a plurality of training models can be performed.

[0069] Next, the prediction unit 34 predicts the feature quantities of the first biometric information at the prediction timing using a prediction model 42A that generates feature quantities of the first biometric information at any time based on the feature quantities of the multiple pieces of first biometric information (2 in FIG. 12). As the prediction model 42A, it is preferable to apply a model that can make predictions by taking into account temporal changes (time-series data) of the first biometric information indicated by the multiple pieces of first biometric information, such as a recurrent neural network (RNN) or a long short-term memory (LSTM). In particular, applying LSTM can reflect long-term changes in the first biometric information, which can contribute to appropriate diagnosis.

[0070] Next, the prediction unit 34 uses an encoder 40B that extracts features from the second biometric information to extract features for at least one piece of second biometric information acquired by the acquisition unit 30 (3 in FIG. 12). As the encoder 40B, for example, a trained model that has been trained in advance to use the second biometric information as input and the features as output can be applied. As such a training model, CNN, ResNet, etc. can be applied, or ensemble training that integrates multiple training models can be performed.

[0071] Next, the prediction unit 34 predicts the feature amount of the second biometric information at the prediction timing using a prediction model 42B that generates the feature amount of the second biometric information at any time point based on the feature amount of the second biometric information (4 in FIG. 12). As the prediction model 42B, it is preferable to apply a model that can make predictions by taking into account temporal changes (time-series data) of the second biometric information indicated by a plurality of pieces of second biometric information, such as an RNN or LSTM. In particular, applying LSTM can reflect long-term changes in the second biometric information, which can contribute to appropriate diagnosis.

[0072] Next, the prediction unit 34 corrects the feature amount of the second biometric information at the prediction timing predicted by the prediction model 42B based on the feature amount of the first biometric information at the prediction timing predicted by the prediction model 42A (5 in FIG. 12). That is, the prediction unit 34 predicts the second biometric information by taking into account the temporal change in the first biometric information indicated by the multiple pieces of first biometric information. The correction of the feature amount of the second biometric information at the prediction timing can be performed, for example, by multimodal deep learning using the feature amount of the first biometric information and the feature amount of the second biometric information.

[0073] Next, the prediction unit 34 restores the first biometric information at the prediction timing from the feature amount of the first biometric information at the prediction timing predicted by the prediction model 42A, using a decoder 44A that restores the first biometric information from the feature amount of the first biometric information (6 in FIG. 12). That is, the prediction unit 34 can also predict the first biometric information at the prediction timing. Note that if the first biometric information is acquired at the same timing as the prediction timing, the prediction unit 34 may evaluate the prediction accuracy by comparing the acquired first biometric information with the predicted first biometric information, and present the reliability of the prediction.

[0074] Furthermore, the prediction unit 34 restores the second biometric information at the prediction timing from the feature amount of the second biometric information at the prediction timing predicted by the prediction model 42B, using a decoder 44B that restores the second biometric information from the feature amount of the second biometric information (7 in FIG. 12). Through the above process, the prediction unit 34 predicts the first biometric information and the second biometric information at the prediction timing.

[0075] The control unit 36 ​​controls the display 24 to display the second biological information at the predicted timing predicted by the prediction unit 34. FIG. 13 shows an example of a screen D2 displayed on the display 24 by the control unit 36. In FIG. 13, the control unit 36 ​​displays the second biological information (cardiac image) at the predicted timing (March 18, 2021) predicted by the prediction unit 34. Note that the control unit 36 ​​may also control the display 24 to display the first biological information at the predicted timing predicted by the prediction unit 34 and an evaluation result of the prediction accuracy based on a comparison between the acquired first biological information and the predicted first biological information.

[0076] Next, the operation of the information processing device 10 according to this exemplary embodiment will be described with reference to Fig. 14. In the information processing device 10, the CPU 21 executes the information processing program 27, thereby executing the prediction process shown in Fig. 14. The prediction process is executed, for example, when a user issues an instruction to start the execution via the input unit 25.

[0077] In step S50, the acquisition unit 30 acquires multiple pieces of first biological information from the first measurement device 11 and acquires at least one piece of second biological information from the second measurement device 12. In step S52, the prediction unit 34 receives a designation of a prediction timing indicating a desired timing for prediction of the second biological information.

[0078] In step S54, the prediction unit 34 predicts the second biological information at the predicted timing received in step S52, based on the plurality of pieces of first biological information and at least one piece of second biological information acquired in step S50. In step S56, the control unit 36 ​​controls the display 24 to display the second biological information at the predicted timing predicted in step S54, and ends this prediction process.

[0079] As described above, the information processing device 10 includes at least one processor, which acquires a plurality of first biological information measured over time on a subject, accepts designation of a prediction timing indicating a desired timing for prediction of second biological information on the subject, the second biological information being of a different type from the first biological information and correlated with the first biological information, and predicts the second biological information at the prediction timing based on the plurality of first biological information. In other words, the second biological information can be predicted and presented at a timing suitable for diagnosis, thereby supporting appropriate diagnosis.

[0080] In the second exemplary embodiment, the acquiring unit 30 acquires the first biometric information and the second biometric information, and the predicting unit 34 predicts the second biometric information based on the first biometric information and the second biometric information. However, the present invention is not limited to this. The prediction of the second biometric information may be based at least on the first biometric information. For example, the predicting unit 34 may predict the first biometric information at the prediction timing and compare the predicted first biometric information with correlation data between the first biometric information and the second biometric information to predict the second biometric information at the prediction timing. In this case, the actual measurement value of the second biometric information is not used to predict the second biometric information, so the acquiring unit 30 does not need to acquire the second biometric information.

[0081] In the second exemplary embodiment, the prediction unit 34 predicts the feature amount of the first biometric information at the prediction timing and uses the predicted feature amount of the first biometric information to correct the prediction of the second biometric information. However, the present invention is not limited to this. For example, if an actual measurement value of the first biometric information is measured at the prediction timing, the prediction of the first biometric information at the prediction timing may be omitted, and the actual measurement value may be used to correct the prediction of the second biometric information.

[0082] In the second exemplary embodiment, the prediction timing is the date and time when the first biometric information satisfies a predetermined condition for each pattern determined by the determination unit 32. However, this is not limiting. For example, the prediction unit 34 may accept a user's specification of a condition to be satisfied by the first biometric information and specify the date and time when the first biometric information satisfies the condition as the prediction timing. FIG. 15 shows an example of a screen D3 that accepts a user's specification of a condition to be satisfied by the first biometric information. As shown in FIG. 15, the condition to be satisfied by the first biometric information may be specified by, for example, the control unit 36 ​​displaying screen D3 on the display 24 and accepting the user's specification via the input unit 25. As shown in FIG. 15, the prediction unit 34 may directly accept, as the prediction timing, a specification of a date and time at which the second biometric information is desired to be predicted. In these cases, the information processing device 10 does not need to have the functionality of the determination unit 32.

[0083] The configuration of the information processing system 1 in each of the exemplary embodiments is not limited to the example shown in FIG. 1 . For example, some or all of the information processing device 10, the first measuring device 11, and the second measuring device 12 included in the information processing system 1 may be the same device. For example, the information processing system 1 may include a plurality of first measuring devices 11 and / or a plurality of second measuring devices. The plurality of first measuring devices 11 may each measure the same type of first biological information, or may each measure different types of first biological information. Similarly, the plurality of second measuring devices 12 may each measure the same type of second biological information, or may each measure different types of second biological information.

[0084] Furthermore, in each of the exemplary embodiments described above, the following various processors can be used as the hardware structure of processing units that perform various processes, such as the acquisition unit 30, the determination unit 32, the prediction unit 34, and the control unit 36. As described above, the various processors include a CPU, which is a general-purpose processor that executes software (programs) and functions as various processing units, as well as dedicated electrical circuits that are processors having a circuit configuration specifically designed to perform specific processes, such as a programmable logic device (PLD), a processor whose circuit configuration can be changed after manufacture, such as an FPGA (Field Programmable Gate Array), and an ASIC (Application Specific Integrated Circuit).

[0085] A single processing unit may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, multiple processing units may be configured with a single processor.

[0086] Examples of configuring multiple processing units with a single processor include, first, a form in which one processor is configured with a combination of one or more CPUs and software, and this processor functions as multiple processing units, as typified by computers such as client and server. Second, a form in which a processor is used to realize the functions of the entire system including multiple processing units with a single IC (Integrated Circuit) chip, as typified by systems on chips (SoCs). In this way, various processing units are configured using one or more of the above-mentioned various processors as a hardware structure.

[0087] Furthermore, more specifically, the hardware structure of these various processors can be an electric circuit that combines circuit elements such as semiconductor elements.

[0088] In addition, in the above exemplary embodiments, the information processing program 27 is described as being pre-stored (installed) in the storage unit 22, but this is not limiting. The information processing program 27 may be provided in a form recorded on a recording medium such as a CD-ROM (Compact Disc Read Only Memory), a DVD-ROM (Digital Versatile Disc Read Only Memory), or a USB (Universal Serial Bus) memory. The information processing program 27 may also be downloaded from an external device via a network. Furthermore, the technology of the present disclosure extends to not only information processing programs but also storage media that non-temporarily store information processing programs.

[0089] The technology of the present disclosure can also be achieved by appropriately combining the above-described exemplary embodiments. The above-described description and illustrations are detailed descriptions of the parts related to the technology of the present disclosure and are merely examples of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is a description of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or new elements may be substituted from the above-described description and illustrations, within the scope of the gist of the technology of the present disclosure.

[0090] The disclosure of Japanese Patent Application No. 2021-077887, filed on April 30, 2021, is incorporated herein by reference in its entirety. All documents, patent applications, and technical standards described herein are incorporated herein by reference to the same extent as if each individual document, patent application, and technical standard was specifically and individually indicated to be incorporated by reference.

Claims

1. at least one processor; The processor: acquiring a plurality of first biological information measured over time on the subject; receiving a designation of a prediction timing indicating a timing at which prediction is desired for second biological information related to the subject, the second biological information being of a different type from the first biological information and correlated with the first biological information; predicting the first biological information at the predicted timing based on the plurality of pieces of first biological information; The second biological information at the predicted timing is predicted based on the first biological information at the predicted timing. Information processing device.

2. a measurement date and time of the first biological information is assigned to each of the plurality of first biological information; The processor: The second biological information is predicted taking into consideration a temporal change in the first biological information indicated by the plurality of pieces of first biological information. The information processing device according to claim 1 .

3. The processor: The second biological information is predicted using a recurrent neural network (RNN) or a long short-term memory (LSTM). The information processing device according to claim 2 .

4. The processor: acquiring at least one second biological information measured on the subject; The second biological information is predicted based on the acquired first biological information and the acquired second biological information. The information processing device according to claim 1 .

5. The processor: As the prediction timing, a date and time at which the second biological information is desired to be predicted is accepted. The information processing device according to claim 1 .

6. The processor: Accepting a designation of a condition that the first biometric information should satisfy; The date and time when the first biological information satisfies the condition is specified as the predicted timing. The information processing device according to claim 1 .

7. a measurement date and time of the first biological information is assigned to each of the plurality of first biological information; The processor: acquiring a plurality of pieces of second biological information measured on the subject, the pieces of second biological information being assigned measurement dates and times; interpolating the second biological information at a time when the first biological information is measured but the second biological information is not measured based on the second biological information before and after the time; determining a pattern according to a relationship between the first biometric information at the time point and the interpolated second biometric information; A predetermined condition for each pattern is designated as a condition that the first biometric information must satisfy. The information processing device according to claim 6 .

8. The predicted timing is in the past than the present time. The information processing device according to claim 1 .

9. The first biological information is measured more frequently than the second biological information. The information processing device according to claim 1 .

10. The first biological information and the second biological information vary non-periodically in accordance with the behavior of the subject. The information processing device according to claim 1 .

11. the first biological information indicates at least one of body temperature, heart rate, electrocardiogram, electromyogram, blood pressure, arterial blood oxygen saturation, blood glucose level, and lipid level; The second biological information indicates at least one of electrocardiogram, electroencephalogram, medical image taken by a medical imaging device, and the result of at least one of hematological test, infectious disease test, biochemical test, and urinalysis. The information processing device according to claim 1 .

12. acquiring a plurality of first biological information measured over time on the subject; receiving a designation of a prediction timing indicating a timing at which prediction is desired for second biological information related to the subject, the second biological information being of a different type from the first biological information and correlated with the first biological information; predicting the first biological information at the predicted timing based on the plurality of pieces of first biological information; The second biological information at the predicted timing is predicted based on the first biological information at the predicted timing. An information processing method in which processing is performed by a computer.

13. acquiring a plurality of first biological information measured over time on the subject; receiving a designation of a prediction timing indicating a timing at which prediction is desired for second biological information related to the subject, the second biological information being of a different type from the first biological information and correlated with the first biological information; predicting the first biological information at the predicted timing based on the plurality of pieces of first biological information; The second biological information at the predicted timing is predicted based on the first biological information at the predicted timing. An information processing program that causes a computer to execute a process.

Images