HK1225941B - Estimation device, recording medium, estimation system - Google Patents

Description

Estimation device, recording medium, and estimation system

Technical Field

The present invention relates to an estimating device, a program, an estimating method, and an estimating system for estimating a disease condition of a subject.

Background Art

In the past, the following technology has been proposed: measuring the subject's sound signals or electrical signals representing the activity of organs such as the heart, obtaining the subject's emotions and organ activities based on the measured signals, and inferring the subject's condition based on the time changes of the obtained emotions and organ activities (for example, refer to cited documents 1 and 2).

Prior art literature

Patent Literature

Patent Document 1: International Publication No. 2006/132159

Patent Document 2: Japanese Patent Application Laid-Open No. 2012-61057

Summary of the Invention

Problems to be solved by the invention

However, conventionally, estimating the condition of a subject based on temporal changes in the subject's emotions or organ activity obtained from measured signals requires the user to have specialized medical knowledge.

On the one hand, the estimation device, program, estimation method, and estimation system of the present application are intended to easily estimate the condition of a subject even without specialized knowledge.

Means for solving problems

An estimation device based on one viewpoint comprises: an extraction unit that extracts, from information representing the subject's physiology, first information representing the subject's physiological state and second information representing at least one of the subject's emotions and organ activities; a calculation unit that determines the similarity between the time changes shown by the extracted first information and the second information, and calculates the deviation amount from a prescribed state for maintaining homeostasis of the subject based on the determined similarity; and an estimation unit that estimates the subject's condition based on the calculated deviation amount.

A program based on other viewpoints causes a computer to perform the following processing: extracting, from information representing the subject's physiology, first information representing the subject's physiological state and second information representing at least one of the subject's emotions and organ activities; obtaining the similarity between the time changes shown by the extracted first information and the second information, and calculating the deviation from the subject's prescribed state for maintaining homeostasis based on the obtained similarity; and inferring the subject's condition based on the calculated deviation.

In an inference method based on another viewpoint, an extraction unit extracts first information representing the physiological state of the subject and second information representing at least one of the subject's emotions and organ activities from information representing the subject's physiology; an operation unit calculates the similarity between the time changes shown by the extracted first information and the second information, and calculates the deviation amount from the subject's prescribed state for maintaining homeostasis based on the calculated similarity; and an inference unit infers the subject's condition based on the calculated deviation amount.

An estimation system based on another viewpoint comprises: a measuring device for measuring the physiology of a subject; an estimation device for estimating the subject's condition using information indicating the subject's physiology measured by the measuring device; and an output device for outputting the result of the condition estimated by the estimation device, the estimation device comprising: an extraction unit for extracting, from the information indicating the subject's physiology, first information indicating the subject's physiological state and second information indicating at least one of the subject's emotions and organ activity; a calculation unit for determining the similarity between time changes indicated by the extracted first information and the second information, and calculating a deviation from a prescribed state for maintaining homeostasis of the subject based on the determined similarity; and an estimation unit for estimating the subject's condition based on the calculated deviation.

The estimation device, program, estimation method, and estimation system of the present application can easily estimate the condition of a subject even without professional knowledge.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a diagram showing one embodiment of an estimation device.

FIG2 is a diagram showing another embodiment of the estimation device.

FIG. 3 is a diagram showing an example of a decision tree showing the relationship between the fundamental frequency of an utterance of a subject and the subject's emotion.

FIG. 4 is a diagram showing an example of temporal changes in the intensity of the physician's excitement level, normal, sad, angry, and happy.

FIG. 5 is a diagram showing an example of temporal changes in the intensity of excitement, normality, sadness, anger, and joy in a patient with depression.

FIG. 6 is a diagram showing an example of temporal changes in the intensity of excitement, normality, sadness, anger, and joy of ordinary people.

FIG. 7 is a diagram showing an example of temporal changes in the intensity of excitement, normality, sadness, anger, and joy of ordinary people different from the ordinary people shown in FIG. 6 .

FIG. 8 is a diagram showing an example of the result of cross-correlation processing between the excitement level of the physician shown in FIG. 4 and each emotion, performed by the calculation unit shown in FIG. 2 .

FIG. 9 is a diagram showing an example of the result of cross-correlation processing between the excitement level and each emotion of the depression patient shown in FIG. 5 , performed by the calculation unit shown in FIG. 2 .

FIG. 10 is a diagram showing an example of the result of cross-correlation processing between the excitement level of the average person shown in FIG. 6 and each emotion, performed by the calculation unit shown in FIG. 2 .

FIG. 11 is a diagram showing an example of the result of cross-correlation processing between the excitement level of the average person shown in FIG. 7 and each emotion, performed by the calculation unit shown in FIG. 2 .

FIG. 12 is a diagram showing an example of the emotional homeostasis of a subject.

FIG. 13 is a diagram showing an example of a temporal change in the amount of deviation in the homeostasis of the physician shown in FIG. 8 , obtained by the calculation unit shown in FIG. 2 .

FIG. 14 is a diagram showing an example of a temporal change in the amount of deviation from the homeostasis of the depression patient shown in FIG. 9 , obtained by the calculation unit shown in FIG. 2 .

FIG. 15 is a diagram showing an example of temporal changes in the amount of deviation in homeostasis of the average person shown in FIG. 10 , obtained by the calculation unit shown in FIG. 2 .

FIG. 16 is a diagram showing an example of a temporal change in the amount of deviation in homeostasis of the average person shown in FIG. 11 , obtained by the calculation unit shown in FIG. 2 .

FIG. 17 is a diagram showing an example of estimation processing performed by the estimation device shown in FIG. 2 .

FIG. 18 is a diagram showing an example of a decision tree for a subject's heart rate and heartbeat fluctuations and the subject's emotions.

FIG. 19 is a diagram showing another example of the emotional homeostasis of a subject.

FIG20 is a diagram showing another embodiment of the estimation device.

FIG. 21 is a diagram schematically showing an example of the linkage of PA homeostasis in a subject.

FIG. 22 is a diagram showing an example of a calculation model of the circulatory system used in the simulation of homeostasis of a subject in the experimental section shown in FIG. 20 .

FIG. 23 is a diagram showing an example of data on displacement of each circulatory system of a subject.

FIG. 24 is a diagram showing an example of an estimation process performed by the estimation device shown in FIG. 20 .

FIG. 25 is a diagram showing another embodiment of the estimation device.

FIG. 26 is a diagram showing an example of a calculation model of the circulatory system used by the experimental section shown in FIG. 25 to simulate the homeostasis of a subject.

FIG. 27 is a diagram showing an example of data on the rotational speed of each circulatory system of a test subject.

FIG. 28 is a diagram showing an example of a medical condition table.

FIG. 29 is a diagram showing an example of estimation processing performed by the estimation device shown in FIG. 25 .

FIG30 is a diagram showing another embodiment of the estimation device.

FIG. 31 is a diagram showing an example of an utterance table.

FIG32 is a diagram showing an example of estimation processing performed by the estimation device shown in FIG30.

DETAILED DESCRIPTION

Hereinafter, embodiments will be described using the drawings.

FIG1 shows an embodiment of an estimation device.

The estimation device AM shown in Figure 1 is a computer device, for example, having a processing unit such as a CPU (Central Processing Unit) and a storage device such as a hard disk drive. The estimation device AM includes an extraction unit EU, a calculation unit CU, and an estimation unit AU. The functions of the extraction unit EU, calculation unit CU, and estimation unit AU can be implemented by a program executed by the CPU or by hardware.

The extraction unit EU extracts first information representing the subject's physiological state, such as the pitch frequency and fundamental frequency of the sound, or the body temperature and heart rate, from information representing the subject's physiological state, including sound data uttered by the subject, or data on body temperature and heartbeat, stored in the storage device of the estimation device AM. Furthermore, the extraction unit EU extracts second information representing at least one of the subject's emotions, such as anger and sadness, and the activity of the subject's organs, such as the heart and intestines, from the information representing the subject's physiological state.

The calculation unit CU obtains the similarity between the extracted first information and the time changes indicated by the second information, and calculates the deviation of the subject from a predetermined state for maintaining homeostasis (hereinafter also referred to as the deviation from homeostasis) based on the obtained similarity.

The estimating unit AU estimates the condition of the subject based on the calculated deviation from homeostasis. The estimating device AM then outputs information indicating the condition estimated by the estimating unit AU to an external display such as an organic EL (Organic Electro-Luminescence) or liquid crystal.

In the embodiment shown in FIG1 , the deviation from the subject's homeostasis is calculated using first information representing the subject's physiological state and second information representing at least one of the subject's emotions and organ activity. Thus, the estimation device AM can easily estimate the subject's condition even without medical expertise by referring to a single indicator, the deviation from homeostasis.

FIG2 shows another embodiment of the estimation device.

The estimation device 100 shown in FIG2 is a computer device having a processing unit such as a CPU and a storage device such as a hard disk drive. The estimation device 100 is connected to the measurement device 1 and the output device 2 via a wired or wireless connection via an interface unit included in the estimation device 100. Thus, for example, the estimation device 100, the measurement device 1, and the output device 2 operate as an estimation system SYS.

The measuring device 1 includes at least a microphone, for example, and measures physiological information representing the subject PA. For example, the measuring device 1 measures a speech signal of the subject PA via the microphone and outputs the measured speech signal to the estimating device 100 as physiological information representing the subject PA.

Output device 2 includes, for example, an organic EL display or a liquid crystal display. Output device 2 receives the estimated condition of subject PA from estimation device 100 and displays the received estimated condition on the organic EL display. Alternatively, output device 2 may be provided within estimation device 100.

2 includes an extraction unit 10, a calculation unit 20, and an estimation unit 30. The functions of the extraction unit 10, the calculation unit 20, and the estimation unit 30 may be implemented by a program executed by a CPU or by hardware.

Extraction unit 10 extracts first information representing subject PA's physiological state and second information representing at least one of subject PA's emotions and the activity of organs such as the heart and intestines from physiological information representing subject PA measured by measurement device 1. Extraction unit 10 outputs the extracted first and second information to calculation unit 20. The operation of extraction unit 10 is described with reference to Figures 3 to 7.

The calculation unit 20 calculates the similarity between the temporal variations of the extracted first and second information. For example, the calculation unit 20 performs cross-correlation processing on the temporal variations of the extracted first and second information and calculates the cross-correlation coefficient as the similarity. The calculation unit 20 uses the multiple calculated similarities to determine the deviation from the homeostasis of the subject's PA. The operation of the calculation unit 20 and the homeostasis are explained with reference to Figures 8 to 12.

The estimating unit 30 estimates the condition of the subject PA based on the obtained deviation from the homeostasis of the subject PA. The estimating unit 30 outputs information indicating the estimated condition of the subject PA to the output device 2. The operation of the estimating unit 30 is described with reference to FIG. 12 to FIG. 16 .

FIG3 shows an example of a decision tree representing the relationship between the fundamental frequency of a subject PA's speech and the subject PA's emotion. The decision tree shown in FIG3 is generated, for example, based on each speech of multiple subjects PA (e.g., more than 100 subjects PA), based on the emotion of each subject PA, which is subjectively evaluated as "normal," "sad," "angry," or "joyful," and the height of the extracted fundamental frequency. That is, the decision tree shown in FIG3 represents the relationship between each emotion of normal, sad, angry, and joyful, and the height, intensity, and average intensity of the fundamental frequency at the time of speech. For example, a normal emotion refers to a fundamental frequency height of less than 150 Hz and a fundamental frequency intensity of 100 or more. A sad emotion refers to a fundamental frequency height of less than 150 Hz and a fundamental frequency intensity of less than 100. An angry emotion refers to a fundamental frequency height of more than 150 Hz and an average fundamental frequency intensity of 80 or more. A joyful emotion refers to a fundamental frequency height of more than 150 Hz and a fundamental frequency intensity of less than 80.

The decision tree shown in FIG3 is previously stored in the storage device of estimation device 100. Furthermore, while the decision tree shown in FIG3 shows normal, sad, angry, and happy emotions for subject PA, other emotions such as anxiety and pain may also be included. Furthermore, estimation device 100 may also include a decision tree that represents the relationship between voice parameters such as pitch and frequency and emotions.

For example, the extraction unit 10 performs frequency analysis such as an FFT (Fast Fourier Transform) on the audio signal of the speech of the subject PA received from the measurement device 1 to determine the height of the fundamental frequency, etc. Based on the height of the fundamental frequency, etc. determined from each speech of the subject PA and the decision tree shown in FIG3 , the extraction unit 10 determines the ratio of the emotions of normal, sad, angry, and happy that appeared in the subject PA at the moment of each speech, for example, within a range of values from 0 to 10. The total value of the ratio of the emotions of normal, sad, angry, and happy is a fixed value, for example, set to 10. Furthermore, the ratios of normal, sad, angry, and happy may also be values in a range other than the range of 0 to 10.

The extraction unit 10 also obtains intonation, pitch frequency, and other information from the voice signal of subject PA. For example, the extraction unit 10 detects regions of identical frequency components from the intensity variation pattern within each utterance of the voice signal and obtains the time intervals during which these regions of identical frequency components appear as the intonation. Furthermore, the extraction unit 10 obtains a frequency spectrum, for example, based on frequency analysis of the voice signal. The extraction unit 10 performs autocorrelation processing while shifting the obtained frequency spectrum on the frequency axis to obtain the waveform of the autocorrelation coefficient. The extraction unit 10 obtains the pitch frequency based on the peak-to-peak or trough-to-trough intervals in the obtained autocorrelation coefficient waveform. The extraction unit 10 then compares the obtained intonation and pitch frequency with predetermined intervals and frequencies to determine the degree of excitement of subject PA (hereinafter also referred to as excitement) within a range of 0 to 10. The shorter the intervals between identical frequency components indicated by the intonation are compared to the predetermined interval, or the higher the pitch frequency is compared to the predetermined frequency, the higher the excitement. In other words, since the physiological excitement of the subject PA is closely related to the activity of the subject PA's cranial nerves, such as the sympathetic and parasympathetic nerves, the relationship between the activity of the subject PA's cranial nerves and the subject PA's emotions can be studied using the excitement level. Furthermore, the excitement level may also be a value in a range other than the range of 0 to 10.

The extraction unit 10 multiplies the obtained excitement by the ratio of normal, sad, angry, and happy to obtain the intensity of each of normal, sad, angry, and happy. The excitement is an example of the first information, and the intensity of normal, sad, angry, and happy is an example of the second information.

Figures 4 to 7 show an example of the temporal changes in the intensity of excitement, normal, sad, angry, and joy for each subject PA. The horizontal axis of Figures 4 to 7 represents the order of the subject PA's speech units as a time axis, and the vertical axis of Figures 4 to 7 represents the excitement and the intensity of normal, sad, angry, and joy. The solid line represents the temporal changes in excitement, and the dotted line represents the temporal changes in the sum of the intensities of normal and angry (hereinafter referred to as "normal plus angry"). In addition, the dotted line represents the temporal changes in the intensity of sadness, and the dashed line represents the temporal changes in the intensity of joy.

4 to 7 show temporal changes in the intensity of excitement, normal plus anger, sadness, and joy, for example, values obtained by performing a moving average with a window width of 10 utterance units by the calculation unit 20 .

The reason for combining the intensity of normal and angry emotions is that the inventors assumed that characteristic changes in sadness and joy occurred when estimating whether the subject PA had a mental illness, and treated normal and angry emotions as separate emotions. Furthermore, normal and angry emotions can also be studied separately, similar to sadness and joy.

Figure 4 shows the temporal changes in the intensity of excitement, calmness, sadness, anger, and joy for subject PA, a healthy psychiatrist without a mental illness, while examining a patient with depression. As shown in Figure 4, subject PA's excitement fluctuated between 1 and 3.5 during the duration of the patient's speech. Furthermore, the intensity of calmness plus anger was greater than sadness and joy during the patient's speech. Because the patient was examining a patient with depression, the intensity of joy was generally lower than sadness.

Figure 5 shows the temporal changes in the intensity of excitement, calmness, sadness, anger, and joy for subject PA, a patient with depression, while being examined by the physician shown in Figure 4. As shown in Figure 5, the excitement of subject PA, a patient with depression, fluctuated between 2 and 5 during her speech, showing values higher than the excitement of the physician shown in Figure 4. Furthermore, during her speech, the intensity of calmness plus anger was greater than that of sadness and joy, and the intensity of sadness was greater than that of joy. Furthermore, the intensities of sadness and joy for the patient with depression were greater than those for the physician shown in Figure 4.

FIG6 shows the temporal changes in the intensity of excitement, normal, sad, angry, and joy when the subject PA is a healthy ordinary person A who does not suffer from a mental illness. As shown in FIG6 , the excitement of ordinary person A as subject PA shows a fluctuation in the range of 1.5 to 4.5. In addition, the emotions of ordinary person A, similar to the case of the physician shown in FIG4 and the case of the depression patient shown in FIG5 , show that the intensity of normal plus anger is greater than sadness and joy during the speech. On the other hand, the intensity of joy of ordinary person A shows a value greater than the intensity of sadness. Furthermore, as shown in FIG6 , the intensity of joy of ordinary person A is distributed within a larger range of values than the case of the physician shown in FIG4 and the case of the depression patient shown in FIG5 , and the intensity of sadness of ordinary person A is distributed within a lower range of values than the case of the physician shown in FIG4 and the case of the depression patient shown in FIG5 .

Figure 7 shows the temporal changes in the intensity of excitement, calmness, sadness, anger, and joy for a healthy, mentally ill person named B, who is different from the person named A in Figure 6. As shown in Figure 7, the excitement level of person named B, who is the subject PA, fluctuates between 3 and 7. Furthermore, similar to the case of person named A in Figure 6, the intensity of calmness and anger during the speech period is greater than that of sadness and joy. Furthermore, similar to the case of person named A in Figure 6, the intensity of joy in person named B is greater than that of sadness.

For example, the calculation unit 20 performs cross-correlation processing on the temporal variation of excitement and the temporal variation of the intensity of anger, sadness, and joy, both normal and normal, for each subject PA shown in Figures 4 to 7. The calculation unit 20 calculates the cross-correlation coefficient between the excitement level and the intensity of anger, sadness, and joy, both normal and normal, for each subject PA. The window width for the cross-correlation processing performed by the calculation unit 20 is set to 150 utterances, for example, but can be set based on the individual subject PA, the required processing speed, the estimation accuracy, and so on.

Figures 8 to 11 show an example of the results of cross-correlation processing between subject PA's excitement and various emotions, performed by calculation unit 20 shown in Figure 2 . The horizontal axes of Figures 8 to 11 represent the time axis in the order of subject PA's utterances, while the vertical axes represent cross-correlation coefficients. The dashed-dotted line shows the temporal variation in the cross-correlation coefficient between excitement and the intensity of a normal state plus anger; the dotted line shows the temporal variation in the cross-correlation coefficient between excitement and the intensity of sadness; and the dashed line shows the temporal variation in the cross-correlation coefficient between excitement and the intensity of joy.

FIG8 shows the temporal changes in the correlation coefficients between the physician's excitement level and the intensity of each of anger, sadness, and joy, as shown in FIG4. In the case of the physician shown in FIG8 , after 40 speech units, the correlation coefficient for anger and joy shows a larger value than for joy and sadness, and the correlation coefficient for sadness shows the smallest value. Furthermore, from the start of speech until 40 speech units, the physician's excitement level and the amount of data for each emotion within the window width of the cross-correlation processing (e.g., 150 speech units) are relatively small. Therefore, the values of the correlation coefficients between the excitement level and each emotion calculated by the calculation unit 20 are unstable, and the reliability of the calculation results is low. Therefore, in the following description, the correlation coefficients after 40 speech units are used for the physician shown in FIG8 .

FIG9 shows the temporal changes in the correlation coefficients between the excitement level and the normal level plus the intensity of anger, sadness, and joy for the depressed patient shown in FIG5 . In the case of the depressed patient shown in FIG9 , after 100 speech units, the correlation coefficient for sadness shows the maximum value, and the correlation coefficient for joy shows the minimum value. Furthermore, similarly to the case of FIG8 , in FIG9 , the values of the correlation coefficients between the excitement level and each emotion calculated by the calculation unit 20 from the start of the speech to 100 speech units are unstable, and the reliability of the results is low. Therefore, in the following description, the correlation coefficients after 100 speech units are used for the depressed patient shown in FIG9 .

Figure 10 shows the temporal changes in the cross-correlation coefficients between the excitement level and the intensities of normal, anger, sadness, and joy for person A shown in Figure 6. In the case of person A shown in Figure 10 , the cross-correlation coefficient for joy reaches its maximum value, while the cross-correlation coefficient for sadness reaches its minimum value, after 70 speech units. Furthermore, as in the cases of Figures 8 and 9 , in Figure 10 , the cross-correlation coefficients between the excitement level and each emotion calculated by calculation unit 20 are unstable between the start of speech and 70 speech units, resulting in low reliability. Therefore, in the following description, the cross-correlation coefficients after 70 speech units are used for the case of person A shown in Figure 10 .

Figure 11 shows the temporal changes in the correlation coefficients between the excitement level and the intensity of each of the normal, anger, sadness, and joy levels for the person B shown in Figure 7. In the case of the person B shown in Figure 11, the correlation coefficient for joy reaches its maximum value, while the correlation coefficient for sadness reaches its minimum value, after 70 speech units. Furthermore, as in the cases of Figures 8 to 10, the reliability of the correlation coefficients between the excitement level and each emotion calculated by the calculation unit 20 from the start of the speech to 70 speech units in Figure 11 is low. Therefore, the correlation coefficients after 70 speech units are used for the person B.

As shown in Figures 8 to 11, when subjects PA were healthy individuals—physicians, ordinary people A, and ordinary people B—expressed emotions such as anger or joy, which correlated most strongly with excitement, while sadness exhibited the weakest correlation. This suggests that healthy subjects PA were in a state of mind where they could express their emotions openly as their excitement increased. Furthermore, this state of mind often involved relatively primitive emotions such as anger. On the other hand, when subjects PA were depressed, sadness exhibited the strongest correlation with excitement, while joy exhibited the weakest correlation. This suggests that even when depressed subjects PA were in an excited state, they were, in contrast, in a state of mind frozen from within.

The calculation unit 20 uses the correlation coefficients between the excitement level of each subject's PA and the intensity of the normal state plus anger, sadness, and joy, as shown in Figures 8 to 11, to determine the equilibrium between the subject's PA's normal state plus anger, sadness, and joy. Specifically, because living organisms, such as humans, possess the property of maintaining a predetermined physiological and psychological state across the entire organism regardless of changes in internal and external environmental factors, the calculation unit 20 determines the equilibrium between emotions. The property of maintaining a predetermined state across the entire organism is called "homeostasis" or "homeostasis."

FIG12 shows an example of the emotional homeostasis of subject PA. FIG12(a) shows a coordinate system in which the coordinate axes representing, for example, normal plus anger, sadness, and joy intersect at an angle of 120 degrees. FIG12(a) represents, for example, the mutual correlation coefficients of normal plus anger, sadness, and joy calculated by the operation unit 20 as shown in FIG8 to FIG11 as vectors in each coordinate direction as the intensity of each emotion of subject PA. The operation unit 20 seeks the balance between emotions based on the vectors of each emotion shown in FIG12(a). In addition, the range of the intensity of each emotion of normal plus anger, sadness, and joy is equal to the range of the mutual correlation coefficients, which is a range of -1 to 1.

Figure 12(b) shows the equilibrium position P1 of each emotional equilibrium of subject PA, calculated by the calculation unit 20, for the vector shown in Figure 12(a) for subject PA's normal state plus the intensity of anger, sadness, and joy. As shown in Figure 12(b), the calculated equilibrium position P1 of subject PA's emotions deviates from the center of the coordinate system. Therefore, the calculation unit 20 calculates the distance between the center of the coordinate system and the equilibrium position P1 of subject PA's emotions as the deviation from homeostasis. As shown in Figure 12(c), the calculation unit 20 calculates the deviation from homeostasis as, for example, the values α, β, and γ on the coordinate axes for normal state plus anger, sadness, and joy. By using the calculated mutual correlation coefficients of each emotion as components of the vector to calculate the deviation from subject PA's homeostasis, the calculation speed can be increased compared to calculating the deviation from homeostasis using, for example, differentiation or integration.

Figures 13 to 16 show an example of the temporal changes in the deviations α, β, and γ of the homeostasis of each subject PA, as calculated by the calculation unit 20 shown in Figure 2 . The vertical axes of Figures 13 to 16 represent the deviations for each emotion, while the horizontal axes of Figures 13 to 16 represent the order of the subject PA's utterances as a time axis. Furthermore, the dashed line shows the temporal changes in the deviation α along the coordinate axis for normal plus anger, the dotted line shows the temporal changes in the deviation β along the coordinate axis for sadness, and the dotted line shows the temporal changes in the deviation γ along the coordinate axis for joy.

Figure 13 shows the temporal changes in the deviations from the homeostatic state of the emotions of the physician shown in Figure 8. Furthermore, Figure 13 shows the temporal changes in the deviations α, β, and γ after 40 utterances, when the correlation coefficients between excitement and each emotion were stable. As shown in Figure 13, in the physician's case, the deviation α for normal anger is positive and larger than the deviations β and γ for sadness and joy. Furthermore, the deviation β for sadness is a smaller negative value than the deviation γ for joy.

Figure 14 shows the temporal changes in the deviations from the homeostatic state of emotions for the depressed patient shown in Figure 9. Furthermore, Figure 14 shows the temporal changes in the deviations α, β, and γ after 100 utterances, when the correlation coefficients between excitement and each emotion were stable. As shown in Figure 14, for depressed patients, the deviation β for sadness is positive and larger than the deviations α and γ for normal expression plus anger and joy. Furthermore, the deviation γ for joy in depressed patients is negative and smaller than the deviation α for normal expression plus anger.

Figure 15 shows the temporal changes in the deviations from the homeostatic state of emotions for person A shown in Figure 10. Furthermore, Figure 15 shows the temporal changes in the deviations α, β, and γ after 70 utterances, when the correlation coefficients between excitement and each emotion are stable. For person A shown in Figure 15, the deviation γ for joy is positive and larger than the deviations α and β for normal state plus anger and sadness. Furthermore, the deviation β for sadness is negative and smaller than the deviation α for normal state plus anger.

Figure 16 shows the temporal changes in the deviations from the homeostatic state of emotions for person B shown in Figure 11. Furthermore, Figure 16 shows the temporal changes in the deviations α, β, and γ after 70 utterances, when the correlation coefficients between excitement and each emotion are stable. In the case of person B shown in Figure 16, as in the case of person A shown in Figure 15, the deviation γ for joy is positive and larger than the deviations α and β for normal state plus anger and sadness. Furthermore, the deviation β for sadness for person B is a negative value, smaller than the deviation α for normal state plus anger.

For example, the estimation unit 30 calculates the distance between the coordinate center shown in FIG12(b) and the equilibrium position P1 based on the deviations from homeostasis shown in FIG12 through FIG16 . The estimation unit 30 estimates the condition of the subject PA based on the deviations α, β, and γ and the calculated distance from the equilibrium position P1. For example, in the case of the physician shown in FIG13 , if the deviation α for normal and angry states is positive and the deviation β for sadness is negative and smaller than the deviations α and γ, and the distance from the equilibrium position P1 is less than a predetermined value, the estimation unit 30 estimates that the subject PA is healthy (or normal). However, if the distance from the equilibrium position P1 is greater than a predetermined value, even though the deviation α for normal and angry states is positive and the deviation β for sadness is negative and smaller than the deviations α and γ, the estimation unit 30 estimates that the subject PA is in a manic state.

Furthermore, for example, as in the case of a typical person B shown in FIG16 , if the deviation γ of joy is positive, the deviation β of sadness is negative and smaller than the deviations α and γ, and the distance to the equilibrium position P1 is less than a predetermined value, the estimation unit 30 estimates that the subject PA is healthy (or normal). However, if the deviation γ of joy is positive and the deviation β of sadness is negative and smaller than the deviations α and γ, but the distance to the equilibrium position P1 is greater than the predetermined value, the estimation unit 30 estimates that the subject PA is in a manic state. On the other hand, for example, as in the case of a depressed patient shown in FIG14 , if the deviation β of sadness is positive and larger than the normal state plus the deviations α and γ of anger and joy, the estimation unit 30 estimates that the subject PA is in a depressed state.

Furthermore, the relationship between the magnitudes of the deviations α, β, and γ, as well as the relationship between the prescribed value corresponding to the distance from the equilibrium position P1 and the disease condition, can be determined based on, for example, the 10th edition of the International Statistical Classification of Diseases and Related Health Problems (ICD-10). The determined magnitudes of the deviations α, β, and γ, as well as the relationship between the prescribed value corresponding to the distance from the equilibrium position P1 and the disease condition, are pre-stored in the storage device of the estimation device 100. Here, ICD stands for International Statistical Classification of Diseases and Related Health Problems. Furthermore, the prescribed value can be adjusted to account for individual differences in the PA of the subjects.

Furthermore, the estimating unit 30 may also consider not only the deviation amounts α, β, γ and the distance from the equilibrium position P1, but also the orientation of the deviation of the equilibrium position P1 relative to the coordinate center, thereby making a detailed assessment of the condition of the subject PA. Furthermore, the estimating unit 30 may estimate the condition of the subject PA based on the deviation amounts α, β, γ. Alternatively, the estimating unit 30 may estimate the condition of the subject PA based on, for example, the rate of stabilization or change of the deviations indicated by the deviation amounts α, β, γ of the subject PA's homeostasis.

Alternatively, the estimating unit 30 may estimate the condition of the subject PA using the deviations α, β, and γ calculated by the computing unit 20 over a long period of time, such as two weeks. By using the deviation data over a long period of time, the estimating unit 30 can estimate the condition of the subject PA with high accuracy.

FIG17 illustrates an example of an estimation process performed by the estimation device 100 shown in FIG2 . Steps S10 to S40 are performed by the CPU mounted on the estimation device 100 executing the estimation program. That is, FIG17 illustrates another embodiment of the program and the estimation method. In this case, the extraction unit 10, the calculation unit 20, and the estimation unit 30 shown in FIG2 are implemented by executing the program. Alternatively, the process shown in FIG17 can be implemented by hardware mounted on the estimation device 100. In this case, the extraction unit 10, the calculation unit 20, and the estimation unit 30 shown in FIG2 are implemented by circuits configured within the estimation device 100.

In step S10 , the extraction unit 10 extracts first information indicating the physiological state of the subject PA and second information indicating at least one of emotion and organ activity based on the physiological information indicating the subject PA measured by the measurement device 1 as described in FIG. 2 to FIG. 7 .

In step S20 , the calculation unit 20 performs a cross-correlation process on the extracted temporal changes of the first information and the second information as described with reference to FIG. 4 to FIG. 11 , and calculates a cross-correlation coefficient indicating the degree of similarity.

In step S30 , the calculation unit 20 obtains the amount of deviation in the homeostasis of the subject PA based on the obtained cross-correlation coefficient as described with reference to FIG. 12 to FIG. 16 .

In step S40 , the estimating unit 30 estimates the condition of the subject PA based on the deviation amount of the homeostasis of the subject PA obtained by the calculating unit 20 , as described with reference to FIG. 12 to FIG. 16 .

Then, the estimation process performed by the estimation device 100 ends. The process shown in Figure 17 can be repeatedly executed each time an instruction is received from the doctor or the subject PA, or it can be executed at a predetermined frequency. In addition, the estimation device 100 outputs the estimation result to the output device 2. The output device 2 displays the result of the estimated disease state and the deviation from homeostasis. In addition, the output device 2 can also indicate the magnitude of the deviation from homeostasis, that is, the degree of the symptoms of the estimated disease state or the degree of health of the subject PA, by using colors or expressions of animated characters, animals, etc., and display them on the display. In addition, the output device 2 can also display suggestions for treatment methods for the estimated disease state based on the magnitude of the deviation from homeostasis.

In the embodiments shown in Figures 2 to 17 , the deviation from homeostasis of subject PA is calculated using first information representing the physiological state of subject PA and second information representing at least one of subject PA's emotions and organ activity. Thus, by referring to the deviation from homeostasis as an indicator, the estimation device 100 can easily estimate the condition of subject PA, even without specialized medical knowledge.

In addition, the calculation unit 20 can also use, for example, a decision tree that represents the relationship between heart rate and heartbeat fluctuations and emotions, instead of the decision tree shown in Figure 3 that represents the relationship between the basic frequency of speech and emotions, to obtain the intensity of the normal, sad, angry, and happy emotions of the subject PA.

FIG18 shows an example of a judgment tree for the heart rate and heartbeat variation of subject PA and the emotion of subject PA. In addition, RRV (R-R Variance) represents the variance of the intervals between R waves in the electrocardiogram. As shown in FIG18 , for example, a normal emotion is defined as a situation where the heart rate is less than 80 bps and the RRV is greater than 100. In addition, a sad emotion is defined as a situation where the heart rate is less than 80 bps and the RRV is less than 100. An angry emotion is defined as a situation where the heart rate is greater than 80 bps and the power of the low-frequency component LF of the heartbeat variation is greater than 80. A joyful emotion is defined as a situation where the heart rate is greater than 80 bps and the power of the low-frequency component LF is less than 80.

Furthermore, the calculation unit 20 obtains the deviation amounts α, β, and γ of the homeostasis of the subject PA as shown in FIG12( c ). However, the deviation amounts α, β, and γ may be obtained as shown in FIG19 , for example.

Figure 19 shows another example of emotional homeostasis for subject PA. The coefficient h shown in Figure 19 is an indicator indicating which of the following is greater: the deviation γ of joy in the coordinate axis or the deviation β of sadness in the coordinate axis, in the vector V1 extending from the center of the coordinate system toward the equilibrium position P1. Specifically, the coefficient h takes a positive value when the deviation γ of joy is greater than the deviation β of sadness, and a negative value when the deviation β of sadness is greater than the deviation γ of joy. Furthermore, if the deviation γ of joy is equal to the deviation β of sadness, the coefficient h becomes zero.

For example, to determine the coefficient h, the calculation unit 20 determines the angle θ between vector V1 and the joy axis. Furthermore, when the joy deviation γ is greater than the sadness deviation β, the angle θ takes a smaller value, closer to 0 degrees (i.e., vector V1 is oriented toward the joy axis). On the other hand, when the sadness deviation β is greater than the joy deviation γ, the angle θ takes a larger value, indicating that vector V1 is oriented toward the sadness axis. As shown in Figure 19, the calculation unit 20 uses the angle θ and the length L of vector V1 to determine the coefficient h, depending on whether the equilibrium position P1 is in the region between joy and sadness (counterclockwise) (hereinafter referred to as region A) or in the region between joy and sadness (clockwise) (hereinafter referred to as region B). The calculation unit 20 then sets the determined coefficient h as the joy deviation γ of vector V1, and any negative coefficient h as the sadness deviation β. That is, β + γ = 0.

Furthermore, in the case of region A shown in FIG19(a), when the coefficient h is close to 0 (i.e., the angle θ is π/3), the orientation of vector V1 is in the negative direction of the coordinate axis for normal plus anger. That is, the deviation γ of joy and the deviation β of sadness are equal and exhibit a larger deviation than the deviation α of normal plus anger. In other words, the deviation α of normal plus anger is smaller than the deviation γ of joy and the deviation β of sadness. Therefore, when vector V1 is in region A, the calculation unit 20 calculates |h|-L as the deviation α of normal plus anger. On the other hand, in the case of region B shown in FIG19(b), when the coefficient h is close to 0 (i.e., the angle θ is 2π/3), the orientation of vector V1 is in the positive direction of the coordinate axis for normal plus anger. That is, the deviation γ of joy and the deviation β of sadness are equal and exhibit a smaller deviation than the deviation α of normal plus anger. In other words, the deviation α of normal plus anger is larger than the deviation γ of joy and the deviation β of sadness. Therefore, when vector V1 is in region B, the calculation unit 20 calculates L-|h| as the normal plus angry deviation α. Thus, the calculation unit 20 can calculate a positive deviation α when the equilibrium position P1 is near the positive normal plus angry axis, and a negative deviation α when the equilibrium position P1 is near the negative normal plus angry axis.

For example, using the deviations α, β, and γ shown in FIG19 , the estimation unit 30 estimates that subject PA is in a depressive state if the deviation β of sadness is a positive value greater than 0 and the deviations α and γ of normal plus anger and joy are small values close to 0. Furthermore, the estimation unit 30 estimates that subject PA is in a manic state if the deviation γ of joy is a positive value greater than 0 and the deviations α and β of normal plus anger and sadness are close to 0. Furthermore, the estimation unit 30 estimates that subject PA is in a manic state if the deviation α of the normal plus anger component is less than 0 (close to -1) and the deviations β and γ of sadness and joy are equal and equivalent.

Fig. 20 shows another embodiment of the estimation device and the estimation process. Elements having the same or equivalent functions as those described in Fig. 2 are denoted by the same or equivalent reference numerals, and detailed description thereof is omitted.

The estimation device 100a shown in FIG20 is a computer device including a CPU or other processing unit and a storage device such as a hard disk drive. The estimation device 100a is connected to the measurement device 1a and the output device 2 via a wired or wireless connection via an interface unit included in the estimation device 100a. Consequently, the estimation device 100a, the measurement device 1a, and the output device 2 operate as an estimation system SYS.

The measuring device 1a includes multiple devices, such as a microphone, a heart rate monitor, an electrocardiograph, a blood pressure monitor, a thermometer, a skin resistance meter, a camera, and an MRI (Magnetic Resonance Imaging) device, to measure physiological information representing the subject's PA. The measuring device 1a outputs the measured physiological information representing the subject's PA to the estimation device 100a. Alternatively, the measuring device 1a may include an acceleration sensor or an electronic gyroscope.

The physiological information representing the subject's PA measured by the measuring device 1a includes, among other things, sound signals, heart rate (pulse rate), heart rate variability, blood pressure, body temperature, sweating (skin resistance, potential skin electrical potential), eye movement, pupil diameter, and blink rate. Furthermore, other physiological information measured includes, for example, exhaled breath, body secretions such as hormones and biomolecules, brain waves, and fMRI (functional magnetic resonance imaging) information.

The estimation device 100a includes an extraction unit 10a, a calculation unit 20a, an estimation unit 30a, an experiment unit 40, and a storage unit 50. The functions of the extraction unit 10a, the calculation unit 20a, the estimation unit 30a, and the experiment unit 40 may be implemented by a program executed by a CPU or by hardware.

Extraction unit 10a, similarly or equivalently to extraction unit 10 shown in FIG2 , extracts first information representing the physiological state of subject PA from the physiological information representing subject PA measured by measurement device 1a. Furthermore, extraction unit 10a, similarly or equivalently to extraction unit 10 shown in FIG2 , extracts second information representing at least one of the emotions of subject PA and the activity of an organ such as the heart or intestines of subject PA from the physiological information representing subject PA measured by measurement device 1a.

The extraction unit 10a extracts, for example, the heart rate (pulse rate) measured by a cardiometer or the like included in the measurement device 1a as second information representing the emotions and organ activity of the subject PA. Furthermore, excitement and stress increase adrenaline secretion in the body, which in turn increases the heartbeat and the heart rate (pulse rate).

Furthermore, the extraction unit 10a performs frequency analysis, such as FFT, on the electrocardiogram waveform of the subject PA measured by the electrocardiograph included in the measurement device 1a, thereby obtaining the cardiac fluctuations of the subject PA. Furthermore, the extraction unit 10a compares the amounts of the low-frequency component LF (e.g., 0.04 to 0.14 Hz) and the high-frequency component HF (e.g., 0.14 to 0.5 Hz) of the obtained cardiac fluctuations, extracting the level of excitement and tension of the subject PA as first information indicating the physiological state of the subject PA. Furthermore, the low-frequency component LF of the cardiac fluctuations has the following properties: it increases primarily with sympathetic nerve activity, while the high-frequency component HF increases with parasympathetic nerve activity.

Furthermore, the extraction unit 10a extracts, for example, the blood pressure value measured by a sphygmomanometer included in the measurement device 1a as second information representing the emotions and organ activity of the subject PA. Blood pressure has the following properties: when blood vessels constrict due to excitement or tension, blood flow resistance increases, leading to an increase in blood pressure.

Furthermore, the extraction unit 10a extracts, for example, the body temperature value measured by a thermometer or the like included in the measurement device 1a as second information representing the emotions and organ activity of the subject PA. Furthermore, body temperature has the following properties: excitement, nervousness, increased heart rate, elevated blood sugar levels, muscle tension, etc., generate heat in the body, causing the body temperature to rise.

Furthermore, the extraction unit 10a extracts, for example, the value of sweating (skin resistance, skin potential) measured by a skin resistance meter or the like included in the measurement device 1a as second information representing the emotions and organ activity of the subject PA. Furthermore, sweating (skin resistance, skin potential) has the following properties: excitement and tension promote sweating, which decreases skin resistance.

Furthermore, the extraction unit 10a extracts, for example, eye movements, pupil diameter, and the number of blinks measured by the electrooculometer, camera, etc., of the measurement device 1a as second information representing the emotions and internal organ activity of the subject PA. Alternatively, the extraction unit 10a may perform facial recognition processing on images captured by the camera, etc., and extract the recognized facial expression and its temporal changes as second information representing the emotions and internal organ activity of the subject PA. Furthermore, eye movements have the following properties: excitement and nervousness cause them to move more violently, excitement and nervousness cause pupil diameter to dilate, and excitement and nervousness cause the number of blinks to increase.

Furthermore, the extraction unit 10a extracts, for example, the number, rate, and volume of exhalations measured from respiratory volume and breathing sounds using a respirometer (respiratory flow meter), a spirometer, or a microphone included in the measurement device 1a as second information representing the subject PA's emotions and organ activity. Exhalation has the property that the number, rate, and volume of exhalations increase with excitement or nervousness.

Furthermore, the extraction unit 10a extracts, for example, body secretions such as hormones and biomolecules measured by the analysis device included in the measurement device 1a as second information representing the subject PA's emotions and organ activity. Furthermore, body secretions such as hormones and biomolecules are measured by chemical analysis of saliva, blood, lymph, sweat, digestive fluid, or urine collected from the subject PA by the analysis device of the measurement device 1a. Alternatively, body secretions can be measured by the measurement device 1a from the subject PA's peripheral blood vessels, digestive system, myoelectric potential, skin temperature, blood flow, or immune system. Furthermore, body secretions have the following properties: the amount or quality of hormones and biomolecules secreted in the body changes due to excitement or stress.

Furthermore, the extraction unit 10a extracts, for example, the amount of change in the electroencephalogram (EEG) measured over time using an optical, magnetic, or potentiometric EEG activity meter included in the measurement device 1a as first information indicating the excitement or tension of the subject PA. EEG waves have the property that their waveform changes with excitement or tension.

Furthermore, the extraction unit 10a extracts, for example, blood flow and the distribution of oxygenated hemoglobin in various brain activity sites contained in the fMRI information measured by the MRI device included in the measurement device 1a as second information representing the subject's PA's emotions and organ activity. Furthermore, the measured fMRI information has the property that the brain activity sites change due to excitement and stress. For example, emotional excitement and stress manifest as changes in blood flow in the limbic system (amygdala), hypothalamus, cerebellum, brainstem, or hippocampus. These changes in blood flow alter the brain distribution of oxygenated hemoglobin.

Furthermore, when the measurement device 1 a includes an acceleration sensor or an electronic gyroscope, the extraction unit 10 a may extract the motion of the subject PA as the second information indicating the emotion or the activity of the internal organs of the subject PA.

The calculation unit 20a calculates the similarity between the temporal variations of the first and second information extracted by the extraction unit 10a. For example, the calculation unit 20a performs cross-correlation processing on the temporal variations of the extracted first and second information and calculates the cross-correlation coefficient as the similarity. The calculation unit 20a uses the multiple similarities calculated between the subject PA's emotions and organ movements to determine the deviation from the homeostasis of the subject PA. The operation of the calculation unit 20a and the homeostasis are explained using FIG21.

Based on the deviation from homeostasis calculated by the computing unit 20a, the experimental unit 40 calculates the energy acting on the subject PA's emotions and organ activity. The experimental unit 40 inputs the calculated energy into a computational model representing the subject PA's biological body, thereby simulating the subject PA's homeostasis. The computational model and the operation of the experimental unit 40 are described using Figures 22 and 23.

The storage unit 50 is a hard disk device, a memory, etc. The storage unit 50 stores programs for execution by the CPU. The storage unit 50 also stores data 60 representing the results of simulations performed by the experiment unit 40. The data 60 will be described using FIG. 23 .

The program for executing the estimation process can be distributed by recording it on a removable disc such as a CD (Compact Disc) or DVD (Digital Versatile Disc). Furthermore, the estimation device 100a can download the program for executing the estimation process from a network via a network interface included in the estimation device 100a and store it in the storage unit 50.

The estimating unit 30a estimates the condition of the subject PA based on the pattern of changes in homeostasis simulated by the experimenting unit 40. The operation of the estimating unit 30a will be described with reference to FIG22 and FIG23.

Figure 21 schematically illustrates an example of the linkage of homeostasis in subject PA. In Figure 21, for example, the balance of homeostasis in the entire biological body of subject PA is represented by a rotating circular figure, which is set as circulatory system 200. Circulatory system 200, for example, further includes multiple circulatory systems K (K1-K10), which form the substances and organs of subject PA. In Figure 21, circulatory systems K1-K10 are represented by rotating circles smaller than circulatory system 200, which are linked to each other to maintain homeostasis. For example, circulatory system K1 represents the homeostasis of subject PA's emotions based on the sound signals emitted by subject PA via the vocal cords. Circulatory system K2 represents the homeostasis of subject PA's heart based on, for example, heart rate and heartbeat fluctuations. Circulatory system K3 represents the homeostasis of subject PA's digestive system, including the stomach, small intestine, and large intestine. Circulatory system K4 represents the homeostasis of subject PA's immune system, which protects subject PA from disease. The circulatory system K5 represents, for example, the homeostasis of hormones that transmit information regulating the activities of organs included in the biological body of the subject PA.

Furthermore, the circulatory system K6, for example, represents the homeostasis of biomolecules such as various proteins produced by the genetic factors of the subject PA. The circulatory system K7, for example, represents the homeostasis of the genetic factors of the subject PA. The circulatory system K8, for example, represents the homeostasis of the activity of the cells that form the subject PA. The circulatory system K9, for example, represents the homeostasis of the activity of the limbic system of the subject PA, including the amygdala, in the brain, which is closely related to emotion. The circulatory system K10, for example, represents the homeostasis of neurotransmitters that transmit information between synapses.

The circulatory system 200 includes ten circulatory systems K1-K10, but is not limited thereto and may include a plurality of circulatory systems other than ten. Furthermore, each circulatory system K may further include multiple circulatory systems. For example, the vocal cord circulatory system K1 may include multiple circulatory systems representing the subject PA's emotions, such as anger, calmness, sadness, and joy. Furthermore, the cardiac circulatory system K2 may include multiple circulatory systems representing, for example, the subject PA's heart rate and heartbeat variability.

For example, the calculation unit 20a uses the calculated multiple similarities between the subject PA's emotions and organ activity to determine the deviation from the homeostasis of each circulatory system K of the subject PA, as illustrated in FIG12 . The calculation unit 20a calculates the deviation from the homeostasis of the subject PA's emotions based on the subject PA's voice signal, similar to the calculation unit 20 shown in FIG2 . Furthermore, the calculation unit 20a performs cross-correlation processing on, for example, the degree of excitement or tension obtained from the ratio of the low-frequency component LF to the high-frequency component HF of the heartbeat fluctuations measured by an electrocardiograph, and the temporal changes in heart rate, blood pressure, and the like. Furthermore, the calculation unit 20a calculates the deviation from the homeostasis of the subject PA's heart, for example, based on the cross-correlation coefficient between the degree of excitement or tension and the temporal changes in heart rate, blood pressure, and the like.

Furthermore, the calculation unit 20 a calculates the deviation amount of the homeostasis of all the circulatory systems K1 to K10 , but the deviation amount of the homeostasis of a part of the circulatory system K may be calculated.

FIG22 shows an example of a computational model of the circulatory system 200 used by the experimental unit 40 shown in FIG20 to simulate the homeostasis of subject PA. The computational model of the circulatory system 200 shown in FIG22 represents, for example, the circulatory systems K1-K10 included in the circulatory system 200 shown in FIG21 as axes SH (SH1-SH10), and is constructed in a virtual space such as a computer device. The length, pitch width, and thread orientation of each axis SH1-SH10 are determined based on the biological characteristics of subject PA. Furthermore, the axes SH1-SH10 are connected by a joint B1 so that their centers align, thereby forming the circulatory system 200. Furthermore, nuts NT1-NT10 are placed on the axes SH1-SH10 of the circulatory systems K1-K10, respectively. The experimental unit 40 simulates the homeostasis of the circulatory system 200 by, for example, rotating the axis SH and detecting the homeostasis of each circulatory system K1-K10 based on changes in the position of the nuts NT1-NT10.

The length, pitch width, and thread orientation of the vocal cord circulatory system K1 are determined based on, for example, the frequency distribution, intonation, or pitch frequency characteristics of the sound signal of the subject PA's speech. Furthermore, the length, pitch width, and thread orientation of the cardiac circulatory system K2 are determined based on, for example, the time interval between heart beats and the frequency distribution of heartbeat fluctuations of the subject PA. The length, pitch width, and thread orientation of the digestive system circulatory system K3 are determined based on, for example, the length of the subject PA's small and large intestines, or the speed of the contraction waves associated with peristalsis. The length, pitch width, and thread orientation of the immune system circulatory system K4 are determined based on, for example, the number of white blood cells in the subject PA's blood, including neutrophils, eosinophils, basophils, lymphocytes, and monocytes. The length, pitch width, and thread orientation of the hormone circulation system K5 are determined based on characteristics such as the amount of hormones synthesized or secreted by various organs of the subject PA and the rate at which hormones circulate in the body through body fluids such as blood.

Furthermore, the length, pitch width, and thread orientation of the biomolecule circulatory system K6 are determined based on, for example, the intake of nucleic acids, proteins, polysaccharides, amino acids (constituent elements of these substances), various sugars, lipids, and vitamins contained in the food consumed by the subject PA. The length, pitch width, and thread orientation of the genetic factor circulatory system K7 are determined based on, for example, the frequency of genetic factor division and genetic factor length characteristics of the subject PA. Furthermore, the length, pitch width, and thread orientation of the cellular circulatory system K8 are determined based on, for example, the amount of sugars, lipids, proteins (amino acids), nucleic acids, etc. contained in the cells of the subject PA, as well as the lifespan of the cells. The length, pitch width, and thread orientation of the brain circulatory system K9 are determined based on, for example, the temporal fluctuations and frequency distribution of brain activity in the subject PA's brain, including the amygdala. The length, pitch width, and thread orientation of the neurotransmitter circulatory system K10 are determined based on, for example, the secretion levels and characteristic reaction rates of amino acids, peptides, monoamines, etc., which transmit information between synapses in the subject PA.

Information indicating the length, pitch width, and thread orientation of each of the axes SH1-SH10 that have been set is stored in advance in the storage unit 50 for each subject PA. Alternatively, the experimenter 40 may receive information indicating the length, pitch width, and thread orientation of each of the axes SH1-SH10 of the subject PA via an input device such as a keyboard or touch panel included in the estimation device 100a.

The experiment unit 40 calculates the energy acting on the subject PA's emotions and organ activity based on the deviations from homeostasis of each of the circulatory systems K1-K10 calculated by the calculation unit 20a. For example, as shown in FIG12(b), similar to the calculation unit 20 shown in FIG2, if the equilibrium position P1 of the subject PA's emotions calculated by the calculation unit 20a differs from the center of the coordinate system, this indicates that the subject PA's emotions, i.e., the homeostasis of the circulatory system K1, have shifted or deviated from a predetermined state. This deviation from homeostasis manifests itself on the subject PA in the form of, for example, stress, affecting not only the subject PA's circulatory system K1 but also other circulatory systems K2-K10, such as the heart and digestive system. Therefore, the experiment unit 40 calculates the energy acting on the subject PA's emotions and organ activity as stress, etc., based on the deviations from homeostasis calculated by the calculation unit 20a for each of the circulatory systems K1-K10. For example, the experimenter 40 calculates energy E(K1) in the vocal cord circulatory system K1 using equation (1) based on the deviation amounts α, β, and γ of the emotional homeostasis calculated by the calculation unit 20a.

E(K1)=sqrt(α×α+β×β+γ×γ)…(1)

In addition, as shown in formula (1), the experimental unit 40 calculates the energy E(K1) generated in the circulatory system K1 of the vocal cords based on the deviation amounts α, β, and γ of the homeostasis of emotions, but the energy E(K1) can also be calculated using a function F(α, β, γ) in which the deviation amounts α, β, and γ of the homeostasis of emotions are set as variables.

The experiment unit 40 also calculates the energy E(K2)-E(K10) for each of the circulatory systems K2-K10, based on the deviation from homeostasis calculated by the calculation unit 20a. The experiment unit 40 calculates the total energy calculated for each of the circulatory systems K1-K10, based on the stress, calories consumed by exercise, and food intake. The total energy E(K2)-E(K10) is calculated using equation (2).

TE＝E(K1)+E(K2)+E(K3)+E(K4)+E(K5)+E(K6)+E(K7)+E(K8)+E(K9)+E(K10)…(2)

Here, E(K2), E(K3), E(K4), E(K5), E(K6), E(K7), E(K8), E(K9), and E(K10) represent the energy generated in the circulation systems K2-K10. TE represents the total energy. Furthermore, while the experimental unit 40 calculates the energy TE by summing the energies E(K1)-E(K10) generated in the circulation systems K1-K10, it is also possible to calculate the energy TE by weighted addition of the energies E(K1)-E(K10). Alternatively, the experimental unit 40 may calculate the energy TE by multiplying the energies E(K1)-E(K10).

The experimental unit 40 inputs the calculated energy TE into the circulatory system 200 and rotates the shaft SH at a rotational speed corresponding to the magnitude of the energy TE. Furthermore, the experimental unit 40 rotates the shaft SH clockwise when the energy TE is positive, and counterclockwise when the energy TE is negative. Furthermore, the input energy TE is controlled by the experimental unit 40 so that the displacements L1-L10 of the nuts NT1-NT10 in response to the rotation of the shaft SH fall within the length range of the respective shafts SH1-SH10.

Furthermore, the energy TE may be positive or negative because, for example, energy is generated in the circulatory system K1-K10 to rotate the shafts SH clockwise or counterclockwise, depending on the orientation of the threads of the shafts SH1-SH10. Specifically, for example, if the shafts SH rotate clockwise depending on the orientation of the threads, positive energy is generated in the circulatory system K, while if the shafts SH rotate counterclockwise, negative energy is generated in the circulatory system K. Alternatively, the experimental unit 40 may rotate the shafts SH1-SH10 clockwise as a whole when the energy TE is negative, and rotate the shafts SH1-SH10 counterclockwise when the energy TE is positive.

The experimental unit 40 rotates the shafts SH1-SH10 using energy TE, thereby displacing the positions of the nuts NT1-NT10. The experimental unit 40 detects the displacements L1-L10 of each nut NT1-NT10 from the respective centers C1-C10 of the shafts SH1-SH10 as changes in the homeostasis (or deviations in the homeostasis) of each circulatory system K1-K10. The experimental unit 40 stores the detected displacements L1-L10 as data 60 in the storage unit 50. Furthermore, based on the displacements L1-L10, the experimental unit 40 detects the speed at which each nut NT1-NT10 moves along the axial direction of the shafts SH1-SH10. The experimental unit 40 inputs the speeds detected in the circulatory system K1-K10 as newly generated energies E(K1)-E(K10) into the circulatory system 200.

Furthermore, when the calculation unit 20a calculates the amount of deviation from homeostasis of a portion of the circulatory system K among the circulatory systems K1-K10, the experimental unit 40 may also determine the energy TE based on the amount of deviation from homeostasis of the portion of the circulatory system K calculated by the calculation unit 20a, and simulate the homeostasis of the circulatory system 200 based on the determined energy TE. Furthermore, the experimental unit 40 may also detect the displacements L1-L10 of all of the circulatory systems K1-K10 through the simulation. Since the experimental unit 40 detects the displacement L of all of the circulatory system K through the simulation, the condition of the subject PA can be estimated with higher accuracy compared to a case where the estimation device 100a uses the amount of deviation from homeostasis of the circulatory system K calculated by the calculation unit 20a.

In the experimental unit 40, the displacements L1-L10 of the circulatory systems K1-K10 are determined by the distances from the centers C1-C10 of the shafts SH1-SH10, but the present invention is not limited thereto. For example, the displacements L1-L10 may be the distances between the nuts NT1-NT10 or the distances from the joint B1.

23 shows an example of data 60 of displacement amounts L1 to L10 of the circulatory systems K1 to K10 of the subject PA. The data 60 has storage areas for dates and circulatory systems K1 to K10, respectively.

The date storage area stores the date and time when the experiment unit 40, for example, simulated the change in the homeostasis of the circulatory system 200 and detected the displacements L1 to L10 of the circulatory systems K1 to K10 (e.g., October 29, 2013, 9:10:00). The experiment unit 40 detects the displacements L1 to L10 at intervals of, for example, one minute, one hour, one day, one week, or one month. In the case of the data 60 shown in FIG. 23 , the intervals are, for example, one hour.

Each storage area of the circulation systems K1 to K10 stores, for example, the displacement amounts L1 to L10 of the nuts NT1 to NT10 detected by the testing unit 40. The units of the displacement amounts L1 to L10 are centimeters or millimeters.

The estimating unit 30a reads the date of data 60 and the displacements L1-L10 of the circulatory systems K1-K10 from the storage unit 50. The estimating unit 30a estimates the medical condition of the subject PA based on the read temporal variation patterns of the displacements L1-L10. For example, data on typical temporal variation patterns of the displacements L1-L10 of the circulatory systems K1-K10, as seen when the subject PA is healthy, is previously stored in the storage unit 50. The estimating unit 30a then compares the temporal variation of the displacements L1-L10 detected by the testing unit 40 with the typical temporal variation of the displacements L1-L10 when the subject PA is healthy, and estimates the medical condition of the subject PA based on the comparison results. For example, the estimating unit 30a calculates the difference between the temporal variation of the displacements L1-L10 detected by the testing unit 40 and the typical temporal variation of the displacements L1-L10 when the subject PA is healthy, and compares the calculated difference with a predetermined threshold value indicating each medical condition. Specifically, for example, in the case of the cardiac circulatory system K2, the estimating unit 30a calculates the difference between the temporal variation of the displacement L2 detected by the testing unit 40 and the typical temporal variation of the displacement L2 when the subject PA is healthy. The estimating unit 30a compares the calculated difference with a predetermined threshold value indicating a heart disease such as myocardial infarction or angina pectoris, thereby estimating whether the subject PA suffers from a heart disease such as myocardial infarction or angina pectoris.

FIG24 shows an example of an estimation process performed by the estimation device 100a shown in FIG20 . Steps S100 to S160 are implemented by the CPU mounted on the estimation device 100a executing the estimation program. That is, FIG24 shows another embodiment of the estimation program and the estimation method. In this case, the extraction unit 10a, the calculation unit 20a, the estimation unit 30a, and the experimental unit 40 shown in FIG20 are implemented by executing the estimation program. In addition, the process shown in FIG24 can also be implemented by hardware mounted on the estimation device 100a. In this case, the extraction unit 10a, the calculation unit 20a, the estimation unit 30a, and the experimental unit 40 shown in FIG20 are implemented by circuits configured in the estimation device 100a.

In step S100 , as described with reference to FIG. 20 , the extraction unit 10 a extracts first information indicating the physiological state of the subject PA and second information indicating the state of emotions and organ activity based on the physiological information indicating the subject PA measured by the measurement device 1 a .

In step S110 , as described with reference to FIG. 21 , the calculation unit 20 a performs a cross-correlation process on the extracted temporal changes of the first information and the second information, and calculates a cross-correlation coefficient indicating the degree of similarity.

In step S120 , as described with reference to FIG. 12 and FIG. 21 , the calculation unit 20 a obtains the amount of deviation in homeostasis of each of the circulatory systems K1 to K10 of the subject PA based on the obtained cross-correlation coefficients.

In step S130, as described in FIG22 , the experimenter 40 calculates the energies E(K1)-E(K10) based on the deviations from the homeostasis of each of the circulatory systems K1-K10 calculated by the computational unit 20a. The experimenter 40 calculates the energy TE by summing the calculated energies E(K1)-E(K10) using equation (2).

In step S140 , as described with reference to FIG. 22 , the experiment section 40 inputs the energy TE totaled in step S130 into the circulatory system 200 , thereby simulating the homeostasis of the circulatory system 200 of the subject PA.

In step S150, as described in FIG22 , the experiment unit 40 detects the displacements L1 to L10 of the circulatory systems K1 to K10 based on the homeostasis simulation executed in step S140. The experiment unit 40 stores the detected displacements L1 to L10 of the circulatory systems K1 to K10 as data 60 in the storage unit 50.

In step S160, as described in FIG23 , the estimating unit 30a estimates the condition of the subject PA based on the temporal variation patterns of the displacements L1-L10 of the circulatory systems K1-K10. For example, the estimating unit 30a compares the temporal variation pattern of the displacements L1-L10 detected by the testing unit 40 with a typical temporal variation pattern of the displacements L1-L10 when the subject PA is healthy, and estimates the condition of the subject PA based on the comparison results.

Then, the estimation process performed by the estimation device 100a is completed. The process shown in Figure 24 can be repeatedly executed each time an instruction is received from the doctor or the subject PA, or it can be executed at a predetermined frequency. In addition, the estimation device 100a outputs the estimation result to the output device 2. The output device 2 displays the result of the estimated disease state and the deviation amount of homeostasis. In addition, the output device 2 can also use colors or expressions of animated characters, animals, etc. to indicate the size of the deviation amount of homeostasis, that is, the degree of the symptoms of the estimated disease state or the degree of health of the subject PA, and display it on the display. In addition, the output device 2 can also display suggestions for treatment methods for the estimated disease state based on the size of the deviation amount of homeostasis.

In the embodiments shown in Figures 20 to 24 , the deviation from the homeostasis of subject PA is calculated using first information representing the physiological state of subject PA and second information representing the subject PA's emotions and organ activity. Thus, by referring to the deviation from the homeostasis as an indicator, the estimation device 100a can easily estimate the condition of subject PA, even without medical expertise. Furthermore, the estimation device 100a performs a simulation of the homeostasis of subject PA, using the deviation from the homeostasis of each circulatory system K as input energy. The estimation device 100a compares the temporal variation in homeostasis detected by the simulation with the temporal variation in homeostasis displayed when subject PA is healthy, thereby enabling a more accurate estimation of the condition of subject PA than is conventionally possible.

FIG25 shows another embodiment of an estimation device. Elements having the same or equivalent functions as those described in FIG20 are given the same or equivalent reference numerals, and their detailed descriptions are omitted. For example, the estimation device 100b, the measuring device 1a, and the output device 2 operate as an estimation system SYS.

Estimation device 100b shown in FIG25 is a computer device including a CPU or other processing unit and a storage device such as a hard disk drive. Estimation device 100b is connected to measurement device 1a and output device 2 via a wired or wireless connection via an interface unit included in estimation device 100b. Thus, estimation device 100b, measurement device 1a, and output device 2 operate as estimation system SYS.

The estimation device 100b includes an extraction unit 10a, a calculation unit 20a, an estimation unit 30b, an experiment unit 40a, and a storage unit 50a. The functions of the extraction unit 10a, the calculation unit 20a, the estimation unit 30b, and the experiment unit 40a may be implemented by a program executed by a CPU or by hardware.

The storage unit 50a is a hard disk device, memory, or the like. The storage unit 50a stores programs for execution by the CPU. Furthermore, the storage unit 50a stores data 60a representing the results of simulations performed by the experiment unit 40a, and a condition table 70 used by the estimation unit 30b to determine the condition of the subject PA using the data 60a. The data 60a and the condition table 70 are described using Figures 27 and 28.

The program for executing the estimation process can be recorded on a removable disk such as a CD or DVD and distributed. Alternatively, the estimation device 100b can download the program for executing the estimation process from a network via a network interface included in the estimation device 100b and store it in the storage unit 50.

Based on the deviation from homeostasis calculated by the computing unit 20a, the experimental unit 40a calculates the energy acting on the emotions and organ activity of the subject PA. The experimental unit 40a inputs the calculated energy into a computational model representing the biological body of the subject PA, thereby simulating the homeostasis of the subject PA. The operation of the computational model and the experimental unit 40a will be described using FIG26 .

The estimating unit 30b estimates the condition of the subject PA based on the change pattern of the homeostasis simulated by the experimental unit 40a. The operation of the estimating unit 30b will be described with reference to FIG27 and FIG28.

FIG26 shows an example of a calculation model of the circulatory system 200a used by the experimental unit 40a shown in FIG25 to simulate the homeostasis of the subject PA. The calculation model of the circulatory system 200a shown in FIG26 includes, for example, four circulatory systems Ka (Ka1-Ka4) contained within the circulatory system 200a. The circulatory system 200a and the circulatory systems Ka1-Ka4 contained within the circulatory system 200a are represented by gears MG and gears Ga1-Ga2, Gb1-Gb2, Gc1, and Gd1-Gd2, and are constructed in a virtual space such as a computer device. Gears MG rotate based on energies E(Ka1)-E(Ka4), which are calculated based on the deviations in homeostasis of each circulatory system Ka1-Ka4 determined by the calculation unit 20a. Rotation of gears MG causes rotation of gears Ga1-Ga2, Gb1-Gb2, Gc1, and Gd1-Gd2 within each circulatory system Ka1-Ka4. As shown in Figure 26, circulatory systems Ka1, Ka2, and Ka4 each have two gears: Ga1-Ga2, Gb1-Gb2, and Gd1-Gd2, while circulatory system Ka3 has one gear: Gc1. The diameter and number of teeth of the gears MG, as well as the number, diameter, and number of teeth of the gears included in each of the circulatory systems Ka1-Ka4, are determined based on the biological characteristics of the subject PA. The experimental unit 40a simulates the homeostasis of the circulatory system 200a by, for example, rotating the gears MG. The homeostasis of each of the circulatory systems Ka1-Ka4 is then detected based on the rotational speeds of the gears Ga2, Gb2, Gc1, and Gd2.

Furthermore, if the circulatory system Ka represents the vocal cords, the number, diameter, and number of teeth of the gears included in the circulatory system Ka are determined based on, for example, the frequency distribution, intonation, or pitch frequency of the sound signal of the subject PA's speech. Furthermore, if the circulatory system Ka represents the heart, the number, diameter, and number of teeth of the gears included in the circulatory system Ka are determined based on, for example, the time interval between heartbeats and the frequency distribution of heartbeat fluctuations. If the circulatory system Ka represents the digestive system, the number, diameter, and number of teeth of the gears included in the circulatory system Ka are determined based on, for example, the length of the small intestine, large intestine, or the speed of the contraction wave associated with peristalsis. If the circulatory system Ka represents the immune system, the number, diameter, and number of teeth of the gears included in the circulatory system Ka are determined based on, for example, the number of white blood cells, including neutrophils, eosinophils, basophils, lymphocytes, and monocytes, in the blood of the subject PA.

Furthermore, if the circulatory system Ka represents hormones, the number, diameter, and number of teeth of the gears contained in the circulatory system Ka are determined based on, for example, the amount of hormones synthesized or secreted by the various organs of the subject PA, the rate at which the hormones circulate in the body through body fluids such as blood, and other properties. If the circulatory system Ka represents biomolecules, the number, diameter, and number of teeth of the gears contained in the circulatory system Ka are determined based on, for example, the amount of nucleic acids, proteins, polysaccharides, amino acids as components of these substances, various sugars, lipids, vitamins, and the like contained in the food consumed by the subject PA. If the circulatory system Ka represents genetic factors, the number, diameter, and number of teeth of the gears contained in the circulatory system Ka are determined based on, for example, the frequency of division of the genetic factors of the subject PA, the length of the genetic factors, and other properties. Furthermore, if the circulatory system Ka represents cells, the number, diameter, and number of teeth of the gears contained in the circulatory system Ka are determined based on, for example, the amount of sugars, lipids, proteins (amino acids), nucleic acids, and the like contained in the cells, or the lifespan of the cells, and other properties. When the circulatory system Ka represents the brain, the number, diameter, and number of teeth of the gears included in the circulatory system Ka are determined based on characteristics such as temporal fluctuations and frequency distribution of brain activity in the brain of the subject PA, including, for example, the amygdala. When the circulatory system Ka represents a neurotransmitter, the number, diameter, and number of teeth of the gears included in the circulatory system Ka are determined based on, for example, the secretion levels and characteristic reaction speeds of amino acids, peptides, monoamines, and the like that transmit information between synapses.

The storage unit 50 of the estimation device 100b stores, in advance, information indicating the diameter and number of teeth of the gear MG, as well as the number, diameter, and number of teeth of the gears Ga1-Ga2, Gb1-Gb2, Gc1, and Gd1-Gd2, for each subject PA. Furthermore, the experimenter 40 may receive information indicating the diameter and number of teeth of the gear MG, as well as the number, diameter, and number of teeth of the gears Ga1-Ga2, Gb1-Gb2, Gc1, and Gd1-Gd2, via an input device such as a keyboard included in the estimation device 100b.

Furthermore, while circulatory system 200a is assumed to include four circulatory systems Ka1-Ka4, this is not limited to this and may include multiple circulatory systems other than four. Furthermore, each circulatory system Ka may further include multiple circulatory systems. For example, if circulatory system Ka represents the vocal cords, it may include multiple gears representing multiple circulatory systems of subject PA's emotions, such as anger, calmness, sadness, and joy. Furthermore, if circulatory system Ka represents the heart, it may include multiple gears representing multiple circulatory systems, for example, where these multiple circulatory systems represent subject PA's heart rate, heartbeat variability, and the like.

Similar to the experimental unit 40 shown in FIG20 , the experimental unit 40a calculates energy TE based on the deviation from the homeostasis of each of the circulatory systems Ka1-Ka4 calculated by the calculation unit 20a using equations (1) and (2). The experimental unit 40a inputs the calculated energy TE into the circulatory system 200a and rotates the gear MG at a rotational speed corresponding to the magnitude of the energy TE. For example, the experimental unit 40a rotates the gear MG clockwise when the energy TE is positive, and rotates the gear MG counterclockwise when the energy TE is negative. Alternatively, the experimental unit 40a may rotate the gear MG counterclockwise when the energy TE is positive, and rotate the gear MG clockwise when the energy TE is negative.

The test unit 40a simulates the homeostasis of the circulation system 200a by rotating the gear MG. For example, the homeostasis state of each of the circulation systems Ka1-Ka4 is detected as the gear rotational speed. The test unit 40a stores the detected rotational speeds R1-R4 in the storage unit 50a. Furthermore, the test unit 40a inputs the rotational speeds R1-R4 detected in each of the circulation systems Ka1-Ka4 into the circulation system 200a as newly generated energies E(Ka1)-E(Ka4).

Furthermore, when the calculation unit 20a calculates the deviation from homeostasis of a portion of the circulatory systems Ka1-Ka4, the experimental unit 40a may determine the energy TE based on the deviation from homeostasis of the portion of the circulatory system Ka calculated by the calculation unit 20a, and simulate the homeostasis of the circulatory system 200a based on the determined energy TE. Furthermore, the experimental unit 40a may detect the rotational speeds R1-R4 of all of the circulatory systems Ka1-Ka4 based on the simulation. Since the experimental unit 40a detects the rotational speed R of all of the circulatory systems Ka based on the simulation, the condition of the subject PA can be estimated with higher accuracy compared to a case where the estimation device 100b uses the deviation from homeostasis of the circulatory system Ka calculated by the calculation unit 20a.

27 shows an example of data 60a of the rotational speeds R1 to R4 of the circulatory systems Ka1 to Ka4 of the subject PA. The data 60a has storage areas for dates and circulatory systems Ka1 to Ka4, respectively.

The date storage area stores the date and time when the experiment unit 40a, for example, simulated the homeostatic changes of the circulatory system 200 and detected the rotational speeds R1 to R4 of the circulatory systems Ka1 to Ka4 (e.g., October 29, 2013, 9:10:00). The experiment unit 40a detects the rotational speeds R1 to R4 at intervals of, for example, one minute, one hour, one day, one week, or one month. In the case of the data 60a shown in FIG27 , the intervals are, for example, one minute.

The storage areas of the circulation systems Ka1 to Ka4 store, for example, the rotation speeds R1 to R4 (eg, 20 revolutions per minute) of the gears Ga2 , Gb2 , Gc1 , and Gd2 detected by the test unit 40 a .

28 shows an example of the disease condition table 70. The disease condition table 70 has storage areas for disease conditions and circulatory systems Ka1 to Ka4.

The condition storage area stores conditions such as severe depression, depression, normal (i.e., the subject's PA is healthy), manic-depressive disorder, and personality disorder. Furthermore, while the condition table 70 shown in FIG28 shows mental illness as a condition, heart disease such as myocardial infarction or brain disease such as cerebral infarction may also be included.

The storage areas for circulatory systems Ka1-Ka4 store conditions used by the estimation unit 30b to estimate each condition stored in the condition storage area. Storage areas containing "-" indicate conditions not included in the corresponding condition estimation. For example, if circulatory systems Ka1-Ka4 represent the emotions of anger, calmness, sadness, and joy, respectively, and the rotational speeds R1-R4 for all of these emotions are 0 (no rotation), the estimation unit 30b estimates that subject PA is severely depressed. In other words, severe depression indicates a homeostatic bias, where none of the emotions of anger, calmness, sadness, and joy occur in subject PA. Furthermore, if circulatory systems Ka1-Ka4 represent the emotions of anger, calmness, sadness, and joy, respectively, and the rotational speed R3 for sadness is less than a threshold value α, the estimation unit 30b estimates that subject PA is depressed, regardless of the rotational speeds for anger, calmness, and joy. In other words, depression indicates a homeostatic bias, where the frequency of sad emotions occurring in subject PA is low. The threshold value α is set in advance and stored in the storage unit 50a. The threshold value α may be set to a different value for each subject PA.

Furthermore, when circulatory systems Ka1-Ka4 express the emotions of anger, calmness, sadness, and joy, respectively, and the rotational speed R3 representing sadness is between threshold α and threshold β (β>α), the estimation unit 30b estimates that the subject PA is in a normal state (i.e., the subject PA is healthy). In other words, the normal condition indicates a state in which sadness and other emotions appropriately occur in the subject PA, and homeostasis is not deviated. The threshold β is pre-set and stored in the storage unit 50a. Alternatively, the threshold β may be set to a different value for each subject PA.

Furthermore, when circulatory systems Ka1-Ka4 indicate the emotions of anger, calmness, sadness, and joy, respectively, and the rotational speed R3 of sadness is greater than threshold β, the estimating unit 30b estimates that the subject PA is bipolar. In other words, bipolar indicates a state in which sadness frequently occurs in the subject PA, resulting in a deviation from homeostasis. Furthermore, the estimating unit 30b estimates a personality disorder when the rotational speed R1 of anger and the rotational speed R4 of joy are equal, regardless of the rotational speeds of calmness and sadness. In other words, a personality disorder indicates a state in which the opposing emotions of anger and joy coexist in the subject PA.

In addition, the circulatory systems Ka1-Ka4 are set to the emotions of anger, normal, sadness, and joy respectively. However, when the condition is panic disorder, the circulatory systems of emotions such as anger, normal, sadness, and joy and the circulatory systems such as heartbeat can be set.

The estimation unit 30b reads data 60a and the medical condition table 70 from the storage unit 50a. Using the read data 60a, the estimation unit 30b calculates the frequency of rotational speeds that satisfy the conditions for each of the circulatory systems Ka1-Ka4 indicated by each medical condition stored in the medical condition table 70, over a predetermined period, such as one day or two weeks. Specifically, for example, if the circulatory systems Ka1-Ka4 represent the emotions of anger, calm, sadness, and joy, the estimation unit 30b calculates the frequency of rotational speeds R1-R4 being 0 (no rotation) for each circulatory system Ka over the predetermined period. Furthermore, the estimation unit 30b calculates the frequency of rotational speeds R3 of the circulatory system Ka3 representing sadness being less than a threshold value α, between threshold values α and β, and greater than threshold value β over the predetermined period. Furthermore, the estimation unit 30b calculates the frequency of rotational speeds R1 of the circulatory system Ka1 representing anger and R4 of the circulatory system Ka4 representing joy being equal over the predetermined period. The frequency of occurrence of the rotation speed of each circulation system Ka within a predetermined period is an example of a pattern of change in the homeostatic state.

For example, the estimating unit 30b extracts conditions indicating a frequency of occurrence exceeding a threshold value Th from the calculated frequency of occurrence. Using the extracted conditions and the medical condition table 70, the estimating unit 30b infers the medical condition that satisfies the combination of the extracted conditions as the medical condition of the subject PA. The prescribed period is determined based on psychiatric standards such as ICD-10. Furthermore, the threshold value Th is pre-set and stored in the storage unit 50a. Furthermore, the threshold value Th may be set to a different value for each subject PA and medical condition.

Furthermore, while the estimating unit 30b calculates the frequency of occurrence of the rotational speeds R1-R4 of each of the circulatory systems Ka1-Ka4, the estimating unit 30b may also calculate the average value and the deviation of the rotational speeds R1-R4 of each of the circulatory systems Ka1-Ka4 over a predetermined period. Furthermore, the estimating unit 30b may compare the time variation of the calculated average value and the deviation of the rotational speeds R1-R4 of each of the circulatory systems Ka1-Ka4 with the typical time variation of the average value and the deviation when the subject PA is healthy, and estimate the condition of the subject PA based on the comparison results.

Figure 29 shows an example of an estimation process performed by the estimation device 100b shown in Figure 25. In addition, steps in the process shown in Figure 29 that represent the same or equivalent processing as the steps shown in Figure 24 are assigned the same step numbers, and detailed descriptions are omitted. Steps S100 to S140, step S150a, and step S160a are implemented by the CPU installed in the estimation device 100b executing the estimation program. In other words, Figure 29 shows another embodiment of the estimation program and estimation method. In this case, the extraction unit 10a, calculation unit 20a, estimation unit 30b, and experimental unit 40a shown in Figure 25 are implemented by executing the estimation program. Alternatively, the process shown in Figure 29 can be implemented by hardware installed in the estimation device 100b. In this case, the extraction unit 10a, calculation unit 20a, estimation unit 30b, and experimental unit 40a shown in Figure 25 are implemented by circuits configured within the estimation device 100b.

After executing the processes of step S100 to step S140 shown in FIG. 29 , the estimation device 100 b executes the process of step S150 a .

In step S150a, the experiment unit 40a detects the rotational speeds R1 to R4 of each of the circulatory systems Ka1 to Ka4 based on the homeostatic simulation executed in step S140, as described in FIG26. The experiment unit 40a stores the detected rotational speeds R1 to R4 of each of the circulatory systems Ka1 to Ka4 as data 60a in the storage unit 50a.

In step S160 a , the estimating unit 30 b estimates the medical condition of the subject PA based on the data 60 a of the rotational speeds R1 - R4 of the circulatory systems Ka1 - Ka4 and the medical condition table 70 , as described with reference to FIG. 27 and FIG. 28 .

Then, the estimation process performed by the estimation device 100b is completed. The process shown in Figure 29 can be repeatedly executed each time an instruction is received from the doctor or the subject PA, or it can be executed at a predetermined frequency. In addition, the estimation device 100b outputs the estimation result to the output device 2. The output device 2 displays the result of the estimated disease state and the deviation from homeostasis. In addition, the output device 2 can also use colors or expressions of animated characters, animals, etc. to indicate the magnitude of the deviation from homeostasis, that is, the degree of the symptoms of the estimated disease state or the degree of health of the subject PA, and display it on the display. In addition, the output device 2 can also display suggestions for treatment methods for the estimated disease state based on the magnitude of the deviation from homeostasis.

In the embodiments shown in Figures 25 to 29 , the deviation from homeostasis of subject PA is calculated using first information representing the physiological state of subject PA and second information representing subject PA's emotions and organ activity. Thus, by referring to the deviation from homeostasis as an indicator, the estimation device 100b can easily estimate the condition of subject PA, even without medical expertise. Furthermore, the estimation device 100b performs a simulation of subject PA's homeostasis using the deviation from homeostasis of each circulatory system Ka as input energy. The estimation device 100b compares the frequency of occurrence of the rotational speed of each circulatory system Ka, which indicates changes in homeostasis, detected during the simulation, with the frequency of occurrence of the rotational speed of each circulatory system Ka when subject PA is healthy. Furthermore, by comparing the comparison results with the condition table 70, the estimation device 100b can estimate the condition of subject PA with higher accuracy than is conventionally possible.

FIG30 illustrates another embodiment of an estimation device. Elements having the same or equivalent functions as those illustrated in FIG25 are designated with the same or equivalent reference numerals, and detailed descriptions thereof are omitted. The estimation device 100c is a computer device having a processing unit such as a CPU and a storage device such as a hard disk. The estimation device 100c is connected to the measuring device 1a and the output device 2a via a wired or wireless interface included in the estimation device 100c. Consequently, the estimation device 100c, the measuring device 1a, and the output device 2a operate as an estimation system SYS.

The output device 2a includes, for example, an organic EL or liquid crystal display, and a speaker for outputting sound. The output device 2a receives the estimated condition of the subject PA from the estimation device 100c and displays the received estimation result on the organic EL display. Furthermore, the output device 2a outputs sound, such as advice, corresponding to the condition estimated by the estimation device 100c. Alternatively, the output device 2a may be located within the estimation device 100c.

The estimation device 100c includes an extraction unit 10a, a calculation unit 20a, an estimation unit 30c, an experiment unit 40a, and a storage unit 50b. The functions of the extraction unit 10a, the calculation unit 20a, the estimation unit 30c, and the experiment unit 40a may be implemented by a program executed by a CPU or by hardware.

The storage unit 50b is a hard disk device, memory, or the like. It stores programs executed by the CPU, data 60a representing the results of simulations performed by the experiment unit 40a, and a medical condition table 70 used by the estimation unit 30c to estimate the medical condition of the subject PA using the data 60a. Furthermore, the storage unit 50b stores an utterance table 80 containing voice data based on the medical condition estimated by the estimation unit 30c, as well as advice to the subject PA. The utterance table 80 is described using FIG31.

The program for executing the estimation process can be recorded on a removable disk such as a CD or DVD and distributed. Alternatively, the estimation device 100c can download the program for executing the estimation process from the network via a network interface included in the estimation device 100c and store it in the storage unit 50b.

The estimating unit 30c estimates the condition of the subject PA based on the pattern of homeostasis changes simulated by the experimenting unit 40a. Furthermore, the estimating unit 30c selects voice data, such as advice, for the subject PA based on the estimated condition of the subject PA and the speech table 80. The operation of the estimating unit 30c will be described using FIG31.

Fig. 31 shows an example of the speech table 80. The speech table 80 has storage areas for disease conditions and speeches.

The condition storage area stores conditions such as severe depression, depression, personality disorder (male), and personality disorder (female). In the case of personality disorder, since males and females are handled differently, utterance table 80 has separate storage areas for personality disorders for males and females. Furthermore, while utterance table 80 lists mental illness as a symptom, it may also include storage areas for heart diseases such as myocardial infarction or other brain diseases such as cerebral infarction.

The speech storage area stores voice data, corresponding to each medical condition stored in the medical condition storage area, including advice for subject PA based on psychiatric standards such as ICD-10. For example, when estimation unit 30c estimates subject PA to be severely depressed, it is estimated that subject PA's depressive symptoms have progressed more severely. Therefore, in order for estimation device 100c to function as a teacher or trainer for subject PA, voice data instructing subject PA, such as "Go to the hospital as soon as possible," is stored in the speech storage area. Furthermore, when estimation unit 30c estimates subject PA to be depressed, it is estimated that subject PA is in a depressed state. Therefore, in order for estimation device 100c to function as a teacher or trainer for subject PA, voice data, such as "You can't stay at home all the time, go for a walk outside once in a while," is stored in the speech storage area to accompany subject PA and strengthen her spirit. Specifically, when subject PA is severely depressed or depressed, the depressive state of subject PA can be improved and the personality of subject PA can be strengthened by storing the voice data of the estimation device 100c as a teacher or trainer of subject PA in the speech storage area.

Furthermore, if subject PA is male and is presumed to have a personality disorder, subject PA tends to be unilaterally aggressive. Therefore, to enable estimation device 100c to function as a counselor for subject PA, voice data, such as "Don't just think about yourself; consider the other person's feelings," is stored in the speech storage area, providing guidance and instruction to subject PA to cultivate empathy. On the other hand, if subject PA is female and is presumed to have a personality disorder, subject PA is likely to engage in self-harm, such as by cutting her wrists. Therefore, to enable estimation device 100c to function as a counselor for subject PA, voice data, such as "I've been working hard, so don't do that," is stored in the speech storage area, providing encouragement and instruction to cultivate empathy. In other words, in cases where subject PA has a personality disorder, storing voice data in the speech storage area, such as "I've been working hard, so don't do that," encourages and instructs subject PA to cultivate empathy.

In addition, the address indicating the area of the storage unit 50b storing the voice data may be stored in the utterance storage area instead of the voice data.

Furthermore, the speech data stored in the speech storage area of the speech table 80 may also be a plurality of speech data with different speech contents for a single medical condition based on psychiatric standards such as ICD-10. For example, the extraction unit 10a extracts the division of each phoneme from the speech signal of the subject PA. That is, when the sound "The weather is nice today (Japanese pronunciation: kyouwaiitenkidesune)" is input, the extraction unit 10a extracts the division of each phoneme such as "kyo/u/wa/i/i/te/n/ki/de/su/ne". Furthermore, the extraction unit 10a extracts the division of each word from the speech signal of the subject PA. For example, when the sound "The weather is nice today" is input, the extraction unit 10a extracts the division of each word such as "kyou/wa/ii/tenki/desune".

The estimation unit 30c then performs recognition and syntactic analysis of each word contained in the subject PA's voice based on information indicating the phoneme and word divisions in the subject PA's voice extracted by the extraction unit 10a. Specifically, the estimation unit 30c identifies the 5W1H information representing "who," "what," "when," "where," "why," and "how" from the subject PA's voice, thereby understanding the content of the subject PA's voice as natural language. Furthermore, based on the understood content of the voice, the estimation unit 30c determines the condition or situation of the subject PA from the voice. Based on the determined condition or situation, the estimation unit 30c selects one piece of voice data from multiple sources, such as advice regarding the estimated medical condition. This enables the estimation device 100c to perform extremely sophisticated processing of the subject PA compared to conventional methods.

Furthermore, by understanding the content of the subject PA's speech, the estimation unit 30c can handle subjects PA experiencing communication difficulties. For example, the estimation unit 30c can infer whether subject PA has a communication difficulty based on the subject PA's emotions extracted by the extraction unit 10a when subject PA uttered a predetermined word. For example, if the extraction unit 10a extracts no or only minimal amounts of anger from subject PA when subject PA utters a predetermined word expressing an emotion such as anger, the estimation unit 30c infers that subject PA is unable to understand the surrounding atmosphere and has a communication difficulty. If a communication difficulty is inferred, the estimation unit 30c reads audio data from the speech storage area, such as "Please pay attention to the surrounding atmosphere," to provide guidance to improve subject PA's communication skills, in order for the estimation device 100c to function as a teacher. Thus, the estimation device 100c can address subject PA's communication difficulty, enabling subject PA to communicate with understanding of the surrounding atmosphere.

Figure 32 shows an example of an estimation process performed by the estimation device 100c shown in Figure 30. In addition, steps in the process shown in Figure 32 that represent the same or equivalent processing as the steps shown in Figure 29 are assigned the same step numbers, and detailed descriptions are omitted. Steps S100 to S140, step S150a, step S160a, and step S170 are implemented by the CPU installed in the estimation device 100c executing an estimation program. In other words, Figure 32 shows another embodiment of the estimation program and estimation method. In this case, the extraction unit 10a, calculation unit 20a, estimation unit 30c, and testing unit 40a shown in Figure 30 are implemented by executing the estimation program. Alternatively, the process shown in Figure 32 can be implemented by hardware installed in the estimation device 100c. In this case, the extraction unit 10a, calculation unit 20a, estimation unit 30c, and testing unit 40a shown in Figure 30 are implemented by circuits configured within the estimation device 100c.

After executing the processes of steps S100 to S140 , S150 a , and S160 a shown in FIG. 32 , the estimation device 100 c executes the process of step S170 .

In step S170, the estimating unit 30c reads voice data of advice etc. to the subject PA based on the medical condition estimated in step S160a and the speech table 80 as described in Fig. 31. The estimating unit 30c outputs the read voice data to the output device 2a.

The estimation process performed by the estimation device 100c then ends. The output device 2a displays the estimated condition and the deviation from homeostasis. Furthermore, the output device 2a outputs the audio data received from the estimation device 100c through a speaker, thereby providing audio advice, etc., corresponding to the estimated condition for the subject PA. Furthermore, the output device 2a may also display the magnitude of the deviation from homeostasis, i.e., the degree of the estimated condition's symptoms or the degree of the subject PA's health, on the display using colors or the expressions of animated characters, animals, etc. Furthermore, the output device 2a may also display animated characters, animals, etc. on the display, and output the received audio data as if the displayed characters or animals were speaking.

Note that the flow shown in FIG. 32 may be repeatedly executed each time an instruction is received from the doctor or the subject PA, or may be executed at a predetermined frequency.

In the embodiments shown in Figures 30 to 32 , the deviation from homeostasis of subject PA is calculated using first information representing the physiological state of subject PA and second information representing subject PA's emotions and organ activity. Thus, by referring to the deviation from homeostasis as an indicator, the estimation device 100c can easily estimate the condition of subject PA, even without medical expertise. Furthermore, the estimation device 100c performs a simulation of subject PA's homeostasis, using the deviation from homeostasis of each circulatory system Ka as input energy. The estimation device 100c compares the frequency of occurrence of the rotational speed of each circulatory system Ka, which indicates changes in homeostasis, detected from the simulation, with the frequency of occurrence of the rotational speed of each circulatory system Ka when subject PA is healthy. Furthermore, by comparing the comparison results with the condition table 70, the estimation device 100c can estimate the condition of subject PA with higher accuracy than conventional methods.

Furthermore, after a suggestion is made, the estimation device 100c can measure the subject PA's physiological state again to estimate the subject PA's condition. Furthermore, the estimation device 100c can evaluate the effectiveness of the suggestion based on the estimation results and, based on the evaluation, modify the content of the suggestion stored in the speech storage area of the speech table 80. This allows the estimation device 100c to perform much more sophisticated processing on the subject PA than is currently possible.

Furthermore, the estimation device 100 (100a, 100b, 100c) is shown as being applicable to psychological counseling, such as psychoanalysis, behavior prediction, and behavioral analysis, psychiatric care, and general medical interviews and prescriptions, but is not limited thereto. For example, the estimation device 100 can also be applied to robotics, artificial intelligence, automobiles, or applications and services for mobile devices such as customer service centers, entertainment, the internet, smartphones, and tablet computers, as well as search systems. Furthermore, the estimation device 100 can also be used in diagnostic devices, automated interview systems, and disaster treatment triage. Furthermore, the estimation device 100 can also be applied to financial credit management systems and behavior prediction, information analysis in businesses, schools, government agencies, police and military, information collection activities, psychological analysis related to lie detection, and organizational management. Furthermore, the estimation device 100 can also be applied to systems for managing the mental health and behavior prediction of organizational members, researchers, practitioners, and managers, systems for controlling environments such as homes and offices, aircraft, and spacecraft, and means for understanding the mental state and behavior prediction of family and friends. Furthermore, the estimation device 100 can also be applied to music and movie distribution, general information retrieval, information analysis management and information processing, or customer preference market analysis, and systems for managing them on a network or a stand-alone computer.

The above detailed description clarifies the features and advantages of the embodiments. It is intended that the claims relate to the features and advantages of the above-mentioned embodiments without departing from their spirit and scope. Furthermore, any technician with ordinary knowledge in this technical field will be able to easily conceive of all improvements and modifications. Therefore, there is no intention to limit the scope of the inventive embodiments to the above-mentioned scope, and appropriate improvements and equivalents included within the scope disclosed in the embodiments may also be used.

Description of Reference Numerals

1. 1a…measuring device; 2. 2a…output device; EU 10. 10a…extraction unit; CU 20. 20a…calculation unit; AU 30. 30a. 30b. 30c…estimation unit; 40. 40a…experimental unit; 50. 50a. 50b…storage unit; 60. 60a…data; 70…patient status table; 80…speech table; AM 100. 100a. 100b. 100c…estimation device; 200. 200a. K1-K10. Ka1-Ka4…circulatory system; B1…joint; PA…subject; SH1-SH10…shaft; NT1-NT10…nut; MG. Ga1-Ga2. Gb1-Gb2. Gc1. Gd1-Gd2…gear; SYS…estimation system.

Claims (7)

1. A estimation device, characterized in that it comprises: The extraction unit extracts, from information representing the subject's physiology, first information representing the subject's physiological state and second information representing at least one of the subject's emotions and organ activity; The computation unit calculates the similarity between the extracted first information and the time changes shown by the second information, and based on the calculated similarity, calculates the deviation from the predetermined state of homeostasis of the subject; and The estimation unit estimates the subject's condition based on the calculated deviation. 2. The estimation device as described in claim 1, characterized in that, The estimation device also includes an experimental unit, which calculates the energy acting on the subject's emotions and organ activities based on the deviation calculated by the calculation unit, and uses the calculated energy as input to simulate the subject's homeostasis. The estimation unit estimates the subject's condition based on the simulated pattern of changes in the homeostasis. 3. The estimation device as described in claim 1 or 2, characterized in that, The estimation device includes an input unit that receives sound signals from the subject as information representing the subject's physiology. 4. The estimation device as described in claim 1 or 2, characterized in that, The estimation device includes an input unit that receives the activity of the subject's organs as information representing the subject's physiology. 5. The estimation device as described in claim 1 or 2, characterized in that, The estimation device includes a storage unit that stores audio data of suggested treatments for the subject for each condition. The estimation unit selects audio data representing recommendations for the subject based on the estimated condition of the subject, and outputs the selected audio data to an external output device. 6. A recording medium, characterized in that it records a program that causes a computer to perform the following processes: From the physiological information of the subject, extract first information representing the subject's physiological state and second information representing at least one of the subject's emotions and organ activity. The similarity between the extracted first information and the second information over time is calculated, and based on the calculated similarity, the deviation from the specified state of homeostasis of the subject is calculated. The subject's condition is inferred based on the calculated deviation. 7. A estimation system, characterized in that it comprises: The measuring device is used to measure the physiological functions of the test subject. A presumption device that presumes the condition of a subject by utilizing physiological information representing the subject as measured by the measuring device; and The output device outputs the result of the condition estimated by the estimation device. The estimation device includes: The extraction unit extracts, from information representing the subject's physiology, first information representing the subject's physiological state and second information representing at least one of the subject's emotions and organ activity; The computation unit calculates the similarity between the extracted first information and the time changes shown by the second information, and based on the calculated similarity, calculates the deviation from the predetermined state of homeostasis of the subject; and The estimation unit estimates the subject's condition based on the calculated deviation.