Health management support system, health management support device, health management support method, and program

Abstract

To classify the types of capillaries by deep learning.SOLUTION: The health management support system includes: an image acquisition unit configured to acquire a nail-bed capillary image of a user; a capillary type classification unit configured to classify which type, among a plurality of types, the capillary in the nail-bed capillary image belongs to; and a health condition determination unit configured to determine the health condition of the user on the basis of the classified type.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a health management support system, a health management support device, a health management support method, and a program. [Background technology]

[0002] In clinical practice, capillary blood flow observations, such as those in the nail bed, are performed when diagnosing autoimmune diseases such as scleroderma and rheumatoid arthritis. Furthermore, capillary images are highly correlated with lifestyle habits, and are expected to be applied to the diagnosis and treatment of pre-disease conditions. Observations of capillaries are recorded by photograph using a microscope, and recently quantification has been promoted by combining this with analytical software.

[0003] For example, Patent Document 1 discloses a capillary imaging system that includes an image acquisition unit that acquires multiple images from an imaging unit that captures multiple images of capillaries in an organism, a focus evaluation value calculation unit that calculates a focus evaluation value that quantifies the degree of focus for each of the multiple images, and a focused image selection unit that selects an image with a good degree of focus from the multiple images as a focused image by comparing each focus evaluation value with a judgment threshold for determining whether the degree of focus is good or not.

[0004] Furthermore, for example, Patent Document 2 discloses a blood abnormality prediction device that includes at least an image receiving unit that receives image information of the capillary heads, and a prediction unit that predicts the presence or absence of blood lipid abnormalities in a subject based on the image information, and the prediction unit measures the arterial limb width and / or venous limb width of the capillary heads based on the image information, and predicts the presence or absence of blood lipid abnormalities in the subject based on the measurement results.

[0005] Furthermore, for example, Patent Document 3 discloses a diagnosis support system that supports the diagnosis of a subject's health condition based on a capillary blood vessel image of the subject, and includes a server, a first terminal connectable to the server via a network, and a second terminal that may be the same as or different from the first terminal and connectable to the server via the network, the first terminal having an imaging means for capturing an image of the subject's capillaries to obtain a capillary blood vessel image, and an image transmitting means for transmitting the capillary blood vessel image to the server, the server having a questionnaire transmitting means for transmitting a questionnaire about the subject's lifestyle habits to the second terminal in response to a request from the second terminal, and an answer receiving means for receiving answers to the questionnaire by the subject transmitted from the second terminal, and The diagnostic support system disclosed includes an image receiving means for receiving a transmitted capillary image, a feature extraction means for extracting features of the capillaries from the capillary image received by the image receiving means, a class determination means for determining which of a plurality of pre-set classes the state of the capillaries represented by the received capillary image falls into based on the extracted features and the responses to the received questionnaire, and a data transmitting means for transmitting data including the result of the class determination to a first terminal as transmission data, wherein the first terminal further includes a data receiving means for receiving the transmission data transmitted from the data transmitting means and a data display means for displaying the received transmission data.

[0006] In these prior arts, the presence or absence of a subpapillary vascular plexus running laterally across the finger's longitudinal direction (vertical direction), the degree of turbidity of the body fluid, and the degree of curvature of the tip of the blood vessels running longitudinally of the finger are determined based on the capillary vessel image, and the health condition is evaluated based on the determination results and the result is displayed. It is also disclosed that in order to improve the accuracy of the capillary vessel diagnosis, the accuracy can be improved by combining the subject's lifestyle questionnaire data with the feature amount to diagnose the subject's health condition. [Prior art documents] [Patent documents]

[0007] [Patent Document 1] Japanese Patent Publication No. 2023-120845 [Patent Document 2] Japanese Patent Publication No. 2022-98405 [Patent Document 3] Patent Publication No. 2021-189828 Summary of the Invention [Problem to be solved by the invention]

[0008] However, although the capillary shape is evaluated by measuring the venous limb width, arterial limb width, loop diameter, etc. of the capillaries, the measurement values ​​of the capillary shape from these measurements are limited to the images used to judge the results, and there is a large difference in the measurement values ​​between different parts of the same subject, so the accuracy of the judgment is insufficient. It is also desirable to be able to maintain and manage health based on biological information, and to be able to provide suggestions for effective health promotion and disease prevention.

[0009] The present invention has been made in consideration of this situation, and aims to provide a health management support system, a health management support device, a health management support method, and a program that can classify capillary shape types using deep learning. [Means for solving the problem]

[0010] In order to solve the above-mentioned problems of the present invention, one aspect of the present invention is a health management support system comprising an image acquisition unit that acquires an image of capillaries in the nail bed of a user, a capillary type classification unit that classifies the capillaries in the nail bed capillary image as to which of multiple types they are, and a health condition determination unit that determines the health condition of the user based on the classified type.

[0011] Another aspect of the present invention is a health management support system that includes a finger image acquisition unit that acquires a finger image of a user, a capillary type classification unit that classifies the user's capillaries into one of multiple types based on the finger image, and a health condition determination unit that determines the health condition of the user based on the classified type.

[0012] Another aspect of the present invention is a health management support device that includes an image acquisition unit that acquires an image of capillaries in the nail bed of a user, a capillary type classification unit that classifies the capillaries in the image of capillaries in the nail bed as to which of multiple types they belong to, and a health condition determination unit that determines the health condition of the user based on the classified type.

[0013] Another aspect of the present invention is a health management support device that includes a finger image acquisition unit that acquires a finger image of a user, a capillary type classification unit that classifies the user's capillaries into one of multiple types based on the finger image, and a health condition determination unit that determines the health condition of the user based on the classified type.

[0014] Another aspect of the present invention is a health management support method executed by a computer, the health management support method including an image acquisition process for acquiring a capillary image of a user's nail bed, a capillary type classification process for classifying the capillaries in the nail bed capillary image as to which of multiple types the capillaries belong to, and a health condition determination process for determining the health condition of the user based on the classified type.

[0015] Another aspect of the present invention is a health management support method executed by a computer, the health management support method including: a finger image acquisition process for acquiring a finger image of a user; a capillary type classification process for classifying the user's capillaries into one of multiple types based on the finger image; and a health condition determination process for determining the health condition of the user based on the classified type.

[0016] Another aspect of the present invention is a program for causing a computer to execute an image acquisition means for acquiring an image of capillaries in the nail bed of a user, a capillary type classification means for classifying the capillaries in the image of capillaries in the nail bed as to which of multiple types they belong to, and a health condition determination means for determining the health condition of the user based on the classified type.

[0017] Another aspect of the present invention is a program for causing a computer to execute a finger image acquisition means for acquiring an image of a user's finger, a capillary type classification means for classifying the user's capillaries into one of multiple types based on the finger image, and a health condition determination means for determining the health condition of the user based on the classified type. [Effects of the Invention]

[0018] According to the present invention, the type of capillary shape can be classified by deep learning. [Brief explanation of the drawings]

[0019] [Figure 1] 1 is a block diagram showing an example of the configuration of a health management support device 1 according to an embodiment of the present invention. [Figure 2] FIG. 2 is a diagram showing an example of a nail bed capillary image according to the present embodiment. [Figure 3] FIG. 10 is a diagram showing another example of the nail bed capillary image according to the present embodiment. [Figure 4] FIG. 10 is a diagram showing an example of nail bed capillary type classification according to the present embodiment. [Figure 5] FIG. 10 is a diagram showing an example of changes over time in the type of nail bed capillaries according to the present embodiment. [Figure 6] FIG. 10 is a diagram showing an example of the relationship between nail bed capillary types, age, and gender according to the present embodiment. [Figure 7] FIG. 10 is a diagram showing an example of the relationship between nail bed capillary types and age according to the present embodiment. [Figure 8] FIG. 10 is a diagram showing an example of the relationship between the type of nail bed capillaries and gender according to the present embodiment. [Figure 9] FIG. 10 is a diagram showing an example of dietary improvement suggestions according to the type of nail bed capillaries according to the present embodiment. [Figure 10] 10 is a flowchart showing an example of a nail bed capillary type classification process and a health condition determination process in the health management support device 1 according to the present embodiment. [Figure 11] FIG. 10 is a diagram showing an example of nail bed capillary type classification by analysis based on measurement classification according to the present embodiment. [Figure 12] FIG. 10 is a diagram showing another example of nail bed capillary type classification by analysis based on measurement classification according to the present embodiment. [Figure 13] FIG. 10 is a diagram showing another example of nail bed capillary type classification by analysis based on measurement classification according to the present embodiment. [Figure 14] FIG. 10 is a diagram showing an example of the relationship between nail bed capillary types and menopausal symptoms based on cross-tabulation analysis according to the present embodiment. [Figure 15] FIG. 10 is a diagram showing an example of the relationship between the type of nail bed capillaries and each hormone according to the present embodiment. [Figure 16] 1 is a block diagram showing an example of a hardware configuration of a health management support device 1 according to the present embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0020] Hereinafter, embodiments of the present invention will be described with reference to the drawings.

[0021] [Overview of the Health Management Support System SYS] The health management support system SYS according to this embodiment is a system that acquires an image of a target user TU, classifies the capillaries into one of multiple types based on the acquired image, and determines the health condition of the target user TU based on the classified type. The image is, for example, an image of capillaries in a nail bed or an image of a finger including a nail bed.

[0022] In the following description, an example will be described in which the health management support system SYS is configured with one device as the health management support device 1, but the functions of the health management support device 1 may be realized by multiple devices.

[0023] [Configuration of health management support device 1] FIG. 1 is a block diagram showing an example of the configuration of a health management support device 1 according to this embodiment. The health management support device 1 includes a communication unit 10, a control unit 12, and a storage unit . The communication unit 10 has a function of communicating with other devices via a network, either wired or wirelessly. The communication unit 10 outputs signals or data received from other devices in the health management support system SYS to the control unit 12. The communication unit 10 transmits signals or data output from the control unit 12 to other devices in the health management support system SYS.

[0024] Control unit 12 has a function of controlling each unit of health management support device 1. Control unit 12 is configured to include a user information acquisition unit 120, an image acquisition unit 121, a capillary type classification unit 122, a health condition determination unit 123, a recommendation unit 124, a notification unit 125, and a learning unit 126.

[0025] The user information acquisition unit 120 acquires user information of the target user TU from the target user TU. The user information is, for example, information indicating the age and gender of the target user. The user information acquisition unit 120 stores the acquired user information in the storage unit 14 in association with identification information that identifies the user.

[0026] The image acquisition unit 121 acquires an image of the target user TU. The image acquisition unit 121 will be described more specifically. The image acquisition unit 121 includes a capillary image acquisition unit 1211 and a finger image acquisition unit 1212 .

[0027] The capillary image acquisition unit 1211 acquires a nail bed capillary image of the target user TU. The nail bed capillary image acquired by the capillary image acquisition unit 1211 may be an image captured by an imaging device, or a nail bed capillary image captured in advance may be acquired from a storage medium connected to the health management support device 1 or from another device via a network. The capillary image acquisition unit 1211 stores the acquired nail bed capillary image in the storage unit 14 in association with identification information that identifies the user.

[0028] The finger image acquisition unit 1212 acquires a finger image including a nail bed of the target user TU. The finger image acquired by the finger image acquisition unit 1212 may be a finger image captured by an imaging device, or a finger image captured in advance may be acquired from a storage medium connected to the health management support device 1 or from another device via a network. The finger image acquisition unit 1212 stores the acquired finger image in the storage unit 14 in association with identification information that identifies the user.

[0029] The capillary type classification unit 122 classifies the shape of the capillaries in the capillary image into one of multiple types based on the nail bed capillary image of the target user TU read from the storage unit 14. Specifically, the capillary type classification unit 122 inputs the nail bed capillary image of the target user TU into a learning model that has learned the correspondence between capillary shapes and classified types, thereby classifying the capillary type by indicating the probability of which of multiple capillary types the capillary type is. Specifically, the capillary type classification unit 122 can classify one or more types and their probabilities, such as type 〇 being 85% and type 〇〇 being 12%, rather than simply presenting one type. The capillary type classification unit 122 stores the classified capillary type and its probability in the storage unit 14 in association with identification information that identifies the user.

[0030] Health condition determination unit 123 reads out user information and one or more types (capillary types) classified by capillary type classification unit 122 and their probabilities from storage unit 14. Health condition determination unit 123 determines the health condition based on the user information and one or more types (capillary types) and their probabilities. The health condition determination by the health condition determination unit will be described later. Health condition determination unit 123 stores the determination result in storage unit 14 in association with identification information that identifies the user.

[0031] The recommendation unit 124 generates health improvement proposal information for proposing health improvements to the target user TU based on the determination result by the health condition determination unit 123. The health improvement proposals include proposals for improving disease risks, proposals for improving lifestyle habits, proposals for improving diet, proposals for improving exercise, and the like.

[0032] The notification unit 125 notifies the target user TU and / or other users designated by the target user of the determination result by the health condition determination unit 123. The other users may be, for example, the target user TU's family, a care manager, a healthcare provider, a medical institution, a medical examination provider, etc. Furthermore, the notification unit 125 notifies the determination result by the health condition determination unit 123 and the health improvement proposal information by the recommendation unit 124 to the target user TU and / or other users designated by the target user.

[0033] The learning unit 126 generates a learning model. The learning model according to this embodiment is a learning model that learns the correspondence between the shape of capillaries in a learning nail bed capillary image and the type of capillaries to be classified. When a nail bed capillary image is input, the learning model outputs one or more types (capillary types) according to the shape of the capillaries and their probabilities. The learning model according to this embodiment is a learning model that learns the correspondence between the skin color and nail color in a learning finger image including a nail bed and the type of capillary blood vessels to be classified. When a finger image including a nail bed is input, the learning model outputs one or more types (capillary blood vessel types) and their probabilities. The learning unit 126 stores the generated learning model in the storage unit 14.

[0034] The storage unit 14 stores a learning model 141 , capillary image information 142 , finger image information 143 , health condition data 144 , recommendation information 145 , and user information 146 . The learning model 141 is a learning model generated by the learning unit 126 . The capillary vessel image information 142 is information in which the nail bed capillary vessel image acquired by the capillary vessel image acquisition unit 1211 is associated with identification information for identifying the target user TU. The finger image information 143 is information in which a finger image including a nail bed portion acquired by the finger image acquisition unit 1212 is associated with identification information for identifying the target user TU. The health condition data 144 is data in which identification information for identifying the target user TU and the determination results by the health condition determination unit 123 are associated in chronological order. The recommendation information 145 is various information used by the recommendation unit 124 to generate health improvement proposal information. The user information 146 is information in which identification information for identifying the target user TU is associated with the age and sex of the target user TU.

[0035] [Classification of capillary shape types] Next, the classification of capillary shape types will be described.

[0036] Fig. 2 is a diagram showing an example of a nail bed capillary image according to this embodiment, and Fig. 3 is a diagram showing another example of a nail bed capillary image according to this embodiment. The capillary vessel image TO is a mature, normal image (type O), and is a type 0 capillary vessel image classified as normal. The capillary image T1 is a mature, dilated, branched, and curved image of a capillary vessel, classified as a torsion type, or Type 1. The capillary image T2 is a mature, macroscopic image of dilated and enlarged capillaries, and is a type 2 capillary image classified as a stasis type. Capillary vessel image T3 is a mountain-shaped image of intermediate hoof base dilation, and is a type 3 capillary vessel image classified as underdeveloped. Capillary vessel image T4 is an intermediate subpapillary vascular plexus image of arteriovenous plexus formation, and is a type 4 capillary vessel image classified as a horizontal line type. The capillary vessel image T5 is a primitive type image of only the head of the capillary vessel (head of the capillary vessel image), and is a type 5 capillary vessel image classified as a tip type. To classify capillary vessel images, for example, it is possible to use Keras, a deep learning library, to read capillary vessel images labeled with six types, create a model, and perform training.

[0037] The characteristics of each type will be explained with reference to Figure 4. Large-scale health-related data collected in 2021 and 2022 [n=536 (2021 data), n=737 (2022 data)] were used as the analysis data. The parentheses in the figure indicate the test items on which the results were based. FIG. 4 is a diagram showing an example of the classification of nail bed capillary types according to this embodiment. Users classified as Type 0 have characteristics such as flexible blood vessels (CAVI), high mental and physical QOL (SF36), no tendency to overeat and a well-balanced diet (FFQ), low alcohol consumption, a preference for wine (FFQ), a preference for stewed meat (especially men), and a risk of low blood pressure (women). Users classified as Type 1 tend to overeat and are affected by stress (ACTH, IgA), drink a lot (sake and beer) and are easily affected by alcohol (glycation index), are overweight (BMI), and have issues with nutritional balance.

[0038] Users classified as Type 2 have characteristics such as polycythemia (red blood cells, Hb, Ht), cardiovascular risk (homocysteine), a tendency to overeat and have low levels of beta-carotene and antioxidant vitamins and high fat intake (FFQ), a protruding belly (body type), issues with nutritional balance, a tendency to exercise less, and a high risk of high blood pressure. Users classified as type 3 tend to be anemic (red blood cells, Hb, Ht), have inflammation in men, low iron levels in women, low blood 25OH vitamin D3 levels and blood calcium levels, have low mineral intake (FFQ), are physically inactive, and are at high risk of arteriosclerosis and hyperlipidemia.

[0039] Users classified as Type 4 tend to have characteristics such as high blood pressure and other factors that inhibit peripheral blood flow, which leads to angiogenesis (VEGF), stress (cortisol), urinary problems (men), a tendency toward low physical QOL (SF36), susceptibility to the effects of glycation and oxidation due to overeating (FFQ) (glycation and oxidation index), a tendency toward lack of exercise, a high risk of anemia, a high risk of hyperlipidemia, and a high risk of fractures. Users classified as Type 5 tend to have high levels of arterial stiffness (CAVI), are at risk of lifestyle-related diseases, are stressed (ACTH), are prone to allergies (IgE), tend to have a low QOL (SF36, women), tend to overeat and prefer strong alcohol (especially men), consume little dairy products (FFQ), are prone to obesity (body type), have issues with nutritional balance, tend to exercise less, and have a high risk to the immune system.

[0040] [Changes in capillary type over time] Next, we will explain the change in capillary type over time. Capillary types change over time depending on diet, exercise, etc. The factors related to each type and the effects of nutrition and diet are explained with reference to Figure 5. FIG. 5 is a diagram showing an example of changes in the type of nail bed capillaries over time according to this embodiment.

[0041] Factors related to the change from type 0 to type 1 include hyperemia (red blood cells, MCV), inhibition of blood coagulation and vasoconstriction (serotonin), and basement membrane synthesis (type IV collagen 7s). The nutritional and dietary influence related to the change from type 0 to type 1 is an increase in salt intake. Factors related to the transition from Type 0 and Type 1 to Type 2 include fluid metabolism (increased urinary sodium) and a decrease in serum zinc. Nutritional and dietary influences related to the transition from Type 0 and Type 1 to Type 2 include a decrease in dietary fiber intake and an increase in meat intake.

[0042] Factors associated with the progression from Type 0, Type 1, and Type 2 to Type 3 include anemia (MCV), increased bone metabolism (total P1NP), and electrolytes (increased urinary sodium). The nutritional and dietary influence associated with the progression from Type 0, Type 1, and Type 2 to Type 3 is a decrease in electrolyte intake. Factors related to the change from type 0, type 1, type 2, and type 3 to type 4 include blood lipids (apoprotein E) and iron metabolism (ferritin). There are no particular nutritional or dietary influences related to the change from type 0, type 1, type 2, and type 3 to type 4.

[0043] Factors related to the progression from type 0, type 1, type 2, type 3, and type 4 to type 5 include obesity (BMI), lifestyle-related diseases (multiple indicators), inflammation and immunity (neutrophil / lymphocyte ratio, granulocytes), etc. Nutritional and dietary influences related to the progression from type 0, type 1, type 2, type 3, and type 4 to type 5 include decreased calcium intake, increased iron intake, decreased dairy intake, and increased alcohol intake. Factors related to the change from Type 1, Type 2, Type 3, Type 4, and Type 5 to Type 0 include electrolyte and fluid retention (serum chloride). Nutritional and dietary influences related to the change from Type 1, Type 2, Type 3, Type 4, and Type 5 to Type 0 include increased protein intake, decreased rice intake, and decreased alcohol intake.

[0044] [Relationship between capillary type, age, and gender] Next, the relationship between capillary type, age, and gender will be explained. FIG. 6 is a diagram showing an example of the relationship between the type of nail bed capillaries, age, and sex according to this embodiment. Type 0 users under the age of 50 have a high female ratio, good QOL (SF36), flexible blood vessels (CAVI), and do not overeat. Type 0 users over the age of 50 have characteristics such as a good quality of life (SF36), flexible blood vessels (CAVI), and do not overeat.

[0045] Type 1 users under the age of 50 are characterized by high levels of 25OHVD3 and heavy alcohol consumption. Type 1 users over the age of 50 are characterized by a tendency toward inflammation, overeating, low intake of vitamin K, and heavy alcohol consumption.

[0046] Type 2 users under the age of 50 were characterized by a high male ratio (significant difference in test values), overeating, low intake of vitamin K, and low intake of beta-carotene equivalents and alpha-tocopherol. Type 2 users over the age of 50 are characterized by increased inflammation and metabolism, low intake of vitamin K, low intake of beta-carotene equivalents and alpha-tocopherol, and heavy alcohol consumption.

[0047] Type 3 users under the age of 50 are characterized by a high proportion of women (low red blood cell and hemoglobin levels), low serum calcium levels, low growth factor levels, small meals, and low niacin intake. Type 3 users aged 50 or over are characterized by high LOX-index (registered trademark) (sLOX-1), low serum calcium, low growth factor levels, small meals, and high dietary fiber intake.

[0048] Type 4 users under the age of 50 are characterized by a high female ratio (low red blood cell and hemoglobin levels), high TSH levels (hypothyroidism), low LOX-index (sLOX-1), overeating, low niacin intake, high dietary fiber intake, and high dairy intake. Type 4 users aged 50 and over are characterized by high LOX-index (sLOX-1), a tendency for high TSH levels, overeating, high intake of dietary fiber, and high intake of dairy products.

[0049] Type 5 users under the age of 50 have characteristics such as a high LOX-index (sLOX-1), a tendency to obesity, decreased liver function, high growth factor levels, good QOL (SF36), low calcium and phosphorus intake, low dairy intake, and heavy alcohol consumption. Type 5 users over the age of 50 tend to have increased inflammation and metabolism, high serum osmolality, low growth factor levels, a reduced QOL (SF 36, CAVI vascular age, locomotive syndrome 25), low calcium and phosphorus intake, low dairy intake, and a preference for strong alcohol.

[0050] [Relationship between capillary type and age] Next, the relationship between the capillary type and significant items of disease risk based on age will be explained. FIG. 7 is a diagram showing an example of the relationship between nail bed capillary types and age according to this embodiment. The significant risk items by type shown in Figure 7 are the results of a cross-tabulation analysis. Type 0 users as a whole are at risk of hypotension*, etc. Here, hypotension* is defined as, for example, a systolic blood pressure of less than 100. Type 0 users over 50 years of age are at risk for low blood pressure*.

[0051] Type 1 users over the age of 50 tend to not pay attention to nutritional balance.

[0052] Type 2 users as a whole tend to have high plasma renin and inactivity**, where inactivity** is defined as not even doing light exercise, for example. Type 2 users under the age of 50 are at risk for high plasma renin and other conditions. Type 2 users over the age of 50 tend to be physically inactive.

[0053] Type 3 users as a whole tend to be at risk of spinal stenosis and lack of exercise. Type 3 users under the age of 50 are at risk for low MCH and other conditions. Type 3 users over the age of 50 are at risk of high LOXindex, hyperlipidemia, and spinal stenosis.

[0054] Type 4 users as a whole tend to have high levels of cystatin C, are at risk of fractures, lower back pain, and nerve paralysis, lack exercise, and do not snack. Type 4 users under the age of 50 tend to be at risk for low MCH and lack of exercise. Type 4 users over the age of 50 tend to have high ALT levels, hyperlipidemia, and a higher risk of fractures and back pain, and tend not to snack.

[0055] Type 5 users as a whole tend to have high cystatin C, high lymphocyte counts, high CD4+CD8+, high CD4+CD8-, heart disease (angina pectoris), risk of malignant neoplasms, and lack of exercise. Type 5 users under the age of 50 tend to be at risk for high CD4+CD8- and lack of exercise. Type 5 users over the age of 50 tend to have high cystatin C, high white blood cell count, high CD4+CD8-, heart disease (angina pectoris), risk of malignant neoplasms, and a lack of consideration for nutritional balance.

[0056] [Relationship between capillary type and gender] Next, the relationship between the capillary type and the significant items of disease risk based on gender will be explained. FIG. 8 is a diagram showing an example of the relationship between the type of nail bed capillaries and gender according to this embodiment. The significant risk items by type shown in Figure 8 are the results of a cross-tabulation analysis.

[0057] Male Type 0 users are at risk for high white blood cell counts and other conditions. Female users of Type 0 are at risk of developing a body temperature below 36 degrees and low blood pressure. Male users of type 1 tend to smoke.

[0058] Type 2 male users are at risk of developing rough hands. Type 2 female users are at risk for high cystatin C, high osmolality, and high total PAI-1. Type 3 male users are at risk of developing rough hands and low white blood cell counts.

[0059] Type 4 male users are at risk for high C3, high pentosidine, and rough handling. Type 4 female users tend to be at risk of nerve paralysis, high RCAVI, fractures and back pain, and do not snack.

[0060] Male users of type 5 are at risk for high CD4-CD8- and other conditions. Type 5 female users are at risk for heart disease (angina pectoris), osteoarthritis, knee and hip joint disease, malignant neoplasms, high lymphocyte counts, high CD4-CD8+ counts, and high sLOX-1 levels.

[0061] [Relationship between capillary types and dietary improvement suggestions] Next, the relationship between capillary types and dietary improvement suggestions will be explained. FIG. 9 is a diagram showing an example of dietary improvement suggestions according to the type of nail bed capillaries according to this embodiment. Type 0 users are advised to maintain their current diet. Type 1 users are advised to be careful about eating and drinking too much. Type 2 users are advised to be careful not to overeat or drink too much. Type 2 users are also advised to consume foods rich in vitamin K (dark green vegetables, seaweed, green tea), beta-carotene (green and yellow vegetables and citrus fruits), and alpha-tocopherol (nuts and seeds such as almonds, oils and fats, grains, seafood, beans, and vegetables).

[0062] For Type 3 users, we recommend that you take adequate nutritional advice and strive to eat a balanced diet that includes calcium supplements and foods rich in vitamins like niacin, such as seafood, meats including liver, mushrooms, grains, and legumes.

[0063] Type 4 users are advised to be careful about overeating and to maintain a balanced diet. Type 4 users are also advised to consume seafood, meat including liver, mushrooms, grains, and beans, which are rich in vitamins such as niacin.

[0064] Type 5 users are advised to consume dairy products that are rich in calcium and phosphorus, and to consume animal products such as seafood and meat in addition to dairy products.

[0065] [Processing flow of health management support device 1] Next, the flow of processing by the health management support device 1 (health management support system SYS) will be described. FIG. 10 is a flowchart showing an example of the nail bed capillary type classification process and health condition determination process in the health management support device 1 according to this embodiment.

[0066] In step S101, the image acquisition unit 121 acquires a nail bed capillary image or a finger image including a nail bed of the target user TU. Next, the health management support apparatus 1 executes the process of step S103.

[0067] In step S103, the user information acquisition unit 120 acquires user information indicating the age and sex of the target user TU. Next, the health management support apparatus 1 executes the process of step S105.

[0068] In step S105, the capillary type classification unit 122 inputs the nail bed capillary image or the finger image including the nail bed of the target user TU into the learning model, and obtains as output a classification result indicating which capillary type the target user TU is. Next, the health management support apparatus 1 executes the process of step S107.

[0069] In step S107, health condition determination unit 123 determines the health condition of target user TU by referring to the type classification result of the capillary image based on the nail bed capillary image or the finger image including the nail bed, the user information, and the various relationships between capillary type and age and gender shown in Figures 5 to 8. Next, health management support device 1 executes the process of step S109.

[0070] In step S109, the notification unit 125 notifies the target user TU and / or another user designated by the target user of the determination result by the health condition determination unit 123. Next, the health management support apparatus 1 executes the process of step S111.

[0071] In step S111, the recommendation unit 124 generates health improvement suggestion information for suggesting health improvements to the target user TU based on the judgment result by the health condition judgment unit 123, by referring to various correspondence relationships such as the correspondence relationship between capillary types and dietary improvement suggestions in Figure 9. The notification unit 125 notifies (recommends) the health improvement proposal information provided by the recommendation unit 124 to the target user TU and / or other users designated by the target user. After that, the nail bed capillary type classification process and health condition determination process shown in FIG. 10 are terminated.

[0072] It is also possible to skip the generation and notification of health improvement suggestion information in step S111 and end the nail bed capillary type classification process and health condition determination process in FIG. 10 after the process in step S109.

[0073] As described above, the health management support device 1 (health management support system SYS) according to this embodiment includes an image acquisition unit (capillary image acquisition unit 1211) that acquires a capillary image of the user's nail bed, a capillary type classification unit 122 that classifies the capillaries in the nail bed capillary image as to which of multiple types they belong to, and a health condition determination unit 123 that determines the user's health condition based on the classified type.

[0074] This allows for classification of capillary types using deep learning. Furthermore, subtle changes in capillary images over time can be used to understand treatment effects and prognosis. Therefore, deep learning can be used to classify the shape of nail bed capillaries, extract the causes of shape changes, and present individualized treatment methods. This makes it possible to predict the treatment effects and prognosis of diseases corresponding to the condition of nail bed capillaries, and present effective ways to improve quality of life (QOL), promote health, and prevent risk diseases.

[0075] Furthermore, the health management support device 1 (health management support system SYS) according to this embodiment includes a finger image acquisition unit 1212 that acquires a finger image of the user, a capillary type classification unit 122 that classifies the user's capillaries into one of multiple types based on the finger image, and a health condition determination unit 123 that determines the health condition of the user based on the classified type.

[0076] This allows for classification of capillary types using deep learning. Furthermore, subtle changes in capillary images accompanying changes in finger images over time can be used to understand treatment effects and prognosis. Therefore, deep learning can be used to classify the shape of finger images of the nail bed, extract the causes of shape changes, and present individualized treatment methods. This makes it possible to predict the treatment effects and prognosis of diseases corresponding to the state of the nail bed capillaries, and present effective ways to improve quality of life (QOL), promote health, and prevent risk diseases.

[0077] In addition, the health management support device 1 (health management support system SYS) according to this embodiment further includes a user information acquisition unit 120 that acquires the user's age and gender as user information, and a health condition determination unit 123 that determines the user's health condition based on the type of capillaries and the user information.

[0078] This allows the type of capillary blood vessels, along with gender and age, to be used to classify health status, disease diagnoses and potential risks.

[0079] Furthermore, in the health management support device 1 (health management support system SYS) according to this embodiment, the health condition determination unit 123 obtains an output representing the health condition of the user by inputting the image of the capillary vessels in the nail bed of the user into a learning model that has learned the correspondence between the image of the capillary vessels in the nail bed and data representing the health condition.

[0080] Moreover, the health management support device 1 (health management support system SYS) according to this embodiment further includes a notification unit that notifies the user, another user designated by the user, or both of an output indicating the user's health condition.

[0081] This allows the user and other users designated by the user to be notified of the results.

[0082] Furthermore, in the health management support device 1 (health management support system SYS) according to this embodiment, the capillary type classification unit 122 classifies the type of the user's capillaries by inputting the user's finger image into a learning model that has learned the correspondence between the nail bed capillary image and the finger image.

[0083] This allows for classification of capillary types using deep learning.

[0084] Moreover, the health management support device 1 (health management support system SYS) according to this embodiment further includes a recommendation unit 124 that makes recommendations to the user according to the user's health condition.

[0085] This allows for suggestions for improvement based on the health condition.

[0086] [Modification] Next, a modified example will be described. The health condition determination unit 123 may input the user's nail bed capillary image into a learning model that has learned the correspondence between the nail bed capillary image or the finger image including the nail bed and the data representing the time series health condition, thereby obtaining an output representing the effect of treatment on the user's health condition. In this case, a learning model that learns the correspondence between nail bed capillary images or finger images including nail beds and data representing time-series health conditions may be generated in the learning unit 126. Specifically, the model may learn the correspondence between changes in capillary types over time and various related factors, as shown in Fig. 5, and output the effects of nutrition and diet according to changes in capillary types.

[0087] In this way, in the health management support device 1 (health management support system SYS) according to this embodiment, the health condition determination unit 123 can input the image of the capillaries in the nail bed of the user into a learning model that has learned the correspondence between the image of the capillaries in the nail bed and data representing the time series of the health condition, thereby obtaining an output representing the effect of treatment on the health condition of the user.

[0088] The recommendation unit 124 may recommend at least one of supplements, Chinese herbal medicine, medicine, food, meal menu, and exercise menu to the user depending on the health condition and capillary type of the user. In this case, the recommendation unit 124 may recommend to the user based on the correspondence between the type of capillary blood vessel and the recommended item, such as the correspondence between the type of capillary blood vessel and supplements, the correspondence between the type of capillary blood vessel and Chinese herbal medicine, or the correspondence between the type of capillary blood vessel and medicine, as shown in FIG.

[0089] In this way, in the health management support device 1 (health management support system SYS) of this embodiment, the recommendation unit 124 can recommend at least one of supplements, Chinese medicine, medicines, food, meal menus, and exercise menus to the user depending on the user's health condition.

[0090] This allows for recommendations that are suitable for improving health status.

[0091] In the above description, the capillary type classification unit 122 classifies capillaries into six types, namely, type 0, type 1, type 2, type 3, type 4, and type 5, but the present invention is not limited to this. For example, it may be possible to classify into multiple types, such as nine types, as follows: The following description will be given with reference to FIGS.

[0092] Fig. 11 is a diagram showing an example of the classification of nail bed capillary types by analysis based on the measurement classification according to this embodiment. Fig. 12 is a diagram showing another example of the classification of nail bed capillary types by analysis based on the measurement classification according to this embodiment. Fig. 13 is a diagram showing another example of the classification of nail bed capillary types by analysis based on the measurement classification according to this embodiment.

[0093] Capillary vessel image G10 is a type 10 capillary vessel image classified as a retracted type. The capillary vessel image G21 is a type 21 capillary vessel image classified as a stasis type. Capillary vessel image G22 is a type 22 capillary vessel image classified as a long stasis and regression type. The capillary vessel image G31 is a capillary vessel image of type 31, which is classified as a tip type. Capillary vessel image G32 is a capillary vessel image of type 32, which is classified as a tip twist type. Capillary vessel image G40 is a type 40 capillary vessel image classified as a hairpin type. Capillary vessel image G50 is a type 50 capillary vessel image classified as a horizontal line type. The capillary vessel image G61 is a type 61 capillary vessel image classified as a torsion type. The capillary vessel image G62 is a capillary vessel image of type 62, which is classified as a long twisted type (including a meandering type). These types were classified by setting discrimination rules using measurement values ​​and shape data of the length, thickness, and horizontal lines of the capillaries through analysis based on measurement classification.

[0094] Furthermore, the health condition determination unit 123 may determine a specific disease or symptom, such as a menopausal symptom, depending on the type of capillary blood vessel. This will be described with reference to Figures 14 and 15. Note that large-scale health-related data collected in 2018 and 2019 [n = 1051 (2018 data), n = 1059 (2019 data)] was used as the analysis data.

[0095] [Relationship between capillary types and menopausal symptoms] FIG. 14 is a diagram showing an example of the relationship between the type of nail bed capillaries and menopausal symptoms based on the cross-tabulation analysis according to this embodiment.

[0096] Type 10 tends to have menopausal symptoms of sweating. Type 21 tends to have menopausal symptoms of hot flashes. Type 31 tends to have menopausal symptoms such as irritability and stiff necks. Type 50 is prone to menopausal symptoms such as depression, insomnia, and irregular menstruation. Type 61 tends to have menopausal symptoms of stiff hands. Type 62 is at risk for menopausal symptoms, such as sweating, headaches, and insomnia.

[0097] FIG. 15 is a diagram showing an example of the relationship between the type of nail bed capillaries and the blood concentration of each hormone according to this embodiment. The average blood concentrations and significance levels of each hormone, including estradiol, FSH, ACTH, cortisol, DHEA, and VEGF, are shown according to the type of capillary. For example, type 62 tends to have low estradiol levels and high cortisol levels, while type 50 tends to have low ACTH levels.

[0098] In this way, the capillary type classification unit 122 may classify the capillaries into the types shown in Figures 11, 12, and 13 based on the images acquired by the image acquisition unit 121, the health condition determination unit 123 may determine the menopausal symptoms by referring to the correspondence between the capillary types and menopausal symptoms shown in Figure 14, and the recommendation unit 124 may make recommendations for symptom improvement by referring to the correspondence between the capillary types and hormones shown in Figure 15.

[0099] [Hardware configuration] Next, the hardware configuration of the health management support device 1 will be described. FIG. 16 is a block diagram showing an example of the hardware configuration of the health management support device 1 according to this embodiment. The health management support device 1 is composed of a CPU 101, a drive unit 102, a storage medium 103, an input unit 104, an output unit 105, a ROM 106 (Read Only Memory), a RAM 107 (Random Access Memory), an auxiliary storage unit 108, and an interface unit 109.

[0100] The CPU 101, drive unit 102, input unit 104, output unit 105, ROM 106, RAM 107, auxiliary storage unit 108, and interface unit 109 are interconnected via a bus. The CPU 101 referred to here refers to a processor in general, and includes not only a device called a CPU in the narrow sense, but also, for example, a GPU, a DSP, etc. Furthermore, the CPU 101 referred to here is not limited to being realized by a single processor, but may be realized by combining multiple processors of the same or different types.

[0101] CPU 101 reads and executes programs stored in auxiliary storage unit 108, ROM 106, and RAM 107, and also reads various data stored in auxiliary storage unit 108, ROM 106, and RAM 107 and writes the various data to auxiliary storage unit 108 and RAM 107, thereby controlling health management support device 1. CPU 101 also reads various data stored in storage medium 103 via drive unit 102 and writes the various data to storage medium 103. Storage medium 103 is a portable storage medium such as a magneto-optical disk, a flexible disk, or a flash memory, and stores various data. The drive unit 102 is a device that reads and writes data from and to a storage medium 103 such as an optical disk drive, a flexible disk drive, or a flash memory.

[0102] The input unit 104 is an input device such as a mouse, a keyboard, a touch panel, a channel button, a power button, a setting button, and an infrared receiver. The output unit 105 is an output device such as a display unit, a speaker, or the like. The ROM 106 and RAM 107 store programs and various data for operating the various functional units of the health management support device 1.

[0103] The auxiliary storage unit 108 is a hard disk drive, a flash memory, or the like, and stores programs for operating the functional units of the health management support device 1 and various data. The interface unit 109 has a communication interface and is connected to the network NW or other devices in the health care support system SYS by wire or wirelessly.

[0104] For example, the control unit 12 in the functional configuration of the health management support device 1 in the above-mentioned FIG. 1 corresponds to the CPU 101 in FIG. 16, the communication unit 10 in FIG. 1 corresponds to the interface unit 109 in FIG. 16, and the memory unit 14 in FIG. 1 corresponds to the memory medium 103, ROM 106, RAM 107, auxiliary memory unit 108, etc. in FIG. 16.

[0105] At least a part of the health management support device 1 in the above-described embodiment may be implemented by a computer. In this case, a program for implementing this control function may be recorded on a computer-readable recording medium, and the program recorded on the recording medium may be read into a computer system and executed. The "computer system" referred to here is a computer system built into the health management support device 1, and includes hardware such as an OS and peripheral devices.

[0106] Furthermore, "computer-readable recording media" refers to portable media such as flexible disks, optical magnetic disks, ROMs, and CD-ROMs, as well as storage devices such as hard disks built into computer systems. Furthermore, "computer-readable recording media" may also include devices that dynamically store programs for a short period of time, such as communication lines used when transmitting programs over networks like the Internet or communication lines like telephone lines, or devices that store programs for a fixed period of time, such as volatile memory within computer systems that serve as servers or clients in such cases. Furthermore, the above-mentioned programs may be programs that realize some of the aforementioned functions, or may be programs that can realize the aforementioned functions in combination with programs already stored in the computer system.

[0107] Furthermore, part or all of the health management support device 1 in the above-described embodiment may be realized as an integrated circuit such as an LSI (Large Scale Integration). Each functional block of the health management support device 1 may be individually implemented as a processor, or part or all of them may be integrated into a processor. The integrated circuit implementation method is not limited to LSI, and may be implemented using a dedicated circuit or a general-purpose processor. Furthermore, if an integrated circuit implementation technology that can replace LSI emerges due to advances in semiconductor technology, an integrated circuit based on that technology may be used.

[0108] One embodiment of the present invention has been described in detail above with reference to the drawings, but the specific configuration is not limited to that described above, and various design changes and the like are possible within the scope that does not deviate from the gist of the present invention. [Explanation of symbols]

[0109] 1 Health management support device 10. Communications Department 12 Control Unit 120 User information acquisition unit 121 Image acquisition unit 1211 Capillary image acquisition unit 1212 Finger image acquisition unit 122 Capillary Type Classification Section 123 Health Status Assessment Department 124 Recommendation Department 125 Notification Department 126 Learning Department 14 Storage section 141 Learning Model 142 Capillary Image Information 143 Finger Image Information 144 Health Status Data 145 Recommendation Information 146 User Information 101 CPU 102 Drive section 103 Storage medium 104 Input section 105 Output section 106 ROM 107 RAM 108 Auxiliary storage 109 Interface section

Claims

1. an image acquisition unit that acquires an image of capillaries in the nail bed of a user; a capillary type classification unit that classifies whether the shape of the capillaries in the nail bed capillary image conforms to one or more of a plurality of types, and outputs one or more conforming types and a conforming probability for each conforming type; a health condition determination unit that determines a health condition of the user based on the classified compatibility types, compatibility probabilities for each compatibility type, and correspondence relationships related to capillary blood vessel types; Equipped with Health management support system.

2. a user information acquisition unit that acquires the age and sex of the user as user information; Furthermore, the health condition determination unit determines the health condition of the user based on the compatible types of capillaries and the compatibility probabilities for each compatible type, correspondence relationships related to the capillary types, and the user information; The health care support system according to claim 1 .

3. The health condition determination unit By inputting the nail bed capillary image of the user into a learning model that has learned the correspondence between the shape of the nail bed capillary image and multiple types corresponding to the capillary shape, the health condition of the user is determined based on one or more compatible types for the user and the compatibility probability for each compatible type that are output, and the correspondence related to the capillary type, which is data representing the health condition. The health care support system according to claim 1 .

4. The health condition determination unit determining the effect of treatment on the health condition of the user by extracting factors of changes in the compatibility type according to changes in the compatibility type and the compatibility probability for each compatibility type using time-series nail bed capillary images of the user and correspondence relationships related to the capillary types; The health care support system according to claim 1 .

5. a notification unit that notifies the user, another user designated by the user, or both of the user and the other user of an output indicating the user's health condition; Further provided with The health care support system according to claim 3 .

6. a recommendation unit that makes recommendations to the user according to the health condition of the user; Further provided with The health care support system according to claim 5 .

7. The recommendation unit Recommending at least one of supplements, Chinese medicine, medicine, food, meal menus, and exercise menus to the user according to the user's health condition; The health care support system according to claim 6.

8. an image acquisition unit that acquires an image of capillaries in the nail bed of a user; a capillary type classification unit that classifies whether the shape of the capillaries in the nail bed capillary image conforms to one or more of a plurality of types, and outputs one or more conforming types and a conforming probability for each conforming type; a health condition determination unit that determines a health condition of the user based on the classified compatibility types, compatibility probabilities for each compatibility type, and correspondence relationships related to capillary blood vessel types; Equipped with Health management support device.

9. A computer-implemented health management support method, comprising: an image acquisition step of acquiring an image of the capillaries in the nail bed of the user; a capillary type classification process for classifying whether the shape of the capillaries in the nail bed capillary image conforms to one or more of a plurality of types, and outputting one or more conforming types and a conformance probability for each conforming type; a health condition determination step of determining a health condition of the user based on the classified compatibility types, the compatibility probabilities for each compatibility type, and correspondences related to capillary types; having Health management support method.

10. On the computer, image acquisition means for acquiring an image of capillaries in the nail bed of a user; a capillary type classification means for classifying whether the shape of the capillaries in the nail bed capillary image conforms to one or more of a plurality of types, and outputting one or more conforming types and a conformance probability for each conforming type; a health condition determination means for determining a health condition of the user based on the classified compatibility types, the compatibility probabilities for each compatibility type, and a correspondence relationship related to capillary blood vessel types; A program to execute.

Images