JP7468892B2 - DIAGNOSTIC SUPPORT SYSTEM, DIAGNOSTIC SUPPORT SERVER, DIAGNOSTIC SUPPORT TERMINAL, DIAGNOSTIC SUPPORT PROGRAM, AND DIAGNOSTIC SUPPORT SERVER MAINTENANCE METHOD - Google Patents

Description

The present invention relates to a diagnostic support system, a diagnostic support server, a diagnostic support terminal, and a diagnostic support program that provide diagnostic support based on capillary blood vessel images, as well as a maintenance method for the diagnostic support server.

In general, capillaries change depending on age and physical condition, so it is said that by observing the condition of subcutaneous capillaries in fingertips, the fundus of the eye, etc., it is possible to understand health conditions, lifestyle-related diseases, illnesses, etc. Capillaries are very delicate and cannot be observed with the naked eye, so diagnosis is performed using observation devices such as microscopes. In this application, the term "capillaries" includes not only subcutaneous blood vessels, but also submucosal blood vessels, etc.

Conventionally, the technologies disclosed in Patent Documents 1 to 3 are known as systems for supporting diagnosis based on images of capillaries. The system disclosed in Patent Document 1 processes an image of the capillaries to recognize the shape of the capillaries in the image, and calculates the length, thickness, clarity, degree of abnormality, etc. of the capillaries in the form of single numerical values from the recognized capillary shape (see paragraphs 0006 to 0016 and Figure 3 of the specification of Patent Document 1).

The system disclosed in Patent Document 2 quantifies the density, thickness, shape, and blood flow velocity of blood vessels from color video images of capillaries in the nail cuticle of the fingertip, ranks the values by comparing them with numerical data in a reference table pre-stored in the system, and obtains an overall assessment of the health condition, such as "mild abnormality" or "severe abnormality," by comparing all the ranked data with empirical data in another reference table (see paragraphs 0005, 0006, 0025, and Figure 11 of the specification of Patent Document 2).

The system disclosed in Patent Document 3 by the applicant of the present application is a client-server system in which a capillary image is obtained by photographing the capillaries in the nail cuticle at a client terminal, and the server judges the presence or absence of a subpapillary vascular plexus running horizontally across the finger's length (vertical direction), the degree of turbidity of the body fluid, and the degree of curvature of the tips of the blood vessels running in the longitudinal direction of the finger based on the capillary image, and further evaluates the health condition based on the results of the judgment, such as "stress or strain on the heart is suspected," "poor metabolism or lack of exercise is suspected," or "mainly poor lifestyle habits are suspected," and displays the results of the evaluation on the client terminal (see paragraphs 0008-0010 and Figure 8 of the specification of Patent Document 3).

However, even though all of these systems output certain results to assist in diagnosis, the data on which the results are based is limited to images of capillaries, so there seems to be room for improvement in the accuracy of the results. In addition, the results provided are limited to numerical values and simple information, so there seems to be room for improvement in the usefulness of the results.

JP 2015-157071 A JP 2006-325714 A Patent No. 6186663

The present invention has been made in consideration of the above problems, and relates to a diagnostic support system, a diagnostic support server, a diagnostic support terminal, a diagnostic support program, and a diagnostic support server maintenance method, and aims to improve the accuracy of the judgment results that support the diagnosis of the subject's health condition, and to increase the usefulness of the judgment results.

In order to achieve the above-mentioned object, a first aspect of the present invention is a diagnostic support system that supports diagnosis of a subject's health condition based on a capillary image of the subject, and includes a server, a first terminal connectable to the server via a network, and a second terminal that is the same as or different from the first terminal and connectable to the server via the network.
The first terminal has an imaging means for capturing an image of the capillaries of the subject to obtain the capillary image, and an image transmitting means for transmitting the capillary image to the server.
The server has a questionnaire sending means, an answer receiving means, an image receiving means, a feature extraction means, a class determination means, and a data sending means. The questionnaire sending means sends a questionnaire on the lifestyle habits of the subject to the second terminal in response to a request from the second terminal. The answer receiving means receives the answer to the questionnaire by the subject sent from the second terminal. The image receiving means receives the capillary image sent by the first terminal. The feature extraction means extracts capillary features from the capillary image received by the image receiving means. The class determination means performs class determination based on the extracted feature and the received answer to the questionnaire to determine which of a plurality of preset classes the state of the capillaries represented by the received capillary image corresponds to. The data sending means sends data including the result of the class determination to the first terminal as transmission data.
Moreover, the first terminal further comprises 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.

According to this configuration, the result of the class determination, which indicates which of a plurality of pre-set classes the state of the subject's capillaries falls into, is displayed on the first terminal, and the displayed data can be used to diagnose the subject's health condition. The class determination is performed based on the feature values extracted from the captured image of the subject's capillaries as well as the subject's responses to a questionnaire about the subject's lifestyle habits, improving the accuracy of the determination results that support the diagnosis of the subject. In each aspect of the present invention, the capillary image may be a still image, a moving image, a monochrome image, or a color image. The server may be a distributed server such as a cloud server. Furthermore, the feature values may be, for example, the length, thickness, number, degree of turbidity, curvature, etc. of the capillaries.

The second aspect of the present invention is a diagnostic support system according to the first aspect, in which the second terminal is different from the first terminal.

With this configuration, the subject can send responses to a questionnaire about his/her lifestyle habits to the server from another terminal without the knowledge of the operator of the first terminal. In other words, the subject can send responses to the questionnaire to the server while protecting privacy about his/her lifestyle habits from the operator of the first terminal.

A third aspect of the present invention is a diagnostic support system according to the first or second aspect, in which the data transmission means of the server transmits the transmission data to the first terminal as editable data in a format editable by the first terminal. The first terminal further has an editing means for editing the editable data received by the data receiving means in accordance with an input operation to the first terminal.

With this configuration, the data displayed on the first terminal that is useful for diagnosing the subject can reflect the knowledge and experience of the operator who inputs the data, or the results of the operator's interviews with the subject.

A fourth aspect of the present invention is a diagnostic support system according to the third aspect, wherein the first terminal further has an edited data transmission means for transmitting the edited editable data to the server as edited data.
The server further includes an edited data receiving means for receiving the edited data transmitted by the edited data transmitting means, a confirmed data creating means for creating confirmed data that cannot be edited based on the received edited data, and a confirmed data transmitting means for transmitting the created confirmed data to the first terminal.
Furthermore, the first terminal further comprises a confirmed data receiving means for receiving the confirmed data transmitted by the confirmed data transmitting means, and a confirmed data display means for displaying the received confirmed data.

According to this configuration, finalized data reflecting the content of editing by the operator of the first terminal is displayed on the first terminal.

According to a fifth aspect of the present invention, the diagnostic support system according to any one of the first to fourth aspects is such that the server further comprises a constitution data storage device that stores constitution data, which is data expressing the constitution of the subject and corresponds to each of the plurality of classes, in association with the corresponding class, and a constitution data reading means that reads, from the constitution data storage device, constitution data corresponding to the class indicated by the result of the class determination among the plurality of classes. The data transmission means then transmits the transmission data including the read constitution data to the first terminal.

According to this configuration, constitutional data useful for diagnosing the subject's constitution is displayed on the first terminal. In other words, the usefulness of the data displayed on the first terminal is improved. The constitutional data includes, for example, high blood pressure, lack of exercise, stress, and hyperglycemia. In each aspect of the present invention, the storage device is, for example, a hard disk device. In addition, the storage devices in each aspect of the present invention may be the same storage device or may be separate storage devices.

According to a sixth aspect of the present invention, in the diagnostic support system according to any one of the first to fifth aspects, the server further comprises a constitution improvement data storage device that stores constitution improvement data expressing advice for improving the constitution of the subject corresponding to each of the plurality of classes in association with the corresponding class, and constitution improvement data reading means that reads, from the constitution improvement data storage device, constitution improvement data corresponding to the class indicated by the result of the class determination among the plurality of classes. The data transmission means then transmits the read constitution improvement data to the first terminal in the transmission data.

According to this configuration, constitution improvement data that is useful for giving advice on improving the subject's constitution is displayed on the first terminal. In other words, the usefulness of the data displayed on the first terminal is improved. The constitution improvement data includes, for example, encouragement to rest, encouragement to moderate, encouragement to exercise, etc.

A seventh aspect of the present invention is a diagnostic support system according to any one of the first to sixth aspects, in which the data transmission means of the server transmits the transmission data to the first terminal, including the extracted feature amount.

This configuration makes it possible to diagnose a subject's health condition in more detail based on the characteristics of the subject's capillaries.

According to an eighth aspect of the present invention, the server further comprises a subject feature storage device, a subject feature storage means for storing the extracted features in the subject feature storage device in association with the subject, and a past feature reading means for reading past features, which are past features associated with the subject, from the subject feature storage device. The data transmission means of the server then transmits the transmission data including the read past features of the subject to the first terminal.

This configuration makes it possible to diagnose the subject's health condition while also taking into account the feature history of the same subject. In other words, the usefulness of the data displayed on the first terminal is improved.

A ninth aspect of the present invention is a diagnostic support system according to any one of the first to eighth aspects, wherein the server further includes a reference data storage device that stores, as reference data, a range or a representative value of capillary feature quantities corresponding to each of the plurality of classes, and a coefficient data storage device that stores coefficient data, which is a set of coefficients representing the possibility that each of a plurality of possible answers to the questionnaire belongs to each of the plurality of classes.
The class determination means includes answer determination means, coefficient data selection means, feature correction means, and feature determination means. The answer determination means determines which of the multiple expected answers the answer received by the answer receiving means corresponds to. The coefficient data selection means selects coefficient data stored in the coefficient data storage device and corresponding to the answer determined by the answer determination means. The feature correction means corrects the feature extracted by the feature extraction means by displacing it in a feature space in which each of the multiple classes is represented as a spatial region, according to the magnitude of each of the coefficient sets contained in the coefficient data selected by the coefficient data selection means. The feature determination means performs the class determination by determining which of the multiple classes the feature corrected by the feature correction means corresponds to by referring to the reference data stored in the reference data storage device.

This configuration makes class determination relatively easy.

According to a tenth aspect of the present invention, in the diagnosis support system according to the ninth aspect, the reference data storage device stores a representative value of the capillary feature corresponding to each of the plurality of classes as the reference data. The feature correction means corrects the feature extracted by the feature extraction means by displacing the feature extracted by the feature extraction means in the feature space so as to approach the representative value of the feature of each of the plurality of classes according to the size of each of the coefficient sets contained in the coefficient data selected by the coefficient data selection means. Furthermore, the feature determination means determines that the class of the plurality of classes whose representative value in the feature space is closest to the feature corrected by the feature correction means is the class to which the corrected feature corresponds.

This configuration makes class determination even easier.

An eleventh aspect of the present invention is a diagnostic support system according to the ninth or tenth aspect, in which the server further includes a feature answer storage device and a feature answer storage means for storing the feature extracted by the feature extraction means and the answer to the questionnaire received by the answer receiving means in the feature answer storage device in association with each other.

With this configuration, as the diagnostic support system is used, the features and responses to the questionnaire that are stored in the feature response storage device in association with each other can be used to reset the capillary conditions into multiple classes and to reset the coefficient data. This allows for more accurate class determination based on more data.

A twelfth aspect of the present invention is a diagnostic support system that supports diagnosis of a subject's health condition based on a capillary image of the subject, and includes a server and a first terminal connectable to the server via a network.
The first terminal has an imaging means for capturing an image of the capillaries of the subject to obtain the capillary image, and an image transmitting means for transmitting the capillary image to the server.
The server has an image receiving means, a feature extraction means, a class determination means, and a data transmission means. The image receiving means receives the capillary image transmitted by the first terminal. The feature extraction means extracts features of the capillaries from the capillary image received by the image receiving means. The class determination means performs class determination based on the extracted features to determine which of a plurality of preset classes the state of the capillaries represented by the received capillary image corresponds to. The data transmission means transmits data including the result of the class determination to the first terminal as transmission data.
The first terminal further comprises 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. Furthermore, the data transmitting means of the server transmits the transmission data to the first terminal as editable data in a format editable by the first terminal. The first terminal further comprises an editing means for editing the editable data received by the data receiving means in accordance with an input operation to the first terminal.

According to this configuration, the result of the class determination, which indicates which of multiple predefined classes the state of the subject's capillaries falls into, is displayed on the first terminal, so this displayed data can be used to diagnose the subject's health condition. Furthermore, according to this configuration, the data displayed on the first terminal that is useful for diagnosing the subject's health condition can reflect the knowledge and experience of the operator who inputs the data, or the results of the operator's interviews with the subject. In other words, the usefulness of the data displayed on the first terminal is improved.

A thirteenth aspect of the present invention is a diagnostic support system according to the twelfth aspect, wherein the first terminal further has an edited data transmission means for transmitting the edited editable data to the server as edited data.
The server further includes an edited data receiving means for receiving the edited data transmitted by the edited data transmitting means, a confirmed data creating means for creating confirmed data that cannot be edited based on the received edited data, and a confirmed data transmitting means for transmitting the created confirmed data to the first terminal.
The first terminal further comprises a confirmed data receiving means for receiving the confirmed data transmitted by the confirmed data transmitting means, and a confirmed data display means for displaying the received confirmed data.

With this configuration, the finalized data reflecting the edits made by the operator is displayed on the first terminal.

A fourteenth aspect of the present invention provides a diagnostic support system for supporting diagnosis of a subject's health condition based on a capillary image of the subject, comprising a server and a first terminal connectable to the server via a network.
The first terminal has an imaging means for capturing an image of the capillaries of the subject to obtain the capillary image, and an image transmitting means for transmitting the capillary image to the server.
The server has an image receiving means, a feature extraction means, a class determination means, and a data transmission means. The image receiving means receives the capillary image transmitted by the first terminal. The feature extraction means extracts features of the capillaries from the capillary image received by the image receiving means. The class determination means performs class determination based on the extracted features to determine which of a plurality of preset classes the state of the capillaries represented by the received capillary image corresponds to. The data transmission means transmits data including the result of the class determination to the first terminal as transmission data.
The first terminal further comprises 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. The server further comprises a constitution data storage device for storing constitution data, which is data expressing the constitution of the subject and corresponds to each of the plurality of classes, in association with the corresponding class, and constitution data reading means for reading, from the constitution data storage device, constitution data corresponding to the class indicated by the result of the class determination among the plurality of classes. The data transmitting means transmits the transmission data including the read constitution data to the first terminal.

According to this configuration, the result of the class determination, which indicates which of a number of predefined classes the state of the subject's capillaries falls into, is displayed on the first terminal, and this displayed data can be used to diagnose the subject's health condition. Also, according to this configuration, constitution data useful for diagnosing the subject's constitution is displayed on the first terminal. In other words, the usefulness of the data displayed on the first terminal is improved. The constitution data includes, for example, high blood pressure, lack of exercise, stress, hyperglycemia, etc.

A fifteenth aspect of the present invention provides a diagnostic support system for supporting diagnosis of a subject's health condition based on a capillary image of the subject, comprising a server and a first terminal connectable to the server via a network.
The first terminal has an imaging means for capturing an image of the capillaries of the subject to obtain the capillary image, and an image transmitting means for transmitting the capillary image to the server.
The server has an image receiving means, a feature extraction means, a class determination means, and a data transmission means. The image receiving means receives the capillary image transmitted by the first terminal. The feature extraction means extracts features of the capillaries from the capillary image received by the image receiving means. The class determination means performs class determination based on the extracted features to determine which of a plurality of preset classes the state of the capillaries represented by the received capillary image corresponds to. The data transmission means transmits data including the result of the class determination to the first terminal as transmission data.
The first terminal further comprises 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.
The server further comprises a constitution improvement data storage device for storing constitution improvement data expressing advice for improving the constitution of the subject corresponding to each of the plurality of classes in association with the corresponding class, and constitution improvement data reading means for reading, from the constitution improvement data storage device, constitution improvement data corresponding to the class indicated by the result of the class determination among the plurality of classes. Furthermore, the data transmission means transmits the read constitution improvement data to the first terminal together with the transmission data.

According to this configuration, the result of class determination, which indicates which of multiple preset classes the state of the subject's capillaries falls into, is displayed on the first terminal, and this displayed data can be used to diagnose the subject's health condition. Also, according to this configuration, constitution improvement data that is useful for giving advice on improving the subject's constitution is displayed on the first terminal. In other words, the usefulness of the data displayed on the first terminal is improved. The constitution improvement data includes, for example, encouragement to rest, encouragement to moderate, encouragement to exercise, etc.

A sixteenth aspect of the present invention is a diagnostic support system according to any one of the twelfth to fifteenth aspects, in which the server further includes a feature storage device and a feature storage means for storing the feature extracted by the feature extraction means in the feature storage device.

With this configuration, the features stored in the feature storage device as the diagnostic support system is used can be used to reassign the capillary conditions to multiple classes. This allows for more accurate class determination based on more data.

A seventeenth aspect of the present invention provides a diagnostic support system for supporting diagnosis of a subject's health condition based on a capillary image of the subject, comprising a server and a first terminal connectable to the server via a network.
The first terminal has an imaging means for capturing an image of the capillaries of the subject to obtain the capillary image, and an image transmitting means for transmitting the capillary image to the server.
The server has an image receiving means, a feature extraction means, a class determination means, and a data transmission means. The image receiving means receives the capillary image transmitted by the first terminal. The feature extraction means extracts features of the capillaries from the capillary image received by the image receiving means. The class determination means performs class determination based on the extracted features to determine which of a plurality of preset classes the state of the capillaries represented by the received capillary image corresponds to. The data transmission means transmits data including the result of the class determination to the first terminal as transmission data.
The first terminal further comprises 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.
The server further includes a subject feature storage device, a subject feature storage means for associating the extracted features with the subject and storing them in the subject feature storage device, and a past feature reading means for reading past features, which are past features associated with the subject, from the subject feature storage device.
Furthermore, the data transmission means of the server transmits the transmission data including the extracted feature amount and the read past feature amount of the subject to the first terminal.

According to this configuration, the result of the class determination, which indicates which of multiple predefined classes the state of the subject's capillaries falls into, is displayed on the first terminal, and this displayed data can be used to diagnose the subject's health condition. This configuration also enables a more detailed diagnosis of the subject's health condition based on the characteristics of the subject's capillaries. Furthermore, this configuration also enables a diagnosis of the subject's health condition that takes into account the history of the characteristics of the same subject. In other words, the usefulness of the data displayed on the first terminal is improved.

The eighteenth aspect of the present invention is a diagnostic support server that can be connected to a network and supports diagnosis of a subject's health condition based on a capillary image of the subject, the diagnostic support server being the server provided in the diagnostic support system according to any one of the first to seventeenth aspects.

This configuration corresponds to the server provided in the diagnostic support system according to each aspect of the present invention.

A nineteenth aspect of the present invention is a diagnostic support terminal that can be connected to a network and supports diagnosis of a subject's health condition based on a capillary image of the subject, the diagnostic support terminal being the first terminal provided in a diagnostic support system according to any one of the first to seventeenth aspects.

This configuration corresponds to the first terminal provided in the diagnostic support system according to each aspect of the present invention.

A twentieth aspect of the present invention is a diagnostic assistance program that assists in the diagnosis of a subject's health condition based on a capillary image of the subject, the diagnostic assistance program being a program that causes a computer to function as a diagnostic assistance server according to the eighteenth aspect.

This configuration corresponds to a program that causes a computer to function as a diagnostic support server according to one aspect of the present invention.

A twenty-first aspect of the present invention is a diagnostic assistance program that assists in the diagnosis of a subject's health condition based on a capillary image of the subject, the diagnostic assistance program being a program that causes a computer to function as a diagnostic assistance terminal according to the nineteenth aspect.

This configuration corresponds to a program that causes a computer to function as a diagnostic support terminal according to one aspect of the present invention.

A 22nd aspect of the present invention relates to a diagnostic support server that is connectable to a network and supports diagnosis of a subject's health condition based on a capillary image of the subject, and is a diagnostic support server maintenance method for maintaining the diagnostic support server that is a server provided in a diagnostic support system according to the 11th aspect, and includes the following actions (a) to (f):
(a) reading out the stored group of features from the feature answer storage device; (b) resetting the plurality of classes for the state of capillaries based on the read-out group of features; (c) updating the stored reference data by storing, as the reference data, a range or representative value of the features corresponding to each of the reset classes in the reference data storage device; (d) reading out, from the feature answer storage device, a group of responses to the questionnaire stored in association with the group of features; (e) calculating the coefficient data based on the read-out group of responses to the questionnaire; and (f) updating the stored coefficient data by storing the calculated coefficient data in the coefficient data storage device.

According to this configuration, the features and answers to the questionnaires stored in the feature answer storage device as the diagnostic support server is used can be used to calculate more accurate feature ranges or representative values and more accurate coefficient data, which can be used by the diagnostic support server in the future to help diagnose health conditions more accurately. Note that actions (a) to (f) are not limited to being performed in the order described.

The twenty-third aspect of the present invention relates to a diagnostic support server that can be connected to a network and supports diagnosis of a subject's health condition based on a capillary image of the subject, and a diagnostic support server maintenance method for maintaining the diagnostic support server, which is a server provided in a diagnostic support system according to the sixteenth aspect, including reading the group of stored features from the feature storage device and reconfiguring the multiple classes regarding the state of the capillaries based on the group of read features.

According to this configuration, the features accumulated in the feature storage device as the diagnostic support server is used can be used to reset multiple classes with higher accuracy, and the reset classes can be provided for future use by the diagnostic support server, which can be useful for more accurate diagnosis of health conditions.

The twenty-fourth aspect of the present invention relates to a diagnostic support server that can be connected to a network and supports diagnosis of a subject's health condition based on a capillary image of the subject, and a diagnostic support server maintenance method for maintaining the diagnostic support server, which is a server provided in a diagnostic support system according to the seventeenth aspect, including reading out the group of stored features from the subject feature storage device, and reconfiguring the multiple classes regarding the state of the capillaries based on the group of features that has been read out.

With this configuration, features stored in the subject feature storage device as the diagnostic support server is used can be used to re-define multiple classes with higher accuracy, and the re-define multiple classes can be made available for future use by the server, which can be useful for more accurate diagnosis of health conditions.

As described above, according to the present invention, the diagnostic support system, diagnostic support server, diagnostic support terminal, diagnostic support program, and diagnostic support server maintenance method improve the accuracy of the judgment results that support the diagnosis of the subject's health condition, and also enhance the usefulness of the judgment results.

1 is a schematic explanatory diagram illustrating a configuration of a diagnosis support system according to an embodiment of the present invention; 2 is a block diagram illustrating an example of the hardware configuration of a diagnostic support server, a diagnostic base terminal, and a subject terminal that constitute the diagnostic support system of FIG. 1. 2 is a flowchart in the form of a sequence diagram illustrating the process flow and the signal flow in the diagnosis support system of FIG. 1 . 4 is a flowchart illustrating an example of an expanded flow of the class determination process of FIG. 3. 13 is a screen diagram illustrating a questionnaire response acceptance screen displayed on a display device of a subject terminal. FIG. FIG. 11 is an explanatory diagram illustrating an example of a procedure for calculating coefficient data for each of a plurality of possible responses to a questionnaire. 11 is an explanatory diagram illustrating an example of a method of feature amount correction executed in class determination; 13 is a graph illustrating a scoring function used to calculate a capillary vessel rank. 13 is a graph illustrating another scoring function used to calculate the capillary vessel rank. 13 is a screen diagram illustrating a diagnosis suggestion screen displayed on a display device of the diagnosis base terminal. FIG. 13 is a screen diagram illustrating an example of a diagnostic element selection screen displayed on a display device of the diagnostic site terminal. FIG. 13 is a screen diagram illustrating an example of a diagnosis result screen displayed on a display device of the diagnosis site terminal. FIG. 2 is a flowchart illustrating an example of a procedure for constructing, operating, and maintaining the diagnostic support server of FIG. 1 .

[1. Overview of the diagnosis support system]
1 is a schematic diagram illustrating the configuration of a diagnosis support system according to an embodiment of the present invention. This diagnosis support system 101 is a system that supports diagnosis of the health condition of a subject 1 based on a capillary blood vessel image of the subject 1, and includes a diagnosis support server 102, a diagnosis base terminal 103, and a subject terminal 104. The diagnosis base terminal 103 and the subject terminal 104 are connected to the diagnosis support server 102 via a network 5 such as the Internet.

The diagnostic base terminal 103 obtains a capillary image by photographing the capillaries of the subject 1, and transmits the obtained capillary image to the diagnostic support server 102. The subject terminal 104 displays a questionnaire prepared by the diagnostic support server 102, and transmits the answers to the questionnaire entered by the subject 1 to the diagnostic support server 102. The diagnostic support server 102 creates diagnostic support data useful for diagnosing the subject's health condition based on the capillary image sent from the diagnostic base terminal 103 and the questionnaire answers sent from the subject terminal 104, and transmits the created diagnostic support data to the diagnostic base terminal 103. The subject 1 can receive the diagnostic support data through the diagnostic base terminal 103. It is also possible to add or modify content to the diagnostic support data based on the specialized knowledge and experience of the operator 9 of the diagnostic base terminal 103.

The diagnostic support server 102 may be a personal computer, or may be a server provided by a server provider as in the illustrated example, and may be a distributed server such as a cloud server.

The diagnostic base terminal 103 is, for example, a personal computer, and is connected to a capillary scope 3 that obtains a capillary image by photographing the capillaries of the subject 1. The capillary scope 3 has, for example, a microscope that optically magnifies the image of the capillaries, and an element that converts the magnified image into an electrical signal using a light receiving element such as a CCD. The capillary image obtained by the capillary scope 3 may generally be a still image, a moving image, a monochrome image, or a color image, but as an example, it is a still image. The capillary scope 3 itself is well known in the art, so a detailed description thereof will be omitted. The diagnostic base terminal 103 is installed at a diagnostic base 7. The diagnostic base 7 is a place or facility that provides diagnostic services to unspecified subjects 1, and is assumed to be, for example, a public facility such as a pharmacy, a hospital, a clinic, or a government office, or a hotel lobby. The diagnostic base terminal 103 is operated by an operator 9 located at the diagnostic base 7.

The subject terminal 104 is a terminal that belongs to the subject 1, who is a user (consumer) of the diagnostic support system 101, and is managed by the subject 1. The subject terminal 104 may be, for example, a personal computer placed in the residence 8 of the subject 1, or, as in the illustrated example, may be a smartphone owned by the subject 1.

Figure 2 is a block diagram illustrating the hardware configuration of the diagnostic support server 102, diagnostic base terminal 103, and subject terminal 104 that constitute the diagnostic support system 101. In the illustrated example, these devices 102, 103, and 104 are configured as personal computers, and have a main circuit section 12, an input device 11, a display device 13, and a hard disk drive (HDD) 14. The main circuit section 12 also has a CPU (Central Processing Unit) 21, a ROM (Read Only Memory) 22, a RAM (Random Access Memory) 23, an interface (I/F) 24, and a reading device 25.

The CPU 21 executes programs stored in the ROM 22 and the RAM 23. The ROM 22 stores, for example, a Basic Input Output System (BIOS) that starts the Operating System (OS) and configures the hardware immediately after the devices 102, 103, and 104 are started up. The OS, applications, and the like stored in the hard disk device 14 are deployed to the RAM 23 as necessary and are made available for execution by the CPU 21. The RAM 23 is also used as a work memory that temporarily stores data generated when the CPU 21 executes processing.

The interface (I/F) 24 has an IDE (Integrated Device Electronics) controller, a LAN controller, an input/output controller, etc. (not shown), and external devices such as the input device 11 are connected to the CPU 21 through these circuits. The CPU 21, ROM 22, RAM 23, and interface 24 are connected to each other through a bus line 26.

The input device 11 includes, for example, a keyboard and a mouse, and is a device for an operator to manually input data. The display device 13 is, for example, an LCD (Liquid Crystal Display) device, and is a device for displaying image data as a visible image. The hard disk device 14 stores the OS, applications, etc., as well as data required for processing. The reading device 25 is a device for reading the contents recorded on a recording medium 27, such as a CD-ROM, on which programs and data are recorded. The devices 102, 103, and 104 are connected to a network 5 (Figure 1), such as the Internet, for example, via a LAN 15, and realize communication, for example, according to the TCP/IP protocol. The LAN 15 is connected to a bus line 26 via a LAN controller included in the interface 24, and is connected to the network 5 via a router (not shown).

Figure 3 is a flow chart in the form of a sequence diagram illustrating the process flow and the signal flow in the diagnostic support system 101 at the same time. The process flow is indicated by solid arrows, and the signal flow is indicated by dotted arrows. The diagnostic support server 102 executes processes S1 to S17. The diagnostic base terminal 103 executes processes S21 to S27. The subject terminal 104 executes processes S31 to S37. Each process corresponds to a means for executing each process. That is, the diagnostic support server 102 has a means for executing each of the processes S1 to S17, and has, for example, a feature storage means for executing process S6 that stores features in a storage device (for example, the hard disk device 14 in Figure 2). Similarly, the diagnostic base terminal 103 has a means for executing each of the processes S21 to S27, and has, for example, a capillary imaging means for imaging capillaries by operating the capillary scope 3. The subject terminal 104 has a means for executing each of the processes S31 to S37, and has, for example, a questionnaire request means for requesting a questionnaire. Therefore, FIG. 3 also serves as a functional block diagram of the diagnostic support server 102, diagnostic base terminal 103, and subject terminal 104.

The subject terminal 104 first requests a questionnaire on lifestyle habits from the diagnostic support server 102 (S31). As an example, the subject terminal 104 first accesses the diagnostic support server 102 through the operation of subject 1, downloads a login screen, and transmits a pre-registered user ID (user identification code) and password to the diagnostic support server 102, thereby downloading a menu screen. When subject 1 selects "Request a questionnaire" on the menu screen, the subject terminal 104 transmits a signal requesting the questionnaire to the diagnostic support server 102.

The diagnostic assistance server 102 receives a questionnaire request signal sent by the subject terminal 104 (S1). When the diagnostic assistance server 102 receives the questionnaire request signal, it reads out a questionnaire previously stored in a storage device (e.g., hard disk device 14 in FIG. 2) and sends it to the subject terminal 104 (S2). The subject terminal 104 receives the sent questionnaire (S32). As an example, the questionnaire includes several questions about the lifestyle habits of the subject 1 and options for answering the questions. When the subject terminal 104 receives the questionnaire, it displays a questionnaire answer acceptance screen on a display device (e.g., display device 13 in FIG. 2) and accepts answering operations by the subject 1 (S33).

When subject 1 performs an operation to confirm the answers on the questionnaire answer reception screen displayed on the subject terminal 104, the subject terminal 104 ends the reception of the answers (S33) and transmits the confirmed questionnaire answers to the diagnostic support server 102 (S34). The diagnostic support server 102 receives the transmitted questionnaire answers (S3). The diagnostic support server 102 stores the received questionnaire answers in a storage device (e.g., hard disk drive 14 in FIG. 2) in association with subject 1 (S3A). To identify subject 1, for example, a user ID is used.

Subject 1 visits diagnostic site 7, which may be, for example, a pharmacy, from time to time to have the subcutaneous capillaries, for example, of the fingertip, photographed using capillary scope 3 operated by operator 9 (S21). At the same time, subject 1 can input, for example, his/her own user ID into diagnostic site terminal 103. This allows diagnostic support server 102 to recognize the identity of subject 1 who has had the diagnostic site terminal 103 acquire a capillary image, and subject 1 who has entered questionnaire responses via subject terminal 104. Diagnostic site terminal 103 transmits the acquired capillary image of subject 1 to diagnostic support server 102 (S22).

The diagnostic support server 102 receives the transmitted capillary image of the subject 1 (S4). The diagnostic support server 102 then extracts features from the received capillary image (S5). The features are, for example, the length, thickness, number, degree of turbidity, and curvature of the capillaries. That is, in process S5, the features such as the length of the capillaries are quantified from the received capillary image. A conventionally known technique can be used to extract the features. The diagnostic support server 102 stores the extracted features in a storage device (for example, the hard disk device 14 in FIG. 2) in association with the subject 1 (for example, a user ID) (S6). The part of the storage device that stores the features in association with the subject 1 corresponds to a specific example of the "subject feature storage device" of the present invention. In addition, the part of the storage device that stores the questionnaire answers in association with the subject 1 by process S3A and the part that stores the features in association with the subject 1 by process S6 together correspond to a specific example of the "feature answer storage device" of the present invention.

The diagnostic support server 102 further performs class determination based on the extracted features and the questionnaire responses of the same subject 1 to determine which of multiple pre-defined classes the capillary state represented by the received capillary image corresponds to (S7). The multiple classes are defined so that each class shows a common tendency regarding the capillary state (or the features of the capillaries).

Before or after process S7, the diagnostic support server 102 calculates the capillary rank of the subject 1 based on the feature of the subject 1 extracted in process S5 (S7A). The capillary rank is an index showing how close the extracted feature is to a normal value. After the class determination process S7 is completed, the diagnostic support server 102 reads out the constitution data of the subject 1 (S8). In order to execute process S8, the diagnostic support server 102 stores constitution data, which is data expressing the constitution of the subject 1 corresponding to each of the multiple classes, in a storage device (for example, the hard disk device 14 in FIG. 2) in advance in association with the corresponding class. The part of the storage device that stores the constitution data corresponds to one specific example of the "constitution data storage device" of the present invention. In process S8, the diagnostic support server 102 reads out constitution data corresponding to the class indicated by the result of class determination (S7) from the constitution data storage device.

The diagnostic assistance server 102 reads out the constitution improvement data of the subject 1 (S9) before or after steps S7A and S8. In order to execute step S9, the diagnostic assistance server 102 stores constitution improvement data, which is data expressing advice for improving the constitution of the subject 1 corresponding to each of a plurality of classes, in a storage device (e.g., the hard disk device 14 in FIG. 2) in advance, in association with the corresponding class. The part of the storage device that stores the constitution improvement data corresponds to one specific example of the "constitution improvement data storage device" of the present invention. In step S9, the diagnostic assistance server 102 reads out the constitution improvement data corresponding to the class indicated by the result of the class determination (S7) from the constitution improvement data storage device.

The diagnostic support server 102 reads out past features, which are past features associated with the subject 1, from the subject feature storage device around the time of steps S5 to S9 (S10). The diagnostic support server 102 creates diagnostic support data including the features of the subject 1 extracted in the feature extraction process S5, the result of the class determination made in the class determination process S6, the capillary rank calculated in the capillary rank calculation process S7A, the constitution data read in the constitution data read process S8, the constitution improvement data read in the constitution improvement data read process S9, and the past features read in the past feature read process S10 as transmission data to be sent to the diagnostic base terminal 103 (S11). The diagnostic support server 102 creates the diagnostic support data to be sent as editable data in a format that can be edited by the diagnostic base terminal 103. The diagnostic support server 102 transmits the created editable data to the diagnostic base terminal 103 (S12).

The diagnostic site terminal 103 receives the editable data sent from the diagnostic support server 102 (S23). The diagnostic site terminal 103 edits the editable data based on the operation of the operator 9 at the diagnostic site 7 (S24). This makes it possible to reflect the knowledge and experience of the operator 9, or the results of the questioning of the subject 1 by the operator 9, in the editable data that is useful for diagnosing the subject 1. When editing is complete, the diagnostic site terminal 103 transmits the edited data to the diagnostic support server 102 (S25).

The diagnostic support server 102 receives the transmitted edited data (S13). The diagnostic support server 102 further creates uneditable finalized data based on the received edited data (S14). While the editable data has content and a format intended for an operator, the finalized data is converted to content and a format suitable for presentation to the subject 1. The diagnostic support server 102 transmits the created finalized data to the diagnostic base terminal 103 (S15).

The diagnostic site terminal 103 receives the transmitted confirmation data (S26). The diagnostic site terminal 103 displays the received confirmation data on a display device for the subject 1 (S27). The display device is, for example, the display device 13 illustrated in FIG. 2 or a printing device (printer). This allows the subject 1 to know the diagnosis results regarding his or her own health condition.

The subject 1 can also request the diagnostic support server 102 to send the finalized data to the subject terminal 104. When the subject terminal 104 sends a finalized data request to the diagnostic support server 102 by the subject 1 (S35), the diagnostic support server 102 receives the sent finalized data request (S16). As in the questionnaire request process (S31), the subject terminal 104 can send, for example, the ID and password of the subject 1, so that the diagnostic support server 102 can identify the subject 1. When the diagnostic support server 102 receives the finalized data request, it sends the finalized data of the subject 1 to the subject terminal 104 (S17). The subject terminal 104 receives the sent finalized data (S36). The subject terminal 104 displays the received finalized data on the display device (S37). This allows the subject 1 to know the diagnosis results regarding his/her health condition using the subject terminal 104.

The diagnostic support system 101 operates as described above, and thus has the following advantages. The class determination process (S7) is performed based on the feature values extracted from the captured image of the capillaries of the subject 1, as well as the subject 1's response to a questionnaire about the subject's lifestyle, so that the accuracy of the determination result that supports the subject's diagnosis is improved. Furthermore, the subject 1 can send the response to the questionnaire about his/her lifestyle to the diagnostic support server 102 by himself/herself, without the operator 9 of the diagnostic base terminal 103 knowing it, by using a terminal that he/she manages. That is, the subject 1 can send the response to the questionnaire to the diagnostic support server 102 while protecting the privacy of his/her lifestyle from the operator 9. Furthermore, in process S12, the editable data is sent to the diagnostic base terminal 103, so that the knowledge and experience of the operator 9, or the results of the operator's interview with the subject, can be reflected in the data displayed on the diagnostic base terminal 103 and useful for the diagnosis of the subject 1.

Furthermore, since constitution data useful for diagnosing the constitution of subject 1 and constitution improvement data useful for giving advice for improving the constitution of subject 1 are displayed on the diagnostic base terminal 103, the usefulness of the data displayed on the diagnostic base terminal 103 is improved. Furthermore, since the transmission data, which is diagnostic support data, including the capillary features of subject 1 extracted in process S5 is sent from the diagnostic support server 102 to the diagnostic base terminal 103 (S12), a more detailed diagnosis of the health condition of subject 1 based on the capillary features of subject 1 is possible. Furthermore, since the transmission data, including the past features of subject 1 read out in process S10, is sent from the diagnostic support server 102 to the diagnostic base terminal 103 (S12), it is possible to diagnose the health condition of subject 1 taking into account the history of the features of subject 1.

Figure 4 is a flow chart illustrating an example of the class determination process S7. In order to execute the class determination process S7, the diagnostic support server 102 stores in advance in a storage device (e.g., the hard disk device 14 in Figure 2) the range or representative value of the capillary feature amount corresponding to each of the multiple classes as reference data, and stores coefficient data, which is a set of coefficients that indicate the possibility that each of the multiple possible answers to the questionnaire belongs to each of the multiple classes. The part of the storage device that stores the reference data corresponds to one specific example of the "reference data storage device" of the present invention, and the part that stores the coefficient data corresponds to one specific example of the "coefficient data storage device" of the present invention.

When the diagnostic support server 102 starts the class determination process S7, it first determines which of the multiple expected responses the questionnaire response of subject 1 corresponds to (S41). The diagnostic support server 102 then selects coefficient data corresponding to the determined response from among the coefficient data stored in the storage device (S42). The diagnostic support server 102 further corrects the feature extracted in process S5 by displacing it in a feature space in which each of the multiple classes is represented as a spatial region according to the size of each of the sets of coefficients contained in the coefficient data selected in process S42 (S43). The diagnostic support server 102 further performs class determination by determining which of the multiple classes the feature corrected in process S43 corresponds to by referring to the reference data stored in the storage device (S44).

The questionnaire responses of subject 1 received by the questionnaire response receiving process S3 (FIG. 3) are stored in the storage device in association with subject 1 by the questionnaire response storage process S3A, and the capillary features of subject 1 extracted by the feature extraction process S5 are stored in the storage device in association with subject 1 by the feature storage process S6. Thus, the features and the questionnaire responses are stored in association with each other in the storage device. As the diagnostic support system 101 is used, the features and the questionnaire responses stored in association with each other in the storage device can be used to reset the capillary conditions into multiple classes and to reset the coefficient data. This enables more accurate class determination based on more data.

[2. Detailed examples of diagnostic support systems]
A more detailed example of the procedure illustrated in FIGS. 3 and 4 will be described below.

(2-1. Survey acceptance S33)
5 is a screen diagram illustrating a questionnaire response reception screen displayed on the display device of the subject terminal 104. In the illustrated example, one of several questions and answer options for that question are displayed. As illustrated, in an example in which the subject terminal 104 is a smartphone, the subject 1 can select one of the multiple options as an answer by touching with his/her finger. Table 1 illustrates questions included in the questionnaire and the answer options for each question.

(2-2. Class Determination S7, S41 to S44)
To obtain the reference data used in the class determination S7, for example, a group of capillary images obtained by sampling from a group of subjects is first prepared. The larger the group of capillary images, the more accurate the reference data and coefficient data that can be obtained. From the group of capillary images, similarly to process S5 (FIG. 3), the capillary feature values are extracted. For example, the feature values include five types of parameters, namely, the length, number, curvature, thickness, and turbidity of the capillaries. Therefore, the feature values can be made to correspond to feature vectors, which are vectors in a feature space, which is a five-dimensional space with each of these five parameters as its coordinate axis.

Since the distance between feature vectors can be defined in the feature space, cluster analysis can be performed on a group of feature vectors obtained from a group of capillary images based on the distance between the feature vectors. The group of feature vectors is clustered (also referred to as "classification" in this application) into multiple clusters (also referred to as "classes" in this application) by cluster analysis. Since the cluster analysis itself is a well-known technique, the procedure will not be described here. The cluster analysis employed is, for example, hierarchical cluster analysis. Table 2 illustrates the results of classification obtained by extracting features from a group of capillary images prepared as samples and applying hierarchical cluster analysis to the group of feature vectors obtained by the extraction. The data illustrated in Table 2 was obtained based on 8,491 capillary images experimentally collected by the applicant of this application from 660 subjects 1.

In the classification shown in Table 2, the blood vessels are divided into six classes tentatively named classes A through F. As an example of the representative values of parameters such as length and number, i.e., the representative values of features, the medians are calculated for each class and are shown in Table 2. The set of medians of capillary length, number, etc. can be made to correspond to a median vector, which is a five-dimensional vector whose components are these. For example, the median vector of class A has components (759.07, 4, 4.92762, 48.8672, 0.445833).

The representative values of these parameters are stored in the reference data storage device in association with each of classes A to F. That is, Table 2 corresponds to an example of the contents stored in the reference data storage device in tabular form. As the diagnostic support system 101 is used, feature values are accumulated in the storage device by process S6 (see FIG. 3). By redoing the cluster analysis using the accumulated feature values, it becomes possible to perform cluster analysis based on a larger group of feature values. This further improves the accuracy of the cluster analysis and the accuracy of the representative values. The contents stored in the reference data storage device can be updated with the newly obtained representative values with higher accuracy. These processes are performed, for example, by an administrator or user of the diagnostic support server 102.

The coefficient data stored in the coefficient data storage device is used by the diagnostic support server 102 to correct the feature quantities of subject 1 extracted in process S5 (see FIG. 3) based on the questionnaire responses received in process S3. To obtain the coefficient data, questionnaire responses are simultaneously obtained from each of a group of subjects from whom a group of capillary images were taken. Table 3 illustrates the contents of the corresponding questionnaire responses and the results of class determination for each subject 1. The data illustrated in Table 3 was obtained based on 104 capillary images experimentally taken from 104 subjects 1 other than the above 660 subjects, and responses to questions equivalent to the questionnaire.

In the example of Table 3, subject 1, from whom a capillary image with image ID "3" was taken, is not physically inactive (physical inactivity = 0), is taking medication (taking medication = 1), is not smoking (smoking = 0), and is determined to be in "Class D" for the capillary class. The class determination here means a determination as to which of classes A to F shown in Table 2 the feature vector of the capillary image corresponding to each image ID belongs. As an example, a class is selected in which the feature vector of the capillary image of each image ID corresponds to the closest median vector (i.e., the closest distance in feature space) of the median vectors of classes A to F shown in Table 2. For example, the capillary image with image ID "3" is determined to belong to class D because its feature vector is closest to the median vector of class D among the median vectors of classes A to F.

The distance between vectors in the feature space is defined as follows, for example. As mentioned above, the feature vector is a five-dimensional vector whose components include different types of parameters, such as the length, number, curvature, thickness, and turbidity of the capillaries. Therefore, the variance of a group of feature vectors generally differs for each dimension (i.e., each component). To balance the effects of dimensions with large and small variances, the Mahalanobis distance can be used as an example of the distance between the feature vector corresponding to each image ID and the median vector of each capillary class A to F.

If the feature vector corresponding to a certain image ID is x ^→ = (x_1, x_2, x_3, x_4, x_5) ^T , the median vector of a certain capillary class is y ^→ = (y_1, y_2, y_3, y_4, y_5) ^T , and the variance vector with the variance of each dimension as components is σ ^→ = (σ_1, σ_2, σ_3, σ_4, σ_5) ^T , the Mahalanobis distance d between vector x ^→ and vector y ^→ is given by Equation 1. That is, the Mahalanobis distance d corresponds to the distance in the feature space normalized by the variance of each dimension. Here, the symbol " ^→ " represents a vector, and the symbol " ^T " represents the interchange of rows and columns, that is, the transposition of a matrix.

Coefficient data can be obtained by, for example, performing decision tree analysis on the contents of the questionnaire responses for each subject 1 and the data on capillary classes, as shown in Table 3. The decision tree is created so that the purity of the classes contained in the nodes (uniqueness of the classes contained) increases with each progression of the questionnaire questions.

Figure 6 is an explanatory diagram that visualizes a decision tree as an example of a procedure for calculating coefficient data for each of multiple possible responses to a questionnaire. In Figure 6, the numbers in the boxes at the root, nodes, and leaves indicate the number of classes A, B, C, D, E, and F, from left to right, and the alphabet in the box indicates the most common class. In the example shown, of the 104 subjects 1, there are 7 people who are taking medication ((medicine = 0) = no; in other words, "NO" is selected at the branch for "medicine = 0"), do not have lower back pain or stiff shoulders ((lower back pain and stiff shoulders = 1) = no), and are not lacking in exercise ((lack of exercise = 1) = no), and one of them is in class B capillary class, and the remaining 6 are in class F capillary class.

In this way, in a decision tree that classifies into six classes from class A to class F, the proportion of each class p ^→ = (p_A, p_B, p_C, p_D, p_E, p_F) ^T is calculated for each leaf. Coefficient data can be expressed as a coefficient vector by vector p ^→ . The coefficient vector corresponding to the answer in the above example is p ^→ = (p_A, p_B, p_C, p_D, p_E, p_F) ^T = (0, 1/7, 0, 0, 0, 6/7) ^T. A coefficient vector p ^→ is determined for each of the multiple answers to the questionnaire, i.e., for each leaf of the decision tree.

As the diagnostic support system 101 is used, the questionnaire responses are accumulated in the storage device along with the features already mentioned (process S3A). By using the accumulated features to redo the cluster analysis and redo the decision tree analysis based on the questionnaire responses, it is possible to obtain a larger set of features and coefficient data based on the questionnaire responses. This further improves the accuracy of the coefficient data. The content stored in the coefficient data storage device can be updated with the newly obtained coefficient data with high accuracy. These processes are performed, for example, by an administrator or user of the diagnostic support server 102.

In process S41 (FIG. 4), it is determined whether the questionnaire response of subject 1 corresponds to any one of a plurality of responses illustrated in FIG. 6. For example, it is determined that the questionnaire response of subject 1 corresponds to the above-mentioned example, that is, taking medicine ((medicine=0)=no), not having back pain and stiff shoulders ((back pain and stiff shoulders=1)=no), and not lacking in exercise ((lack of exercise=1)=no). In this case, in process S42, the coefficient vector p ^→ =(0, 1/7, 0, 0, 0, 6/7) ^T is selected as the coefficient data corresponding to the questionnaire response of subject 1. Next, an example of the procedure of process S43 (FIG. 4) for correcting the feature amount extracted in process S5 (FIG. 3) and process S44 for determining the class to which the corrected feature amount belongs will be described in detail.

Let the median vector of a capillary class Z (Z = any of A to F) be (y_Z) ^→ = (y_1Z, y_2Z, y_3Z, y_4Z, y_5Z) ^T , and define the median matrix Y of classes A to F as shown in Equation 2.

Furthermore, the correction vector c ^→ = (c_1, c_2, c_3, c_4, c_5) is defined by Equation 3.

The correction vector c ^→ represents the similarity of the leaf of the decision tree corresponding to the questionnaire response of subject 1, represented by the coefficient vector p ^→ , to each of the capillary classes A to F. For example, when the coefficient vector p ^→ = (1, 0, 0, 0, 0, 0) ^T , whose only component is class A, the correction vector is c ^→ = (y_1A, y_2A, y_3A, y_4A, y_5A), which coincides with the median vector of class A.

Clustering using capillary vessel images uses only information on capillary vessel images. Therefore, a correction that reflects the user's lifestyle is performed using a correction vector c ^→ generated from information based on questionnaire responses. The feature vector representing the feature extracted in process S5 is set as x ^→ = (x_1, x_2, x_3, x_4, x_5) ^T. The corrected feature vector (x') ^→ is expressed by Equation 4, for example, using the pre-correction feature vector x ^→ and the correction vector c ^→ . Equation 4 means that the corrected feature vector is determined midway between the feature vector and the correction vector in the feature space.

Figure 7 is an explanatory diagram illustrating the procedure for feature correction executed according to Equation 4. In Figure 7, for simplification, the five-dimensional feature space is replaced with a two-dimensional space in which the only dimensions are the length and thickness of the capillaries. It is also assumed that the state of the capillaries is classified into three classes, A, B, and C. In Figure 7, the solid circle marked with the symbol "?" represents the position of the pre-correction feature vector corresponding to the feature newly extracted in process S5. Similarly, the solid circle marked with the symbols "A", "B", and "C" represent the positions of the median vectors of classes A, B, and C.

In the illustrated example, the feature vector before correction belongs to class A but is close to class B in the area of classes A, B, and C, and the positions of the feature vector after correction in the cases where the questionnaire response is class A, B, or C are indicated by dotted circles. For example, "the questionnaire response is class B" means that the coefficient vector is p ^→ = (p_A, p_B, p_C) ^T = (0, 1, 0) ^T. When the questionnaire response is class B, the feature vector before correction approaches the median vector of class B (solid circle with the symbol "B") through correction, and is at the position indicated by the dotted circle with the symbol "B". In this manner, process S43 is executed to correct the feature extracted in process S5.

Which class the corrected feature vector belongs to can be determined, for example, by which class the median vector closest to the corrected feature vector is. For example, in Equation 1, the corrected feature vector is taken as vector x ^→ , and the distance between the corrected feature vector and the median vector y ^→ of each class is calculated, and it can be determined that the corrected feature vector x ^→ belongs to the class corresponding to the median vector y ^→ with the smallest distance.

In the example of FIG. 7, the corrected feature vector represented by the dotted circle marked with the letter "B" is closest to the median vector of class B among classes A, B, and C, and therefore the corrected feature is determined to belong to class B. In this manner, process 44 is executed to determine the class to which the corrected feature belongs. By applying corrections to the feature vector in this manner, a multifaceted evaluation can be made that takes into account not only the shape of the capillaries but also the questionnaire responses of subject 1.

(2-3. Capillary rank calculation S7A)
The capillary rank is designed so that the closer the parameters (i.e., features) of the capillaries, such as length, thickness, turbidity, curvature, and number, are to the standard values of normal capillaries, the higher the rank. Each parameter is scored, for example, in the range of 0 to 100 points, and the ranking is determined by the total score. The threshold value for determining the rank can be updated, for example, by analyzing the features accumulated in the storage device by process S6.

As one method of scoring, scoring based on normal distribution is possible. In this method, the deviation from the normal value is quantified using a normal distribution function as exemplified in FIG. 8. If the input value is x, the normal value is μ, and the variance of the population is σ, the score p _score is given by Equation 5. The normal value μ can be obtained, for example, from the median of a group of normal samples.

However, Equation 5 can only be applied to parameters that are evaluated low whether their values are higher or lower than the normal value. For example, if the parameter is the thickness of the capillary, the Gaussian function shown in FIG. 8 can be used as the scoring function. On the other hand, for parameters that are evaluated higher as their values increase, such as the length of the capillary, a hyperbolic function such as the hyperbolic tangent shown in FIG. 9 can be used.

The total score for each parameter (for example, a range of 0 to 500 points) may be used as the capillary rank, or the total score may be divided into several ranges to determine the capillary rank. For example, a total score range of 0 to 200 points may be assigned to rank C, a range of 201 to 400 points to rank B, and a range of 401 to 500 points to rank A.

(2-4. Creating editable data S11)
In a preferred embodiment, the editable data created by the editable data creation process S11 and transmitted to the diagnostic site terminal 103 by the editable data transmission process S12 includes data representing two types of screens to be displayed on the display device of the diagnostic site terminal 103: a diagnostic suggestion screen and a diagnostic element selection screen.

Figure 10 is a screen diagram illustrating a diagnosis suggestion screen displayed on the display device of the diagnostic site terminal 103. The diagnosis suggestion screen is a page on which the operator 9 can confirm the analysis results of the capillary blood vessel image, and cannot be edited. The operator 9 uses the information on this page to help formulate questions about the user's constitution.

Figure 11 is a screen diagram illustrating a diagnostic element selection screen displayed on the display device of the diagnostic site terminal 103. The diagnostic element selection screen is an editable page for selecting elements that will ultimately constitute the diagnostic statement to be presented to the user, based on the content presented on the diagnostic suggestion screen and new information obtained through questions. The operator 9 can display the data of these screens received by the diagnostic site terminal 103 by the editable data reception process S23 on the display device of the diagnostic site terminal 103 by switching pages as needed.

In the diagnostic suggestion screen illustrated in FIG. 10, the upper left column displays the ID of subject 1, as well as the capillary rank and capillary type (synonymous with capillary class) determined by the diagnostic support server 102. The upper right column displays the scores for each capillary parameter in the form of a radar chart. The lower right column displays the capillary image captured by the capillary imaging process S21. In this way, the editable data transmitted to the diagnostic base terminal 103 by the editable data transmission process S12 can also include the scores for each capillary parameter and the capillary image.

The contents of the diagnostic suggestions for each item (i.e., suggestions that assist the operator 9 in making a diagnosis) are displayed in the lower left column. In the illustrated example, suggestions regarding the shape of the capillaries, suggestions regarding tendencies in physical constitution, and suggestions regarding advice for improving physical constitution are included. The contents of these suggestions correspond to default diagnostic results created based on the capillary class determined by the diagnostic support server 102. These default diagnostic results can be edited by the operator 9 on the diagnostic element selection screen. In other words, the final diagnosis is left to the operator 9.

In order to create suggestions for the shape of capillaries, the diagnostic support server 102 stores shape representation data that represents the shape of capillaries corresponding to each of a plurality of classes in a storage device (e.g., the hard disk device 14 in FIG. 2) in association with the corresponding class. The diagnostic support server 102 reads out shape representation data that corresponds to the capillary class determined for the subject 1 from the storage device, and includes the read shape representation data in the diagnostic suggestion screen as a suggestion for the shape of capillaries. The suggestions for constitution correspond to the constitution data read out by the constitution data read out process S8. The suggestions for advice for improving constitution correspond to the constitution improvement data read out by the constitution improvement data read out process S9. As already mentioned, in order to execute processes S8 and S9, the diagnostic support server 102 stores constitution data and constitution improvement data that correspond to each of a plurality of classes in a storage device (e.g., the hard disk device 14 in FIG. 2) in association with the corresponding class.

Table 4 explains the names and roles of the components of the diagnostic element selection screen shown in Figure 11.

The selection buttons located in the shape selection area (1), trend selection area (2), and advice selection area (3) are displayed in different colors so that selected and unselected buttons can be visually identified. In the illustrated example, for example, "low," "muddy," and "bypass" are selected in the shape selection area (1). Each time the button is clicked, it switches between selected and unselected states.

The selected/non-selected state of the selection button in the shape selection area (1), in the default state before the operator 9 operates it, corresponds to a suggestion about the shape of capillaries as shown in FIG. 10. Similarly, the selected/non-selected state of the selection button in the tendency selection area (2), in the default state before the operator 9 operates it, corresponds to a suggestion about physical constitution tendencies as shown in FIG. 10. Furthermore, the selected/non-selected state of the selection button in the advice selection area (3), in the default state before the operator 9 operates it, corresponds to a suggestion about advice for improving physical constitution as shown in FIG. 10.

There is a limit to the number of buttons that can be selected. For example, the limit is three for shape selection, three for trend selection, and one for advice selection. The currently selected number is displayed in fractional form at the top right corner of each area (1) to (3), along with the limit. If you click a button in an attempt to select more than the limit, the click will be invalid.

When the diagnosis generation button (7) is clicked if none of the shape selection, trend selection, or advice selection buttons are selected, a message is displayed in red. For example, the message "Please select a diagnosis element button" is displayed below the diagnosis generation button (7).

The three elements of the shape of the capillaries, the constitution of the subject 1, and advice on improving the constitution are collectively defined as diagnostic elements. Details of each element are exemplified below.
The shape selection in the shape selection area (1) selects the shape characteristics of the capillary. There are 12 options, as shown in Table 5, for example.

The constitution selected in the tendency selection area (2) is the one that is thought to apply to the constitution of subject 1. As an example, there are 12 options, as shown in Table 6.

The advice selection selected in the advice selection area (3) selects the content of advice to be given to the subject 1 based on the operator 9's specialized knowledge and experience. As an example, there are six options, as shown in Table 7.

The diagnostic sentence (4) is automatically generated from the selected diagnostic elements. It is also possible to further edit the text. The operator 9 may create the entire sentence from the beginning. Editing and creation of the entire sentence can be performed using an input device such as a keyboard or mouse (for example, the input device 11 illustrated in FIG. 2). For example, bullets or a synthesis algorithm are used to automatically generate the diagnostic sentence (4), that is, to synthesize the sentences. For example, the data illustrated in Table 8 is used as the synthesis algorithm. For example, if "thick" is selected for the shape of the capillaries, the sentence "The capillaries are thick" is synthesized into the diagnostic sentence (4). The diagnostic support server 102 stores the contents of Table 8 in advance in a storage device (for example, the hard disk drive 14 in FIG. 2).

Regarding the contents of the diagnostic suggestions for each capillary class, a diagnostic suggestion table as shown in Table 9 is created in advance and stored in the storage device of the diagnostic support server 102 (for example, the hard disk drive 14 in FIG. 2). That is, the diagnostic support server 102 also stores the contents of Table 9 in advance in the storage device. When the diagnostic support server 102 determines the capillary class of subject 1 by the class determination process S7, it refers to the diagnostic suggestion table and reads out the diagnostic suggestions corresponding to the determined capillary class (processes S8, S9, etc.). The contents of Table 9 can be improved based on the data collected in the storage device when the diagnostic support server 102 is maintained or upgraded. These operations are performed, for example, by the administrator or user of the diagnostic support server 102.

The function of generating diagnostic sentence (4) will be further explained. The structure of the sentence is, for example, "capillary shape → tendency that can be read from it → advice based on them." A template can be used to generate the sentence. For example, the template shown in Table 10 is stored in the storage device of the diagnostic support server 102 (for example, the hard disk drive 14 in FIG. 2). The diagnostic support server 102 reads out diagnostic suggestion sentence fragments corresponding to the capillary class by referring to the diagnostic suggestion table shown in Table 9, and fills in each element of the template shown in Table 10 with the read out sentence fragments. In Table 10, elements A1, A2, and A3 are assigned to the shape of the capillaries, elements B1, B2, and B3 to tendencies of the constitution, and element C1 to sentence fragments related to advice.

It is desirable to design the sentence fragments so that they match the template and can be synthesized without any awkwardness. The table shown in Table 11 shows an example of more sophisticated sentence fragments that correspond to the diagnostic suggestions shown in Table 8. The table shown in Table 11 is also stored in advance in the storage device of the diagnostic support server 102 (for example, the hard disk drive 14 in FIG. 2). By referring to the diagnostic suggestion table shown in Table 9, the diagnostic support server 102 reads out diagnostic suggestions corresponding to the capillary class, and reads out sentence fragments corresponding to the read out diagnostic suggestions from Table 11. The diagnostic support server 102 then fills in each element of the template shown in Table 10 with the read out sentence fragments.

Table 12 shows an example of a diagnostic statement (4) created in this way.

(2-5. Creating Final Data S14)
12 is a screen diagram illustrating a diagnosis result screen displayed on the display device of the diagnostic site terminal 103. The diagnosis result screen is created by the diagnostic assistance server 102 as a screen for displaying the finalized data through the finalized data creation process S14. In the illustrated example, the diagnosis result screen is configured in a format that can be printed on A4 portrait paper, and includes the following seven elements.

(1) Header The header is located at the top of the screen and displays information such as the date and location.
(2) Basic Data Section The basic data section displays basic data and information that is a summary of the diagnosis results. It includes the following items:
Identification code (ID) of subject 1. This is input by subject 1 or operator 9 to diagnostic base terminal 103 before or after capillary blood vessel imaging process S21.
Gender: The subject 1 or the operator 9 inputs this into the diagnostic base terminal 103 before or after the capillary blood vessel imaging process S21.
Age: The subject 1 or the operator 9 inputs this into the diagnostic base terminal 103 before or after the capillary blood vessel imaging process S21.
・Capillary rank.
・Capillary class.
(3) Capillary Image Section The capillary image section displays the capillary image of the subject 1 that was captured in the capillary imaging process S21 and used for analysis by the diagnosis assistance server 102.
(4) Two-dimensional barcode unit The two-dimensional barcode unit displays a two-dimensional barcode for displaying the diagnosis result on the subject terminal 104. When the subject terminal 104 is a smartphone, the subject terminal 104 reads the two-dimensional barcode, thereby executing the finalized data request process S35 ( FIG. 3 ), and then processes S16, S17, S36, and S37, whereby the finalized data illustrated in FIG. 12 is displayed on the display device of the subject terminal 104.
(5) Time-series data section The time-series data section displays a line graph with the horizontal axis representing the date and the vertical axis representing the score of the capillary parameter. That is, the time-series data section displays the change in the score of the capillary parameter over time. Displaying the capillary parameter (i.e., the feature of the capillary) as a score is a preferred form of displaying the feature of the capillary.
(6) Parameter Section The parameter section displays the current capillary parameters, which have been converted into scores, in a radar chart format.
(7) Analysis Result Section The analysis result section displays text summarizing the results of the analysis and diagnosis. The illustrated example is composed of three types of text.
In the analysis result section, the capillary rank description is created by referring to Table 13, for example.

The description of the capillary class in the analysis results section is created by referring to the table in Table 14, for example.

The tables exemplified in Tables 13 and 14 are also stored in advance in the storage device of the diagnostic support server 102 (e.g., the hard disk drive 14 in FIG. 2). The contents of Tables 13 and 14 can also be improved based on the data collected in the storage device when the diagnostic support server 102 is maintained or upgraded.

In the analysis result section, the advice text displays the contents of the diagnostic text after editing on the diagnostic element selection screen illustrated in FIG. 11. In this way, the diagnostic support server 102 can create finalized data (S14) by making appropriate changes to the screen format and text style/content of the edited data edited (S24) by the diagnostic base terminal 103, and changing it into a form that is easy for the subject 1 to receive the information.

[3. Construction, operation, and maintenance of diagnostic support server 102]
Next, the procedure for constructing, operating, and maintaining the diagnostic support server 102 will be summarized as above. FIG. 13 is a flowchart illustrating the procedure for constructing, operating, and maintaining the diagnostic support server 102. The work according to the procedure illustrated in FIG. 13 is performed, for example, by an administrator of the diagnostic support server 102 or a user of the diagnostic support server 102. Before operating the diagnostic support server 102, the diagnostic support server 102 is naturally constructed. The diagnostic support server 102 is constructed by creating a program that specifies the operation of the diagnostic support server 102 and storing it in the storage device (for example, the hard disk device 14 in FIG. 2) of the diagnostic support server 102 (S51), and creating various reference data referenced by the program and storing it in the storage device (for example, the hard disk device 14 in FIG. 2) of the diagnostic support server 102 (S52).
Table 15 shows examples of reference data. All of the reference data D1 to D10 have already been described, so detailed explanations of each will be omitted. Table 15 also shows the correspondence between each of the reference data D1 to D10 and specific examples such as tables already shown.

When construction of the diagnostic support server 102 is completed, the diagnostic support server 102 is put into operation (S53). During operation, the diagnostic support server 102 operates automatically according to a program. The operation of the diagnostic support server 102 during operation is illustrated in FIG. 3 as steps S1 to S17. Unless the administrator of the diagnostic support server 102 or the like determines that maintenance should be performed on the diagnostic support server 102 or that operation of the diagnostic support server 102 should be terminated (S54), the diagnostic support server 102 automatically repeats the operation (S53), for example steps S1 to S17. As operation (S53) is repeated, data such as feature values and questionnaire responses is accumulated in the storage device of the diagnostic support server 102.

When the administrator or the like determines that the diagnostic assistance server 102 should be maintained (S54), the administrator or the like moves on to the maintenance work. In the maintenance work, new clustering analysis and decision tree analysis are performed based on data such as feature quantities and questionnaire responses stored in the storage device, and new reference data D1 to D10 is created based on the results. The reference data D1 to D10 stored in the storage device is replaced with the newly created data from the reference data D1 to D10 (S52). In this way, the reference data D1 to D10 stored in the storage device is updated as appropriate during repeated operation. This makes it possible to provide diagnostic assistance data that is useful for diagnosing the health condition of the subject 1 with higher accuracy.

When the maintenance work of the diagnostic support server 102 is completed by completing the update of the reference data D1 to D10 (S52), the diagnostic support server 102 is put back into operation (S53). When the administrator or the like determines that the operation of the diagnostic support server 102 should be terminated (S54), the operation of the diagnostic support server 102 is terminated. The construction work and maintenance work of the diagnostic support server 102 may be performed automatically according to a program. For example, a maintenance program may be incorporated into the diagnostic support server 102, and when performing maintenance, the diagnostic support server 102 itself may newly perform clustering analysis and decision tree analysis based on data such as feature amounts and questionnaire responses stored in the storage device, and based on the results, newly create at least a portion of the reference data D1 to D10 and store it in the storage device.

[4. Other embodiments]
In the diagnostic support system 101, the questionnaire is displayed on the subject terminal 104, and the questionnaire responses are sent to the diagnostic support server 102 via the subject terminal 104. Alternatively, the questionnaire may be displayed on the diagnostic site terminal 103, and the questionnaire responses may be input to the diagnostic site terminal 103 by the operation of the subject 1 or the operator 9, and then sent from the diagnostic site terminal 103 to the diagnostic support server 102. In this case, the subject terminal 104 is not necessary.

The data created by process S11 (FIG. 3) as data including the results of class determination, etc., and sent from the diagnostic support server 102 to the diagnostic base terminal 103 by process S12 may be in a format that is determined from the beginning, instead of an editable format. In this case, the processes following edit process S24 (S24, S25, S13, S14, S15, S26) are unnecessary.

The class determination process (S7) can also be performed without answering the questionnaire. In this case, the class of the features can be determined in the same manner as in process 44 (Figure 4) without correcting the features extracted in process S5.

In the diagnosis result screen illustrated in FIG. 12, the analysis result section includes a capillary rank description (contents of Table 13) and a capillary class description (contents of Table 14). In contrast, it is also possible to adopt a form in which the capillary rank and capillary class are not accompanied by their description. These descriptions can be conveyed to the subject 1 by another medium, for example, a printed matter such as a pamphlet, a website provided as part of the diagnosis support server 102 or separately therefrom, or a verbal explanation by the operator 9. Conversely, for example, there may be no display of the capillary class (for example, a display such as "Class B"), and only the capillary class description (for example, a description corresponding to Class B) may be displayed. In either case, the result of the class determination is displayed, and the display format is different. In addition, any display format can be useful for diagnosing the health condition of the subject 1.

In the explanation of the specific example of the class determination process S7, the "representative value" of the capillary feature corresponding to each of the multiple classes is used as the "reference data", and in particular, an example of using the "median" as the representative value is shown. In contrast to this, it is also possible to use other representative values, such as the "average value" or the "mode" (the coordinate value at which the features belonging to each class are most dense in the feature space), instead of the "median". Furthermore, instead of the "representative value", the range of the capillary feature corresponding to each of the multiple classes may be used as the "reference data" instead of the "representative value". In the example of FIG. 7, the "feature range" corresponds to a polygonal region surrounded by the boundaries of each class. In this case, the class determination of the feature vector for determining the coefficient vector (for example, the determination of the capillary class in Table 3) or the class determination of the corrected feature vector can be achieved by determining which region these feature vectors belong to. The "coefficient data" can be derived in the same way as the coefficient vector p ^→ , for example, by the procedure already exemplified. To derive a corrected feature vector from the feature vector, for example, the median value of each region is calculated, and a median value matrix based on these median values is used in place of Equation 2, and further Equations 3 and 4 can be used.

1 Subject, 3 Capillary scope, 5 Network, 7 Diagnostic base, 8 Residence, 9 Operator, 11 Input device, 12 Main circuit section, 13 Display device, 14 Hard disk drive (HDD), 15 LAN, 21 CPU (Central Processing Unit), 22 ROM (Read Only Memory), 23 RAM (Random Access Memory), 24 Interface (I/F), 25 Reading device, 26 Bus line, 101 Diagnostic support system, 102 Diagnostic support server (server), 103 Diagnostic base terminal (first terminal, diagnostic support terminal), 104 Subject terminal (second terminal).

Claims (22)

A diagnostic support system for supporting a diagnosis of a health condition of a subject based on a capillary image of the subject, comprising:
A system comprising: a server; a first terminal connectable to the server through a network; and a second terminal which is the same as or different from the first terminal and connectable to the server through the network,
The first terminal is
an imaging means for imaging the capillaries of the subject to obtain the capillary image;
an image transmitting means for transmitting the capillary blood vessel image to the server,
The server,
a questionnaire sending means for sending a questionnaire regarding the subject's lifestyle habits to the second terminal in response to a request from the second terminal;
A response receiving means for receiving responses to the questionnaire by the subject, the responses being transmitted from the second terminal;
an image receiving means for receiving the capillary blood vessel image transmitted by the first terminal;
a feature extraction means for extracting a feature of the capillaries from the capillary image received by the image receiving means;
a class determination means for determining to which of a plurality of preset classes the state of the capillaries represented by the received capillary image corresponds, based on the extracted feature amount and the received response to the questionnaire; and
a data transmission means for transmitting data including the result of the class determination to the first terminal as transmission data;
The first terminal is
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. The diagnostic support system according to claim 1, wherein the second terminal is different from the first terminal. The data transmission means of the server transmits the transmission data to the first terminal as editable data in a format editable by the first terminal;
The first terminal is
3. The diagnostic support system according to claim 1, further comprising an editing unit that edits the editable data received by the data receiving unit in accordance with an input operation to the first terminal. The first terminal is
The method further includes an edited data transmission means for transmitting the editable data that has been edited to the server as edited data,
The server,
a post-edit data receiving means for receiving the post-edit data transmitted by the post-edit data transmitting means;
a finalized data creating means for creating finalized data that cannot be edited based on the received edited data;
The method further includes a finalized data transmission means for transmitting the finalized data to the first terminal,
The first terminal is
a confirmed data receiving means for receiving the confirmed data transmitted by the confirmed data transmitting means;
4. The diagnosis support system according to claim 3, further comprising a confirmed data display means for displaying the received confirmed data. The server,
a constitution data storage device that stores constitution data, which is data expressing the constitution of a subject and corresponds to each of the plurality of classes, in association with the corresponding class;
a constitution data reading means for reading, from the constitution data storage device, constitution data corresponding to a class indicated by a result of the class determination, among the plurality of classes;
5. The diagnostic support system according to claim 1, wherein the data transmission means transmits the transmission data including the read-out constitution data to the first terminal. The server,
a constitution improvement data storage device that stores constitution improvement data, which corresponds to each of the plurality of classes and expresses advice for improving the constitution of the subject, in association with the corresponding class;
a constitution improvement data reading means for reading, from said constitution improvement data storage device, constitution improvement data corresponding to the class indicated by the result of said class determination, among said plurality of classes;
6. The diagnostic support system according to claim 1, wherein said data transmitting means transmits said transmission data including said read constitution improvement data to said first terminal. The diagnostic support system according to any one of claims 1 to 6, wherein the data transmission means of the server transmits the transmission data including the extracted feature amount to the first terminal. The server,
A subject feature storage device;
a subject feature storage means for storing the extracted feature in the subject feature storage device in association with the subject;
and a past feature reading means for reading, from the subject feature storage device, a past feature that is a past feature associated with the subject,
8. The diagnosis support system according to claim 7, wherein the data transmission means of the server transmits the transmission data to the first terminal, the transmission data including the read past feature amount of the subject. The server,
a reference data storage device that stores a range or a representative value of the capillary feature amount corresponding to each of the plurality of classes as reference data;
a coefficient data storage device that stores coefficient data, which is a set of coefficients representing the possibility that each of a plurality of possible answers to the questionnaire belongs to each of the plurality of classes;
The class determination means
an answer determination means for determining which of the plurality of expected answers the answer received by the answer receiving means corresponds to;
a coefficient data selection means for selecting coefficient data stored in the coefficient data storage device and corresponding to the answer determined by the answer determination means;
a feature amount correction means for correcting the feature amount extracted by the feature amount extraction means by displacing the feature amount within a feature amount space in which each of the plurality of classes is represented as a spatial domain in accordance with the magnitude of each of the sets of coefficients included in the coefficient data selected by the coefficient data selection means;
and a feature determination means for determining which of the plurality of classes the feature corrected by the feature correction means belongs to by referring to the reference data stored in the reference data storage device, thereby making the class determination. the reference data storage device stores, as the reference data, representative values of the capillary feature amounts corresponding to each of the plurality of classes;
the feature amount correction means corrects the feature amount extracted by the feature amount extraction means by displacing the feature amount within the feature amount space so as to approach the representative value of the feature amount of each of the plurality of classes in accordance with a magnitude of each of the sets of coefficients included in the coefficient data selected by the coefficient data selection means;
10. The diagnostic support system according to claim 9, wherein the feature determination means determines, among the plurality of classes, a class whose representative value in the feature space is closest to the feature corrected by the feature correction means, as the class to which the corrected feature corresponds. The server,
A feature answer storage device;
11. The diagnostic support system according to claim 9, further comprising a feature answer storage means for storing the feature extracted by the feature extraction means and the answers to the questionnaire received by the answer receiving means in the feature answer storage device in association with each other. A diagnostic support system for supporting a diagnosis of a health condition of a subject based on a capillary image of the subject, comprising:
A server and a first terminal connectable to the server through a network,
The first terminal is
an imaging means for imaging the capillaries of the subject to obtain the capillary image;
an image transmitting means for transmitting the capillary blood vessel image to the server,
The server,
an image receiving means for receiving the capillary blood vessel image transmitted by the first terminal;
a feature extraction means for extracting a feature of the capillaries from the capillary image received by the image receiving means;
a class determination means for determining to which of a plurality of preset classes the state of the capillaries represented by the received capillary image corresponds based on the extracted feature amount; and
a data transmission means for transmitting data including the result of the class determination to the first terminal as transmission data;
The first terminal is
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.
The data transmission means of the server transmits the transmission data to the first terminal as editable data in a format editable by the first terminal;
The first terminal is
The diagnostic support system further comprises an editing means for editing the editable data received by the data receiving means in accordance with an input operation to the first terminal. The first terminal is
The method further includes an edited data transmission means for transmitting the editable data that has been edited to the server as edited data,
The server,
a post-edit data receiving means for receiving the post-edit data transmitted by the post-edit data transmitting means;
a finalized data creating means for creating finalized data that cannot be edited based on the received edited data;
The method further includes a finalized data transmission means for transmitting the finalized data to the first terminal,
The first terminal is
a confirmed data receiving means for receiving the confirmed data transmitted by the confirmed data transmitting means;
13. The diagnosis support system according to claim 12, further comprising a confirmed data display means for displaying the received confirmed data. A diagnostic support system for supporting a diagnosis of a health condition of a subject based on a capillary image of the subject, comprising:
A server and a first terminal connectable to the server through a network,
The first terminal is
an imaging means for imaging the capillaries of the subject to obtain the capillary image;
an image transmitting means for transmitting the capillary blood vessel image to the server,
The server,
an image receiving means for receiving the capillary blood vessel image transmitted by the first terminal;
a feature extraction means for extracting a feature of the capillaries from the capillary image received by the image receiving means;
a class determination means for determining to which of a plurality of preset classes the state of the capillaries represented by the received capillary image corresponds based on the extracted feature amount; and
a data transmission means for transmitting data including the result of the class determination to the first terminal as transmission data;
The first terminal is
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.
The server,
a constitution data storage device that stores constitution data, which is data expressing the constitution of a subject and corresponds to each of the plurality of classes, in association with the corresponding class;
a constitution data reading means for reading, from the constitution data storage device, constitution data corresponding to a class indicated by a result of the class determination, among the plurality of classes;
The data transmission means transmits the transmission data including the read constitution data to the first terminal. A diagnostic support system for supporting a diagnosis of a health condition of a subject based on a capillary image of the subject, comprising:
A server and a first terminal connectable to the server through a network,
The first terminal is
an imaging means for imaging the capillaries of the subject to obtain the capillary image;
an image transmitting means for transmitting the capillary blood vessel image to the server,
The server,
an image receiving means for receiving the capillary blood vessel image transmitted by the first terminal;
a feature extraction means for extracting a feature of the capillaries from the capillary image received by the image receiving means;
a class determination means for determining to which of a plurality of preset classes the state of the capillaries represented by the received capillary image corresponds based on the extracted feature amount; and
a data transmission means for transmitting data including the result of the class determination to the first terminal as transmission data;
The first terminal is
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.
The server,
a constitution improvement data storage device that stores constitution improvement data, which corresponds to each of the plurality of classes and expresses advice for improving the constitution of the subject, in association with the corresponding class;
a constitution improvement data reading means for reading, from said constitution improvement data storage device, constitution improvement data corresponding to the class indicated by the result of said class determination, among said plurality of classes;
The data transmission means transmits the transmission data including the read constitution improvement data to the first terminal. The server,
A feature storage device;
16. The diagnosis support system according to claim 12, further comprising: feature amount storage means for storing the feature amount extracted by the feature amount extraction means in the feature amount storage device. A diagnostic support system for supporting a diagnosis of a health condition of a subject based on a capillary image of the subject, comprising:
A server and a first terminal connectable to the server through a network,
The first terminal is
an imaging means for imaging the capillaries of the subject to obtain the capillary image;
an image transmitting means for transmitting the capillary blood vessel image to the server,
The server,
an image receiving means for receiving the capillary blood vessel image transmitted by the first terminal;
a feature extraction means for extracting a feature of the capillaries from the capillary image received by the image receiving means;
a class determination means for determining to which of a plurality of preset classes the state of the capillaries represented by the received capillary image corresponds based on the extracted feature amount; and
a data transmission means for transmitting data including the result of the class determination to the first terminal as transmission data;
The first terminal is
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.
The server,
A subject feature storage device;
a subject feature storage means for storing the extracted feature in the subject feature storage device in association with the subject;
and a past feature reading means for reading, from the subject feature storage device, a past feature that is a past feature associated with the subject,
The data transmission means of the server transmits the transmission data to the first terminal, the transmission data including the extracted feature amount and the read-out past feature amount of the subject. A diagnostic support server that can be connected to a network and supports diagnosis of a subject's health condition based on a capillary image of the subject, the diagnostic support server being the server provided in the diagnostic support system according to any one of claims 1 to 17. A diagnostic assistance program that assists in the diagnosis of a subject's health condition based on a capillary image of the subject, the diagnostic assistance program being a program that causes a computer to function as the diagnostic assistance server described in claim 18. 12. A diagnostic support server maintenance method for maintaining a diagnostic support server that is connectable to a network and supports diagnosis of a health condition of a subject based on a capillary blood vessel image of the subject, the diagnostic support server being a server included in the diagnostic support system according to claim 11, comprising:
reading out the stored group of features from the feature answer storage device;
redefining the plurality of classes regarding the states of capillaries based on the group of the read feature amounts; and
updating the stored reference data by storing the range or representative value of the feature amount corresponding to each of the reset classes in the reference data storage device as the reference data;
reading out a group of responses to the questionnaire stored in association with the group of features from the feature response storage device;
calculating the coefficient data based on the group of responses to the questionnaire that have been read;
and updating the stored coefficient data by storing the calculated coefficient data in the coefficient data storage device. 17. A diagnostic support server maintenance method for maintaining a diagnostic support server that is connectable to a network and supports diagnosis of a health condition of a subject based on a capillary image of the subject, the diagnostic support server being a server included in the diagnostic support system according to claim 16, comprising the steps of:
reading out the group of stored features from the feature storage device;
and resetting the plurality of classes regarding the state of capillaries based on the group of the read feature amounts. 18. A diagnostic support server maintenance method for maintaining a diagnostic support server that is connectable to a network and supports diagnosis of a health condition of a subject based on a capillary image of the subject, the diagnostic support server being a server included in the diagnostic support system according to claim 17, comprising:
reading out the group of stored features from the subject feature storage device;
and resetting the plurality of classes regarding the state of capillaries based on the group of the read feature amounts.