JP2016198197A - Diagnosis support system and application program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a diagnosis support technique capable of highly reliably providing a disease name corresponding to input lesion information.SOLUTION: A diagnosis support system includes: a first terminal 10 for transmitting a lesion image obtained by imaging a lesion; a server 14 for receiving the lesion image from the first terminal 10 and transmitting information related to a disease name corresponding to the lesion image to the first terminal 10; and second terminals 16 each of which transmits information related to a disease name indicating a diagnosis result by a doctor to the server. The first terminal 10 transmits the necessity of the diagnosis of a doctor to the server 14 together with the lesion image. The server 14 includes: an image analysis section for deriving a lesion name corresponding to the lesion image by image analysis; a diagnosis request processing section for transmitting the lesion image to the second terminal 16 when the diagnosis of a doctor is requested from the first terminal 10; and a disease information generation section for receiving the information related to the disease name derived by the image analysis section and the information related to the disease name transmitted from the second terminal 16, and transmitting the information to the first terminal 10 when the pieces of information are received.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a diagnosis support system and an application program using affected area information obtained by imaging an affected area.

  The proportion of doctors specializing in dermatology is small, and doctors who do not specialize in dermatology may diagnose skin diseases. In addition, when making a diagnosis in a disaster area or in a medical field such as a depopulated area, a doctor may make a diagnosis outside the field of specialization. Thus, when a doctor makes a diagnosis outside the field of expertise, it is desired to provide a diagnosis support technique for increasing the reliability of the doctor's diagnosis.

  In Patent Document 1, an input diagnostic image is analyzed to extract an image of a region of interest, and the image of the region of interest is converted into an image having the same dynamic range as a case image registered in a database. A similar image search device is disclosed that performs a similar image search based on a comparison between a feature amount calculated from the image and a search feature amount registered in a database.

JP 2013-200590 A

  In the technique disclosed in Patent Document 1, an image similar to the input image of the affected part is output as a search result. However, providing a similar image may be insufficient as support.

  Therefore, an object of the present invention is to provide a diagnosis support technology that can provide a disease name corresponding to an input diseased part image with high reliability.

  In order to solve the above problems, a diagnosis support system according to an aspect of the present invention relates to a first terminal that transmits an affected part image obtained by imaging the affected part, a disease name that receives the affected part image from the first terminal, and corresponds to the affected part image A server that transmits information to the first terminal; and a second terminal that transmits information related to a disease name indicating a diagnosis result by a doctor to the server. The first terminal transmits the necessity of diagnosis by the doctor together with the affected part image to the server, and the server transmits an image analyzing part for deriving a disease name corresponding to the affected part image by image analysis, and the affected part image to the second terminal. A diagnosis request processing unit; and a disease information generation unit that transmits information related to a disease name derived by the image analysis unit and information related to a disease name transmitted from the second terminal to the first terminal.

  ADVANTAGE OF THE INVENTION According to this invention, the diagnostic assistance technique which can provide the disease name according to the input affected part image with high reliability can be provided.

It is a figure for demonstrating the structure of the diagnosis assistance system concerning an Example. It is a figure for demonstrating the function structure of a 1st terminal. It is a figure which shows the example of a screen of the application for terminals displayed on a 1st terminal. It is a figure which shows the example of a screen of the disease information which a 1st terminal displays. It is a figure for demonstrating the function structure of a server. It is a figure for demonstrating the structure of the memory | storage part of a server. It is a figure for demonstrating the function structure of a 2nd terminal. It is a figure for demonstrating operation | movement of a diagnostic assistance system.

  FIG. 1 is a diagram for explaining a configuration of a diagnosis support system 1 according to the embodiment. The diagnosis support system 1 provides support for improving the quality of diagnosis when the user 11 as a doctor diagnoses the affected part 18 of the patient 19.

  When the user 11 transmits an affected area image obtained by imaging the affected area 18 of the patient 19 to the server 14 via the network 12, the support for providing the disease name by image analysis and the support for providing the disease name by the doctor diagnosis are received. It is done.

  The server 14 has a function of deriving a disease name by analyzing an affected part image and a function of supporting diagnosis by the specialist 17. When receiving a request for diagnosis from the server 14, the specialist 17 returns a disease name corresponding to the affected part image from the second terminal 16 to the server 14.

  The user 11 of the first terminal 10 can use the support function 20 for receiving the disease name derived by the server 14 free of charge, and can use the support function 22 for receiving the disease name diagnosed by the specialist 17 for a fee. The diagnosis support system 1 can provide quick diagnosis support to the user 11 with the free support function 20, and can provide highly reliable diagnosis support from the specialist 17 with the paid support function 22. The patient 19 can receive high-quality medical care by receiving a highly reliable diagnosis. In addition, by providing the disease name and its prescription from the affected area image, it is possible to support diagnosis even in a region where there is no doctor or in a disaster region. Moreover, the server 14 can collect the affected part image and can improve the accuracy of deriving the disease name.

  As a user of the second terminal 16, a doctor who has qualified as a doctor but has lost his job as a full-time housewife, a doctor who is on maternity leave, etc. can make a diagnosis using his spare time. In this way, it is possible to effectively utilize human resources who have doctor qualifications but are unemployed as doctors.

  In the diagnosis support system 1, the first terminal 10 and the second terminal 16 are connected to the server 14 wirelessly or by wire via a network 12 such as the Internet. The first terminal 10 is a mobile terminal, for example, and has a camera function for imaging the affected part 18 of the patient 19. An application program for using the diagnosis support system 1 is installed in the first terminal 10. In the modification, the first terminal 10 may be a personal computer, and the affected part image obtained by imaging the affected part 18 from a separate camera may be received by the personal computer and transmitted to the server 14. The second terminal 16 is a portable terminal or a personal computer, receives affected area information from the server 14, and transmits the disease name to the server 14 as a diagnosis result.

  FIG. 2 is a diagram for explaining a functional configuration of the first terminal 10. An application program is installed in the first terminal 10, and information regarding a disease name corresponding to the affected part image can be received from the server 14 by transmitting the affected part image obtained by imaging the affected part 18 to the server 14. The first terminal 10 includes a user authentication unit 32, a support request unit 31, a disease information reception unit 33, a transmission / reception unit 42, and an advertisement display unit 46. The support request unit 31 includes a patient information input unit 34, an affected part image input unit 36, an affected part state input unit 38, and a doctor diagnosis selection unit 40. The disease information reception unit 33 includes a disease name display unit 44 and an evaluation input. Part 48. In terms of hardware components, these configurations are realized by a CPU (Central Processing Unit) or GPU (Graphics Processing Unit) of any computer, a memory, a program loaded in the memory, etc. Depicts functional blocks realized by. Accordingly, those skilled in the art will understand that these functional blocks can be realized in various forms by hardware only, software only, or a combination thereof.

  The user authentication unit 32 authenticates whether the user name and password input to the user 11 are the same as the user account data stored in the server 14. If the user does not have a user account, the user authentication unit 32 displays a screen requesting creation of a user account. When creating a user account, the user authentication unit 32 inputs an email address, a user's account name, a password, a user's work place, a user's name, a user's name, the presence or absence of a doctor's qualification, and a user's medical department, Create a user account. The input user information is stored in the server 14 as user information in association with the user account ID.

  The patient information input unit 34 notifies the user 11 to input the patient name, patient gender, patient age, patient residence, and patient history as patient information. The affected part image input unit 36 inputs an affected part image obtained by imaging the affected part 18. The affected part image input unit 36 displays a screen for inputting a plurality of types of affected part images such as an affected part image captured so that the entire affected part 18 can be seen and an affected part image obtained by enlarging the affected part 18 in the vicinity of the affected part 18. To do.

  The affected part state input unit 38 allows the user 11 to input the state of the affected part 18. As information indicating the symptoms of the affected area 18, the site of the affected area 18, the onset period, the state of swelling, the presence or absence of persistence of the symptom, the presence or absence of disappearance when the affected area 18 is pressed, the presence or absence of a hot sensation of the affected area 18 Presence or absence of tenderness, presence or absence of itch, presence or absence of spread around the affected area 18, cause of the affected area 18 (trauma, insect bites, etc.), presence or absence of medication, presence or absence of increased C-reactive protein, increased white blood cell count At least one of presence / absence and presence / absence of fever of the patient is displayed as an input screen.

  The doctor diagnosis selection unit 40 displays a screen that allows the user 11 to select whether to perform diagnosis by the specialist 17. When the user 11 requests diagnosis by the specialist 17, the doctor diagnosis selection unit 40 displays a screen indicating that there is a charge. The patient information, the affected area information, and the presence / absence of the doctor diagnosis input by the user 11 are transmitted to the server 14 by the transmitting / receiving unit 42.

  The disease information receiving unit 33 receives information about the disease name corresponding to the diseased part information of the affected part 18 from the transmission / reception part 42 and displays it to the user 11 as diagnosis support from the server 14. The disease name display unit 44 displays the disease name received from the server 14 on the screen together with a typical disease image corresponding to the disease name. The disease name display unit 44 displays a prescription corresponding to the disease name. The evaluation input unit 48 may allow the user 11 to evaluate whether the derived disease name is correct or whether the diagnosis result of the specialist 17 is correct. For example, the evaluation input unit 48 may request the user 11 to evaluate by mail. The evaluation input unit 48 transmits the evaluation result of the user 11 to the server 14.

  The advertisement display unit 46 displays at least an advertisement corresponding to the medical department of the user 11 and an advertisement of a medicine corresponding to the disease name.

  FIG. 3 is a diagram illustrating a screen example of the terminal application 30 displayed on the first terminal 10. As shown in FIG. 3A, the user 11 has a user account for identifying himself / herself, and signs in to the diagnosis support service provided by the application program using the user account. The user authentication unit 32 authenticates whether the user name and password are the same as the user account data stored in the server 14.

  As shown in FIG. 3B, when the sign-in is successful, the first terminal 10 shifts to a screen for starting diagnosis. As shown in FIG. 3C, on the patient information input screen, the user 11 inputs the patient information and touches “Next”.

  As shown in FIG. 3D, the affected area image input unit 36 inputs affected area information obtained by imaging the affected area 18. The user 11 inputs diseased part information stored in the first terminal 10 and touches “Next”.

  As shown in FIG. 3 (e), the affected area state input unit 38 causes the user 11 to input the disease state of the affected area 18. The user 11 operates the touch pad of the first terminal 10 with a finger and inputs answers to a plurality of questions. A diagnosis support request may not be completed unless an inquiry response is input.

  As shown in FIG. 3 (f), the doctor diagnosis selection unit 40 allows the user 11 to make a diagnosis by the doctor. When the user 11 desires a doctor's diagnosis and selects “Yes”, as shown in FIG. 3G, the user terminal 11 shifts to a charging screen, and the first terminal 10 notifies the user 11 that there is a charge.

  As shown in FIG. 3H, when the user 11 touches “input data transmission”, the transmission / reception unit 42 inputs the affected part information and the affected part information including the disease state input by the user 11 and the doctor diagnosis selection unit 40. The necessity of diagnosis of the selected doctor is transmitted to the server 14. Thereby, the support request process of the user 11 is completed at the first terminal 10, and the server 14 executes the derivation of the disease name based on the affected part information and the diagnosis request to the specialist 17. As shown in FIG. 3 (h), the advertisement display unit 46 displays an advertisement in the advertisement space, for example, an advertisement of a medicine or a medical device corresponding to the medical department of the user 11. Thereby, an effective advertisement according to the medical care of the user 11 can be displayed.

  After the user 11 finishes the support request, information about the disease name by the image analysis processing is transmitted from the server 14 to the first terminal 10 several minutes to five minutes later. The information related to the disease name transmitted from the server 14 includes a typical disease image corresponding to the disease name, and a prescription and an advertisement of a medicine corresponding to the prescription.

  FIG. 4 is a diagram illustrating a screen example of support information received by the first terminal 10. As shown in FIG. 4A, the disease name display unit 44 of the first terminal 10 receives information on the disease name from the server 14 from the transmission / reception unit 42, and displays it on the screen together with typical disease images according to the disease name. Let As the disease names derived by the server 14, a plurality of disease names can be displayed in descending order of similarity of the transmitted diseased part information.

  When the user 11 touches the disease name, the screen shifts, and as shown in FIG. 4B, the disease name display unit 44 displays a prescription corresponding to the disease name. The advertisement display unit 46 displays an advertisement of a medicine corresponding to the disease name on the prescription display screen shown in FIG. Thereby, an effective advertisement can be displayed according to the disease name. Thus, the advertisement display unit 46 can display an effective advertisement according to different screen displays by displaying not only an advertisement according to the user's department but also an advertisement according to the disease name.

  FIG. 5 is a diagram for explaining a functional configuration of the server 14. FIG. 6 is a diagram for explaining the configuration of the storage unit 73 of the server 14. The server 14 includes a transmission / reception unit 50, a user management unit 52, an image analysis unit 54, a diagnosis request processing unit 56, and a disease information generation unit 58, and extracts information from the storage unit 73 or writes information in the storage unit 73. The storage unit 73 holds user information 74, advertisement information 76, acquired diseased part information 78, and accumulated disease information 80. In terms of hardware components, these configurations are realized by a CPU or GPU of an arbitrary computer, a memory, a program loaded in the memory, or the like.

  The storage unit 73 retains information input at the time of account creation as user information 74 in association with a user account ID, and stores information on medicines and medical devices classified as clinical departments as advertisement information 76 and medicines associated with disease names. The information of is retained. In addition, the storage unit 73 temporarily stores the diseased part image 82 and the symptom information 84 transmitted from the first terminal 10 as the acquired diseased part information 78, and a plurality of accumulated disease images 88 and a plurality of accumulated symptoms as accumulated disease information 80. Information 90 is stored in association with disease name / prescription information 86. A plurality of accumulated disease images 88 are associated with one disease name / prescription information 86, and a combination thereof is used for analysis in the image analysis unit 54.

  The user management unit 52 manages users of the first terminal 10 and the second terminal 16. The user management unit 52 includes the user name and password transmitted from the first terminal 10 or the second terminal 16 and the user information stored in the storage unit 73 when the first terminal 10 or the second terminal 16 is signed in. 74 is collated, and the collation result is transmitted to the first terminal 10 or the second terminal 16. The advertisement generation unit 60 of the user management unit 52 extracts an advertisement corresponding to the user's medical department from the advertisement information 76, generates an advertisement, and transmits the advertisement to the first terminal 10 or the second terminal 16.

  The image analysis unit 54 derives a disease name based on the acquired diseased part information. The image analysis unit 54 includes a similarity calculation unit 62, a disease name selection unit 64, and a learning unit 66.

  The similarity calculation unit 62 analyzes the affected part image 82 and the symptom information 84 transmitted from the first terminal 10 and derives a disease name corresponding to the affected part image 82 and its similarity degree. The similarity calculating unit 62 learns the accumulated disease information 80 as a teacher in deep learning of machine learning in the learning unit 66, and learns in advance a method for deriving a disease name from the diseased part image 82 based on the accumulated disease information 80. ing. The disease name of the main skin disease is 100 or less, and the similarity calculation unit 62 analyzes the affected part image 82 by using a convolutional neural network technique, and the affected part image 82 may correspond to any of a plurality of disease names. Identify if is high.

  The similarity calculation unit 62 repeats the convolution process and the sampling process a plurality of times, and hierarchically convolves the local features of the input affected part image 82, thereby identifying which disease name the affected part image 82 corresponds to. The similarity calculation unit 62 calculates the similarity of each disease name with respect to the affected part image 82. The similarity degree of the disease name may be a probability that the affected part image 82 corresponds to the disease name. By using the convolutional neural network technique, analysis can be performed without setting features on all the affected part images 82, and local features can be captured from the affected part images 82 to derive highly reliable disease names.

  The similarity calculation unit 62 generates a plurality of first maps that have been subjected to convolution processing from the affected part image 82. The convolution process is executed using at least one of R value, G value, B value, and luminance included in the pixel value. The similarity calculation unit 62 generates a first map by weighting and combining the pixel values included in the weighted region (n × n pixels) of the affected part image 82 with adjacent pixel values. In the convolution region, features are extracted using the weight region as a filter. Note that the weighting parameter in the convolution process is updated by the deep learning process in the learning unit 66.

  Next, the similarity calculation unit 62 performs a sampling process for generating a second map from the first map. An average sampling process using the 2 × 2 space area of the first map calculated by the convolution process as a sampling area or a maximum value sampling process using the 2 × 2 space area of the first map as a sampling area is executed. The second map is a reduced version of the first map. By the sampling process, the error of the scale and position of the diseased part of the affected part image 82 can be reduced. The second map is obtained by extracting features of image data by convolution processing and sampling processing, and a plurality of second maps may be generated from one input image data.

  The similarity calculation unit 62 further performs a convolution process and a sampling process on the second map, and calculates an output node by repeating the convolution process and the sampling process a plurality of times. The calculated output node indicates the degree of similarity of the disease name, and the disease name having the highest degree of similarity corresponds to the affected part image 82. In the embodiment, the similarity calculation unit 62 uses the neural network technique that repeats the convolution process and the sampling process. However, the present invention is not limited to this mode, and a multilayer neural network technique may be used. .

  The similarity calculation unit 62 derives the similarity degree of the disease name using the symptom information 84 in addition to the affected part image 82. For example, after calculating the output node, the similarity calculation unit 62 corrects the similarity of the disease name with the symptom information 84. By using the symptom information, the derivation accuracy can be improved as compared with the case where the analysis is performed using only the image. In addition, after deriving the degree of similarity of disease names using a convolutional neural network technique that does not require the assumption of image feature amounts, the disease names can be derived with high accuracy by correcting the symptoms using symptom information.

  The disease name selection unit 64 receives the derivation result from the similarity calculation unit 62, derives the disease name having the highest degree of similarity and the disease name having a similarity degree equal to or greater than a predetermined value, and uses information related to the disease name as the derivation result for the disease The information is sent to the information generator 58. The information related to the disease name sent from the image analysis unit 54 to the disease information generation unit 58 may be a code corresponding to the disease name.

  When a plurality of disease names having a degree of similarity equal to or greater than a predetermined value are derived, the disease name selection unit 64 selects a predetermined number in descending order of the degree of similarity and sends information related to the selected disease name to the disease information generation unit 58. The disease name selection unit 64 can send a disease name of a type equal to or less than the maximum number to the disease information generation unit 58 with a predetermined maximum number as an upper limit. The similarity calculation unit 62 and the disease name selection unit 64 send the derivation result for the acquired diseased part information 78 to the learning unit 66.

  The learning unit 66 receives the acquired diseased part information 78, the derivation results of the similarity calculation unit 62 and the disease name selection unit 64, and the diagnosis result of the specialist 17 transmitted from the second terminal 16, and the diagnosis result of the specialist 17 Learning is performed by the deep learning process so that the correct disease name is derived from the diseased part image 82 input in the similarity calculation unit 62 with a certain disease name as a correct answer. The learning unit 66 updates the weighting parameters and the like of the convolution processing of each layer in the similarity calculation unit 62 so that the derivation result when the affected part image 82 is input to the similarity calculation unit 62 is correct.

  The learning unit 66 receives the derivation result of the similarity calculation unit 62 and the disease name selection unit 64 and the evaluation result transmitted from the first terminal 10, and learns the evaluation result transmitted from the first terminal 10 as a correct answer. You can do it. Further, the learning unit 66 has a function of acquiring a disease name for the affected part image 82 from the specialist 17 even when the user 11 uses the free support function 20. You may learn the disease name as a correct answer. Thus, by providing the learning unit 66, the derivation accuracy of the similarity calculation unit 62 can be increased as it is used.

  When the diagnosis of the specialist 17 is requested from the first terminal 10, the diagnosis request processing unit 56 requests the diagnosis to the second terminal 16, and if a reply that can be diagnosed is returned, the affected part image 82 is transmitted via the transmission / reception unit 50. To the second terminal 16. The diagnosis request processing unit 56 includes an execution determining unit 68, a billing processing unit 70, a doctor selecting unit 72, and a request information generating unit 69. The execution determination unit 68 determines the execution of the doctor support when the diagnosis of the specialist 17 is requested from the first terminal 10, and does not determine the execution of the doctor support unless the diagnosis of the specialist 17 is requested.

  The billing processing unit 70 performs billing processing of the diagnostic usage fee for the first terminal 10 every time the diagnosis of the specialist 17 is requested. Further, the billing processing unit 70 performs a process of giving a reward to the specialist 17 in accordance with the diagnosis of the specialist 17. When the billing process is completed, the process proceeds to selection of the specialist 17 by the doctor selection unit 72. The doctor selection unit 72 selects any one of the specialists 17 registered in the user information 74.

  When the doctor selection unit 72 requests diagnosis of the specialist 17 from the first terminal 10, the doctor selection unit 72 selects the second terminal 16 of any specialist 17 who has shown that diagnosis is possible among the plurality of specialists 17 and performs diagnosis. Ask.

  The doctor selection unit 72 notifies the plurality of second terminals 16 of whether or not diagnosis is possible in the order of the connection timing of the second terminals 16 to the server 14, and there is a response indicating that the diagnosis is possible. 16 is determined to request diagnosis, and the acquired diseased part information 78 is transmitted to the determined second terminal 16. If there is no response from the second terminal 16 that has first notified whether diagnosis is possible or not for a predetermined time, the second terminal 16 with the next connection time is notified of diagnosis availability. By requesting doctor diagnosis in order from the second terminal 16 having the latest access time to the server 14, the possibility of obtaining a quick reply can be increased. In addition, by limiting the response period of the second terminal 16 after the request, it is possible to shorten the time to determine which specialist 17 to decide.

  In the selection method of the specialist 17 according to the modified example, the doctor selecting unit 72 notifies the plurality of second terminals 16 that there is a request for diagnosis, and the second terminal 16 that responds the earliest that the diagnosis is possible is diagnosed. The request is determined, and the acquired diseased part information 78 is transmitted to the determined second terminal 16. The doctor selecting unit 72 may transmit the acquired diseased part information 78 together with the request when notifying the second terminal 16 of the request for diagnosis.

  The request information generation unit 69 generates request information and a request ID to be transmitted to the second terminal 16 from the acquired diseased part information 78 transmitted from the first terminal 10. The request information is obtained by removing the personal information of the user 11 of the first terminal 10 and the personal information of the patient from the acquired affected part information 78, and includes an affected part image 82 and symptom information 84. The request ID includes information on the first terminal 10, information on the second terminal 16, and a request identification number. The request information generation unit 69 transmits request information to the second terminal 16 and transmits a request ID to the disease information generation unit 58. With the request ID, the disease information generation unit 58 can transmit the diagnosis result transmitted from the second terminal 16 to the first terminal 10 corresponding to the request ID.

  When the disease information generation unit 58 receives information on the disease name as a diagnosis result from the second terminal 16 or receives information on the disease name as a derivation result from the image analysis unit 54, the disease information corresponding to the disease name Is generated. The disease information generation unit 58 generates a disease name, a representative disease image corresponding to the disease name, a prescription corresponding to the disease name, and an advertisement corresponding to the disease name as the disease information. The disease information generation unit 58 transmits the disease information for the acquired diseased part information 78 to the first terminal 10 via the transmission / reception unit 50. Thus, the server 14 can send back information related to the disease name corresponding to the disease information transmitted from the first terminal 10 to the first terminal 10.

  The disease information generation unit 58 may derive a medical disease name that can be the cause from the skin disease name, generate the medical disease name as the disease information, and generate an advertisement of the medicine according to the medical disease name as the disease information. It's okay. The storage unit 73 holds information on internal medicine disease names associated with skin disease names and information on advertisements associated with internal disease names. That is, the memory | storage part 73 hold | maintains the combination with another disease name which has a high correlation with a disease name, or the correlation of a disease name.

  In addition, the disease information generation unit 58 may display an advertisement according to a disease name closely related to the user 11 that is inferred from a set of past histories of the patient group that the user 11 examines. For example, the disease information generation unit 58 estimates the disease names that are common to the patients diagnosed by the user 11 based on the patient history information transmitted from the first terminal 10 by the user 11 and corresponds to the estimated disease names. An advertisement may be generated and transmitted to the first terminal 10 as disease information. In addition, the disease information generation unit 58 accumulates the disease name transmitted to the user 11, generates an advertisement corresponding to the disease name having a relatively large number of transmissions based on the accumulated transmission information, and stores the disease information in the first terminal 10. You may send as Thereby, an effective advertisement with high relevance to the medical care of the user 11 can be displayed. Moreover, although the aspect which produces | generates these advertisements with the server 14 was shown, you may produce | generate with the application program of the 1st terminal 10. FIG.

  FIG. 7 is a diagram for explaining a functional configuration of the second terminal 16. An application program for the diagnosis support system 1 is installed in the second terminal 16. The second terminal 16 includes a transmission / reception unit 92, a user authentication unit 93, an advertisement display unit 94, a diagnosis acceptance unit 95, a disease name input unit 96, and a reward unit 97. In terms of hardware components, these configurations are realized by a CPU or GPU of an arbitrary computer, a memory, a program loaded in the memory, or the like.

  The user authentication unit 93 authenticates whether the user name and password are the same as the user account data stored in the server 14. The advertisement display unit 94 displays an advertisement corresponding to the user account. If the dermatologist is 17, the advertisement display unit 94 displays an advertisement for a dermatological drug. As a modification, the user authentication unit 93 may use biometric authentication such as fingerprint authentication or palm vein authentication at the time of login to confirm that the user is a doctor, and personal information necessary for biometric authentication is stored in the server 14. It's okay. As described above, in the user authentication of the second terminal 16, the security is higher than that of the user authentication of the first terminal 10, and authentication that requires identity verification is executed.

  The diagnosis acceptance unit 95 receives a notification of whether diagnosis is possible from the server 14, and transmits the presence / absence of diagnosis acceptance or acceptance of the diagnosis to the server 14 via the transmission / reception unit 92. The disease name input unit 96 receives the acquired diseased part information 78 from the server 14, inputs a disease name as a diagnosis result by the specialist 17, and the transmission / reception unit 92 transmits information related to the disease name to the server 14. There is one disease name as the diagnosis result transmitted from the second terminal 16 to the server 14, and the feedback is fed back to the learning unit 66 of the server 14.

  The reward unit 97 provides a reward to the specialist 17 every time a disease name is sent from the disease name input unit 96. When the disease name input unit 96 is transmitted to the server 14, the remuneration unit 97 performs remuneration processing in the charging processing unit 70 of the server 14, receives remuneration grant information from the charging processing unit 70, and rewards the specialist 17. I will provide a. The server 14 may store personal information necessary for payment of remuneration, such as the user's address and account number, as user information 74 as necessary. The operation of the above configuration will be described below.

  FIG. 8 is a diagram for explaining the operation of the diagnosis support system 1. First, the user 11 logs in to the application program of the first terminal 10 (S10), and the first terminal 10 transmits the input user name and password to the server 14 (S12).

  The server 14 compares the received user name and password with the stored user information 74 (S14), and if they match, transmits information indicating that the login has been successful to the first terminal 10 (S16). The specialist 17 logs in to the application program of the second terminal 16 (S30), and the second terminal 16 transmits the input user name and password to the server 14 (S32). The server 14 compares the received user name and password with the stored user information 74 (S34), and if they match, transmits information indicating that the login is successful to the second terminal 16 (S36).

  When the login is successful, the affected area information including the affected area information and the symptom information is input to the first terminal 10 by the user 11 (S18), and a request for a doctor diagnosis is input (S18). The first terminal 10 transmits the input diseased part information and information indicating a request for doctor diagnosis to the server 14 (S20).

  The diagnosis request processing unit 56 of the server 14 receives information indicating a request for doctor diagnosis from the first terminal 10, starts doctor support processing, and selects the specialist 17 (S22). The diagnosis request processing unit 56 transmits information for requesting a doctor diagnosis to the selected second terminal 16 (S24). The diagnosis request processing unit 56 may transmit a doctor diagnosis request using a function of another application program, for example, a mail function.

  When the image analysis unit 54 of the server 14 receives the affected part information, the image analysis unit 54 executes image analysis processing (S26). Note that the image analysis process may be executed prior to the doctor support process, or may be executed simultaneously in parallel. The image analysis unit 54 derives a disease name corresponding to the diseased part information, generates disease information corresponding to the disease name in the disease information generation unit 58 (S27), and stores the disease information as an analysis result in the first terminal 10. Send.

  The first terminal 10 displays the analysis result received from the server 14 (S29). Thus, the user of the first terminal 10 can acquire the disease name corresponding to the affected area information of the affected area of the patient in a short time. In addition, receiving a disease name can provide effective support.

  The second terminal 16 receives an input of diagnosis acceptance from the specialist 17 (S38), and transmits information indicating the diagnosis acceptance to the server 14 (S40). Thereby, the doctor selection part 72 of the server 14 determines the specialist 17 (S42), and transmits the acquired affected part information 78 to the second terminal 16 (S43).

  The second terminal 16 receives an input of a disease name corresponding to the acquired diseased part information 78 as a diagnosis result by the specialist 17 (S44), and transmits information related to the disease name to the server 14 (S46). When the disease information generation unit 58 of the server 14 receives information on the disease name from the second terminal 16, the disease information generation unit 58 generates disease information corresponding to the disease name and transmits the disease information as a diagnosis result via the transmission / reception unit 50 ( S50).

  The first terminal 10 displays the received diagnosis result of the specialist 17 (S51). Based on the diagnosis result of the specialist 17, a highly reliable disease name can be provided to the user 11. The learning unit 66 of the server 14 learns that the disease name of the diagnosis result of the specialist 17 is correct by the machine learning process, and reflects it in the analysis result of the image analysis unit 54 according to the acquired diseased part information 78 (S52). . Thereby, the derivation | leading-out precision of the disease name in the server 14 can be improved.

  The present invention is not limited to the above-described embodiment, and various modifications such as design changes can be added based on the knowledge of those skilled in the art. The configuration shown in each figure is for explaining an example, and any configuration that can achieve the same function can be changed as appropriate, and the same effect can be obtained.

  In the embodiment, the doctor 11 who is not a dermatologist is the user 11 and the dermatologist 17 uses the doctor. However, the present invention is not limited to this. For example, it may be surgery or internal medicine. In addition, the user 11 may receive support regarding the state of bones or lesions by transmitting the captured image from the first terminal 10 to the server 14 by X-ray imaging.

  In the embodiment, the application program of the first terminal 10 and the application program of the second terminal 16 are described as different ones, but the present invention is not limited to this mode. For example, the application program of the second terminal 16 may be common with the application program of the first terminal 10. The shared application program is provided with a function for selecting whether it is a support request side or a diagnosis support side.

  The first terminal 10 and the second terminal 16 may be able to use the diagnosis support service with a web browser instead of installing the application program.

  In the embodiment, when the first terminal 10 transmits the affected part image, the aspect in which the diagnosis request by the specialist 17 is transmitted to the server 14 as paid support is shown, but the present invention is not limited to this aspect. For example, every time an affected area image is transmitted from the first terminal 10, it is not necessary to transmit information requesting diagnosis by the specialist 17, and the user 11 performs a monthly charge or a limited number of charges in advance using the application program of the first terminal 10. In this case, if the first terminal 10 transmits the affected part image, it is possible to receive support by image analysis and diagnosis by the specialist 17 without requesting diagnosis by the specialist 17.

  Moreover, in another modification, it may be an aspect in which support by the diagnosis of the specialist 17 can be received without fail. For example, it may be a diagnosis support system in which the charge varies depending on the response period of the support by the diagnosis of the specialist 17. There is a charge for receiving an answer based on the diagnosis of the specialist 17 within 30 minutes, but no charge is required if an answer based on the diagnosis of the specialist 17 is received within one day. In this case, when the first terminal 10 transmits the affected part image, the result of selecting paid support and free support is transmitted instead of the presence or absence of a diagnosis request by the specialist 17.

  DESCRIPTION OF SYMBOLS 1 Diagnosis support system, 10 1st terminal, 11 User, 12 Network, 14 Server, 16 2nd terminal, 17 Specialist, 18 Affected part, 19 Patient, 31 Support request part, 32 User authentication part, 33 Disease information receiving part, 34 Patient information input section, 36 affected area image input section, 38 affected area status input section, 40 doctor diagnosis selection section, 42 transmission / reception section, 44 disease name display section, 46 advertisement display section, 48 evaluation input section, 50 transmission / reception section, 52 user Management unit, 54 image analysis unit, 56 diagnosis request processing unit, 58 disease information generation unit, 60 advertisement generation unit, 62 similarity calculation unit, 64 disease name selection unit, 66 learning unit, 68 execution determination unit, 69 request information generation , 70 charging processing unit, 72 doctor selection unit, 73 storage unit, 74 user information, 76 Advertisement information, 78 Acquired diseased part information, 80 Accumulated disease information, 82 Diseased part image, 84 Symptom information, 86 Disease name / prescription information, 88 Accumulated disease image, 90 Accumulated symptom information, 92 Transmission / reception part, 93 User authentication part, 94 Advertisement display Department, 95 diagnosis acceptance department, 96 disease name input department, 97 reward section.

Claims (6)

A first terminal that transmits an image of an affected area obtained by imaging the affected area;
A server that receives the affected area image from the first terminal and transmits information on a disease name corresponding to the affected area image to the first terminal;
A second terminal that transmits information related to a disease name indicating a diagnosis result by a doctor to the server,
The first terminal transmits a doctor's diagnosis necessity together with the affected part image to the server,
The server
An image analysis unit for deriving a disease name corresponding to the affected part image by image analysis;
A diagnosis request processing unit for transmitting the affected part image to the second terminal;
A disease information generation unit that transmits information about a disease name derived by the image analysis unit and information about a disease name transmitted from the second terminal to the first terminal. Diagnosis support system.   The diagnosis support system according to claim 1, wherein the image analysis unit includes a learning unit that performs machine learning using information related to a disease name corresponding to the diseased part image received from the second terminal. The first terminal notifies the user that there is a charge when the user requests a doctor's diagnosis,
The diagnosis support system according to claim 1, wherein the server includes a billing processing unit that performs billing processing of a diagnosis usage fee by a user of the first terminal. The first terminal is
A user authentication unit for inputting information of a user of the first terminal;
A disease information input unit for inputting the affected part image;
The diagnosis support system according to claim 1, further comprising: an advertisement display unit that displays an advertisement according to the information on the user or the information on the disease name.   When the diagnosis request processing unit is requested by the first terminal to diagnose a doctor, the diagnosis request processing unit selects any one of the second terminals indicating that diagnosis is possible from the plurality of second terminals and performs diagnosis. The diagnosis support system according to claim 1, further comprising a doctor selection unit that requests. An application program installed in a terminal capable of transmitting an affected area image obtained by imaging an affected area to a server,
Means for inputting the user's department as user information;
Means for inputting an affected part image obtained by imaging the affected part;
Means for inputting the disease state of the affected area;
Means for inputting the patient's medical history;
Means for inputting the necessity of diagnosis by a doctor;
Means for displaying information on a disease name corresponding to the affected part image received from the server;
In an advertisement according to the department of the user, an advertisement according to the disease name, an advertisement according to another disease name presumed to have caused the disease name, or a set of past histories of patient groups examined by the user Means for displaying an advertisement corresponding to a disease name that is presumed based on an application program.

Images