Proposal System

Abstract

We provide a suggestion system that can easily propose individualized treatments tailored to each patient. [Solution] The terminal device of the proposed system includes: an examination result receiving unit that receives at least numerical data obtained from an ophthalmic diagnosis and image data which is the result of an image diagnosis of the eyeball; an image data analysis unit that uses a trained model to analyze diseases contained in the image data based on the similarity between the features of the image data and the features of existing ophthalmic image samples; and an optimal treatment proposal unit that determines and proposes the optimal treatment for the patient based on the numerical data and the results of the image data analysis by the image data analysis unit.

Description

Technical Field

[0001] The present disclosure relates to a proposal system that proposes individualized treatments and prescription drug selections tailored to patients.

Background Art

[0002] Currently, especially in cancer treatment, the concept of personalized medicine is being introduced. Personalized medicine refers to performing treatments and prescriptions tailored to the constitution and disease type of each patient. For example, after examining in more detail information such as data related to the patient's constitution and disease, and data related to genes, treatments are performed according to individual patients.

[0003] Patent Document 1 related to personalized medicine below is a device that utilizes machine learning. In a medical information processing device, a first acquisition unit acquires a plurality of training samples. Each training sample includes a feature amount representing the state of a subject, a type label of an event (a medical action performed by a medical worker or an action performed by the subject himself), and an effect label of the event. Further, a second acquisition unit acquires a knowledge base from the plurality of training samples.

[0004] An assignment unit assigns a knowledge label (correct data in machine learning) to at least a part of the plurality of training samples based on this knowledge base. Then, a learning unit trains a model that infers the effect for each type of event based on the training samples to which the knowledge label has been assigned. That is, at least a part of the training samples includes a feature amount, a type label, an effect label, and a knowledge label. Since not only the training samples but also the knowledge base are taken into account to infer the effect for each event type, the accuracy of the causal inference model can be improved with a small number of training samples. This inference is very useful for patient treatment (see Patent Document 1 / paragraphs 0006, 0066, 0067).

Prior Art Documents

Patent Documents

[0005] [Patent Document 1] Japanese Patent Publication No. 2024-074287 [Overview of the project] [Problems that the invention aims to solve]

[0006] In cases of eye diseases such as glaucoma, age-related macular degeneration (AMD), and diabetic retinopathy (DR), the same medication was often prescribed to patients with the same disease and symptom level. Furthermore, while genetic testing and ophthalmic medications make personalized medicine possible in the field of ophthalmology, there has been a problem with low cost-effectiveness.

[0007] This disclosure is made in light of these circumstances and aims to provide a proposal system that facilitates individualized treatment tailored to each patient. [Means for solving the problem]

[0008] To achieve the above objectives, this disclosure is intended to A proposed system comprising a terminal device installed within a hospital, and a trained model connected to the terminal device to support the determination of disease detection, The aforementioned terminal device is A means for receiving examination results that accepts at least numerical data obtained from ophthalmic diagnosis and image data resulting from an imaging diagnosis of the eyeball, An image data analysis means that uses the aforementioned trained model to analyze diseases contained in the image data based on the similarity between the features of the image data and the features of existing ophthalmic image samples, An optimal treatment proposal means that determines and proposes the optimal treatment for a patient based on the numerical data and the results of the image data analysis by the image data analysis means, It has. [Effects of the Invention]

[0009] According to the proposed system of the present disclosure, it is possible to easily perform an optimal individual treatment for a patient.

Brief Description of the Drawings

[0010] [Figure 1] It is an overall view of the proposed system according to an embodiment of the present disclosure. [Figure 2] It is a diagram for explaining the details of each component of the proposed system. [Figure 3] It is a flowchart of the optimal treatment proposal process by the terminal device. [Figure 4] It is a diagram for explaining the analysis method of image data. [Figure 5] It is a diagram for explaining the display content of the terminal display unit (electronic medical record). [Figure 6] It is a diagram for explaining an analysis example of a fundus image and an example of proposal information. [Figure 7] It is a diagram for explaining an analysis example of a fundus image and an example of proposal information (other aspect). [Figure 8] It is a diagram for explaining an analysis example of an OCT image and a proposal. [Figure 9] It is an overall view of the proposed system according to a modified embodiment of the present disclosure. [Figure 10] It is a diagram for explaining the details of each component of the proposed system (modified embodiment).

Modes for Carrying Out the Invention

[0011] Hereinafter, an example of the present disclosure will be described with reference to the drawings. The scope of the present disclosure is not limited to the embodiment described here, and various modifications are possible without departing from the spirit. Also, when a plurality of upper and lower limit values are described for a specific parameter, the numerical range can be set to a suitable range by combining any upper limit value and lower limit value among these upper and lower limit values.

[0012] FIG. 1 is an overall view of the proposed system 1 according to an embodiment of the present disclosure.

[0013] The proposed system 1 mainly consists of a terminal device 10 installed in examination rooms and the like within a hospital, a learned model 20, an ophthalmic device 30, and a server device 40 mainly installed outside the hospital. Since the ophthalmic device 30 is network-connected to the terminal device 10, various data obtained in ophthalmic diagnosis can be confirmed by doctor D on the terminal device 10. The ophthalmic device 30 mainly includes a fundus camera and an OCT (Optical Coherence Tomography) device, but other devices such as a scanning laser ophthalmoscope (SLO) may also be included.

[0014] The terminal device 10 is a PC, a notebook PC, a tablet terminal, etc. owned by a medical institution such as a hospital. For example, the case where patient P undergoes a glaucoma examination in the hospital will be described. When patient P undergoes an intraocular pressure measurement by the ophthalmic device 30, the intraocular pressure value is transmitted to the terminal device 10 as numerical data X. Also, when the ophthalmic device 30 takes a fundus image, the fundus image is transmitted to the terminal device 10 as image data Y. The intraocular pressure value and the fundus image are stored in the terminal device 10 and can be confirmed by doctor D.

[0015] Also, since the server device 40 is network-connected to the terminal device 10, doctor D can always access an ophthalmic image sample Sa (for example, a fundus image including glaucoma symptoms), which is the result of image diagnosis stored in the server device 40. Doctor D can confirm the items recorded in the electronic medical record in addition to the results of ophthalmic diagnosis obtained in other medical institutions, and thus judge the presence or absence of eye diseases and the like considering each piece of information.

[0016] Doctor D judges the glaucoma symptoms of patient P based on at least the intraocular pressure value and the fundus image. Also, doctor D can receive assistance in discovering lesions and diseases by means of the terminal device 10 and the learned model 20.

[0017] The terminal device 10 uses a trained model 20 to compare the fundus image of patient P with multiple ophthalmic image samples Sa to determine the possibility of glaucoma. Glaucoma can generally be estimated from intraocular pressure values, but there are also cases of glaucoma with normal intraocular pressure. The proposed system 1 in this disclosure, with the support of the trained model 20, can make a quick and accurate diagnosis without overlooking various symptoms of glaucoma.

[0018] Based on the above determination by the terminal device 10, for example, the presence and progression of glaucoma in patient P will be displayed on the display unit. In addition, suggestions for individualized treatment and prescribed medications tailored to patient P will be displayed.

[0019] Figure 2 is a diagram illustrating the details of each component of the proposed system 1. Below, the internal configurations of the terminal device 10 and the server device 40, and the details of the trained model 20 will be described.

[0020] The terminal device 10 includes a terminal control unit 11, a terminal display unit 12, and a terminal storage unit 13. The terminal control unit 11 is a processor (CPU, GPU, FPGA, etc.) capable of primarily analyzing image data Y.

[0021] The terminal control unit 11 consists of an examination result receiving unit 11a, an image data analysis unit 11b, and an optimal treatment proposal unit 11c. The examination result receiving unit 11a receives numerical data X and image data Y transmitted from the ophthalmic device 30.

[0022] Furthermore, the image data analysis unit 11b analyzes the image data Y while referring to the ophthalmic image sample Sa to determine the presence and progression (stage) of an eye disease. In doing so, the image data analysis unit 11b utilizes the trained model 20.

[0023] Here, the trained model 20 is a machine learning database that has been trained to determine the likelihood of eye disease by inputting the relationship between multiple ophthalmic image samples Sa and eye diseases as training data. For example, the trained model 20 takes a large number of fundus images as training data, adds results for glaucoma, age-related macular degeneration, etc., as training data, and performs machine learning. As a result, the image data analysis unit 11b can output information such as the current progression of glaucoma and the probability of future occurrence. There are no particular restrictions on the machine learning method, and various methods such as unsupervised learning and deep learning can be employed.

[0024] The trained model 20 may also be configured to be connected to the server device 40. In that case, the server device 40 will determine the possibility of lesions and diseases, and the results will be sent to the terminal device 10.

[0025] The optimal treatment proposal unit 11c, while referring to numerical data X, determines and proposes the optimal treatment for patient P based on the analysis results by the image data analysis unit 11b. The proposal includes a treatment plan for the eye disease, as well as prescribed medications and the reasons for them.

[0026] The terminal display unit 12 is a display that shows the results of the examination or image diagnosis. Doctor D can check the ophthalmological diagnosis results as well as the above-mentioned optimal treatment suggestions on the terminal display unit 12. The display can be of any type, such as liquid crystal, plasma, or organic EL, and may also be a touch panel type.

[0027] The terminal storage unit 13 is a storage medium such as a semiconductor memory, optical disk, or magnetic disk on which data can be written. The ophthalmic image sample Sa from the server device 40 may be downloaded by physician D and stored in the terminal storage unit 13, and accessed when needed.

[0028] Next, the server device 40 includes a server storage unit 41 internally. The server storage unit 41 is a storage medium such as a semiconductor memory, optical disk, or magnetic disk on which data can be written. The server storage unit 41 stores previously obtained ophthalmic diagnostic results (numerical values ​​of examinations and ophthalmic image samples Sa). The ophthalmic image samples Sa in the server storage unit 41 are separated into groups based on eye diseases such as glaucoma and age-related macular degeneration, and also into groups based on the progression of a single eye disease.

[0029] There are no particular restrictions on the type of server device 40; for example, it may be a cloud-type server. In this embodiment, the server device 40 is located outside the hospital, assuming a cloud-type server, but it may also be a conventional server installed inside the hospital.

[0030] Next, with reference to Figure 3, the flowchart for the optimal treatment suggestion process will be explained. The optimal treatment suggestion process is performed on the terminal device 10 and presupposes that an ophthalmic diagnosis has been performed on the ophthalmic device 30.

[0031] First, in step S10, numerical data X for ophthalmic diagnosis is input to the terminal device 10. Numerical data X such as intraocular pressure and visual field are received by the examination result reception unit 11a of the terminal device 10. After that, the optimal treatment suggestion process proceeds to step S20.

[0032] In step S20, ophthalmic diagnostic image data Y is input to the terminal device 10. Image data Y, such as fundus images and OCT images, is received by the examination result reception unit 11a. After that, the optimal treatment suggestion process proceeds to step S30.

[0033] In step S30, the terminal device 10 analyzes the image data Y. Specifically, the image data analysis unit 11b of the terminal device 10 uses the trained model 20 to analyze the eye diseases contained in the image data Y based on the similarity between the features of the image data Y and the features of the existing ophthalmic image sample Sa. The image data analysis unit 11b then determines the presence or absence of an eye disease and the stage of the disease that can be read from the image data Y.

[0034] Figure 4 illustrates the analysis method for image data Y in this step. Note that the following analysis is an internal process of the image data analysis unit 11b and is not displayed on the terminal display unit 12, etc.

[0035] Image data Y1 in the figure is one of the image data Y, and is a fundus image of patient P obtained through ophthalmological diagnosis. The image data analysis unit 11b extracts features from image data Y1. For example, it extracts features from image data Y1 that indicate the presence or absence of eye diseases such as glaucoma, age-related macular degeneration, and diabetic retinopathy, as well as features specific to the stage of each eye disease.

[0036] Group GrpA is a group that collected ophthalmic image samples Sa from glaucoma stage 1, and features specific to stage 1 have already been extracted. Similarly, Group GrpB is a group that collected ophthalmic image samples Sa from glaucoma stage 2, and features specific to stage 2 have already been extracted. In addition, numerous other groups are available, including groups for glaucoma stages 3 and 4, age-related macular degeneration stages 1-4, and diabetic retinopathy stages 1-4.

[0037] The image data analysis unit 11b sequentially compares the features of image data Y1 with those of each group to determine the group with the highest similarity. In the example in Figure 4, the similarity between image data Y1 and GrpA is 0.05, while the similarity between image data Y1 and GrpB is a high 0.86. Therefore, the image data analysis unit 11b determines that image data Y1 is close to the features of glaucoma stage 2 in GrpB, and that this is highly probable.

[0038] Returning to Figure 3, in step S40, the terminal device 10 determines and proposes the optimal treatment. Specifically, the optimal treatment proposal unit 11c of the terminal device 10 determines and proposes the optimal treatment for patient P based on the numerical data X and the analysis results of the image data Y1 by the image data analysis unit 11b (step S30). A specific example of the proposal displayed on the terminal display unit 12 will be described later. This concludes the optimal treatment proposal process.

[0039] Next, an example of the display on the terminal device 10 (terminal display unit 12) will be explained with reference to Figures 5 to 8.

[0040] Figure 5 shows the electronic medical record screen of the proposed system 1 as displayed on the terminal display unit 12. Area 12a is the Patient Information area, where information such as the patient's (subject's) name, age, gender, and address is displayed. Area 12b is the Visit History area, where information such as the medical institution where the examination was performed, the date and time, and the medication prescribed at that time is displayed.

[0041] Area 12c is a data panel (Data Panel 1) where the test results are displayed when a specific test date is designated. Area 12c displays information such as visual acuity, intraocular pressure, and any diseases discovered during the test on that day. It may also display the patient's surgical history.

[0042] Area 12d is also a data panel (Data Panel 2) where test results are displayed, and it displays test results other than those displayed in Area 12c. Area 12d may display information such as visual acuity and intraocular pressure values ​​from a different day than Area 12c for comparison, or it may display other test results from the same day as Area 12c.

[0043] Area 12e is the area where the electronic medical record entries (Medical Record Content) are displayed. Because area 12e is a relatively large area, it is also possible to display the results of image data (Image View).

[0044] Area 12f is where the Name of Disease is displayed. Area 12g is the Function Panel, which contains buttons for switching between various displays. It is also possible to superimpose image data and progress graphs onto areas 12a-12h for comparison.

[0045] Area 12h is the suggestion area, where the optimal treatment suggestion unit 11c displays individualized treatment and prescription drug suggestions tailored to patient P. Area 12h may also display the reasons for the suggestion and information on side effects.

[0046] Figure 6 illustrates an example of fundus image analysis and an example of proposed information.

[0047] First, area 12d of the terminal display unit 12 displays the results of patient P's ophthalmological examination for today (2024-09-01). Specifically, it displays the results of the visual acuity test, intraocular pressure test, laser flare, etc. (numerical data X), as well as the results of the medical interview and the status of pupil dilation.

[0048] Next, the area 12e of the terminal display unit 12 displays the image data Y2 of patient P's ophthalmological diagnosis. Window Wa displays today's fundus image (right eye), and window Wb displays the analysis results by the image data analysis unit 11b. This display indicates that, based on the area enclosed by curve Z on the image data Y2 (fundus image), the possibility of DR (diabetic retinopathy) stage 2 is highest (Similarity: 0.86).

[0049] Area 12h of the terminal display unit 12 displays suggested information determined by the optimal treatment suggestion unit 11c from numerical data X and image data Y2. Specifically, in area 12h, in addition to the diagnosis (Diagnosis) labeled "DR (Stage 2)," the most appropriate treatment (Treatment) and prescription (Prescription) for patient P are suggested. The reasons for selecting the treatment and prescription, as well as information on side effects, may also be displayed. Doctor D makes the final decision on the treatment and prescription for patient P, referring to the suggestions in area 12h.

[0050] The optimal treatment suggestion unit 11c preferably determines and proposes the optimal treatment for patient P by prioritizing the analysis results of the image data Y2 (fundus image of window Wa) from the image data analysis unit 11b. This is because prioritizing numerical data X would result in the same suggestion for patients with similar numerical values ​​(e.g., intraocular pressure). Furthermore, in terms of analyzing image data Y2, current image analysis technology is more rapid and accurate than human analysis, and is also beneficial because it is not influenced by the biases of the attending physician.

[0051] Figure 7 illustrates an example of fundus image analysis and an example of proposed information (other embodiments).

[0052] First, area 12d of the terminal display unit 12 displays the results of patient P's ophthalmological examination today (2024-09-01) and the results of the ophthalmological examination on the previous visit date (2024-08-01). Specifically, it displays visual acuity test results, intraocular pressure test results, laser flare results, etc. (numerical data X).

[0053] Next, area 12e of the terminal display unit 12 displays the ophthalmic diagnostic image data Y3 and Y3' of patient P. Window Wa displays today's fundus image (image data Y3 of the right eye), and window Wb displays the fundus image from the previous visit (image data Y3' of the right eye). This analysis focuses on the area of ​​abnormal structures (abnormal volume), and the area from the previous examination date (image data Y3') is 5.10 mm². 2 In contrast, the area of ​​today's image data (Y3) is 32.96 mm². 2 It is increasing. The image data analysis unit 11b has the function of easily calculating and quantifying this area, and can obtain more accurate results than a doctor P could judge by visual inspection.

[0054] Area 12h of the terminal display unit 12 displays suggested information determined by the optimal treatment suggestion unit 11c from numerical data X and image data Y3. Specifically, in area 12h, in addition to the diagnosis of "AMD (Age-Related Macular Degeneration)," the most appropriate treatment and prescription for patient P are suggested.

[0055] Next, with reference to Figure 8, we will explain examples of OCT image analysis and proposed information.

[0056] First, area 12d of the terminal display unit 12 displays the results of patient P's ophthalmological examination today (2024-09-01) and the results of the ophthalmological examination on the previous visit date (2024-08-01). Specifically, it displays visual acuity test results, intraocular pressure test results, laser flare results, etc. (numerical data X).

[0057] Next, area 12e of the terminal display unit 12 displays the ophthalmic diagnostic image data Y4 and Y4' of patient P. Window Wa displays today's OCT image (image data Y4 of the right eye), and window Wb displays the OCT image (image data Y4' of the right eye) from the previous visit. In this analysis, attention is focused on the edematous area of ​​the retina as the segmented volume of the unique structure, and the area on the previous examination day (image data Y4') was 12.55 mm². 2 In contrast, the area of ​​today's image data (Y4) is 35.10 mm². 2 It is increasing. The image data analysis unit 11b has a function that can easily calculate this area, and differences can be recognized even if the apparent size is the same.

[0058] In area 12h of the terminal display unit 12, suggested information determined by the optimal treatment suggestion unit 11c from numerical data X and image data Y4 is displayed. Specifically, in area 12h, in addition to the diagnosis of "AMD (Age-Related Macular Degeneration)," the most appropriate treatment and prescription for patient P are suggested. In this way, the terminal device 10 of this disclosure places importance on the analysis of image data Y4 and, with the support of a trained model 20, selects the most appropriate individual treatment and prescription for patient P.

[0059] As described above, the proposed system 1 in this disclosure allows the terminal device 10 to easily propose the most suitable individual treatment and prescription drug selection for patient P. In the case of eye diseases, personalized medicine can be realized by the proposed system 1 using image data Y, without the patient P having to undergo costly genetic testing, etc. This disclosure is not limited to the above-described embodiments and can be implemented in various forms without departing from its gist. For example, the proposed system 1 in this disclosure can be applied to diseases other than ophthalmology that require image diagnosis.

[0060] In the modified embodiment shown in Figure 9, the proposed system 100 is configured such that the ophthalmic device 30 and the cloud server 50 (with a pre-trained model) are connected via a network. Furthermore, the cloud server 50 has an image data analysis unit and can utilize the pre-trained model.

[0061] As shown in the diagram, numerical data X and image data Y are transmitted from the ophthalmic device 30 to the cloud server 50, where the possibility of lesions and diseases is determined. Furthermore, the terminal device 10' and the cloud server 50 are connected via a network, and the analysis results Z are transmitted to the terminal device 10'.

[0062] Doctor D will examine at least numerical data X and image data Y to determine the likelihood of eye disease in patient P (glaucoma, age-related macular degeneration, diabetic retinopathy, etc.). In addition, Doctor D can receive assistance in detecting eye disease through the image analysis function (analysis results Z) of the cloud server 50.

[0063] Figure 10 is a block diagram of each component constituting the proposed system 100 according to the modified embodiment. In the following description, the same reference numerals are used for components that are the same as those in the embodiment shown in Figure 2, and some explanations may be omitted.

[0064] If the ophthalmic device 30 is a fundus camera, its imaging unit (not shown) captures an image of the patient P's fundus. The captured fundus image is transmitted as image data Y to the cloud server 50 by the communication unit (not shown). In addition to numerical data X and image data Y, the communication unit may also transmit the patient's personal information (age, gender, etc.).

[0065] The cloud server 50 includes a server storage unit 51, a server communication unit 52, and a server image data analysis unit 53 internally. In the modified embodiment, an external server computer is assumed, but a conventional server computer installed inside or outside the hospital may also be used.

[0066] The server storage unit 51 is a storage medium such as a semiconductor memory, optical disk, or magnetic disk on which data can be written. The server storage unit 51 stores previously obtained ophthalmic diagnostic results (numerical values ​​of examinations and ophthalmic image samples Sa).

[0067] The server communication unit 52 transmits and receives data with the terminal device 10' and the ophthalmic device 30. The server communication unit 52 receives image data Y etc. transmitted from the ophthalmic device 30. The server communication unit 52 also transmits analysis results Z etc. to the terminal device 10'.

[0068] The server image data analysis unit 53 analyzes image data Y while referring to ophthalmic image sample Sa to estimate eye diseases and determine their progression (stage). In doing so, the server image data analysis unit 53 utilizes the trained model 20. The server image data analysis unit 53 is a processor (CPU, GPU, FPGA, etc.) that is mainly capable of performing AI-based image analysis of image data.

[0069] The terminal device 10' includes a terminal control unit 11, a terminal display unit 12, and a terminal storage unit 13. The terminal control unit 11 mainly consists of an inspection result receiving unit 11a and an optimal treatment suggestion unit 11c. The terminal control unit 11 is configured using a processor such as a CPU (Central Processing Unit). This processor works in cooperation with the memory (including the terminal storage unit 13) of the terminal device 10' to enable each of the processes.

[0070] Here, the test result receiving unit 11a receives the analysis result Z transmitted from the cloud server 50. The optimal treatment proposal unit 11c determines and proposes the optimal treatment for patient P based on the numerical data X and the analysis result Z. The terminal display unit 12 also displays the proposed information determined by the optimal treatment proposal unit 11c.

[0071] Thus, the terminal device 10' of the proposed system 100 in this disclosure can easily propose the most suitable individual treatment and prescription drug selection for patient P. Furthermore, patient P can receive personalized medicine through the proposed system 100 by utilizing image data Y, without having to undergo expensive genetic testing, etc.

[0072] The proposed system in this disclosure will have the following functions and effects:

[0073] (1) A proposed system comprising a terminal device installed within a hospital, and a trained model connected to the terminal device to support the determination of disease detection, The aforementioned terminal device is A means for receiving examination results that accepts at least numerical data obtained from ophthalmic diagnosis and image data resulting from an imaging diagnosis of the eyeball, An image data analysis means that uses the aforementioned trained model to analyze diseases contained in the image data based on the similarity between the features of the image data and the features of existing ophthalmic image samples, The system includes an optimal treatment proposal means that determines and proposes the optimal treatment for a patient based on the numerical data and the results of the image data analysis by the image data analysis means.

[0074] The proposed system in this disclosure involves a test result reception means that, when a patient undergoes an ophthalmological examination, receives at least numerical data and image data from the ophthalmological examination results. The numerical data is used to understand the patient's disease and symptom level.

[0075] The image data analysis means analyzes whether the patient's image data contains lesions or signs of disease by referring to the current image data and ophthalmic image samples (past diagnostic results). Furthermore, the optimal treatment suggestion means determines and proposes the optimal treatment for the patient based on the numerical data and the analysis results of the image data analysis means. In this way, the proposed system can easily propose the most appropriate individual treatment for a patient by making decisions based on the analysis results of image data in addition to the patient's numerical data.

[0076] (2) In the proposed system of this disclosure, it is preferable that the trained model is trained to use machine learning to determine the possibility of the disease by inputting the relationship between the ophthalmic image sample and the disease as training data.

[0077] The terminal device is connected to a pre-trained model, which has been trained using machine learning to determine the likelihood of disease by inputting the relationship between ophthalmic image samples and diseases as training data. The image data analysis means uses the pre-trained model to compare the patient's image data with multiple ophthalmic image samples to determine the likelihood of disease, thus enabling an unbiased and objective determination of the presence, progression, etc., of the disease.

[0078] (3) Furthermore, in the proposed system of this disclosure, it is preferable that the optimal treatment proposal means prioritizes the analysis results of the image data by the image data analysis means in determining and proposing the optimal treatment for the patient.

[0079] Some diseases can be overlooked if numerical and image data are evaluated fairly, or if numerical data is given too much weight. Therefore, the optimal treatment suggestion system prioritizes the analysis results of image data to determine and propose the optimal treatment. This allows the proposed system to prevent overlooking certain diseases.

[0080] (4) Furthermore, in the proposed system of this disclosure, it is preferable that the image data includes a fundus image, and the image data analysis means quantifies the unique structure of the fundus image to determine the possibility of disease.

[0081] If the image data includes fundus images, the image data analysis means quantifies, for example, the shape of the optic disc and other unique structures of the fundus (number, area, etc.) to determine the possibility of disease. This allows the proposed system to detect various diseases that appear in fundus images.

[0082] (5) Furthermore, in the proposed system of this disclosure, it is preferable that the image data includes OCT images, and that the image data analysis means quantifies the unique structure of the OCT images to determine the possibility of disease.

[0083] When image data includes OCT images, the image data analysis means quantifies specific structures contained in the layered structure of the fundus (such as the size of edema) to determine the possibility of disease. This allows the proposed system to detect various diseases that appear in OCT images.

[0084] (6) Another proposed system comprises a terminal device installed within a hospital, and a server (e.g., a cloud server) connected to the terminal device and having a trained model that supports disease detection decisions. The server has an image data analysis means that uses the trained model to analyze the diseases contained in the image data based on the similarity between the features of the image data and the features of existing ophthalmic image samples, using image data which is the result of an image diagnosis of the eye obtained in an ophthalmic diagnosis. The terminal device has an optimal treatment proposal means that determines and proposes the optimal treatment for the patient based on numerical data obtained from ophthalmic diagnosis and the results of the analysis of the image data by the image data analysis means.

[0085] Another proposed system in this disclosure involves sending image data of an ophthalmic diagnosis to a cloud server for analysis when a patient undergoes an ophthalmic diagnosis. The image data analysis means on the cloud server analyzes whether the patient's image data contains lesions or signs of disease by referring to the current image data and ophthalmic image samples (past diagnosis results). The analysis results are sent to a terminal device, which receives the examination result reception unit of the terminal device.

[0086] The terminal device's optimal treatment suggestion mechanism determines and proposes the optimal treatment for the patient based on numerical data and analysis results transmitted from the cloud server. In this way, the proposed system can easily propose the most suitable individual treatment for each patient by making decisions based on analysis results obtained from analyzing image data in addition to the patient's numerical data. [Explanation of Symbols]

[0087] 1,100 Proposal Systems 10,10' Terminal device 11 Terminal Control Unit 11a Test Result Reception Department 11b Image Data Analysis Department 11c Optimal Treatment Proposal Department 12 Terminal display unit 13 Terminal Storage Unit 20 Pre-trained Models 30 Ophthalmological equipment 40 Server Devices 41 Server Storage Unit 50 Cloud Servers 51 Server Storage Unit 52 Server Communication Unit 53 Server Image Data Analysis Department Doctor D P patient Sa Ophthalmic Image Samples X Numerical data Y image data Z analysis results

Claims

1. A proposed system comprising a terminal device installed within a hospital, and a trained model connected to the terminal device to support the determination of disease detection, The aforementioned terminal device is A means for receiving examination results that accepts at least numerical data obtained from ophthalmic diagnosis and image data resulting from an imaging diagnosis of the eyeball, An image data analysis means that uses the aforementioned trained model to analyze diseases contained in the image data based on the similarity between the features of the image data and the features of existing ophthalmic image samples, A proposal system comprising an optimal treatment proposal means that determines and proposes the optimal treatment for a patient based on the numerical data and the results of the analysis of the image data by the image data analysis means.

2. The aforementioned trained model is trained to use machine learning to determine the likelihood of the disease by inputting the relationship between the ophthalmic image sample and the disease as training data. The proposed system according to claim 1.

3. The aforementioned optimal treatment proposal means determines and proposes the optimal treatment for the patient, prioritizing the analysis results of the image data by the image data analysis means. The proposed system according to claim 1 or 2.

4. The aforementioned image data includes a fundus image, The aforementioned image data analysis means quantifies the specific structure of the fundus image to determine the possibility of disease. The proposed system according to claim 1.

5. The aforementioned image data includes an OCT image, The image data analysis means quantifies the unique structure of the OCT image to determine the possibility of disease. The proposed system according to claim 1.

6. The proposed system comprises a terminal device installed within a hospital, and a server connected to the terminal device that has a trained model to assist in the determination of disease detection, The server has an image data analysis means that uses the trained model to analyze the diseases contained in the image data based on the similarity between the features of the image data and the features of existing ophthalmic image samples, using image data which is the result of an image diagnosis of the eye obtained in an ophthalmic diagnosis. The terminal device is a proposal system that has an optimal treatment proposal means that determines and proposes the optimal treatment for a patient based on numerical data obtained from an ophthalmic diagnosis and the results of the analysis of the image data by the image data analysis means.

Images