Medical diagnosis support device and program

Abstract

To enable determination of a priority order of inspections or treatments of a patient by estimating a possibility of diseases from a plurality of pieces of vital information by using AI.SOLUTION: A medical diagnosis support device 100 includes: an acquisition unit 111 that acquires a plurality of pieces of biological information of each patient; an estimation unit 112 that estimates a disease and a possibility of affection of the disease by using a learned model 103 from the plurality of pieces of biological information of each patient that have been acquired by the acquisition unit 111; and a control unit 113 that determines modality recommended to use for the patient on the basis of the disease and the possibility of affection of each patient estimated by the estimation unit 112.SELECTED DRAWING: Figure 6

Description

Technical Field

[0001] The present invention relates to a medical diagnosis support device Place and a program.

Background Art

[0002] As a device for determining the urgency of an injured or sick person, a triage support device has been proposed that uses AI (Artificial Intelligence) to determine the urgency of an injured or sick person from the characteristics of a captured image captured by a drone (see, for example, Patent Document 1). In addition, a device related to a rescue plan presentation system has also been disclosed that calculates the degree of danger indicating that the body is in a dangerous state for each of a plurality of rescue targets who need rescue and determines the priority order of rescue (for example, Patent Document 2).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, in Patent Document 1, since the degree of disease is determined from a captured image (photo), it has been difficult to determine the urgency of a diseased person. Further, in Patent Document 2, since the priority order of rescue is determined according to the degree of danger of the rescue target, it has not been possible to recognize the urgency of whether or not it is necessary to rescue the rescue target at an early stage.

[0005] Therefore, an object of the present invention is to provide a medical diagnosis support device and and a program that can determine the priority order of patient examinations and treatments by estimating the possibility of each disease from a plurality of vital signs using AI.

Means for Solving the Problem

[0006] That is, the above problems of the present invention are solved by the following configuration. (1) An acquisition unit that acquires a plurality of biological information of each patient, An estimation unit that estimates a disease and its likelihood of occurrence using a learned discrimination device from the plurality of biological information of each patient acquired by the acquisition unit, A control unit that determines a modality recommended for use with the patient based on the disease and its likelihood of occurrence of each patient estimated by the estimation unit, A medical diagnosis support device comprising: (2) On a computer, A step of acquiring a plurality of biological information of each patient, A step of estimating a disease and its likelihood of occurrence using a learned discrimination device from the plurality of biological information of each patient, A step of determining a modality recommended for use with the patient based on the disease and its likelihood of occurrence of each patient, A program characterized by causing the above steps to be executed.

Advantages of the Invention

[0007] According to the present invention, by using AI to estimate a disease and its likelihood of occurrence in a subject, it is possible to determine the priority order of examinations and treatments for a patient.

[0008] As a result, it is possible to expect an improvement in the prognosis of each patient, and to reduce medical costs and improve the patient's QOL (Quality Of Life).

Brief Description of the Drawings

[0009]

Fig. 1A

Fig. 1B

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Fig. 6

Fig. 7

Fig. 8

Modes for Carrying Out the Invention

[0010] Hereinafter, the modes for carrying out the present invention will be described in detail. Note that the embodiments described below are examples for realizing the present invention, and should be appropriately modified or changed according to the configuration of the device to which the present invention is applied and various conditions. The present invention is not limited to the following embodiments. Also, some of the following embodiments may be appropriately combined and configured. Note that the same members are denoted by the same reference numerals, and the description will be omitted as appropriate.

[0011] <The First Embodiment> [Configuration of the Entire Medical Diagnosis Support Device] FIG. 1A is a diagram for explaining a configuration example of a medical diagnosis support apparatus 100 according to the first embodiment. The medical diagnosis support apparatus 100 according to the first embodiment includes a CPU 101, an operation reception unit 102, a learned model 103, an output unit 104, a display unit 105, a ROM (Read Only Memory) 106, a RAM (Random Access Memory) 107, and an external storage device 108. The external storage device 108 includes a disease estimation program 109.

[0012] As shown in FIG. 1A, the CPU 101 is an arithmetic processing unit that comprehensively controls the entire medical diagnosis support apparatus 100. The CPU 101 performs various controls by executing programs stored in the ROM 106 and the external storage device 108. For example, by executing the disease estimation program 109 stored in the external storage device 108, the CPU 101 realizes an acquisition unit 111, an estimation unit 112, a control unit 113, and a calculation unit 114 shown in FIG. 1B.

[0013] FIG. 1B is a functional block diagram showing the configuration of the CPU 101 of the medical diagnosis support apparatus 100 according to the first embodiment.

[0014] The acquisition unit 111 has a function of acquiring a plurality of biometric information (vital data) of each patient. Vital data is, for example, information such as age, gender, blood pressure, respiratory rate, pulse, electrocardiogram, body temperature, oxygen saturation, medical history, smoking history, purpose of visit, and method of visit. The acquisition unit 111 automatically acquires measurement results or manually acquires them based on user input, for example.

[0015] The estimation unit 112 has a function of estimating diseases and their likelihoods of occurrence from the plurality of biometric information of each patient acquired by the acquisition unit 111 using a learned model (learned discrimination device) 103. For example, for each patient, the estimation unit 112 estimates diseases and their likelihoods of occurrence as follows: cerebral aneurysm: 60%, diabetes: 55%, asthma: 40%, lung cancer: 30%, fracture: 25%, pneumothorax: 10%, calculus: 3%, etc.

[0016] The control unit 113 has a function of outputting to the output unit 104 the diseases of each patient estimated by the estimation unit 112 and the likelihood of suffering from those diseases.

[0017] The calculation unit 114 has a function of calculating the risk level of each disease based on the likelihood of suffering from each disease. The calculation unit 114 can also calculate the overall risk level according to the risk level (score) of each disease. Note that the risk level indicates the degree of urgency requiring immediate treatment, encompassing both the meaning of risk degree and emergency degree.

[0018] In this embodiment, the CPU 101 controls the entire medical diagnosis support device 100 in an overall manner. Instead of controlling the entire medical diagnosis support device 100, for example, a plurality of hardware components may share the processing to control the entire medical diagnosis support device 100.

[0019] The operation reception unit 102 receives the input of patient information and patient biometric information. The operation reception unit 102 is composed of operation members such as a keyboard, a mouse, various switches, buttons, a touch panel, etc., which receive various operations from the user.

[0020] The learned model 103 is configured to include a pre-learned discriminator and stores and saves various parameters. Details of the learned model 103 will be described later.

[0021] The output unit 104 outputs the diseases and the likelihood of suffering from them to the display unit 105. The output unit 104 is composed of, for example, a diagnostic terminal used by a doctor or the like.

[0022] The display unit 105 displays the diseases and the likelihood of suffering from them. The display unit 105 is composed of, for example, a display device such as a liquid crystal display, an organic EL display, a printer, or software (viewer) for display.

[0023] The ROM 106 stores, for example, a control program for controlling the medical diagnosis support device 100. The RAM 107 is composed of, for example, a DRAM (Dynamic Random Access Memory), and functions as a working memory that temporarily stores data and image data necessary for the CPU 101 to execute a control program and a disease estimation program 109.

[0024] The external storage device 108 is composed of, for example, a hard disk drive, and stores the risk level of diseases of each patient calculated by the calculation unit 114.

[0025] [Operation of Medical Diagnosis Support Device] Next, the operation of the medical diagnosis support device 100 having the above configuration will be described using a flowchart with reference to FIGS. 1A and 1B.

[0026] FIG. 2 is a flowchart showing the operation of the medical diagnosis support device 100 according to the first embodiment.

[0027] First, the CPU 101 of the medical diagnosis support device 100 acquires a plurality of biometric information of each patient in the acquisition unit 111 (step S001). The acquisition unit 111 receives, for example, patient information and the input of patient biometric information (vital data) of each patient. Note that when acquiring a plurality of biometric information of each patient, the CPU 101 may automatically acquire the measurement results of each patient, or may receive the input from the operation reception unit 102.

[0028] The CPU 101 of the medical diagnosis support device 100 estimates one or more diseases and their likelihood of occurrence for each patient using the learned model 103 from the plurality of biometric information of each patient acquired by the acquisition unit 112 in the estimation unit 112 (step S003).

[0029] The CPU 101 of the medical diagnosis support device 100 calculates the risk level of each disease based on the likelihood of occurrence of each disease in the calculation unit 114 (step S005). In this case, in the calculation unit 114, the CPU 101 calculates the risk level for each patient from the priority score and the likelihood of occurrence (simply referred to as possibility) of each disease.

[0030] Figure 3 is an explanatory diagram showing the risk level for each disease based on the disease and its possibility of a certain patient (e.g., Patient C). As shown in Figure 3, columns of "Priority Score", "Possibility", and "Risk Level" are provided on the horizontal axis, and "Stroke", "Brain Aneurysm", "Fracture", "Lung Cancer", "Diabetes", etc. are provided on the vertical axis.

[0031] The priority score is external data determined as a default value. This priority score can adopt the value if there are national standards, for example, and can also be arbitrarily edited according to the authority of the doctor and the regulations of the hospital, etc. In addition, this priority score can be made variable according to the priority order of each hospital, the prevalence of infectious diseases, etc.

[0032] The possibility indicates the estimated probability of each disease being affected in the estimation unit 112. The risk level is a value calculated by using the priority score and the possibility of that disease. In this embodiment, this value is described using the value obtained by integrating the priority score and the possibility of that disease, but this embodiment is not limited to this, and in addition to the priority score and the possibility of the disease, biological information such as age, gender, and blood pressure can also be used for correction, and various methods can be adopted as long as it is a value that can be calculated using the priority score and the possibility of that disease.

[0033] In the case of Mr. C shown in Fig. 3, for "stroke", the priority score of "100" and the probability of "0.8" are multiplied, and the risk level is calculated as "80". Similarly, for "cerebral aneurysm", the priority score of "90" and the probability of "0.6" are multiplied, and the risk level is calculated as "54". For "fracture", the priority score of "60" and the probability of "0.2" are multiplied, and the risk level is calculated as "12". For "lung cancer", the priority score of "90" and the probability of "0.2" are multiplied, and the risk level is calculated as "18". For "diabetes", the priority score of "20" and the probability of "0.4" are multiplied, and the risk level is calculated as "8". Note that in this embodiment, the calculation unit 114 not only calculates the "risk level", but also may display the reason and details for which the disease with high severity is output when the disease is output.

[0034] Returning to the flowchart of Fig. 2, the CPU 101 of the medical diagnosis support device 100 calculates the overall risk level for each patient from the risk levels of the diseases of each patient in the calculation unit 114 (step S007).

[0035] In this case, the calculation unit 114 calculates the sum of the risk levels for each disease for each patient, and calculates the overall risk level for each patient. For example, in the case of Mr. C in Fig. 3, the calculation unit 114 adds the risk level of "80" for stroke, the risk level of "54" for cerebral aneurysm, the risk level of "12" for fracture, the risk level of "18" for lung cancer, the risk level of "8" for diabetes, etc., and calculates the overall risk level as "172".

[0036] The CPU 101 of the medical diagnosis support device 100 outputs the overall risk level for each patient calculated by the calculation unit 114 to the output unit 104. The output unit 104 displays the overall risk level for each patient on the display unit 105 (step S009). In this case, the calculation unit 114 of the CPU 101 performs triage for multiple patients from the overall risk levels of the multiple patients, and displays the triage result on the display unit 105 to end the process of the medical diagnosis support device 100.

[0037] Note that triage refers to determining and selecting the priority of treatment according to the severity of the patient.

[0038] FIG. 4 is an explanatory diagram showing the result of performing triage on a plurality of patients from the comprehensive risk levels of the plurality of patients. The CPU 101 can display the patients in descending order of the comprehensive risk level by performing triage in the calculation unit 114.

[0039] As shown in FIG. 4, for example, columns for the comprehensive risk level and the usage modality are provided for each patient's name, and four patients are displayed in the column for the name in descending order of the comprehensive risk level. Specifically, following the order of the highest comprehensive risk level, patient C has a comprehensive risk level of "172", patient B has a comprehensive risk level of "135", patient D has a comprehensive risk level of "96", and patient A has a comprehensive risk level of "47".

[0040] Thereby, medical staff can determine the priority order of examinations and treatments in order from patient C with the highest comprehensive risk level. Further, the medical diagnosis support device 100 may display the disease and the likelihood of suffering of the selected patient when the medical staff selects the patient name, and display an explanatory diagram (see, for example, FIG. 3) in which the risk level for each disease is calculated.

[0041] Note that in the column for the usage modality, the modality that each patient is scheduled to use is displayed. The usage modality will be described in the second embodiment.

[0042] As described above, the CPU 101 of the medical diagnosis support device 100 according to the first embodiment includes an acquisition unit 111, an estimation unit 112, a control unit 113, and a calculation unit 114. The CPU 101 estimates the disease and the likelihood of suffering thereof from the plurality of biological information of each patient acquired by the acquisition unit 111 using the learned model 103 in the estimation unit 112. Further, the CPU 101 calculates the risk level of each disease based on the likelihood of suffering from each disease in the calculation unit 114, and calculates the comprehensive risk level for each patient from the risk levels of the diseases of each patient.

[0043] As a result, the medical diagnosis support device 100 according to the first embodiment can prevent overlooking of diseases at an early stage by grasping the diseases of the subject and the likelihood of their occurrence, and can extract the subjects who should be preferentially diagnosed in a situation where there are a large number of test subjects at an emergency site or the like.

[0044] Furthermore, the medical diagnosis support device 100 according to the first embodiment can be expected to improve the prognosis of each patient, and can reduce medical costs and improve the QOL of patients.

[0045] <Second Embodiment> The CPU 101 of the medical diagnosis support device 100 according to the second embodiment calculates, in the calculation unit 114, a list of modalities (hereinafter, also referred to as usage modalities) recommended for use for each patient from the risk levels of each disease of each patient. Note that this usage modality also includes the meaning of the modality that the patient in the first embodiment is planned to use.

[0046] According to the CPU 101 of the medical diagnosis support device 100 according to the second embodiment, since a list of usage modalities can be displayed for each patient, medical staff such as doctors and nurses can determine the priority order of patient examinations and treatments.

[0047] [Operation of Medical Diagnosis Support Device] Next, the operation of the medical diagnosis support device 100 according to the second embodiment will be described using the flowchart shown in FIG. 5. In the operation of the medical diagnosis support device 100 according to the second embodiment, the same operations as those of the medical diagnosis support device 100 according to the first embodiment shown in FIG. 2 are denoted by the same reference numerals, and the description thereof will be omitted as appropriate.

[0048] FIG. 5 is a flowchart showing a process in which the medical diagnosis support device 100 according to the second embodiment calculates a list of modalities recommended for use for each patient.

[0049] First, in steps S001 and S003 shown in FIG. 5, the same processes as steps S001 and S003 shown in FIG. 2 are performed.

[0050] Next, the CPU 101 of the medical diagnosis support device 100 calculates, in the calculation unit 114, the risk level of each disease based on the likelihood of each disease, and calculates a list of modalities to be used for each patient from the risk levels of each disease of each patient (step S101).

[0051] For example, referring to FIG. 3, from the possibility of stroke and the possibility of cerebral aneurysm, "MRI (Magnetic Resonance Imaging)" is determined as the modality recommended for use, and from the possibility of fracture, "CT (Computed Tomography)" is determined as the modality recommended for use. Also, from the possibility of lung cancer, "X-ray" is determined as the modality recommended for use, and from the possibility of diabetes, "US (UltraSonography)" is determined as the modality recommended for use.

[0052] Thereby, in step S009, as shown in FIG. 4, the modality to be used can be displayed for each patient. Also, in FIG. 4, an index is displayed according to the ratio of the modality recommended for use from the likelihood of disease. For example, it shows that the possibility of selecting (using) MRI as the modality is high.

[0053] As described above, the CPU 101 of the medical diagnosis support device 100 according to the second embodiment calculates, in the calculation unit 114, a list of modalities recommended for use for each patient from the risk levels of each disease of each patient.

[0054] Thereby, the CPU 101 of the medical diagnosis support device 100 according to the second embodiment can display a list of modalities to be used for each patient, so that medical staff such as doctors and nurses can determine the priority order of patient examinations and treatments.

[0055] <Third Embodiment> The CPU 101 of the medical diagnosis support device 100 according to the third embodiment executes by combining the operations of the medical diagnosis support device 100 according to the first embodiment and the operations of the medical diagnosis support device 100 according to the second embodiment.

[0056] [Operation of Medical Diagnosis Support Device] Next, the operation of the medical diagnosis support device 100 according to the third embodiment will be described using the flowchart shown in FIG. 6. The operation of the medical diagnosis support device 100 according to the third embodiment executes the same process as the process in which the process of step S101 of the flowchart shown in FIG. 5 is added to the flowchart shown in FIG. 2. The same reference numerals are given to the same processes, and the description will be omitted as appropriate.

[0057] FIG. 6 is a flowchart showing the operation of the medical diagnosis support device 100 according to the third embodiment.

[0058] The medical diagnosis support device 100 according to the third embodiment first acquires a plurality of biological information of each patient in the acquisition unit 111 (step S001). The acquisition unit 111 receives, for example, patient information and input of patient biological information (vital data) of each patient.

[0059] The CPU 101 of the medical diagnosis support device 100 estimates one or more diseases and their likelihood of occurrence for each patient using the learned model 103 from the plurality of biological information of each patient acquired by the acquisition unit 112 in the estimation unit 112 (step S003).

[0060] The CPU 101 of the medical diagnosis support device 100 calculates the risk level of each disease based on the likelihood of occurrence of each disease in the calculation unit 114 (step S005). In the calculation unit 114, the CPU 101 calculates the risk level for each patient from the priority score and the likelihood (probability) of each disease.

[0061] The CPU 101 of the medical diagnosis support device 100 calculates the overall risk level for each patient from the risk levels of the diseases of each patient in the calculation unit 114 (step S007). The calculation unit 114 calculates the sum of the risk levels for each disease for each patient and calculates the overall risk level for each patient.

[0062] Also, the CPU 101 of the medical diagnosis support device 100 calculates (determines) a list of modalities to be used for each patient from the overall risk level of each patient in the calculation unit 114 (step S101). Note that when the modality can be calculated (determined) from the risk level of each disease of each patient, the calculation unit 114 may calculate (determine) the modality from the risk level of each disease of each patient.

[0063] The CPU 101 of the medical diagnosis support device 100 displays, on the display unit 105, the result of performing the triage shown in FIG. 4 according to the overall risk level for each patient calculated by the calculation unit 114 (step S009), and displays a list of modalities to end the processing of the medical diagnosis support device 100.

[0064] As described above, the CPU 101 of the medical diagnosis support device 100 according to the third embodiment includes an acquisition unit 111, an estimation unit 112, a control unit 113, and a calculation unit 114. In the calculation unit 114, the CPU 101 estimates a disease and its likelihood of occurrence from a plurality of biological information of each patient acquired by the acquisition unit 111 using the learned model 103. The CPU 101 calculates the risk level of each disease based on the likelihood of occurrence of each disease, and calculates the overall risk level and the list of modalities to be used for each patient from the risk levels of the diseases of each patient.

[0065] Thereby, the medical diagnosis support device 100 according to the first embodiment can select patients with a high priority for treatment by grasping the diseases of the subjects and their likelihood of occurrence, and can also easily arrange the modalities to be used, so that many patients can be efficiently rescued and treated.

[0066] Next, the generation of the learned model of the medical diagnosis support device 100 will be described.

[0067] [Generation of Learned Model] The learned model 103 uses, as learning sample data, a large number (i sets, where i is, for example, several thousand to several hundred thousand) of data sets with pre-prepared biological information (such as vital data) as input and the possibility of disease as output, and performs machine learning thereby. As the learning device (not shown), a stand-alone high-performance computer using processors of a CPU and a GPU, or a cloud computer can be used.

[0068] In this embodiment, the learning device (not shown) will be described using a flowchart for a learning method using a neural network configured by combining perceptrons. However, this embodiment is not limited thereto, and various methods can be adopted as long as it is supervised learning.

[0069] Specifically, for the learning device (not shown), random forest, support vector machine (SVM), boosting, Bayesian, network linear discrimination method, non-linear discrimination method, etc. can be applied.

[0070] [Learning Method of Learned Model] FIG. 7 is a flowchart showing the processing of the learning method of the learned model 103 of the medical diagnosis support device 100.

[0071] The learning device (not shown) reads the learning sample data that is teacher data. For example, at the first time, the first set of learning sample data (biological information, possibility of disease) is read, and at the i-th time, the i-th set of learning sample data is read (step S201).

[0072] FIG. 8 is an explanatory diagram showing biometric information (vital data) prepared in advance for an arbitrary patient X. As shown in FIG. 8, the biometric information (vital data) is composed of, for example, input data such as age, gender, blood pressure, respiratory rate, pulse, electrocardiogram, body temperature, oxygen saturation, past history, smoking history, purpose of visit, method of visit, and others, and teacher data such as cerebral aneurysm, fracture, lung cancer, diabetes, and others. Then, the trained model 103 reads these biometric information (vital data) in order as learning sample data.

[0073] The learning device (not shown) inputs the input data among the read learning sample data into the neural network (step S203).

[0074] The learning device (not shown) compares the estimation result of the neural network, that is, the estimated disease, with the input teacher data (step S205).

[0075] The learning device (not shown) adjusts the parameters from the comparison result (step S207). For example, the learning device (not shown) performs a process called back-propagation (error backpropagation method) to adjust and update the parameters so that the error of the comparison result becomes smaller.

[0076] When the learning device (not shown) has completed all processing of all data up to the 1st to i-th sets (YES in step S209), it proceeds to step S211. On the other hand, when all processing of all data has not been completed (NO in step S209), it returns to step 201, reads the next learning sample data, and repeats the processing below step S201.

[0077] The learning device (not shown) stores the trained model 103 constructed by the processing up to step S207 (step S211) and ends the processing.

[0078] As a result, the CPU 101 of the medical diagnosis support device 100 according to the fourth embodiment can estimate a disease and its likelihood of occurrence from a plurality of biometric information of each patient acquired by the acquisition unit 111 using the learned model 103 in the estimation unit 112.

Explanation of Reference Numerals

[0079] 100 Medical diagnosis support device 101 CPU 102 Operation reception unit 103 Learned model 104 Output unit 105 Display unit 106 ROM 107 RAM 108 External storage device 109 Disease estimation program 111 Acquisition unit 112 Estimation unit 113 Control unit 114 Calculation unit

Claims

1. An acquisition unit that acquires a plurality of biological information of each patient; An estimation unit that estimates a disease and its likelihood of occurrence from the plurality of biological information of each patient acquired by the acquisition unit using a learned identification device; A control unit that determines a modality recommended for use for the patient based on the disease and its likelihood of occurrence of each patient estimated by the estimation unit; A medical diagnosis support device comprising the above.

2. A program for causing a computer to perform a step of acquiring a plurality of biological information of each patient; perform a step of estimating a disease and its likelihood of occurrence from the plurality of biological information of each patient using a learned identification device; perform a step of determining a modality recommended for use for the patient based on the disease and its likelihood of occurrence of each patient; characterized by the above.

Images