KR102168216B1 - Method for predicting pneumonia - Google Patents

Abstract

A method of determining the probability of occurrence of pneumonia using an apparatus for determining the probability of occurrence of pneumonia according to an embodiment of the present invention includes: acquiring information on a plurality of items and environment information about a patient to be diagnosed; Determining a node value of at least one input node included in an input layer by using information on a plurality of items and an environment on the patient to be diagnosed; Determining a node value of at least one output node included in an output layer from a learned pneumonia incidence probability calculator based on the node value of the at least one input node; And determining a probability of developing pneumonia in the patient to be diagnosed based on the node value of the at least one output node.

Description

How to determine the probability of developing pneumonia{METHOD FOR PREDICTING PNEUMONIA}

Embodiments of the present invention relate to a method of determining the onset probability of pneumonia, and to a method of determining the onset probability of pneumonia using an artificial neural network learned based on a plurality of learning data.

Pneumonia is an inflammation of the lungs. The most common cause is infection through bacteria, and viruses, fungi, or other microorganisms can also be the cause. Rarely, it can be caused by an allergic reaction or inhalation of irritating chemicals. Inhaled pneumonia is more common among the elderly, young children, patients with poor overall condition, or people with weak cough reflexes.

Pneumonia such as this is the most common cause of death worldwide today. In particular, pneumonia is the last disease that causes death in patients with other serious chronic diseases. In the United States, 2-3 million people get pneumonia each year, of which about 60,000 die. In the United States, pneumonia, along with influenza, is the eighth leading cause of death and the first infectious cause of death. Pneumonia is the most common cause of death among infections that occur during hospital stays, and the most common overall cause of death in developing countries.

As described above, pneumonia is a fatal disease despite the possibility of curing it, so it is necessary to more accurately predict the occurrence probability of pneumonia and actively deal with it.

The present invention attempts to reasonably infer the probability of onset of pneumonia from patient information and environmental information.

In addition, the present invention intends to infer the probability of death and the occurrence of related diseases with high reliability in addition to the above-described pneumonia occurrence probability from patient information.

A method of determining the probability of occurrence of pneumonia using an apparatus for determining the probability of occurrence of pneumonia according to an embodiment of the present invention includes: acquiring information on a plurality of items and environment information about a patient to be diagnosed; Determining a node value of at least one input node included in an input layer by using information on a plurality of items and an environment on the patient to be diagnosed; Determining a node value of at least one output node included in an output layer from a learned pneumonia incidence probability calculator based on the node value of the at least one input node; And determining a probability of developing pneumonia in the patient to be diagnosed based on the node value of the at least one output node.

In this case, the pneumonia incidence probability calculator includes information on a plurality of items for each of a plurality of patients, information on a plurality of environments, and learning data including a probability of onset of pneumonia for each of the plurality of patients. A neural network learned about the correlation between information and information on a plurality of environments and the probability of onset of pneumonia, wherein the neural network is a hidden defined between the input layer, the output layer, and the input layer and the output layer. A layer, a first function between the input layer and the hidden layer, and a second function between the hidden layer and the output layer.

The information of the plurality of items on the patient includes at least one of the patient's personal information, the patient's vital sign information, and the patient's medical information, and the patient's medical information includes the patient's past medical history, It may include at least one of the patient's medical image, the patient's drug taking history, and the patient's symptom history.

The input layer may include a first input node to which the patient's personal information is input, a second input node to which vital sign information of the patient is input, and a third input node to which medical information of the patient is input.

The information on the environment includes at least one of information on a concentration of atmospheric fine dust, humidity, and ultraviolet rays during a predetermined period, and the input layer further includes a fourth input node into which the concentration of the fine dust is input. can do.

The method for determining the probability of pneumonia incidence further comprises: learning the neural network based on the learning data, wherein the learning of the neural network includes information on a plurality of items about a patient received from an external device, Classifying patient information including the occurrence diagnosis code as first valid learning data; Classifying information of a patient, including a medical image related to pneumonia, as second effective learning data, in a plurality of items of information about a patient received from the external device; Classifying information of a patient including a keyword suggesting pneumonia in a plurality of items of information about a patient received from the external device as third effective learning data; Classifying information of a patient including a medical history of a specific disease within a predetermined period before the onset of pneumonia in a plurality of items of information about a patient received from the external device as fourth effective learning data; And determining at least some of the first valid learning data, the second valid learning data, the third valid learning data, and the fourth valid learning data as the learning data.

In the determining of the learning data, learning data belonging to all of the first valid learning data, the second valid learning data, the third valid learning data, and the fourth valid learning data may be determined as the learning data.

The neural network may be learned by further including a probability of death of the patient according to time and a probability of occurrence of a first disease of the patient. In this case, the output layer may further include a second output node corresponding to a probability of death of the patient over time and a third output node corresponding to a probability of the patient's first disease.

The first disease may be a plurality, and the output layer may include a third output node for each of the plurality of first diseases.

Other aspects, features, and advantages other than those described above will become apparent from the following drawings, claims, and detailed description of the invention.

According to embodiments of the present invention, it is possible to reasonably infer a probability of developing pneumonia from patient information.

In addition, in addition to the above-described probability of occurrence of pneumonia, the probability of death and the probability of occurrence of related diseases can be inferred with high reliability.

1 schematically illustrates a system for determining a probability of onset of pneumonia according to an embodiment of the present invention.
2 schematically shows the configuration of an apparatus for determining an onset probability of pneumonia according to an embodiment of the present invention.
3 is a diagram illustrating a structure of a neural network learned by a controller according to an embodiment of the present invention.
4 is a flowchart illustrating a method of determining the probability of occurrence of pneumonia performed by the apparatus for determining the probability of occurrence of pneumonia according to an embodiment of the present invention.
5 is an example of a screen displayed on a user terminal according to an embodiment of the present invention.

Since the present invention can apply various transformations and have various embodiments, specific embodiments are illustrated in the drawings and will be described in detail in the detailed description. Effects and features of the present invention, and a method of achieving them will be apparent with reference to the embodiments described later in detail together with the drawings. However, the present invention is not limited to the embodiments disclosed below and may be implemented in various forms.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings, and when describing with reference to the drawings, the same or corresponding constituent elements are assigned the same reference numerals, and redundant descriptions thereof will be omitted. .

In the following embodiments, terms such as first and second are not used in a limiting meaning, but for the purpose of distinguishing one component from another component. In the following examples, the singular expression includes the plural expression unless the context clearly indicates otherwise. In the following embodiments, terms such as include or have means that the features or elements described in the specification are present, and do not preclude the possibility of adding one or more other features or elements in advance. In the drawings, components may be exaggerated or reduced in size for convenience of description. For example, the size and shape of each component shown in the drawings are arbitrarily shown for convenience of description, and thus the present invention is not necessarily limited to what is shown.

1 schematically illustrates a system for determining a probability of onset of pneumonia according to an embodiment of the present invention.

Referring to FIG. 1, the system for determining the probability of occurrence of pneumonia according to an embodiment of the present invention may include a server 100, a user terminal 200, an external device 300, and a communication network 400 connecting them. .

The server 100 of the system for determining the probability of pneumonia according to an embodiment of the present invention receives information on a plurality of items and environment information about a patient from the user terminal 200 and/or the external device 300, Based on the received information, the probability of developing pneumonia of the patient may be determined.

In addition, the server 100 includes information on a plurality of items for a plurality of patients from the user terminal 200 and/or the external device 300, information on a plurality of environments, and the probability of onset of pneumonia for each of a plurality of patients (or Neural network can be trained based on the fact of the outbreak). A detailed description of this will be described later.

In the present invention,'information of a plurality of items on a patient' (or'patient information') may include information on various items of a corresponding patient, such as its name. For example, the information of a plurality of items on the patient may include at least one of personal information of the patient, vital sign information of the patient, and medical information of the patient.

The patient's personal information may include general information about the patient, such as the patient's name, the patient's gender, the patient's age, the patient's date of birth, the patient's name, the patient's residence address, and the patient's guardian information.

The patient's vital sign information may include, for example, blood pressure information, pulse information, oxygen saturation information, body temperature information, and the like according to the passage of time of the patient.

The patient's medical information includes, for example, the patient's past medical history (injury code, etc.), the patient's medical image (MRI image, CT image, X-ray image, etc.), the patient's medication history, the patient's symptom history, and the patient's It may include a diagnosis history and the like.

However, the above-described items are exemplary, and the spirit of the present invention is not limited thereto. Therefore, in addition to the listed items, general information about the patient may be included in the patient's personal information, information about the patient's vital signs may be included in the patient's vital signs information, and the patient's medical information may be included in the patient's medical information.

In the present invention,'environmental information' (or'environmental information') may include various items of information on the natural environment. For example, the information on the environment may include information on the concentration of fine air dust, humidity, and ultraviolet rays level by region for a predetermined period.

In an optional embodiment, the'information on the environment' may include only various items of information on the environment around the patient's residence. However, this is merely an example and the spirit of the present invention is not limited thereto.

In the present invention, the'probability of pneumonia' refers to the probability of developing pneumonia in a specific patient, and is a fixed value (for example, 70%, etc.) regardless of the passage of time, or a value that changes with the passage of time. (For example, 18% after 1 week, 25% after 2 weeks, 11% after 3 weeks, etc.).

The'neural network' included in the'pneumonia incidence probability calculator' described later in the present invention may mean an artificial neural network that is learned by machine learning or deep learning. At this time, the artificial neural network refers to a model in which artificial neurons (hereinafter referred to as'nodes') that form a network through the combination of synapses change the strength of the synapses through learning to learn problem solving ability. I can. A description of the structure of the neural network will be described later with reference to FIG. 3.

The user terminal 200 according to an embodiment of the present invention may refer to various devices capable of transmitting/receiving patient information and environmental information to and from the server 100. In this case, the terminal may be a portable terminal 201 or a personal computer 202.

According to an embodiment of the present invention, the user terminal 200 includes a display means for displaying a screen on which the probability of occurrence of pneumonia received from the server 100 is displayed, and an input means for acquiring a user's input to the screen. Can be equipped. At this time, the input means and the display means may be configured in various ways. For example, the input means may include a keyboard, a mouse, a trackball, a microphone, a button, and a touch panel, but are not limited thereto.

In the present invention, the external device 300 may mean various devices that transmit and receive data through the server 100 and/or the user terminal 200 and the communication network 400.

According to an embodiment of the present invention, the external device 300 may be a device that provides learning data for learning a pneumonia incidence probability calculator provided in the server 100. For example, the external device 300 may be a server of a medical institution that provides information on a plurality of items for a plurality of patients.

In addition, the external device 300 may be a weather information providing server that provides information on environments of various regions in various time zones. In this way, the external device 300 may be singular or plural.

In the present invention, the communication network 400 serves to connect the server 100, the user terminal 200, and the external device 300. For example, the communication network 400 provides an access path so that the user terminal 200 can transmit and receive packet data after accessing the server 100. The communication network 400 is, for example, a wired network such as LANs (Local Area Networks), WANs (Wide Area Networks), MANs (Metropolitan Area Networks), ISDNs (Integrated Service Digital Networks), wireless LANs, CDMA, Bluetooth, satellite communication, etc. It may cover wireless networks of, but the scope of the present invention is not limited thereto.

The server 100 according to an embodiment of the present invention receives information about a plurality of items about a patient and information about the environment from the user terminal 200 and/or the external device 300, and based on the received information, The probability of developing pneumonia in a patient can be determined.

In addition, the server 100 includes information on a plurality of items for a plurality of patients from the user terminal 200 and/or the external device 300, information on a plurality of environments, and the probability of onset of pneumonia for each of a plurality of patients (or Neural network can be trained based on the fact of the outbreak). To this end, the server 100 may include a device for determining the probability of onset of pneumonia as shown in FIG. 2.

2 schematically shows the configuration of the apparatus 110 for determining the probability of onset of pneumonia according to an embodiment of the present invention.

Referring to FIG. 2, the apparatus 110 for determining the probability of occurrence of pneumonia according to an embodiment of the present invention may include a communication unit 111, a control unit 112, and a memory 113. In addition, although not shown in the drawing, the apparatus 110 for determining the probability of occurrence of pneumonia according to the present embodiment may further include an input/output unit, a program storage unit, and the like.

The communication unit 111 is for transmitting and receiving a signal such as a control signal or a data signal through a wired or wireless connection with another network device such as the user terminal 200 and/or the external device 300 by the apparatus 110 for determining the probability of pneumonia. It may be a device that includes the necessary hardware and software.

The controller 112 may include all types of devices capable of processing data, such as a processor. Here, the'processor' may refer to a data processing device embedded in hardware having a circuit physically structured to perform a function expressed by, for example, a code or command included in a program. As an example of such a data processing device built into the hardware, a microprocessor, a central processing unit (CPU), a processor core, a multiprocessor, and an application-specific integrated (ASIC) Circuit) and processing devices such as a Field Programmable Gate Array (FPGA) may be covered, but the scope of the present invention is not limited thereto.

The memory 113 performs a function of temporarily or permanently storing data processed by the apparatus 110 for determining the probability of occurrence of pneumonia. The memory 113 may include a magnetic storage medium or a flash storage medium, but the scope of the present invention is not limited thereto.

Hereinafter, description will be made on the premise that the apparatus 110 for determining the probability of occurrence of pneumonia is provided in the server 100, but the apparatus 110 for determining the probability of occurrence of pneumonia may be provided separately from the server 100 according to role allocation.

The apparatus 110 for determining the probability of occurrence of pneumonia provided in the server 100 according to an embodiment of the present invention provides information on a plurality of items about a patient and environment from the user terminal 200 and/or the external device 300. Information (hereinafter referred to as “patient and environmental information”) may be received, and a probability of developing pneumonia of a corresponding patient may be determined based on the received information.

In addition, the apparatus 110 for determining the probability of pneumonia incidence includes information on a plurality of items for a plurality of patients, information on a plurality of environments, and pneumonia for each of a plurality of patients from the user terminal 200 and/or the external device 300. The neural network may be trained based on the onset probability (or onset fact) of (hereinafter, described as'learning data').

As described above, the'pneumonia incidence probability calculator' in the present invention may include a neural network learned by the control unit 112 of the incidence probability determination device 110. In this case, the'neural network' may mean an artificial neural network that is learned by machine learning or deep learning. Here, the'artificial neural network' refers to a model in which artificial neurons (hereinafter referred to as'nodes'), which form a network by combining synapses, change the strength of synapses through learning to learn problem solving ability. It can mean.

3 is a view for explaining the structure of the'neural network' learned by the control unit 112 according to an embodiment of the present invention.

3, a neural network according to an embodiment of the present invention includes an input layer L1 including at least one input node N1, a hidden layer L2 including a plurality of hidden nodes N2, and at least An output layer L3 including one output node N3 may be included. At this time

Values corresponding to patient and environmental information acquired by the controller 112 may be input to at least one input node N1 of the input layer L1. For example, a value corresponding to the patient's personal information may be input to the first input node, and a value corresponding to the patient's vital sign may be input to the second input node.

As illustrated, the hidden layer L2 may include one or more layers that are fully connected. When the hidden layer L2 includes a plurality of layers, the neural network may include a function (not shown) defining a relationship between each hidden layer.

At least one output node N3 of the output layer L3 may include an output value generated by a neural network from an input value of the input layer L1 under the control of the controller 112. For example, the first output node may include a value corresponding to the probability of developing pneumonia according to the passage of time of the patient.

In this case, a value included in each node of each layer may be a vector. In addition, each node may include a weight corresponding to the importance of the node.

Meanwhile, the neural network includes a first function (F1) that defines the relationship between the input layer (L1) and the hidden layer (L2), and a second function (F2) that defines the relationship between the hidden layer (L2) and the output layer (L3). can do.

The first function F1 may define a connection relationship between the input node N1 included in the input layer L1 and the hidden node N2 included in the hidden layer L2. Similarly, the second function F2 may define a connection relationship between the hidden node N2 included in the hidden layer L2 and the output node N2 included in the output layer L2.

The first function F1, the second function F2, and the functions between the hidden layer may include a recurrent neural network (RNN) model that outputs a result based on an input of a previous node.

In the process of learning the neural network by the control unit 112, the first function F1 and the second function F2 may be learned based on a plurality of training data. Of course, while the neural network is trained, functions between a plurality of hidden layers may also be learned in addition to the above-described first function F1 and second function F2.

In the present invention,'learning' means that the controller 112 updates the above-described functions (F1, F2, functions between hidden layers, etc.) based on a plurality of learning data and stores them in the memory 113. can do. For example, the controller 112 may appropriately adjust coefficients of elements constituting the first function based on a plurality of learning data, and store the coefficients in the memory 113. Of course, similarly to this, the control unit 112 may appropriately adjust the coefficients of the elements constituting the second function and store the coefficients in the memory 113. However, this is merely an example and the spirit of the present invention is not limited thereto.

The neural network according to an embodiment of the present invention may be trained in a supervised learning method based on labeled learning data. In this case, the learning data may be data in which the probability of onset of pneumonia (or fact of onset) of a corresponding patient is marked in information on a plurality of items about a patient and information on the environment. At this time, the'probability of pneumonia' is a fixed value (e.g., 70%, etc.) regardless of the passage of time as described above, or a value that changes with the passage of time (e.g., 18% after 1 week, 2 weeks). It may be 25% after 3 weeks, 11% after 3 weeks, etc.). 'Onset fact' may be a value indicating whether pneumonia has occurred.

The control unit 112 according to an embodiment of the present invention uses a plurality of training data, so that the output value generated by inputting any one training data into a neural network approaches the probability of onset of pneumonia marked in the corresponding training data. A neural network can be trained by repeatedly performing a process of updating functions (F1, F2, functions between hidden layers, etc.).

At this time, the controller 112 according to an embodiment of the present invention may update the above-described functions (F1, F2, functions between hidden layers, etc.) according to a back propagation algorithm. However, this is merely an example and the spirit of the present invention is not limited thereto.

Of course, the neural network according to an embodiment of the present invention may be trained in an unsupervised learning method based on unmarked learning data. However, this is merely an example and the spirit of the present invention is not limited thereto.

The controller 112 according to an embodiment of the present invention may obtain patient and environment information, and determine a node value of at least one input node N1 included in the above-described input layer L1 by using this.

For convenience of explanation, the neural network provides a first input node corresponding to the patient's personal information, a second input node corresponding to the patient's vital sign information, and a third input node corresponding to the patient's medical information in the input layer L1. Suppose to include.

In this case, the controller 112 according to an embodiment of the present invention may determine a node value of each of the first to third input nodes from the obtained information. In this case, the control unit 112 according to an embodiment of the present invention may extract feature values for each item by applying a convolutional neural network (CNN) to information of each item, and the extracted feature values may be in the form of vectors. .

In other words, the controller 112 according to an embodiment of the present invention may determine a first vector corresponding to the patient's personal information as the node value of the first input node by applying a convolutional neural network to the patient's personal information. . Similarly, the controller 112 may determine the node value of the second input node from the patient's vital sign information and the node value of the third input node from the patient's medical information.

The input layer L1 may include a fourth input node corresponding to information on the environment in addition to the information on the patient described above. In this case, the fourth input node may be a node having a node value corresponding to the concentration of fine dust. The node value of the fourth input node may be generated by the controller 112 by applying a convolutional neural network to information about the environment.

As a method of determining the node values of the first to fourth input nodes of the above-described input layer, it has been exemplarily described that the control unit 112 applies a convolutional neural network to information of each item, but such a convolutional neural network It is not necessary to use. Therefore, various algorithms capable of deriving the feature values of each item according to the characteristics of the data and/or the design of the system may be used.

Meanwhile, node values of the first to fourth input nodes may be n-dimensional, m-dimensional, i-dimensional, and j-dimensional vectors, respectively.

According to an embodiment of the present invention, the control unit 112 includes at least one included in the output layer from the pneumonia incidence probability calculator based on the input layer (that is, node values of input nodes included in the input layer) determined by the above-described process. The node value of the output node of can be determined.

In other words, the controller 112 according to an embodiment of the present invention may determine the node value of the output node output by the neural network when the node values of the input nodes are input.

Looking at this process in detail, the control unit 112 according to an embodiment of the present invention may determine node values of nodes of the first hidden layer by inputting and outputting values of input nodes to the first function F1. Similarly, the controller 112 may determine the node values of the nodes of the second hidden layer by inputting and outputting the node values of the nodes of the first hidden layer to a function defining the relationship between the hidden layers. In this case, the above-described'function defining a relationship between hidden layers' may be a function defining a relationship between a first hidden layer and a second hidden layer. In the same way, the controller 112 may determine the node values of the output nodes by inputting and outputting the node values of the nodes of the last hidden layer to the second function. In this way, the controller 112 may determine the output of the neural network corresponding to the node values of the input nodes, that is, the node value of the output node.

The control unit 112 according to an embodiment of the present invention may determine the probability of developing pneumonia of the patient based on the node value of the above-described output node. In this case, the output layer L3 according to an embodiment of the present invention may include a first output node corresponding to a probability of developing pneumonia of a patient. The node value of the first output node may be in the form of a vector such as [0, 1, 0, 0, 0, 0, 0, 0, 0, 0].

Meanwhile, the neural network according to an embodiment of the present invention may be learned by further including a probability of death of a patient according to time and a probability of occurrence of a first disease of the patient.

For example, the controller 112 according to an embodiment of the present invention may learn the probability of death of the patient according to time and the probability of occurrence of the patient's first disease in a supervised learning method. In this case, in addition to the probability of developing pneumonia, the probability of death and the probability of developing the patient's first disease may be further marked in the learning data. Of course, the controller 112 may learn the probability of death of the patient according to time and the probability of occurrence of the patient's first disease in an unsupervised learning method. However, this is merely an example and the spirit of the present invention is not limited thereto.

The output layer L3 according to an embodiment of the present invention includes a second output node corresponding to the probability of death of the patient over time, along with the aforementioned first output node, and a third output node corresponding to the probability of the patient's first disease. It may further include an output node.

At this time, the first disease is a disease related to pneumonia, and may be ascites. When there are a plurality of first diseases, the output layer L3 may include a third output node for each of the plurality of first diseases.

In this way, the present invention can reasonably infer the probability of onset of a specific disease, particularly pneumonia, from the patient's information, and also can infer the probability of death of the patient.

Hereinafter, a process of generating learning data from raw patient information and/or environment information by the control unit 112 according to an embodiment of the present invention will be described.

The controller 112 according to an embodiment of the present invention may receive a plurality of patient information and/or environment information from the user terminal 200 and/or the external device 300. For example, the controller 112 may receive a plurality of patient information from a server (not shown) of a medical institution. At this time, the received plurality of patient information may or may not be related to pneumonia. In other words, the plurality of patient information may include both information on a patient related to pneumonia and information on a patient related to other diseases.

In addition, the control unit 112 may receive environmental information of a plurality of time zones in a plurality of regions from a weather information providing server (not shown). At this time, the environmental information may also include environmental information of a region and/or a time point related to the patient, or may include environmental information of a region and/or a time point not related to the patient.

Accordingly, in order to use the received information for learning in a neural network, it may need to be properly processed.

The control unit 112 according to an embodiment of the present invention may generate learning data by extracting patient information related to pneumonia from a plurality of patient information. At this time, the controller 112 may extract patient information related to pneumonia based on whether or not a diagnosis code for pneumonia occurrence (eg, ICD10 standard, J100/J110/J12 ~ 18/J851, etc.) is included in the patient's medical information. Also, the controller 112 may extract patient information related to pneumonia based on whether or not a medical image related to pneumonia is included in the medical information of the patient. In addition, the control unit 112 may treat pneumonia on the basis of the presence or absence of keywords (Pneumonia, Pneumonic, Consolidation, Infiltration, Ground glass opacity (opacities) = GGO, Parapneumonic, etc.) that are related to pneumonia or suggest pneumonia in the patient's medical information. Related patient information can be extracted. In addition, the controller 112 may extract patient information related to pneumonia based on the presence or absence of a medical history of a specific disease within a predetermined period before the onset of pneumonia.

In an optional embodiment, the control unit 112 may extract patient information corresponding to any one of the above-described four criteria as patient information related to pneumonia, and only the patient information corresponding to all four criteria described above is pneumonia. It can also be extracted as patient information related to.

The controller 112 according to an embodiment of the present invention may generate learning data based on information on a plurality of patients extracted according to the above-described criteria. At this time, the controller 112 according to an embodiment of the present invention may generate learning data including appropriate environmental information by referring to the personal information of the patient included in each patient information.

For example, the control unit 112 may generate learning data by identifying a patient's residence area and a time when pneumonia occurs in the corresponding patient, and mapping environmental information of the corresponding region and period with the patient's degree.

As described above, the present invention can generate appropriate learning data for learning a neural network, especially for learning a neural network that predicts the probability of pneumonia.

4 is a flowchart illustrating a method of determining the probability of occurrence of pneumonia performed by the apparatus 110 for determining the probability of occurrence of pneumonia according to an embodiment of the present invention. Hereinafter, descriptions of contents overlapping with those described in FIGS. 1 to 3 will be omitted, but will be described with reference to FIGS. 1 to 3.

The apparatus 110 for determining the occurrence probability of pneumonia according to an embodiment of the present invention may learn a pneumonia occurrence probability calculator based on a plurality of learning data (S41).

In the present invention, the'pneumonia incidence probability calculator' may include an artificial neural network that is learned by machine learning or deep learning. In this case, the artificial neural network may refer to a model in which artificial neurons, which form a network by combining synapses, change the strength of synapses through learning to learn problem solving ability.

Referring back to FIG. 3, the neural network according to an embodiment of the present invention includes an input layer L1 including at least one input node N1, a hidden layer L2 including a plurality of hidden nodes N2, and An output layer L3 including at least one output node N3 may be included. At this time

Values corresponding to patient and environmental information acquired by the apparatus 110 for determining an onset of pneumonia may be input to at least one input node N1 of the input layer L1. For example, a value corresponding to the patient's personal information may be input to the first input node, and a value input to the patient's vital sign may be input to the second input node.

As illustrated, the hidden layer L2 may include one or more layers that are fully connected. When the hidden layer L2 includes a plurality of layers, the neural network may include a function (not shown) defining a relationship between each hidden layer.

At least one output node N3 of the output layer L3 may include an output value generated by a neural network from an input value of the input layer L1 under control of the apparatus 110 for determining the probability of pneumonia. For example, the first output node may include a value corresponding to the probability of developing pneumonia according to the passage of time of the patient.

In this case, a value included in each node of each layer may be a vector. In addition, each node may include a weight corresponding to the importance of the node.

Meanwhile, as shown in FIG. 3, the neural network defines the relationship between the first function F1 and the hidden layer L2 and the output layer L3, which defines the relationship between the input layer L1 and the hidden layer L2. It may include a second function F2. The first function F1 may define a connection relationship between the input node N1 included in the input layer L1 and the hidden node N2 included in the hidden layer L2. Similarly, the second function F2 may define a connection relationship between the hidden node N2 included in the hidden layer L2 and the output node N2 included in the output layer L2.

The first function F1, the second function F2, and the functions between the hidden layer may include a recurrent neural network (RNN) model that outputs a result based on an input of a previous node.

While the neural network is trained by the apparatus 110 for determining the probability of occurrence of pneumonia, the first function F1 and the second function F2 may be learned based on a plurality of training data. Of course, in the course of learning a neural network, functions between a plurality of hidden layers in addition to the above-described first function (F1) and second function (F2) can also be learned, and furthermore, each layer (L1, L2, L3) Weights of the plurality of nodes N1, N2, and N3 may also be learned.

In the present invention, it is'learned' that the apparatus 110 for determining the probability of occurrence of pneumonia is based on a plurality of learning data, and the above-described functions (F1, F2, functions between hidden layers, etc.) and nodes N1, N2 , N3) may mean updating the weights and storing them in the memory 113 or the like. For example, the apparatus 110 for determining the probability of occurrence of pneumonia may appropriately adjust coefficients of elements constituting the first function based on a plurality of learning data, and store the coefficients in the memory 113. Of course, similarly to this, the apparatus 110 for determining the probability of occurrence of pneumonia may appropriately adjust the coefficients of elements constituting the second function and/or weights of each node, and store the coefficients in the memory 113. However, this is merely an example and the spirit of the present invention is not limited thereto.

The neural network according to an embodiment of the present invention may be learned in a supervised learning method based on labeled learning data. In this case, the learning data may be data in which the probability of onset of pneumonia (or fact of onset) of a corresponding patient is marked in information on a plurality of items about a patient and information on the environment. At this time, the'probability of pneumonia' is a fixed value (e.g., 70%, etc.) regardless of the passage of time as described above, or a value that changes with the passage of time (e.g., 18% after 1 week, 2 weeks). It may be 25% after 3 weeks, 11% after 3 weeks, etc.). 'Onset fact' may be a value indicating whether pneumonia has occurred.

The apparatus 110 for determining the probability of occurrence of pneumonia according to an embodiment of the present invention uses a plurality of training data to input any one of the training data into a neural network, and the output value generated is the probability of developing pneumonia marked in the corresponding training data. The neural network may be trained by repeatedly performing the process of updating the above-described functions (F1, F2, functions between hidden layers, etc.) and/or the weights of the nodes N1, N2, N3 so as to be close to. At this time, the apparatus 110 for determining the probability of occurrence of pneumonia according to an embodiment of the present invention includes the above-described functions (F1, F2, functions between hidden layers, etc.) and/or nodes according to a back propagation algorithm. Weights of (N1, N2, N3) can be updated. However, this is merely an example and the spirit of the present invention is not limited thereto.

Of course, the neural network according to an embodiment of the present invention may be trained in an unsupervised learning method based on unmarked learning data. However, this is merely an example and the spirit of the present invention is not limited thereto.

Meanwhile, the apparatus 110 for determining the probability of occurrence of pneumonia according to an embodiment of the present invention may generate learning data from patient information and/or environment information that has not been processed for the above-described learning.

First, the apparatus 110 for determining the probability of occurrence of pneumonia according to an embodiment of the present invention may receive a plurality of patient information and/or environment information from the user terminal 200 and/or the external device 300. For example, the apparatus 110 for determining the probability of occurrence of pneumonia may receive information on a plurality of patients from a server (not shown) of a medical institution. At this time, the received plurality of patient information may or may not be related to pneumonia. In other words, the plurality of patients may be related to pneumonia, or may be related to other diseases.

In addition, the apparatus 110 for determining the probability of occurrence of pneumonia may receive environmental information of a plurality of times in a plurality of regions from a weather information providing server (not shown). At this time, the environmental information may also include environmental information of a region and/or a time point related to the patient, or may include environmental information of a region and/or a time point not related to the patient. Accordingly, in order to use the received information for learning in a neural network, it may need to be properly processed.

The apparatus 110 for determining the probability of occurrence of pneumonia according to an embodiment of the present invention may generate learning data by extracting patient information related to pneumonia from information on a plurality of patients. At this time, the device 110 for determining the probability of pneumonia may extract patient information related to pneumonia based on the presence or absence of a diagnosis code for pneumonia occurrence (for example, ICD10 standards, J100/J110/J12~18/J851, etc.) in the patient's medical information. I can. In addition, the apparatus 110 for determining the probability of occurrence of pneumonia may extract patient information related to pneumonia based on whether the medical information of the patient includes a medical image of the chest. In addition, the device 110 for determining the probability of pneumonia incidence is related to pneumonia in the patient's medical information, or whether or not keywords that suggest pneumonia (Pneumonia, Pneumonic, Consolidation, Infiltration, Ground glass opacity (opacities) = GGO, Parapneumonic, etc.) are included. Patient information related to pneumonia can be extracted based on. In addition, the apparatus 110 for determining the probability of occurrence of pneumonia may extract patient information related to pneumonia based on the presence or absence of treatment within a predetermined period before the onset of pneumonia.

In an optional embodiment, the apparatus 110 for determining the probability of occurrence of pneumonia may extract information of a patient corresponding to any one of the above four criteria as patient information related to pneumonia, or corresponding to all of the above four criteria. Only patient information can be extracted as patient information related to pneumonia.

The apparatus 110 for determining the probability of occurrence of pneumonia according to an embodiment of the present invention may generate learning data based on information on a plurality of patients extracted according to the above-described criteria. At this time, the apparatus 110 for determining the probability of occurrence of pneumonia according to an embodiment of the present invention may generate learning data including appropriate environmental information by referring to personal information of a patient included in each patient information.

For example, the apparatus 110 for determining the probability of pneumonia incidence checks the residential area of a patient and the time when pneumonia occurs, and maps the corresponding region and period environmental information with the patient's degree of learning data. Can be generated.

As described above, the present invention can generate appropriate learning data for learning a neural network, especially for learning a neural network that predicts the probability of pneumonia.

The apparatus 110 for determining the probability of occurrence of pneumonia according to the present invention may determine the probability of occurrence of pneumonia with higher accuracy by learning a neural network based on the learning data generated by the above-described process.

Subsequently, the apparatus 110 for determining the probability of occurrence of pneumonia according to an embodiment of the present invention acquires patient and environmental information (S42), and uses it to obtain at least one input node N1 included in the above-described input layer L1. A node value of) can be determined (S43)

For convenience of explanation, the neural network provides a first input node corresponding to the patient's personal information, a second input node corresponding to the patient's vital sign information, and a third input node corresponding to the patient's medical information in the input layer L1. Suppose to include.

In this case, the apparatus 110 for determining the probability of occurrence of pneumonia according to an embodiment of the present invention may determine a node value of each of the first to third input nodes from the obtained information. At this time, the apparatus 110 for determining the probability of occurrence of pneumonia according to an embodiment of the present invention may extract feature values for each item by applying a convolutional neural network (CNN) to information of each item, and use the extracted feature values as vectors. It can be in the form of.

In other words, the apparatus 110 for determining the probability of pneumonia incidence according to an embodiment of the present invention applies a convolutional neural network to the patient's personal information, and applies a first vector corresponding to the patient's personal information to the node of the first input node. Can be determined by value. Similarly, the apparatus 110 for determining the probability of occurrence of pneumonia may determine the node value of the second input node from the patient's vital sign information and the node value of the third input node from the patient's medical information.

The input layer L1 may include a fourth input node corresponding to information on the environment in addition to the information on the patient described above. In this case, the fourth input node may be a node having a node value corresponding to the concentration of fine dust. Of course, the node value of the fourth input node may be generated by the apparatus 110 for determining the probability of occurrence of pneumonia by applying a convolutional neural network to information about the environment.

As a method of determining the node values of the first to fourth input nodes of the above-described input layer, it has been exemplarily described that the apparatus 110 for determining the probability of occurrence of pneumonia applies a convolutional neural network to information of each item. It is not necessary to use a convolutional neural network like this. Therefore, various algorithms capable of deriving the feature values of each item according to the characteristics of the data and/or the design of the system may be used.

Meanwhile, node values of the first to fourth input nodes may be n-dimensional, m-dimensional, i-dimensional, and j-dimensional vectors, respectively.

Meanwhile, the apparatus 110 for determining the probability of occurrence of pneumonia may be obtained by receiving the above-described patient and environment information from the user terminal 200 and/or the external device 300, or read from the memory 113. Patient and environmental information can also be obtained in a way.

The apparatus 110 for determining the probability of occurrence of pneumonia according to an embodiment of the present invention includes a node of at least one output node included in the output layer from the pneumonia occurrence probability calculator based on the node values of the input nodes determined by the above-described process. The value can be determined (S44)

In other words, the apparatus 110 for determining the probability of occurrence of pneumonia according to an embodiment of the present invention may determine the node value of the output node output by the neural network when the node values of the input nodes are input.

Looking at this process in detail, the apparatus 110 for determining the probability of pneumonia incidence according to an embodiment of the present invention inputs and outputs the values of input nodes to the first function F1, thereby calculating the node values of the nodes of the first hidden layer. You can decide. Similarly, the apparatus 110 for determining the probability of occurrence of pneumonia may determine the node values of the nodes of the second hidden layer by inputting and outputting the node values of the nodes of the first hidden layer to a function defining the relationship between the hidden layers. In this case, the above-described'function defining a relationship between hidden layers' may be a function defining a relationship between a first hidden layer and a second hidden layer. Also, in the same way, the apparatus 110 for determining the probability of occurrence of pneumonia may determine the node values of the output nodes by inputting and outputting the node values of the nodes of the last hidden layer to the second function.

In this way, the apparatus 110 for determining the probability of occurrence of pneumonia may determine the output of the neural network corresponding to the node values of the input nodes, that is, the node value of the output node.

The apparatus 110 for determining the probability of onset of pneumonia according to an embodiment of the present invention may determine the probability of onset of pneumonia of the patient based on the node value of the above-described output node. (S45) At this time, according to an embodiment of the present invention The output layer L3 may include a first output node corresponding to a patient's probability of developing pneumonia. The node value of the first output node may be in the form of a vector such as [0, 1, 0, 0, 0, 0, 0, 0, 0, 0], and each component of the vector corresponds to a probability. I can. For example, each of the 10 components described above may correspond to a probability of developing pneumonia of 100%, 90%, 80% 70%, 60%, 50%, 40%, 30%, 20%, and 10%.

Meanwhile, the neural network according to an embodiment of the present invention may be learned by further including a probability of death of a patient according to time and a probability of occurrence of a first disease of the patient.

For example, the apparatus 110 for determining the probability of occurrence of pneumonia according to an embodiment of the present invention may learn the probability of death of the patient according to time and the probability of occurrence of the patient's first disease in a supervised learning method. In this case, in addition to the probability of developing pneumonia, the probability of death and the probability of developing the patient's first disease may be further marked in the learning data. Of course, the apparatus 110 for determining the probability of occurrence of pneumonia may learn the probability of death of the patient over time and the probability of occurrence of the patient's first disease in an unsupervised learning method. However, this is merely an example and the spirit of the present invention is not limited thereto.

The output layer L3 according to an embodiment of the present invention, together with the above-described first output node, includes a second output node corresponding to the patient's probability of death over time, and a second output node corresponding to the patient's first disease occurrence probability. It may further include 3 output nodes.

Meanwhile, such a first disease is a disease related to pneumonia, and may be ascites. When there are a plurality of first diseases, the output layer L3 may include a third output node for each of the plurality of first diseases.

In this way, the present invention can reasonably infer the probability of onset of a specific disease, particularly pneumonia, from the patient's information, and also can infer the probability of death of the patient.

5 is an example of a screen 500 displayed on the user terminal 200 according to an embodiment of the present invention.

Hereinafter, for convenience of explanation, the user of the user terminal 200 transmits a plurality of items of information on a patient (a patient to be diagnosed) who wants to know the probability of developing pneumonia to the apparatus 110 for determining the probability of developing pneumonia, and It is assumed that the result is received from the apparatus 110 for determining the probability of occurrence of and displayed as shown on the screen 500.

Referring to FIG. 5, a screen 500 includes an area 510 displaying patient information, an area 520 displaying environmental information, an area 530 displaying a probability of pneumonia, and an area displaying a related probability. It may include 540.

In the area 510 in which patient information is displayed, personal information of the patient, vital sign information, and medical information may be displayed. In other words, the apparatus 110 for determining the probability of occurrence of pneumonia according to an embodiment of the present invention may provide the user terminal 200 with main information about a patient.

In the area 520 in which the environmental information is displayed, information on the level of fine dust at a region and time point that is expected to be related to the onset of pneumonia of the patient may be displayed. In addition, in the area 520 where environmental information is displayed, the critical fine dust value 521, which is expected to have a significant correlation with the invention of pneumonia, and the critical fine dust value 521, are expected to contribute to the onset of pneumonia. A time period 522 may be displayed. Accordingly, the apparatus 110 for determining the probability of pneumonia incidence according to an embodiment of the present invention extracts numerical information on the area and time point that is expected to be related to the patient's pneumonia symptom from environmental information, ) Can be provided. For example, the apparatus 110 for determining the probability of pneumonia incidence according to an embodiment of the present invention may provide numerical information of fine dust at a predetermined time in a predetermined area in consideration of a patient's residence area and hospitalization time.

In the region 530 in which the probability of occurrence of pneumonia is displayed, a degree of the probability of occurrence of pneumonia over time may be displayed. In this case, a risk section 531 may be further displayed in the area 530 in which the probability of developing pneumonia is displayed. The probability of occurrence of pneumonia displayed in the region 530 in which the probability of occurrence of pneumonia is displayed may be based on the node value of the above-described first output node.

In the area 540 in which the related probability is displayed, the probability of death of the patient and the occurrence probability of the first disease over time may be further displayed. In this case, the probability of death of the patient and the probability of occurrence of the first disease may also be based on a node value of an output node corresponding to each included in the output layer L3.

The embodiment according to the present invention described above may be implemented in the form of a computer program that can be executed through various components on a computer, and such a computer program may be recorded in a computer-readable medium. In this case, the medium may store a program executable by a computer. Examples of media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical recording media such as CD-ROMs and DVDs, magnetic-optical media such as floptical disks, and And ROM, RAM, flash memory, and the like, and may be configured to store program instructions.

Meanwhile, the computer program may be specially designed and configured for the present invention, or may be known and usable to those skilled in the computer software field. Examples of the computer program may include not only machine language codes produced by a compiler but also high-level language codes that can be executed by a computer using an interpreter or the like.

The specific implementations described in the present invention are examples, and do not limit the scope of the present invention in any way. For brevity of the specification, descriptions of conventional electronic configurations, control systems, software, and other functional aspects of the systems may be omitted. In addition, the connection or connection members of the lines between the components shown in the drawings exemplarily represent functional connections and/or physical or circuit connections, and in an actual device, various functional connections that can be replaced or additionally It may be referred to as a connection, or circuit connections. In addition, if there is no specific mention such as “essential” or “importantly”, it may not be an essential component for the application of the present invention.

Accordingly, the spirit of the present invention is limited to the above-described embodiments and should not be defined, and all ranges equivalent to or equivalently changed from the claims to be described later as well as the claims to be described later are the scope of the spirit of the present invention. It will be said to belong to.

100: server
110: device for determining the probability of developing pneumonia
111: communication department
112: control unit
113: memory
200: user terminal
300: external device
400: communication network

Claims (6)

In the method of determining the probability of onset of pneumonia using a device for determining the onset of pneumonia,
Learning a pneumonia incidence probability calculator based on the learning data;
Acquiring information on a plurality of items and environment information on a patient to be diagnosed;
Determining a node value of at least one input node included in an input layer by using information on a plurality of items and an environment on the patient to be diagnosed;
Determining a node value of at least one output node included in an output layer from a learned pneumonia incidence probability calculator based on the node value of the at least one input node; And
Determining a probability of developing pneumonia of the patient to be diagnosed based on the node value of the at least one output node; and
The pneumonia incidence probability calculator includes information on the plurality of items based on learning data including information on a plurality of items for each of a plurality of patients, information on a plurality of environments, and a probability of onset of pneumonia for each of the plurality of patients. And a neural network learned about a correlation between information on a plurality of environments and a probability of onset of pneumonia,
The learning step,
Classifying patient information including a pneumonia occurrence diagnosis code as first valid learning data in a plurality of items of information about a patient received from an external device;
Classifying information of a patient, including a medical image related to pneumonia, as second effective learning data, in a plurality of items of information about a patient received from the external device;
Classifying information of a patient including a keyword suggesting pneumonia in a plurality of items of information about a patient received from the external device as third effective learning data;
Classifying information of a patient including a medical history of a specific disease within a predetermined period before the onset of pneumonia in a plurality of items of information about a patient received from the external device as fourth effective learning data; And
Determining at least some of the first valid learning data, the second valid learning data, the third valid learning data, and the fourth valid learning data as the learning data; including, a method for determining the probability of occurrence of pneumonia. The method of claim 1
Multiple items of information on the patient
Including at least one of the patient's personal information, the patient's vital sign information, and the patient's medical information,
The patient's medical information is
A method for determining a probability of onset of pneumonia comprising at least one of the patient's past medical history, the patient's medical image, the patient's drug taking history, and the patient's symptom history. According to claim 2
The input layer is
And a first input node to which the patient's personal information is input, a second input node to which the patient's vital sign information is input, and a third input node to which the patient's medical information is input. delete delete The method of claim 1
The step of determining with the training data is
The first effective learning data, the second valid learning data, the third valid learning data, and learning data belonging to all of the fourth valid learning data is determined as the learning data.

Images