WO2024080450A1 - Method and device for assessing mental health disorder by using behavioral pattern and medical images - Google Patents

Abstract

The present disclosure relates to a method and device for assessing a mental health disorder by using a behavioral pattern and medical images. The method according to an embodiment of the present disclosure may comprise the steps of: inputting at least one brain image of different modalities of a user into at least one pre-trained first analysis model to assess a first risk level of at least one mental health disorder; collecting, through a user terminal, behavioral data, which is a plurality of pieces of time-series data about behavioral patterns of the user, related to the at least one mental health disorder; preprocessing, on the basis of the first risk level, the behavioral data by applying different weights to at least a part of the behavioral data; and inputting the behavioral data into a pre-trained second analysis model to assess a second risk level of the at least one mental health disorder.

Description

Method and device for assessing mental health disorders using behavioral patterns and medical images

The technical idea of the present disclosure relates to a method and device for assessing mental health disorders using behavioral patterns and medical images.

Previously, in order to determine mental health disorders, the patient had relied on tests based on questionnaire answers to describe the patient's current emotional state and the surrounding circumstances that would cause the emotion. Recently, computer-based machine learning and user interfaces have been used. Therefore, rather than simply adding up questionnaire scores, a technology (e.g., KR 10-2022-0099190) is being developed that prepares a wider variety of questions and provides preliminary diagnosis of mental illness information based on trends in the answers.

However, both traditional questionnaires and machine learning survey methods not only have the potential to be subjective in the patient's answers, but also because the patient's condition changes frequently, the results of a single test are not accurate in determining the direction of diagnosis and treatment. It may not be possible. In addition, the pathogenesis of the disease in different patients with the same symptoms may be different, and treatment methods (type of counseling treatment, type of drug and dosage) may vary accordingly. For example, compared to depression caused by psychotic disorders (bipolar disorder, etc.), there is a clear difference in treatment methods for depression caused by seasonal affective disorder, and it is very difficult to accurately classify and diagnose it using questionnaire methods.

When it comes to pre-diagnosing mental illness through a smartphone app, a theoretically sound method is to analyze text messages sent and received by smartphone users to people around them or content posted on SNS through a process of natural language processing (NLP) to identify specific words. There is a way to analyze the frequency with which sentences are detected. However, this method has the serious problem of exposing personal information.

Meanwhile, technology to visualize the form and function of the brain is gradually being developed. Psychiatric diseases caused by congenital or metabolic abnormalities also affect the form and function of the brain, which can be diagnosed by visualizing the form and function of the brain through medical imaging. However, this is a burden to patients due to the large amount of time and cost, and because patients have a lot of anxiety and discomfort during the test itself, frequent image acquisition is avoided. Therefore, the time interval for acquiring images is inevitably long, and the patient's healing progress cannot be frequently checked.

The technical idea of the present disclosure was devised to solve the above problems, and its purpose is to provide a method and device for evaluating mental health disorders using behavioral patterns and medical images.

The technical problems to be achieved by the method and device according to the technical idea of the present disclosure are not limited to the problems mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the description below.

According to an aspect according to the technical idea of the present disclosure, a method for assessing mental health disorders using behavior patterns and medical images includes inputting at least one brain image of different modalities of a user into at least one pre-trained first analysis model. , assessing a first risk of at least one mental health disorder; Collecting behavioral data, which is a plurality of time-series data about the user's behavioral patterns related to at least one mental health disease, through a user terminal; preprocessing the behavioral data by applying different weights to at least a portion of the behavioral data based on the first risk; and inputting the behavioral data into a pre-trained second analysis model to evaluate a second risk of at least one mental health disorder.

According to an exemplary embodiment, the second analysis model includes a clustering model, and the step of evaluating the second risk of mental health disease includes inputting the behavioral data into the pre-trained clustering model, classifying into at least one cluster associated with at least one said mental health disorder; and evaluating a second risk of at least one mental health disorder based on the cluster.

According to an exemplary embodiment, the second analysis model further includes a classification model, and the step of evaluating the second risk of mental health disease includes inputting the behavioral data into the pre-trained classification model; , may be performed by evaluating the second risk of at least one mental health disease based on the output value of the classification model.

According to an exemplary embodiment, one of the clustering model and the classification model is selected as an application model based on at least one of user input, the type of behavioral data, the amount of behavioral data, and the number of brain images of different modalities. It further includes the step of evaluating the second risk of mental health disease, and may be performed through the selected application model.

According to an exemplary embodiment, the brain images include magnetic resonance imaging (MRI), functional magnetic resonance imaging (fMRI), positron emission tomography (PET-CT) imaging, diffusion-weighted imaging (DWI), and diffusion tensor imaging (DTI). ) may include at least one of

According to an exemplary embodiment, the behavioral data includes data on at least one of heart rate, application switching frequency, repeated use frequency of the same application, movement pattern, and sleep pattern acquired through the user terminal over a predetermined period of time. can do.

According to an exemplary embodiment, the preprocessing of the behavioral data includes applying a predetermined weight to at least some of the behavioral data of a type related to at least one of the mental health diseases according to the ranking of the first risk. , can be performed.

According to an exemplary embodiment, the step of collecting the behavioral data may be repeatedly performed at a predetermined period of time.

According to one aspect according to the technical idea of the present disclosure, an apparatus for evaluating mental health disorders using behavior patterns and medical images includes at least one processor; a memory storing a program executable by the processor; and the processor, by executing the program, inputs at least one brain image of different modalities of the user into at least one first pre-trained analysis model to evaluate a first risk of at least one mental health disorder, Collecting behavioral data, which is a plurality of time-series data about the user's behavioral patterns related to at least one mental health disease, through a user terminal, and applying different weights to at least a portion of the behavioral data based on the first risk level. By doing so, the behavioral data can be pre-processed and the behavioral data can be input into a pre-trained second analysis model to evaluate the second risk of at least one mental health disease.

According to the method and device for evaluating mental health disorders using behavioral patterns and medical images according to embodiments of the technical idea of the present disclosure, the risk of mental health disorders can be more accurately determined by comprehensively analyzing the user's behavioral patterns and medical images. You can evaluate and monitor the progress of the disease.

The effects that can be obtained by the method and device according to the technical idea of the present disclosure are not limited to the effects mentioned above, and other effects not mentioned can be obtained by those skilled in the art from the description below. can be clearly understood.

In order to more fully understand the drawings cited in this disclosure, a brief description of each drawing is provided.

1 is a diagram illustrating a mental health disease evaluation system using behavior patterns and medical images according to an embodiment of the present disclosure.

FIG. 2 is a flowchart illustrating a method for evaluating mental health disorders using behavior patterns and medical images according to an embodiment of the present disclosure.

FIG. 3 is a flowchart illustrating a method for assessing mental health disorders using behavioral patterns and medical images according to an embodiment of the present disclosure.

Figure 4 is a flowchart for explaining an embodiment of step S350 of Figure 3.

FIG. 5 is a flowchart illustrating a method for evaluating mental health disorders using behavioral patterns and medical images according to an embodiment of the present disclosure.

FIG. 6 is a flowchart illustrating a method for evaluating mental health disorders using behavioral patterns and medical images according to an embodiment of the present disclosure.

Figure 7 is a diagram for explaining a first analysis model according to an embodiment of the present disclosure.

Figure 8 is a diagram for explaining a second analysis model according to an embodiment of the present disclosure.

FIG. 9 is a block diagram briefly illustrating the configuration of a mental health disease evaluation device using behavior patterns and medical images according to an embodiment of the present disclosure.

Since the technical idea of the present disclosure can be subject to various changes and can have several embodiments, specific embodiments will be illustrated in the drawings and described in detail. However, this is not intended to limit the technical idea of the present disclosure to specific embodiments, and should be understood to include all changes, equivalents, and substitutes included in the scope of the technical idea of the present disclosure.

In explaining the technical idea of the present disclosure, if it is determined that a detailed description of related known technology may unnecessarily obscure the gist of the present disclosure, the detailed description will be omitted. Additionally, numbers (eg, first, second, etc.) used in the description of the present disclosure are merely identifiers to distinguish one component from another component.

In addition, in the present disclosure, when a component is referred to as "connected" or "connected" to another component, the component may be directly connected or directly connected to the other component, but specifically Unless there is a contrary description, it should be understood that it may be connected or connected through another component in the middle.

In addition, terms such as “unit”, “unit”, “character”, and “module” described in the present disclosure refer to a unit that processes at least one function or operation, which refers to a processor, micro Processor (Micro Processer), Micro Controller, CPU (Central Processing Unit), GPU (Graphics Processing Unit), APU (Accelerate Processor Unit), DSP (Digital Signal Processor), ASIC (Application Specific Integrated Circuit), FPGA It can be implemented by hardware or software such as (Field Programmable Gate Array) or a combination of hardware and software.

In addition, we would like to clarify that the division of components in this disclosure is merely a division according to the main function each component is responsible for. That is, two or more components, which will be described below, may be combined into one component, or one component may be divided into two or more components for more detailed functions. In addition to the main functions that each component is responsible for, each of the components described below may additionally perform some or all of the functions that other components are responsible for, and some of the main functions that each component is responsible for may be performed by other components. Of course, it can also be carried out exclusively by .

The method according to an embodiment of the present disclosure may be performed on a personal computer, work station, or server computer device equipped with computing power, or may be performed on a separate device for this purpose.

Additionally, the method may be performed on one or more computing devices. For example, at least one step of the method according to an embodiment of the present disclosure may be performed in a client device, and other steps may be performed in a server device. In this case, the client device and the server device are connected through a network and can transmit and receive calculation results. Alternatively, the method may be performed by distributed computing techniques.

Additionally, throughout this specification, network function may be used synonymously with neural network and/or neural network. A neural network may generally consist of a set of interconnected computational units, which may be referred to as nodes, and these nodes may be referred to as neurons. A neural network is generally composed of a plurality of nodes, and the nodes constituting the neural network may be interconnected by one or more links. At this time, some of the nodes constituting the neural network may form one layer based on the distances from the first input node. For example, a set of nodes with a distance n from the initial input node may constitute n layers. The neural network may include a deep neural network (DNN) that includes multiple hidden layers in addition to the input layer and output layer.

Hereinafter, embodiments of the present disclosure will be described in detail one by one.

1 is a diagram illustrating a mental health disease evaluation system using behavior patterns and medical images according to an embodiment of the present disclosure.

Referring to FIG. 1, a system according to an embodiment may operate including a user terminal 10,...,N, a device 20, a database server 30, and a network 40.

The user terminal 10,...,N may include any device that can access the network 40. For example, the user terminal 10,...,N may include a smartphone, tablet, PC, laptop, home appliance device, medical device, camera, wearable device, etc. In an embodiment, the user terminal 10,...,N may collect learning data, which is time-series data about a user's behavior pattern related to at least one mental illness, over a certain period of time and transmit it to the device 20.

In the embodiment, the user terminal 10,...,N provides a predetermined user interface, receives data through user input from the user, and transmits it to the device 20 and/or the database server ( 30).

The device 20 provides computing resources, storage resources, etc. to provide risk assessment and follow-up of mental health disorders to clients through the network 40. Here, the client may include a user terminal (10,...,N). In embodiments, device 20 may include various types of servers, such as application servers, control servers, data storage servers, and servers for providing specific functions. Additionally, the device 20 may process the process alone, or multiple devices may process the process together.

In an embodiment, the device 20 determines the risk of mental health disorders through at least one analysis model based on the user's behavior data received from the user terminal 10 and the user's brain image received from the database server 30. can be evaluated. Information regarding risk may be provided to the user terminal 10 or a medical institution terminal.

In an embodiment, behavioral data may be periodically updated and collected at predetermined intervals, and the device 20 may track the risk of mental health disease based on the analysis model and report this to the user terminal 10 or a medical institution. It can be provided to the terminal.

The database server 30 stores data necessary for system operation. In an embodiment, the database server 30 may be operated and managed directly or indirectly by at least one medical institution (e.g., hospital, etc.) for data maintenance and repair, and may be operated by a user with certain qualifications or It may be configured to deliver the data to business operators, etc. Information stored in the database server 30 can be delivered at the request of the device 20 and used in the service provision process.

The network 40 is connected to terminals 10,...,N, such as the Internet, intranet, extranet, LAN (Local Area Network), MAN (Metropolitan Area Network), and WAN (Wide Area Network). , may include all networks that the device 20 and the database server 30 can access.

FIG. 2 is a flowchart illustrating a method for evaluating mental health disorders using behavior patterns and medical images according to an embodiment of the present disclosure.

In step S210, the device 20 may input at least one brain image of different modalities of the user into at least one pre-trained first analysis model to evaluate the first risk of at least one mental health disease.

For example, the device 20 connects to an external (medical institution, etc.) database server 30 to obtain a brain image corresponding to the user, or searches for the user's brain image among a plurality of brain images stored in its own database. can do. In embodiments, the brain imaging is at least one of magnetic resonance imaging (MRI), functional magnetic resonance imaging (fMRI), positron emission tomography (PET-CT) imaging, diffusion-weighted imaging (DWI), and diffusion tensor imaging (DTI). It may include, but is not limited to this.

The device 20 extracts features causing mental health diseases (i.e., features corresponding to dementia, stress, degenerative depression, anxiety disorders, etc.) from brain images of different modalities through the first analysis model, and based on this, The risk (or severity) of health diseases can be assessed.

The first analysis models may each include at least one network function, extracting features corresponding to a mental health disorder through learning data (e.g., brain images of patients known to have a mental health disorder, etc.), It can be trained in advance to calculate the risk of mental health disorders.

In embodiments, the mental health condition may include at least one of depression, anxiety disorder, Attention-Deficit/Hyperactivity Disorder (ADHD), Post Traumatic Stress Disorder (PTSD), and schizophrenia. You can.

In step S220, the device 20 may collect behavioral data, which is time-series data about the user's behavior pattern related to at least one mental health disease, through the user terminal 10.

For example, behavioral data regarding the user's behavior patterns for 2 to 4 weeks is collected from the user terminal 10, such as a smartphone and/or a wearable device, and the collected behavioral data may be transmitted to the device 20. there is.

In an embodiment, the behavioral data may include data on at least one of heart rate, application switching frequency, repeated use frequency of the same application, movement pattern, and sleep pattern acquired through the user terminal 10 over a predetermined period of time. , but is not limited to this.

In step S230, the device 20 may preprocess the behavior data by applying different weights to at least a portion of the behavior data based on the first risk.

In embodiments, device 20 may adjust weights for types of behavioral data associated with high risk mental health disorders based on the first ranking of risk.

Meanwhile, although not shown, in an embodiment, the method 200 may further include the step of correcting behavioral data collected through the user terminal 10.

For example, when doing exercise such as jogging or tracking, the heart rate changes, but if multiple other applications are not used at the same time, the amount of movement may increase at the same time. Accordingly, the device 20 can exclude factors caused by exercise from heart rate data based on this. Additionally, for example, heart rate decreases during sleep, and this factor can also be ruled out through sleep pattern analysis data.

Additionally, in an embodiment, the method 200 may further include encoding behavioral data into a form corresponding to a second analysis model or performing preprocessing such as dimensionality reduction.

In step S240, the device 20 may input the preprocessed behavioral data into the pre-trained first analysis model to evaluate the second risk of at least one mental health disorder.

That is, the device 10 may extract at least one feature corresponding to a mental health disorder from behavioral data through the second analysis model and evaluate a second risk of at least one mental health disorder based on the feature.

At this time, the second analysis model may include at least one network function and is used to calculate the risk of mental health disease in advance through learning data (e.g., behavioral data of patients known to have mental health disease, etc.). It can be learned.

In an embodiment, the second analysis model may calculate the risk of mental health disease based further on the user's disease history information (age, gender, family history, comorbidities, etc.). Disease history information may be input through the user terminal 10 or obtained from the database server 30.

FIG. 3 is a flowchart illustrating a method for assessing mental health disorders using behavioral patterns and medical images according to an embodiment of the present disclosure.

Steps S310, S320, and steps S340 to S350 of the method 300 are similar to steps S210 to S240 described above with reference to FIG. 2, and the description will focus on the differences between the two embodiments.

In method 300, the second analysis model includes a classification model and a clustering model, and may include selecting an application model from among the plurality of second analysis models.

Specifically, in step S330, the device 20 collects user input, type of behavioral data (i.e., number of behavioral patterns subject to collection), amount of behavioral data (i.e., collection period of behavioral data), amount, and different Based on at least one of the number of brain images in the modality (i.e., the number of types of brain images), a classification model or a clustering model among the second analysis models may be selected as an application model.

For example, classification models may be more effective at making proactive predictions. It can be more useful in determining whether a mental health disorder has developed temporarily due to external or environmental factors, or whether long-term treatment is needed. However, since relatively more data must be predicted for many target diseases, there may be a disadvantage in that prediction takes a long time.

On the other hand, if the clustering model is trained with sufficient training data and the centroid for each cluster is sufficiently reliable, the clustering model may be able to predict the risk of mental health disorders more quickly than the classification model. However, since the clustering model may not be suitable for predicting multiple diseases simultaneously, if the user is aware of the disease to some extent due to history, genetic factors, medical factors, etc., or based on brain imaging Therefore, when a high-risk disease is identified, it can be effective to analyze the current situation or observe treatment progress using a clustering model.

Once the application model is selected, step S350 may be performed based on the analysis model selected as the application model.

In an embodiment, when a classification model is selected as an application model, behavioral data may be input into a pre-trained classification model, and the risk of at least one mental health disorder may be evaluated based on the output value of the classification model.

In an embodiment, when the clustering model is selected as the application model, as shown in FIG. 4, in step S351, the behavioral data is input into the pre-trained clustering model to determine the characteristics of the behavioral data with at least one mental health disorder. The cluster may be classified into at least one related cluster, and then, in step S352, the risk of mental health disease may be assessed based on the classified cluster. In this case, the risk of a specific mental health disease can be calculated based on the distance from the center of the cluster.

The clustering model may, for example, apply the Louvain algorithm, but is not limited thereto, and includes at least one Euclidean distance, Manhattan distance, etc. in a low-dimensional representation method using Jaccard similarity, correlation, UMAP, and t-SNE. Various types of algorithms applied above can be applied.

FIG. 5 is a flowchart illustrating a method for evaluating mental health disorders using behavioral patterns and medical images according to an embodiment of the present disclosure.

Steps S510, S520, and S540 of the method 500 are similar to steps S210, S220, and S240 described above with reference to FIG. 2, and the description will focus on the differences between the two embodiments.

In step S530, the device 20 may adjust the second analysis model based on the first risk assessed through brain imaging.

That is, based on the risk of mental health disease based on brain imaging, the second analysis model is adjusted to be more suitable for the user, and as described in detail below with reference to FIG. 6, the user's mental health disease is followed up using this. can do.

In an embodiment, step S530 may be performed by adjusting the bias of the second analysis model for at least one mental health disorder. For example, the evaluation target of the first analysis model can be reduced to one or two mental health diseases evaluated as having a high first risk, and the degree of bias for each disease can be adjusted.

Behavioral data collected from the user terminal 10 may have limitations in accurately classifying user patterns due to severe or more severe mental illness and patterns due to temporary anxiety. Additionally, time series data must be collected over a significant period of time to draw reliable conclusions based on heart rate, sleep patterns, movement patterns, etc. In the present disclosure, by supplementing the second analysis model based on the analysis results of the brain image, the collection period of behavioral data can be reduced, the processing time can be shortened by reducing the amount of calculation required, and the evaluation accuracy using the second analysis model can be improved. can be further improved.

FIG. 6 is a flowchart illustrating a method for evaluating mental health disorders using behavioral patterns and medical images according to an embodiment of the present disclosure.

Method 600 may be performed to follow the course of a mental health disorder after method 200 of FIG. 2, method 300 of FIG. 3, and/or method 500 of FIG. 5.

In steps S610 and S620, the device 20 updates and collects the user's behavior data from the user terminal 10 at a predetermined period of time and inputs the periodically collected behavior data into the second analysis model to detect mental health disorders. The risk can be continuously assessed.

The risk assessment results may be provided to the user terminal 10 and/or the medical institution terminal. Through this, it is possible to accurately inform users, doctors, etc. of the treatment progress of mental health disorders or provide appropriate prescriptions.

That is, according to the present disclosure, whether a mental health disorder with a high risk or severity is worsening or whether the disorder is improving while undergoing treatment for the disorder (e.g., drug treatment, psychological treatment, etc.) is continuously monitored through behavioral data. Monitoring is possible.

In an embodiment, when a new brain image of the user is captured, the device 20 may re-evaluate the first risk of mental health disease through a first analysis model based on the brain image. In this case, the weight of the behavioral data can be adjusted or the second analysis model can be adjusted based on the re-evaluated second risk.

Meanwhile, although not shown, the method 600 may further include the step of preprocessing or correcting updated and collected behavioral data.

Figure 7 is a diagram for explaining a first analysis model according to an embodiment of the present disclosure.

Referring to Figure 7, the first analysis model extracts features related to mental health diseases (degenerative diseases such as dementia, vascular diseases, stress, anxiety disorders, etc.) from brain images of at least one different modality, and synthesizes them to The risk (or risk score) of mental health disorders can be assessed.

To this end, the first analysis model may be configured to include a plurality of analysis models and evaluation models, and each analysis model and evaluation model may be implemented through at least one network function.

For example, the first analysis model may be a structural analysis model for extracting features from magnetic resonance images, a functional analysis model for extracting features from functional magnetic resonance images, a biochemical analysis model for extracting features from positron emission tomography images, and a diffusion model. It may include a diffusion image analysis model that extracts features from the highlighted image and diffusion tensor image, and an evaluation model that synthesizes them to evaluate the risk of mental health disease.

Features extracted from brain images of different modalities output from the analysis model may be encoded in a certain format (for example, a feature vector of a certain dimension) and transmitted to the evaluation model.

In an embodiment, at least one of the plurality of analysis models included in the first analysis model is applied based on the risk assessment result of the first analysis model based on behavioral data and/or the type of user's brain image acquired through a database server. It can be configured to be selected as a model.

However, the configuration of this first analysis model is illustrative and may be modified in various ways depending on the embodiment.

Figure 8 is a diagram for explaining a second analysis model according to an embodiment of the present disclosure.

Referring to FIG. 8, the second analysis model receives behavioral data about the user's behavior patterns collected over a certain period of time through the user terminal 10, and calculates a risk (or risk score) of one or more mental health disorders from this. can be evaluated.

For example, in anxiety-type symptoms (anxiety disorder, ADHD, etc.), the amount and frequency of heart rate changes are very large, and in depressive-type symptoms (depression, PTSD, etc.), the heart rate is relatively constant and tends to be lower than the normal average. can be shown. Additionally, for example, in anxiety-type symptoms, smartphone users tend to habitually switch applications frequently until they feel comfortable, and in depressive-type symptoms, symptoms include being obsessed with social media or video media and not being able to escape easily. It can happen.

Additionally, for example, if anxiety symptoms are severe, there may be a pattern of continuously moving to the same place, and if depressive symptoms are severe, there may be a strong tendency to stay in one place. Additionally, for example, both depressed and anxious people have difficulty maintaining normal sleep patterns and may wake up frequently in the middle of the night and show a pattern of tossing and turning or moving around.

The second analysis model may be pre-trained to calculate the risk of mental health disease based on this behavioral data.

FIG. 9 is a block diagram briefly illustrating the configuration of a mental health disease evaluation device using behavior patterns and medical images according to an embodiment of the present disclosure.

The communication unit 910 may receive input data (behavioral data, brain imaging, etc.) for evaluating the risk of mental health disorders. The communication unit 910 may include a wired or wireless communication unit. When the communication unit 910 includes a wired communication unit, the communication unit 910 includes a local area network (LAN), a wide area network (WAN), a value added network (VAN), and a mobile communication network ( It may include one or more components that enable communication through a Mobile Radio Communication Network, a satellite communication network, and a combination thereof. In addition, when the communication unit 910 includes a wireless communication unit, the communication unit 910 can transmit and receive data or signals wirelessly using cellular communication, wireless LAN (e.g., Wi-Fi), etc. You can. In an embodiment, the communication unit may transmit and receive data or signals with an external device or external server under the control of the processor 940.

The input unit 920 can receive various user commands through external manipulation. To this end, the input unit 920 may include or connect one or more input devices. For example, the input unit 920 may be connected to various input interfaces such as a keypad and mouse to receive user commands. To this end, the input unit 920 may include not only a USB port but also an interface such as Thunderbolt. Additionally, the input unit 920 may include or be combined with various input devices such as a touch screen and buttons to receive external user commands.

The memory 930 may store programs and/or program instructions for operating the processor 940, and temporarily or permanently store input/output data. The memory 930 is a flash memory type, a hard disk type, a multimedia card micro type, a card type memory (for example, SD or XD memory, etc.), and RAM. , SRAM, ROM, EEPROM, PROM, magnetic memory, magnetic disk, and optical disk may include at least one type of storage medium.

In addition, the memory 930 can store various network functions and algorithms, and can store various data, programs (one or more instructions), applications, software, commands, codes, etc. for driving and controlling the device 900. there is.

The processor 940 may control the overall operation of the device 900. The processor 940 may execute one or more programs stored in the memory 930. The processor 940 may refer to a central processing unit (CPU), a graphics processing unit (GPU), or a dedicated processor that performs methods according to the technical idea of the present disclosure.

In an embodiment, the processor 940 inputs at least one brain image of different modalities of the user into at least one first pre-trained analysis model to evaluate the first risk of at least one mental health disorder, and outputs the user terminal By collecting behavioral data, which is a plurality of time-series data about the user's behavior pattern related to at least one mental health disease, and applying different weights to at least a portion of the behavioral data based on the first risk, The behavioral data may be pre-processed, and the behavioral data may be input into a pre-trained second analysis model to evaluate a second risk of at least one mental health disorder.

In an embodiment, the second analysis model includes a clustering model, and processor 940 inputs the behavioral data into the pre-trained clustering model to determine at least one condition associated with the at least one mental health condition. Classification into clusters may be performed, and a second risk of at least one mental health disease may be assessed based on the clusters.

In an embodiment, the second analysis model further includes a classification model, and the processor 940 inputs the behavioral data into the pre-trained classification model, and based on the output value of the classification model, at least one The second risk of the mental health disorder can be assessed.

In an embodiment, the processor 940 applies one of the clustering model and the classification model based on at least one of user input, the type of behavioral data, the amount of behavioral data, and the number of brain images of different modalities. , and the second risk of the mental health disease can be evaluated through the selected application model.

In an embodiment, the processor 940 may preprocess the behavioral data by applying a predetermined weight to at least some of the behavioral data of a type related to at least one of the mental health disorders according to the ranking of the first risk. You can.

In an embodiment, the processor 940 may repeatedly collect the behavioral data at predetermined intervals.

The method according to an embodiment of the present disclosure may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc., singly or in combination. Program instructions recorded on the medium may be those specifically designed and configured for this disclosure, or may be known and usable by those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic media such as floptical disks. -Includes optical media (magneto-optical media) and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, flash memory, etc. Examples of program instructions include machine language code, such as that produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc.

Additionally, the method according to the disclosed embodiments may be provided and included in a computer program product. Computer program products are commodities and can be traded between sellers and buyers.

A computer program product may include a S/W program and a computer-readable storage medium in which the S/W program is stored. For example, a computer program product may include a product in the form of a S/W program (e.g., a downloadable app) distributed electronically by the manufacturer of an electronic device or through an electronic marketplace (e.g., Google Play Store, App Store). there is. For electronic distribution, at least part of the S/W program may be stored in a storage medium or temporarily created. In this case, the storage medium may be a manufacturer's server, an electronic market server, or a relay server's storage medium that temporarily stores the SW program.

A computer program product, in a system comprised of a server and a client device, may include a storage medium of a server or a storage medium of a client device. Alternatively, if there is a third device (eg, a smartphone) in communication connection with the server or client device, the computer program product may include a storage medium of the third device. Alternatively, the computer program product may include the S/W program itself, which is transmitted from a server to a client device or a third device, or from a third device to a client device.

In this case, one of the server, the client device, and the third device may execute the computer program product to perform the method according to the disclosed embodiments. Alternatively, two or more of a server, a client device, and a third device may execute the computer program product and perform the methods according to the disclosed embodiments in a distributed manner.

For example, a server (eg, a cloud server or an artificial intelligence server, etc.) may execute a computer program product stored on the server and control a client device connected to the server to perform the method according to the disclosed embodiments.

Although the embodiments have been described in detail above, the scope of rights of the present disclosure is not limited thereto, and various modifications and improvements made by those skilled in the art using the basic concept of the present disclosure defined in the following claims are also included in the scope of rights of the present disclosure. belongs to

Claims (10)

1. In behavioral patterns and medical imaging methods for assessment of mental health disorders,

   Inputting at least one brain image of different modalities of the user into at least one first pre-trained analysis model to evaluate a first risk of at least one mental health disorder;

   Collecting behavioral data, which is a plurality of time-series data about the user's behavioral patterns related to at least one mental health disease, through a user terminal;

   preprocessing the behavioral data by applying different weights to at least a portion of the behavioral data based on the first risk; and

   Inputting the behavioral data into a second pre-trained analytic model to assess a second risk of at least one mental health disorder.
2. According to claim 1,

   The second analysis model includes a clustering model,

   The step of assessing the second risk of mental health disease is,

   Inputting the behavioral data into the pre-trained clustering model to classify the behavioral data into at least one cluster related to at least one mental health disease; and

   Assessing a second risk of at least one mental health disorder based on the cluster.
3. According to claim 2,

   The second analysis model further includes a classification model,

   The step of assessing the second risk of mental health disease is,

   The method is performed by inputting the behavioral data into the pre-trained classification model and assessing a second risk of at least one mental health disorder based on the output of the classification model.
4. According to claim 1,

   Further comprising selecting one of the clustering model and the classification model as an application model based on at least one of user input, the type of behavioral data, the amount of behavioral data, and the number of brain images of different modalities,

   The method of claim 1 , wherein the step of assessing the second risk of a mental health disorder is performed via the selected application model.
5. According to claim 1,

   The brain image includes at least one of magnetic resonance imaging (MRI), functional magnetic resonance imaging (fMRI), positron emission tomography (PET-CT) imaging, diffusion-weighted imaging (DWI), and diffusion tensor imaging (DTI). method.
6. According to claim 1,

   The behavioral data includes data on at least one of heart rate, application switching frequency, repeated use frequency of the same application, movement pattern, and sleep pattern acquired through the user terminal over a predetermined period of time.
7. According to claim 1,

   The step of preprocessing the behavioral data is,

   The method is performed by applying a predetermined weight to at least some of the behavioral data of a type related to at least one of the mental health disorders according to the ranking of the first risk.
8. According to claim 1,

   The method of collecting the behavioral data is performed repeatedly at a predetermined period of time.
9. In behavioral data and medical imaging data devices for monitoring mental health disorders,

   at least one processor;

   a memory storing a program executable by the processor; and

   The processor, by executing the program, inputs at least one brain image of different modalities of the user into at least one first pre-trained analysis model to evaluate a first risk of at least one mental health disorder, and Collecting behavioral data, which is a plurality of time-series data about the user's behavior pattern related to at least one mental health disease, through a terminal, and applying different weights to at least some of the behavioral data based on the first risk , Preprocessing the behavioral data, and inputting the behavioral data into a second pre-trained analysis model to assess a second risk of the at least one mental health disorder.
10. A computer program stored in a recording medium for executing the method of any one of claims 1 to 8.

Images