KR20190133900A - System and method for diagnosing brain neuronal disease - Google Patents

Abstract

According to one embodiment of the present invention, a system for diagnosing cranial nerve disorder at low cost comprises: data measuring units measuring heart rate and activity; a data collection unit collecting heart rate data and activity data measured from the data measuring units; a data preprocessing unit preprocessing the heart rate data and activity data collected from the data collection unit; a data storage unit storing reference data for determining disorder; and a disorder determination unit determining for disorder based on the data preprocessed in the data preprocessing unit and the reference data stored in the data storage unit. The plurality of data measuring units may be provided. The data collection unit may collect the heart rate data and activity data measured by the data measuring units according to individual protocols.

Description

CENTRAL AND METHOD FOR DIAGNOSING BRAIN NEURONAL DISEASE}

The present invention relates to a system and method for diagnosing cranial nerve disease, and more particularly, to a system and method for diagnosing cranial nerve disease based on heart rate and activity.

Recently, more and more people are visiting hospitals because of abnormalities such as headaches, dizziness, memory problems, peripheral nerve disorders, and the like.

Cranial nerve disease refers to various diseases that can occur in the peripheral nerves and muscles, including the brain and spinal cord. Cerebral neurological disorders range from dementia, memory loss, headache, dizziness, epilepsy, numbness or paralysis due to peripheral neuropathy, and stroke due to cerebrovascular disorders (cerebral hemorrhage, cerebral infarction).

In particular, mild cognitive impairment is a state in which cognitive function, especially memory is far off compared to the same age, and the ability to perform daily life is preserved, it means a state not yet dementia. In other words, it is an intermediate step between normal aging and dementia. Epidemiological studies have shown that mild cognitive impairment is a high risk group for the transition to Alzheimer's disease. This condition is also clinically important in that it is the earliest stage of Alzheimer's disease and can maximize the therapeutic effect.

For this reason, there is a need for a way to track disease progression at all times.

The prior patent "Method for diagnosing early Alzheimer's disease or mild cognitive impairment (Registration No. KR10-1734645)" relates to a method for diagnosing Alzheimer's disease or mild cognitive impairment. At this time, the method of diagnosing Alzheimer's disease or mild cognitive impairment confirms the expression level of miR-206 in olfactory tissue, shows a high diagnosis rate for diagnosing Alzheimer's disease and mild cognitive impairment, can be diagnosed at low cost, and stability of biopsy is stable. It is characterized by high remarkable progress in diagnosing Alzheimer's disease or mild cognitive impairment.

Prior patent "Method and system for distinguishing functional change pattern of default mode network according to progression from normal person to mild cognitive impairment, Alzheimer's patient" (Registration No. KR10-1317040) "is related to progression from normal person to mild cognitive impairment and Alzheimer's patient The present invention relates to a method and a system for distinguishing functional change patterns of a default mode network. A method for distinguishing functional change patterns of a computer-implemented default mode network may include extracting an area having a difference greater than or equal to a predetermined value between default mode networks of a normal, mild cognitive impaired patient and an Alzheimer's patient. Extracting functional connectivity information of the default mode network from the extracted region, and analyzing the relationship between the extracted functional connectivity information and information on each dementia line indicating the degree of dementia to distinguish functional change patterns of the default mode network according to the progression of dementia. Characterized in that can be included.

Sentence comprehension and expressiveness test device and method for the diagnosis of mild cognitive impairment and Alzheimer's dementia (Registration No .: KR10-1562105) "Sentence comprehension and expressiveness tester is based on the color character with reference to the test question DB A test item data generation unit for generating a series of item data including graphic data and audio data of an answer sentence, an incorrect answer sentence, an ignition sentence or a view sentence for each item of an understanding task and a sentence expression task and a verb understanding task and a verb expression task; By using the generated item data, for each item of the task of each department, the graphic data is visually displayed to the examinee and the audio data is transmitted audibly, and the test progress control unit and the test process collecting the response data and the response data. In response, the response data input from the examinee is processed to correct the answer. Determining an incorrect answer, and a determination result; and a result processing unit for storing the test result DB.

As described above, the preceding patents are characterized by a procedure and method for diagnosing mild cognitive impairment based on image data obtained from a magnetic resonance imaging apparatus, data of olfactory tissue, or sentence understanding power.

Such procedures and methods require separate data collection procedures and costs. In particular, since the methods are data collected at a specific point in time, if the state of the data to be collected has a characteristic that can vary with time, the result may vary depending on the inspection point. In general, patients with mild cognitive impairment may have a low degree of cognitive impairment, and may produce different results depending on conditions at the time of examination. That is, it is not possible to constantly monitor the transition from the normal state to the mild cognitive impairment state.

In addition, test tools such as MMSE, CDR, GDS, etc., which are widely used clinically, have a problem that a patient learns test tools through repetitive tests.

The background art described above is possessed or acquired by the inventors in the derivation process of the present invention, and is not necessarily a publicly known technology disclosed to the general public before the application of the present invention.

The purpose of one embodiment is to classify the normal group and the disease group (for example, mild cognitive impairment group) based on the heart rate and the amount of activity (steps) and easily collectable from the wearable device with less discomfort in everyday life brain nerves It is to provide a system and method for diagnosing brain neurological disease which can diagnose the disease early.

An object according to an embodiment is to provide a system and method for diagnosing cranial nerve disease that can pre-process heart rate data and activity data to easily compare with reference data.

An object according to an embodiment is to provide a system and method for diagnosing cranial nerve disease that can convert data having a relatively long unit time into data having a relatively short unit time.

An object according to an embodiment is to provide a system and method for diagnosing cranial nerve disease, which enables follow-up management of cranial nerve disease without a hospital visit through constant monitoring.

An object according to one embodiment is to provide a system and method for diagnosing cranial nerve disease, which may be simple and inexpensive to query or test procedures.

According to an embodiment of the present invention, there is provided a system for diagnosing a neurological disease in a brain, the data measuring unit measuring a heart rate and an activity amount; A data collector configured to collect heart rate data and activity data measured by the data measurer; A data preprocessor configured to preprocess heart rate data and activity data collected in the data collector; A data storage unit for storing reference data for disease determination; And a disease discrimination unit determining whether a disease is based on data preprocessed by the data preprocessing unit and reference data stored in the data storage unit, wherein the data measuring unit is provided in plurality. Heart rate data and activity data measured by the measurement unit may be collected according to individual protocols.

According to one side, the data preprocessor may combine the heart rate data and activity data collected in the data collection unit into a single tuple on the basis of the same time to generate a plurality of tuples, and remove missing data or outlier data. .

According to one side, the data preprocessor may generate a data density image for each particular period based on the plurality of tuples.

According to one side, the reference data stored in the data storage unit is provided with a data density image of a normal group or a data density image of a disease group, the disease determination unit is a data density image of the user generated by the data pre-processing unit and the normal The similarity of the data density image of the group or the data density image of the disease group may be determined to determine whether the user has a disease.

According to one side, the plurality of data measuring unit has a different unit of measurement time, the data pre-processing unit converts the heart rate data and activity data collected in the data collection unit to the same unit of measurement time, based on the same time By generating a plurality of tuples by combining into one tuple, or the data pre-processing unit combines the heart rate data and activity data collected in the data collection unit into a single tuple based on the same time and converts the same to the same unit of measurement time You can create tuples.

According to one side, the data pre-processing unit, the data collection unit by generating a virtual data by dividing the data having a long unit of measurement time from the heart rate data and activity data collected in the data collection unit into a plurality of data according to a normal distribution The heart rate data and the activity data collected in the data may be converted into data having the same unit time as the shorter unit of time data.

According to one side, further comprising a reference updating unit for updating the reference data stored in the data storage unit, the reference data is a learning data model, the reference updating unit, based on the data collected in the data collection unit The reference data may be learned to actively update the reference data.

According to one side, the data storage unit is further stored in the data pre-processed by the data pre-processing unit, the result determined by the disease determination unit, the reference data stored in the data storage unit and the pre-processed data stored in the data storage unit It may further include a service provider for providing at least one of the.

Cerebral neurological disease diagnosis method according to an embodiment for achieving the above object, the step of measuring the amount of activity and heart rate in the data measuring unit; Collecting activity data and heart rate data measured by the data measurer; Pre-processing activity data and heart rate data collected by the data collection unit; And determining whether the disease is based on pre-stored reference data and the preprocessed activity data and heart rate data, wherein the activity data and heart rate data collected in the data collection unit are preprocessed. In the step of determining whether an image is generated based on the pre-stored reference data, the pre-processed activity amount data, and the heart rate data, the image may be determined based on the data density image for each specific period.

According to one side, the step of pre-processing the activity data and the heart rate data collected in the data collection unit, the step of determining whether the activity unit data and the measurement unit time of the heart rate data is determined; includes the activity data and If the unit time of measurement of the heart rate data coincides, the activity amount data and the heart rate data are combined into one tuple based on the same time to generate a plurality of tuples, and if the unit time of the activity amount data and the heart rate data do not match After the activity data and the heart rate data are converted into the same unit time of measurement, a plurality of tuples may be generated by combining them into one tuple based on the same time.

According to an embodiment of the system and method for diagnosing cerebral nerve disease, a normal group and a disease group (eg, mild or mild) are based on an activity amount (steps) and a heart rate that can be easily collected from a wearable device having less inconvenience in daily life. Group of disorders) can be used to diagnose cerebral neuropathy early.

According to the system and method for diagnosing neurological disease according to an embodiment, the heart rate data and the activity data may be preprocessed and easily compared with the reference data.

According to a system and method for diagnosing cranial nerve disease according to an embodiment, data having a relatively long unit time may be converted into data having a relatively short unit time.

According to a system and method for diagnosing cerebral nerve disease according to an embodiment of the present disclosure, it may be possible to track and manage the neurological disorder without visiting a hospital through constant monitoring.

According to a system and method for diagnosing cerebral nerve disease according to an embodiment, a question or test procedure may be simple and inexpensive.

1 is a block diagram of a system for diagnosing a neurological disease according to an embodiment.
2A to 2C show data density images of a normal group, data density images of a disease group, and differences in data density of a normal group and a disease group, respectively.
3 is a data density image of a normal group of actual specific periods.
4 is a data density image of a disease group of actual specific periods.
5 shows an example of a disease discriminating unit.
6 is a result of evaluating the accuracy of classifying a normal group and a disease group based on activity and heart rate.
7 is a flowchart illustrating a method for diagnosing cranial nerve disease according to an embodiment.

Hereinafter, the embodiments will be described in detail with reference to the accompanying drawings. In adding reference numerals to the components of each drawing, it should be noted that the same reference numerals are assigned to the same components as much as possible even though they are shown in different drawings. In addition, in describing the embodiment, when it is determined that the detailed description of the related well-known configuration or function interferes with the understanding of the embodiment, the detailed description thereof will be omitted.

In addition, in describing the components of the embodiment, terms such as first, second, A, B, (a), and (b) may be used. These terms are only for distinguishing the components from other components, and the nature, order or order of the components are not limited by the terms. If a component is described as being "connected", "coupled" or "connected" to another component, that component may be directly connected or connected to that other component, but between components It will be understood that may be "connected", "coupled" or "connected".

Components included in one embodiment and components including common functions will be described using the same names in other embodiments. Unless stated to the contrary, the description in any one embodiment may be applied to other embodiments, and detailed descriptions thereof will be omitted in the overlapping range.

1 illustrates the configuration of a system for diagnosing a neurological disease according to an embodiment, and FIGS. 2A to 2C show data density images of a normal group, data density images of a disease group, and data density differences between a normal group and a disease group, respectively. 3 is a data density image of a normal group of a specific period of time, FIG. 4 is a data density image of a disease group of a specific period of time, FIG. 5 shows an example of a disease discriminating unit, and FIG. 6 is based on activity and heart rate. This is the result of evaluating the accuracy of classifying the normal group and the disease group.

Referring to FIG. 1, the neurological disease diagnosis system 10 according to an embodiment includes a data measuring unit 100, a data collecting unit 200, a data preprocessor 300, a disease determining unit 400, and a data storage unit. 500, a reference update unit 600, and a service provider 700 may be included.

The data measuring unit 100 may measure a user's heart rate and activity amount.

Existing studies have shown that activity sufferers and sleep disorders have been shown in patients with cerebral neurological disorders such as mild cognitive impairment. Based on the background of the study, the heart rate and activity of the user may be measured to determine whether the user has a neurological disease.

In detail, the data measuring unit 100 may be provided in plurality.

For example, the plurality of data measuring units 100 may include a first data measuring unit 110 and a second data measuring unit 120.

The first data measuring unit 110 may include a first heart rate measuring unit 112 and a first activity measuring unit 114 to measure the heart rate and the activity of the first user.

The first heart rate monitor 112 may be, for example, a wearable device such as a smart watch or a wrist band (smart band) capable of constantly measuring the heart rate of the first user. Activity meter 114 may be a wearable device, such as a smart watch or wrist band (smart band) that can, for example, constantly measure the activity of a first user (eg, steps, etc.). Can be

In addition, the second data measuring unit 120 may include a second heart rate measuring instrument 122 and a second activity measuring instrument 124 to measure the heart rate and the amount of activity of the second user.

The second heart rate monitor 122 may be, for example, a wearable device such as a smart watch or a wrist band (smart band) capable of constantly measuring the heart rate of the second user. Activity meter 124 is a wearable device, such as a smart watch or wrist band (smart band), for example, which can constantly measure the amount of activity (eg, number of steps, etc.) of a second user. Can be

As described above, the data measuring unit 100 may be worn in the form of a watch in everyday life such as a wearable device, thereby reducing the feeling of rejection of wearing the product.

Meanwhile, the first data measuring unit 110 and the second data measuring unit 120 may be manufactured by the same or different manufacturers, and the first heart rate monitor 112 and the first activity meter 114 may be identical to each other. The second heart rate monitor 122 and the second activity meter 124 may also be manufactured by the same or different manufacturers.

As a result, the first heart rate measurer 112 and the first activity measurer 114 of the first data measurer 110 may have the same or different measurement unit time, and the second of the second data measurer 120 The heart rate monitor 122 and the second activity meter 124 may have the same or different unit time of measurement.

Alternatively, the first heart rate measuring unit 112 of the first data measuring unit 110 and the second heart rate measuring unit 122 of the second data measuring unit 120 may have the same or different measurement unit time, and the first The first activity number measuring unit 114 of the data measuring unit 110 and the second activity number measuring unit 124 of the second data measuring unit 120 may have the same or different measurement unit time.

For example, the first heart rate measuring unit 112 of the first data measuring unit 110 or the second heart rate measuring unit 122 of the second data measuring unit 120 measures the heart rate in units of 1 minute, and the first data. The first activity measurer 114 of the measurement unit 110 or the second activity measurer 124 of the second data measurer 120 measures the activity in units of 1 minute, or in addition to 1 minute, 10 minutes and 1 hour You can measure your activity with.

Alternatively, the first heart rate measuring unit 112 and the first activity measuring unit 114 of the first data measuring unit 110 measure the heart rate and the activity amount in units of one minute, and the second heart rate of the second data measuring unit 120. The measuring unit 122 and the second activity measuring unit 124 may measure heart rate and activity in units of 1 minute, or may measure heart rate and activity in units of 10 minutes and 1 hour in addition to 1 minute.

As an example, a manufacturer selling a smart band (wrist band) provides the heart rate data and activity data of a specific user in units of 1 minute, 1 second, and the like as follows.

1, a plurality of data measuring units 100 includes a first data measuring unit 110 measuring heart rate and activity of a first user and a second data measuring unit 120 measuring heart rate and activity of a second user. Although illustrated as including, it is obvious that the plurality of data measuring units 100 may include a heart rate meter for measuring a heart rate of a specific user and an activity meter for measuring an activity amount of a specific user.

As described above, the heart rate data and the activity data measured by the data measuring unit 100 may be collected by the data collecting unit 200.

The data collection unit 200 may collect heart rate data and activity data measured by the plurality of data measuring units 100 according to individual protocols.

In detail, heart rate data and activity data measured by the plurality of data measuring units 100 may be directly connected to the data collection unit 200 through a Bluetooth (BLE), a wireless communication network (WIFI), and the like. Alternatively, heart rate data and activity data measured by the plurality of data measuring units 100 are primarily stored in a data server of a corresponding manufacturer through a dedicated app of the corresponding manufacturer of the data measuring unit 100 and collected in a cloud service form. can do.

1, heart rate data and activity data measured by the first data measurer 110 are collected in the form of a Klaus service in the data collector 200, and heart rate data and activity data measured by the second data measurer 120. Although illustrated as being collected in the data collection unit 200 through Bluetooth (BLE), wireless communication network (WIFI), etc., heart rate data and activity data measured by the data measuring unit 100 is collected by the data collection unit 200 The protocol is not limited thereto, and any of the heart rate data and activity data measured by the data measuring unit 100 may be collected by the data collecting unit 200.

As described above, heart rate data and activity data collected by the data collector 200 may be transmitted to the data preprocessor 300.

The data preprocessor 300 may convert the heart rate data and the activity data collected by the data collector 200 into a data form that can be used by the disease determiner 400 to determine whether a disease exists.

In detail, the data preprocessor 300 may combine heart rate data and activity data collected by the data collector 200 into one tuple based on the same time.

For example, heart rate data can be collected as follows.

[Serial number 25, 2018-03-15 12:30, heart rate 76]

In addition, activity data may be collected as follows.

[Serial number 10, 2018-03-15 12:30, 25 steps]

In this case, when heart rate data and activity data are combined into one tuple based on the same time, it may be as follows.

[Serial number 37, 2018-03-15 12:30, heart rate 76 steps 25]

As such, in the process of combining the heart rate data and the activity data, missing value data without the heart rate data or the activity data may be excluded.

In addition, in the process of combining the heart rate data and the activity data, outlier data outside the range of multiples (2 times, 3 times, etc.) of the standard variance value based on the average value may also be excluded.

The data preprocessor 300 may generate a plurality of tuples by combining the heart rate data and the activity data collected by the data collector 200 into one tuple based on the same time.

For example, a plurality of tuples may be generated as follows.

[Serial number 37, 2018-03-15 12:30, heart rate 76 steps 25]

[Serial number 47, 2018-03-15 12:40, heart rate 80 steps 30]

Then, the data preprocessor 300 may generate a data density image for a specific period based on the plurality of tuples.

In detail, the data preprocessor 300 may have a data density based on a plurality of tuples for a specific period of time (one week, one day, three days, etc.).

For example, the week 40 data density table for user A is:

In the table above, step indicates the user's activity (steps), hr indicates the user's heart rate, and count indicates the frequency or frequency at which the activity and heart rate were measured during the 40th week.

Then, the data preprocessor 300 may generate a data density image from the data density table.

In particular, as shown in FIGS. 2A and 2B, a data density image may be generated in the data preprocessor 300 from heart rate data and activity data collected by the data collector 200.

In FIGS. 2A-2C, the X-axis indicates heart rate and the Y-axis may indicate activity (steps).

Specifically, FIG. 2A is a data density image of a normal group, and it can be seen that the heart rate data is high in the range of 85 to 100 and the activity data is high in the range of 0 to 125. In addition, Figure 2b is a data density image of the disease group, it can be seen that the heart rate data is a high frequency in the range of 55 to 80. Figure 2c is an image showing the data density difference between the normal group and the disease group, through which it can be seen that there are a lot of patterns of behavior in the form of cardio exercise in the normal group.

Therefore, it is possible to determine whether the user's disease is based on the data density image or the data density table generated by the data preprocessor 300, and can distinguish the disease group from the normal group.

In addition, the neurological disease diagnosis system 10 according to one embodiment actually generated data density images of the normal group and the disease group for a specific period of time.

3 and 4, the data density image of the normal group and the data density image of the disease group were generated differently during a specific period of time. It was confirmed that there are many.

As described above, the data preprocessing unit 300 in the cranial nerve disease diagnosis system 10 according to an embodiment not only has a plurality of data density images for a specific period of time (eg, 1 day, 3 days, 1 week, etc.) A data density image may be generated for a particular time period for a user of.

On the other hand, as described above, when the plurality of data measuring units 100 have different unit time of measurement, the data preprocessor 300 is converted into various granularities (1 minute, 10 minutes, 1 hour, etc.) Multiple tuples can be created.

In detail, when the first heart rate monitor 112 and the first activity meter 114 have different unit time of measurement, the data preprocessor 300 collects the heart rate data and the activity data collected by the data collector 200. After converting to the same unit of measurement time, a plurality of tuples may be generated by combining into one tuple based on the same time.

Alternatively, when the first heart rate measurer 112 and the first activity measurer 114 have different unit time units of measurement, the data preprocessor 300 may collect the heart rate data and the activity amount data collected by the data collector 200. After generating a plurality of tuples by combining one tuple on the basis of the same time, the plurality of tuples may be secondarily generated by converting the plurality of tuples into the same unit of measurement time.

On the other hand, when the first data measuring unit 110 and the second data measuring unit 120 have different measurement unit times, the data preprocessor 300 may be connected to the first user collected by the data collection unit 200. After converting the heart rate data and activity data for the second user and the heart rate data and activity data for the second user into the same unit of measurement time, the plurality of tuples may be generated by combining them into one tuple based on the same time.

Alternatively, when the first data measuring unit 110 and the second data measuring unit 120 have different unit time of measurement, the data preprocessor 300 may be connected to the first user collected by the data collection unit 200. After combining the heart rate data and the activity data for the second user and the heart rate data and the activity data for the second user individually into one tuple on the same time basis for the first user and the second user, the plurality of tuples are generated first. Tuples may be converted to the same unit of measurement time to generate a plurality of tuples.

In this case, the data preprocessor 300 divides data having a long unit of measurement time among heart rate data and activity data collected by the data collector 200 into a plurality of data according to a normal distribution to generate virtual data, thereby generating one tuple. Alternatively, a plurality of tuples may be converted into the same unit of measurement time.

For example, if the heart rate data is measured in 1 minute units and the activity data is measured in 10 minutes units, the activity data is divided into 10 data according to the user's data distribution (for example, a normal distribution). To generate virtual one minute data.

Similarly, if the heart rate data is measured in units of 10 minutes, and the activity data is measured in units of 1 hour, the activity data is divided into six data according to the user's data distribution (for example, a normal distribution). Virtual 10 minute data can be generated.

Alternatively, the heart rate data or activity data of the first user measured by the first data measuring unit 110 is data measured in units of 1 minute, and the heart rate data of the second user measured by the second data measuring unit 120 or When the activity data is data measured in units of 10 minutes, the 10 heart rate data or the activity data of the second user measured by the second data measuring unit 120 is 10 data according to the data distribution (for example, a normal distribution) of the user. The virtual data may be generated by dividing into.

Similarly, heart rate data or activity data of the first user measured by the first data measuring unit 110 are data measured in units of 10 minutes, and heart rate data of the second user measured by the second data measuring unit 120. Alternatively, when the amount of activity data is data measured in units of 1 hour, the heart rate data or the amount of activity data of the second user measured by the second data measuring unit 120 may be determined according to the data distribution (eg, normal distribution) of the user. Can be divided into data to generate virtual 10-minute data

For example, when heart rate data and activity data acquired for a plurality of users are changed based on the same unit of measurement time, they may be as follows.

As such, the data preprocessor 300 may generate a user-id step and heart rate hr for each specific timestamp, and further, the data preprocessor 300 may be configured for a specific user. A data density table for each user and a specific time period can be generated as well as a data density table for a specific time period, and data density images for a plurality of users can be generated individually.

As described above, when the plurality of data measuring units 100 have different unit time units of measurement, various data measuring units are generated by generating data density images for a specific period based on a plurality of tuples converted to the same unit time as described above. Heart rate data and activity data measured at 100 may be used to generate more reliable data density images.

As described above, the data preprocessed by the data preprocessor 300 may be transmitted to the disease determiner 400.

The disease determining unit 400 may determine whether the user has a disease based on the data preprocessed by the data preprocessor 300.

In this case, reference data for disease determination may be stored in advance in the data storage unit 500, and the disease determination unit 400 may pre-process the data pre-processed by the data preprocessor 300 and reference data stored in the data storage unit 500. Based on the disease can be determined.

As described above, the data preprocessed by the data preprocessor 300 may be a data density image, and the reference data stored in the data storage 500 may be provided as a data density image of a normal group or a data density image of a disease group. Can be.

Therefore, the disease determination unit 400 determines the similarity between the data density image of the user generated by the data preprocessor 300 and the data density image of the normal group or the data density image of the disease group stored in the data storage unit 500. It can be determined whether the disease or the state of the user (or patient).

In this case, the disease determination unit 400 may utilize a CNN (Convolutional Neural Network) model, which is a deep learning technique optimized for image-based classification, as shown in FIG. 5.

In detail, the data density image of the user generated by the data preprocessor 300 and training data and validation data using the user's state as a label (normal or disease) may be used as an input of the CNN model. have.

In particular, referring to FIG. 6, the disease determination unit 400 may check an evaluation result of the accuracy of classifying the normal group and the disease group based on the heart rate and the activity amount, and as a result, the heart rate and activity amount may be determined by the disease determination unit 400. The accuracy of classifying the normal group and the disease group was very high (98%).

In this case, the disease determination unit 400 may determine whether the user has a disease or the state of the user (or patient) based on the data density image of the user generated by the data preprocessor 300 based on the reference data updated by the reference update unit 600. Can be determined.

Specifically, the reference data may be a learning data model as reference data for classifying a normal group and a disease group in consideration of age, gender, region, and the like.

The reference updating unit 600 may actively update the reference data by learning the collected data in addition to the previously learned learning data model. In this case, the pre-learned data model may learn data collected additionally by age, gender, and region.

In this case, the reference data updated by the reference update unit 600 may be stored in the data storage unit 500.

As described above, the disease determining unit 400 may determine whether the user has a disease or a state of the user (or a patient) according to a data model learned based on recently collected data. Or patient).

The data storage unit 500 may further store data preprocessed by the data preprocessor 300, for example, a plurality of tuples, data density tables, and data density images.

Therefore, the data storage unit 500 may store at least one of data preprocessed by the data preprocessor 300 and reference data updated by the reference updater 600.

In addition, the service provider 700 may provide a service to an external operator or a user. For example, the service provider 700 may be stored in the data storage unit 500 as a result of the disease determination unit 400. At least one of reference data and preprocessed data stored in the data storage unit 500 may be provided.

As described above, the brain neurological disease diagnosis system 10 may be classified into a normal group and a disease group based on the amount of activity and heart rate of the user, thereby enabling early diagnosis of the disease, and easy tracking and management without visiting a hospital through regular monitoring. can do.

It has been described with respect to the neurological disease diagnosis system according to one embodiment, below will be described with respect to the neurological disease diagnosis method.

7 is a flowchart illustrating a method for diagnosing cranial nerve disease according to an embodiment.

Referring to FIG. 7, a method for diagnosing cranial nerve disease according to an embodiment may be performed as follows.

First, the user's activity and heart rate are measured in the data measuring unit (S10).

In this case, the amount of activity and heart rate of the user may be measured by a wearable device such as a smart watch or a wrist band (smart band).

Then, activity data and heart rate data measured by the data measuring unit is collected in the data collection unit (S20).

In this case, the activity data and the heart rate data may be collected in the data collection unit through various protocols such as a communication network or a cloud service.

Subsequently, activity data and heart rate data collected in the data collection unit are preprocessed (S30).

Specifically, first, it is determined whether the unit time of the activity data and the heart rate data is matched (S32). When the unit time of the activity data and the heart rate data is matched, the activity data and the heart rate data are one tuple based on the same time. Combined into a plurality of tuples are generated (S34).

On the other hand, if the unit time of activity data and heart rate data is inconsistent, the activity data and heart rate data are converted to the same unit time of measurement (S36), and then the activity data and heart rate data are combined into one tuple based on the same time, and thus, Two tuples are generated (S36).

In this case, the active amount data and heart rate data may generate virtual data by dividing the data having a long unit of measurement time among the heart rate data and the active amount data into a plurality of data according to a normal distribution, and the virtual data may be one of the heart rate data and the active amount data. The unit time of measurement may have the same unit time as the short data.

Although not specifically illustrated in FIG. 7, in the step of preprocessing the activity data and the heart rate data collected in the data collection unit, a data density image of a specific period of time may be additionally generated based on a plurality of tuples.

In addition, the step of pre-processing the activity data and heart rate data collected in the data collection unit is not limited to the flow chart shown in Figure 7, any one can be determined if the disease can be determined by judging the similarity with the previously stored reference data. .

Finally, whether or not the disease is determined based on the pre-stored reference data, the preprocessed activity data and the heart rate data (S40).

Since the activity data and the heart rate data processed as described above are generated as data density images of a specific period of time of the user, prestored reference data is also provided as a data density image to determine the similarity between the data density image of the user and the previously stored reference data. Thus, whether or not the user's disease can be determined.

Additionally, as a result of analyzing the subject homogeneity of the data utilized in the present invention, the Korean version of Mini-Mentel State Exam (K-) between the healthy group (HC) and the control group (MCI) as follows. All of the demographics except for the MMSE) test showed no significant difference (p> 0.05), thereby confirming the homogeneity of the subjects.

In addition, various dementia related tests in Seoul Neuropsychological Screening Battery (2nd Edition, SNSB-II) showed significant differences among groups. In addition, there was a slight difference between the normal group and the control group in the language domain and the memory domain in the cognitive domain test of the SNSB-II test.

As described above, since the homogeneity was shown in the normal group and the disease group, the difference was significantly different in the dementia test and the cognitive area test. It can determine whether the user's condition or disease through.

Although the embodiments have been described with reference to the accompanying drawings as described above, various modifications and variations are possible to those skilled in the art from the above description. For example, the described techniques may be performed in a different order than the described method, and / or components of the described structure, apparatus, etc. may be combined or combined in a different form than the described method, or may be combined with other components or equivalents. Appropriate results can be achieved even if they are replaced or substituted.

10: Neurological Disease Diagnosis System
100: data measurement unit
110: first data measurement unit
112: first heart rate monitor
114: first activity meter
120: second data measurement unit
122: second heart rate monitor
124: second activity meter
200: data collector
300: data preprocessor
400: disease discrimination unit
500: data storage
600: standard update unit
700: service provider

Claims (10)

A data measuring unit measuring heart rate and activity;
A data collector configured to collect heart rate data and activity data measured by the data measurer;
A data preprocessor configured to preprocess heart rate data and activity data collected in the data collector;
A data storage unit for storing reference data for disease determination; And
A disease discrimination unit determining whether a disease is based on data preprocessed by the data preprocessor and reference data stored in the data storage unit;
Including,
The data measuring unit is provided in plurality,
The data collecting unit collects heart rate data and activity data measured by a plurality of data measuring units according to individual protocols, neurological disease diagnosis system.
The method of claim 1,
The data preprocessor is configured to combine the heart rate data and activity data collected in the data collection unit into a single tuple on the basis of the same time to generate a plurality of tuples, and remove missing data or outlier data, brain neurological disease diagnosis system.
The method of claim 2,
The data preprocessing unit generates a data density image for a specific period of time based on the plurality of tuples.
The method of claim 3,
The reference data stored in the data storage unit is provided as a data density image of a normal group or a data density image of a disease group, and the disease discrimination unit is a data density image of a user generated by the data preprocessor and a data density image of the normal group. Or judging the similarity of the data density image of the disease group to determine whether the user's disease.
The method of claim 1,
The plurality of data measuring units have different unit time of measurement,
The data preprocessor converts the heart rate data and the activity data collected in the data collector into the same unit of measurement time, and then combines them into one tuple based on the same time to generate a plurality of tuples,
The data preprocessing unit merges heart rate data and activity data collected in the data collection unit into one tuple on the basis of the same time and converts the same unit of measurement time to generate a plurality of tuples.
The method of claim 5,
The data preprocessing unit divides the data having a longer unit of measurement time from the heart rate data and the activity data collected in the data collection unit into a plurality of data according to a normal distribution to generate virtual data by collecting the heart rate data collected in the data collection unit. And a neurological disease diagnosis system for converting the activity unit data into data having the same unit time as the unit of measure time.
The method of claim 1,
A reference updater for updating the reference data stored in the data storage;
More,
The reference data is a training data model,
The reference update unit,
Neurological disease diagnosis system for actively updating the reference data by learning the reference data based on the data collected in the data collection unit.
The method of claim 1,
The data storage unit further stores the data preprocessed by the data preprocessor,
And a service provider configured to provide at least one of reference data stored in the data storage unit and preprocessed data stored in the data storage unit as a result of the disease determination unit.
Measuring the amount of activity and heart rate in the data measuring unit;
Collecting activity data and heart rate data measured by the data measurer;
Preprocessing the activity amount data and the heart rate data collected in the data collection unit; And
Determining whether a disease is based on previously stored reference data and the preprocessed activity data and heart rate data;
Including,
In the step of preprocessing the activity data and the heart rate data collected in the data collection unit, a data density image for each specific period is generated,
In the step of determining whether the disease based on the pre-stored reference data, the pre-processed activity data and heart rate data, whether the disease is determined based on the data density image for each particular period, brain nerve disease diagnosis method.
The method of claim 9,
Pre-processing the activity data and the heart rate data collected in the data collection unit,
Determining whether the unit time of the activity data and the heart rate data coincide with each other;
Including,
When the unit of measurement time of the activity data and the heart rate data coincide, the activity data and the heart rate data are combined into one tuple on the basis of the same time to generate a plurality of tuples,
When the unit time of measurement of the activity data and the heart rate data is inconsistent, the activity data and the heart rate data are converted into the same unit of time, and then merged into one tuple on the basis of the same time to generate a plurality of tuples How to diagnose the disease.

Images