KR20200137536A - Automatic analysis system for picture for cognitive ability test and recording medium for the same - Google Patents

Abstract

According to an embodiment of the present invention, an automatic picture analysis system for a cognitive ability test includes a server and a user device. The server includes a control unit that provides the pentagonal drawing test data to a user device to determine whether there is a cognitive impairment, obtains response information of the user corresponding to the pentagonal drawing test, analyzes the obtained response information, and scores an analysis result to perform cognitive impairment diagnosis. The control unit includes a feature of an intersection area of two objects in a picture drawn by a user as an analysis item, wherein the characteristic of the intersection area includes information on the presence or absence of the intersection area and a distance between two objects. The user device receives information required to perform a pentagon drawing test from the server through communication with the server.

Description

Automatic analysis system for picture for cognitive ability test and recording medium for the same

The present invention relates to a system for automatically analyzing pictures for cognitive ability and to a recording medium therefor.

Recently, as the importance of early diagnosis of cognitive disorders such as dementia has been recognized, a drawing test, which is a relatively simple method of discriminating whether there is a cognitive disorder, has been widely used. Such types of drawing inspection include clock drawing inspection and pentagon drawing inspection.

Among them, the pentagon drawing test is a test in which two pentagons partially overlapped are presented to the subject, and the subject draws according to the presented image. Observing the patient's examination in actual clinical practice, it can be found that depending on the person who draws, the direction of the pentagon, whether it intersects, and the degree to which the two pentagons are close vary.

However, in the existing scoring method, performance ability is scored with only 0 or 1 point, and since both cases where no picture is drawn and cases where errors are shown only in small areas receive 0 points in common, the subject's accurate cognitive ability is identified. There is a lack of side to do.

Accordingly, in performing the pentagon drawing test, it is necessary to evaluate each item such as whether or not two pentagons cross each other, as well as the cross shape and the degree to which the two pentagons are close to each other, and diagnose cognitive impairment based on this. Is being raised.

Unexamined Patent Publication No. 10-2019-0015649 (A method and prediction system for predicting the onset of dementia using a neuropsychological test)

In the present invention, in the pentagon drawing test, which is a type of cognitive ability test for determining cognitive impairment such as dementia, the test result of the test subject is automatically scored, and an image of an intersection area in which two figure objects overlap. The purpose is to automatically detect and calculate the scoring results in various score bands accordingly.

The automatic analysis system according to an embodiment of the present invention includes a server and a user device, wherein the server provides pentagon drawing test data to the user device to determine whether or not a cognitive impairment exists, and a user corresponding to the pentagon drawing test A control unit for acquiring the response information of, analyzing the obtained response information, and performing a scoring on the analysis result to perform cognitive disorder diagnosis, wherein the control unit is used as an analysis item when analyzing the response information. In the drawn picture, the feature of the intersection area of two objects is included, and the feature of the intersection area includes information on the presence or absence of the intersection area and the distance between the two objects, and the user device communicates with the server to the server. It is characterized in that the information required to perform the pentagon drawing inspection is received from and executed.

The present invention provides an automatic scoring method for a pentagon drawing test, so that a simple cognitive ability test opportunity can be provided to more users.

The present invention may provide detailed information on the severity of cognitive impairment in response to various user types determined through a pentagonal drawing test.

1 is a diagram showing the configuration of an automatic analysis system for a picture for a cognitive ability test according to an embodiment of the present invention.
2 is a block diagram of a server configuring an automatic analysis system for a picture for a cognitive ability test according to an embodiment of the present invention.
3 is a block diagram showing the configuration of a test execution unit according to an embodiment of the present invention.
4 is a block diagram showing a configuration of a preprocessor according to an embodiment of the present invention.
5 is a block diagram showing the configuration of an information analysis unit according to an embodiment of the present invention.
6 is a block diagram illustrating a configuration of a feedback providing unit according to an embodiment of the present invention.
7 to 22 are exemplary views for explaining a data processing process for a cognitive ability test according to an embodiment of the present invention.

In the present invention, various modifications may be made and various embodiments may be provided, and specific embodiments will be illustrated in the drawings and described in detail.

However, this is not intended to limit the present invention to a specific embodiment, it is to be understood to include all changes, equivalents, and substitutes included in the spirit and scope of the present invention. In describing each drawing, similar reference numerals have been used for similar elements.

When an element is referred to as being'connected' or'connected' to another element, it is understood that it may be directly connected or connected to the other element, but other elements may exist in the middle. Should be. On the other hand, when a component is referred to as being'directly connected' or'directly connected' to another component, it should be understood that there is no other component in the middle.

The terms used in the present application are only used to describe specific embodiments, and are not intended to limit the present invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In the present application, terms such as'comprise' or'have' are intended to designate the presence of features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, but one or more other features It is to be understood that the presence or addition of elements or numbers, steps, actions, components, parts, or combinations thereof, does not preclude in advance.

In the present specification, the cognitive ability test may include a cognitive ability test for various purposes. Preferably, the test performed by the automatic analysis system for a picture for cognitive ability test according to an embodiment of the present invention is a pentagon drawing test (PDT), which is one of the mini mental state examinations (MMSE). Can mean In addition, in the present specification,'response information' may mean picture-related information input by a user in response to a request for a pentagon drawing inspection request from a server, and'image data' which means an image of a picture created by the user as a test, and the picture It may include'sensor data' which is information that senses the order and speed of creation.

In addition,'cognitive disorder' in the present specification may be interpreted as meaning dementia. And accordingly, the cognitive ability test in the present specification can be interpreted as meaning a dementia test.

1 is a diagram showing the configuration of an automatic analysis system for a picture for a cognitive ability test according to an embodiment of the present invention.

The automatic analysis system for pictures for cognitive ability test according to an embodiment of the present invention may include a server 100 and a user device 200 communicating with the server 100. The number of user devices 200 may be plural (eg, 201 to 20n), and the types may also be various such as smart phones, tablets, PCs, and wearable devices. In order to support the inspection function for various types of user devices, the server 100 may be implemented to perform the inspection operation in the form of a web or an app.

The user device 200 may refer to an electronic device including a sensor that receives a drawing input through a touch or a smart pen from a user.

The user device 200 according to various embodiments of the present document includes, for example, a smartphone, a tablet personal computer, a mobile phone, a video phone, and a display function and an input function. E-book reader, desktop personal computer, laptop personal computer, netbook computer, workstation, server, PDA (personal digital assistant), PMP ( portable multimedia player), an MP3 player, a mobile medical device, a camera, or a wearable device.

A detailed description of the server 100 will be described with reference to the following drawings.

2 is a block diagram of a server configuring an automatic analysis system for a picture for a cognitive ability test according to an embodiment of the present invention.

As shown in FIG. 2, the server 100 may include a communication unit 110, a storage unit 120, and a control unit 130. In addition, the control unit 130 may include a test execution unit 131, a preprocessor 132, an information analysis unit 133, and a feedback providing unit 134.

The communication unit 110 may use a network to transmit and receive data between a user device and a server, and the type of the network is not particularly limited. The network may be, for example, an Internet Protocol (IP) network that provides a service for transmitting and receiving large amounts of data through an Internet Protocol (IP) or an All IP network integrating different IP networks. In addition, the network includes a wired network, a wireless broadband (Wibro) network, a mobile communication network including WCDMA, a high speed downlink packet access (HSDPA) network and a mobile communication network including a long term evolution (LTE) network, and LTE advanced (LTE-A). ), a mobile communication network including 5G (Five Generation), a satellite communication network, and a Wi-Fi network, or at least one or more of them may be combined.

The communication unit 110 according to an embodiment of the present invention may transmit information required to perform a pentagon drawing inspection to a user device. In addition, the communication unit 110 is for receiving the user's response information (image data related to a pentagonal drawing, sensor data, etc.), personal information of the user, and lifestyle information (eating habits, lifestyle information) from the user device. Communication function can be performed.

According to various embodiments, the communication unit 110 may receive current location information or user residence information determined by the user device from the user device side. In response, the communication unit 110 may transmit information on the nearest medical institution to the user device 200 based on the calculated current location with the user device or the user residence information.

According to various embodiments, the communication unit 110 may communicate with a web server for search in addition to the user device. Alternatively, the communication unit 110 may communicate with a server of an associated institution or company. For example, the communication unit 110 automatically transmits the test results of a specific user (subject) to the linked medical institution as needed, and determines based on the opinion of the person in charge of the subject and MRI test data previously performed by the medical institution. Confirmation information about cognitive ability test can be received. Confirmation information on the cognitive ability test received from the medical institution for one user (subject) may be transmitted to the user device 200 again.

The storage unit 120 may include, for example, an internal memory or an external memory. The internal memory includes, for example, volatile memory (e.g., dynamic RAM (DRAM), static RAM (SRAM), synchronous dynamic RAM (SDRAM), etc.)), non-volatile memory (e.g., OTPROM (one time programmable ROM), programmable ROM (PROM), erasable and programmable ROM (EPROM), electrically erasable and programmable ROM (EEPROM), mask ROM, flash ROM, flash memory (such as NAND flash or NOR flash), hard drive, Alternatively, it may include at least one of a solid state drive (SSD).

External memory is a flash drive, for example, compact flash (CF), secure digital (SD), micro secure digital (Micro-SD), mini secure digital (mini-SD), extreme digital (XD), It may further include a multi-media card (MMC) or a memory stick. The external memory may be functionally and/or physically connected to the electronic device through various interfaces.

The storage unit 120 according to an embodiment of the present invention may store cognitive ability test data such as a problem for drawing a pentagon, a scoring criterion, a test score, and matching information of a cognitive impairment and severity.

In addition, the storage unit 120 may store an algorithm related to a segmentation operation based on U-net, and may store secondary images generated as a result of the segmentation operation. In addition, the storage unit 120 may store personal information directly input by the user and questionnaire information related to lifestyle.

The control unit 130 of the present invention may also be referred to as a processor, a controller, a microcontroller, a microprocessor, a microcomputer, and the like. Meanwhile, the control unit 130 may be implemented by hardware, firmware, software, or a combination thereof.

In the case of implementation by firmware or software, an embodiment of the present invention may be implemented in the form of a module, procedure, or function that performs the functions or operations described above. The software code may be stored in a memory and driven by the control unit. The memory may be located inside or outside the user terminal and the server, and may exchange data with the controller through various known means.

The controller 130 may perform overall operations related to performing a cognitive ability test according to an embodiment of the present invention. For example, the controller 130 acquires response information, such as an image of a picture drawn by a user, An operation of analyzing response information and scoring, an operation of determining whether there is a cognitive impairment according to the scoring result, and the like may be performed.

Specifically, the control unit 130 may include a test execution unit 131, a preprocessor 132, an information analysis unit 133, and a feedback providing unit 134.

The test execution unit 131 may provide a guide and an operation related to test execution so that a user (subject) may perform a pentagon drawing inspection.

A detailed configuration of the test execution unit 131 will be described below with reference to FIG. 3.

The preprocessor 132 may perform preprocessing on response information of a user obtained as a result of executing a test on the user by the test execution unit 131. Specifically, the preprocessor 132 obtains the user's response information obtained through the test operation provided to the user device 200 by the test execution unit 131, and performs a preprocessing operation of processing the obtained response information into information easy for analysis. Can be done. More specific information on the preprocessor 132 will be described below with reference to FIG. 4.

The information analysis unit 133 may analyze the user's response information to the test and calculate the user's score. Furthermore, the information analysis unit 133 may determine whether the user (the subject) has a cognitive impairment or determine the severity of the cognitive impairment based on the calculated score. As illustrated in FIG. 10, the information analysis unit 133 may perform an analysis operation based on a result obtained by inputting image data into an artificial neural network-based algorithm.

A detailed description of the information analysis unit 133 will be described below with reference to FIG. 5.

The feedback providing unit 134 may provide a feedback corresponding to a test result to a user (subject). The feedback providing unit 134 may provide, for example, a function related to testing or training for preventing the occurrence of a user's cognitive disorder or alleviating symptoms of a cognitive disorder to the user. In addition, in the present invention, the feedback providing unit 134 may select data related to learning of an artificial neural network that analyzes a test result of a user (subject) and perform an artificial neural network learning operation. More detailed information about the feedback providing unit 134 will be described below with reference to FIG. 6.

3 is a block diagram showing the configuration of a test execution unit according to an embodiment of the present invention.

The test execution unit 131 may perform an operation related to overall test execution, such as providing pentagon drawing inspection information (eg, a problem) to a user. To this end, the test execution unit 131 may include a problem execution unit 131a, a response information acquisition unit 131b, and a response information classification unit 131c, as shown in FIG. 3.

The problem execution unit 131a may transmit problems and instructions to the user device 200 so that the user (subject) performs a pentagon drawing test. The problem execution unit 131a may induce a user to draw a pentagon by presenting a screen as shown in FIG. 8.

More specifically, as shown in FIG. 8, the problem execution unit 131a may display the presented image on one side of the test providing screen so that the user can draw along. The presented image is a form in which two pentagons are overlapped, and this may be an image mainly used in a pentagon drawing inspection. In addition, the problem execution unit 131a may provide a drawing area in which the user draws a pentagon along the presented image in the remaining space of the test providing screen. The problem execution unit 131a may display a trajectory of a user's drawing input (eg, touch input, smart pen input, etc.) on the screen. Accordingly, the user can draw a test pentagon on the screen (or a separate input device) while grasping factors such as the shape and length of the line he wants to draw.

According to various embodiments, the problem execution unit 131a may inform the user of the type of hand to draw a pentagon (eg, whether it is a left hand or a right hand). According to an embodiment, the pentagonal drawing test is based on the type of hand to be tested or the order of the hand to perform the test based on the user's main hand type information (e.g., a right-handed user uses a left hand after a drawing test using a right hand). You can specify and tell) to perform drawing checks.

According to various embodiments, the problem execution unit 131a may perform a problem providing operation based on various user personal information. The problem execution unit 131a may check user personal information input to the user account before providing the problem to the user device 200. The items of the user's personal information may include, for example, a name, an age, a gender, a primary hand (eg, right-handed or left-handed). The problem execution unit 131a may calculate a probability of occurrence of a typical cognitive impairment based on the user's gender and age. Thereafter, the problem execution unit 131a may change the test condition for a user group corresponding to an age group and a gender whose normal cognitive impairment probability percentile is equal to or greater than a reference value. For example, the problem execution unit 131a may change conditions such as a change in the number of times a test proceeds and a decrease in a test time limit.

According to various embodiments, the item of personal information of the user may further include information on eating habits information, alcohol intake, etc., which are items related to occurrence of cognitive impairment. Before the test is performed, the user may be asked to enter information related to his or her eating habits (eg, the number of times of consumption of meat, the number of times of alcohol consumption, etc.) in the user's personal information. In response to this, the user can fill in the eating habit information, and in response, the problem execution unit 131a stores the risk of developing cognitive disorders (eg, dementia) corresponding to the user’s eating habit information (eg,'Food habit information- It can be calculated based on the risk level matching table'). In addition, the problem execution unit 131a may change test conditions, such as the number of times the test is performed and a test time limit, based on the risk level of developing cognitive disorders based on the user's eating habit information.

The response information acquisition unit 131b, which is a second configuration of the test execution unit 131, may perform an operation of acquiring response information input by a user. In this case, the response information may include'image data' of a pentagonal picture drawn by the user and'sensor data' which is data about an input element sensed when the user draws a picture (pentagon).

The response information acquisition unit 131b may acquire user input (touch input or smart pen input, etc.) information detected by a touch sensor for each predetermined period while an app on the user device 200 is being executed. For example, the response information acquisition unit 131b may extract and store x and y coordinate values of the user input sensed by the touch sensor every 20 ms sampling period.

Accordingly, the'sensor data' acquired by the response information acquisition unit 131b may include elements such as speed and acceleration when the user draws a pentagon. As illustrated in 911 of FIG. 9, the'sensor data' may include items related to a test execution date, a time required for the test, a speed of drawing a pentagon, a coordinate position of line generation by time slot, and the like. Later, the'sensor data' may be used to calculate a graph as shown in 912 and 913.

4 is a block diagram showing a configuration of a preprocessor according to an embodiment of the present invention.

The preprocessor 132 may preprocess the response information (sensor data, image data) obtained for one inspection in a form that is easy to analyze. As shown in FIG. 4, the preprocessor 132 may include a response information classification unit 132a and a useful data conversion unit 132b.

The response information classification unit 132a may classify the obtained response information into image data and sensor data. In addition, the response information classifying unit 132a may classify response information for each test time when a test is performed multiple times according to various embodiments. In addition, the response information classification unit 132a may classify response information according to a user type or group.

When the response information is classified by the response information classification unit 132a, the useful data conversion unit 132b may perform a data conversion operation of converting the response information into a form suitable for analysis.

Specifically, when image data and sensor data are classified by the response information classification unit 132a, the useful data conversion unit 132b may determine validity of the image with respect to the image data. For example, when the thickness of the line included in the image is greater than or equal to a preset value, the useful data conversion unit 132b may touch a palm by mistake, not a touch using a normal smart pen or a finger, and multi-touch multiple fingers by mistake. It is regarded as such, and can be judged as invalid accordingly.

The useful data conversion unit 132b may request the test execution unit 131 to perform a problem providing operation again when it is determined that the obtained image data is invalid. Alternatively, if the test is performed multiple times and there is response information classified for each episode in the response information classification unit 132a, the useful data conversion unit 132b determines that one image data is invalid, the next image You can judge the validity of the data. When the validity of the image data is determined to be greater than or equal to the reference value, the useful data conversion unit 132b may determine the image data as data for analysis and request the information analysis unit 133 to analyze the image data.

The useful data conversion unit 132b may convert sensor data into data for analysis in addition to image data. The useful data conversion unit 132b may generate a graph related to a drawing speed based on sensor data, and determine the generated graph as data for analysis. The graph may be generated as shown in FIG. 11 to extract information such as a plurality of peaks and average speeds indicated by drawing speed. For example, as shown in FIG. 9, the useful data conversion unit 132b may convert sensor data 911 into graphs such as 912 and 913. In addition, the useful data conversion unit 132b may determine graphs such as 811 and 813 as analysis data and request the information analysis unit 133 to analyze the graph.

5 is a block diagram showing the configuration of an information analysis unit according to an embodiment of the present invention.

The information analysis unit 133 may include an angle determination unit 133a, a cross determination unit 133b, a closure determination unit 133c, an additional feature determination unit 133d, and an artificial intelligence support unit 133e.

The information analysis unit 133 may perform an artificial intelligence-based score calculation operation for response information through the angle determination unit 133a, the cross determination unit 133b, and the closing determination unit 133c. Preferably, the information analysis unit 133 may perform a score calculation operation through U-net, which is one of CNN types.

As shown in FIG. 7, the U-net is an artificial neural network model in which the network shape is similar to that of the alphabet U, and can be used to extract the same element as the model image of a specific shape from the target image. In the case of U-net, it is mainly used for image segmentation in the medical field, and the contracting path corresponding to the U-shaped left algorithm (the part that gradually decreases the image) and the right algorithm (for increasing the image) based on the center line in the U-shaped. It can be composed of Expanding path corresponding to part). The U-net can improve the slow speed, which is a problem of the existing segmentation operation, because the overlap ratio of the image is small when performing the operation of recognizing the image in a predetermined unit (Patch) for the target image. to be. That is, since the image recognition of the overlapped area is not performed, the U-net-based segmentation operation can be expected to increase the speed compared to the conventional method. In addition, in the case of the segmentation operation using U-net, problems that occur when the recognition unit (patch) size is increased or decreased can be prevented, and accordingly, localization while simultaneously verifying the output of several layers. It is possible to obtain an appropriate effect in both and in the context.

Based on the score calculated by the information analysis unit 133, not only the user's cognitive impairment but also the severity of the cognitive impairment can be determined, and the rate of deterioration of cognitive impairment symptoms (e.g., dementia) can be determined through periodic examination results comparison. I can.

The information analysis unit 133 may synthesize result values analyzed by each of the angle determination unit 133a, the cross determination unit 133b, and the closure determination unit 133c in order to calculate the user's test result as a score. Hereinafter, the angle determination unit 133a, the cross determination unit 133b, and the closing determination unit 133b will be described in order. According to an embodiment, the information analysis unit 133 sequentially performs an angle determination operation performed by the angle determination unit 133a, a cross determination operation performed by the cross determination unit 133b, and a closing determination operation performed by the closing determination unit 133b. In addition, a score for each operation step may be assigned according to a predetermined standard.

According to various embodiments of the present disclosure, the information analysis unit 133 may include a separate operation unit that calculates by summing scores calculated in each of the three operation steps.

1. Angle judgment unit

The angle determination unit 133a may perform an angle determination step (angle determination operation) of determining the number of all angles in the image drawn by the user. If the image is drawn the same as the presented image, the number of angles composed of two pentagons must be 10. The angle determination unit 133a may assign a higher score as the number of all angles displayed in the image data approaches 10.

1-1. Sensor data information extraction operation

The angle determination unit 133a may primarily perform an operation of extracting sensor data information to determine the number of angles on the image. First, the angle determination unit 133a may extract data such as time, coordinates, speed, and acceleration related to drawing a picture from sensor data. The angle determination unit 133a may temporarily store the sampled data in a storage unit for later comparison.

1-2. Segmentation behavior

In addition, the angle determination unit 133a may perform a segmentation operation based on deep learning on image data. The angle determination unit 133a may generate a second image (segmented image) as a result of the segmentation operation on the image data. In addition, only a pentagonal object may be detected and displayed on the secondary image. In this case, the segmentation operation may be implemented to detect each pentagonal object instead of detecting an image in which two pentagons overlap.

12 and 13 illustrate examples of a result of extracting a secondary image from the angle determination unit 133a according to an embodiment. First, FIG. 12 shows a secondary image when two pentagons are detected as a result of performing a segmentation simulation.

Meanwhile, FIG. 13 shows secondary images when two pentagons are not detected. In 1301 of FIG. 13, there is no pentagon detected in the secondary image because there is no pentagon in the picture drawn by the user. In 1302, since the user drew a picture in which the pentagon and the square overlap, only one pentagon is detected in the secondary image. Is shown. Similarly, in 1303 and 1304, the secondary image in which only the object corresponding to the pentagon is detected among the pictures drawn by the user is shown.

As such, the segmentation operation performed by the angle determination unit 133a may extract only a pentagon from the original image using a U-net model, which is an algorithm based on deep learning.

1-3. Coordinate comparison operation

When the second image is generated, the angle determination unit 133a may perform a coordinate comparison operation of comparing coordinates on the second image and coordinates of sensor data. Through this, the angle determination unit 133a can determine the presence or absence of a pentagon and the number of pentagons.

First, as illustrated in FIG. 14, the angle determination unit 133a may apply a ruler for coordinate detection to a secondary image (segmented image) generated as a result of the segmentation operation, and extract coordinate information. Thereafter, the angle determination unit 133a may overlay and display the coordinate information extracted from the sensor data and the coordinate information extracted from the secondary image as shown in 1401. The blue information shown in 1401 may refer to coordinate information extracted from sensor data, and the red information may refer to coordinate information (segmented data) of a secondary image as a result of segmentation.

The angle determination unit 133a may determine a ratio at which both coordinate data overlap after overlaying the coordinates of both sides (segmented data coordinates and sensor data side coordinates). In this case, the angle determination unit 133a may determine an overlap ratio by separating the left object and the right object, respectively.

As shown in 1501 of FIG. 15, the angle determination unit 133a may determine an overlapping ratio for both objects. In addition, the angle determination unit 133a may determine an overlap ratio for the left object (see 1502) and an overlap ratio for the right object (see 1503), respectively. When the result of the U-net-based extraction operation in which only the pentagon is learned to be extracted by the angle determination unit 133a and the overlapping ratio of the sensor data is greater than or equal to a reference value, the angle determination unit 133a may distinguish whether or not a pentagon is present.

The angle determination unit 133a is a diagram for a case where 1502 is a diagram for calculating a ratio of a coordinate value having a small distance from the total number of left data, and 1503 is a diagram corresponding to a case of calculating a ratio of a small distance sitting value from the total number of data on the right. It is a drawing.

For example, in 1502, the overlap ratio of coordinate values with a small distance in the total number of data on the left object (eg (total distance made by sensor data coordinates)/(total distance made by segmented data coordinates)) is 91.208%, and In 1503, a coordinate value having a small distance from the total number of data of the right object may have an overlap ratio of 90.0.

Assuming that the reference value of the overlapping ratio is 88%, since the overlapping ratios of the two objects determined in 1502 and 1503 are both equal to or greater than the reference value, both objects may be determined to be pentagons. According to various embodiments of the present disclosure, the reference value of the overlapping ratio is not specified and may be changed and set by an administrator. For example, the reference value of the overlapping ratio may be set to 95% or more.

1-4. Operation of judging the number of pentagons

The angle determination unit 133a may determine an object having an overlapping ratio equal to or greater than a reference value as a pentagon based on the coordinate comparison operation (overlay display operation, overlap ratio determination operation). Accordingly, the angle determination unit 133a may determine whether the number of pentagons present in the image drawn by the user and the number of pentagons in the presented image are the same (whether the number of pentagons is two).

If the angle determination unit 133a determines that the number of pentagons in the picture drawn by the user is the same (2) as the presented image, the number of angles becomes 10 accordingly, and thus scores corresponding to 10 angle conditions may be calculated. The score calculated according to an embodiment may follow the criteria shown in FIG. 22. For reference, in FIG. 22, when the number of angles is 10, it is shown that 4 points are assigned.

1-5. Extraction of the number of angles of each left and right figure

When it is determined that the number of pentagons is not two through a coordinate comparison operation, the angle determination unit 133a may extract the number of angles of each of the left and right figures and determine the number of angles of each object.

Specifically, the angle determination unit 133a may perform a smoothing operation on the speed graph to extract a peak from the speed graph converted from sensor data. In this case, the speed graph may be generated based on x,y coordinate values (or moving distance) for each sampling period. For convenience, a speed graph converted from sensor data before a smoothing operation is performed will be referred to as a first speed graph, and a speed graph in which the smoothing operation is performed will be referred to as a second speed graph. The smoothing operation may be performed to extract the correct number of peaks by removing unnecessary noise on the graph.

The angle determination unit 133a may classify a left object (a figure) and a right object according to a characteristic of a peak extracted from the second speed graph after the smoothing operation.

For example, the angle determination unit 133a determines whether the object is the same based on the density of individual peaks (e.g., if the interval between peaks has a certain density, it is determined as the same object, and the interval between peaks is separated by an interval greater than or equal to a reference value). In this case, it can be determined as an object that is not the same.

In addition, as illustrated in FIG. 16, the angle determination unit 133a may determine the number of angles of the figure based on the number of peaks present in the second speed graph.

Specifically, the angle determination unit 133a may calculate each of the number of angles of the left object and the number of angles of the right object based on the number of left and right peaks starting from the center of the graph.

The operation of calculating the number of angles according to the number of peaks is a calculation method based on the feature that the speed decreases when most people draw an angled part of a figure and then increases the speed again near a straight line. Accordingly, the angle determination unit 133a can detect the number of angles in a situation in which two pentagons are not detected.

In this way, after first extracting the number of pentagons, the reason for determining the number of angles by the number of peaks on the speed graph is that the number of peaks on the speed graph is equal to the perfect score of 10. This is to prevent errors that can be provided with the same perfect score even for a picture consisting of a single square.

During scoring, the angle determination unit 133a may determine the number of angles as 9 (10ㅁ1) in the case of a picture in which, for example, a pentagon and a square are drawn, and 3 points are assigned according to the criteria shown in FIG. I can. Similarly, the angle determination unit 133a may assign 2 points when the number of angles to be checked corresponds to 10ㅁ2, 1 point when the number of angles to be checked is between 7 and 5, and 0 points when the number of angles is less than 5 and 13 or more. .

2. Cross judgment

Hereinafter, the cross determination step performed by the cross determination unit 133b will be described. Specifically, an operation of determining information related to the intersection area of the picture drawn by the user in the intersection determination unit 133b and assigning points accordingly will be described. The intersecting determination step may include a first intersecting determination step of determining whether an intersecting area exists, and a second intersecting determination step of determining a distance between two objects.

2-1. Sensor data information extraction operation

The intersection determination unit 133b may extract drawing-related information including time, coordinate data, speed, and acceleration sensed while drawing from the sensor data. Furthermore, the intersection determination unit 133b may extract information such as a relationship between a sampling time and a coordinate (change of position coordinates according to the lapse of measurement time) from sensor data.

2-2. Segmentation behavior

After the sensor data information extraction operation, the intersection determination unit 133b may perform a segmentation operation of extracting an image of an area where two pentagons overlap from the image data of the response information through a deep learning-based (e.g., U-net) algorithm. have. The model image trained by the deep learning-based algorithm in order for the intersection determination unit 133b to extract the image through the segmentation operation is, for example, as shown in FIG. 17, the rightmost angle of the left pentagon and the leftmost angle of the right pentagon. This may be an image including a closed rhombus shape.

When the intersection determination unit 133b performs segmentation from the upper left image data shown in FIG. 18, as a result, a secondary image at the upper right may be obtained. In the case of the upper left image, since there is an overlapped region (intersection region) in the center of both pentagons having similar characteristics to the training image shown in FIG. 17, the image of the intersection region can be detected as a secondary image as in the upper right image. .

Meanwhile, when the cross determination unit 133b performs segmentation on the lower left image of FIG. 18, a valid secondary image may not be generated as in the lower right image. As described above, if the same shape as the model image does not exist in the image data to be segmented, the intersection area where the two pentagons overlap is not detected.

2-3. Coordinate comparison operation

When the intersection determination unit 133b generates the secondary image by performing segmentation, the coordinates of the secondary image may be extracted thereafter. In addition, the intersection determination unit 133b performs an operation of comparing the coordinates of the secondary image with respect to the intersection area with the coordinates of the sensor data, and determines whether the two pentagons intersect based on the result. As with the coordinate comparison operation performed by the angle determination unit 133a, the coordinate comparison operation performed by the intersection determination unit 133b is whether the overlapping ratio of the coordinate data (coordinates extracted from the secondary image, the coordinates extracted from the sensor data) is greater than or equal to the reference value. It may be an operation to check. When the overlapping ratio is greater than or equal to the reference value as a result of the coordinate comparison operation, the intersection determination unit 133b may determine that an intersection region, which is an area where two pentagons overlap each other, exists. When it is determined that the intersection region 133b exists, the intersection determination unit 133b may assign a perfect score (eg, 4 points) as a score in the intersection determination step.

According to various embodiments of the present disclosure, the intersection determination unit 133b may subdivide and classify the shape of the intersection region, and may differentially give a score according thereto. The intersection determination unit 133b may determine whether the detected intersection area has the same shape as the example image (eg, a rhombus shape) or a different shape (eg, a triangular shape or a rectangular shape).

The intersection determination unit 133b may determine whether to detect a'same shape intersection area' through a segmentation operation and a coordinate comparison operation. That is, even if there is an intersection area where the left pentagon and the right pentagon overlap each other, the intersection determination unit 133b does not have the same shape as the model image that is the reference for segmentation (e.g., the straight part of the left pentagon and the angle part of the right pentagon In the case of an overlapping triangular area), it may not be detected.

In this case, the intersection determination unit 133b may assign a perfect score (eg, 4 points) when it is determined that the same “same shape intersection area” as the model image exists. On the other hand, when it is determined that the'different shape intersection area', which is an intersection area different from the model image, is determined to exist, the intersection determination unit 133b can assign a first subtraction score (eg, 3 points) by subtracting a certain value from the perfect score. have. In order to determine that there is a different shape intersection region, the model image used for the segmentation operation may include another example shape of the intersection region that can be created by overlapping two pentagons.

2-4. Distance judgment operation

When it is determined that the intersection region 133b does not exist, the intersection determination unit 133b may perform an operation of determining a distance between two objects (Figures).

The intersection determination unit 133b is a first comparative difference situation in which two figures are in contact without an intersection area (eg, a distance between two figures is 1 mm or less), and a second comparison difference situation in which a distance between two figures without an intersection area is less than a reference value (eg, 1 cm). , It is possible to determine a third comparative difference situation in which a distance between two figures without an intersection area exceeds a reference value (eg, 1 cm). In addition, the cross determination unit 133b may assign a differential score to each of the first to third comparative situations.

For example, the intersection determination unit 133b is a second subtraction score (eg, 2 points) in a first comparative difference situation, a third subtraction score (eg, 1 point) in a second comparison difference situation, and a third difference score in the third comparison difference situation. Deduction points (eg 0 points) can be given.

In order to specifically describe the distance determination operation performed by the intersection determination unit 133b, reference will be made to FIG. 19.

When there is no intersection area, the intersection determination unit 133b may perform a'sensor data imaging operation' that displays and image the coordinate position identified from the sensor data for each predetermined time period (e.g., 20 ms) as shown in FIG. have.

Thereafter, the intersection determination unit 133b reads the sensor data image generated by the'sensor data imaging operation'. As shown in FIG. 19, when the arrangement of two objects is parallel to the x-axis, the distance between the two objects is x It is determined based on the value coordinates.

The intersection determination unit 133b extracts the largest x-coordinate value (the left object maximum x value) from the left object (pentagon) among two figures arranged side by side, and the smallest x value (the right object Minimum x value) can be extracted. The intersection determination unit 133b may perform distance calculation from a difference in coordinates between the minimum x value of the right object and the maximum x value of the left object.

And accordingly, when the distance difference between the two objects is 0, less than or equal to the reference value (1cm), and exceeding the reference value, the intersection determination unit 133b corresponds to the first comparison difference situation, the second comparison difference situation, and the third comparison difference situation. I can judge. And it is possible to give a score corresponding to each non-difference situation.

3. Closed judgment unit

Hereinafter, a closing determination step performed by the closing determination unit 133c will be described. The closing determination unit 133c may determine whether the shape of the figure drawn by the user is a closed type. The closing determination unit 133c may differentially give a score according to whether the figure drawn by the user is a closed form or an open form.

Referring to FIG. 20, examples of a closed figure 2001 in which all solid lines are connected and an open figure 2002 in which all solid lines are connected are shown.

3-1. Graph peak check operation

The closing determination unit 133c may first perform an operation of converting sensor data into a speed change graph to determine whether the shape of the picture drawn by the user is closed. Thereafter, the closing determination unit 133c may determine whether to close/open based on the shape and number of peaks on the speed change graph.

The closing determination unit 133c may determine whether there is one point in the speed graph generated based on sensor data where a peak indicating a change in speed is calculated larger than a reference value. In this case, when there is one peak extracted with a value that is more than the reference value, the closing determination unit 133c may determine the point indicated by the peak as a boundary point between the two figures.

This is the case of the closed type drawing one side at a time when drawing a figure, compared to the degree of speed change that occurs while trying to draw the second side after drawing one side. This is because the degree of speed change will be perceived much larger. In other words, the degree of change in the drawing speed that occurs when the hand is released will have a much larger value than when the hand is not released, so if only one peak above the reference value appears, the user releases one hand while drawing. Can be inferred. Accordingly, when the number of peaks equal to or greater than the reference value is 1, the closing determination unit 133c may determine that a closed figure is drawn.

On the other hand, when the number of peaks above the reference value is not one, the closing determination unit 133c means that the number of times the hand is removed during the drawing process is many times, so that the picture to be analyzed can be determined to correspond to an open-type figure.

Further, when a plurality of peaks above the reference value on the graph are found, the closure determination unit 133c refers to a peak (hereinafter, a boundary peak), which means a boundary point between two figures based on the peak intensity and the x-axis position of the peak occurrence point. Can be judged. Further, the closing determination unit 133c may determine whether one of the two figures is closed based on whether a peak equal to or greater than the remaining reference value is concentrated on one side (left or right) based on the boundary peak. The closing determination unit 133c may determine that one figure is drawn as a closed shape and the other one figure is drawn as an open shape when peaks greater than or equal to the remaining reference values are concentrated on the left or right of the boundary peak. On the other hand, when one or more peaks equal to or greater than the reference value are distributed on both sides of the boundary peak, the closing determination unit 133c may determine that both figures are drawn as open.

The closed figure 2001 and the corresponding speed change graph and the open figure 2002 and the corresponding speed change graph are shown in FIG. 21. As shown in FIG. 21, it can be seen that in the graph corresponding to the closed shape, peaks are found only at the boundary of two pentagons, and in the graph corresponding to the open shape 2002, three peaks above the reference value are found. In the graph corresponding to the open figure 2002, one peak of similar intensity is found on both sides of the boundary of the two pentagons, which may mean that there is one open portion of each of the two pentagons.

When there is one peak equal to or greater than the reference value on the graph, the closing determination unit 133c may determine that both figures are closed and give a perfect score (eg, 2 points). In addition, when it is determined that only one figure is of a closed type, the closing determination unit 133c may assign a first subtraction score (eg, 1 point) by subtracting a predetermined value from the perfect score, and it is determined that both figures are open type. If so, a secondary deduction score (eg 0) can be given.

As such, the information analysis unit 133 may diagnose a user's cognitive impairment (dementia) based on scores assigned through the angle determination unit 133a, the cross determination unit 133b, and the closure determination unit 133c.

Although not shown in the drawing, the information analysis unit 133 may include a synthesis result calculation unit (not shown) that synthesizes the scores calculated by the angle determination unit 133a, the cross determination unit 133b, and the closing determination unit 133c. I can. When the total score of the user (subject) is equal to or higher than the reference value, the total result calculation unit may determine that it is normal, and when the total score is less than the reference value, it may determine that he suffers from a cognitive disorder (eg, dementia). In addition, when it is determined that the user is suffering from a cognitive disorder (eg, dementia), the comprehensive result calculator may determine a severity level according to a score range and guide the determined severity of the user's cognitive impairment.

According to various embodiments, the information analysis unit 133 of the present invention performs an additional score determination operation in addition to the score determination operation through the angle determination unit 133a, the cross determination unit 133b, and the closure determination unit 133c, and is calculated as a result of the score determination. Based on the total score, it is possible to diagnose the user's cognitive impairment. For example, the information analysis unit 133 may analyze and score additional features such as a rotation direction (rotation) of each figure and an arrangement direction of two figures through the additional feature determination unit 133b. According to an embodiment, the additional feature determination unit 133d may be configured not to consider a rotation less than a preset angle (eg, 45 degrees).

In addition, the additional feature determination unit 133d is also additionally performed according to the results of three-step analysis of the angle determination step, the cross determination step, and the closure determination step performed by the angle determination unit 133a, the cross determination unit 133b, and the closing determination unit 133c. It is also possible to determine whether to perform a feature determination operation. For example, the additional feature determination unit 133d may perform the additional step determination only when it is determined that the total score calculated as a result of a three-step analysis of a picture drawn by a user is less than (or above) a reference value. In addition, the determination of the additional step may include a drawing inspection and analysis operation of a figure other than a pentagon drawing inspection. Similarly, the additional feature determination unit 133d may perform a cognitive test in the form of a multiple-choice quiz, such as a language recognition test and a color determination test, in the determination of an additional step.

The additional feature determination unit 133d separately separates users who are determined to be a high-risk group for cognitive disorders based on the eating habit information and lifestyle information previously inputted by the user, even if it is determined that the user does not correspond to cognitive impairment as a result of the three-step analysis. Can be classified. In addition, the additional feature determination unit 133d may perform an additional test and analysis operation for users in a high-risk group of cognitive impairments classified separately. In this case, the additional feature determination unit 133d may provide a test problem with an increased difficulty level to users in a high-risk group of cognitive impairments classified separately. Through this, even if the user does not currently have a cognitive impairment disease, it is possible to perform a high-difficulty test for quickness and cognitive ability. When a high-risk user who has obtained a score less than the average score occurs, the additional feature determination unit 133d may later request the feedback providing unit 134 to provide preventive feedback for the high-risk user.

The artificial intelligence support unit 133e may support an artificial intelligence-based operation required for a 3-step analysis operation. In particular, the artificial intelligence support unit 133e may manage a pentagonal detection operation in the angle determination unit 133a and the intersection determination unit 133b, and a calculation operation of a CNN model (eg, U-net) required for the intersection region detection operation.

Specifically, in order to perform a pentagon detection operation in the angle determination unit 133a, if image data is provided to the artificial intelligence support unit 133e, the artificial intelligence support unit 133e performs a segmentation operation to determine the angle of the secondary image in which only the pentagon is detected. It can be provided back to part 133a. Similarly, the artificial intelligence support unit 133e may perform a segmentation operation required for extracting a cross region required by the cross determination unit 133b.

In addition, the artificial intelligence support unit 133e may set the type of image to be detected by the segmentation operation according to various embodiments. For example, if the presented image suggested to be drawn by the user is partially changed (e.g., the shape of the intersection area is changed), the artificial intelligence support unit 133e changes and sets the model image for segmentation (the image for detection) accordingly. You can ask for it. In addition, when a model image is set as a new image according to a manager's setting, the artificial intelligence support unit 133e may automatically generate a plurality of duplicate model images by rotating, reversing, and resizing the original data of the model image. Thereafter, the artificial intelligence support unit 133e may perform training based on a new model image and a plurality of duplicate model images input by an administrator. The artificial intelligence support unit 133e may apply a new model image through this process and perform a segmentation operation of detecting an image corresponding to the new model image.

6 is a block diagram showing a configuration of a feedback providing unit according to an embodiment of the present invention.

The feedback providing unit 134 according to an embodiment of the present invention may provide a feedback for preventing or delaying deterioration of a cognitive disorder to a user based on a result of analyzing and scoring a picture drawn by the user by the information analysis unit 133. In addition, the feedback providing unit 134 may perform a feedback operation to improve accuracy of an artificial intelligence-based response information analysis operation.

The feedback providing unit 134 may include a prevention training performing unit 134a, a learning data determining unit 134b, and an artificial neural network learning unit 134c.

The preventive training execution unit 134a may provide appropriate preventive training information corresponding to whether the user has a cognitive impairment, a severity level of the user's cognitive impairment, or a scoring score itself. According to various embodiments, the preventive training performing unit 134a calculates a total score obtained by combining scores scored on lifestyles such as eating habits and lifestyles of the user with scores scored by the information analysis unit (pentagonal drawing test score). , It is possible to provide information on this by calculating the type of prevention training that is determined to be suitable for the user.

For example, the prevention training performing unit 134a may provide recommended information on the user's exercise pattern, eating pattern, and drinking pattern.

Specifically, the recommended information related to the exercise pattern may be information on an exercise time, an exercise amount, and an exercise type recommended to a user. Furthermore, the preventive training performing unit 134a may be linked with a user's wearable device so that the user actually performs an exercise, checks the user's walking information, heart rate, etc., and evaluates whether or not a target amount of exercise is achieved.

Furthermore, the prevention training performing unit 134a may display the risk of cognitive impairment of the corresponding user, but when a target amount of exercise is achieved, the risk of cognitive impairment may be reduced by a preset value and displayed. Thereby, through the real-time cognitive impairment risk provision operation performed by the prevention training execution unit 134a, high-risk users with a high probability of cognitive impairment may be stimulated and at the same time motivated to establish a good lifestyle for cognitive impairment prevention. Can be given.

In addition, the prevention training performing unit 134a may set various target values for correcting a user's eating pattern (meal time, meal menu, amount of meal, etc.) in a direction favorable to prevention of cognitive impairment, and may request the user to achieve this.

In addition, the prevention training performing unit 134a may set target values for items such as the amount of alcohol consumed and the number of monthly alcohol consumption in order to correct the user's drinking pattern in a direction favorable to the prevention of cognitive impairment, and may request the user to achieve this.

The prevention training performing unit 134a may display a cognitive impairment risk by reducing a predetermined value as target values for the eating pattern and the drinking pattern are multiplied.

According to various embodiments, the prevention training performing unit 134a provides useful information for patients suspected of cognitive impairment in addition to providing cognitive impairment prevention training and prevention information to users determined to be not cognitive impairment patients. Can provide.

For example, the prevention training execution unit 134a may guide medical institution information to a user (eg, a person suspected of having a cognitive impairment) whose pentagonal drawing test score is less than or equal to a preset value.

Alternatively, the preventive training execution unit 134a provides the user with a comprehensive result calculated by comprehensively determining the additional test and pentagonal drawing test scores when performing an additional cognitive ability test through detailed examination at a hospital where a person suspected of having a cognitive impairment has been guided. Can provide. The overall result may include a more detailed description of the type of cognitive impairment the user has suffered and the severity of the cognitive impairment.

The preventive training execution unit 134a includes the type of cognitive impairment that the user has suffered in response to the comprehensive result provided by the medical institution, appropriate treatment information corresponding to the severity of the cognitive impairment of the user, method guide for delaying the deterioration rate, and attention to daily life Can provide. The preventive training execution unit 134a also provides treatment information corresponding to the severity of the cognitive impairment of the user, a method for delaying the deterioration rate, and daily life caution information to the previously registered parental device in preparation for the possibility that the user's cognitive ability is deteriorated. Can provide.

The learning material determination unit 134b may determine response information when an error occurs as a result of scoring or specific response information designated by an administrator or a user as learning material.

Image data determined as learning data, sensor data, and various graphs generated based on the sensor data may be converted to a format separately required for artificial neural network learning through a preprocessing process.

The artificial neural network learning unit 134c may perform artificial intelligence learning based on the learning data selected by the learning data determination unit 134b and subjected to standard conversion. Accordingly, the accuracy of the segmentation operation for extracting the pentagon and the intersection area may be improved.

In addition, according to various embodiments, the artificial neural network learning unit 134c may be configured with a level of shaking less than a certain level found in a line drawn by a user (e.g., a level that can be determined as shaking by hand shaking, a period of occurrence of curvature, a curve size, an angle, etc.) Case) can be performed as a straight line. For example, the artificial neural network learning unit 134c may generate an additional image (shaking version image) in which a straight line area constituting a conventional model image is changed to a curve with an acceptable level of tremor in a specific situation (determining the user's receipt). In addition, the shaking version image may be trained as a model image of U-net.

In short, in the automatic analysis method of a picture for cognitive ability test according to an embodiment of the present invention, the server transmits information required for performing a pentagon drawing test, which is a test method for determining whether or not a cognitive failure, to a user device, the Receiving, by the server, response information to the user's pentagon drawing test from the user device, performing scoring according to the result of analyzing the response information, and diagnosing the user's cognitive impairment according to the scoring result. However, the step of performing the scoring includes determining the number of angles in the image drawn by the user and assigning a score corresponding to the number of angles. A first cross determination step of determining and assigning a score corresponding thereto, a second cross determination step of determining a distance between two objects and giving a score corresponding to the distance when the cross region does not exist, the user It may include a closing determination step of determining whether the object is a closed type in the image drawn by and giving a score corresponding to the closed type.

The steps of a method or algorithm described in connection with an embodiment of the present invention may be implemented directly in hardware, implemented as a software module executed by hardware, or a combination thereof. Software modules include Random Access Memory (RAM), Read Only Memory (ROM), Erasable Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), Flash Memory, hard disk, removable disk, CD-ROM, or It may reside on any type of computer-readable recording medium well known in the art to which the present invention pertains.

Components of the present invention may be implemented as software programming or software elements, and similarly, embodiments include various algorithms implemented with a combination of data structures, processes, routines or other programming elements, including C, C++ , Java, assembler, or the like may be implemented in a programming or scripting language. Functional aspects can be implemented with an algorithm running on one or more processors.

Although the present invention has been described in detail with reference to the above-described examples, those skilled in the art can make modifications, changes, and modifications to these examples without departing from the scope of the present invention. In short, in order to achieve the intended effect of the present invention, it is not necessary to separately include all functional blocks shown in the drawings or to follow all the sequences shown in the drawings as shown in the order shown. Note that it may fall within the range.

100: server 110: communication unit
120: storage unit 130: control unit
131: test execution unit 131a: problem execution unit
131b: response information acquisition unit 132: preprocessor
132a: response information classification unit 132b: useful data conversion unit
133: information analysis unit 133a: angle determination unit
133b: cross determination unit 133c: closure determination unit
133d: additional feature determination unit 133e: artificial intelligence support unit
134: feedback providing unit 200: user device

Claims (8)

In the automatic analysis system of a picture for cognitive ability test for determining whether there is a cognitive impairment comprising a server and a user device,
The server
In order to determine whether there is a cognitive impairment, the pentagon drawing test data is provided to the user device, the user's response information corresponding to the pentagon drawing test is obtained, the obtained response information is analyzed, and the analysis result is scored. Including; a control unit for performing cognitive disorder diagnosis,
The control unit
When analyzing the response information, the analysis item includes a feature of the intersection area of two objects in the picture drawn by the user,
The characteristic of the intersection region includes information on the presence or absence of the intersection region and the distance between two objects,
The user device
An automatic analysis system for a picture for a cognitive ability test, characterized in that receiving and executing information required to perform a pentagon drawing inspection from the server through communication with the server. The method of claim 1,
The control unit
Including; an information analysis unit that performs an artificial intelligence-based score calculation operation on the response information,
The information analysis unit
An intersection determination unit that determines information related to an intersection area, which is a point where two objects intersect in the user-drawn picture, and gives a score according to the characteristics of the determined intersection area; Automatic analysis system. The method of claim 2,
The cross determination unit
Perform a segmentation operation of extracting an image of an intersection area where two objects overlap from the image data of the response information through a deep learning-based U-net algorithm, and the coordinates of the intersection area extracted through the segmentation and the response information An automatic analysis system for a picture for cognitive ability test, characterized in that it compares coordinates extracted from sensor data and determines that an intersection area exists in a picture drawn by a user when the overlapping ratio is greater than or equal to a reference value. The method of claim 3,
The cross determination unit
When it is determined that the intersection area does not exist, a sensor data imaging operation is performed to image by displaying the coordinate position identified from the sensor data at a predetermined time period, and the two objects on the image generated by the sensor data imaging operation are performed. A picture for cognitive ability test characterized by determining the distance between two objects on a picture drawn by a user by calculating the distance from the difference between the minimum x-coordinate of the right object and the maximum x-coordinate of the left object when the arrangement direction is parallel to the x-axis. Automatic analysis system. The method of claim 2
The information analysis unit
Including; an additional feature determination unit for performing analysis and scoring for additional features including the rotation direction of each figure on the picture drawn by the user and the arrangement direction of the two figures, and performing a cognitive impairment diagnosis including the result of the scoring; Automatic analysis system of pictures for cognitive ability test, characterized in that to. In the server that performs automatic analysis of pictures for cognitive ability test,
A communication unit for transmitting information required to perform a pentagon drawing test, which is a test method for determining whether a cognitive impairment exists, to a user device, and receiving response information for the user's pentagon drawing test from the user device;
A storage unit for storing cognitive ability test data including a problem for a pentagonal drawing test, a scoring criterion, a test score, and matching information of the presence of cognitive impairment and severity; And
Including; a control unit for analyzing the response information obtained from the user device and performing a scoring on the analysis result to perform cognitive disorder diagnosis; and
The control unit
When analyzing the response information, the analysis item includes a feature of the intersection area of two objects in the picture drawn by the user,
The server, characterized in that the characteristic of the intersection region includes information on the presence or absence of the intersection region and a distance between two objects. In the automatic analysis method of pictures for cognitive ability test,
Transmitting, by the server, information required for performing a pentagon drawing test, which is a test method for determining whether a cognitive impairment exists, to a user device;
Receiving, by the server, response information for a user's pentagon drawing inspection from a user device;
Performing scoring according to a result of analyzing the response information;
Including; Including,
The step of performing the scoring is
An angle determination step of determining the number of angles in the image drawn by the user and giving a score corresponding to the number of angles;
A first intersection determination step of determining whether there is an intersection area where two objects intersect in the image drawn by the user and giving a score corresponding thereto;
A second intersection determination step of determining a distance between two objects and assigning a score corresponding to the distance when the intersection area does not exist;
A closing determination step of determining whether the object is a closed type in the image drawn by the user and giving a score corresponding to the closed type; Automatic analysis method comprising a. A computer-readable recording medium storing a program for executing the method of claim 7 on a computer.

Images