System for early screening of autism spectrum disorder using eye tracking

Abstract

The present invention relates to a system for early screening of autism spectrum disorder using eye tracking. According to the present invention, autism spectrum disorder may be diagnosed early by applying response data of a user's gaze, which reacts to images or videos outputted on a display, to an autism spectrum disorder screening model. In addition, the present invention may be used to easily check a degree of suspicion of autism spectrum disorder at home, and thus may be utilized in the fields of early diagnosis, prevention, and treatment of autism in children.

Description

System for early screening of autism spectrum disorder using eye tracking

[0001] The present invention relates to a system for early screening of autism spectrum disorder using eye tracking.

[0002] Recently, the number of children with developmental disabilities, typified by the autism spectrum, has been increasing and is emerging as a social issue. While some argue that this phenomenon is caused by advancements in diagnostic techniques and stricter criteria rather than an actual increase in numbers compared to the past, there is no disagreement that early screening and diagnosis are crucial as they significantly influence the development of social and communication skills, given the nature of developmental disabilities where the likelihood of improvement increases the earlier they are detected and treated.

[0003] Korea is currently conducting national screenings for infants and toddlers to identify children with developmental disabilities, including autism spectrum disorder, and similar developmental screenings are being carried out at both national and private levels in other countries overseas.

[0004] While children exhibit behaviors and language appropriate for their age during their growth, it is difficult for parents, who lack knowledge and experience regarding developmental processes, to identify specific developmental anomalies. This is particularly challenging for experts to make a judgment when the degree of anomaly is ambiguous or the child is young. Furthermore, even specialized institutions often rely on surveys completed by caregivers and observe the child for a relatively short period to provide an opinion; as there are variations in developmental speed among children, some cases may simply involve delayed development, making it difficult to hastily diagnose autism.

[0005] Furthermore, due to insufficient social systems for developmental disabilities such as autism spectrum disorder and a shortage of relevant institutions and specialists, there are often significantly long waiting periods for professional screening and diagnostic evaluations. This can lead to missed opportunities for timely treatment, and the substantial costs involved place a heavy burden on the individual.

[0006] In addition, early intervention through early screening is associated with better intervention effectiveness and positive prognosis, and there is a continuous demand for screening tools for early screening before the age of two. However, despite this situation, there is currently no early screening tool capable of performing validated early screening on the domestic population that possesses excellent discriminative power and cost-effectiveness.

[0007] The inventors have developed an economical early diagnosis method for autism spectrum disorder with high reliability and discriminability by identifying early warning signs of autism spectrum disorder that may appear around the age of 3 through specificity in gaze patterns regarding social stimuli such as faces and social communication situations in children at high risk and diagnosed with autism spectrum disorder, and by analyzing social behaviors and restricted and repetitive behavior patterns through motion analysis technology.

[0008] One aspect is a method for screening autism performed by an autism screening device,

[0009] (a) acquiring the user's gaze coordinates for calibration through a camera while K points of different locations are output on a display; and

[0010] (b) a step of screening for autism spectrum disorder by applying the data obtained in step (a) above to a pre-established autism spectrum disorder screening model; the present invention provides a method for screening for autism spectrum disorder comprising: (b) a step of screening for autism spectrum disorder by applying the data obtained in step (a) above to a pre-established autism spectrum disorder screening model.

[0011] (a) acquiring the user's gaze coordinates for calibration through a camera while K points of different locations are output on a display; and

[0012] (b) a step of acquiring response data by recording the point of gazing at unique content capable of screening for autism spectrum disorder; and

[0013] The present invention provides a method for screening autism spectrum disorder, comprising the step of screening autism spectrum disorder by applying gaze coordinates obtained in step (a) and response data obtained in step (b) to a pre-established autism spectrum disorder screening model.

[0014] Another aspect is in an autism spectrum disorder screening device based on user eye-tracking,

[0015] A user monitoring unit that monitors the user's gaze and movement through a camera while an image or video is displayed on a screen and acquires the user's real-time gaze coordinates;

[0016] A time series data derivation unit that derives time series response data by recording the time intervals during which the user gazed at an element of interest on the display based on the above real-time gaze coordinates; and

[0017] The present invention provides an autism spectrum disorder diagnostic device comprising: a determination unit that determines whether there is an autism spectrum disorder by applying time-series response data derived in response to the above-mentioned user to a pre-established autism spectrum disorder screening model.

[0018] According to the present invention, early diagnosis of autism spectrum disorder can be achieved by applying response data regarding a user's gaze responding to an image or video output on a display to an autism spectrum disorder screening model.

[0019] Furthermore, this invention enables the relatively easy checking of the suspected degree of autism spectrum disorder in infants and toddlers at home or in medical settings, making it usable as a tool for the early diagnosis of childhood autism, which is a major social issue.

[0020] Figure 1 is a diagram showing a system for early screening of autism spectrum disorder using eye tracking.

[0021] Figure 2 is a diagram showing the configuration of an early screening method for autism spectrum disorder using eye tracking.

[0022] Figure 3 is a figure showing the K point locations for acquiring calibration data for eye tracking.

[0023] Figure 4 is a flowchart of a calibration method for autism spectrum disorder according to an embodiment.

[0024] Figure 5 is a flowchart of an inference method for autism spectrum disorder according to an embodiment.

[0025] Figure 6 is an example of a social / non-social object test content screen.

[0026] Figure 7 is an example of an inspection screen for LAI (Low autism interest objects) / HAI (High autism interest objects) content.

[0027] Figure 8 is an example of a test screen of the Joint Attention Test content.

[0028] The present invention will be explained in more detail below through examples. However, these examples are intended to illustrate the invention and the scope of the invention is not limited to these examples. Furthermore, in order to clearly explain the invention in the drawings, parts unrelated to the explanation have been omitted, and similar parts throughout the specification have been given similar reference numerals.

[0029] Throughout the specification, when a part is described as being "connected" to another part, this includes not only cases where they are "directly connected," but also cases where they are "electrically connected" with other components interposed between them. Furthermore, when a part is described as "including" a certain component, this means that, unless specifically stated otherwise, it does not exclude other components but may include additional components.

[0030]

[0031] The present invention relates to a method and device for diagnosing autism spectrum disorder based on user gaze monitoring. It proposes a technique capable of determining and diagnosing the possibility of autism spectrum disorder by tracking the position of a user's gaze while viewing a display using AI, and applying the user's gaze response data, which reacts to an image or video being displayed on the display, to a pre-trained autism spectrum disorder diagnostic model.

[0032] One aspect is a method for screening autism performed by an autism screening device,

[0033] (a) acquiring the user's gaze coordinates for calibration through a camera while K points of different locations are output on a display; and

[0034] (b) a step of acquiring response data by recording the point of gazing at unique content capable of screening for autism spectrum disorder; and

[0035] (c) a step of screening for autism spectrum disorder by applying gaze coordinates obtained in step (a) and response data obtained in step (b) to a pre-established autism spectrum disorder screening model; the present invention provides a method for screening for autism spectrum disorder, comprising: (c) a step of screening for autism spectrum disorder by applying gaze coordinates obtained in step (a) and response data obtained in step (b) to a pre-established autism spectrum disorder screening model.

[0036] FIG. 1 is a diagram illustrating an autism spectrum disorder screening system based on user gaze monitoring according to an embodiment of the present invention.

[0037] As shown in FIG. 1, an autism spectrum disorder screening system based on user gaze monitoring according to an embodiment of the present invention includes a camera (100), a display (200), and an autism spectrum disorder screening system (300), and may further include a network or a user terminal.

[0038] The autism spectrum disorder screening system (300) can be connected to the camera (100) and the display (200) via a wired, wireless, or combined wired-wireless network to transmit and receive information to and from each other. The wireless network may include at least one of RF, WLAN, Wi-Fi, and Bluetooth methods, and various known wireless network methods may be used.

[0039] The autism spectrum disorder screening system (300) may correspond to a management server (Server) for providing a service to determine and diagnose the possibility of autism spectrum disorder from time-series response data including the gaze of a user watching image or video content.

[0040] The autism spectrum disorder screening system (300) can operate in conjunction with the camera (100) and the display (200), and can also control the operation of the camera (100) and the display (200). In addition, the autism spectrum disorder screening system (300) can provide the user with time-series response data collected while viewing an image or video, and an analysis result of the possibility of autism spectrum disorder derived based thereon.

[0041] The autism spectrum disorder screening system (300) can provide autism diagnosis service platform implemented as an application or web to a network-connected user, and can also provide it through a display (200). In addition, the service platform may be an application running in a mobile app or web environment.

[0042] The autism spectrum disorder screening system (300) can transmit and receive information through a user terminal and a network, and the user terminal may include a device capable of exchanging information by connecting to a wired or wireless network, such as a PC, desktop, smartphone, tablet, or notebook. The display (200) may be a standard TV, display device, or portable display, as well as a display device connected to a network and linked with a set-top box.

[0043] The autism spectrum disorder screening system (300) can register personal information about a subject through a display (200) or a user terminal, and can provide the results of determining the possibility of autism spectrum disorder for each video content derived by providing at least one video content to the subject in conjunction with the camera (100) and the display (200) to the display (200) or the user terminal.

[0044] Here, the autism spectrum disorder screening system (300) may provide a result of determining the possibility of autism spectrum disorder and a diagnosis result as a probability value for the possibility of autism spectrum disorder, and may provide a result of determining autism spectrum disorder or normal depending on whether the probability value is greater than or equal to a threshold value.

[0045] The autism spectrum disorder screening system (300) may be physically configured and implemented as a computer device including a processor, memory, user interface input / output device and storage device, network input / output unit, etc., or may be implemented as an application program executed on a computer device or user terminal.

[0046] In one embodiment of the present invention, the camera (100) can photograph a user in front of the display (200) while installed on the display (200) and provide the captured image to an autism spectrum disorder screening system (300). Here, the camera (100) may be configured to be integrated with the display (200) or may be separately mounted on the outside of the display (200).

[0047] The autism spectrum disorder screening system (300) can monitor the user's gaze and movement using images captured by the camera (100) and obtain the user's real-time gaze coordinates toward the display screen. The acquisition of gaze coordinates can be implemented by including an eye tracker function. Of course, in the case of an embodiment of the present invention, the camera may be implemented in the form of an eye tracker.

[0048] Here, for accurate eye tracking, coordinate calibration of the position of the camera (100) based on the position of the display (200) can be performed in advance. Since techniques for tracking the gaze of a user looking at a display screen using the camera's image are known in various ways, the known techniques can be used.

[0049] In one embodiment of the present invention, the eye tracker may be a Tobii eye tracker. By using the Tobii eye tracker to check the user's pupil-center corneal reflection (PCCR), the user's eye movements can be detected to obtain data such as where the person is looking and whether they are present.

[0050] In one embodiment of the present invention, when using a Tobii eye tracker, a camera and lighting integrated device measuring 28.5 cm x 1.5 cm x 2.2 cm is attached to the bottom of the monitor to illuminate near-infrared light invisible to the human eye, and gaze data such as the user's gaze coordinates can be obtained by tracking the gaze at 60 Hz through two infrared cameras.

[0051] In one embodiment, the gaze coordinates obtained in step (a) and the response data obtained in step (b) may be obtained by a camera by shining near-infrared light on the user to acquire pattern data reflected from the pupil and cornea.

[0052] In an embodiment of the present invention, the autism spectrum disorder screening system (300) first determines whether a user is present by detecting the movement of a user in front of the display through the image of the camera (100) while video content is being displayed on the display (200). For motion detection, various known video-based motion detection techniques may be used.

[0053] If user movement is detected above a threshold in front of the display screen, it can be determined that a user is present. In this case, the autism spectrum disorder screening system (300) detects the user's eyes within the image and secondarily determines whether the coordinates of the gaze point are located within the viewing range (display screen area) of the screen.

[0054] If the user's gaze point is within the viewing range, the autism spectrum disorder screening system (300) obtains the user's real-time gaze coordinates while performing gaze tracking based on an eye tracking algorithm. In addition, if the gaze point is located within the viewing range, the user's real-time movement values ​​detected in the image of the camera (100) can be additionally obtained along with the gaze tracking.

[0055] In an embodiment of the present invention, the autism spectrum disorder screening system (300) can monitor the user's gaze and movement while K points of different locations are displayed on the display (200) through the camera (100) and obtain the user's real-time gaze coordinates.

[0056] In an embodiment of the present invention, the autism spectrum disorder screening system (300) can monitor the user's gaze and movement toward one or more elements of interest selected from a group consisting of shapes, objects, characters, or people appearing on the display (200) at each hour while image or video content is output on the display (200) through the camera (100), and can obtain the user's real-time gaze coordinates.

[0057] Additionally, the autism spectrum disorder screening system (300) can derive visual response and pattern data of the user by extracting the entire area and a partial area as elements of interest corresponding to each shape or object appearing on the display (200), or by extracting a plurality of areas selected from the character or person area, character or person torso area, head area, face area, eye area, mouth area, cheek area, forehead area, ear area, hand area, and shoulder area corresponding to each character or person.

[0058] At this time, it is self-evident that the area that can be extracted as an element of interest from one or more selected from a group consisting of shapes, objects, characters, or figures is not necessarily limited to the listed elements, and any area on the area where shapes, objects, characters, or figures are displayed can be the subject.

[0059] In one embodiment of the present invention, the element of interest may be a K point. A K point may mean a shape or coordinate displayed at a certain position on a display (200).

[0060] For example, as illustrated in Fig. 4, the autism spectrum disorder screening system (300) can analyze the K-point screen currently being output on the display (200) at each time interval (e.g., frame, unit time) to derive visual response and pattern data of the user at that time.

[0061] In one embodiment of the present invention, the spectrum disorder diagnosis method may further include the step of (d) obtaining result data by applying the real-time gaze coordinates to a fine-tune model.

[0062] In one embodiment, the fine-tuning model in step (d) may be a model that is pre-trained with images for object recognition and then additionally trained with images for autism spectrum disorder diagnosis. This is a task to accurately track the user's gaze coordinates on the monitor.

[0063] The above fine-tuning model may be a model trained through pre-training, which involves training an AI model with a large dataset (e.g., numerous web texts, images) to acquire general knowledge such as language structures or image features, and fine-tuning, which involves additionally training the model with a very small amount of specific dataset while keeping the weights of the model equipped with general knowledge fixed.

[0064] For example, with reference to FIG. 5, the autism spectrum disorder screening system (300) can analyze i) a screen containing social images and non-social images (see FIG. 6), ii) a screen containing LAI objects and HAI objects (see FIG. 7), or iii) a video for joint attention testing (see FIG. 8) currently being output on the display (200) on an hourly basis to derive visual response and pattern data of the user at that time.

[0065] In one embodiment of the present invention, the spectrum disorder screening method

[0066] (e) A step of monitoring the user's gaze while social and non-social images are simultaneously displayed on a display and acquiring the user's real-time gaze coordinates;

[0067] (f) A step of monitoring the user's gaze while images of LAI objects and HAI objects are simultaneously displayed on a display and acquiring the user's real-time gaze coordinates;

[0068] (g) A step of monitoring the user's gaze while a joint attention test image is displayed on a display and acquiring the user's real-time gaze coordinates;

[0069] (h) A step of diagnosing autism spectrum disorder by applying the real-time gaze coordinates of (e) to (g) above to a pre-established autism spectrum disorder screening model; may be further included.

[0070] In one embodiment, steps (e) to (h) may be performed when the accuracy determined by step (c) is 0.50 to 1.0.

[0071] In one embodiment, the social image of step (e) is an image of a person appearing with various facial expressions, and the non-social image may be a figure or pattern image. For example, the social image may be displayed as an image of a person smiling, and the non-social image may be displayed as a simple figure or pattern, and may be composed of 30 to 60 seconds.

[0072] In one embodiment of the present invention, step (f) may be a step of obtaining data on the user's gaze by measuring how often and for how long the user looks at specific targets on the display (200).

[0073] In one embodiment, the step (f) may be a step of obtaining data on the gaze of a user exploring an image in which LAI (Low autism interest objects) and HAI (High autism interest objects) objects are displayed on a display (200) to the user when LAI objects of the same size are hidden and when HAI objects are shown at different ratios.

[0074] In this invention, LAI (Low autism interest objects) refers to objects that individuals with autism spectrum disorder do not look at well or have no interest in.

[0075] In this invention, HAI (High autism interest objects) refers to objects of high interest to individuals with autism spectrum disorder.

[0076] Examples of LAI and HAI in the present invention may be the objects presented in Table 1 below, but are not limited thereto (see Table 1).

[0077] HAILAI Transportation: Car, train, bus, airplane Food: Cookies, pizza, pasta, fried rice, chicken Electronics / Machines: Mouse, laptop, printer, gas stove, speaker Clothing and Accessories: Shoes, t-shirt, pants, socks, gloves Symbols: Watch, calendar, sign, license plate, alphabet Furniture: Chair, dining table, sofa Animals: Dinosaur, lion Plants: Flowerpot, wood Toys: Ball, puzzle, blocks Stationery / Bathroom Supplies: Notebook, scissors, toothbrush, towel, cup

[0078] In one embodiment, the above step (f) may additionally obtain data on i) whether the reaction time to the image of the HAI object is fixed at 600ms or more compared to the reaction time to the image of the LAI object, or ii) whether the HAI is looked at 0.2 times more frequently than the LAI.

[0079] In one embodiment, data can be obtained regarding i) whether the reaction time to the image of the HAI object is fixed for 600ms or more compared to the reaction time to the image of the LAI object, for example, 600ms to 3000ms, 600ms to 1500ms, or 600ms to 1000ms.

[0080] In one embodiment, data can be obtained regarding whether the time spent gazing at the image of the HAI object is 1.2 times or more than the time spent gazing at the image of the LAI object, for example, 1.2 to 3.0 times, 1.2 to 2.0 times, or 1.2 to 1.5 times more often.

[0081] In one embodiment, the above step (g) may be a step of obtaining the user's real-time gaze coordinates for an object that a person in the video for a Joint Attention Test speaks to, looks at, or points to.

[0082] In this invention, the Joint Attention Test is a test that verifies the ability of one or more people to recognize that they are sharing attention to a single external object (thing, event).

[0083] In addition, by applying the user's visual response and pattern data separated from the video on an hourly basis in this manner to a pre-established autism spectrum disorder screening model, historical data on the user's response to the corresponding image or video can be collected.

[0084] Generally, children with autism spectrum disorder may exhibit abnormal reactions while watching or gazing at images or videos, such as fixating their gaze on a single object rather than the content, focusing their gaze on areas unrelated to the topic, or being unresponsive or overresponsive to music or sound.

[0085] Considering these points, if the characteristics of response data collected from typical children and autistic children while viewing specific images or videos are learned in advance, it becomes possible to screen for or diagnose the possibility of autism spectrum disorder in a specific child based on the recorded response data when various arbitrary images or videos are provided to that child.

[0086] In one embodiment of the present invention, in addition to gaze response, whether a user's movement responds to the output of sound (e.g., voice, horn sound, etc.) in the video can be additionally utilized as data for determining autism spectrum disorder.

[0087] To this end, the autism spectrum disorder screening system (300) may include time-series sound response data, which records time intervals in which movement is detected while a sound value greater than a threshold is output, based on sound data of a video being output on a display (200) and a real-time movement value of a user monitored through a camera (100), in the time-series response data and provide it as input data for an autism spectrum disorder screening model. At this time, the interval in which a movement value greater than a preset threshold is observed can be determined as the interval in which movement is detected. In this way, the time-series response data may include visual response and sound response data.

[0088] Next, the autism spectrum disorder screening system (300) can diagnose autism spectrum disorder by applying time-series response data derived in response to the user to a pre-established autism spectrum disorder screening model.

[0089] Here, the autism spectrum disorder screening model may be a model that tracks gaze using time-series user response data collected individually during the viewing of image or video content as input data, targeting multiple subjects, namely those with and without autism spectrum disorder, respectively.

[0090] The autism spectrum disorder screening system (300) can provide result data including whether or not there is an autism spectrum disorder as well as the probability of possibility, and can provide the determination result by outputting it through a display (200) or a user terminal.

[0091] FIG. 2 is a diagram illustrating a method for diagnosing autism spectrum disorder according to an embodiment of the present invention.

[0092] First, the autism spectrum disorder screening system (300) monitors the user's gaze and movement through the camera (100) while K points are output through the display (200) (S110).

[0093] Here, the autism spectrum disorder screening system (300) detects the movement value of a user in front of the display through the image of the camera (100) to first determine whether the user is present, and if the user is present as a result of the determination, it performs eye detection of the user and can determine whether the point of gaze of the eye is located within the viewing range of the screen.

[0094] When the autism spectrum disorder screening system (300) confirms that the user's gaze point is within the viewing range of the screen, it performs gaze tracking to obtain the user's real-time gaze coordinates (S120).

[0095] In this S120 step, real-time movement values ​​of the user can be obtained from the camera footage along with eye tracking.

[0096] And, the autism spectrum disorder screening system (300) extracts multiple interest elements related to shapes, objects, characters, or people appearing in the image or video of the display (200) every hour and applies the user's visual response and pattern data to a pre-built autism spectrum disorder screening model to determine whether there is a possibility of autism spectrum disorder (S110 to S140).

[0097] An autism spectrum disorder screening system (300) can obtain time-series response data of a user by checking and recording which part of the screen the user is looking at at every moment while viewing an image or video based on the user's real-time gaze coordinate data. This time-series response data may have a time-series pattern form, and time-series feature patterns collected from people with and without autism spectrum disorder can be learned in advance based on deep learning in an autism spectrum disorder screening model.

[0098] Additionally, the autism spectrum disorder screening system (300) can provide the determination result by outputting it through a display (200) or a user terminal. Furthermore, the autism spectrum disorder screening system (300) can generate a diagnostic assistance report for the user based on the determination result and transmit it to an affiliated external institution or a relevant clinician.

[0099]

[0100] According to the present invention, the presence of autism spectrum disorder can be conveniently screened or diagnosed simply by inputting the user's time-series response data obtained while watching video content into a pre-established screening model.

[0101] According to the present invention, autistic children can be easily diagnosed at home, medical staff can provide high-quality medical services more efficiently, and the problem of treating autistic children can be solved through more universalized diagnostic services.

[0102] Furthermore, by equipping televisions or video devices used daily at home with such devices, it is possible to naturally screen for children at high risk of autism spectrum disorder in each household. In addition, if autism spectrum disorder, which has a prevalence rate of about 2%, can be screened so easily, it will not only reduce individual suffering by providing opportunities for early treatment but also provide enormous benefits to the national economy and public health.

[0103] Furthermore, if a child is diagnosed with developmental delays during an infant screening, the proposed device allows parents to self-assess the suspected degree of autism spectrum disorder at home. In particular, while many parents worry about their babies watching too many videos from a young age, the proposed device offers the significant advantage of allowing parents to naturally monitor the baby's condition during routine video viewing and restrict viewing early if any abnormal reactions are detected.

[0104] The present invention has been described with reference to the embodiments illustrated in the drawings, but this is merely illustrative, and those skilled in the art will understand that various modifications and equivalent alternative embodiments are possible therefrom. Accordingly, the true technical scope of protection of the present invention should be determined by the technical spirit of the appended claims.

[0105] [Explanation of the symbol]

[0106] 100 : Camera

[0107] 200 : Display

[0108] 300: Autism Spectrum Disorder Screening System

Claims

1. A method for screening autism performed by an autism screening device, (a) acquiring the user's gaze coordinates for calibration through a camera while K points of different locations are output on a display; and (b) a step of acquiring response data by recording the point of gazing at unique content capable of screening for autism spectrum disorder; and (c) a step of screening for autism spectrum disorder by applying gaze coordinates obtained in step (a) and response data obtained in step (b) to a pre-established autism spectrum disorder screening model; comprising a method for screening for autism spectrum disorder.

2. In Claim 1, A screening method for autism spectrum disorder, wherein the gaze coordinates obtained in step (a) and the response data obtained in step (b) are obtained by using a camera to acquire pattern data reflected from the pupil and cornea by shining near-infrared light on the user.

3. In Claim 1, (d) a step of obtaining result data by applying the above real-time gaze coordinates to a fine-tune model; further comprising a method for screening autism spectrum disorder.

4. In Claim 3, A screening method for autism spectrum disorder, wherein in step (d) above, the fine-tuning model is a model that is pre-trained with images for object recognition and then additionally trained with images for autism spectrum disorder diagnosis.

5. In Claim 1, (e) A step of monitoring the user's gaze and movement while social and non-social images are simultaneously displayed on a display and acquiring the user's real-time gaze coordinates; (f) A step of monitoring the user's gaze and movement while images of LAI objects and HAI objects are simultaneously displayed on a display and acquiring the user's real-time gaze coordinates; (g) A step of monitoring the user's gaze and movement while a joint gaze test image is displayed on a display and acquiring the user's real-time gaze coordinates; (h) a step of diagnosing autism spectrum disorder by applying the real-time gaze coordinates of (f) to (h) above to a previously constructed autism spectrum disorder screening model; further comprising an autism spectrum disorder screening method.

6. In Claim 5, A screening method for autism spectrum disorder, wherein the social image in step (e) above is an image featuring one or more people, and the non-social image is a figure or pattern image.

7. In Claim 5, A screening method for autism spectrum disorder, wherein in step (f) above, additional data is obtained regarding i) whether the reaction time to the image of the HAI object is fixed at 600ms or more compared to the reaction time to the image of the LAI object, or ii) whether the HAI is gazed at 0.2 times more frequently than the LAI.

8. In Claim 5, A screening method for autism spectrum disorder, wherein the above step (g) is a step of obtaining the user's real-time gaze coordinates for an object that a person in a video for a joint gaze test speaks to, looks at, or points to.

9. In an autism spectrum disorder screening device based on user eye gaze monitoring, A user monitoring unit that monitors the user's gaze and movement through a camera while an image or video is displayed on a screen and acquires the user's real-time gaze coordinates; A time series data derivation unit that derives time series response data by recording the time intervals during which the user gazed at an element of interest on the display based on the above real-time gaze coordinates; and A screening device for autism spectrum disorder comprising: a discrimination unit that determines whether there is an autism spectrum disorder by applying time-series response data derived in response to the above-mentioned user to a pre-established screening model for autism spectrum disorder.

Images