Generative artificial intelligence-based autism family intervention scheme generation method

Abstract

The invention belongs to the technical field of artificial intelligence, and discloses an autism family intervention scheme generation method based on generative artificial intelligence. The method comprises the steps of performing multi-round questions and answers with objects with different intimacy degrees through a Transform model, extracting key behaviors and frequency information, evaluating a social communication capability level, combining an evaluation result of a CNN model on a social interaction video, comprehensively obtaining the capability level, enabling the capability level to correspond to a development milestone, and generating an intervention scheme containing elements such as an execution object and time arrangement. Meanwhile, the ability level is adjusted according to identity information such as patient age. In the execution process, real-time behavior data are collected, the real-time social communication capability level is evaluated through the first neural network model, an intervention scheme is adjusted and updated through the second neural network model according to the deviation value between the real-time behavior data and an intervention target, and accurate and scientific autism behavior correction is achieved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method for generating an autism family intervention program based on generative artificial intelligence. Background Art

[0002] Autism spectrum disorder (ASD) is a neurodevelopmental disorder characterized by social interaction disorders, language expression disorders, and repetitive stereotyped behaviors. For preschool children with ASD, social communication disorders have a significant impact on their long-term development, making it difficult for them to establish normal social relationships with children of the same age, affecting their social adaptability.

[0003] Traditional methods of generating autism intervention programs rarely refer to the natural development laws of children in setting goals, and rarely follow the natural development sequence of social communication skills to conduct interventions. As a result, they face problems such as difficulty tracking intervention progress, difficulty quantifying intervention effects, and difficulty applying intervention results in the real world. At the same time, it is often difficult to customize programs based on the child's situation and the needs of each family, and it is also difficult to update intervention programs in a timely and efficient manner. Summary of the invention

[0004] The purpose of the present invention is to provide a method, medium and device for generating autism family intervention plan based on generative artificial intelligence, so as to at least partially solve the above-mentioned problems.

[0005] According to one aspect of the present application, a method for generating autism family intervention plans based on generative artificial intelligence is proposed, comprising: conducting multiple rounds of questions and answers with a first subject through a Transformer model, extracting a first key behavior of the child with autism and its frequency information based on the first subject's answer, and evaluating a first ability level of the social communication ability of the child with autism based at least on the first key behavior and its frequency information; matching the first ability level with a milestone in the development of social communication ability, and generating corresponding intervention goals, and generating a corrective intervention plan at least based on the intervention goals.

[0006] Optionally, the method also includes conducting multiple rounds of questions and answers with the second object through the Transformer model, extracting the second key behavior of the autistic child and its frequency information based on the second object's answer, and evaluating the second ability level of the social communication ability of the autistic child at least based on the second key behavior and its frequency information; wherein the first object and the second object are objects with different degrees of intimacy with the autistic child respectively; taking a weighted average of the first ability level and the second ability level to obtain a third ability level; wherein the higher the intimacy between the object and the child, the higher the weight corresponding to the ability level obtained based on the evaluation of the object; matching the third ability level with the milestones of social communication ability development, and generating corresponding intervention goals, and generating a corrective intervention plan at least based on the intervention goals.

[0007] Optionally, the method also includes obtaining video information of social interaction behavior of children with autism, and evaluating the fourth ability level of the social communication ability of children with autism through a CNN model, integrating the fourth ability level with the first ability level or the third ability level to obtain a fifth ability level, corresponding the fifth ability level with the milestones of social communication ability development, and generating corresponding intervention goals, and at least generating a corrective intervention plan based on the intervention goals; the intervention plan at least includes the objects of the intervention plan, the time schedule for the implementation of the intervention plan, the intervention goals for the children, the intervention scenarios, and the intervention doses.

[0008] Optionally, the method further includes obtaining identity information of the autistic patient, the identity information including at least age; and adjusting the ability level based on the identity information, wherein age is negatively correlated with the ability level.

[0009] Optionally, the method also includes recording real-time behavioral data of executing the intervention plan, and based on the real-time behavioral data, evaluating the real-time social communication ability level through a first neural network model, wherein the real-time behavioral data includes at least the child's behavioral data, the actual effective delivery dose, the intervention video clips, and the actual self-assessment data.

[0010] Optionally, the method also includes adjusting and updating the intervention plan through a second neural network model based on the deviation value between the real-time social communication ability level and the intervention target and the intervention plan.

[0011] The present application also provides a computer-readable storage medium storing a computer program, including the steps in the above method executed when the computer program is executed by a processor.

[0012] An embodiment of the present application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the steps in the above method by calling the computer program stored in the memory.

[0013] The present application discloses a method for generating autism family intervention programs based on generative artificial intelligence. The method includes: using the Transformer model to conduct multiple rounds of questions and answers with objects of different intimacy levels, extracting key behaviors and frequency information, evaluating the level of social communication ability, and then combining the evaluation results of the social interaction video with the CNN model to comprehensively obtain the ability level and correspond to the development milestones, and generate an intervention program containing elements such as the execution object and time schedule. At the same time, the ability level is adjusted according to the patient's age and other identity information. During execution, real-time behavioral data is collected, and the real-time social communication ability level is evaluated by the first neural network model. Then, according to the deviation value from the intervention target, the intervention plan is adjusted and updated through the second neural network model to achieve accurate and scientific autism behavior correction. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 A schematic diagram of a method for generating an autism family intervention plan based on generative artificial intelligence provided in an embodiment of the present application.

[0015] Figure 2 A schematic diagram of a system for generating autism family intervention plans based on generative artificial intelligence provided in an embodiment of the present application.

[0016] Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0017] The specific embodiments of the present invention are further described below in conjunction with the accompanying drawings. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0018] It should be noted that, in the present application, "first", "second" and various numerical numbers indicate distinctions made for the convenience of description, and are not used to limit the scope of the embodiments of the present application. For example, different classification results are distinguished, rather than being used to describe a specific order or sequence. It should be understood that the objects described in this way can be interchanged where appropriate, so as to be able to describe solutions other than the embodiments of the present application.

[0019] Specifically, Figure 1 The following is a flowchart showing a method for generating autism family intervention plan based on generative artificial intelligence according to an embodiment of the present application. Figure 1The present application provides a method for generating an autism family intervention plan based on generative artificial intelligence, and the specific steps are as follows: S101, conducting multiple rounds of questions and answers with a first subject through a Transformer model, extracting a first key behavior of autistic children and its frequency information based on the first subject's answer, and evaluating a first ability level of the social communication ability of the autistic child based on at least the first key behavior and its frequency information.

[0020] In this embodiment, the first object can be the parent of a child with autism, such as a mother. Through the Transformer model, multiple rounds of questions and answers are conducted with the mother of the child, such as asking "If the child wants to eat the cake on the table, but he can't reach it, will he look at you? What is the probability that the child will take the initiative to look at you in similar situations every day?" (i.e., the probability of eye contact), "Does the child take the initiative to initiate a conversation when he sees something he is interested in every day? What is the probability of taking the initiative to find you for a conversation in similar situations?" (i.e., the probability of active conversation), for example, the mother replied: "Similar situations occur an average of 5 times a day, the child looks at me for help an average of about 3 times, and takes the initiative to initiate a conversation 2 times." Based on the mother's reply, the child's eye contact 3 times when in need is extracted, and he can take the initiative to initiate a conversation 2 times. The Transformer model can generate new questions based at least on the mother's reply, such as "When you call the child's name, is there a verbal or non-verbal response? How often?" The mother replied: "There is a 50% probability of turning your head when calling the name." Until enough features are extracted. It is understandable that the key behavior can be a variety of behaviors related to the evaluation of social communication ability, which is not limited in this embodiment. The first ability level of the social communication ability of the autistic child is obtained by at least evaluating the information through a pre-set scoring rule.

[0021] S102, matching the first ability level with the social communication ability development milestone, and generating corresponding intervention goals, and generating a corrective intervention plan at least based on the intervention goals.

[0022] In one embodiment, an example of predefined scoring rules is as follows: Table 1 Predefined scoring rules:

[0023] In this embodiment, the total score is 3 points, which is mapped to the milestones of social communication ability development. The milestones are divided into: L1 (0-1 points): basic social awakening intervention is required.

[0024] L2 (2-3 points): Non-verbal communication reinforcement is required.

[0025] L3 (4 points): Requires training in complex social scenarios.

[0026] Evaluation result: The first ability level is L2.

[0027] Based on the L2 level, for example, a goal is generated: "Increase the probability of non-verbal response to 70%", and a plan is output: Execution object: primary caregiver, such as a professional rehabilitation therapist, parent of the child, etc.

[0028] Time arrangement: From Monday to Friday, there is 30 minutes of one-on-one interaction after lunch every day, and family interaction activities are arranged on Saturdays and Sundays.

[0029] Intervention settings: structured games (such as asking for help with a puzzle), or activities in outdoor settings, etc.

[0030] Intervention dose: 7 times per week, 10 trigger opportunities each time.

[0031] In some possible embodiments, in order to avoid the overly subjective judgment caused by the special emotional bias of relatives, especially parents, towards their children, which leads to the distortion of the assessment of the ability level of the child, it is necessary to refer to the question and answer data of others, such as kindergarten teachers, for evaluation. Therefore, this embodiment, on the basis of the previous embodiment, also includes multiple rounds of questions and answers with the second object. It is assumed that the second object is the kindergarten teacher of the child, and the intimacy with the child is lower than that of the mother. Similarly, multiple rounds of questions and answers are conducted with the teacher through the Transformer model, such as "In class, will the child take the initiative to ask others for his needs when he has needs, and what is the probability of asking for his needs when encountering such a situation?", "Will the child share his interests with other children, and what is the probability of sharing interests?". The second key behavior of the child with autism and its frequency information are extracted based on the teacher's reply. For example, the probability of the child taking the initiative to ask his needs every day within a week is 20%, and the probability of taking the initiative to share his interests with the children is 50%. Assuming that the second ability level is L1 based on the predefined scoring rules, the scoring rules of this embodiment can be predefined with reference to the method of the previous embodiment, and this embodiment is not limited. In addition, due to the high degree of intimacy between the mother and the child, the weight of the ability level obtained based on the mother's conversation is relatively high, and the weight of the ability level obtained based on the teacher's conversation is relatively low. For example, the level weight corresponding to the mother's conversation is 0.7, and the level weight corresponding to the teacher's conversation is 0.3. The first ability level and the second ability level are weighted averaged to obtain the third ability level, that is, the third ability level = mother data weight (0.7) × L2 + teacher data weight (0.3) × L1 = 0.7 × 2 + 0.3 × 1 = 1.7 → L1.7 (rounded down to L1). Generate corresponding intervention targets based on the L1 level, and at least generate corrective intervention plans based on the intervention targets. For specific intervention targets and corresponding corrective intervention plans, please refer to the previous embodiment for setting, which will not be repeated here.

[0032] By collecting longitudinal behavioral data of children provided by parents and / or primary caregivers, the limitations of traditional one-on-one assessments that only collect cross-sectional data and assessment results that are easily affected by the current state of the children are avoided, ensuring the accuracy of behavioral data and assessments of children's social communication abilities.

[0033] In another embodiment, in order to further improve the accuracy of evaluating the social ability level of children, in addition to the subjective question and answer data of the caregivers of the children in the aforementioned embodiment, this embodiment also introduces the behavioral video information data of the children. The behavioral video information data of the children is relatively objective and can eliminate the deviation caused by some subjective factors. Video information of social interaction behaviors of children with autism in family gatherings and / or kindergartens is obtained. The video is analyzed by a trained CNN model, for example, the duration and method of interaction between the child and others in the video are detected. For example, the frequency of non-verbal communication behaviors (such as pointing to objects): 2 times / 10 minutes, and the duration of shared attention: an average of 8 seconds. The fourth ability level of the social communication ability of children with autism is evaluated, for example, level L2. The fourth capability level is integrated with the third capability level. The integration method can adopt the existing technical method, which is not limited in this embodiment. For example, it can be linear integration. The weight of the third capability level is 0.4, and the weight of the fourth capability level is 0.6. After integration, the fifth capability level is obtained, that is, the fifth capability level = the weight of the third capability level (0.4) × L1 + the weight of the fourth capability level (0.6) × L2 = 0.4 × 1 + 0.6 × 2 = 1.6 → L1.6 (rounded down to L1). The corresponding intervention target is generated based on the L1 level, and at least a corrective intervention plan is generated based on the intervention target. The specific intervention target and the corresponding corrective intervention plan can be set with reference to the previous embodiment, and will not be repeated here.

[0034] By collecting longitudinal behavioral data of children provided by parents and / or primary caregivers of children, as well as cross-sectional behavioral data of children provided by video information of children, the accuracy of behavioral data and assessment of children's social communication ability can be further improved.

[0035] In some possible embodiments, the identity information of the autistic patient is obtained, where the identity information can be age, for example, Xiao Ming, an 8-year-old patient. Based on age, since the older the patient is, the stronger the social communication ability should be in theory, and 8 years old is relatively old in the autistic patient group, it has a certain negative impact on the ability level. For example, for an 8-year-old patient, it is assumed that the original assessment is L2. According to the adjustment rule: Age adjustment coefficient : 0 (age < 5 years), -0.3 (age between 5 and 7 years), -0.5 (age > 7 years).

[0036] The calibration formula is: ,in, Indicates the level after calibration, For the original level, is the age adjustment factor.

[0037] Based on the calibration formula, the calibrated level (rounded down to L1), a corresponding intervention target is generated based on the L1 level, and a correction intervention plan is generated based on the intervention target. The specific intervention target and the corresponding correction intervention plan can be set with reference to the previous embodiment, and will not be repeated here.

[0038] In some embodiments, the identity information includes the patient's gender and personality in addition to age. Generally speaking, male patients are more proactive than female patients, and patients with strong social motivation are more willing to socialize than those with weak social motivation. Therefore, male patients or patients with strong social motivation need to adjust their ability level downward. For example, the patient Xiao Xia is 5 years old, male, and has a strong social motivation personality. Assume that the original assessment is L2. According to the adjustment rule: Age adjustment coefficient : 0 (age < 5 years), -0.3 (age between 5 and 7 years), -0.5 (age > 7 years).

[0039] Gender adjustment factor :-0.2.

[0040] Personality Adjustment Factor :-0.1.

[0041] The calibration formula is: ,in, Indicates the level after calibration, For the original level, is the age adjustment factor, is the gender adjustment coefficient, is the personality adjustment factor.

[0042] Based on the calibration formula, the calibrated level (rounded down to L1), a corresponding intervention target is generated based on the L1 level, and a correction intervention plan is generated based on the intervention target. The specific intervention target and the corresponding correction intervention plan can be set with reference to the previous embodiment, and will not be repeated here.

[0043] In some possible embodiments, real-time tracking of intervention effects is also included. Exemplarily, during the implementation of the intervention plan, the following data is collected in real time through the combination of IoT devices and manual records: Child behavior data: Non-verbal response delay time (such as the time from the issuance of instructions to eye contact), example: average delay of 5 seconds (standard deviation ± 2 seconds).

[0044] Completion rate of target behavior (such as the number of times toys are actively shared), example: 3 times / hour (the preset target is 5 times / hour).

[0045] Practical effective delivery dose: the number of intervention triggers actually completed per day (i.e. the effective delivery dose, such as the number of times parents successfully guide their children to make requests). Example: 12 times / day were planned, but 9 times were actually completed.

[0046] Duration of a single intervention session, example: average 8 minutes (target 10 minutes).

[0047] Analysis of intervention video clips: Key scenes (such as social game interactions) are captured through edge computing devices, and a lightweight CNN model is used to extract: the facial orientation angle of the child (an angle >45° with the interactor is defined as distraction), example: distraction frequency 40%.

[0048] Practical self-assessment data: The intervention implementer fills out the execution log every day. Quantitative indicators include but are not limited to: the correct rate of use of intervention strategies (based on self-assessment), example: 78%.

[0049] These rich real-time behavior data are input into the trained first neural network model. The model contains multiple hidden layers, and each hidden layer extracts and converts the input data through a specific activation function. After layer-by-layer calculation, the real-time social communication ability level is finally evaluated as L2 through the output layer. The first neural network model can adopt an existing model, which is not limited in this embodiment.

[0050] Calculate the deviation between the real-time social communication ability level L1 and the ability level corresponding to the intervention target (assuming that the ability level corresponding to the intervention target is L3). Input the deviation value and the current intervention plan into the trained second neural network model. The second neural network model comprehensively adjusts the intervention plan according to the deviation, the current intervention plan and the intervention target. The second neural network model can adopt an existing model, which is not limited in this embodiment.

[0051] The adjusted plan, for example, includes but is not limited to: In terms of training time, in addition to increasing the daily intervention time from 30 minutes to 40 minutes, the training time will be reasonably allocated according to the patient's fatigue and attention span. For example, a 40-minute game can be divided into two 20-minute sections with a 5-minute break in between to improve the training effect.

[0052] As the difficulty of interactive games increases, they will be carefully designed according to the current ability level of the children. For example, the original interactive game was a simple building block game, which has now been upgraded to a building block game with illusion attributes, requiring children to not only build blocks, but also give certain real-world meanings to the blocks they build and describe them, so as to better promote the improvement of children's social communication skills.

[0053] Expand the scenarios, such as adding a new community supermarket shopping scenario, and increase participation in more novel environmental stimuli.

[0054] In addition, the training content will be adjusted based on the interests and shortcomings of the children as reflected in the real-time behavioral data. If the children are found to be highly interested in topics related to animals, more content about animal knowledge and habits will be added in the subsequent intervention; if the children have difficulty expressing complex sentences, special training will be designed specifically, such as demonstrating more complex sentences and providing different types of assistance for different types of complex sentences.

[0055] Corresponding to the method for generating autism family intervention plan based on generative artificial intelligence in the above embodiment, Figure 2 A structural block diagram of an autism behavior correction intervention program generation system based on generative artificial intelligence provided in an embodiment of the present application is shown. For ease of explanation, only the parts related to the embodiment of the present application are shown.

[0056] See also Figure 2 As shown, an autism behavior correction intervention program generation system 200 based on generative artificial intelligence provided by an embodiment of the present application includes: a first evaluation unit, used to conduct multiple rounds of questions and answers with a first subject through a Transformer model, extract a first key behavior of an autistic child and its frequency information based on the answer of the first subject, and evaluate and obtain a first ability level of the social communication ability of the autistic child based on at least the first key behavior and its frequency information; an intervention program generation unit, used to correspond the first ability level with the development milestone of social communication ability, and generate corresponding intervention goals, and generate a correction intervention plan at least based on the intervention goals.

[0057] Optionally, the system also includes a second evaluation unit, which is used to conduct multiple rounds of questions and answers with the second object through the Transformer model, extract the second key behavior of the autistic child and its frequency information based on the second object's answer, and evaluate the second ability level of the social communication ability of the autistic child based on at least the second key behavior and its frequency information; wherein the first object and the second object are objects with different degrees of intimacy with the autistic child respectively; a third unit, which is used to weighted average the first ability level and the second ability level to obtain a third ability level; wherein the higher the intimacy between the object and the child, the higher the weight corresponding to the ability level obtained based on the evaluation of the object; a second intervention plan generation unit, which is used to correspond the third ability level with the social communication ability development milestone, and generate corresponding intervention goals, and at least generate a corrective intervention plan based on the intervention goals.

[0058] Accordingly, an embodiment of the present application further provides an electronic device, which may be a terminal or a server. Figure 3 As shown, Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. For ease of explanation, only the parts related to the embodiment of the present application are shown.

[0059] The electronic device 300 includes a processor 301 having one or more processing cores, a memory 302 having one or more computer-readable storage media, and a computer program stored in the memory 302 and executable on the processor. The processor 301 is electrically connected to the memory 302. Those skilled in the art will appreciate that the electronic device structure shown in the figure does not constitute a limitation on the electronic device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0060] The processor 301 is the control center of the electronic device 300. It uses various interfaces and lines to connect various parts of the entire electronic device 300, and executes various functions of the electronic device 300 and processes data by running or loading software programs (computer programs) and / or units stored in the memory 302, and calling data stored in the memory 302, thereby monitoring the electronic device 300 as a whole.

[0061] In the embodiment of the present application, the processor 301 in the electronic device 300 will load the instructions corresponding to the processes of one or more applications into the memory 302 according to the following steps, and the processor 301 will run the application stored in the memory 302 to achieve various functions: conduct multiple rounds of questions and answers with the first object through the Transformer model, extract the first key behavior and frequency information of the autistic child based on the answer of the first object, and evaluate the first ability level of the social communication ability of the autistic child based on at least the first key behavior and its frequency information; correspond the first ability level with the development milestone of social communication ability, and generate corresponding intervention goals, and generate a corrective intervention plan at least based on the intervention goals.

[0062] The specific implementation of the above operations can be found in the aforementioned embodiments, which will not be described in detail here.

[0063] Optional, such as Figure 3 As shown, the electronic device 300 further includes: an intervention plan generation module 303, a communication module 304, an input unit 305 and a power supply 306. The processor 301 is electrically connected to the intervention plan generation module 303, the communication module 304, the input unit 305 and the power supply 306 respectively. Those skilled in the art can understand that Figure 3 The electronic device structure shown in the figure does not constitute a limitation of the electronic device, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0064] The intervention program generation module 303 can be used to generate an autism behavior correction intervention program based on generative artificial intelligence.

[0065] The communication module 304 can be used to communicate with other devices.

[0066] The input unit 305 may be used to receive input numbers, character information or user feature information (such as fingerprint, iris, facial information, etc.), and generate keyboard, mouse, joystick, optical or trackball signal input related to user settings and function control.

[0067] The power supply 306 is used to supply power to various components of the electronic device 300. Optionally, the power supply 306 can be logically connected to the processor 301 through a power management system, so that the power management system can manage charging, discharging, power consumption, and other functions. The power supply 306 can also include one or more DC or AC power supplies, recharging systems, power failure detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0068] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0069] A person of ordinary skill in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by instructions, or by controlling related hardware through instructions. The instructions may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0070] To this end, an embodiment of the present application provides a computer-readable storage medium, which stores multiple computer programs. The computer programs can be loaded by a processor to execute the steps of a method for generating an autism family intervention plan based on generative artificial intelligence provided by an embodiment of the present application.

[0071] The specific implementation of the above operations can be found in the previous embodiments, which will not be described in detail here.

[0072] The computer-readable storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0073] Since the computer program stored in the storage medium can execute the steps in any one of the methods for generating autism family intervention programs based on generative artificial intelligence provided in the embodiments of the present application, the beneficial effects of any one of the methods for generating autism family intervention programs based on generative artificial intelligence provided in the embodiments of the present application can be achieved. For details, please see the previous embodiments and will not be repeated here.

[0074] The present application is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present application. These computer program instructions may be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the flowchart. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0075] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0076] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0077] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. For those skilled in the art, various changes, modifications, substitutions and variations of these embodiments are made without departing from the principles and spirit of the present invention, and still fall within the scope of protection of the present invention.

Claims

1. A method for generating autism family intervention programs based on generative artificial intelligence, characterized in that: include: Conducting multiple rounds of question-and-answer sessions with the first subject through a Transformer model, extracting first key behaviors of the child with autism and information about their frequency based on the first subject's responses, and evaluating a first ability level of the child with autism's social communication ability based at least on the first key behaviors and information about their frequency; The first ability level is matched with a social communication ability development milestone, and a corresponding intervention goal is generated, and a corrective intervention plan is generated at least based on the intervention goal.

2. According to claim 1, a method for generating autism family intervention programs based on generative artificial intelligence is characterized in that: The method also includes: conducting multiple rounds of questions and answers with a second object through the Transformer model, extracting a second key behavior of the autistic child and its frequency information based on the second object's answer, and evaluating a second ability level of the social communication ability of the autistic child based at least on the second key behavior and its frequency information; wherein the first object and the second object are objects with different degrees of intimacy with the autistic child respectively; taking a weighted average of the first ability level and the second ability level to obtain a third ability level; wherein the higher the intimacy between the object and the child, the higher the weight corresponding to the ability level obtained based on the object's evaluation; matching the third ability level with a milestone in the development of social communication ability, and generating corresponding intervention targets, and generating the correction intervention plan at least based on the intervention targets.

3. The method for generating autism family intervention programs based on generative artificial intelligence according to claim 2, characterized in that: The method also includes: obtaining video information of social interaction behavior of children with autism, and obtaining a fourth ability level of the social communication ability of the children with autism through CNN model evaluation, fusing the fourth ability level with the first ability level or the third ability level to obtain a fifth ability level, corresponding the fifth ability level with a milestone in the development of social communication ability, and generating corresponding intervention goals, and generating the corrective intervention plan at least based on the intervention goals; the intervention plan at least includes the implementation object of the intervention plan, the execution schedule of the intervention plan, the intervention goal of the child, the intervention scenario, and the intervention dose.

4. The method for generating autism family intervention plan based on generative artificial intelligence according to claim 3, characterized in that: The method further includes: obtaining identity information of the autistic patient, the identity information including at least age; and adjusting the ability level based on the identity information, wherein age is negatively correlated with the ability level.

5. The method for generating autism family intervention plan based on generative artificial intelligence according to claim 4, characterized in that: The method also includes: the intervention implementer records the real-time behavioral data of executing the intervention plan, and based on the real-time behavioral data, obtains the real-time social communication ability level through the first neural network model evaluation, wherein the real-time behavioral data at least includes the child's behavioral data, the actual effective delivery dose, the intervention video clips, and the actual self-assessment data.

6. The method for generating autism family intervention programs based on generative artificial intelligence according to claim 5, characterized in that: The method further includes: based on the deviation value between the real-time social communication ability level and the intervention target and the intervention plan, adjusting and updating the intervention plan through a second neural network model.

7. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 6 is executed.

8. An electronic device, characterized in that: It comprises a memory storing executable program code and a processor coupled to the memory; wherein the processor calls the executable program code stored in the memory to execute the method according to any one of claims 1-6.

Images