Cognitive level parameter evaluation method and device, computer equipment and storage medium

Abstract

The invention relates to a cognitive level parameter evaluation method and device, computer equipment and a storage medium. Comprising the following steps: determining a cognitive score of a tester, and determining a brain activity sequence of the tester according to test functional magnetic resonance imaging data; according to the brain activity sequence of the testee, constructing an emotion memory network of the testee, and determining a test brain state corresponding to a network activation mode of the emotion memory network; determining mode similarity between network activation modes according to the tested brain state, and determining a predictive characterization map according to the mode similarity; determining a test receptive field and a predictive index of the predictive characterization map according to the predictive characterization map, and determining a cognitive iconography marker of the tester according to the predictive index; constructing a cognitive level prediction model according to the cognitive score and the cognitive iconography mark through a machine learning algorithm; and determining cognitive level parameters of the target person through the cognitive level prediction model. The accuracy of the cognitive evaluation result of the target person can be improved.

Description

Technical Field

[0001] The present application relates to the field of cognitive imaging, and in particular to a cognitive level parameter evaluation method, apparatus, computer equipment and storage medium. Background Art

[0002] Cognitive impairment, with its high incidence, high disability rate and high recurrence rate, has placed a huge burden on society and individuals. Cognitive impairment may manifest as persistent low mood, decreased vitality, decreased appetite, decreased sleep quality, and slower perception and movement. In addition, the learning, memory and attention of people with cognitive impairment may also be impaired. However, the current assessment of cognitive impairment in individuals mainly relies on the individual's cognitive performance and scale scores. The assessment cycle is long and lacks objective quantitative indicators, and the accuracy of the assessment of individual cognitive level parameters is low. Therefore, how to improve the efficiency of the assessment of individual cognitive level parameters and improve the accuracy of the assessment of cognitive level parameters is a problem that needs to be solved. Summary of the invention

[0003] Based on this, it is necessary to provide a cognitive level parameter evaluation method, device, computer equipment and storage medium that can improve the evaluation efficiency of individual cognitive level parameters and improve the accuracy of cognitive level parameter evaluation in response to the above technical problems.

[0004] In a first aspect, the present application provides a method for evaluating cognitive level parameters, the method comprising: Determining a cognitive score of a test person, and test functional magnetic resonance imaging data of the test person when the test person watches a candidate movie, and determining a brain activity sequence of the test person based on the test functional magnetic resonance imaging data; Constructing an emotional memory network of the tester according to the brain activity sequence of the tester, determining a network activation pattern of the emotional memory network, and determining a test brain state corresponding to the network activation pattern; Determining a test activation sequence of the emotional memory network according to the test brain state, determining pattern similarity between network activation patterns according to the test activation sequence, generating a high-dimensional representation relationship structure according to the pattern similarity, and generating a predictive representation map according to the high-dimensional representation relationship structure; Determining a test receptive field according to the predictive representation map, determining a predictive index of the predictive representation map according to the test receptive field, and determining a cognitive imaging marker of the test person according to the predictive index; Constructing a cognitive level prediction model based on the cognitive score and the cognitive imaging markers of the test subject through a machine learning algorithm; The cognitive level parameters of the target person are determined by the cognitive level prediction model according to the target functional magnetic resonance imaging data when the target person watches the target movie clip.

[0005] In one embodiment, the cognitive level parameter evaluation method further includes: Determine the vocabulary feature vectors of the candidate movies, and construct a correlation network structure between the vocabulary feature vectors based on the Pearson correlation coefficients between the vocabulary feature vectors; A correlation analysis is performed on the correlation network structure and the predictive representation map, and a regulatory paradigm of the film genre of the candidate film on the cognitive level is constructed according to the analysis results.

[0006] In one embodiment, determining a test receptive field according to the predictive representation map, and determining a predictive index of the predictive representation map according to the test receptive field, includes: Determine a test receptive field according to the predictive representation map, determine a curve window of an activation curve corresponding to the test receptive field by a sliding window method, determine the kurtosis in the window corresponding to the curve window, and determine a candidate window from the curve window according to the kurtosis in the window; The candidate windows are de-overlapped to determine a target window, and a predictive index of the predictive representation map is determined according to the skewness value of the target window.

[0007] In one embodiment, the machine learning algorithm includes a support vector machine and a ridge regression model, and the cognitive level prediction model is constructed according to the cognitive score and the cognitive imaging markers of the test person through the machine learning algorithm, including: Building a classification model based on the cognitive score and the cognitive imaging markers of the test person by a support vector machine; Constructing a model to be tested according to the cognitive score and the cognitive imaging markers of the test person through a ridge regression model; The reliability of the model to be detected is verified by a cross-validation method. If the verification is passed, the model to be detected is used as the target model, and a cognitive level prediction model is constructed according to the classification model and the target model.

[0008] In one embodiment, a classification model is constructed based on the cognitive score and the cognitive imaging markers of the test person by using a support vector machine, including: Determine a training data set and a test data set according to the cognitive score and the cognitive imaging markers of the test person; The training data set is used to train a support vector machine to determine a candidate model, and the accuracy of the candidate model is verified using the test data set. If the verification passes, the candidate model is used as a classification model.

[0009] In one embodiment, the cognitive level parameters of the target person are determined by the cognitive level prediction model according to the target functional magnetic resonance imaging data when the target person watches the target movie clip, including: determining a brain activity sequence of the target person based on target functional magnetic resonance imaging data when the target person watches a target movie clip; constructing a target memory network of the target person according to the brain activity sequence of the target person, determining a target activation pattern according to the target memory network, and determining a target brain state corresponding to the target activation pattern; Determining a target activation sequence of the target memory network according to the target brain state, and generating a target prediction representation map according to the target activation sequence; Determine a target receptive field according to the target prediction representation map, determine a target prediction index according to the target receptive field, and determine a cognitive imaging marker of the target person according to the target prediction index; The cognitive level parameters of the target person are determined according to the cognitive imaging markers of the target person through the cognitive level prediction model.

[0010] In one embodiment, generating a predictive representation graph according to the high-dimensional representation relationship structure includes: Determine a diagonal degree matrix according to the high-dimensional representation relationship structure, and determine a transfer matrix according to the high-dimensional representation relationship structure and the diagonal degree matrix; A predictive representation map is determined based on the transfer matrix, the identity matrix, and the discount factor.

[0011] In a second aspect, the present application also provides a cognitive level parameter evaluation device, the device comprising: A brain activity sequence determination module, used to determine the cognitive score of the test person and the test functional magnetic resonance imaging data of the test person when watching the candidate movie, and determine the brain activity sequence of the test person according to the test functional magnetic resonance imaging data; A brain state determination module, used to construct an emotional memory network of the tester according to the brain activity sequence of the tester, determine a network activation pattern of the emotional memory network, and determine a test brain state corresponding to the network activation pattern; a representation map determination module, configured to determine a test activation sequence of the emotional memory network according to the test brain state, determine pattern similarity between network activation patterns according to the test activation sequence, generate a high-dimensional representation relationship structure according to the pattern similarity, and generate a predictive representation map according to the high-dimensional representation relationship structure; An imaging marker determination module, used to determine a test receptive field according to the predictive representation map, determine a predictive index of the predictive representation map according to the test receptive field, and determine a cognitive imaging marker of the test person according to the predictive index; A model training module, used for constructing a cognitive level prediction model according to the cognitive score and the cognitive imaging markers of the test person through a machine learning algorithm; The cognitive level parameter evaluation module is used to determine the cognitive level parameters of the target person according to the target functional magnetic resonance imaging data when the target person watches the target movie clip through the cognitive level prediction model.

[0012] In a third aspect, the present application further provides a computer device, the computer device comprising a memory and a processor, the memory storing a computer program, and the processor implementing the following steps when executing the computer program: Determining a cognitive score of a test person, and test functional magnetic resonance imaging data of the test person when the test person watches a candidate movie, and determining a brain activity sequence of the test person based on the test functional magnetic resonance imaging data; Constructing an emotional memory network of the tester according to the brain activity sequence of the tester, determining a network activation pattern of the emotional memory network, and determining a test brain state corresponding to the network activation pattern; Determining a test activation sequence of the emotional memory network according to the test brain state, determining pattern similarity between network activation patterns according to the test activation sequence, generating a high-dimensional representation relationship structure according to the pattern similarity, and generating a predictive representation map according to the high-dimensional representation relationship structure; Determining a test receptive field according to the predictive representation map, determining a predictive index of the predictive representation map according to the test receptive field, and determining a cognitive imaging marker of the test person according to the predictive index; Constructing a cognitive level prediction model based on the cognitive score and the cognitive imaging markers of the test subject through a machine learning algorithm; The cognitive level parameters of the target person are determined by the cognitive level prediction model according to the target functional magnetic resonance imaging data when the target person watches the target movie clip.

[0013] In a fourth aspect, the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented: Determining a cognitive score of a test person, and test functional magnetic resonance imaging data of the test person when the test person watches a candidate movie, and determining a brain activity sequence of the test person based on the test functional magnetic resonance imaging data; Constructing an emotional memory network of the tester according to the brain activity sequence of the tester, determining a network activation pattern of the emotional memory network, and determining a test brain state corresponding to the network activation pattern; Determining a test activation sequence of the emotional memory network according to the test brain state, determining pattern similarity between network activation patterns according to the test activation sequence, generating a high-dimensional representation relationship structure according to the pattern similarity, and generating a predictive representation map according to the high-dimensional representation relationship structure; Determining a test receptive field according to the predictive representation map, determining a predictive index of the predictive representation map according to the test receptive field, and determining a cognitive imaging marker of the test person according to the predictive index; Constructing a cognitive level prediction model based on the cognitive score and the cognitive imaging markers of the test subject through a machine learning algorithm; The cognitive level parameters of the target person are determined by the cognitive level prediction model according to the target functional magnetic resonance imaging data when the target person watches the target movie clip.

[0014] The above-mentioned cognitive level parameter evaluation method, device, computer equipment and storage medium determine the cognitive score of the tester, as well as the test functional magnetic resonance imaging data of the tester when watching the candidate movie, and determine the brain activity sequence of the tester based on the test functional magnetic resonance imaging data; construct the emotional memory network of the tester based on the brain activity sequence of the tester, and determine the network activation mode of the emotional memory network, and determine the test brain state corresponding to the network activation mode; determine the test activation sequence of the emotional memory network based on the test brain state, determine the pattern similarity between the network activation patterns based on the test activation sequence, and generate a high-dimensional representation relationship structure based on the pattern similarity, and generate a predictive representation map based on the high-dimensional representation relationship structure; determine the test receptive field based on the predictive representation map, and determine the predictive index of the predictive representation map based on the test receptive field, and determine the cognitive imaging markers of the tester based on the predictive index; construct a cognitive level prediction model based on the cognitive score and the cognitive imaging markers of the tester through a machine learning algorithm; determine the cognitive level parameters of the target person based on the target functional magnetic resonance imaging data when the person watches the target movie clip through the cognitive level prediction model. This solves the problem that the current assessment of cognitive impairment in individuals mainly relies on the individual's cognitive performance and scale scores, the assessment cycle is long and lacks objective quantitative indicators, and the accuracy of the assessment of the individual's cognitive level parameters is low. The above scheme, based on the fMRI data of the emotional and memory circuits of the tester's brain, combines modeling methods such as brain network dynamic analysis and cognitive map representation to explore the potential imaging markers of individual cognition and study the neural mechanisms that affect the changes in cognitive level parameters. Through machine learning algorithms, the cognitive level prediction model is trained based on the tester's fMRI data and the tester's cognitive score. The cognitive score of the target person is predicted through the cognitive level prediction model, which can improve the efficiency of cognitive assessment and the accuracy of cognitive assessment results when conducting cognitive assessment on the target person. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A diagram showing an application environment of a cognitive level parameter evaluation method in one embodiment; Figure 2 Schematic diagram of a process of evaluating a cognitive level parameter in one embodiment; Figure 3 An example diagram of determining a test receptive field according to a predictive representation map in one embodiment; Figure 4 is a flow chart of a cognitive level parameter evaluation method in another embodiment; Figure 5 is a flow chart of a cognitive level parameter evaluation method in another embodiment; Figure 6 is a flow chart of a cognitive level parameter evaluation method in another embodiment; Figure 7 is a flow chart of a cognitive level parameter evaluation method in another embodiment; Figure 8 is a structural block diagram of a cognitive level parameter evaluation device in one embodiment; Fig. 9 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0017] The cognitive level parameter evaluation method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown, the terminal 102 communicates with the server 104 through a network. The data storage system can store data that the server 104 needs to process. The data storage system can be integrated on the server 104, or placed on the cloud or other network servers. The server 104 determines the cognitive score of the tester and the test functional magnetic resonance imaging data when the tester watches the candidate movie, and determines the brain activity sequence of the tester based on the test functional magnetic resonance imaging data; constructs the emotional memory network of the tester based on the brain activity sequence of the tester, determines the network activation pattern of the emotional memory network, and determines the test brain state corresponding to the network activation pattern; determines the test activation sequence of the emotional memory network based on the test brain state, determines the pattern similarity between the network activation patterns based on the test activation sequence, generates a high-dimensional representation relationship structure based on the pattern similarity, and generates a predictive representation map based on the high-dimensional representation relationship structure; determines the test receptive field based on the predictive representation map, determines the predictive index of the predictive representation map based on the test receptive field, determines the cognitive imaging marker of the tester based on the predictive index, and constructs a cognitive level prediction model based on the cognitive score and the cognitive imaging marker of the tester through a machine learning algorithm; determines the cognitive level parameters of the target person based on the target functional magnetic resonance imaging data when the target person watches the target movie clip through the cognitive level prediction model, and sends the relevant information of the cognitive level parameter evaluation to the terminal 102 through the communication network. The terminal 102 may be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart car-mounted devices, etc. The portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers.

[0018] In one embodiment, Figure 2 As shown, a method for evaluating cognitive level parameters is provided. This embodiment uses the method applied to a terminal as an example. It can be understood that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps: S210, determining the cognitive score of the test person and the test functional magnetic resonance imaging data when the test person watches the candidate movie, and determining the brain activity sequence of the test person based on the test functional magnetic resonance imaging data.

[0019] Among them, the tester is a right-handed person with normal vision, and the age of the tester is greater than 22 years old and less than 35 years old. The number of testers can be set according to actual needs, for example, it can be 50 people. The test functional magnetic resonance imaging data is the fMRI (functional MRI, fMRI, functional magnetic resonance imaging) data of the tester. It is an emerging neuroimaging method. Its principle is to use magnetic resonance imaging to measure the changes in blood dynamics caused by neuronal activity. It is mainly used in the study of the brain or spinal cord of humans and animals. The candidate movies can include different types of movie clips. For example, the movie types corresponding to the candidate movies can include romantic types, suspense types, documentary types, historical drama types and situation comedy types. Among them, the candidate movies of the romantic type and situation comedy type are positive emotional tendency movie clips, the suspense type is a negative emotional tendency movie clip, and the documentary type and historical drama type are neutral emotional tendency movie clips.

[0020] It should be noted that when selecting test subjects, it is necessary to exclude people with mental disorders, people with a history of organic brain diseases, people who abuse or are dependent on substances, people who are undergoing short-term medication for mental disorder-related diseases, people whose vision, hearing, and comprehension abilities cannot meet the data collection requirements of the project, people who have contraindications to magnetic resonance imaging, and people who have obvious impulsive behavior and cannot cooperate with the examination.

[0021] Specifically, the required number of testers are selected according to the tester selection criteria, and the Hamilton Depression Rating Scale, Hamilton Anxiety Rating Scale and Cognitive Mini-Mental State Examination scale of the testers are collected; the fluid intelligence-related behavioral performance of the testers is determined according to the testers' episodic memory ability, cognitive flexibility, executive function, inhibitory control ability, information processing speed and working memory ability; the crystallized intelligence-related behavioral performance of the testers is determined according to the language ability, reading decoding ability, language comprehension ability and vocabulary comprehension ability; the negative emotion score, mental health score, stress and self-efficacy score and social relationship score of the testers are determined; the cognitive score of the tester is determined according to the Hamilton Depression Rating Scale, Hamilton Anxiety Rating Scale and Cognitive Mini-Mental State Examination scale, fluid intelligence-related behavioral performance, crystallized intelligence-related behavioral performance, negative emotion score, mental health score, stress and self-efficacy score and social relationship score of the testers. The fMRI data of the test subjects when watching the candidate movies are collected as test functional magnetic resonance imaging data. According to the test functional magnetic resonance imaging data, the brain activity sequences at different positions in the brain regions of the test subjects' emotional memory circuit are extracted as the brain activity sequences of the test subjects, wherein the brain regions of the emotional memory circuit include key areas such as the hippocampus, amygdala, parahippocampal gyrus, medial and dorsolateral prefrontal lobes, anterior cingulate gyrus, posterior cingulate gyrus and precuneus.

[0022] S220. Construct an emotional memory network of the tester according to the brain activity sequence of the tester, determine a network activation pattern of the emotional memory network, and determine a test brain state corresponding to the network activation pattern.

[0023] Specifically, based on the brain activity sequences at different locations in the brain region of the tester's emotional memory circuit, the synergy between activations at different locations in the brain region is evaluated, and the network decomposition method is used to construct an individualized brain function network of the brain's emotional and memory circuits, namely the emotional memory network. Through the hidden Markov model method, by detecting the co-activation patterns at various locations in the emotional memory network, the brain states with different activation patterns in the dynamic activities of the emotional memory network are identified, namely the test brain states corresponding to the network activation patterns.

[0024] S230. Determine a test activation sequence of the emotional memory network according to the test brain state, determine pattern similarity between network activation patterns according to the test activation sequence, generate a high-dimensional representation relationship structure according to the pattern similarity, and generate a predictive representation map according to the high-dimensional representation relationship structure.

[0025] Among them, the test activation sequence of the emotional memory network is the activation sequence of the emotional memory network extracted from each test brain state of different activation modes.

[0026] Specifically, a test activation sequence of the emotional memory network is extracted from each test brain state of different activation patterns. The pattern similarity between the test brain states corresponding to different time periods, that is, the pattern similarity between the network activation patterns, is calculated through the searchlight multi-voxel analysis method. A matrix of high-dimensional representation relationship structure is generated according to the pattern similarity between the network activation patterns. Through the reinforcement learning method of the successor representation model, the reinforcement learning model of the successor representation is applied to the matrix corresponding to the high-dimensional representation relationship structure to generate a predictive representation map, thereby obtaining the predictive conversion relationship between the representations of each test brain state in different time periods.

[0027] S240, determining a test receptive field according to the predictive representation map, determining a predictive index of the predictive representation map according to the test receptive field, and determining a cognitive imaging marker of the test person according to the predictive index.

[0028] Among them, the test receptive field is extracted from the column elements in the matrix corresponding to the predictive representation map, and the position field is a typical cognitive map representation pattern.

[0029] Specifically, Figure 3 As shown, the column elements in the matrix corresponding to the predictive representation map are mapped to the time period representation distribution map of the two-dimensional brain space, and the test receptive field is determined according to the time period representation distribution map of the two-dimensional brain space. The predictive index of the predictive representation map is calculated according to the test receptive field. The difference between the predictive representation maps is determined according to the test receptive field and the predictive index, and the cognitive imaging marker of the test person is determined according to the difference between the predictive representation maps.

[0030] S250. Through machine learning algorithms, a cognitive level prediction model is constructed based on cognitive scores and cognitive imaging markers of test subjects.

[0031] S260. Determine the cognitive level parameters of the target person according to the target functional magnetic resonance imaging data when the person watches the target movie clip through the cognitive level prediction model.

[0032] The target movie clip may correspond to a movie genre such as a romance genre, a suspense genre, a documentary genre, a historical drama genre, or a situation comedy genre.

[0033] Specifically, the cognitive imaging markers of the target person are determined based on the target functional magnetic resonance imaging data when the person watches the target movie clip, and the cognitive imaging markers of the target person are input into the cognitive level prediction model to determine the cognitive level parameters of the target person.

[0034] In the above-mentioned cognitive level parameter evaluation method, the cognitive score of the tester and the test functional magnetic resonance imaging data of the tester when watching the candidate movie are determined, and the brain activity sequence of the tester is determined based on the test functional magnetic resonance imaging data; the emotional memory network of the tester is constructed based on the brain activity sequence of the tester, and the network activation mode of the emotional memory network is determined, and the test brain state corresponding to the network activation mode is determined; the test activation sequence of the emotional memory network is determined based on the test brain state, the pattern similarity between the network activation patterns is determined based on the test activation sequence, and a high-dimensional representation relationship structure is generated based on the pattern similarity, and a predictive representation map is generated based on the high-dimensional representation relationship structure; the test receptive field is determined based on the predictive representation map, and the predictive index of the predictive representation map is determined based on the test receptive field, and the cognitive imaging markers of the tester are determined based on the predictive index; a cognitive level prediction model is constructed based on the cognitive score and the cognitive imaging markers of the tester through a machine learning algorithm; through the cognitive level prediction model, the cognitive level parameters of the target person are determined based on the target functional magnetic resonance imaging data when the person watches the target movie clip. This solves the problem that the current assessment of cognitive impairment in individuals mainly relies on the individual's cognitive performance and scale scores, the assessment cycle is long and lacks objective quantitative indicators, and the accuracy of the assessment of the individual's cognitive level parameters is low. The above scheme, based on the fMRI data of the emotional and memory circuits of the tester's brain, combines modeling methods such as brain network dynamic analysis and cognitive map representation to explore the potential imaging markers of individual cognition and study the neural mechanisms that affect the changes in cognitive level parameters. Through machine learning algorithms, the cognitive level prediction model is trained based on the tester's fMRI data and the tester's cognitive score. The cognitive score of the target person is predicted through the cognitive level prediction model, which can improve the efficiency of cognitive assessment and the accuracy of cognitive assessment results when conducting cognitive assessment on the target person.

[0035] In one embodiment, Figure 4 As shown, the above cognitive level parameter evaluation method also includes: S310, determining the vocabulary feature vectors of the candidate movies, and constructing a correlation network structure between the vocabulary feature vectors based on the Pearson correlation coefficients between the vocabulary feature vectors.

[0036] Specifically, the lexical feature vectors of the candidate movies are extracted through the WordNet lexicon, and the Pearson correlation coefficient between the lexical feature vectors is calculated, and the correlation network structure between the lexical feature vectors is constructed according to the Pearson correlation coefficient.

[0037] S320. Perform correlation analysis on the correlation network structure and the predictive representation map, and construct a regulatory paradigm of the film genre of the candidate film on the cognitive level based on the analysis results.

[0038] Among them, the regulatory paradigm of the film genre of the candidate film on the cognitive level refers to the regulatory paradigm of the regulatory effect of the film genre on the parameters of the individual cognitive level.

[0039] Specifically, a correlation analysis is performed on the correlation network structure and the predictive representation map to determine the association between the correlation network structure corresponding to the candidate movie and the predictive representation map determined by the test subjects watching the candidate movie, and based on the association between the correlation network structure corresponding to the candidate movie and the predictive representation map determined by the test subjects watching the candidate movie, a regulatory paradigm of the film type of the candidate movie on the cognitive level is established.

[0040] The above method, combined with the natural stimulation paradigm, studies the relationship between lexical features and cognitive imaging markers corresponding to different types of movies, which can reveal the potential mechanism of changes in cognitive level parameters and provide a theoretical basis for subsequent research on the regulation and intervention strategies of cognitive level parameters.

[0041] In one embodiment, Figure 5 As shown, the test receptive field is determined according to the predictive representation map, and the predictive index of the predictive representation map is determined according to the test receptive field, including: S410, determining a test receptive field according to the predictive representation map, and determining a curve window of an activation curve corresponding to the test receptive field by a sliding window method, determining an in-window kurtosis corresponding to the curve window, and determining a candidate window from the curve window according to the in-window kurtosis.

[0042] Specifically, the test receptive field is determined according to the predictive representation map, the activation curve corresponding to the test receptive field is processed by the sliding window method, the curve window of the activation curve corresponding to the test receptive field is determined, the kurtosis within the window corresponding to each curve window is determined, and the curve windows with the top 20% kurtosis values ​​can be selected as candidate windows.

[0043] S420, performing de-overlapping processing on the candidate windows, determining a target window, and determining a predictive index of the predictive representation map according to the skewness value of the target window.

[0044] Specifically, the candidate windows are de-overlapped, the target windows are determined, the skewness values ​​of the target windows are calculated, the negative numbers are determined from the skewness values, and the absolute values ​​of the negative numbers are taken as the predictive indicators of the predictive representation map.

[0045] The above scheme provides a method for determining the predictive index of the predictive characterization map, thereby improving the accuracy of the predictive index of the predictive characterization map.

[0046] In one embodiment, Figure 6As shown, the machine learning algorithm includes a support vector machine and a ridge regression model. Through the machine learning algorithm, a cognitive level prediction model is constructed based on the cognitive score and the cognitive imaging markers of the tester, including: S510, constructing a classification model according to the cognitive scores and the cognitive imaging markers of the test persons through a support vector machine.

[0047] Among them, the classification model is the classifier corresponding to the testers with different emotions and cognitive scores.

[0048] Specifically, a classification model is constructed based on cognitive scores and cognitive imaging markers of testers through support vector machines, that is, a classifier is trained using support vector machines to distinguish testers with different cognitive scores, and a screening model for testers with different cognitive scores is established.

[0049] S520. Construct a model to be tested according to the cognitive scores and the cognitive imaging markers of the test persons through a ridge regression model.

[0050] Specifically, the cognitive imaging markers of the testers are used to predict the cognitive scores of the testers, and a target data set is formed. Some data are randomly extracted from the target data set as model training data, and the data in the target data set other than the model training data is used as model testing data. The ridge regression model is trained using the model training data, and the number of model training times can be set according to actual needs, for example, it can be iterated 100 times, and the model to be tested is determined based on the model training results.

[0051] S530, using a cross-validation method to verify the reliability of the model to be detected. If the verification is successful, the model to be detected is used as the target model, and a cognitive level prediction model is constructed based on the classification model and the target model.

[0052] Specifically, the leave-p-out cross-validation method is used to verify the reliability of the model to be tested according to the model test data. If the verification passes, the model to be tested is used as the target model, and a cognitive level prediction model is constructed based on the classification model and the target model.

[0053] The above scheme trains the support vector machine according to the cognitive score and the cognitive imaging markers of the tester to determine the classifier, trains the ridge regression model according to the cognitive score and the cognitive imaging markers of the tester, verifies the reliability of the trained ridge regression model through the cross-validation method, and determines the cognitive level prediction model according to the classifier and the ridge regression model after training, which can improve the model prediction accuracy of the cognitive level prediction model.

[0054] In one embodiment, a classification model is constructed based on cognitive scores and cognitive imaging markers of test subjects by using a support vector machine, including: The training data set and the test data set are determined according to the cognitive scores and the cognitive imaging markers of the testers; the support vector machine is trained with the training data set to determine the candidate model, and the accuracy of the candidate model is verified with the test data set. If the verification passes, the candidate model is used as the classification model.

[0055] Specifically, the cognitive imaging markers of the test persons predict the cognitive scores of the test persons and organize them into a target data set, randomly extract some data from the target data set as a training data set, and use the data in the target data set other than the training data set as a test data set. The support vector machine is iteratively trained using the training data set to determine the candidate model. The number of iterative trainings can be 100 times. The accuracy of the candidate model is verified by the test data set. If the verification is passed, the candidate model is used as a classification model.

[0056] The above scheme can improve the reliability of the classification model.

[0057] Exemplarily, the cognitive level parameters of the target person are determined by the cognitive level prediction model according to the target functional magnetic resonance imaging data when the target person watches the target movie clip, including: Determine the target person's brain activity sequence based on the target functional magnetic resonance imaging data when the target person watches the target movie clip; construct the target person's target memory network based on the target person's brain activity sequence, determine the target activation mode based on the target memory network, and determine the target brain state corresponding to the target activation mode; determine the target activation sequence of the target memory network based on the target brain state, and generate a target prediction representation map based on the target activation sequence; determine the target receptive field based on the target prediction representation map, determine the target prediction index based on the target receptive field, and determine the target person's cognitive imaging markers based on the target prediction index; determine the target person's cognitive level parameters based on the target person's cognitive imaging markers through a cognitive level prediction model.

[0058] Among them, the target memory network is the emotional memory network of the target person, the target activation mode refers to the network activation mode of the target memory network, the target prediction representation map refers to the predictive representation map corresponding to the target memory network, the target receptive field refers to the receptive field corresponding to the target prediction representation map, and the target prediction index refers to the predictive index corresponding to the target receptive field.

[0059] The above scheme, when evaluating the cognitive score of the target person, can determine the cognitive imaging markers of the target person based on the target functional magnetic resonance imaging data when the target person watches the target movie clip, and determine the cognitive score of the target person based on the cognitive imaging markers of the target person through a cognitive level prediction model, so as to evaluate the cognitive level parameters of the target person based on the cognitive score of the target person, thereby improving the efficiency of evaluating the cognitive level parameters of the target person.

[0060] In one embodiment, Figure 7 As shown, a predictive representation graph is generated based on the high-dimensional representation relationship structure, including: S610: Determine a diagonal degree matrix according to the high-dimensional representation relationship structure, and determine a transfer matrix according to the high-dimensional representation relationship structure and the diagonal degree matrix.

[0061] The values ​​in the diagonal degree matrix are the sum of the values ​​of the corresponding rows in the high-dimensional matrix representing the relational structure.

[0062] Specifically, the diagonal matrix is ​​determined according to the values ​​of each row in the matrix representing the high-dimensional relationship structure and the values ​​of each row in the diagonal matrix. The calculation formula of the transfer matrix is ​​shown in formula (1): (1) Among them, T is a diagonal matrix, D is a diagonal degree matrix, and W is a high-dimensional representation relationship structure.

[0063] S620, determining a predictive representation map according to the transfer matrix, the identity matrix and the discount coefficient.

[0064] Specifically, the calculation formula of the predictive representation map is shown in formula (2): (2) Where M is the predictive representation map, I is the identity matrix, is the discount factor.

[0065] The above scheme provides a method for calculating a predictive characterization map, which can improve the accuracy of the predictive characterization map.

[0066] In one embodiment, the cognitive level parameter evaluation method further includes: According to the tester selection criteria, the required number of testers were selected, and their Hamilton Depression Rating Scale, Hamilton Anxiety Rating Scale and Cognitive Mini-Mental State Examination scores were collected. The testers' fluid intelligence-related behavioral performance was determined according to their episodic memory ability, cognitive flexibility, executive function, inhibitory control ability, information processing speed and working memory ability. The testers' crystallized intelligence-related behavioral performance was determined according to their language ability, reading decoding ability, language comprehension ability and vocabulary comprehension ability. The testers' negative emotion score, mental health score, stress and self-efficacy score and social relationship score were also determined. The testers' cognitive score was determined according to the testers' Hamilton Depression Rating Scale, Hamilton Anxiety Rating Scale and Cognitive Mini-Mental State Examination, fluid intelligence-related behavioral performance, crystallized intelligence-related behavioral performance, negative emotion score, mental health score, stress and self-efficacy score and social relationship score. The fMRI data of the test subjects when watching the candidate movies are collected as test functional magnetic resonance imaging data. According to the test functional magnetic resonance imaging data, the brain activity sequences at different positions in the brain regions of the test subjects' emotional memory circuit are extracted as the brain activity sequences of the test subjects, wherein the brain regions of the emotional memory circuit include key areas such as the hippocampus, amygdala, parahippocampal gyrus, medial and dorsolateral prefrontal lobes, anterior cingulate gyrus, posterior cingulate gyrus and precuneus.

[0067] According to the brain activity sequences at different locations in the brain region of the tester's emotional memory circuit, the synergy between activations at different locations in the brain region is evaluated, and the network decomposition method is used to construct an individualized brain function network of the brain's emotional and memory circuits, namely the emotional memory network. Through the hidden Markov model method, by detecting the co-activation patterns at various locations in the emotional memory network, the brain states with different activation patterns in the dynamic activities of the emotional memory network are identified, namely the test brain states corresponding to the network activation patterns.

[0068] The test activation sequence of the emotional memory network is extracted from each test brain state of different activation patterns. The pattern similarity between the test brain states corresponding to different time periods, that is, the pattern similarity between the network activation patterns, is calculated through the searchlight multi-voxel analysis method. A matrix of high-dimensional representation relationship structure is generated according to the pattern similarity between the network activation patterns. Through the reinforcement learning method of the successor representation model, the reinforcement learning model of the successor representation is applied to the matrix corresponding to the high-dimensional representation relationship structure to generate a predictive representation map, thereby obtaining the predictive conversion relationship between the representations of each test brain state in different time periods.

[0069] The column elements in the matrix corresponding to the predictive representation map are mapped to the time period representation distribution map of the two-dimensional brain space. The test receptive field is determined according to the predictive representation map. The activation curve corresponding to the test receptive field is processed by the sliding window method to determine the curve window of the activation curve corresponding to the test receptive field. The kurtosis in the window corresponding to each curve window can be determined. The curve windows in the top 20% of the kurtosis value can be selected as candidate windows. The candidate windows are de-overlapped to determine the target window, the skewness value of each target window is calculated, the negative number is determined from the skewness value, and the absolute value of the negative number is taken as the predictive index of the predictive representation map. The difference between the predictive representation maps is determined according to the test receptive field and the predictive index, and the cognitive imaging markers of the tester are determined according to the difference between the predictive representation maps.

[0070] The cognitive imaging markers of the testers are used to predict the cognitive scores of the testers, and some data are randomly extracted from the target data set as the training data set, and the data in the target data set other than the training data set is used as the test data set. The support vector machine is iteratively trained using the training data set to determine the candidate model. The number of iterative training can be 100 times. The accuracy of the candidate model is verified by the test data set. If the verification is passed, the candidate model is used as the classification model.

[0071] Randomly extract some data from the target data set as model training data, and use the data in the target data set other than the model training data as model test data. Use the model training data to train the ridge regression model. The number of model training can be set according to actual needs. For example, it can be iterated 100 times, and the model to be tested is determined according to the model training results. The leave-p-out cross-validation method is used to verify the reliability of the model to be tested based on the model test data. If the verification passes, the model to be tested is used as the target model, and a cognitive level prediction model is constructed based on the classification model and the target model.

[0072] The lexical feature vectors of the candidate movies are extracted through the WordNet lexicon, and the Pearson correlation coefficient between the lexical feature vectors is calculated. The correlation network structure between the lexical feature vectors is constructed based on the Pearson correlation coefficient. The correlation network structure and the predictive representation map are analyzed to determine the correlation between the correlation network structure corresponding to the candidate movie and the predictive representation map determined by the testers watching the candidate movie. Based on the correlation between the correlation network structure corresponding to the candidate movie and the predictive representation map determined by the testers watching the candidate movie, the regulatory paradigm of the movie type of the candidate movie on the cognitive level is established.

[0073] The above scheme determines the cognitive score of the tester and the test functional magnetic resonance imaging data when the tester watches the candidate movie, and determines the brain activity sequence of the tester based on the test functional magnetic resonance imaging data; constructs the emotional memory network of the tester based on the brain activity sequence of the tester, determines the network activation pattern of the emotional memory network, and determines the test brain state corresponding to the network activation pattern; determines the test activation sequence of the emotional memory network based on the test brain state, determines the pattern similarity between the network activation patterns based on the test activation sequence, generates a high-dimensional representation relationship structure based on the pattern similarity, and generates a predictive representation map based on the high-dimensional representation relationship structure; determines the test receptive field based on the predictive representation map, determines the predictive index of the predictive representation map based on the test receptive field, and determines the cognitive imaging markers of the tester based on the predictive index; constructs a cognitive level prediction model based on the cognitive score and the cognitive imaging markers of the tester through a machine learning algorithm; determines the cognitive level parameters of the target person based on the target functional magnetic resonance imaging data when the person watches the target movie clip through the cognitive level prediction model. This solves the problem that the current assessment of cognitive impairment in individuals mainly relies on the individual's cognitive performance and scale scores, the assessment cycle is long and lacks objective quantitative indicators, and the accuracy of the assessment of the individual's cognitive level parameters is low. The above scheme, based on the fMRI data of the emotional and memory circuits of the tester's brain, combines modeling methods such as brain network dynamic analysis and cognitive map representation to explore the potential imaging markers of individual cognition and study the neural mechanisms that affect the changes in cognitive level parameters. Through machine learning algorithms, the cognitive level prediction model is trained based on the tester's fMRI data and the tester's cognitive score. The cognitive score of the target person is predicted through the cognitive level prediction model, which can improve the efficiency of cognitive assessment and the accuracy of cognitive assessment results when conducting cognitive assessment on the target person.

[0074] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0075] Based on the same inventive concept, the embodiment of the present application also provides a cognitive level parameter evaluation device for implementing the cognitive level parameter evaluation method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more cognitive level parameter evaluation device embodiments provided below can refer to the limitations of the cognitive level parameter evaluation method above, and will not be repeated here.

[0076] In one embodiment, Figure 8 As shown, a cognitive level parameter evaluation device is provided, comprising: a brain activity sequence determination module 701, a brain state determination module 702, a representation map determination module 703, an imaging marker determination module 704, a model training module 705 and a cognitive level parameter evaluation module 706, wherein: A brain activity sequence determination module 701 is used to determine the cognitive score of the test subject and the test functional magnetic resonance imaging data when the test subject watches the candidate movie, and determine the brain activity sequence of the test subject according to the test functional magnetic resonance imaging data; The brain state determination module 702 is used to construct the tester's emotional memory network according to the tester's brain activity sequence, determine the network activation mode of the emotional memory network, and determine the test brain state corresponding to the network activation mode; A representation map determination module 703 is used to determine a test activation sequence of the emotional memory network according to the test brain state, determine pattern similarity between network activation patterns according to the test activation sequence, generate a high-dimensional representation relationship structure according to the pattern similarity, and generate a predictive representation map according to the high-dimensional representation relationship structure; An imaging marker determination module 704 is used to determine a test receptive field according to the predictive representation map, determine a predictive index of the predictive representation map according to the test receptive field, and determine a cognitive imaging marker of the test person according to the predictive index; A model training module 705 is used to construct a cognitive level prediction model based on cognitive scores and cognitive imaging markers of test subjects through a machine learning algorithm; The cognitive level parameter evaluation module 706 is used to determine the cognitive level parameters of the target person according to the target functional magnetic resonance imaging data when the target person watches the target movie clip through the cognitive level prediction model.

[0077] Exemplarily, the above-mentioned cognitive level parameter evaluation device further includes: A correlation network determination module is used to determine the vocabulary feature vectors of the candidate movies and to construct a correlation network structure between the vocabulary feature vectors based on the Pearson correlation coefficients between the vocabulary feature vectors; The correlation analysis module is used to perform correlation analysis on the correlation network structure and the predictive representation map, and to construct a regulatory paradigm of the film genre of the candidate film on the cognitive level based on the analysis results.

[0078] Exemplarily, the imaging marker determination module 704 is specifically used for: Determine a test receptive field according to the predictive representation map, determine a curve window of an activation curve corresponding to the test receptive field by a sliding window method, determine the kurtosis in the window corresponding to the curve window, and determine a candidate window from the curve window according to the kurtosis in the window; The candidate windows are de-overlapped, a target window is determined, and a predictive index of the predictive representation map is determined according to the skewness value of the target window.

[0079] Exemplarily, the above machine learning algorithm includes a support vector machine and a ridge regression model. Further, the model training module 705 is specifically used for: A classification model was constructed based on cognitive scores and cognitive imaging markers of test subjects using support vector machines; The model to be tested was constructed based on the cognitive scores and cognitive imaging markers of the testers through the ridge regression model; The cross-validation method is used to verify the reliability of the model to be tested. If the verification passes, the model to be tested is used as the target model, and a cognitive level prediction model is constructed based on the classification model and the target model.

[0080] Furthermore, the model training module 705 is also specifically used for: Determine the training data set and the test data set based on the cognitive scores and cognitive imaging markers of the testers; The training data set is used to train the support vector machine to determine the candidate model, and the accuracy of the candidate model is verified through the test data set. If the verification passes, the candidate model is used as the classification model.

[0081] Exemplarily, the cognitive level parameter evaluation module 706 is specifically used for: determining the target person's brain activity sequence based on the target functional magnetic resonance imaging data when the target person watches the target movie clip; Constructing a target memory network of the target person according to the brain activity sequence of the target person, determining a target activation pattern according to the target memory network, and determining a target brain state corresponding to the target activation pattern; Determine a target activation sequence of a target memory network according to a target brain state, and generate a target prediction representation map according to the target activation sequence; Determine the target receptive field according to the target prediction representation map, determine the target prediction index according to the target receptive field, and determine the cognitive imaging marker of the target person according to the target prediction index; Through the cognitive level prediction model, the cognitive level parameters of the target person are determined according to the cognitive imaging markers of the target person.

[0082] Exemplarily, the characterization spectrum determination module 703 is specifically used for: Determine a diagonal degree matrix according to the high-dimensional representation relationship structure, and determine a transfer matrix according to the high-dimensional representation relationship structure and the diagonal degree matrix; A predictive representation map is determined based on the transfer matrix, the identity matrix, and the discount coefficient.

[0083] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Fig. 9 As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be realized through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a cognitive level parameter evaluation method is realized. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0084] Those skilled in the art will understand that Fig. 9 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0085] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented: Step 1, determining the cognitive score of the tester and the test functional magnetic resonance imaging data when the tester watches the candidate movie, and determining the brain activity sequence of the tester based on the test functional magnetic resonance imaging data; Step 2: construct the tester's emotional memory network according to the tester's brain activity sequence, determine the network activation mode of the emotional memory network, and determine the test brain state corresponding to the network activation mode; Step 3: determining a test activation sequence of the emotional memory network according to the test brain state, determining pattern similarity between network activation patterns according to the test activation sequence, generating a high-dimensional representation relationship structure according to the pattern similarity, and generating a predictive representation map according to the high-dimensional representation relationship structure; Step 4: determining a test receptive field according to the predictive representation map, determining a predictive index of the predictive representation map according to the test receptive field, and determining a cognitive imaging marker of the tester according to the predictive index; Step 5: Using machine learning algorithms, a cognitive level prediction model is constructed based on cognitive scores and cognitive imaging markers of test subjects; Step 6: Determine the cognitive level parameters of the target person based on the target functional magnetic resonance imaging data when the target person watches the target movie clip through the cognitive level prediction model.

[0086] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented: Step 1, determining the cognitive score of the tester and the test functional magnetic resonance imaging data when the tester watches the candidate movie, and determining the brain activity sequence of the tester based on the test functional magnetic resonance imaging data; Step 2: construct the tester's emotional memory network according to the tester's brain activity sequence, determine the network activation mode of the emotional memory network, and determine the test brain state corresponding to the network activation mode; Step 3: determining a test activation sequence of the emotional memory network according to the test brain state, determining pattern similarity between network activation patterns according to the test activation sequence, generating a high-dimensional representation relationship structure according to the pattern similarity, and generating a predictive representation map according to the high-dimensional representation relationship structure; Step 4: determining a test receptive field according to the predictive representation map, determining a predictive index of the predictive representation map according to the test receptive field, and determining a cognitive imaging marker of the tester according to the predictive index; Step 5: Using machine learning algorithms, a cognitive level prediction model is constructed based on cognitive scores and cognitive imaging markers of test subjects; Step 6: Determine the cognitive level parameters of the target person based on the target functional magnetic resonance imaging data when the target person watches the target movie clip through the cognitive level prediction model.

[0087] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps: Step 1, determining the cognitive score of the tester and the test functional magnetic resonance imaging data when the tester watches the candidate movie, and determining the brain activity sequence of the tester based on the test functional magnetic resonance imaging data; Step 2: construct the tester's emotional memory network according to the tester's brain activity sequence, determine the network activation mode of the emotional memory network, and determine the test brain state corresponding to the network activation mode; Step 3: determining a test activation sequence of the emotional memory network according to the test brain state, determining pattern similarity between network activation patterns according to the test activation sequence, generating a high-dimensional representation relationship structure according to the pattern similarity, and generating a predictive representation map according to the high-dimensional representation relationship structure; Step 4: determining a test receptive field according to the predictive representation map, determining a predictive index of the predictive representation map according to the test receptive field, and determining a cognitive imaging marker of the tester according to the predictive index; Step 5: Using machine learning algorithms, a cognitive level prediction model is constructed based on cognitive scores and cognitive imaging markers of test subjects; Step 6: Determine the cognitive level parameters of the target person based on the target functional magnetic resonance imaging data when the target person watches the target movie clip through the cognitive level prediction model.

[0088] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.

[0089] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0090] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0091] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be construed as limiting the scope of the present application. It should be noted that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A cognitive level parameter evaluation method, characterized in that: include: Determining a cognitive score of a test person, and test functional magnetic resonance imaging data of the test person when the test person watches a candidate movie, and determining a brain activity sequence of the test person based on the test functional magnetic resonance imaging data; Constructing an emotional memory network of the tester according to the brain activity sequence of the tester, determining a network activation pattern of the emotional memory network, and determining a test brain state corresponding to the network activation pattern; Determining a test activation sequence of the emotional memory network according to the test brain state, determining pattern similarity between network activation patterns according to the test activation sequence, generating a high-dimensional representation relationship structure according to the pattern similarity, and generating a predictive representation map according to the high-dimensional representation relationship structure; Determining a test receptive field according to the predictive representation map, determining a predictive index of the predictive representation map according to the test receptive field, and determining a cognitive imaging marker of the test person according to the predictive index; Constructing a cognitive level prediction model based on the cognitive score and the cognitive imaging markers of the test subject through a machine learning algorithm; The cognitive level parameters of the target person are determined by the cognitive level prediction model according to the target functional magnetic resonance imaging data when the target person watches the target movie clip.

2. The method according to claim 1, characterized in that include: Determine the vocabulary feature vectors of the candidate movies, and construct a correlation network structure between the vocabulary feature vectors based on the Pearson correlation coefficients between the vocabulary feature vectors; A correlation analysis is performed on the correlation network structure and the predictive representation map, and a regulatory paradigm of the film genre of the candidate film on the cognitive level is constructed according to the analysis results.

3. The method according to claim 1, characterized in that Determining a test receptive field according to the predictive representation map, and determining a predictive index of the predictive representation map according to the test receptive field, including: Determine a test receptive field according to the predictive representation map, determine a curve window of an activation curve corresponding to the test receptive field by a sliding window method, determine the kurtosis in the window corresponding to the curve window, and determine a candidate window from the curve window according to the kurtosis in the window; The candidate windows are de-overlapped to determine a target window, and a predictive index of the predictive representation map is determined according to the skewness value of the target window.

4. The method according to claim 1, characterized in that: The machine learning algorithm includes a support vector machine and a ridge regression model. The machine learning algorithm is used to construct a cognitive level prediction model according to the cognitive score and the cognitive imaging markers of the test person, including: Building a classification model based on the cognitive score and the cognitive imaging markers of the test person by a support vector machine; Constructing a model to be tested according to the cognitive score and the cognitive imaging markers of the test person through a ridge regression model; The reliability of the model to be detected is verified by a cross-validation method. If the verification is passed, the model to be detected is used as the target model, and a cognitive level prediction model is constructed according to the classification model and the target model.

5. The method according to claim 4, characterized in that A classification model is constructed according to the cognitive score and the cognitive imaging markers of the tester by a support vector machine, including: Determine a training data set and a test data set according to the cognitive score and the cognitive imaging markers of the test person; The training data set is used to train a support vector machine to determine a candidate model, and the accuracy of the candidate model is verified using the test data set. If the verification passes, the candidate model is used as a classification model.

6. The method according to claim 1, characterized in that Determining the cognitive level parameters of the target person according to the target functional magnetic resonance imaging data when the target person watches the target movie clip by using the cognitive level prediction model includes: determining a brain activity sequence of the target person based on target functional magnetic resonance imaging data when the target person watches a target movie clip; constructing a target memory network of the target person according to the brain activity sequence of the target person, determining a target activation pattern according to the target memory network, and determining a target brain state corresponding to the target activation pattern; Determining a target activation sequence of the target memory network according to the target brain state, and generating a target prediction representation map according to the target activation sequence; Determine a target receptive field according to the target prediction representation map, determine a target prediction index according to the target receptive field, and determine a cognitive imaging marker of the target person according to the target prediction index; The cognitive level parameters of the target person are determined according to the cognitive imaging markers of the target person through the cognitive level prediction model.

7. The method according to claim 1, characterized in that Generating a predictive representation graph according to the high-dimensional representation relationship structure includes: Determine a diagonal degree matrix according to the high-dimensional representation relationship structure, and determine a transfer matrix according to the high-dimensional representation relationship structure and the diagonal degree matrix; A predictive representation map is determined based on the transfer matrix, the identity matrix, and the discount factor.

8. A cognitive level parameter evaluation device, characterized in that: The cognitive level parameter evaluation device comprises: A brain activity sequence determination module, used to determine the cognitive score of the test person and the test functional magnetic resonance imaging data of the test person when watching the candidate movie, and determine the brain activity sequence of the test person according to the test functional magnetic resonance imaging data; A brain state determination module, used to construct an emotional memory network of the tester according to the brain activity sequence of the tester, determine a network activation pattern of the emotional memory network, and determine a test brain state corresponding to the network activation pattern; a representation map determination module, configured to determine a test activation sequence of the emotional memory network according to the test brain state, determine pattern similarity between network activation patterns according to the test activation sequence, generate a high-dimensional representation relationship structure according to the pattern similarity, and generate a predictive representation map according to the high-dimensional representation relationship structure; An imaging marker determination module, used to determine a test receptive field according to the predictive representation map, determine a predictive index of the predictive representation map according to the test receptive field, and determine a cognitive imaging marker of the test person according to the predictive index; A model training module, used for constructing a cognitive level prediction model according to the cognitive score and the cognitive imaging markers of the test person through a machine learning algorithm; The cognitive level parameter evaluation module is used to determine the cognitive level parameters of the target person according to the target functional magnetic resonance imaging data when the target person watches the target movie clip through the cognitive level prediction model.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images