Mild cognitive impairment identification method and system based on fNIRS data under multi-task normal form fusion

Abstract

The invention discloses a mild cognitive impairment recognition method and system based on fNIRS data under multi-task normal form fusion. The method comprises the steps that fNIRS data, obtained based on a resting state task and an n-back work memory task, of a testee are obtained; based on the obtained fNIRS data, extracting fNIRS related features; performing feature selection on the extracted fNIRS related features based on a feature selection algorithm to form a feature set; and inputting the formed feature set into the trained deep learning model, and outputting a mild cognitive impairment recognition result. According to the method, the data of the fNIRS under the resting state and the n-back work memory cognition task are obtained, the multi-dimensional fNIRS related features are calculated, feature selection is carried out based on a Gradient Boosting algorithm, and finally accurate MCI recognition is carried out in combination with a deep learning model. The method not only can effectively extract high-dimensional features, but also can realize automatic MCI diagnosis through model training and reasoning, and has relatively high accuracy and relatively strong popularization and application potential.

Description

Technical Field

[0001] The present invention relates to a method and system for identifying mild cognitive impairment based on fNIRS data under the fusion of multi-task paradigms. Background Art

[0002] Mild Cognitive Impairment (MCI) refers to a state of mild decline in cognitive function commonly seen in the elderly population, usually manifested as slight impairments in aspects such as memory, attention, and language ability, but not severe enough to reach the level of dementia. MCI is often regarded as a precursor to neurodegenerative diseases such as Alzheimer's disease. Therefore, early and accurate diagnosis of MCI and intervention can effectively delay the progression of the disease and improve the quality of life of patients. Currently, the diagnosis of MCI mainly relies on clinical evaluations, including medical history collection, cognitive function assessment, neuroimaging examinations, and laboratory tests. Commonly used cognitive assessment tools include the Montreal Cognitive Assessment Scale (MoCA), the Mini-Mental State Examination Scale (MMSE), etc. These scales assist doctors in making a preliminary diagnosis by evaluating multiple dimensions of the patient's memory, language ability, attention, and executive function. However, although these tools are widely used in clinical practice, there are still some problems and limitations.

[0003] On the one hand, traditional cognitive assessment scales rely on doctors' experience and judgment and are easily affected by factors such as the patient's mental state, educational background, and cultural differences, which may lead to misdiagnosis or missed diagnosis. Especially in the early stage of MCI, the decline in the patient's cognitive function is relatively mild, and traditional assessment methods may not be able to detect mild cognitive impairment in a timely manner. On the other hand, although existing neuroimaging techniques (such as magnetic resonance imaging MRI, positron emission tomography PET, etc.) play an important role in the diagnosis of neurodegenerative diseases, these techniques are not only costly, complex to operate, but also have high requirements for equipment, limiting their application in the general population. Especially in the diagnosis of early MCI, imaging examinations are often difficult to reveal subtle brain changes, and the diagnostic sensitivity is relatively low.

[0004] Functional near-infrared spectroscopy (fNIRS), as a non-invasive brain function imaging technique, has been widely used in cognitive neuroscience research in recent years. fNIRS can reflect neural activities by measuring blood oxygen changes in the cerebral cortex region, has good spatial and temporal resolution, and the equipment is relatively portable and easy to operate, suitable for large-scale screening.

[0005] The applications of fNIRS in cognitive function detection mainly include two types of task modes: cognitive tasks and resting states. In the study of mild cognitive impairment, the n-back working memory task of working memory can be used to evaluate working memory and attention. When patients perform tasks, fNIRS records the dynamic changes of cerebral blood flow and analyzes the activation intensity of different brain regions during cognitive tasks. The resting state task can be used to evaluate the network connection strength between different brain regions. The brain region activation under the n-back working memory task and the network connection strength between brain regions in the resting state have certain diagnostic value for identifying cognitive impairments such as MCI. However, the fNIRS data under any single task mode usually only has limited discrimination ability, and in-depth fusion analysis and processing of the data are required to extract diagnostically valuable features.

[0006] Currently, the MCI recognition technology based on fNIRS still faces the following challenges and deficiencies:

[0007] 1. Feature selection and high-dimensional data processing problems: fNIRS data usually contains a large amount of time-series signals and blood oxygen change information of multiple brain regions. These data are often high-dimensional and contain redundant information. Traditional methods are difficult to extract effective features from them, resulting in low recognition accuracy.

[0008] 2. Limitations of traditional machine learning methods: Some existing recognition methods based on traditional machine learning algorithms have improved the MCI recognition ability to a certain extent. However, these methods usually rely on manual feature selection and have poor processing effects on high-dimensional data, and cannot effectively capture complex non-linear relationships.

[0009] 3. Lack of integration of multi-level models: Many existing fNIRS data analysis methods lack in-depth integration of multiple features and multi-level data, and cannot fully explore the potential complex relationships in the data. Therefore, although a certain accuracy can be obtained in some tasks, the generality among different individuals is poor.

[0010] 4. Lack of automated and intelligent reasoning methods: Traditional MCI recognition methods often rely on the subjective judgment of experts and relatively cumbersome analysis steps, and lack an intelligent model that can automatically perform fast and accurate reasoning on new samples. Summary of the Invention

[0011] To address the deficiencies of existing MCI diagnosis methods, the present invention provides a method and system for identifying mild cognitive impairment based on fNIRS data under multi-task paradigm fusion. By obtaining fNIRS data in the resting state and the n-back working memory cognitive task, calculating multi-dimensional fNIRS-related features, and performing feature selection based on the Gradient Boosting algorithm, and finally combining with a deep learning model for accurate identification of MCI. This method can not only effectively extract high-dimensional features, but also achieve automated MCI diagnosis through model training and inference, with high accuracy and strong potential for popularization and application.

[0012] According to one aspect of the specification of the present invention, there is provided a method for identifying mild cognitive impairment based on fNIRS data under multi-task paradigm fusion, including:

[0013] Obtaining fNIRS data of a subject obtained based on a resting state task and an n-back working memory task;

[0014] Extracting fNIRS-related features based on the obtained fNIRS data;

[0015] Performing feature selection on the extracted fNIRS-related features based on a feature selection algorithm to form a feature set;

[0016] Inputting the formed feature set into a trained deep learning model and outputting a mild cognitive impairment identification result.

[0017] As a further technical solution, extracting fNIRS-related features based on the obtained fNIRS data includes:

[0018] Respectively extracting fNIRS eigenvalue under the resting state task and the n-back working memory task to form multi-dimensional fNIRS features.

[0019] As a further technical solution, the fNIRS eigenvalue under the extracted resting state task includes: functional connectivity, betweenness centrality, degree centrality, nodal clustering coefficient, nodal efficiency, nodal local efficiency, nodal shortest path length, global efficiency, local efficiency, clustering coefficient, and small-world property.

[0020] As a further technical solution, the fNIRS eigenvalue under the extracted n-back working memory task includes: beta value, mean value, integral, K-activation, and peak value.

[0021] As a further technical solution, performing feature selection on the extracted fNIRS-related features includes: performing feature selection on the extracted fNIRS-related features by using the gradient boosting algorithm.

[0022] As a further technical solution, the training of the deep learning model includes:

[0023] Construct fNIRS data samples under resting-state tasks and n-back working memory tasks;

[0024] Extract fNIRS-related features of the fNIRS data samples;

[0025] Perform feature selection on the extracted fNIRS-related features to form a feature set;

[0026] Label the features in the feature set to form a training data set;

[0027] Use the formed training data set to train a fully connected neural network and output a trained deep learning model.

[0028] As a further technical solution, after obtaining the fNIRS data of the subject obtained based on the resting-state task and the n-back working memory task, it further includes:

[0029] Preprocess the obtained fNIRS data, including converting the original light intensity to optical density, artifact detection, filtering, and converting the optical density to relative concentration.

[0030] According to one aspect of the specification of the present invention, there is provided a mild cognitive impairment recognition system based on fNIRS data under a multi-task paradigm fusion, including:

[0031] A data input module for obtaining the fNIRS data of the subject obtained based on the resting-state task and the n-back working memory task;

[0032] A feature extraction module for extracting fNIRS-related features based on the obtained fNIRS data;

[0033] A feature selection module for performing feature selection on the extracted fNIRS-related features based on a feature selection algorithm to form a feature set;

[0034] A feature recognition module for inputting the formed feature set into the trained deep learning model and outputting a mild cognitive impairment recognition result.

[0035] According to one aspect of the specification of the present invention, there is provided a mild cognitive impairment recognition device based on fNIRS data under a multi-task paradigm fusion, including a processor and a memory; the memory stores at least one instruction, and the at least one instruction is used to be executed by the processor to implement the steps of the mild cognitive impairment recognition method based on fNIRS data under the multi-task paradigm fusion.

[0036] According to one aspect of the specification of the present invention, there is provided a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the method for identifying mild cognitive impairment based on fNIRS data under multi-task paradigm fusion are implemented.

[0037] Compared with the prior art, the present invention combines fNIRS data, feature selection, machine learning and deep learning technologies, and has significant advantages in the early identification of MCI:

[0038] 1. Improve diagnostic accuracy:

[0039] Through efficient feature extraction and selection of fNIRS data, especially the multi-dimensional features obtained by combining the n-back and resting-state tasks, the present invention can capture subtle changes that are difficult to identify by traditional methods, and significantly improve the early diagnostic accuracy of MCI.

[0040] 2. Automated diagnosis:

[0041] The present invention uses a deep learning model for automated reasoning, which can diagnose new samples efficiently and without subjective bias, reduce manual intervention, lower the misdiagnosis rate, and significantly shorten the time required for diagnosis.

[0042] 3. Reduce costs and improve popularity:

[0043] Compared with traditional imaging examinations (such as MRI, PET), fNIRS devices are relatively inexpensive and easy to operate. The present invention can use more popular devices for data collection, reduce the cost of MCI screening, and improve the application potential in places such as primary hospitals and health examination centers.

[0044] 4. Strong adaptability:

[0045] The technical solution of the present invention can adapt to different patient groups. Especially by evaluating cognitive function through multi-task combinations (VFT and n-back), the sensitivity and generalization ability of the method are enhanced, and it can adapt to the early diagnosis of different types of cognitive impairment. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a schematic flowchart of the method for identifying mild cognitive impairment based on fNIRS data under multi-task paradigm fusion provided by an embodiment of the present invention.

[0047] Figure 2 It is a schematic diagram of the feature extraction result provided by an embodiment of the present invention.

[0048] Figure 3 It is a schematic flowchart of the model training process provided by an embodiment of the present invention.

[0049] Figure 4 Schematic diagram of the mild cognitive impairment recognition system based on fNIRS data under the multi-task paradigm fusion provided by the embodiments of the present invention. Detailed implementation manners

[0050] With the rapid development of machine learning and deep learning technologies, the artificial intelligence-based MCI recognition method has gradually become a research hotspot. In recent years, algorithms such as Gradient Boosting, deep neural network (DNN), and convolutional neural network (CNN) have been widely applied to the analysis of various medical data, especially in dealing with large-scale data sets, feature selection, and pattern recognition, showing excellent performance.

[0051] For fNIRS data, machine learning, especially ensemble learning and deep learning methods, has great advantages. Ensemble learning algorithms such as Gradient Boosting can effectively improve the accuracy and stability of the model through multiple rounds of learning and error correction. Deep learning models can automatically learn deeper feature representations in high-dimensional and non-linear data, further improving the generalization ability and reasoning ability of the model.

[0052] Based on this, the present invention proposes a new MCI recognition method. By obtaining the results of fNIRS in the resting state and the n-back working memory cognitive task, calculating multi-dimensional fNIRS-related features, and performing feature selection based on the Gradient Boosting algorithm, and finally combining with a deep learning model for accurate MCI recognition. This method can not only effectively extract high-dimensional features, but also achieve automated MCI diagnosis through model training and reasoning, with high accuracy and strong potential for popularization and application.

[0053] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of the present invention.

[0054] Existing MCI recognition methods mainly rely on clinical evaluation and neuroimaging examinations. Although these methods can provide some help to a certain extent, they still face many problems in practical applications: (1) Traditional cognitive assessment tools are highly subjective and are easily affected by factors such as the patient's mental state and educational background; (2) Imaging examinations such as MRI and PET can provide information on brain structure changes, but their sensitivity for early MCI recognition is relatively low, and the equipment is expensive and the operation is complex, making it impossible to be widely used in large-scale screening. (3) Existing machine learning methods face difficulties in high-dimensional feature selection when processing fNIRS data, lacking effective feature extraction and model training methods, resulting in low diagnostic accuracy.

[0055] To solve the above technical problems, an embodiment of the present invention provides a method for recognizing mild cognitive impairment based on fNIRS data under the fusion of multi-task paradigms, as Figure 1 shown. First, obtain the fNIRS data of the subject based on the resting-state task and the n-back working memory task; then, extract fNIRS-related features based on the obtained fNIRS data; subsequently, based on the feature selection algorithm, perform feature selection on the extracted fNIRS-related features to form a feature set; then, input the formed feature set into the trained deep learning model to output the recognition result of mild cognitive impairment.

[0056] Based on the foregoing technical solution, an embodiment of the present invention combines two types of task modes (resting state and n-back working memory task), evaluates cognitive functions from two dimensions of the resting brain network and cognitive task activation, and extracts feature values. And, based on the feature selection algorithm (such as the Gradient Boosting algorithm), feature selection is performed to avoid feature redundancy.

[0057] The MCI recognition method based on fNIRS data provided by the embodiment of the present invention can achieve early MCI recognition and overcome the accuracy and universality problems of existing methods. In addition, the embodiment of the present invention also performs inference on new samples through a deep learning model to achieve automated and rapid MCI diagnosis.

[0058] The method for recognizing mild cognitive impairment based on fNIRS data under the fusion of multi-task paradigms provided by the embodiment of the present invention mainly includes the following aspects:

[0059] 1. Obtain fNIRS data:

[0060] In the present invention, fNIRS data of subjects are obtained through resting-state tasks and n-back working memory tasks. In the resting-state task, the subjects are required to remain calm, close their eyes and rest without falling asleep for several minutes, which is used to evaluate the strength of the resting network connection between brain regions. In the n-back working memory task, the subjects need to remember a series of numbers and judge whether the current number is the same as the previous number, which is used to evaluate the working memory ability.

[0061] After obtaining the fNIRS data, it also includes data preprocessing of the original fnirs signal, specifically as follows:

[0062] (1) Convert the original light intensity to optical density (OD);

[0063] (2) Use predefined parameters for motion artifact detection (tMotion = 0.5 s, STDEVthresh = 20, tMask = 3 s, AMPthresh = 5), and then perform spline interpolation correction (p = 0.99);

[0064] (3) Perform band-pass filtering between 0.01 and 0.1 Hz to reduce cardiac and respiratory artifacts;

[0065] (4) Convert the optical density (OD) to relative concentration.

[0066] 2. Calculate fNIRS-related features:

[0067] The obtained fNIRS data includes the temporal information of cerebral blood oxygen changes and usually has high-dimensional features. To effectively utilize these data, the present invention extracts multi-dimensional fNIRS features through signal processing methods. As Figure 2 shown, under the n-back working memory task, eigenvalue such as the integral, mean, general linear model beta value, etc. of the oxygenated hemoglobin concentration in the task period of all channels and brain regions; under the resting-state task mode, eigenvalue such as the wavelet coherence value and Pearson correlation value of the change in oxygenated hemoglobin concentration between all channels and brain regions.

[0068] Specifically, the features under the n-back working memory task include: Beta (beta value), Mean (mean), Integration (integration), K-slope (K-activation), Peak (peak).

[0069] The features under the resting-state task include: FC (functional connectivity), aBc (betweenness centrality), aDc (degree centrality), aNCp (node clustering coefficient), aNe (node efficiency), aNLe (node local efficiency), aNLp (node shortest path length), aEg (global efficiency), aEloc (local efficiency), aCp (clustering coefficient), aSigma (small-world property).

[0070] 3. Select the feature set based on the Gradient Boosting Classifier:

[0071] After the fNIRS data feature extraction, the present invention performs feature selection through the Gradient Boosting algorithm. Gradient Boosting is an ensemble learning algorithm. Through multiple rounds of iterative training, it can effectively reduce the redundancy and noise in high-dimensional features and improve the accuracy of the model. This algorithm can automatically select the most representative features and perform weighted combination to generate a feature set suitable for MCI recognition.

[0072] 4. Build a deep learning model:

[0073] After feature selection, the selected features are further processed using a deep learning model. Before inputting into the deep learning model, it also includes feature preprocessing on the features extracted from the two paradigms respectively, that is, normalizing the features in the selected feature set and labeling the label "whether MCI" to form a training data set.

[0074] The present invention uses a deep neural network (DNN) to build a deep learning model. The DNN model is used to capture the complex non-linear relationships between features and improve the accuracy of MCI recognition. After being trained, this model can automatically learn the brain activation patterns under different tasks and perform precise classification in combination with the performance characteristics of MCI. The training of the model is as Figure 3 shown.

[0075] 5. Infer new samples based on the deep learning model:

[0076] Finally, the trained deep learning model can be applied to new samples for inference to automatically identify whether it is MCI. The fNIRS data of the new sample is compared with the training results of the model, and the model will give a diagnostic result. This step requires no manual intervention and can achieve rapid screening and accurate diagnosis of new patients.

[0077] The implementation basis of each embodiment of the present invention is achieved through programmed processing by a device with a processor function. Therefore, in engineering practice, the technical solutions and their functions of each embodiment of the present invention are encapsulated into various modules. Based on this actual situation, on the basis of the above embodiments, an embodiment of the present invention provides a mild cognitive impairment recognition system based on fNIRS data under multi-task paradigm fusion, and this system is used to execute a mild cognitive impairment recognition method based on fNIRS data under multi-task paradigm fusion in the above method embodiments.

[0078] See Figure 4, the system includes: a data input module for acquiring the fNIRS data of the subject obtained based on the resting-state task and the n-back working memory task; a feature extraction module for extracting fNIRS-related features based on the acquired fNIRS data; a feature selection module for performing feature selection on the extracted fNIRS-related features based on a feature selection algorithm to form a feature set; and a feature recognition module for inputting the formed feature set into a trained deep learning model and outputting a mild cognitive impairment recognition result.

[0079] A mild cognitive impairment recognition system based on fNIRS data under multi-task paradigm fusion provided by an embodiment of the present invention, aiming at the deficiencies of existing MCI diagnosis methods, adopts Figure 4 several modules therein. By acquiring the data of fNIRS under the resting state and the n-back working memory cognitive task, calculating multi-dimensional fNIRS-related features, and performing feature selection based on the GradientBoosting algorithm, and finally combining a deep learning model for accurate recognition of MCI. This method can not only effectively extract high-dimensional features, but also realize automated MCI diagnosis through model training and inference, with high accuracy and strong potential for popularization and application.

[0080] It should be noted that the system embodiment provided by the present invention, in addition to being used to implement the method in the above method embodiment, is also used to implement the methods in other method embodiments provided by the present invention. The difference is only in setting corresponding functional modules, and its principle is basically the same as that of the above system embodiment provided by the present invention. As long as those skilled in the art, based on the above system embodiment, refer to the specific technical solutions in other method embodiments, obtain corresponding technical means by combining technical features, and the technical solutions composed of these technical means, and on the premise of ensuring the practicability of the technical solutions, improve the modules in the above system embodiment to obtain corresponding system-like embodiments for implementing the methods in other method-like embodiments. For example:

[0081] Based on the content of the above system embodiment, as a preferred embodiment, in a mild cognitive impairment recognition system based on fNIRS data under multi-task paradigm fusion provided by an embodiment of the present invention, the feature extraction module is further configured to execute the following instructions:

[0082] Extract the fNIRS eigenvalue under the resting-state task and the n-back working memory task respectively to form multi-dimensional fNIRS features.

[0083] Based on the content of the above system embodiments, as a preferred embodiment, a mild cognitive impairment recognition system based on fNIRS data under multi-task paradigm fusion provided in the embodiments of the present invention, the fNIRS eigenvalue extracted by the feature extraction module under the resting state task includes: functional connectivity, betweenness centrality, degree centrality, nodal clustering coefficient, nodal efficiency, nodal local efficiency, nodal shortest path length, global efficiency, local efficiency, clustering coefficient, and small-world property.

[0084] Based on the content of the above system embodiments, as a preferred embodiment, a mild cognitive impairment recognition system based on fNIRS data under multi-task paradigm fusion provided in the embodiments of the present invention, the fNIRS eigenvalue extracted by the feature extraction module under the n-back working memory task includes: beta value, mean value, integral, K-activation, and peak value.

[0085] Based on the content of the above system embodiments, as a preferred embodiment, a mild cognitive impairment recognition system based on fNIRS data under multi-task paradigm fusion provided in the embodiments of the present invention, the feature selection module is further configured to execute the following instructions:

[0086] Adopt the gradient boosting algorithm to perform feature selection on the extracted fNIRS-related features.

[0087] Based on the content of the above system embodiments, as a preferred embodiment, a mild cognitive impairment recognition system based on fNIRS data under multi-task paradigm fusion provided in the embodiments of the present invention, the feature recognition module is further configured to execute the following instructions:

[0088] Construct fNIRS data samples under the resting state task and the n-back working memory task;

[0089] Extract fNIRS-related features of the fNIRS data samples;

[0090] Perform feature selection on the extracted fNIRS-related features to form a feature set;

[0091] Label the features in the feature set to form a training data set;

[0092] Use the formed training data set to train the fully connected neural network and output the trained deep learning model.

[0093] Based on the content of the above system embodiments, as a preferred embodiment, a mild cognitive impairment recognition system based on fNIRS data under multi-task paradigm fusion provided in the embodiments of the present invention further includes:

[0094] The preprocessing module is used to preprocess the acquired fNIRS data, including converting the original light intensity to optical density, artifact detection, filtering, and converting optical density to relative concentration.

[0095] Based on the same inventive concept as the above embodiments, an apparatus for identifying mild cognitive impairment from fNIRS data based on a multi-task paradigm fusion is further provided in an embodiment of the present invention, including a processor and a memory; the memory stores at least one instruction, and the at least one instruction is used to be executed by the processor to implement the steps of the method for identifying mild cognitive impairment from fNIRS data based on the multi-task paradigm fusion.

[0096] In an embodiment of the present invention, the memory may be a non-volatile memory, such as a hard disk drive (HDD) or a solid-state drive (SSD), etc., or may also be a volatile memory, such as a random-access memory (RAM). The memory is any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory in an embodiment of the present invention may also be a circuit or any other device capable of implementing a storage function, for storing program instructions and / or data.

[0097] In an embodiment of the present invention, the processor may be a general-purpose processor, a digital signal processor, an application-specific integrated circuit, a field-programmable gate array, or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, and can implement or execute the various methods, steps, and logic block diagrams disclosed in an embodiment of the present invention. The general-purpose processor may be a microprocessor or any conventional processor, etc. The steps of the method disclosed in combination with an embodiment of the present invention may be directly embodied as being executed by a hardware processor, or implemented by a combination of hardware and software modules in the processor.

[0098] Based on the same inventive concept as the above embodiments, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method for identifying mild cognitive impairment from fNIRS data based on the multi-task paradigm fusion are implemented as follows:

[0099] Obtain the fNIRS data of the subject obtained based on the resting-state task and the n-back working memory task;

[0100] Extract fNIRS-related features based on the obtained fNIRS data;

[0101] Based on the feature selection algorithm, perform feature selection on the extracted fNIRS-related features to form a feature set;

[0102] Input the formed feature set into the trained deep learning model to output the mild cognitive impairment recognition result.

[0103] In summary of the above embodiments, the present invention obtains the results of fNIRS in the resting state and the n-back working memory cognitive task, calculates multi-dimensional fNIRS-related features, performs feature selection based on the Gradient Boosting algorithm, and finally combines the deep learning model for accurate recognition of MCI. This method can not only effectively extract high-dimensional features, but also achieve automated MCI diagnosis through model training and inference, with high accuracy and strong potential for popularization and application.

[0104] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the technical solutions of the embodiments of the present invention.

Claims

1. A method for identifying mild cognitive impairment based on fNIRS data under the fusion of multi-task paradigms, characterized in that, Including: Obtaining the fNIRS data of the subject obtained based on the resting-state task and the n-back working memory task; Extracting fNIRS-related features based on the obtained fNIRS data; Performing feature selection on the extracted fNIRS-related features based on a feature selection algorithm to form a feature set; Inputting the formed feature set into a trained deep learning model and outputting the mild cognitive impairment recognition result.

2. The method for identifying mild cognitive impairment based on fNIRS data under the fusion of multi-task paradigms according to claim 1, wherein Extracting fNIRS-related features based on the obtained fNIRS data, including: Respectively extracting the fNIRS eigenvalue under the resting-state task and the n-back working memory task to form multi-dimensional fNIRS features.

3. The method for identifying mild cognitive impairment based on fNIRS data under multi-task paradigm fusion according to claim 2, wherein The fNIRS eigenvalue extracted under the resting-state task includes: functional connectivity, betweenness centrality, degree centrality, nodal clustering coefficient, nodal efficiency, nodal local efficiency, nodal shortest path length, global efficiency, local efficiency, clustering coefficient, and small-world property.

4. The method for identifying mild cognitive impairment based on fNIRS data under the fusion of multi-task paradigms according to claim 2, wherein, The fNIRS eigenvalue extracted under the n-back working memory task includes: beta value, mean, integral, K-activation, and peak value.

5. The method for identifying mild cognitive impairment based on fNIRS data under the fusion of multi-task paradigms according to claim 1, characterized in that, Performing feature selection on the extracted fNIRS-related features, including: performing feature selection on the extracted fNIRS-related features using the gradient boosting algorithm.

6. The method for identifying mild cognitive impairment based on fNIRS data under the fusion of multi-task paradigms according to claim 1, wherein The training of the deep learning model includes: Constructing fNIRS data samples under the resting-state task and the n-back working memory task; Extracting fNIRS-related features of the fNIRS data samples; Performing feature selection on the extracted fNIRS-related features to form a feature set; Labeling the features in the feature set to form a training data set; Using the formed training data set to train a fully connected neural network and outputting a trained deep learning model.

7. The method for identifying mild cognitive impairment based on fNIRS data under the multi-task paradigm fusion according to claim 1, characterized in that After obtaining the fNIRS data of the subject obtained based on the resting-state task and the n-back working memory task, it further includes: Preprocessing the obtained fNIRS data, including converting the original light intensity to optical density, artifact detection, filtering, and converting the optical density to relative concentration.

8. A mild cognitive impairment recognition system based on fNIRS data under multi-task paradigm fusion, characterized in that, Including: A data input module for obtaining the fNIRS data of the subject obtained based on the resting-state task and the n-back working memory task; A feature extraction module for extracting fNIRS-related features based on the obtained fNIRS data; A feature selection module for performing feature selection on the extracted fNIRS-related features based on a feature selection algorithm to form a feature set; A feature recognition module for inputting the formed feature set into a trained deep learning model and outputting the mild cognitive impairment recognition result.

9. An apparatus for identifying mild cognitive impairment based on fNIRS data under multi-task paradigm fusion, characterized in that, Including a processor and a memory; the memory stores at least one instruction, and the at least one instruction is used to be executed by the processor to implement the steps of the method for recognizing mild cognitive impairment based on fNIRS data under the multi-task paradigm fusion as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method for recognizing mild cognitive impairment based on fNIRS data under the multi-task paradigm fusion as described in any one of claims 1 to 7.

Images