A data prediction device and system based on infrared spectroscopy

Abstract

The application provides a kind of data prediction device and system based on infrared spectrum, it is related to data processing technical field, the data prediction device includes infrared spectrum data acquisition module and prediction model.For any sampling period, infrared spectrum data acquisition module is used to obtain the test parameters of multiple test sample groups under the same test task based on infrared spectroscopy.Subsequently, for any test sample group, the prediction model extracts the target features of different behavior attributes in each test data sample in the current test sample group from the test parameters;And determine the coherence between the same target feature between two test data samples in the current test sample group, obtain the coherence parameter;Finally, based on the coherence parameter, determine the coupling difference between it and the theoretical data, obtain the prediction result.Based on this, the application can reduce the dependence of the existing data acquisition method on language and cognitive ability, and improve the effectiveness of evaluation data.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and more specifically, to a data prediction device and system based on infrared spectroscopy. Background Technology

[0002] Most medical aids in existing technologies, such as research data on autism spectrum disorder (ASD) in children, are obtained through behavioral scales, structured questionnaires, and clinical interviews. However, these methods are easily affected by situational factors and are limited by the language and cognitive abilities of the respondents, which can lead to a lack of objectivity, quantifiability, and validity in the data samples, thereby interfering with the assessor's judgment.

[0003] Therefore, there is an urgent need for a data sample evaluation method to reduce the dependence of existing data collection methods on language and cognitive abilities and to improve the effectiveness of the evaluation data. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a data prediction device and system based on infrared spectroscopy, which can reduce the dependence of existing data acquisition methods on language and cognitive abilities and improve the effectiveness of evaluation data.

[0005] To achieve the above objectives, the technical solutions adopted in the embodiments of the present invention are as follows:

[0006] In a first aspect, the present invention provides a data prediction device based on infrared spectroscopy, the data prediction device comprising: an infrared spectral data acquisition module and a prediction model, wherein the output end of the infrared spectral data acquisition module is connected to the input end of the prediction model;

[0007] The infrared spectral data acquisition module is used to acquire test parameters of multiple test sample groups under the same test task within any sampling period using infrared spectroscopy; wherein each test sample group includes two test data samples, and there is a correlation between the two test data samples;

[0008] For any test sample group, a prediction model is used to extract target features of different behavioral attributes from each test data sample within the current test sample group from the test parameters; wherein, each test parameter includes at least two target features of different behavioral attributes;

[0009] The prediction model is also used to determine the coherence between the same target feature between two test data samples in the current test sample group, and to obtain the coherence parameter;

[0010] The prediction model is also used to determine the coupling difference between the coherence parameters and the theoretical data, and to obtain the prediction results.

[0011] Optionally, the infrared spectral data acquisition module includes a data acquisition component and a data processing component. The data acquisition component includes a light source emitting unit and a detection unit. The detection unit includes multiple detectors, and the light source emitting unit includes multiple laser sources. Each laser source and a detector constitute an acquisition channel. The output of each detection unit is connected to the input of the data processing component.

[0012] The data acquisition component is used to acquire and transmit the raw light intensity data from each acquisition channel to the data processing component to obtain test parameters.

[0013] Optionally, when the test parameters include a vector of changes in hemoglobin concentration, the data processing component includes:

[0014] The preprocessing unit is used to receive and preprocess the raw light intensity data to obtain preprocessed light intensity data.

[0015] The optical density conversion unit is used to convert the preprocessed light intensity data into optical density data;

[0016] The hemoglobin conversion unit is used to determine the vector of hemoglobin concentration change based on optical density data.

[0017] Optionally, the preprocessing unit is also used to calculate and determine whether the mean of the original light intensity data under at least two different preset light wavelengths is within the preset standard deviation range, so as to determine whether each acquisition channel is normal at present;

[0018] Currently, all acquisition channels are functioning normally. The preprocessing unit is also used to filter raw light intensity data that meets preset rules from the raw light intensity data of each acquisition channel according to preset indicators, and obtain preprocessed light intensity data.

[0019] Optionally, when the test parameters include a first target feature and a second target feature with two different behavioral attributes, for any test sample group, the prediction model is also used to obtain the first target feature, the second target feature, and the basic feature under a preset time interval from the test parameters, wherein the value of the preset time interval is determined by the test task; the first target feature, the second target feature, and the basic feature are all in different time intervals;

[0020] The prediction model is also used to obtain the first coherence between two test data samples in the current test sample group under the first target feature, the second coherence under the second target feature, and the third coherence under the basic feature, respectively, to obtain the coherence parameter.

[0021] Optionally, the formula for calculating the first coherence is expressed as:

[0022] ;

[0023] in, Regarding time With wavelet scale The first coherence; Regarding time and wavelet scale The smoothing function; It is a complex conjugate function; The wavelet transform parameters are the parameters corresponding to the first target feature of the first sample in the current test sample group. The wavelet transform parameters corresponding to the first target feature of the second sample in the current test sample group.

[0024] Optionally, the prediction model is also used to determine the target frequency band corresponding to each target feature based on the coherence parameter;

[0025] The prediction model is also used to obtain prediction results based on the coupling difference between the target frequency band corresponding to each target feature and the theoretical data.

[0026] Optionally, for the first target feature, the prediction model is also used to calculate the first enhancement value based on the first coherence and the third coherence;

[0027] The prediction model is also used to perform statistical calculations on the first enhancement value to obtain the first target frequency band.

[0028] Optionally, the prediction model is also used to determine a first difference between the first target feature and the basic feature under the first target frequency band;

[0029] The prediction model is also used to determine the significance level of each sampling channel in the current test sample group based on the first difference using the single sample detection method, and to screen out the target sampling channels with a significance level higher than the preset threshold.

[0030] The prediction model is also used to determine the coupling difference between the first coherence and theoretical data under each target sampling channel, and to obtain the prediction results.

[0031] In a second aspect, the present invention also provides a data prediction system based on infrared spectroscopy, including the data prediction device of any one of the first aspects described above.

[0032] The data prediction device and system based on infrared spectroscopy provided in this invention have the following beneficial effects:

[0033] The infrared spectroscopy-based data prediction device in this application includes an infrared spectral data acquisition module and a prediction model. For any given sampling period, the infrared spectral data acquisition module acquires test parameters for multiple test sample groups under the same test task, based on infrared spectroscopy. Each test sample group includes two test data samples, and these two samples are correlated. Subsequently, for any test sample group, the prediction model extracts target features of different behavioral attributes from the test parameters of each test data sample within the current test sample group; it also determines the coherence between the same target feature between two test data samples in the current test sample group, obtaining a coherence parameter; finally, based on the coherence parameter, it determines the coupling difference between the parameter and the theoretical data, obtaining the prediction result. Based on this, this application can reduce the dependence of existing data acquisition methods on language and cognitive abilities and improve the effectiveness of evaluation data.

[0034] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0035] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 A schematic diagram of the data prediction device provided in an embodiment of the present invention is shown;

[0037] Figure 2 This shows one of the structural schematic diagrams of the infrared spectral data acquisition module provided in an embodiment of the present invention;

[0038] Figure 3 A schematic diagram of the acquisition channel of the data acquisition component provided in an embodiment of the present invention is shown;

[0039] Figure 4 This is a second schematic diagram of the structure of the infrared spectral data acquisition module provided in an embodiment of the present invention;

[0040] Figure 5 A schematic diagram of the structure of the data processing component provided in an embodiment of the present invention is shown.

[0041] Icons: 10-Data prediction device; 101-Infrared spectral data acquisition module; 102-Prediction model; 201-Data acquisition component; 202-Data processing component; 2011-Light source emission unit; 2012-Detection unit; 301-Detector; 302-Laser source; 2021-Preprocessing unit; 2022-Optical density conversion unit; 2023-Hemoglobin conversion unit. Detailed Implementation

[0042] 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 some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0043] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0044] It should be noted that relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0045] Please refer to Figure 1 , Figure 1 The diagram shows the structure of the infrared spectroscopy-based data prediction device provided in this embodiment. The data prediction device 10 includes an infrared spectroscopy data acquisition module 101 and a prediction model 102. The output of the infrared spectroscopy data acquisition module 101 is connected to the input of the prediction model 102.

[0046] In this embodiment, the infrared spectral data acquisition module 101 is used to acquire test parameters of multiple test sample groups under the same test task within any sampling period using infrared spectroscopy.

[0047] Each test sample group includes two test data samples, and there is a correlation between the two test data samples.

[0048] For any test sample group, prediction model 102 is used to extract target features of different behavioral attributes from each test data sample in the current test sample group from the test parameters.

[0049] Each test parameter includes at least two target features with different behavioral attributes.

[0050] Prediction model 102 is also used to determine the coherence between the same target feature between two test data samples in the current test sample group, and to obtain the coherence parameter.

[0051] Prediction model 102 is also used to determine the coupling difference between the coherence parameters and the theoretical data to obtain the prediction results.

[0052] It should be noted that in this embodiment, each test sample group includes two test data samples that are related to each other. This correlation can be characterized as the mutual influence and effect between the data. In one possible implementation, taking medical auxiliary data corresponding to childhood autism spectrum disorder as an example, the test data samples can be data samples with parent-child relationships. Correspondingly, the collected test parameters, i.e., neurological characteristics, will have a certain correlation due to blood relations, environmental factors, etc. Correspondingly, the prediction result can be understood as the degree of deviation between the experimental data (such as the aforementioned coherence parameters) obtained based on the test samples and the theoretical values.

[0053] To reduce the interference of uncertainties such as the external environment on the test data, the corresponding test task in this embodiment can be set in an independent experimental room, and external interference can be minimized by adjusting the ambient lighting, temperature, and humidity. Based on this, the test task in this embodiment can be set as multiple test data samples with parent-child relationships watching a specific movie or television program in the same experimental room. Then, the infrared spectral data acquisition module 101 uses infrared spectroscopy to acquire the test parameters of the test sample group within a sampling period, so that the prediction model 102 can obtain the target features representing different behavioral attributes from the obtained test parameters and perform data evaluation.

[0054] In this embodiment, behavioral attributes may include pain empathy attributes, theory of mind attributes, etc. Among them, pain empathy attributes can be characterized by scenarios such as small animals falling from a height or being pricked by sharp thorns, expressing pain and fright through visual and auditory means, and triggering pain empathy responses in the data samples to be tested; theory of mind attributes can be characterized by: role inferring others' intentions, understanding others' feelings, and shifts in trust and misunderstanding, etc.

[0055] Once the prediction model 102 obtains the target features representing different behavioral attributes, it will calculate the coherence between the target features of the same behavioral attribute among the test data samples with parent-child relationships, thereby obtaining the coherence parameter containing the coherence between multiple target features. Finally, based on the coupling difference between the coherence parameter and the theoretical data, the prediction result is obtained.

[0056] Please refer to Figure 2 , Figure 2 The diagram shows the structure of the infrared spectral data acquisition module provided in this embodiment. In this embodiment, the infrared spectral data acquisition module 101 includes a data acquisition component 201 and a data processing component 202. The data acquisition component 201 includes a light source emitting unit 2011 and a detection unit 2012.

[0057] In this embodiment, please Figure 2 Based on, refer to Figure 3 , Figure 3 This diagram shows another structural schematic of the infrared spectral data acquisition module provided in this embodiment. The detection unit 2012 includes multiple detectors 301, and the light source emission unit 2011 includes multiple laser sources 302. Each laser source 302 and a detector 301 form an acquisition channel. The output terminal of each detector 301 is connected to the input terminal of the data processing component 202.

[0058] Taking any laser source 302 as an example, when the detection unit 2012 includes n detectors 301, the laser source 302 can form n sampling channels with the n detectors 301 under the detection unit 2012.

[0059] In this embodiment, the data acquisition component 201 is used to acquire and transmit the raw light intensity data obtained from each acquisition channel to the data processing component 202.

[0060] In this embodiment, the light source emitting unit 2011 may include two laser sources 302 of different wavelengths for use in predicting the test parameters calculated by the model 102. In one possible implementation, the data acquisition component 201 corresponding to each sample group under test may consist of eight laser sources 302 and eight detectors 301, wherein the wavelengths corresponding to the laser sources 302 may be 760nm and 850nm. The corresponding sampling frequency is 7.81Hz.

[0061] In this embodiment, when the test parameters include the hemoglobin concentration change vector, the data acquisition component 201 can be set as an optical electrode cap for easy sampling. That is, the above-mentioned 8 laser sources 302 and 8 detectors 301 constitute 16 optical electrodes, and each optical electrode is set on the optical electrode cap to correspond to a specific location area of ​​the brain.

[0062] In this embodiment, each laser source 302 and a detector 301 form a channel. Please refer to [reference needed]. Figure 4 , Figure 4 The diagram shows the acquisition channels of the data acquisition component in this embodiment; the numbers in the box area in the diagram represent the corresponding acquisition channel numbers. The sampling area of ​​each acquisition channel can be the area radiated from the midpoint of the line connecting the laser source 302 (Emitter) and the detector 301 (Detector) to the cerebral cortex. Then, the consistency of the test space under one acquisition cycle is determined by the position coordinates of the midpoint.

[0063] Please continue to refer to this. Figure 3 In this embodiment, 12 acquisition channels can be constructed in the prefrontal cortex (PFC), and the corresponding brain regions covered include the frontopolar area, i.e. Figure 4 The area where acquisition channels 1, 5, and 8 are located; the orbitofrontal area, i.e. Figure 4 The area where acquisition channels 2, 9, and 10 are located; the dorsolateral prefrontal cortex, i.e. Figure 4 The areas where acquisition channels 3, 4, 11, and 12 are located.

[0064] Simultaneously, eight acquisition channels were constructed in the right temporal lobe cortex, corresponding to the brain regions covered, including the primary somatosensory cortex. Figure 4 The area where acquisition channel 13 is located; the pre-motor and supplementary motor cortex, i.e. Figure 4 The area where acquisition channels 14 and 16 are located; the subcentral area, i.e. Figure 4 The area where acquisition channel 15 is located; the Broca area (pars opercularis, part of Broca's area), i.e. Figure 4 The area where acquisition channels 17, 19, and 20 are located.

[0065] Please continue to refer to this. Figure 3In one possible implementation, the center of the laser cap in the above embodiment is marked at the midpoint of the line connecting the occipital protuberance to the root of the nose (i.e., the Cz position). Using this as a reference, Fp1 / Fp2 are positioned at the frontal pole, and P8 / T8 are positioned at the right temporoparietal junction. Based on this, the laser source can be aligned during actual testing. Figure 3 Light source calibration point set, detector alignment Figure 3 The calibration point set for the detector, in this embodiment, includes calibration points S1, S2, S3, S4, S5, S6, S7, and S8; the calibration point set for the detector includes calibration points D1, D2, D3, D4, D5, D6, D7, and D8. Based on this, this embodiment can ensure the accuracy of data acquisition by aligning the electrode positions of the above calibration points.

[0066] Based on this, in this embodiment, each data sample to be tested can obtain the original light intensity data corresponding to the current position through 20 sampling channels to obtain test parameters.

[0067] To improve data processing efficiency and ensure the accuracy of data results, please refer to... Figure 5 , Figure 5 The diagram shows the structure of the data processing component in this embodiment. The data processing component 202 includes a preprocessing unit 2021, an optical density conversion unit 2022, and a hemoglobin conversion unit 2023 connected in sequence.

[0068] In this embodiment, the preprocessing unit 2021 is used to receive and preprocess the original light intensity data to obtain preprocessed light intensity data.

[0069] The optical density conversion unit 2022 is used to convert preprocessed light intensity data into optical density data.

[0070] The hemoglobin conversion unit 2023 is used to determine the hemoglobin concentration change vector based on optical density data.

[0071] In this embodiment, the preprocessing unit 2021 is used to calculate and determine whether the mean value of the original light intensity data under at least two different preset light wavelengths is within the preset standard deviation range, so as to determine whether each acquisition channel is normal at present.

[0072] When all acquisition channels are in normal condition, the preprocessing unit 2021 is also used to filter the original light intensity data that meets the preset rules from the original light intensity data of each acquisition channel according to preset indicators, and obtain the preprocessed light intensity data.

[0073] In this embodiment, the preprocessing unit 2021 can obtain the original intensity signals under two light wavelengths and calculate the mean and standard deviation of the original intensity signals under each wavelength to determine whether the corresponding sampling channel is abnormal. If the corresponding acquisition channel is in a normal state, quality detection can be performed based on the original light intensity data corresponding to the current acquisition channel.

[0074] In one possible implementation, the detection method could be as follows: the signal strength, signal-to-noise ratio (SNR), and source-detector (SD) distance of each acquisition channel could be evaluated using the enPruneChannels function, and channels that do not meet the quality standards could be automatically marked and removed.

[0075] Taking the signal strength judgment scheme as an example, it can be set that the light intensity corresponding to the original light intensity data under all acquisition channels must be in the range of [0, 5]. If the original light intensity signal of a certain acquisition channel exceeds this range, it is determined that the signal under that acquisition channel is abnormal, and the original light intensity signal corresponding to that acquisition channel is removed from the test parameters to obtain the preprocessed light intensity data.

[0076] Similarly, for the signal-to-noise ratio (SNR) judgment scheme, after obtaining the SNR value based on the ratio between the mean and standard deviation of the original light intensity signal in each acquisition channel, it is determined whether the signal in the current acquisition channel is abnormal based on the preset SNR parameters. If abnormal, the original light intensity signal corresponding to that acquisition channel is removed from the test parameters to obtain the preprocessed light intensity data.

[0077] Similarly, for the source-detector distance judgment scheme, it is determined whether the detection distance of the source-detector corresponding to the current acquisition channel falls within the preset range, i.e. [0, 60 mm]. If it exceeds the range, the original light intensity signal corresponding to the acquisition channel is removed from the test parameters to obtain the preprocessed light intensity data.

[0078] After obtaining the preprocessed light intensity data, it can be converted into optical density data using the Lambert-Beer law. Wherein, any optical density data... The conversion formula can be expressed as: ;in, express t The light intensity measured at each moment, I 0 represents the baseline intensity. Based on this, this embodiment can eliminate the interference of changes in the light source intensity itself on the data, thereby highlighting the components directly related to changes in hemoglobin concentration.

[0079] To ensure the accuracy of the hemoglobin concentration change vector, this embodiment can also process the aforementioned optical density data. Noise reduction processing is performed.

[0080] In one possible implementation, for example, optical density data can be processed. Perform kurtosis-based wavelet filtering for noise reduction, i.e., first process the optical density data... Discrete wavelet transform is performed, and then the peak values ​​corresponding to the high-frequency detail coefficients at each scale are calculated. If the peak value is greater than 3.3, it is determined to be a noise coefficient, and the noise coefficient is thresholded to reconstruct the signal and remove the optical density data. Mid-to-high frequency artifacts.

[0081] To remove global systemic artifacts, such as body swaying or system drift that occurs during testing, principal component analysis (PCA) can be used to denoise the OD signal first. The denoised OD signal... OD denoised It can be represented as: OD denoised =T k P k ;in, T k Indicates the previous One principal component coefficient matrix, P k is the corresponding space vector matrix.

[0082] In addition, this embodiment can also use a Butterworth filter to perform bandpass filtering on the signal to filter out low-frequency drift (<0.01 Hz, such as instrument thermal drift) and high-frequency physiological noise (>1 Hz, such as heartbeat and electromyography signals).

[0083] It should be noted that this embodiment does not limit the denoising processing scheme used, and may include, but is not limited to, the above-mentioned wavelet filtering denoising based on kurtosis, principal component analysis denoising, and bandpass filtering, and may also include motion artifact detection and correction.

[0084] After obtaining the final optical density data, the hemoglobin conversion unit 2023 can be used to convert the data into a hemoglobin concentration change vector, and the corresponding calculation method can be expressed as follows: ;in, Let the vector be the change in optical density. Here is the extinction coefficient matrix. The source-detector distance is in cm, and DPF is the differential path factor. This represents the vector representing the change in hemoglobin concentration. In this embodiment, the vector represents the change in hemoglobin concentration. Including vector of changes in oxyhemoglobin concentration Vector of changes in deoxyhemoglobin concentration Vector of changes in total hemoglobin concentration .

[0085] In one possible implementation, this embodiment uses the hemoglobin concentration change vector. The vector of changes in oxyhemoglobin concentration is preferred. To conduct subsequent data predictions.

[0086] After obtaining the aforementioned hemoglobin concentration change vector, this embodiment can select two behavioral attributes from the test parameters for analysis, such as the aforementioned pain empathy attribute and theory of mind attribute, and set the pain empathy attribute as the first target feature and the theory of mind attribute as the second target feature. The basic feature can be characterized as the behavioral attributes during the baseline period of the test task, for example, the neural characteristics of the test data sample corresponding to the first 30 seconds of a specific movie clip. It should be noted that this basic feature can provide a reference standard for the first and second target features, but does not possess specific behavioral attributes.

[0087] Specifically, for any test sample group, the prediction model 102 is also used to obtain the first target feature, the second target feature, and the basic feature from the test parameters within a preset time interval, wherein the value of the preset time interval is determined by the test task; the first target feature, the second target feature, and the basic feature are all in different time intervals.

[0088] The prediction model 102 is also used to obtain the first coherence between two test data samples in the current test sample group under the first target feature, the second coherence under the second target feature, and the third coherence under the basic feature, respectively, to obtain the coherence parameter.

[0089] In this embodiment, the specific movie segments can be tagged first. For example, the first 30 seconds of the movie (0-30 seconds) can be set as the baseline period, the 86-second segment containing the pain empathy segment can be set as the pain empathy period, and the 68-second segment containing the theory of mind segment can be set as the theory of mind period. Then, the prediction model 102 extracts the basic features, the first target feature, and the second target feature of each data sample to be tested in the current sampling period from the baseline period, the pain empathy period, and the theory of mind period, respectively.

[0090] Finally, prediction model 102 calculates the coherence of two test data samples within the test sample group that have a parent-child relationship based on the aforementioned basic features, first target features, and second target features.

[0091] In one feasible manner, when the data prediction model of this application is used to obtain medical auxiliary data on childhood autism spectrum disorder, the coherence can be used to characterize the interpersonal neural synchronization (INS) of oxyhemoglobin (HbO) obtained for each test sample group during movie viewing.

[0092] It should be noted that this embodiment does not limit the method of calculating coherence. In one possible implementation, the formula for calculating the first coherence can be expressed as:

[0093] ;

[0094] in, Regarding time With wavelet scale The first coherence; Regarding time and wavelet scale The smoothing function; It is a complex conjugate function; The wavelet transform parameters are the parameters corresponding to the first target feature of the first sample in the current test sample group. The wavelet transform parameters corresponding to the first target feature of the second sample in the current test sample group.

[0095] After obtaining the coherence parameters of a sample group to be tested, the prediction model 102 in this embodiment is also used to determine the target frequency band corresponding to each target feature based on the coherence parameters.

[0096] The prediction model 102 is also used to obtain prediction results based on the coupling difference between the target frequency band corresponding to each target feature and the theoretical data.

[0097] Taking the first target feature as an example, in this embodiment, the prediction model 102 is also used to calculate the first enhancement value based on the first coherence and the third coherence, and then perform statistical calculation on the first enhancement value to obtain the first target frequency band.

[0098] Taking the aforementioned oxyhemoglobin concentration change vector as an example, the oxyhemoglobin concentration change vector of each test sample group under the same sampling channel can be extracted. Within the sampling period, for example, within the frequency range of 0.01-1 Hz, the coherence of each test sample group in each frequency band is calculated, and then the first coherence, second coherence, and third coherence are extracted respectively. Then, the third coherence is subtracted from the first coherence and second coherence, and their mean values ​​are calculated accordingly to obtain the enhancement values ​​of the pain empathy attribute and the theory of mind attribute based on the baseline, namely the first enhancement value and the second enhancement value. Subsequently, Fisher's algorithm is applied to the above-mentioned first enhancement value and second enhancement value. z Transformation was performed to achieve normalization, followed by single-sample processing. t The test determined the final target frequency bands specific to the pain empathy attribute and the theory of mind attribute, namely the first target frequency band and the second target frequency band.

[0099] In one possible implementation, the first target frequency band can be in the range of 0.03-0.06 Hz; the second target frequency band can be in the range of 0.03-0.05 Hz.

[0100] Subsequently, taking the first target feature as an example, prediction model 102 is also used to determine the first difference between the first target feature and the basic feature under the first target frequency band.

[0101] The prediction model 102 is also used to determine the significance level of each sampling channel in the current sample group based on the first difference using the single-sample detection method, and to screen out the target sampling channels with a significance level higher than the preset threshold.

[0102] Prediction model 102 is also used to determine the coupling difference between the first coherence and theoretical data under each target sampling channel, and to obtain the prediction results.

[0103] In this embodiment, the first target feature and the second target feature can be calculated based on their respective target frequency band ranges. The average value of the result after subtracting the coherence value of the third target feature is then used as the first difference corresponding to the first target feature and the second difference corresponding to the second target feature. Fisher's algorithm is then applied again. z Transformation. Taking the first target feature as an example, prediction model 102 can again transform the Fisher feature. z The coherence values ​​of each sampled channel after conversion are used for single-sample processing. t The test determines which sampling channels exhibit significant enhancement during the testing task. One possible approach is to determine if the significance level of each sampling channel is less than 0.05; if so, the current sampling channel is defined as significantly enhanced. Subsequently, a second independent sampling is performed on the sampling channels that are significantly enhanced. t The test results are then compared with theoretical data to determine the coupling differences, and an evaluation result is generated.

[0104] It should be noted that this embodiment does not limit the way the above evaluation results are presented, including but not limited to statistical values ​​such as deviation ratio or deviation mean.

[0105] Similarly, in this embodiment, the prediction model can calculate the second enhancement value based on the second coherence and the third coherence, and then perform statistical processing on the second enhancement value to obtain the second target frequency band, thereby obtaining the prediction result for the second target features, which will not be elaborated here.

[0106] Based on this, the infrared spectroscopy-based data prediction device provided in this application includes an infrared spectral data acquisition module and a prediction model. For any sampling period, the infrared spectral data acquisition module is used to acquire test parameters for multiple test sample groups under the same test task based on infrared spectroscopy. Each test sample group includes two test data samples, and there is a correlation between the two test data samples. Subsequently, for any test sample group, the prediction model extracts target features of different behavioral attributes from the test parameters of each test data sample in the current test sample group; and determines the coherence between the same target feature between the two test data samples in the current test sample group to obtain a coherence parameter; finally, based on the coherence parameter, it determines the coupling difference between the parameter and the theoretical data to obtain the prediction result. Based on this, this application can reduce the dependence of existing data acquisition methods on language and cognitive abilities and improve the effectiveness of evaluation data.

[0107] Following the same approach as the previous embodiment, the present invention also provides a data prediction system based on infrared spectroscopy, including the data prediction device of any of the previous embodiments, to reduce the dependence of existing data acquisition methods on language and cognitive abilities and improve the effectiveness of evaluation data.

[0108] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can also be implemented in other ways. The apparatus embodiments described above are merely illustrative; for example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0109] In addition, the functional modules in the various embodiments of the present invention can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0110] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0111] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A data prediction device based on infrared spectroscopy, characterized in that, The data prediction device includes: an infrared spectral data acquisition module and a prediction model, wherein the output end of the infrared spectral data acquisition module is connected to the input end of the prediction model; The infrared spectral data acquisition module is used to acquire test parameters of multiple test sample groups under the same test task within any sampling period using infrared spectroscopy; wherein each test sample group includes two test data samples, and there is a correlation between the two test data samples; For any test sample group, the prediction model is used to extract target features of different behavioral attributes from each test data sample in the current test sample group from the test parameters; wherein, each test parameter includes at least two target features of different behavioral attributes; The prediction model is also used to determine the coherence between the same target feature between two data samples in the current test sample group, and to obtain the coherence parameter; The prediction model is also used to determine the coupling difference between the coherence parameter and the theoretical data, and to obtain the prediction result; When the test parameters include a first target feature, a second target feature, and a basic feature with two different behavioral attributes, for any test sample group... The prediction model is further used to obtain a first target feature, a second target feature, and a basic feature from the test parameters within a preset time interval, wherein the value of the preset time interval is determined by the test task; the first target feature, the second target feature, and the basic feature are all in different time intervals; wherein the basic feature is used to characterize the behavioral attributes of the baseline period in the test task. The prediction model is also used to obtain the first coherence between two test data samples in the current test sample group under the first target feature, the second coherence under the second target feature, and the third coherence under the basic feature, respectively, to obtain the coherence parameter. The prediction model is also used to determine the target frequency band corresponding to each target feature based on the coherence parameter; The prediction model is also used to obtain prediction results based on the coupling difference between the target frequency band corresponding to each target feature and the theoretical data; For the first target feature, the prediction model is further configured to determine a first enhancement value based on the difference between the first coherence and the third coherence; The prediction model is further used to perform a Fisher transform on the first enhancement value to obtain a transformed first enhancement value, and to determine the first target frequency band based on the transformed first enhancement value using a one-sample t-test.

2. The data prediction device according to claim 1, characterized in that, The infrared spectral data acquisition module includes a data acquisition component and a data processing component. The data acquisition component includes a light source emitting unit and a detection unit. The detection unit includes multiple detectors, and the light source emitting unit includes multiple laser sources. Each laser source and a detector form an acquisition channel. The output terminal of each detection unit is connected to the input terminal of the data processing component. The data acquisition component is used to acquire and transmit the raw light intensity data obtained from each acquisition channel to the data processing component.

3. The data prediction device according to claim 2, characterized in that, When the test parameters include a vector of changes in hemoglobin concentration, the data processing component includes: The preprocessing unit is used to receive and preprocess the original light intensity data to obtain preprocessed light intensity data. Optical density conversion unit is used to convert the preprocessed light intensity data into optical density data; The hemoglobin conversion unit is used to determine the hemoglobin concentration change vector based on the optical density data.

4. The data prediction device according to claim 3, characterized in that, The preprocessing unit is also used to calculate and determine whether the mean value of the original light intensity data under at least two different preset light wavelengths is within the preset standard deviation range, so as to determine whether each acquisition channel is normal at present. Currently, all acquisition channels are functioning normally. The preprocessing unit is also used to filter raw light intensity data that meets preset rules from the raw light intensity data of each acquisition channel according to preset indicators, and obtain preprocessed light intensity data. The preset indicators include the signal strength and / or signal-to-noise ratio and / or source-detector distance of the acquisition channel; the preset rules are used to characterize preset signal strength values ​​and / or preset signal-to-noise ratio values ​​and / or preset source-detector distance values.

5. The data prediction device according to claim 1, characterized in that, The formula for calculating the first coherence is as follows: ； in, Regarding time With wavelet scale The first coherence; Regarding time and wavelet scale The smoothing function; It is a complex conjugate function; The wavelet transform parameters are the parameters corresponding to the first target feature of the first sample in the current test sample group. The wavelet transform parameters corresponding to the first target feature of the second sample in the current test sample group.

6. The data prediction device according to claim 1, characterized in that, The prediction model is also used to determine a first difference between the first target feature and the basic feature under the first target frequency band; The prediction model is also used to determine the significance level of each sampling channel in the current sample group under test based on the first difference using the single sample detection method, and to screen out the target sampling channels with a significance level higher than a preset threshold. The prediction model is also used to determine the coupling difference between the first coherence and the theoretical data under each target sampling channel, and to obtain the prediction results.

7. A data prediction system based on infrared spectroscopy, characterized in that, Includes the data prediction device according to any one of claims 1 to 6.

Images