A method for detecting emotional feedback effect of body posture based on fNIRS

Abstract

The application discloses a detection method for emotional feedback effect of body posture based on fNIRS, and comprises the following steps: collecting a plurality of fNIRS light intensity data F ij , each fNIRS light intensity data F ij comprises fNIRS light intensity data on a plurality of channels; removing inferior data in the plurality of fNIRS light intensity data F ij obtained in step one, and obtaining M pieces of preferred fNIRS light intensity data F ij after optimization; pre-processing the M pieces of fNIRS light intensity data F ij , and obtaining M pieces of fNIRS oxyhemoglobin concentration data f ij ; respectively detecting the M pieces of fNIRS oxyhemoglobin concentration data f ij , and obtaining detection data; and through quantitative and statistical analysis of neural response state and brain function connection network mode of related brain regions when posture-induced emotion occurs, high sensitivity and specificity quantitative analysis of emotional state is realized, the problem of large subjective difference of scorers and insufficient analysis accuracy in previous methods is overcome, and the problem that posture-induced emotional feedback effect cannot be objectively and quantitatively analyzed in the prior art is solved.

Description

Technical Field

[0001] This invention belongs to the field of image processing and relates to image analysis methods, specifically a method for detecting the effect of body posture on emotional feedback based on fNIRS. Background Technology

[0002] Statistics show that many people experience varying degrees of mental health problems, manifesting as depression, lack of purpose, confusion, lack of competence, academic difficulties, a large and ever-increasing amount of confusion, strained interpersonal relationships, and other mental health issues. These widespread problems pose significant challenges to self-regulation and mental health intervention. Research indicates that regulating emotions through body posture and behavior can enhance self-reference, improve mental health, strengthen psychosomatic practices in clinical settings, and regulate emotions in work and learning environments. A growing number of people recognize the dynamic coupling between mind and body, influencing cognition and emotion, significantly enriching our scientific understanding of mental health. Embodied theory proposes that the motor system (i.e., posture and movement) influences our emotional and behavioral responses to the world, emphasizing the importance of physical movement in cognitive processes. It argues that the body is not merely a "container" for mental processes but also provides authentic experiences and feelings, giving the body a decisive role in shaping cognition. Over a century of research has found that the neural circuits of mind-body coupling originate from three distinct networks: motor, cognitive, and emotional networks. These three networks likely provide the neural circuits that link motor, emotional, and cognitive control processes with immediate responses. The motor network provides the link between physical movement and stress regulation; the cognitive network acquires memory and executive functions mediated by the lateral prefrontal cortex and processes cognitive and emotional information from the anterior cingulate cortex; and the emotional network, including the prefrontal and temporal cortices, provides emotional sensation and regulation. Therefore, studying the influence of energy posture on emotion is both representative and significant.

[0003] Embodied cognition theory categorizes body postures into three types: open and relaxed high-energy postures, curled-up low-energy postures, and natural postures. Through numerous empirical studies using emotion scales and a single case of electroencephalography (EEG) neuroimaging, the influence of energy postures on emotional responses has been explored. Energy postures are generally considered to reflect an individual's intention or motivational state, manifested as engagement or withdrawal from the internal or external world. While empirical studies have provided valuable qualitative insights, the method suffers from low sensitivity and specificity, significant inter-rater variability, and fails to elucidate the underlying neural mechanisms of emotional feedback. Furthermore, posture-emotion EEG studies only compare neural activity differences between high-energy and low-energy postures, failing to precisely pinpoint brain regions activated by emotion, thus not meeting the need for imaging brain regions responsible for fine emotional responses. Additionally, they cannot reflect the neural activity states and functional network connectivity patterns of relevant brain regions. Emotion, as a complex interaction, involves internal and external factors, including our ability to perceive bodily signals. Quantifying the perception of internal bodily signals is crucial for understanding the role of energy postures in emotional responses.

[0004] fNIRS offers a promising approach to assess changes in cortical activity related to emotional states by non-invasively measuring hemodynamic responses in the brain. When external stimuli trigger activity in functional brain regions, a complex series of regulatory activities occur, including local vasodilation, increased cerebral blood flow, and increased cerebral blood volume. This leads to an increase in the concentration of oxygenated hemoglobin (HbO2) and a decrease in the concentration of deoxygenated hemoglobin (HbR) in the blood of those brain regions. fNIRS indirectly reflects local brain functional activity and its patterns of change by detecting these slowly changing hemodynamic responses. Compared to the aforementioned empirical studies, fNIRS provides a more direct and sensitive means of neurophysiological monitoring, tracking hemodynamic changes in the brain under specific postures in real time and non-invasively, thereby revealing the neural mechanisms by which energy posture affects emotional states. This technology significantly reduces the subjectivity of human ratings, enhances the objectivity and reliability of the data, and allows researchers to explore the neural basis of emotional responses with greater precision, including activity patterns in key brain regions such as the prefrontal cortex, amygdala, and cingulate cortex.

[0005] Compared to EEG studies of posture-emotion interaction, fNIRS demonstrates significant advantages in assessing the relationship between posture and emotion. Its lower sensitivity to motion artifacts ensures stable recording of brain activity even under experimental conditions with frequent changes in body posture, effectively reducing signal interference. Furthermore, the high spatial resolution of fNIRS on the surface of the cerebral cortex makes it possible to precisely locate brain regions activated by emotion, meeting the need for detailed brain region imaging in emotion research. More importantly, fNIRS can simultaneously monitor the activity of multiple brain regions, revealing the functional connectivity networks between them and reflecting the synergistic effects of multiple brain regions. This is crucial for understanding the complex network mechanisms of emotion generation, making fNIRS highly suitable for research on posture-emotion interaction.

[0006] Therefore, there is an urgent need for a method based on fNIRS to detect the effect of body posture on emotional feedback, so as to detect changes in brain functional activity through fNIRS technology and explore the possible emotional feedback effects of different body postures on people. Summary of the Invention

[0007] To address the shortcomings of existing technologies, the present invention aims to provide a method for detecting the effect of body posture on emotional feedback based on fNIRS, thereby solving the technical problems of existing technologies that cannot objectively and quantitatively analyze the effect of posture-induced emotional feedback and cannot locate brain regions with fine emotional responses.

[0008] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0009] A method for detecting the effect of body posture on emotional feedback based on fNIRS includes the following steps:

[0010] Step 1: Collect multiple fNIRS light intensity data F ij Each fNIRS light intensity data F ij Includes fNIRS light intensity data across multiple channels;

[0011] The channel is an optical transmission path between an exciter and a detector;

[0012] i represents the participant's posture, with values ​​of 1, 2, and 3, corresponding to the natural posture REST, low-energy posture LP, and high-energy posture HP, respectively.

[0013] j represents the eight regions of interest of the participant, with values ​​of 1, 2, ..., 8, corresponding to the left temporal cortex, right temporal cortex, dorsolateral prefrontal cortex, left dorsolateral prefrontal cortex, right dorsolateral prefrontal cortex, ventromedial prefrontal cortex, left ventrolateral prefrontal cortex, and right ventrolateral prefrontal cortex, respectively;

[0014] Step 2: Remove multiple fNIRS intensity data obtained in Step 1. ij From the inferior data, the optimal M fNIRS light intensity data F were obtained. ij ;

[0015] Step 3, process the M fNIRS intensity data obtained in Step 2. ij Preprocessing was performed to obtain M fNIRS oxygenated hemoglobin concentration data. ij ;

[0016] Step four, analyze the M fNIRS oxygenated hemoglobin concentration data obtained in step three. ij Conduct testing to obtain test data;

[0017] The detection data includes the R-BFCN (Radio-Based Brain Functional Connectivity Network) for emotional feedback under different postures. i C-BFCN (Channel Brain Functional Connectivity Network) for Emotional Feedback in Different Postures i The brain functional connectivity network (R-BFCN) of the region of interest during emotional feedback in different postures. i Significantly different ROI connection edges, and the channel brain functional connectivity network (C-BFCN) for emotional feedback under different postures. i The study included channels with significant differences in connectivity, the intensity of neural activity during emotional feedback, the speed of neural activity and the trend of neural responses, as well as channels with significant differences corresponding to different eigenvalues.

[0018] This invention also includes the following technical features:

[0019] Step two specifically includes:

[0020] Calculate the multiple fNIRS light intensity data obtained in step one. ij Each fNIRS light intensity data F ij The signal-to-noise ratio of each channel is used to determine the intensity data F of each fNIRS. ij If the number of channels with a signal-to-noise ratio below the first threshold exceeds one-tenth of the total number of channels, then that fNIRS intensity data is discarded to obtain the optimized fNIRS intensity data F. ij ;

[0021] Alternatively, calculate the multiple fNIRS light intensity data F obtained in step one. ij For each channel of each fNIRS intensity data point, the quotient of the standard deviation and the mean is used to determine whether the number of channels whose quotient exceeds a second threshold exceeds one-tenth of the total number of channels. If so, that fNIRS intensity data point is discarded, resulting in the optimized fNIRS intensity data F. ij .

[0022] Step three specifically involves:

[0023] Step 3.1: A dual filtering mechanism based on a standard deviation threshold and an amplitude threshold, set under a time window, is used to eliminate the M fNIRS light intensity data F obtained in Step 3. ij Interference from motion artifacts in the m-th fNIRS intensity data;

[0024] Step 3.2: Use a bandpass filter to remove baseline drift, artificial noise, physiological noise, and drift noise interference introduced by the instrument's own signal from the m-th fNIRS light intensity data processed in Step 3.1;

[0025] The physiological noises include heartbeat, respiration, and Mayer waves;

[0026] Step 3.3: According to the modified Beer-Lambert law, the m-th fNIRS intensity data F processed in step 3.2 is processed... ij Convert to fNIRS oxygenated hemoglobin concentration data f ij ;

[0027] Step 3.4: Let m = m + 1, return to step 3.1, and continue until m = M, obtaining M fNIRS oxygenated hemoglobin concentration data. ij ;

[0028] M represents the fNIRS light intensity data. ij Total number of records; m represents fNIRS light intensity data F ij The serial number.

[0029] Step four specifically includes the following steps:

[0030] Step 4.1, analyze the M fNIRS oxygenated hemoglobin concentration data obtained in Step 3. ij The Pearson correlation coefficient was calculated for the correlation between regions of interest (ROIs) and channels in the brain to obtain the R-value. Based on the R-value, the brain functional connectivity network (R-BFCN) for emotional feedback under different postures was obtained. i C-BFCN (Channel Brain Functional Connectivity Network) for Emotional Feedback in Different Postures i ;

[0031] Step 4.2: One-way ANOVA was used to statistically analyze the R-values ​​corresponding to the correlation of the region of interest (ROI) and the correlation of the channels obtained in Step 4, respectively. Statistical values ​​were obtained, and the corresponding brain functional connectivity network (R-BFCN) of the ROI during emotional feedback under different postures was obtained based on the statistical values. i Significantly different ROI connection edges and the channel brain functional connectivity network (C-BFCN) for emotional feedback under different postures. i Channel connection edges with significant differences;

[0032] Step 4.3: Extract the M fNIRS oxygenated hemoglobin concentration data obtained in Step 3. ij eigenvalues;

[0033] The eigenvalues ​​include integral values, centroid values, and slope coefficients;

[0034] Step 4.4: Calculate and compare the fNIRS oxygenated hemoglobin concentration data obtained in Step 4.3. 2j and fNIRS oxygenated hemoglobin concentration data f 3j The average integral value, average centroid value, and average slope coefficient of the same channel are used to obtain the intensity of neural activity, the speed of neural activity, and the trend of neural response during emotional feedback.

[0035] Step 4.5: Use one-way ANOVA to analyze each fNIRS oxygenated hemoglobin concentration data obtained in Step 4.3. ij Statistical analysis was performed on the three feature values ​​corresponding to each channel to obtain statistical values. The false positive results in the statistical values ​​were corrected by the false detection rate correction method. The significance level p was set to be less than 0.05 to obtain the channels with significant differences corresponding to different feature values.

[0036] Step 4.1 specifically includes the following steps:

[0037] Step 4.1.1, analyze the fNIRS oxygenated hemoglobin concentration data obtained in Step 3. ij The correlation between any two regions of interest (ROIs) is calculated using the Pearson correlation coefficient to obtain the R-value. The absolute value of the R-value is then checked; if it is greater than zero, an edge connects the two ROIs; otherwise, no edge connects them, and the R-value is set to 0. The resulting network, consisting of all edges and eight ROIs, forms the R-BFCN (Radio-Oriented Brain Functional Connectivity Network) for emotional feedback under different postures. i ;

[0038] Step 4.1.2: Let i = i + 1, return to step 4.1.1, until i > 3, and output the ROI brain functional connectivity network R-BFCN for emotional feedback under different postures. i ;

[0039] Step 4.1.3, analyze the fNIRS oxygenated hemoglobin concentration data. 3jThe correlation between any two channels is calculated using the Pearson correlation coefficient to obtain the R value. The absolute value of the R value is then checked against a first threshold. If the absolute value is greater than a first threshold, a connecting edge exists between the two channels; otherwise, no connecting edge exists. All edges obtained, along with the two channels, form the channel-brain functional connectivity network (C-BFCN) for emotional feedback under different postures. i ;

[0040] Step 4.1.4: Let i = i + 1, return to step 4.1.3, until i > 3, and output the channel brain functional connectivity network (C-BFCN) for emotional feedback under different postures. i .

[0041] Step 4.2 specifically includes the following steps:

[0042] Step 4.2.1: One-way ANOVA was used to statistically analyze all R values ​​of any two groups obtained in Step 4.1.1, obtaining statistical values. The FDR correction method was used to correct for false positive results in the statistical values, and ROI connections with a significance level p greater than 0.05 were removed. This yielded ROI connections with significant differences in the ROI brain functional connectivity network during emotional feedback under different postures, specifically the R-BFCN between high-energy posture (HP) and low-energy posture (LP), between high-energy posture (HP) and natural posture (REST), and between low-energy posture (LP) and natural posture (REST). i ROI connection edges with significant differences;

[0043] Step 4.2.2: One-way ANOVA was used to statistically analyze the R-values ​​of each channel in any two groups obtained in Step 4.1.3, obtaining statistical values. The FDR correction method was used to correct for false positives in the statistical values, and channel connections with a significance level greater than 0.05 were removed. This yielded channel connections with significant differences in the channel-brain functional connectivity network during emotional feedback under different postures, specifically the C-BFCN (Channel-Brain Functional Connectivity Network) between high-energy posture (HP) and low-energy posture (LP), between high-energy posture (HP) and natural posture (REST), and between low-energy posture (LP) and natural posture (REST) ​​during emotional feedback under different postures. i Channel connections with significant differences.

[0044] Step 4.4 specifically includes the following steps:

[0045] Step 4.4.1, respectively, the fNIRS oxygenated hemoglobin concentration data f 2j and fNIRS oxygenated hemoglobin concentration data f 3jThe integral values, centroid values, and slope coefficients of the same channel are accumulated and added together to obtain the total integral value, total centroid value, and total slope coefficient of each channel;

[0046] Step 4.4.2: The total integral value, total centroid value and total slope coefficient of each channel obtained in step 4.4.1 are averaged to obtain the average integral value, average centroid value and average slope coefficient of each channel.

[0047] Step 4.4.3, compare the fNIRS oxygenated hemoglobin concentration data obtained in step 4.4.2 with the fNIRS data. 2j and fNIRS oxygenated hemoglobin concentration data f 3j The average integral value, average centroid value, and average slope coefficient of the same channel;

[0048] The average integral value is used to evaluate the intensity of neural activity during emotional feedback.

[0049] The average center of gravity value is used to evaluate the speed of neural activity during emotional feedback.

[0050] The average slope coefficient is used to evaluate the trend of neural responses to emotional feedback.

[0051] Compared with the prior art, the beneficial technical effects of this invention are:

[0052] (I) In this invention, by quantifying and statistically analyzing the neural response state and brain functional connectivity network pattern of the relevant brain regions when posture induces emotion, a highly sensitive and specific quantitative analysis of emotional state is achieved. This overcomes the problems of large subjective differences among raters and insufficient analysis accuracy in previous methods, and solves the problem that the feedback effect of posture-induced emotion cannot be objectively and quantitatively analyzed in the existing technology.

[0053] (II) In this invention, various features of neural responses induced by different postures are statistically calculated to accurately locate brain regions with different levels and speeds of neural activity during emotional feedback. At the same time, the synergistic effect and differentiated activity of multiple brain regions during emotional feedback are monitored, solving the technical problem of being unable to locate brain regions with fine emotional responses in the prior art.

[0054] (III) This invention can explore the influence of different body postures on emotional state and the underlying neural mechanisms through brain functional imaging, providing a scientific basis for understanding the physiological basis of emotion regulation, while promoting the development of personalized psychological intervention and rehabilitation strategies. Attached Figure Description

[0055] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0056] Figure 2The three body postures illustrated in this embodiment are: (a) a normal sitting posture, with the mind and body relaxed and eyes looking forward; (b) a low-energy posture with the body curled up, shoulders and neck drooping, head down and chest hunched; and (c) a high-energy posture with the head held high, shoulders and neck open, and hands clasped to the chest.

[0057] Figure 3 This is a schematic diagram of the experimental process for the participants in this embodiment;

[0058] Figure 4 The fNIRS channel design array diagram and the related 8 ROI distribution diagrams are shown in this embodiment.

[0059] Figure 5 This is a visual diagram illustrating the brain functional network connections under three different energy postures in this embodiment.

[0060] Figure 6 This is an ANOVA statistical chart of the brain functional network under the three postures of HP, LP, and REST in this embodiment.

[0061] Figure 7 This is a visualization and statistical analysis of brain functional activation based on the integral value (i), centroid value (c), and slope coefficient (s) in this embodiment. (a) Under HP posture, (b) Under LP posture, (c) Brain regions where ANOVA shows statistically significant differences between HP and LP.

[0062] The specific content of the present invention will be further explained in detail below with reference to the embodiments. Detailed Implementation

[0063] It should be noted that, unless otherwise specified, all components in this invention are those known in the art.

[0064] The following are specific embodiments of the present invention. It should be noted that the present invention is not limited to the following specific embodiments. All equivalent modifications made based on the technical solutions of this application fall within the protection scope of the present invention.

[0065] This invention provides a method for detecting the effect of body posture on emotional feedback based on fNIRS, comprising the following steps:

[0066] Step 1: Collect multiple fNIRS light intensity data F ij Each fNIRS light intensity data F ij Includes fNIRS light intensity data across multiple channels;

[0067] The channel is the optical transmission path between an exciter and a detector;

[0068] i represents the participant's posture, with values ​​of 1, 2, and 3, corresponding to the natural posture REST, low-energy posture LP, and high-energy posture HP, respectively.

[0069] j represents the eight regions of interest of the participants, with values ​​of 1, 2, ..., 8, corresponding to the left temporal cortex, right temporal cortex, dorsolateral prefrontal cortex, left dorsolateral prefrontal cortex, right dorsolateral prefrontal cortex, ventromedial prefrontal cortex, left ventrolateral prefrontal cortex, and right ventrolateral prefrontal cortex, respectively.

[0070] Step 2: Remove multiple fNIRS intensity data obtained in Step 1. ij From the inferior data, the optimal M fNIRS light intensity data F were obtained. ij ;

[0071] Step 3, process the M fNIRS intensity data obtained in Step 2. ij Preprocessing was performed to obtain M fNIRS oxygenated hemoglobin concentration data. ij ;

[0072] Step four, analyze the M fNIRS oxygenated hemoglobin concentration data obtained in step three. ij Conduct testing to obtain test data;

[0073] The data includes the R-BFCN (Radio-Brain Functional Connectivity Network) for emotional feedback in different postures. i C-BFCN (Channel Brain Functional Connectivity Network) for Emotional Feedback in Different Postures i The brain functional connectivity network (R-BFCN) of the region of interest during emotional feedback in different postures. i Significantly different ROI connection edges, and the channel brain functional connectivity network (C-BFCN) for emotional feedback under different postures. i The study included channels with significant differences in connectivity, the intensity of neural activity during emotional feedback, the speed of neural activity and the trend of neural responses, as well as channels with significant differences corresponding to different eigenvalues.

[0074] In the above technical solution, participants sequentially assume the postures determined in step one, holding each posture for two minutes. Specifically:

[0075] S1, natural posture for 2 minutes;

[0076] Participants should adopt a normal sitting posture, relax their mind and body, and look straight ahead in a natural position, and maintain this posture for 2 minutes;

[0077] Rest: After data collection is completed, participants are allowed unlimited free movement and stretching.

[0078] S2, low-energy posture for 2 minutes;

[0079] Participants were guided to adopt a low-energy posture with their bodies curled up, shoulders and neck drooping, and heads hunched over, and to hold this posture for 2 minutes;

[0080] Emotional calming: Participants have at least 30 seconds to calm their emotions, during which time they can use free movement, rest, or deep breathing to regulate their emotional state.

[0081] Rest: After calming down, participants are free to move and stretch their bodies for an unlimited amount of time;

[0082] S3, 2 minutes of high-power posture;

[0083] Participants adjusted to a high-energy posture with their head held high, shoulders and neck open, and arms crossed over their chest, and held this posture for 2 minutes.

[0084] In the above technical solutions, rest is intended to ensure the comfort and readiness of participants during the experiment; emotional calming is intended to ensure that participants maintain their baseline emotional state as much as possible before each posture begins.

[0085] By quantifying and statistically analyzing the neural response states and brain functional connectivity patterns of relevant brain regions when posture induces emotions, this method achieves highly sensitive and specific quantitative analysis of emotional states. It overcomes the problems of large subjective differences among raters and insufficient analytical precision in previous methods, and solves the problem that existing technologies cannot objectively quantify the feedback effect of posture-induced emotions.

[0086] Furthermore, by statistically calculating various characteristics of neural responses induced by different postures, we can accurately locate brain regions with different levels and speeds of neural activity during emotional feedback. At the same time, we can monitor the synergistic effect and differentiated activity of multiple brain regions during emotional feedback, thus solving the technical problem of being unable to locate brain regions with fine emotional responses in existing technologies.

[0087] Specifically, step two includes:

[0088] Calculate the multiple fNIRS light intensity data obtained in step one. ij Each fNIRS light intensity data F ij The signal-to-noise ratio of each channel is used to determine the intensity data F of each fNIRS. ij If the number of channels with a signal-to-noise ratio below the first threshold exceeds one-tenth of the total number of channels, then that fNIRS intensity data is discarded to obtain the optimized fNIRS intensity data F. ij ;

[0089] Alternatively, calculate the multiple fNIRS light intensity data F obtained in step one. ijFor each channel of each fNIRS intensity data point, the quotient of the standard deviation and the mean is used to determine whether the number of channels whose quotient exceeds a second threshold exceeds one-tenth of the total number of channels. If so, that fNIRS intensity data point is discarded, resulting in the optimized fNIRS intensity data F. ij .

[0090] Specifically, step three is as follows:

[0091] Step 3.1: A dual filtering mechanism based on a standard deviation threshold and an amplitude threshold, set under a time window, is used to eliminate the M fNIRS light intensity data F obtained in Step 3. ij Interference from motion artifacts in the m-th fNIRS intensity data;

[0092] Step 3.2: Use a bandpass filter to remove baseline drift, artificial noise, physiological noise, and drift noise interference introduced by the instrument's own signal from the m-th fNIRS light intensity data processed in Step 3.1;

[0093] Physiological noise includes heartbeat, respiration, and Mayer waves;

[0094] Step 3.3: According to the modified Beer-Lambert law, the m-th fNIRS intensity data F processed in step 3.2 is processed... ij Convert to fNIRS oxygenated hemoglobin concentration data f ij ;

[0095] Step 3.4: Let m = m + 1, return to step 3.1, and continue until m = M, obtaining M fNIRS oxygenated hemoglobin concentration data. ij ;

[0096] M represents the fNIRS light intensity data. ij Total number of records; m represents fNIRS light intensity data F ij The serial number.

[0097] In the above technical solution, this series of standardized and accurate data preprocessing operations ensures the purity and reliability of experimental data, so as to more accurately reflect the participants' brain activity state.

[0098] Specifically, step four includes the following steps:

[0099] Step 4.1, analyze the M fNIRS oxygenated hemoglobin concentration data obtained in Step 3. ij The Pearson correlation coefficient was calculated for the correlation between regions of interest (ROIs) and channels in the brain to obtain the R-value. Based on the R-value, the brain functional connectivity network (R-BFCN) for emotional feedback under different postures was obtained. iC-BFCN (Channel Brain Functional Connectivity Network) for Emotional Feedback in Different Postures i ;

[0100] Step 4.2: One-way ANOVA was used to statistically analyze the R-values ​​corresponding to the correlation of the region of interest (ROI) and the correlation of the channels obtained in Step 4, respectively. Statistical values ​​were obtained, and the corresponding brain functional connectivity network (R-BFCN) of the ROI during emotional feedback under different postures was obtained based on the statistical values. i Significantly different ROI connection edges and the channel brain functional connectivity network (C-BFCN) for emotional feedback under different postures. i Channel connection edges with significant differences;

[0101] Step 4.3: Extract the M fNIRS oxygenated hemoglobin concentration data obtained in Step 3. ij eigenvalues;

[0102] Eigenvalues ​​include integral values, centroid values, and slope coefficients;

[0103] Step 4.4: Calculate and compare the fNIRS oxygenated hemoglobin concentration data obtained in Step 4.3. 2j and fNIRS oxygenated hemoglobin concentration data f 3j The average integral value, average centroid value, and average slope coefficient of the same channel are used to obtain the intensity of neural activity, the speed of neural activity, and the trend of neural response during emotional feedback.

[0104] Step 4.5: Use one-way ANOVA to analyze each fNIRS oxygenated hemoglobin concentration data obtained in Step 4.3. ij Statistical analysis was performed on the three feature values ​​corresponding to each channel to obtain statistical values. The false positive results in the statistical values ​​were corrected by the false detection rate correction method. The significance level p was set to be less than 0.05 to obtain the channels with significant differences corresponding to different feature values.

[0105] Specifically, step 4.1 includes the following steps:

[0106] Step 4.1.1, analyze the fNIRS oxygenated hemoglobin concentration data obtained in Step 3. ij The correlation between any two regions of interest (ROIs) is calculated using the Pearson correlation coefficient to obtain the R-value. The absolute value of the R-value is then checked; if it is greater than zero, an edge connects the two ROIs; otherwise, no edge connects them, and the R-value is set to 0. The resulting network, consisting of all edges and eight ROIs, forms the R-BFCN (Radio-Oriented Brain Functional Connectivity Network) for emotional feedback under different postures. i ;

[0107] Step 4.1.2: Let i = i + 1, return to step 4.1.1, until i > 3, and output the ROI brain functional connectivity network R-BFCN for emotional feedback under different postures. i ;

[0108] Step 4.1.3, analyze the fNIRS oxygenated hemoglobin concentration data. 3j The correlation between any two channels is calculated using the Pearson correlation coefficient to obtain the R value. The absolute value of the R value is then checked against a first threshold. If the absolute value is greater than a first threshold, a connecting edge exists between the two channels; otherwise, no connecting edge exists. All edges obtained, along with the two channels, form the channel-brain functional connectivity network (C-BFCN) for emotional feedback under different postures. i ;

[0109] Step 4.1.4: Let i = i + 1, return to step 4.1.3, until i > 3, and output the channel brain functional connectivity network (C-BFCN) for emotional feedback under different postures. i .

[0110] In the above technical solution, the R-BFCN (Radio-Oriented Brain Functional Connectivity Network) is used for emotional feedback in different postures. i The R value of each edge in the equation indicates that the larger the absolute value of the R value, the stronger the correlation between the ROIs at both ends of the connected edge, which means that the two ROIs have a stronger synergy when HP induces emotions.

[0111] C-BFCN (Channel Brain Functional Connectivity Network) for Emotional Feedback in Different Postures i The more edges a channel has, the denser the network; conversely, the sparser the network. The number of edges connecting each channel is also calculated; a larger number indicates a more crucial role for that channel in high-energy postural emotional feedback.

[0112] Specifically, step 4.2 includes the following steps:

[0113] Step 4.2.1: One-way ANOVA was used to statistically analyze all R values ​​of any two groups obtained in Step 4.1.1, obtaining statistical values. The FDR correction method was used to correct for false positive results in the statistical values, and ROI connections with a significance level p greater than 0.05 were removed. This yielded ROI connections with significant differences in the ROI brain functional connectivity network during emotional feedback under different postures, specifically the R-BFCN between high-energy posture (HP) and low-energy posture (LP), between high-energy posture (HP) and natural posture (REST), and between low-energy posture (LP) and natural posture (REST). i ROI connection edges with significant differences;

[0114] Step 4.2.2: One-way ANOVA was used to statistically analyze the R-values ​​of each channel in any two groups obtained in Step 4.1.3, obtaining statistical values. The FDR correction method was used to correct for false positives in the statistical values, and channel connections with a significance level greater than 0.05 were removed. This yielded channel connections with significant differences in the channel-brain functional connectivity network during emotional feedback under different postures, specifically the C-BFCN (Channel-Brain Functional Connectivity Network) between high-energy posture (HP) and low-energy posture (LP), between high-energy posture (HP) and natural posture (REST), and between low-energy posture (LP) and natural posture (REST) ​​during emotional feedback under different postures. i Channel connections with significant differences.

[0115] Specifically, step 4.4 includes the following steps:

[0116] Step 4.4.1, respectively, the fNIRS oxygenated hemoglobin concentration data f 2j and fNIRS oxygenated hemoglobin concentration data f 3j The integral values, centroid values, and slope coefficients of the same channel are accumulated and added together to obtain the total integral value, total centroid value, and total slope coefficient of each channel;

[0117] Step 4.4.2: The total integral value, total centroid value and total slope coefficient of each channel obtained in step 4.4.1 are averaged to obtain the average integral value, average centroid value and average slope coefficient of each channel.

[0118] Step 4.4.3, compare the fNIRS oxygenated hemoglobin concentration data obtained in step 4.4.2 with the fNIRS data. 2j and fNIRS oxygenated hemoglobin concentration data f 3j The average integral value, average centroid value, and average slope coefficient of the same channel;

[0119] The average score is used to evaluate the intensity of neural activity during emotional feedback;

[0120] The average center of gravity value is used to evaluate the speed of neural activity during emotional feedback.

[0121] The average slope coefficient is used to evaluate the trend of neural responses to emotional feedback.

[0122] In the above technical solution, the average integral value, average center of gravity value, and average slope coefficient are used to explore the state of brain functional activity under different postures and reveal the relationship between posture and brain function. The average integral value is used to reflect the correlation between neural activity and posture-induced emotional feedback; the average center of gravity value is used to reflect the main activation space of activated brain regions; and the average slope coefficient is used to understand the changing trend of blood oxygenation response speed caused by neural activity.

[0123] The larger the average integral value, the stronger the emotional feedback neural activity induced by the HP group in that channel; the smaller the average centroid value, the more important that the channel is the main area of ​​neural response during emotional feedback; and the larger the average slope coefficient, the greater and faster the neural response in that channel.

[0124] Example:

[0125] This embodiment presents a method for detecting the effect of body posture on emotional feedback based on fNIRS. A total of 52 undergraduate and graduate students were recruited as participants, including 29 males and 23 females, aged between 20 and 27 years (M = 24.44 years, SD = 2.63 years). All participants were in good health, had normal limb motor abilities, were right-handed, and of moderate build. None of the participants had motor function, mobility or muscle impairments, or brain trauma. All participants participated in the experiment in accordance with the Declaration of Helsinki, and this study was approved by the Academic Committee of Shaanxi Normal University (NO: 202409023).

[0126] Specifically:

[0127] Before participating in the experiment, all participants signed a written informed consent form to ensure their rights were fully protected. The researchers showed all participants a picture of the posture they were to assume, such as... Figure 2 As shown, (a) a normal sitting posture, relaxed body and mind, and looking straight ahead, (b) a low-energy posture with the body curled up, shoulders and neck drooping, and head hunched, and (c) a high-energy posture with the head held high, shoulders and neck open, and arms crossed over the chest. The purpose of the experiment and the detailed experimental procedures were explained to mitigate these expected effects. Two practice experiments were then conducted to ensure that each participant was familiar with the experimental procedures.

[0128] Step 1: Collect multiple fNIRS light intensity data F ij Each fNIRS light intensity data F ij Includes fNIRS light intensity data across multiple channels;

[0129] In a laboratory setting, participants sequentially assumed three postures, each held for two minutes. Simultaneously, near-infrared brain imaging (NIRS) was used to acquire fNIRS signals from eight regions of interest, yielding multiple fNIRS intensity data points. ij ;like Figure 3As shown, participants were first asked to assume a normal sitting posture, with their mind and body relaxed and looking straight ahead, for two minutes as the baseline condition for collecting fNIRS signals. After data collection, participants were allowed unlimited free movement and stretching. This protocol was designed to ensure participant comfort and readiness during the experiment. Next, participants were guided to adopt a low-energy posture with their body curled up, shoulders and neck lowered, and head hunched, and held this posture for two minutes while collecting fNIRS signals. After the low-energy posture phase, participants had at least 30 seconds to calm down. During this time, participants engaged in activities such as free movement, rest, or deep breathing to regulate their emotional state. This protocol was designed to ensure that participants maintained their baseline emotional state as much as possible before each posture began. After calming down, participants were allowed unlimited free movement and stretching. This protocol was designed to ensure participant comfort and readiness during the experiment. Finally, participants adjusted to a high-energy posture with their head held high, shoulders and neck open, and arms crossed, and held this posture for two minutes to collect fNIRS signals. According to the Broadman brain structural partitioning system, the prefrontal cortex (PFC) and temporal lobe (TC) brain regions associated with emotion are divided into eight regions of interest (ROIs), such as... Figure 4 As shown, this experiment used the NirSmart-6000A near-infrared brain functional imaging device (Danyang Huichuang Medical Equipment Co., Ltd., Jiangsu, China). (a) 15 exciters and 16 detectors cover PFC and TC, forming 48 channels. (b) Distribution of 8 regions of interest (ROIs) in PFC and TC: TC-R (green), DLPFC-R (orange), VLPFC-R (cinnamon), DMPFC (purple), VMPFC (red), DLPFC-L (orange), VLPFC-L (cinnamon), TC-L (green). (c) Localization of ROIs and channels in 3D brain map.

[0130] Step 2: Remove multiple fNIRS intensity data obtained in Step 1. ij From the inferior data, the optimal M fNIRS light intensity data F were obtained. ij ;

[0131] The signal-to-noise ratio (SNR) of each channel was calculated. If the number of channels with a value 30 dB below a set threshold exceeded one-tenth of the total number of channels, that data point was removed. Alternatively, the percentage of the standard deviation to the mean of the fNIRS signal for each channel was calculated. If the number of channels with a result greater than 5 exceeded one-tenth of the total number of channels, that data point was removed. After careful review, the fNIRS intensity data of 5 participants were excluded because they did not meet the analysis criteria. Therefore, the fNIRS intensity data of 26 males and 21 females were ultimately retained. These data were of excellent quality and could be used for subsequent statistical analysis to ensure the accuracy and reliability of the experimental results.

[0132] Step 3: Process the M fNIRS intensity data obtained in Step 2. ij Preprocessing was performed to obtain M fNIRS oxygenated hemoglobin concentration data. ij ;

[0133] First, a dual filtering mechanism based on a 0.3s time window, using a standard deviation threshold of 6 and an amplitude threshold of 0.5, is employed to eliminate motion artifacts in the fNIRS intensity data obtained in step three. Then, a 0.01-0.1Hz bandpass filter is used to remove baseline drift, artificial noise, physiological noise, and drift noise introduced by the instrument's own signals from the fNIRS intensity data processed in step 4.1. Physiological noise includes heartbeat, respiration, and Mayer waves. Finally, according to the modified Beer-Lambert law, the differential path factor (DPF) is set to 6 to convert the fNIRS intensity data into relative concentration changes of fNIRS hemoglobin. Since the HbO2 concentration parameter has a higher amplitude value and higher SNR than the relative HbR concentration parameter, and HbO2 concentration is relatively sensitive during task response, it is more likely to reflect objective experiments. Therefore, this experiment mainly analyzes the changes in HbO2 concentration throughout the experimental process. Through this series of standardized and accurate data preprocessing operations, fNIRS oxygenated hemoglobin concentration data f0 is obtained. ij This ensured the purity and reliability of the experimental data. Through this series of standardized and accurate data preprocessing operations, the purity and reliability of the experimental data were guaranteed, allowing for a more accurate reflection of the participants' brain activity.

[0134] Step 4: Analyze the M fNIRS oxygenated hemoglobin concentration data obtained in Step 3. ij Conduct testing to obtain test data;

[0135] The data includes the R-BFCN (Radio-Brain Functional Connectivity Network) for emotional feedback in different postures. i C-BFCN (Channel Brain Functional Connectivity Network) for Emotional Feedback in Different Postures iThe brain functional connectivity network (R-BFCN) of the region of interest during emotional feedback in different postures. i Significantly different ROI connection edges, and the channel brain functional connectivity network (C-BFCN) for emotional feedback under different postures. i The study included channels with significant differences in connectivity, the intensity of neural activity during emotional feedback, the speed of neural activity and the trend of neural responses, as well as channels with significant differences corresponding to different eigenvalues.

[0136] Step four specifically includes:

[0137] The M fNIRS oxygenated hemoglobin concentration data obtained in step three ij The Pearson correlation coefficient was calculated for the correlation between regions of interest (ROIs) and channels in the brain to obtain the R-value. Based on the R-value, the brain functional connectivity network (R-BFCN) for emotional feedback under different postures was obtained. i C-BFCN (Channel Brain Functional Connectivity Network) for Emotional Feedback in Different Postures i ;like Figure 5 This study illustrates the connectivity of brain functional networks under three energy postures: Pearson correlation coefficient connection mapping ROI-ROI Network, Channel Network, and Fully Connected (FC) matrix. Colored ellipses represent key brain regions in the functional network, purple dots are defined as key nodes (defined as nodes connected by more than 15 edges in the functional network), and red squares indicate functional connections between channels.

[0138] In the HP posture, connections between different ROIs are very dense, especially in the medial prefrontal cortex and DLPFC. As shown in Figure (b3), specifically, 56 edges form a complex brain functional network structure, indicating strong functional synergy between brain regions in the HP posture. As shown in Figure (b2), channels CH11, CH27, CH28, CH30, CH43, and CH44 in the medial prefrontal cortex, and channels CH24, CH32, CH41, and CH45 in the DLPFC region, each have more than 15 connections, serving as key nodes in the brain functional network, with an average brain functional network connectivity similarity of 0.7362. In contrast, brain functional connectivity in the LP posture appears particularly sparse, mainly limited to connections between VLPFC and TC-L, with only 7 edges. The average brain functional network connectivity similarity is 0.72297, slightly lower than in the HP posture, but still indicating a certain degree of functional synchronicity. Notably, in the LP posture, no single channel has more than 5 connections, suggesting relatively weak functional connectivity between brain regions. Finally, the brain functional network connectivity in the Rest state also exhibited high density, with connections between all brain regions, particularly a tight connection in the PFC (Peripherally Flexible Central Threat) with 63 edges. Compared to the low-energy pose, this formed a more complex network structure, with an average brain functional network connectivity similarity of 0.73674, slightly higher than the HP pose, indicating stronger coordination between brain regions in the natural state, i.e., a more active default mode network. However, the Rest state brain functional network had only 7 key nodes, namely channels CH8, CH11, CH26, CH27, CH28, CH42, and CH43, fewer than the number of key nodes observed in the HP pose.

[0139] One-way ANOVA was used to statistically analyze the R-values ​​corresponding to the correlation of regions of interest (ROIs) and channels obtained in step four, respectively. Statistical values ​​were obtained, and the corresponding brain functional connectivity network (R-BFCN) for emotional feedback under different postures was derived from these statistical values. i Significantly different ROI connection edges and the channel brain functional connectivity network (C-BFCN) for emotional feedback under different postures. i Channel connection edges with significant differences. Figure 6Differences in brain functional network connectivity between different energy postures are shown. Colored ellipses indicate statistically significant ROIs, blue lines indicate a significance level less than 0.05, and red lines indicate a significance level less than 0.01. Statistical significance was defined as p < .05. Differences in brain functional network connectivity between HP and LP postures were mainly concentrated in the right prefrontal cortex and DLPFC (F(1,92)>3.67, p < .05). These brain regions may show increased activity and greater involvement in cognitive and emotional processing when an individual is in an HP state, while their activity may be less correlated in an LP state. Differences in brain functional network connectivity between HP and Res postures were relatively small, with the most significant differences only observed in the connectivity differences between channels CH17 and CH40 (F(1,92)=5.18, p=.0036<.05). Differences in brain functional network connectivity between LP and Rest postures were mainly observed in the medial prefrontal cortex (F(1,92)>5.94, p < .05), with a total of 45 significantly different functional connectivity edges.

[0140] Extracting fNIRS oxygenated hemoglobin concentration data f 2j and fNIRS oxygenated hemoglobin concentration data f 3j The average integral value, average centroid value, and average slope coefficient of the same channel were used to obtain the intensity, speed, and trend of neural activity during emotional feedback. The average integral value describes the total change in hemodynamic response during the 120-second activation task, serving as an intensity indicator representing the sum of signal changes throughout the task; a higher value indicates greater activity related to the cognitive task. The average centroid value quantifies the "center" position of the hemoglobin concentration change curve, reflecting not only the main activation space of the activated brain region but also the average time of blood oxygen concentration change; a smaller centroid value indicates a faster cortical response and relaxation after task completion. The average slope coefficient indicates the amount (and direction) of change in oxygenation response during stimulation for each task, revealing changes in the speed of blood oxygenation response caused by neural activity. For example, a higher (positive) slope value for HbO2 indicates greater and faster cortical activation. The extraction of these features aims to explore brain functional activity under different postures, thereby revealing the connection between posture and brain function.

[0141] Figure 7The changes in brain functional activity of HP and LP across different features c, i, and s are shown, with white arrows pointing to channels showing significant differences between the two groups (p<.05). c<60 indicates more rapid cortical responses and relaxation during the task. HP primarily affects the medial prefrontal cortex, while LP primarily affects the temporal lobe. Specifically, HP is associated with increased blood flow to the DMPFC and VMPFC (i>2), indicating enhanced neural activity in brain regions associated with emotion. Conversely, LP shows increased blood flow to the TC (i>4) and decreased blood flow to the DMPFC (i<0). For feature s, HP shows positive values ​​in the TC and lateral prefrontal cortex (s>0), indicating rapid cortical activation during the experiment. In contrast, LP shows positive values ​​in the medial prefrontal cortex and the DLPFC-L region (s>0), highlighting different cortical activation patterns associated with different postural states.

[0142] Further analysis of variance on brain functional activity between HP and LP revealed significant differences in cortical activation of specific brain regions and channels. Channels CH40 (Fc(1,92)=14.882, p=.010<.05) and CH44 (Fc(1,92)=10.457, p=.027) showed significant differences in activation intensity in the DMPFC region, and channel CH13 (Fc(1,92)=12.142, p=.018) showed significant differences in activation intensity in the TC-L brain region. Furthermore, significant differences also existed in activation intensity in the TC (Fi(1,92)>7.752, p<.028) and DLPFC-L (Fi(1,92)>7.099, p<.036) regions. Finally, there were also significant differences in activation rates between TC-L (Fs(1,92)>8.923, p<.025) and the right medial prefrontal cortex (Fs(1,92)>6.691, p<.045).

Claims

1. A method for detecting the effect of body posture on emotional feedback based on fNIRS, characterized in that, Includes the following steps: Step 1: Collect multiple fNIRS light intensity data. Each fNIRS light intensity data Includes fNIRS light intensity data across multiple channels; The channel is an optical transmission path between an exciter and a detector; This represents the participant's posture, with values ​​of 1, 2, and 3, corresponding to the natural posture (REST), low-energy posture (LP), and high-energy posture (HP). The eight regions of interest of the participants are represented by values ​​1, 2, ..., 8, which correspond to the left temporal cortex, right temporal cortex, dorsolateral prefrontal cortex, left dorsolateral prefrontal cortex, right dorsolateral prefrontal cortex, ventromedial prefrontal cortex, left ventrolateral prefrontal cortex, and right ventrolateral prefrontal cortex, respectively. The Step 2: Remove multiple fNIRS intensity data obtained in Step 1. From the low-quality data, we obtained the optimized M fNIRS light intensity data. ; Step 3: Process the M fNIRS intensity data obtained in Step 2. Preprocessing was performed to obtain M fNIRS oxygenated hemoglobin concentration data. ; Step 4: Analyze the M fNIRS oxygenated hemoglobin concentration data obtained in Step 3. Conduct testing to obtain test data; The detection data includes the brain functional connectivity network of the region of interest (ROI) for emotional feedback under different postures. Brain functional connectivity networks for emotional feedback under different postures ROI brain functional connectivity network during emotional feedback in different postures Significantly different ROI connection edges and channel brain functional connectivity networks during emotional feedback under different postures. The channels with significant differences, the intensity of neural activity during emotional feedback, the speed of neural activity and the trend of neural response, and the channels with significant differences corresponding to different feature values; Step four is as follows: Step 4.1, analyze the M fNIRS oxygenated hemoglobin concentration data obtained in Step 3. The Pearson correlation coefficient was calculated to obtain the R-value by analyzing the correlation between the regions of interest (ROIs) and the correlation between channels in the brain. Based on the R-value, the functional connectivity networks of the ROIs for emotional feedback under different postures were obtained. and the neural pathway connectivity network for emotional feedback in different postures ; Step 4.2: One-way ANOVA was used to statistically analyze the R-values ​​corresponding to the correlation of the regions of interest (ROIs) and the correlation of the channels obtained in Step 4, respectively. Statistical values ​​were obtained, and the corresponding brain functional connectivity networks of the ROIs during emotional feedback under different postures were obtained based on the statistical values. Significantly different ROI connectivity edges and channel brain functional connectivity networks during emotional feedback under different postures Channel connection edges with significant differences; Step 4.3: Extract the M fNIRS oxygenated hemoglobin concentration data obtained in Step 3. eigenvalues; The eigenvalues ​​include integral values, centroid values, and slope coefficients; Step 4.4: Calculate and compare the fNIRS oxygenated hemoglobin concentration data obtained in Step 4.

3. and fNIRS oxygenated hemoglobin concentration data The average integral value, average centroid value, and average slope coefficient of the same channel are used to obtain the intensity of neural activity, the speed of neural activity, and the trend of neural response during emotional feedback. Step 4.5: Use one-way ANOVA to analyze each fNIRS oxygenated hemoglobin concentration data obtained in Step 4.

3. Statistical analysis was performed on the three feature values ​​corresponding to each channel to obtain statistical values. The false positive results in the statistical values ​​were corrected by the false detection rate correction method. The significance level p was set to be less than 0.05 to obtain the channels with significant differences corresponding to different feature values.

2. The method for detecting the effect of body posture on emotional feedback based on fNIRS as described in claim 1, characterized in that, Step two specifically includes: Calculate the multiple fNIRS light intensity data obtained in step one. Each fNIRS light intensity data The signal-to-noise ratio of each channel is used to determine the intensity data of each fNIRS. If the number of channels with a signal-to-noise ratio below the first threshold exceeds one-tenth of the total number of channels, then that fNIRS intensity data is discarded to obtain the optimized fNIRS intensity data. ; Alternatively, calculate the multiple fNIRS intensity data obtained in step one. For each channel of each fNIRS intensity data point, the quotient of the standard deviation and the mean is used to determine whether the number of channels whose quotient exceeds a second threshold exceeds one-tenth of the total number of channels. If so, that fNIRS intensity data point is discarded to obtain the optimized fNIRS intensity data. .

3. The method for detecting the effect of body posture on emotional feedback based on fNIRS as described in claim 1, characterized in that, Step three specifically involves: Step 3.1: A dual filtering mechanism based on a standard deviation threshold and an amplitude threshold, set under a time window, is used to eliminate the M fNIRS light intensity data obtained in Step 3. Interference from motion artifacts in the m-th fNIRS intensity data; Step 3.2: Use a bandpass filter to remove baseline drift, artificial noise, physiological noise, and drift noise interference introduced by the instrument's own signal from the m-th fNIRS light intensity data processed in Step 3.1; The physiological noises include heartbeat, respiration, and Mayer waves; Step 3.3: According to the modified Beer-Lambert law, the m-th fNIRS intensity data processed in step 3.2 is... Converted to fNIRS oxygenated hemoglobin concentration data ; Step 3.4: Let m = m + 1, return to step 3.1, and continue until m = M, obtaining M fNIRS oxygenated hemoglobin concentration data. ; M represents fNIRS light intensity data. Total number of records; m represents fNIRS light intensity data. The serial number.

4. The method for detecting the effect of body posture on emotional feedback based on fNIRS as described in claim 1, characterized in that, Step 4.1 specifically includes the following steps: Step 4.1.1: Analyze the fNIRS oxygenated hemoglobin concentration data obtained in Step 3. The correlation between any two regions of interest (ROIs) is calculated using the Pearson correlation coefficient to obtain the R-value. The absolute value of the R-value is then checked; if it is greater than zero, an edge connects the two ROIs; otherwise, no edge connects them, and the R-value is set to 0. All edges obtained, along with the eight ROIs, form a functional connectivity network for emotional feedback under different postures. ; Step 4.1.2: Let i = i + 1, return to step 4.1.1, until i > 3, and output the ROI brain functional connectivity network for emotional feedback under different postures. ; Step 4.1.3, analyze fNIRS oxygenated hemoglobin concentration data. The correlation between any two channels is calculated using the Pearson correlation coefficient to obtain the R value. The absolute value of the R value is then checked against a first threshold. If the absolute value is greater than a first threshold, a connecting edge exists between the two channels; otherwise, no connecting edge exists. All the edges obtained, along with the two channels themselves, form a functional neural connectivity network for emotional feedback under different postures. ; Step 4.1.4: Let i = i + 1, return to step 4.1.3, until i > 3, and output the channel brain functional connectivity network of emotional feedback under different postures. .

5. The method for detecting the effect of body posture on emotional feedback based on fNIRS as described in claim 1, characterized in that, Step 4.2 specifically includes the following steps: Step 4.2.1: One-way ANOVA was used to statistically analyze all R-values ​​of any two groups obtained in Step 4.1.1, obtaining statistical values. The FDR correction method was used to correct for false positive results in the statistical values, and ROI connections with a significance level p greater than 0.05 were removed. This yielded ROI connections with significant differences in the ROI brain functional connectivity network during emotional feedback under different postures. Specifically, the ROI brain functional connectivity network during emotional feedback under different postures was obtained between high-energy posture (HP) and low-energy posture (LP), between high-energy posture (HP) and natural posture (REST), and between low-energy posture (LP) and natural posture (REST). ROI connection edges with significant differences; Step 4.2.2: One-way ANOVA was used to statistically analyze the R-values ​​of each channel in any two groups obtained in Step 4.1.3, obtaining statistical values. The FDR correction method was then used to correct for false positives in the statistical values, and channel connections with a significance level greater than 0.05 were removed. This yielded channel connections with significant differences in the channel-brain functional connectivity network during emotional feedback under different postures. Specifically, the channel-brain functional connectivity network during emotional feedback under different postures was obtained between high-energy posture (HP) and low-energy posture (LP), between high-energy posture (HP) and natural posture (REST), and between low-energy posture (LP) and natural posture (REST). Channel connections with significant differences.

6. The method for detecting the effect of body posture on emotional feedback based on fNIRS as described in claim 1, characterized in that, Step 4.4 specifically includes the following steps: Step 4.4.1: The fNIRS oxygenated hemoglobin concentration data are then processed. and fNIRS oxygenated hemoglobin concentration data The integral values, centroid values, and slope coefficients of the same channel are accumulated and added together to obtain the total integral value, total centroid value, and total slope coefficient of each channel; Step 4.4.2: The total integral value, total centroid value and total slope coefficient of each channel obtained in step 4.4.1 are averaged to obtain the average integral value, average centroid value and average slope coefficient of each channel. Step 4.4.3: Compare the fNIRS oxygenated hemoglobin concentration data obtained in Step 4.4.

2. and fNIRS oxygenated hemoglobin concentration data The average integral value, average centroid value, and average slope coefficient of the same channel; The average integral value is used to evaluate the intensity of neural activity during emotional feedback. The average center of gravity value is used to evaluate the speed of neural activity during emotional feedback. The average slope coefficient is used to evaluate the trend of neural responses to emotional feedback.

Images