Method and device for calculating coupled physiological indexes based on synchronous respiration, continuous blood pressure and electroencephalogram signals and cognitive function evaluation method based on coupled physiological indexes
By simultaneously processing respiratory and EEG signals, identifying signal feature points, and constructing systolic pressure evoked potentials for inspiratory and expiratory phases, the objectivity and ecological applicability issues of existing cognitive assessment methods are resolved, enabling multi-dimensional and accurate assessment of cognitive function.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-25
- Publication Date
- 2026-03-27
AI Technical Summary
Existing cognitive function assessment methods suffer from poor objectivity, weak mechanistic correlation, and low ecological applicability. They are unable to integrate autonomic and central nervous system signals and cannot fully reflect the dynamic coupling relationship between respiration, blood pressure, and the brain.
By simultaneously processing respiratory, continuous blood pressure, and electroencephalogram (EEG) signals, identifying signal feature points, constructing systolic pressure evoked potentials for inspiratory and expiratory phases, and combining independent component analysis and sample entropy quantification, coupled physiological indicators are obtained to achieve an objective assessment of cognitive function.
It provides a completely objective, multi-dimensional cognitive function assessment method that can reflect the coupling relationship between respiratory phase and blood pressure fluctuations and brain activity, breaking through the limitations of single signal assessment and achieving accurate quantification of cognitive function.
Smart Images
Figure CN121730779A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the processing of physiological signals and the assessment of cognitive functions, and more particularly to a method, apparatus, and cognitive function assessment method based on a coupled physiological index of synchronized respiration, continuous blood pressure, and electroencephalogram (EEG) signals. Background Technology
[0002] Cognitive function is a core human ability to perform daily activities, learn, work, and maintain social interactions, involving multiple dimensions such as attention, memory, language comprehension, and executive function. Cognitive function assessment is a crucial technical support for neuroscience research, early diagnosis of neurodegenerative diseases (such as Alzheimer's disease and dementia), and verification of the effectiveness of cognitive interventions. Currently, mainstream assessment methods mainly include neuropsychological scales, neuroimaging assessments, neurophysiological assessments (such as traditional EEG / event-related potentials), and behavioral experimental assessments. However, these methods all have certain technical limitations and cannot fully meet the needs for objective, accurate, and ecological cognitive assessment.
[0003] Neuropsychological scales, such as the Mini-Mental State Examination (MMSE) and the Montreal Cognitive Assessment (MoCA), are currently the most commonly used tools for assessing cognitive function in clinical practice. However, they rely on subjective answers from test takers and manual scoring by examiners, making them susceptible to interference from emotions, education level, and cultural background, resulting in poor objectivity of assessment results. They are also prone to the "practice effect" and are difficult to detect early, mild cognitive impairments. Furthermore, the tasks on these scales are detached from real-life scenarios, resulting in extremely low ecological validity and making it difficult to reflect the test takers' daily cognitive abilities.
[0004] While cognitive function assessment techniques based on neuroimaging methods, such as functional magnetic resonance imaging (fMRI) and positron emission tomography (PET), can link brain structure / function with cognition, they are expensive, complex to operate, and require subjects to remain still in a closed, noisy environment for 30-60 minutes, making them unsuitable for children, the elderly, claustrophobic patients, or people with movement disorders. Furthermore, fMRI has low temporal resolution (1-2 seconds), making it unable to capture millisecond-level dynamic changes in cognitive processes, while PET relies on radioactive tracers, posing radiation risks and making it difficult to use for high-frequency longitudinal assessments.
[0005] Traditional neurophysiological cognitive function assessment techniques, such as event-related potentials (EG), can provide objective neurophysiological indicators through external auditory or visual stimulation. However, their induction depends on the subject's level of attention and cooperation during the task; for patients with attention deficits or difficulty cooperating, the reliability of the results decreases significantly. Furthermore, it primarily reflects the cognitive processing of external stimuli rather than the brain's intrinsic spontaneous activity. While this method offers high temporal resolution, it focuses only on the activity of a single brain system, severing the regulatory role of the autonomic nervous system (respiration, circulation) in cognition; it also lacks consideration of key regulatory variables such as respiratory phasing, failing to reflect the dynamic coupling relationship between respiration, blood pressure, and the brain, resulting in a limited assessment dimension.
[0006] Heartbeat evoked potentials (HEPs) are an important indicator that has emerged in recent years for assessing the intrinsic perception of brain-heart interaction. It examines the brain's perceptual processing of its own heartbeat by using phase-locked averaging of EEG signals with the ECG R wave as an anchor point. Studies have shown that the amplitude of HEPs is related to interoceptive awareness and exhibits abnormalities in certain cognitive impairments. However, HEPs primarily reflect the interaction between the heart and brain, failing to incorporate respiration, a physiological process that has a strong periodic regulatory effect on brain state and the cardiovascular system. Therefore, its assessment of cognitive function is relatively limited.
[0007] In summary, existing cognitive assessment methods generally suffer from poor objectivity, weak mechanism correlation, low ecological applicability, and lack of multi-system collaborative consideration. There is an urgent need for a new assessment method that can integrate autonomic and central nervous system signals, extract coupling features according to physiological phases, and achieve accurate quantification of cognitive function. Summary of the Invention
[0008] In view of the above problems, this application aims to propose a method and device for calculating coupled physiological indicators based on synchronized respiration, continuous blood pressure, and electroencephalogram (EEG) signals, thereby constructing coupled physiological indicators of respiratory-cardiovascular-brain function. This application also aims to propose a cognitive function assessment method based on these coupled physiological indicators.
[0009] The method for calculating coupled physiological indicators based on synchronized respiration, continuous blood pressure, and electroencephalogram (EEG) signals, as described in this application, includes:
[0010] Synchronous physiological signal preprocessing steps: Respiratory, continuous blood pressure, and electroencephalogram (EEG) signals are preprocessed separately to obtain clean respiratory, continuous blood pressure, and EEG signals;
[0011] Continuous blood pressure signal feature point detection and processing steps; identifying the peak point of systolic blood pressure in each cycle of a clean continuous blood pressure signal; each peak point is a SBP event;
[0012] Respiratory signal feature point detection and processing steps: Identify the peak and trough values of clean respiratory signals using a sliding window method to obtain respiratory peak points and respiratory trough points; Sort the respiratory trough points and respiratory peak points by time, and define the period from the current respiratory trough point to the next respiratory peak point as the inspiratory phase and the period from the current respiratory peak point to the next respiratory trough point as the expiratory phase;
[0013] The steps for classifying respiratory phases of SBP events are as follows: Traverse all SBP events and determine whether each SBP event belongs to the inspiratory or expiratory phase.
[0014] The steps for constructing inspiratory systolic blood pressure evoked potentials (SBPs) are as follows: For the inspiratory phase, using the SBP event corresponding to inspiration as an anchor point and each anchor point as a reference point, clean EEG signals of a predetermined duration are extracted to form EEG signal segments. All EEG signal segments are aligned and averaged, and baseline correction is performed to obtain the inspiratory systolic blood pressure evoked potentials. For the expiratory phase, using the SBP event corresponding to expiration as a reference point, clean EEG signals of a predetermined duration are extracted to form EEG signal segments. All EEG signal segments are aligned and averaged, and baseline correction is performed to obtain the expiratory systolic blood pressure evoked potentials.
[0015] The coupled physiological indicators consist of inspiratory systolic pressure evoked potentials and / or expiratory systolic pressure evoked potentials.
[0016] Preferably, the predetermined duration is 100ms; the predetermined duration is 250-350ms after the reference point.
[0017] Preferably, in the respiratory signal feature point detection and processing steps, after identifying the peak and trough values of the clean respiratory signal through the sliding window method, noise interference is reduced through a multi-window verification mechanism, and false extreme values are filtered in combination with the physiological characteristics of the respiratory signal, and finally the respiratory peak point and respiratory trough point are obtained.
[0018] Preferably, in the synchronous physiological signal preprocessing step,
[0019] For the respiratory signal, a Butterworth bandpass filter with a cutoff frequency of 0.05-1.0Hz was used to remove low-frequency motion artifacts and high-frequency ECG interference, and a 50 Hz notch filter was used to suppress power line interference to obtain a clean respiratory signal.
[0020] For continuous blood pressure signals, baseline drift is removed by a Butterworth high-pass filter with a cutoff frequency of 0.5 Hz, motion artifacts are reduced by a Butterworth low-pass filter with a cutoff frequency of 5 Hz, and finally power frequency interference is eliminated by a 50 Hz notch filter to obtain a clean continuous blood pressure signal.
[0021] For EEG signals, a single-channel EEG signal is processed using a 50Hz notch filter to eliminate power frequency interference, followed by filtering with a 0.05-100Hz bandpass filter to remove non-physiological artifacts. Then, the EEG signal is subjected to wavelet transform using Daubechiesdb4 discrete wavelets to obtain approximate and detail components. These components are then processed using complete empirical mode decomposition with adaptive noise to calculate intrinsic mode functions (EMFs). Next, independent component analysis (ICA) is used to calculate the independent components of the EMFs, and sample entropy quantization is performed. Independent components whose sample entropy values satisfy the Gomez-Herrero condition are considered artifacts and set to zero. Finally, inverse ICA is performed to reconstruct new approximate and detail components, resulting in a clean EEG signal.
[0022] The cognitive function assessment method based on coupled physiological indicators of this application is obtained by the above calculation method; when the coupled physiological indicators are greater than their corresponding thresholds, the subject is considered to have cognitive function impairment.
[0023] The computing device of this application, based on the coupled physiological indicators of synchronized respiration, continuous blood pressure, and electroencephalogram (EEG) signals, includes:
[0024] The synchronous physiological signal preprocessing unit is used to preprocess respiration, continuous blood pressure, and electroencephalogram (EEG) signals respectively to obtain clean respiration, continuous blood pressure, and EEG signals.
[0025] The continuous blood pressure signal feature point detection and processing unit is used to identify the peak point of systolic blood pressure in each cycle of a clean continuous blood pressure signal, with each peak point being a SBP event.
[0026] The respiratory signal feature point detection and processing unit identifies the peak and trough values of a clean respiratory signal through a sliding window method, obtains the respiratory peak point and respiratory trough point, sorts the respiratory trough point and respiratory peak point by time, and defines the period from the current respiratory trough point to the next respiratory peak point as the inspiratory phase and the period from the current respiratory peak point to the next respiratory trough point as the expiratory phase.
[0027] The respiratory phase classification unit of SBP events traverses SBP events and determines whether each SBP event belongs to the inspiratory phase or the expiratory phase.
[0028] The inspiratory systolic blood pressure evoked potential and the inspiratory systolic blood pressure evoked potential construction unit, for the inspiratory phase, use the SBP event corresponding to inspiration as the reference point, extract clean EEG signals of a predetermined duration to form EEG signal segments, align and average all EEG signal segments, and perform baseline correction to obtain the inspiratory systolic blood pressure evoked potential; for the expiratory phase, use the SBP event corresponding to expiration as the reference point, extract clean EEG signals of a predetermined duration to form EEG signal segments, align and average all EEG signal segments, and perform baseline correction to obtain the expiratory systolic blood pressure evoked potential.
[0029] The coupled physiological indicators consist of inspiratory systolic pressure evoked potentials and / or expiratory systolic pressure evoked potentials.
[0030] This application presents a method for acquiring coupled physiological indicators based on synchronized respiration, continuous blood pressure, and electroencephalogram (EEG) signals. This method provides a completely objective cognitive function assessment indicator that is entirely independent of the subject's subjective cooperation and task execution. It breaks through the limitations of traditional single-signal or dual-signal analysis. By simultaneously integrating three core physiological signals—RSP, CBP, and EEG—it constructs a more comprehensive and multidimensional respiratory-cardiovascular-brain function coupled assessment system. This reveals the specific modulation effect of respiratory phase (inhalation / exhalation) on the coupling relationship between blood pressure fluctuations and cerebral cortex activity. This modulation effect is then used to extract physiological indicators that are highly sensitive to changes in cognitive function, thereby achieving an objective and accurate assessment of cognitive function. Attached Figure Description
[0031] Figure 1 This is a flowchart illustrating the method for obtaining physiological indicators based on the coupling of synchronized respiration, continuous blood pressure, and electroencephalogram (EEG) signals according to this application.
[0032] Figure 2 This is a schematic diagram of the original RSP, CBP, and EEG signals in the example.
[0033] Figure 3 This is a schematic diagram of the preprocessed RSP, CBP, and EEG signals in the example.
[0034] Figure 4 This is a schematic diagram of the RSP and CBP signal feature point detection results in the example.
[0035] Figure 5 The getter phase marked in the example and exhalation phase A schematic diagram.
[0036] Figure 6 For example obtained and A schematic diagram.
[0037] Figure 7For example obtained , A schematic diagram.
[0038] Figure 8 For example obtained , A diagram illustrating Spearman correlation analysis of the indicators.
[0039] Figure 9 To utilize , The indicators were used to plot the ROC curves for the plains and plateau populations. Detailed Implementation
[0040] The present application will now be described in detail with reference to the accompanying drawings.
[0041] like Figure 1 The diagram shown is a flowchart illustrating the method for processing synchronized breathing, continuous blood pressure, and electroencephalogram signals according to this application.
[0042] Step 1: Synchronize physiological signal preprocessing;
[0043] The respiratory RSP, continuous blood pressure CBP, and electroencephalogram (EEG) signals acquired simultaneously were processed as follows:
[0044] RSP signal: A Butterworth bandpass filter with a cutoff frequency of (0.05-1.0) Hz is used to remove low-frequency motion artifacts and high-frequency ECG interference, and a 50 Hz notch filter is used to suppress power frequency interference to obtain a clean RSP signal.
[0045] CBP signal: Baseline drift is removed by a Butterworth high-pass filter with a cutoff frequency of 0.5Hz, motion artifacts are reduced by a Butterworth low-pass filter with a cutoff frequency of 5Hz, and finally power frequency interference is eliminated by a 50Hz notch filter to obtain a clean CBP signal.
[0046] EEG Signal: For single-channel EEG signals, a 50Hz notch filter was used to eliminate power frequency interference. EEG data was filtered through a (0.05-100)Hz bandpass filter to remove non-physiological artifacts. Then, a Daubechies db4 discrete wavelet transform was performed on the EEG to obtain approximate and detail components. These components were then processed using complete empirical mode decomposition with adaptive noise to calculate intrinsic mode functions (EMFs). Subsequently, independent component analysis (ICA) was used to calculate the independent components of the EMFs, followed by sample entropy quantization. Independent components whose sample entropy values satisfy the Gomez-Herrero condition were considered artifacts and set to zero. Finally, inverse ICA was performed to reconstruct new approximate and detail components, resulting in a clean EEG signal free of cardiac electric field artifacts, eye movement artifacts, and electromyography (EMG) artifacts.
[0047] Step 2, CBP signal feature point detection and processing;
[0048] The system automatically identifies the peak points (corresponding to systolic blood pressure, SBP) of each blood pressure waveform cycle based on first-order differential analysis and an adaptive threshold method. Specifically, it calculates the first derivative of the clean CBP signal to identify points where the peak slope changes (the zero-crossing point where the slope changes from rising to falling). If no valid peak is detected for two consecutive seconds, the detection threshold is automatically lowered, with a lower limit set (0.4 × average fluctuation range) to prevent oversensitivity. The SBP anchor point position sequence is obtained based on all detected consecutive CBP signal peak points. Each SBP anchor point is defined as an SBP event.
[0049] Step 3, RSP signal feature point detection and processing;
[0050] A local extremum detection algorithm is employed to identify local maxima (peaks) and local minima (valleys) of the signal within a sliding window (50 sampling points). A multi-window verification mechanism reduces noise interference; that is, the same extremum point must be identified in multiple windows, improving detection stability. Furthermore, the physiological characteristics of the respiratory signal (periodic changes in inspiration / expiration) are incorporated to filter out false extrema, ultimately yielding the respiratory peak point. and respiratory valley value point .
[0051] Step 4: Define the inspiratory and expiratory phases;
[0052] Respiratory valley value point and peak respiratory rate Perform time sorting, define the period from the current respiratory trough to the next respiratory peak (trough→peak) as the inspiratory phase, and define the period from the current respiratory peak to the next respiratory trough (peak→trough) as the expiratory phase.
[0053] Step 5, respiratory phase classification of SBP events
[0054] Iterate through the SBP events and determine the position corresponding to each SBP event. Store inspiratory phase event markers for either the inspiratory or expiratory phase. and exhalation phase event markers .
[0055] Step 6: Construct the inspiratory phase systolic blood pressure evoked potential (Inhale_SBPEP) and the expiratory phase systolic blood pressure evoked potential (Exhale_SBPEP).
[0056] For the inspiratory phase, the inspiratory phase event is obtained in step 5. For each anchor point, EEG segments corresponding to 200ms before and 600ms after the anchor point are extracted. All extracted EEG segments are aligned and averaged, and baseline correction is performed based on the 200ms leading time window before the anchor point to remove DC offset, resulting in... For the expiratory phase, the expiratory phase event is obtained in step 5. For each anchor point, EEG segments corresponding to 200ms before and 600ms after the anchor point are extracted. All extracted EEG segments are aligned and averaged, and baseline correction is performed based on the 200ms leading time window before the anchor point to remove DC offset, resulting in... .
[0057] Step 7, Quantify cognitive function evaluation indicators
[0058] Based on the results obtained in step 6 , The signal was located within a time window of 250-350 ms after the anchor point, and the classic P300 paradigm was used to quantify the coupling effect between respiration, cardiovascular system, and brain. The average amplitude of the EEG signal within this window was calculated, and this average amplitude value was defined as the inspiratory phase blood pressure-brain coupling P300 response. Expiratory phase blood pressure brain coupling P300 response This amplitude value is the core coupled physiological indicator used to objectively evaluate cognitive function.
[0059] Example
[0060] A total of 298 healthy subjects enrolled in plains and plateau (>3000 meters) environments were tested. Simultaneous RSP, CBP, and EEG signals were collected while the subjects were lying down. The Psychomotor Vigilance Task (PVT) was used to measure sustained attention and reaction speed in cognitive function, including reaction time (RT), standard deviation of reaction time (SDRT), and minor lapses. A Multi-Attribute Task Battery (MATB), comprising system monitoring, resource management, and target tracking tasks, was used to test executive function. The indicators included resource management error (Resman), target tracking distance error (Track), and system monitoring reaction time (SysmonRT). Results reflecting sustained attention, reaction speed, and executive function were then statistically analyzed. The extracted physiological indicators of quantified cognitive function were used to further analyze the results. , Correlation analysis was performed with cognitive test data to verify the method proposed in this invention.
[0061] Figure 5 In this process, the inspiratory and expiratory phases are defined based on RSP feature points, each SBP event is determined, and the inspiratory phase is marked. and exhalation phase . Figure 6 In, respectively based on and Anchor points are identified, and EEG segments from 200ms before to 600ms after each anchor point are extracted. All extracted EEG segments are aligned and averaged, and baseline correction is performed based on the leading time window to remove DC offset. and . Figure 7 In the middle, after the anchor point is located, a time window of 250-350ms is used to calculate the average amplitude of the EEG signal within this window, and obtain the result. , . Figure 8 In this study, cognitive function results from behavioral tests of 298 participants were combined with physiological signal analysis to obtain... , Spearman correlation analysis was performed on the indicators, and the results showed that the quantitative cognitive function indicators proposed in this invention are significantly correlated with the results of behavioral cognitive function tests such as PVT and MATB, which reflect attention, reaction time, and executive function. Specifically, the correlations are as follows: Compared with RT (R=-0.117, P=0.044), Compared with SDRT (R=-0.119, P=0.039), Compared with Minor Lapses (R=-0.163, P=0.005), Compared with Track (R=0.344, P<0.001), Compared with SysmonRT (R=-0.115, P=0.046), Compared with RT (R=-0.133, P=0.022), Compared with SDRT (R=-0.127, P=0.028), Compared with Minor Lapses (R=-0.185, P=0.001), Compared with Track (R=0.397, P<0.001), The results, along with Resman (R=-0.170, P=0.003), confirm the practicality and accuracy of the cognitive function quantification method proposed in this invention.
[0062] Because prolonged stays in high-altitude (>3000 meters) environments can impair a person's overall cognitive abilities, including reaction time, memory, and executive function, this invention addresses this issue. , The indicators were used to plot ROC curves for the plains and plateau populations, quantifying the ability of the indicators to differentiate between different cognitive function levels, such as... Figure 9 As shown. The area under the curve is 0.7352 (standard error 0.029). When the threshold is set to 0.19 μV, the sensitivity for differentiation is 80.10%. The area under the curve is 0.7289 (standard error 0.029). When the threshold is set to 0.15μV, the sensitivity for differentiation is 80.10%.
[0063] This application presents a method for acquiring coupled physiological indicators based on synchronized respiration, continuous blood pressure, and electroencephalogram (EEG) signals. The resulting coupled physiological indicators are free from subjective scoring interference and exhibit objectivity far superior to neuropsychological scales. This application, by integrating signals from the three systems of respiration (autonomic nervous system), continuous blood pressure (circulatory system), and EEG (central nervous system), proposes for the first time... and These two concepts and their extraction methods capture the dynamic coupling characteristics of blood pressure and brain under respiratory phase regulation, breaking through the limitations of existing methods that only assess the brain or autonomic nervous system individually. This approach is more in line with the physiological nature of the synergistic effect of multiple systems in cognitive function, and provides a more comprehensive assessment dimension. Furthermore, signal acquisition and analysis can be completed in real time, capturing dynamic fluctuations in cognitive function, supporting real-time feedback on the effects of cognitive interventions, and can also be used for long-term longitudinal monitoring of neurodegenerative diseases, providing dynamic evidence for clinical decision-making. More importantly, the quantified indicators in this application directly correspond to the well-defined physiological pathway of "respiratory phase-blood pressure fluctuation-cerebral perfusion-EEG response-cognitive function." That is, respiration affects cerebral blood flow perfusion by regulating blood pressure fluctuations, thereby altering the synchronicity of neuronal electrical activity and ultimately affecting core cognitive processes such as attention and executive function. The mechanism is highly correlated, and the indicators are highly interpretable. The signals required in this application can all be obtained non-invasively, the equipment is relatively widespread, and it is easy to promote and apply in clinical and research settings.
Claims
1. A method for calculating coupled physiological indicators based on synchronized respiration, continuous blood pressure, and electroencephalogram (EEG) signals, comprising: Synchronized physiological signal preprocessing steps; Respiratory, continuous blood pressure, and electroencephalogram (EEG) signals were preprocessed to obtain clean respiratory, continuous blood pressure, and EEG signals. Continuous blood pressure signal feature point detection and processing steps; identifying the peak point of systolic blood pressure in each cycle of a clean continuous blood pressure signal; each peak point is a SBP event; Steps for detecting and processing respiratory signal feature points; The peak and trough values of a clean respiratory signal are identified by a sliding window method to obtain the respiratory peak point and respiratory trough point; the respiratory trough point and respiratory peak point are sorted by time, and the period from the current respiratory trough point to the next respiratory peak point is defined as the inspiratory phase, and the period from the current respiratory peak point to the next respiratory trough point is defined as the expiratory phase. Respiratory phase classification steps for SBP events; Iterate through the SBP events and determine whether each SBP event belongs to the inspiratory or expiratory phase. The steps for constructing inspiratory systolic blood pressure evoked potentials (SBPs) are as follows: For the inspiratory phase, using the SBP event corresponding to inspiration as an anchor point and each anchor point as a reference point, clean EEG signals of a predetermined duration are extracted to form EEG signal segments. All EEG signal segments are aligned and averaged, and baseline correction is performed to obtain the inspiratory systolic blood pressure evoked potentials. For the expiratory phase, using the SBP event corresponding to expiration as a reference point, clean EEG signals of a predetermined duration are extracted to form EEG signal segments. All EEG signal segments are aligned and averaged, and baseline correction is performed to obtain the expiratory systolic blood pressure evoked potentials. The coupled physiological indicators consist of inspiratory systolic pressure evoked potentials and / or expiratory systolic pressure evoked potentials.
2. The method for obtaining physiological indicators based on the coupling of synchronized respiration, continuous blood pressure, and electroencephalogram signals according to claim 1, characterized in that: The predetermined duration is 100ms; this predetermined duration is 250-350ms after the reference point.
3. The method for obtaining physiological indicators based on the coupling of synchronized respiration, continuous blood pressure, and electroencephalogram signals according to claim 1, characterized in that: In the respiratory signal feature point detection and processing steps, after identifying the peak and trough values of the clean respiratory signal through the sliding window method, the noise interference is reduced through the multi-window verification mechanism, and false extreme values are filtered in combination with the physiological characteristics of the respiratory signal, and finally the respiratory peak point and respiratory trough point are obtained.
4. The method for obtaining physiological indicators based on the coupling of synchronized respiration, continuous blood pressure, and electroencephalogram signals according to claim 1, characterized in that: In the synchronous physiological signal preprocessing step For the respiratory signal, a Butterworth bandpass filter with a cutoff frequency of 0.05-1.0Hz was used to remove low-frequency motion artifacts and high-frequency ECG interference, and a 50 Hz notch filter was used to suppress power line interference to obtain a clean respiratory signal. For continuous blood pressure signals, baseline drift is removed by a Butterworth high-pass filter with a cutoff frequency of 0.5 Hz, motion artifacts are reduced by a Butterworth low-pass filter with a cutoff frequency of 5 Hz, and finally power frequency interference is eliminated by a 50 Hz notch filter to obtain a clean continuous blood pressure signal. For EEG signals, a single-channel EEG signal is processed using a 50Hz notch filter to eliminate power frequency interference, followed by filtering with a 0.05-100Hz bandpass filter to remove non-physiological artifacts. Then, the EEG signal is subjected to wavelet transform using Daubechies db4 discrete wavelets to obtain approximate and detail components. These components are then processed using complete empirical mode decomposition with adaptive noise to calculate intrinsic mode functions (EMFs). Next, independent component analysis (ICA) is used to calculate the independent components of the EMFs, and sample entropy quantization is performed. Independent components whose sample entropy values satisfy the Gomez-Herrero condition are considered artifacts and set to zero. Finally, inverse ICA is performed to reconstruct new approximate and detail components, resulting in a clean EEG signal.
5. A method for assessing cognitive function based on coupled physiological indicators, wherein the coupled physiological indicators are obtained by the calculation method according to any one of claims 1-4; characterized in that: When the coupled physiological index is greater than its corresponding threshold, the subject is considered to have cognitive impairment.
6. The cognitive function assessment method based on coupled physiological indicators according to claim 5, characterized in that: The coupled physiological index is obtained by the calculation method described in claim 2; The threshold value for the inspiratory systolic pressure evoked potential in the coupled physiological index is 0.15 μV; The threshold value for the expiratory systolic pressure evoked potential in the coupled physiological indicators is 0.19 μV.
7. A computing device based on coupled physiological indicators of synchronized respiration, continuous blood pressure, and electroencephalogram (EEG) signals, comprising: The synchronous physiological signal preprocessing unit is used to preprocess respiration, continuous blood pressure, and electroencephalogram (EEG) signals respectively to obtain clean respiration, continuous blood pressure, and EEG signals. The continuous blood pressure signal feature point detection and processing unit is used to identify the peak point of systolic blood pressure in each cycle of a clean continuous blood pressure signal, with each peak point being a SBP event. The respiratory signal feature point detection and processing unit identifies the peak and trough values of a clean respiratory signal through a sliding window method, obtains the respiratory peak point and respiratory trough point, sorts the respiratory trough point and respiratory peak point by time, and defines the period from the current respiratory trough point to the next respiratory peak point as the inspiratory phase and the period from the current respiratory peak point to the next respiratory trough point as the expiratory phase. The respiratory phase classification unit of SBP events traverses SBP events and determines whether each SBP event belongs to the inspiratory phase or the expiratory phase. The inspiratory systolic blood pressure evoked potential and the inspiratory systolic blood pressure evoked potential construction unit, for the inspiratory phase, use the SBP event corresponding to inspiration as the reference point, extract clean EEG signals of a predetermined duration to form EEG signal segments, align and average all EEG signal segments, and perform baseline correction to obtain the inspiratory systolic blood pressure evoked potential; for the expiratory phase, use the SBP event corresponding to expiration as the reference point, extract clean EEG signals of a predetermined duration to form EEG signal segments, align and average all EEG signal segments, and perform baseline correction to obtain the expiratory systolic blood pressure evoked potential. The coupled physiological indicators consist of inspiratory systolic pressure evoked potentials and / or expiratory systolic pressure evoked potentials.
8. The calculation device for coupled physiological indicators based on synchronized respiration, continuous blood pressure, and electroencephalogram signals according to claim 7, characterized in that: The predetermined duration is 100ms; this predetermined duration is 250-350ms after the reference point.
9. The calculation device for coupled physiological indicators based on synchronized respiration, continuous blood pressure, and electroencephalogram signals according to claim 7, characterized in that: The respiratory signal feature point detection and processing unit identifies the peak and trough values of clean respiratory signals through a sliding window method, reduces noise interference through a multi-window verification mechanism, and filters out false extreme values by combining the physiological characteristics of respiratory signals, finally obtaining the respiratory peak point and respiratory trough point.
10. The calculation device for coupled physiological indicators based on synchronized respiration, continuous blood pressure, and electroencephalogram signals according to claim 8, characterized in that: The threshold value for the inspiratory systolic pressure evoked potential in the coupled physiological index is 0.15 μV; The threshold value for the expiratory systolic pressure evoked potential in the coupled physiological indicators is 0.19 μV.