A real-time blood oxygen monitoring system based on photoelectric plethysmogram

Abstract

The present application relates to the technical field of medical diagnosis, and discloses a blood oxygen real-time monitoring system based on a photoelectric volume pulse wave, which comprises the following steps: inducing a vasomotor response through periodic thermal excitation, and demodulating a physiological pulse wave from a thermal modulation signal by using a synchronous phase-locked amplification technology. The present application solves the signal acquisition problem in a silent low perfusion state, converts background signals regarded as invalid noise in traditional monitoring into a resolvable vital sign carrier, and realizes non-invasive evaluation of microcirculation function through a thermal response index, thereby providing an early screening means for peripheral vascular diseases. The system maintains medical-grade precision while realizing the engineering feasibility of a consumer-grade device with an extremely simple hardware architecture.

Description

TECHNICAL FIELD

[0001] The present application relates to a real-time blood oxygen monitoring system based on photoelectric plethysmogram, belonging to the technical field of medical diagnosis. BACKGROUND

[0002] In the field of medical monitoring technology, photoelectric plethysmogram (PPG) blood oxygen monitoring technology faces a long-standing challenge, such as its effectiveness depends on the existence of identifiable pulse wave signals in the measured tissue. When patients are in a silent low perfusion state, such as shock, hypothermia or severe peripheral circulation disorder, the physiological signal intensity is lower than the detection threshold, and the traditional system is in a dilemma, or misjudges noise as valid signal and outputs false values, or directly gives up monitoring, resulting in the absence of key vital signs. The current mainstream solution usually relies on multi-wavelength optical filtering, inertial sensor fusion and complex signal processing algorithms to improve the signal-to-noise ratio in low perfusion state. However, this method is mainly based on the technical path of passively extracting pulse signals from background noise. In intensive care and other scenarios, such as peripheral circulation disorders caused by diabetes, pulse signals are easily overwhelmed by noise. In order to improve the recognition ability of weak signals, existing systems often need to integrate high-sensitivity sensors and high-performance processing modules, which to some extent increases the complexity and cost of system design, making it difficult to balance the engineering constraints of hardware resources and energy efficiency for consumer products.

[0003] Although some studies have tried to use deep learning and other methods to improve the robustness to noise, such methods often face problems such as large consumption of computing resources, long inference delay, etc. in actual deployment, making it difficult to achieve real-time processing on resource-constrained edge devices. Further analysis shows that the existing technical system has three deep bottlenecks: 1. physiological signals and noise are regarded as opposite entities, ignoring the convergence of energy levels of the two in extreme conditions, resulting in a lack of signal regeneration ability; 2. focusing on blood oxygen value extraction, failing to tap the dynamic response characteristics of the vascular network to physical excitation, resulting in the loss of clinically valuable data such as microcirculation function evaluation; 3. to improve the detection accuracy of weak signals, falling into a vicious cycle of hardware stacking, power soaring and cost out of control, deviating from the principle of universal health care.

[0004] Especially noteworthy is that in dynamic scenarios such as vascular elasticity degeneration (such as hypertension) or pathological reaction variation (such as Raynaud's syndrome), the existing solution lacks an adaptive mechanism for physiological response characteristics drift, further deteriorating its measurement robustness. This technical bottleneck not only restricts the ability of critical care, but also hinders the feasibility of home health management scenarios. Therefore, how to build a blood oxygen monitoring mechanism that breaks the dependence on signal existence, simultaneously realizes physiological parameter acquisition and vascular function evaluation, and meets the engineering constraints of consumer products, has become a technical problem to be solved by the present application. SUMMARY

[0005] The application provides a blood oxygen real-time monitoring system based on a photoelectric plethysmogram, which mainly aims to solve the problems of invalid capture of vital sign signals, lack of blood vessel function evaluation and insufficient landing of medical equipment engineering in a low perfusion state.

[0006] To achieve the above-mentioned purpose, the application provides a blood oxygen real-time monitoring system based on a photoelectric plethysmogram, which comprises:

[0007] a photoelectric plethysmogram signal acquisition unit configured to emit light signals to a measured tissue region and receive reflected or transmitted light signals from the measured tissue region to obtain an original photoelectric plethysmogram mixed signal;

[0008] a thermal excitation unit arranged adjacent to the photoelectric plethysmogram signal acquisition unit and configured to periodically apply a safe trace amount of thermal excitation to the measured tissue region according to a preset square wave driving frequency, so that the safe trace amount of thermal excitation induces the blood vessels in the measured tissue region to produce periodic dilation and contraction, thereby causing the physiological pulse wave signal contained in the original photoelectric plethysmogram mixed signal to be amplitude-modulated;

[0009] a signal processing unit electrically connected to the photoelectric plethysmogram signal acquisition unit and configured to: receive the original photoelectric plethysmogram mixed signal; input the original photoelectric plethysmogram mixed signal into a phase-locked amplification module, the phase-locked amplification module being configured to perform phase-locked processing on the original photoelectric plethysmogram mixed signal and a reference reference signal having a frequency of the square wave driving frequency and being phase-synchronized with the driving signal of the thermal excitation unit, so as to demodulate the physiological pulse wave signal from the original photoelectric plethysmogram mixed signal and amplify the amplitude thereof; calculate the blood oxygen saturation of the measured tissue region based on the demodulated and amplified physiological pulse wave signal; and quantitatively judge the microcirculation perfusion state of the measured tissue region according to the relationship between the real-time driving energy of the thermal excitation unit and the envelope amplitude of the demodulated and amplified physiological pulse wave signal.

[0010] Preferably, the thermal excitation unit comprises a patch resistor configured to receive a periodic electric pulse current from a pulse current driving module to generate the safe trace amount of thermal excitation.

[0011] Preferably, the thermal excitation unit is configured to generate the safe trace amount of thermal excitation by periodically overdriving the light-emitting diode itself in the photoelectric plethysmogram signal acquisition unit at the square wave driving frequency.

[0012] Preferably, the phase-locked amplification module is configured to realize phase-locked processing by the following steps: performing point-by-point multiplication operation on the original photoelectric plethysmogram mixed signal and the reference reference signal, and then performing digital low-pass filtering processing on the multiplication operation result.

[0013] Preferably, the signal processing unit is further configured to continuously monitor the envelope amplitude of the demodulated and amplified physiological pulse wave signal to calculate in real time a vascular thermal response index reflecting the response strength of the measured tissue region to the thermal stimulation; and the signal processing unit is configured to adjust adaptively the driving energy of the thermal stimulation unit according to the comparison result of the vascular thermal response index and a preset optimal interval, so as to maintain the vascular thermal response index within the preset optimal interval.

[0014] Preferably, the vascular thermal response index is calculated in the following manner: , wherein, represents the vascular thermal response index, represents the average value of the envelope amplitude of the demodulated and amplified physiological pulse wave signal within a preset duration window, represents the average thermal stimulation energy applied to the thermal stimulation unit within the same preset duration window.

[0015] Preferably, the signal processing unit is further configured to subtract the demodulated and amplified physiological pulse wave signal from the original photoplethysmogram mixed signal to obtain a residual noise signal; calculate a complexity index of the residual noise signal within a continuous time window, the complexity index being obtained by calculating the standard deviation of the first-order difference of the residual noise signal; and trigger a warning signal when the complexity index continuously and unilaterally deviates from an established individual baseline fluctuation range, the warning signal indicating that the measured tissue region may have a risk of physiological state instability.

[0016] Preferably, the frequency of the square wave driving signal ranges from 0.1 Hz to 0.5 Hz.

[0017] Preferably, the safe micro-thermal stimulation causes the local temperature of the measured tissue region to fluctuate within a range of 0.05°C to 0.5°C.

[0018] Preferably, the system further comprises an energy management unit configured to dynamically adjust the power supply of the thermal stimulation unit according to the real-time instructions of the signal processing unit, so as to maintain the total power consumption of the system within a preset energy-saving threshold range while ensuring the effectiveness of the safe micro-thermal stimulation.

[0019] Compared with the prior art, the present application has the following advantages:

[0020] 1. By injecting a preset frequency of micro-thermal stimulation into the measured tissue, the blood vessel network in an extremely low perfusion state generates a periodic dilation and contraction response, and the synchronous phase-locked amplification technology is used to capture the thermal stimulation modulated physiological pulse wave, so that the background signal which is regarded as invalid noise in traditional monitoring is converted into an analyzable vital sign carrier. This active detection mechanism enables the system to stably obtain the pulse waveform even in extreme working conditions such as peripheral circulation failure, thereby avoiding the risk of false reporting or missed diagnosis of conventional devices in a signal flooded state.

[0021] 2、System according to the thermal excitation energy and the dynamic coupling relationship of demodulation signal amplitude, generate reflecting vascular reactivity thermal response index, this parameter through monitoring the physiological response intensity of tissue to standard heat disturbance, quantitative characterization of peripheral vascular function state, this will detect energy and physiological feedback combined mechanism, in the process of blood oxygen monitoring synchronous realization of microcirculation perfusion evaluation, for early screening means of diabetic foot, Raynaud's syndrome and other vascular lesions; When the vascular reactivity occurs pathological fluctuation, the system real-time tracking demodulation signal envelope amplitude change, through the closed loop feedback dynamic adjustment of thermal excitation energy, this self-calibration mechanism makes the phase-locked amplification link always maintain in the linear working interval, avoid the idealized dependence of traditional active detection technology on physiological response consistency, ensure that the monitoring continuity is still maintained in the dynamic pathological state such as vasospasm.

[0022] 3、System from the residual signal after phase-locked processing to extract complexity index, through analyzing its deviation from individualized baseline trend, capture the nonlinear dynamics characteristics before the instability of cardiopulmonary compensation system, this early warning mechanism based on signal entropy change can identify the risk of occult hypoxia when the blood oxygen value has not decreased significantly, and strive for the key intervention window for critical care; The light source component or standard patch resistor is used for energy injection in the thermal excitation unit, and the signal demodulation process is completed by using basic multiplication and addition operation; The vascular response evaluation and risk warning function share the original signal processing pipeline, the simple design of hardware architecture and the cross-functional reuse of computing resources make the system maintain the medical level performance while realizing the engineering feasibility of consumer level equipment. BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1 The working flow chart of the blood oxygen real-time monitoring system based on the photoelectric plethysmogram of the application;

[0024] Figure 2 The VTRI (Vascular Thermal Response Index) change curve chart of the application;

[0025] Figure 3 The signal processing and evaluation framework chart of the blood oxygen real-time monitoring system of the application;

[0026] Figure 4 The working flow chart of the blood oxygen real-time monitoring system based on the photoelectric plethysmogram of the application.

[0027] The purpose realization, functional characteristics and advantages of the application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0028] It should be understood that the specific embodiments described herein are only used to explain the application, and are not used to limit the application.

[0029] The embodiment of the application provides a blood oxygen real-time monitoring system based on a photoelectric plethysmogram, in a clinical scene requiring continuous blood oxygen saturation monitoring and microcirculation function evaluation on a patient in a silent low-perfusion state, the system can operate in the following mode, a periodic thermal excitation is actively applied to induce the vasomotor response of the measured tissue region, and a synchronous phase-locked amplification technology is used to accurately demodulate the physiological pulse wave from the thermal modulation signal, thereby solving the challenge of signal acquisition difficulty in the low-perfusion state of the traditional monitoring method, and converting the background signal originally regarded as invalid noise into an analyzable vital sign carrier, and the system can also noninvasively evaluate the microcirculation function through a thermal response index, thereby providing an early screening means for peripheral vascular diseases; the system comprises a photoelectric plethysmogram signal acquisition unit, a thermal excitation unit and a signal processing unit, the photoelectric plethysmogram signal acquisition unit is configured to emit light signals to the measured tissue region and receive reflected or transmitted light signals from the measured tissue region to obtain an original photoelectric plethysmogram mixed signal, the unit usually comprises at least one light-emitting diode and at least one photodetector, the light-emitting diode emits light of a specific wavelength, the light penetrates or reflects the measured tissue and is received by the photodetector to be converted into an electrical signal, thereby forming a mixed signal containing physiological pulse wave information, environmental noise and vasomotor response information induced by the thermal excitation, in engineering implementation, the photoelectric plethysmogram signal acquisition unit is usually designed as a wearable sensor, such as a finger clip or a wristband, to ensure stable light path and comfortable wearing, the thermal excitation unit is arranged adjacent to the photoelectric plethysmogram signal acquisition unit, and the unit is configured to periodically apply a safe trace thermal excitation to the measured tissue region according to a preset square wave driving frequency, the safe trace thermal excitation induces the blood vessels in the measured tissue region to produce periodic dilation and contraction, so that the physiological pulse wave signal contained in the original photoelectric plethysmogram mixed signal is amplitude-modulated; meanwhile, the square wave driving frequency of the periodic thermal excitation mainly affects the effectiveness of the vasomotor response and the real-time performance of the monitoring. If the frequency is too high, the blood vessels may not respond fully due to physiological delay, resulting in weakened modulation effect. If the frequency is too low, the real-time performance of the monitoring is reduced, and the user may feel uncomfortable. In combination, the frequency is set in the range of zero point one hertz to zero point five hertz, so that the effective modulation is ensured, and the requirements of real-time performance and comfort are met, and a specific implementation manner of the thermal excitation unit is to comprise a patch resistor, the patch resistor is configured to receive periodic pulse current from a pulse current driving module to generate the safe trace thermal excitation; another implementation manner is that the thermal excitation unit is configured to generate the safe trace thermal excitation by periodically overdriving the light-emitting diode in the photoelectric plethysmogram signal acquisition unit itself at the square wave driving frequency; this multiplexing light-emitting diode mode can further simplify the hardware architecture and reduce the cost.

[0030] The signal processing unit is electrically connected with the photoplethysmogram signal acquisition unit, and the unit is configured to receive the original photoplethysmogram mixed signal. The received mixed signal is then input to a phase-locked amplification module. The phase-locked amplification module is configured to perform phase-locked processing on the original photoplethysmogram mixed signal and a reference signal with a frequency of the square wave driving frequency and synchronized in phase with the driving signal of the thermal excitation unit, to demodulate the physiological pulse wave signal from the original photoplethysmogram mixed signal and amplify the amplitude thereof. The phase-locked amplification module is configured to perform phase-locked processing by the following steps: performing point-by-point multiplication operation on the original photoplethysmogram mixed signal and the reference signal, and then performing digital low-pass filtering processing on the multiplication result. The combination of multiplication and low-pass filtering is a classic phase-locked amplification technology, which can effectively extract the weak signal in phase with the reference frequency from the noise background. Based on the demodulated and amplified physiological pulse wave signal, the signal processing unit calculates the blood oxygen saturation of the measured tissue region. The blood oxygen saturation is usually calculated by the red light and infrared light two-wavelength light absorption ratio method, and is converted through a pre-set calibration curve or algorithm model. The stability and high signal-to-noise ratio of the demodulated physiological pulse wave signal are the basis for accurate calculation of the blood oxygen saturation. The signal processing unit is further configured to quantitatively judge the microcirculation perfusion state of the measured tissue region according to the relationship between the real-time driving energy of the thermal excitation unit and the envelope amplitude of the demodulated and amplified physiological pulse wave signal. In order to realize this function, the signal processing unit continuously monitors the envelope amplitude of the demodulated and amplified physiological pulse wave signal to calculate the blood vessel thermal response index reflecting the response strength of the measured tissue region to the thermal excitation in real time. The calculation method of the blood vessel thermal response index is: = , wherein, represents the blood vessel thermal response index, represents the average value of the envelope amplitude of the demodulated and amplified physiological pulse wave signal in a pre-set duration windowrepresents the average thermal stimulation energy applied to the thermal stimulation unit within the same preset duration window; the signal processing unit is configured to adaptively adjust the driving energy of the thermal stimulation unit according to the comparison result of the vascular thermal response index and a preset optimal interval, so as to maintain the vascular thermal response index in the preset optimal interval. The fundamental technical consideration for setting the preset optimal interval is to achieve a technical optimization balance between ensuring that the vascular thermal response index can sensitively reflect the change of the microcirculation perfusion state and avoiding excessive thermal stimulation or insufficient stimulation. If the upper limit of the preset optimal interval is set too high, the system may apply excessive thermal stimulation energy in order to reach the high value, thereby excessively sacrificing the comfort and safety of the patient and possibly causing non-physiological vascular response. Conversely, if the lower limit of the preset optimal interval is set too low, the vascular response may be too weak in the low perfusion state, so that the change of the vascular thermal response index is not obvious, thereby excessively sacrificing the sensitivity of the microcirculation perfusion evaluation and the diagnostic accuracy. Therefore, in specific engineering practice, the determination of the optimal interval is not an isolated absolute value, but needs to be set in a reasonable engineering interval that can optimize the overall technical effect according to the physiological characteristics of the measured tissue, the detection sensitivity requirement of the system to the microcirculation change, and combined with clinical experience. The adaptive adjustment mechanism can ensure that the system can maintain the best monitoring performance under different individuals and different physiological states.

[0031] The signal processing unit is further configured to subtract the demodulated and amplified physiological pulse wave signal from the original photoplethysmogram mixed signal to obtain a residual noise signal, and calculate a complexity index of the residual noise signal in a continuous time window, the complexity index being obtained by calculating the standard deviation of the first-order difference of the residual noise signal. The fundamental technical consideration of the calculation of the complexity index is to achieve a technical optimization balance between effectively representing the randomness or disorder of the signal and avoiding misjudgment of normal physiological fluctuations. If the calculation method or threshold setting of the complexity index is too sensitive, the normal physiological noise or slight non-pathological fluctuations may be misjudged as physiological instability, thus excessively sacrificing the specificity of the early warning, leading to frequent false alarms. Conversely, if it is set too blunt, it may not be able to identify in time when the physiological state has shown signs of instability, thus excessively sacrificing the sensitivity and timeliness of the early warning, and may miss the critical intervention window. Therefore, in specific engineering practice, the determination of the complexity index is not an isolated absolute value, but needs to be set in a reasonable engineering interval that can optimize the overall early warning performance according to the statistical characteristics of the residual noise signal, the clinical definition of physiological instability, and the demand of the system for early warning. When the complexity index continuously and unidirectionally deviates from an established individual baseline fluctuation range, the system will trigger an early warning signal, indicating that the measured tissue region may have a risk of physiological instability. This early warning mechanism based on the complexity of the residual noise signal can identify the risk of occult hypoxia when the oxygen saturation value has not yet decreased significantly, thus gaining a critical intervention window for critical care.

[0032] The system further comprises an energy management unit configured to dynamically adjust the power supply of the thermal stimulation unit according to the real-time instructions of the signal processing unit, so as to maintain the total power consumption of the system within a preset energy-saving threshold range while ensuring the effectiveness of the safe micro-thermal stimulation. The fundamental technical consideration of the power supply adjustment of the energy management unit is to achieve a technical optimization balance between ensuring the effectiveness of the safe micro-thermal stimulation and maintaining the total power consumption of the system within the preset energy-saving threshold range. If the power supply is set too high, the total power consumption of the system may exceed the preset energy-saving threshold, thereby excessively sacrificing the portability and battery endurance of the device, which does not meet the engineering constraints of consumer devices. Conversely, if the power supply is set too low, the thermal stimulation may be insufficient, affecting the amplitude and stability of the vasomotor response, thereby excessively sacrificing the effective modulation of the physiological pulse wave and the accuracy of the microcirculation perfusion evaluation. Therefore, in specific engineering practice, the determination of the power supply is not an isolated absolute value, but needs to be set within a reasonable engineering interval that can optimize the overall technical effect, according to the characteristics of the thermal stimulation unit, the power consumption limit of the system, and the demand for thermal stimulation effect. This unit can ensure effective thermal stimulation while maximizing system energy efficiency to meet the low-power requirements of consumer devices.

[0033] Meanwhile, in the specific deployment and execution of the present application, the workflow of the system is strictly divided into an offline calibration phase and a continuous online monitoring phase. In the offline calibration phase, the core objective is to establish a universal optimal working interval for the vascular thermal response index by using an organization simulation body that can accurately control the perfusion rate of the internal blood simulation liquid through a precision pump. After starting the calibration process, the system is driven with a fixed and low-level thermal stimulation energy, and at the same time, the perfusion rate is gradually and linearly decreased from the normal circulation level. The signal processing unit continuously records the demodulated physiological pulse wave signal signal-to-noise ratio and the simultaneously calculated value. When the signal-to-noise ratio decreases to a preset engineering critical point where the basic morphological features of the pulse wave cannot be stably identified, the corresponding value at this moment is defined and stored as the lower limit of the optimal interval , and the upper limit of the interval is set as ninety percent of the average index value measured in a healthy resting state. This setting aims to reserve sufficient dynamic response space and avoid unnecessary adjustment actions for minor normal physiological fluctuations.

[0034] When the system is applied to a specific individual and starts online monitoring, it will first automatically enter a five-minute individual baseline establishment phase, during which the system assumes that the measured individual is in a physiologically stable state and continuously calculates the complexity index of the residual noise signal. The specific calculation process is as follows: first-order difference operation is performed on the residual noise signal, and then the standard deviation of the difference sequence in a two-second sliding time window is calculated. After five minutes, the system will store the arithmetic mean of all complexity index values during this period and the standard deviation . Based on this, a quantitative individual baseline fluctuation range is formally established, with the upper limit being and the lower limit being .

[0035] After the baseline is established, the system seamlessly switches to the real-time closed-loop monitoring and early warning mode. In this mode, the signal processing unit executes a fixed internal algorithm logic with a calculation period of every two seconds. First, it accurately calculates the vascular thermal response index of the previous period, and the calculation formula is: In this formula, is explicitly defined as the arithmetic mean of the physiological pulse wave signal envelope amplitude demodulated and amplified in the past two-second time window; is defined as the total thermal stimulation energy obtained by time integration of the product of the driving current and voltage applied to the thermal stimulation unit in the same time window. After the calculation is completed, the system compares the latest value with the optimal interval established in the offline calibration stage. If is lower than , the energy management unit will instruct the thermal stimulation unit to increase the driving energy by a preset minimum step; if is higher than , it will instruct it to decrease by the same step; if it is within this interval, the current energy will remain unchanged. This adjustment logic ensures that the detection sensitivity is always optimal, while the system simultaneously calculates the latest residual noise complexity index and compares it with the established individual baseline fluctuation range. The trigger condition for the early warning signal is strictly set as follows: when the system continuously detects five or more complexity index values, all of which are on the same side of the baseline fluctuation range (i.e., all are higher than the upper limit or all are lower than the lower limit), the early warning signal is formally triggered. Behind this series of operation processes is the systematic parameter setting of the digital low-pass filter in the phase-locked amplification module. The cutoff frequency of the filter is set to one-tenth of the thermal stimulation driving frequency to completely filter out high harmonic interference, and its equivalent time constant is set to thirty seconds, which aims to ensure the tracking sensitivity of minute-level microcirculation state changes while smoothing out meaningless transient noise, achieving an optimal trade-off between signal smoothness and physiological response tracking speed.

[0036] In one specific implementation scenario, when the patient is in a silent low perfusion state, for example in the early stage of shock or in a hypothermic environment, the peripheral circulation blood flow velocity is significantly slowed down, making it difficult for traditional photoplethysmographic pulse wave detection devices to capture effective physiological pulse wave signals. The system cannot perform blood oxygen saturation monitoring because the signal strength is below the detection threshold, and thus the clinical staff cannot obtain key vital sign data in a timely manner. To address this technical challenge, the system can operate as follows. The thermal excitation unit in the system periodically applies a safe and small amount of thermal excitation to the measured tissue area according to a preset square wave driving frequency. This thermal stimulation is not intended to directly change the blood oxygen saturation, but to convert an otherwise undetectable silent physiological state into a dynamic response that can be actively sensed by the system. Specifically, the small amount of thermal excitation can induce periodic dilation and contraction of the vascular network in a low perfusion state. Even under extremely low perfusion conditions, it can cause weak but periodic volume changes on the blood vessel wall. The periodic dilation and contraction of the original photoplethysmographic pulse wave mixed signal makes the physiological pulse wave signal contained therein amplitude modulated. This modulation effect converts the background signal, which is considered as invalid noise in traditional monitoring, into a vital sign carrier that can be analyzed.

[0037] Based on the demodulated and amplified physiological pulse wave signal, the system can accurately calculate the blood oxygen saturation of the measured tissue area. At the same time, the system does not stop at single blood oxygen monitoring. The signal processing unit further continuously monitors the envelope amplitude of the demodulated and amplified physiological pulse wave signal, and calculates the vascular thermal response index based on the relationship between the real-time driving energy of the thermal excitation unit and the envelope amplitude of the demodulated and amplified physiological pulse wave signal. The calculation method of the vascular thermal response index is as follows: wherein, represents the vascular thermal response index, This represents the average value of the envelope amplitude of the demodulated and amplified physiological pulse wave signal within a preset duration window. This index represents the average thermal excitation energy applied to the thermal excitation unit within the same preset duration window. This index quantifies the physiological response intensity of peripheral blood vessels to standard thermal disturbances, thereby simultaneously assessing the microcirculation perfusion status during blood oxygen monitoring. By unifying signal acquisition and microcirculation function assessment under the same physical excitation and signal processing architecture through thermal excitation, the signal processing unit adaptively adjusts the driving energy of the thermal excitation unit based on the comparison result between the vascular thermal response index and a preset optimal range, so as to stably maintain the vascular thermal response index within the preset optimal range. This closed-loop feedback mechanism ensures that the intensity of thermal excitation is always matched with the real-time reactivity of the measured tissue, avoiding discomfort caused by over-excitation and preventing signal modulation failure caused by insufficient excitation. It achieves adaptation to dynamic changes in vascular reactivity, ensuring the continuity and robustness of monitoring. Furthermore, the system possesses early warning capabilities. The signal processing unit subtracts the demodulated and amplified physiological pulse wave signal from the original photoplethysmography pulse wave mixture signal to obtain a residual noise signal. For the residual noise signal, the system calculates its complexity index within a continuous time window. The complexity index is obtained by calculating the standard deviation of the first difference of the residual noise signal. When the complexity index continuously and unidirectionally deviates from an established individual baseline fluctuation range, the system will trigger an early warning signal. The early warning signal indicates that the measured tissue area may have a risk of physiological instability. This early warning based on the nonlinear dynamic characteristics of the residual signal can provide an earlier warning than the traditional blood oxygen value decline, securing a critical intervention window for critical care. This architecture integrates active detection, synchronous demodulation, physiological assessment, and early warning functions into one unit, achieving medical-grade accuracy with a minimal hardware configuration. It provides a robust and promising paradigm for the accurate monitoring and early intervention of consumer-grade medical devices in complex physiological states.

[0038] In this experiment, the key parameters are set in strict engineering decision logic chain. For example, the frequency of the square wave driving of the thermal excitation unit needs to be optimized between ensuring effective induction of vasomotor response and avoiding too long monitoring period. The fundamental technical consideration for the frequency selection is that if the frequency is too high, the vascular smooth muscle may not fully dilate and contract due to its inherent physiological response delay, resulting in weakened amplitude modulation effect on the physiological pulse wave, and thus reducing the signal-to-noise ratio of subsequent phase-locked demodulation. On the contrary, if the frequency is too low, the single measurement time will be lengthened, sacrificing the real-time tracking ability of rapid changes in vital signs. Based on this technical trade-off, the driving frequency is set to 0.2 Hz, which not only falls within the preferred range of 0.1-0.5 Hz, but also ensures that a complete modulation and demodulation cycle can be completed every five seconds. Similarly, the cut-off frequency of the digital low-pass filter in the phase-locked amplification module is set to balance the smoothness of the output signal and the response speed of the system to physiological state changes. To effectively filter out noise components unrelated to the 0.2 Hz reference frequency, while ensuring the ability to track minute-level microcirculation perfusion state changes, the equivalent time constant of the filter is set to 50 seconds, which is much larger than the excitation signal period, thereby ensuring stable and accurate extraction of the demodulated signal envelope amplitude. After starting the test process, the flow rate of the precision syringe pump is first set to 5.0 ml / min to simulate normal peripheral perfusion in a healthy state. In this stage, the system runs stably for five minutes to establish a physiological baseline, at which time clear pulse waveforms can be observed from the original photoelectric plethysmogram mixed signal. Subsequently, the flow rate of the syringe pump is suddenly reduced to 0.5 ml / min to simulate the transition into a silent low perfusion state, and this state is maintained for continuous monitoring. During this transition, a decisive phenomenon is observed: as the perfusion volume drops sharply, the traditional pulse wave alternating current component driven by the beat in the original signal rapidly decays, and its amplitude is submerged in the background noise within two minutes and cannot be effectively identified. This accurately reproduces the signal loss dilemma that traditional photoelectric plethysmography devices inevitably face in such working conditions. However, the system of the present application successfully extracts the physiological pulse wave signal clearly marked by the 0.2 Hz frequency from the noise through active thermal excitation and synchronous phase-locked demodulation. Although the amplitude has decreased, the shape remains stable and analyzable. The example data in the following table quantitatively solidifies this key evidence.

[0039] Table 1: Comparison of pulse wave signal characteristics and vascular thermal response indices under different perfusion states.

[0040]

[0041] The above data reveals the underlying technical mechanism of this invention. Under low perfusion conditions, the failure of traditional signals is due to the extremely weak pulsatile changes in blood volume. This system, however, actively induces responsive vasoconstriction and vasodilation by applying periodic micro-thermal stimulation, transforming passive, weak physiological pulsations into a signal modulated by a specific frequency amplitude that can be targeted and identified. The lock-in amplification module utilizes this active labeling to accurately demodulate and amplify the signal from the noisy background under extremely low signal-to-noise ratio conditions. Furthermore, the vascular thermal response index drops sharply from 0.95 under normal perfusion to 0.20 under low perfusion. This change is not a simple signal amplitude attenuation, but rather a result of the demodulated physiological response amplitude... With the applied constant excitation energy The quantitative index obtained by normalization calculation objectively reflects the significant decrease in the blood vessel's ability to respond to standard thermal disturbances, thus enabling a direct and non-invasive assessment of the microcirculation perfusion status.

[0042] Example 3: This example combines Figures 1 to 4 This paper describes the implementation of a real-time blood oxygen monitoring system based on photoplethysmography (PPG), such as... Figure 1 As shown, the system's workflow begins with the thermal excitation unit, which periodically applies micro-thermal excitation by generating periodic thermal excitation signals at a frequency of 0.1-0.5 Hz. Next, the PPG acquisition unit acquires the photoplethysmography (PPG) signal, including a raw PPG mixed signal. This signal comprises the physiological pulse wave signal, environmental noise, and vasomotor response information induced by thermal excitation. After comparison with a synchronous reference standard signal, the signal enters the lock-in amplifier module for demodulation and amplification. The demodulated signal yields the physiological pulse wave signal. The signal processing unit further processes the demodulated signal and calculates blood oxygen saturation and the vascular thermal response index (VTI). The system dynamically adjusts the energy of the thermal excitation unit based on real-time monitoring data to assess the microcirculation perfusion status. This ensures the accuracy and stability of signal acquisition. Through this system, reliable blood oxygen monitoring data can be obtained under extremely low perfusion conditions, and microcirculation function can be assessed.

[0043] like Figure 2 As shown in the figure, the horizontal axis represents time (minutes), and the vertical axis represents the vascular thermal response index (VTI). The figure clearly shows the different vascular response states. Changes; under normal perfusion conditions, The value is 0.95, which is within the normal perfusion range. As the perfusion volume gradually decreases, The value drops rapidly and enters a silent hypoperfusion state. The value is approximately 0.20, below the minimum range of 0.4. The curve also indicates the optimal range (0.4-0.8), which represents the ideal range for normal vascular response. Within this range, vascular response is healthy and stable. The data in the figure reflect the dynamic changes of blood vessels under different perfusion states, providing a direct basis for assessing vasomotor function and microcirculation status, especially under low perfusion states. Changes in these values ​​can effectively monitor vascular reactivity and provide data support for early intervention.

[0044] like Figure 3 As shown, in this system, the thermal excitation unit first generates a periodic thermal excitation signal, and the PPG acquisition unit collects reflected or transmitted light signals. These signals include a raw photoplethysmography (PPG) pulse wave mixture signal, which contains a physiological pulse wave signal modulated by thermal excitation. Subsequently, the system uses a synchronous reference signal for support to perform demodulation and phase-locked loop (PLL) processing of the signal, and inputs it into the PPL amplification module. The PPL amplification module extracts the effective physiological pulse wave signal through demodulation and amplification, and finally sends it to the signal processing unit. The signal processing unit further calculates the blood oxygen saturation and vascular thermal response index based on the demodulated signal. The signal processing unit also performs microcirculation function assessment and early warning based on the characteristics of residual noise signals and the signal processing results. One of the key technologies of signal processing is demodulated waveform analysis, which further provides data support for blood oxygen saturation calculation and vascular thermal response index assessment. In addition, through complexity analysis based on residual noise signals, the system can realize early warning function for physiological instability, providing important auxiliary decision-making information for clinical practice.

[0045] like Figure 4As shown, first, the photoelectric plethysmogram signal acquisition unit is responsible for transmitting light signals to the measured tissue area and receiving reflected or transmitted light signals from the area, thereby obtaining the original photoelectric plethysmogram mixed signal, which contains physiological pulse wave signals, background noise and blood vessel dilation response information induced by thermal excitation. In order to process the signal, a thermal excitation unit is provided in the system. The thermal excitation unit applies a safe trace amount of thermal excitation to the measured tissue area periodically through a patch resistor (or LED) at a square wave driving frequency (0.1~0.5Hz), thereby inducing the blood vessels to produce periodic dilation and contraction, modulating the original photoelectric plethysmogram signal. Then, the signal processing unit inputs the acquired original signal to the lock-in amplification module. The lock-in amplification module is synchronized with the reference signal, and through point-by-point multiplication operation and digital low-pass filtering processing, the physiological pulse wave signal is extracted from the noise and amplified in amplitude. The signal processing unit calculates the blood oxygen saturation based on these demodulated and amplified signals, and also evaluates the microcirculation perfusion state using the vascular thermal response index. The signal processing unit will also adjust the driving energy of the thermal excitation unit in real time according to the comparison result of the vascular thermal response index and the preset optimal interval, so as to ensure that the vascular response index is stable within the optimal range. In addition, the system also includes an energy management unit, which dynamically adjusts the power supply of the thermal excitation unit according to the instructions of the signal processing unit, so as to ensure the thermal excitation efficiency while keeping the total power consumption of the system within the energy-saving range. Through this structure, the system realizes real-time processing and efficient detection of blood oxygen monitoring and microcirculation function evaluation in low perfusion state.

[0046] In this embodiment, the optimal interval of vascular thermal response index and its closed-loop feedback adjustment mechanism are first addressed. The detailed implementation procedure can be decomposed into the following steps. Before the system is applied to a certain type of measured tissue or population, a one-time offline calibration procedure can be performed. The procedure utilizes a tissue simulation body to gradually decrease perfusion rate from normal level by precise control. In this process, the system works with an initial, lower constant thermal excitation energy, and records the change of vascular thermal response index with perfusion rate. When the signal-to-noise ratio of the demodulated physiological pulse wave signal drops to a pre-set engineering critical point where the key feature points of the waveform cannot be stably identified, the corresponding vascular thermal response index value is defined as the lower limit of the pre-set optimal interval. The upper limit of the interval is set at a certain percentage position of the average index value measured in the healthy resting state, for example, at the ninety percent position, to retain the redundancy for normal physiological fluctuations and avoid unnecessary adjustment actions. After the system enters real-time monitoring work, the internal logic of the signal processing unit follows a clear, step-by-step adjustment rule: it calculates the average value of the vascular thermal response index in the past ten seconds at fixed time intervals, for example, every ten seconds, and compares the average value with the set optimal interval. If the average value is lower than the lower limit, the signal processing unit generates instructions to increase the driving energy of the thermal excitation unit by a fixed minimum step. If it is higher than the upper limit, it instructs to decrease by the same step. If it is within the interval, it instructs to maintain the current energy unchanged. This series of operations will adapt the functional description to a discrete, rule-based, directly executable control process.

[0047] Secondly, for the physiological instability early warning function based on residual noise complexity, its core individual baseline establishment and early warning trigger logic can be specified as follows: when the system starts monitoring a new individual, it will automatically enter an initialization baseline establishment phase, which lasts a preset time sufficient to cover normal physiological fluctuations, for example, five minutes, during which the system assumes that the subject is in a relatively stable physiological state, and continuously calculates the complexity index of the residual noise signal. After this phase, the system stores the arithmetic mean and standard deviation of all complexity index values during this period. Based on these two statistics, a quantitative individual baseline fluctuation range is established, with the upper limit defined as the average value plus the standard deviation multiplied by a preset multiple (e.g., three times), and the lower limit as the average value minus the standard deviation multiplied by the same multiple. After the baseline is established, the system enters the continuous monitoring and early warning mode, and the early warning trigger judgment logic is strictly defined as follows: when the system detects that the complexity index calculation value is on the same side of the baseline fluctuation range for a certain number of times (e.g., five times) or more, i.e., all above the upper limit or all below the lower limit, the early warning signal is officially generated and triggered. The combination of continuous counting and one-way deviation converts the description of continuous and one-way deviation into a judgment procedure that can be executed by a microprocessor without ambiguity, achieving a technical balance between sensitivity and specificity.

[0048] Furthermore, regarding how the energy management unit cooperates with the signal processing unit to achieve an optimal balance between ensuring effectiveness and maintaining low power consumption, its internal decision-making mechanism is as follows: the signal processing unit calculates a current required target thermal stimulation energy based on its closed-loop feedback algorithm of the vascular thermal response index, and transmits the target value to the energy management unit. The energy management unit has an internal absolute energy-saving threshold representing the upper limit of system power consumption, and its decision-making logic is a priority limiting mechanism: the energy management unit compares the received target thermal stimulation energy with the absolute energy-saving threshold. If the target energy is below or equal to the threshold, the energy management unit will accurately regulate the power supply to the thermal stimulation unit according to the target value. However, if the target energy exceeds the energy-saving threshold, the energy management unit will no longer follow the target, but will forcibly limit the power supply to the level equal to the energy-saving threshold. Through this peak clipping processing mechanism, the system ensures that the total power consumption does not exceed the preset limit in any case, and in extreme working conditions, the energy-saving engineering constraint is prioritized.

[0049] Finally, the parameter setting of the digital low-pass filter in the phase-locked amplification module aims to balance between filtering noise and preserving effective physiological information changes. The core parameter of the filter, i.e., its cutoff frequency, must be significantly lower than the thermal excitation square wave driving frequency as the reference benchmark. Specifically, the cutoff frequency can be set to one-tenth or lower of the driving frequency to ensure complete filtering of the modulated carrier frequency and its higher harmonics. At the same time, the cutoff frequency cannot be too low, otherwise it will be excessively smoothed, thereby masking the true physiological information representing slow changes in microcirculation state. Therefore, the second selection principle is that the equivalent time constant corresponding to the cutoff frequency must be shorter than the shortest effective physiological change time in the clinical application scenario. For example, it can be set to no longer than half of the shortest effective time. Combined with the specific excitation frequency and clinical monitoring needs, the optimal filter parameters can be systematically determined without creative labor, thereby ensuring the smoothness of the demodulated signal and the response speed to physiological changes.

[0050] In this embodiment, to ensure the calculation of the vascular thermal response index has clear engineering implementability, the envelope amplitude of the physiological pulse wave signal is defined as the mean value of the demodulated signal amplitude within a continuous ten-second time window, with units of millivolts, to reflect the instantaneous reaction intensity of the tissue to the standard thermal excitation. The thermal excitation energy is obtained by time integration of the product of the pulse current and voltage, with units of millijoules, representing the actual total energy applied to the thermal excitation unit within the same time window. The sampling frequency is preferably set to one hundred times per second to ensure the dynamic resolution of amplitude changes, and can be flexibly configured at the software level according to different application scenarios. This definition method can realize cross-individual comparison without introducing complex parameters, and has good engineering universality.

[0051] To further clarify the setting principle of the filter parameters in the phase-locked amplification module, the cut-off frequency should be ensured to be much lower than the reference frequency used for modulation, so as to effectively suppress the interference of high-order harmonics and background noise, and the preferred setting mode is below one-tenth of the driving frequency, so as to clearly define the passband range of the phase-locked demodulation. At the same time, in order to balance the response delay and data smoothness, the equivalent time constant needs to be controlled to be less than half of the shortest period of the change of the measured physiological parameter. For example, if the system faces the tracking of microcirculation changes at the sub-minute level, it is recommended to set the time constant between thirty seconds and sixty seconds. This setting path not only has a theoretical basis, but also is convenient to set in a parameterized form during system initialization, and does not need creative labor to deploy in actual hardware. In addition, it needs to be explained that, in order to truly simulate the optical signal acquisition characteristics under the condition of low perfusion, a microchannel structure with an inner diameter of two hundred microns can be configured and a simulated medium with tissue equivalent optical properties can be injected, so as to establish an experimental platform similar to the optical conditions of human subcutaneous capillaries. Scattering particles with a specific particle size can be added to the simulated medium to reproduce the reflection behavior of red blood cells in blood. A dual-band light source with wavelengths in the six hundred and sixty nanometer and nine hundred and forty nanometer intervals is used for signal injection, and a linear response photodetection component is used for receiving, which can effectively restore the influence of thermal modulation on the signal form.

[0052] In another specific embodiment of the present application, first, the calculation of the vascular thermal response index VTRI is defined in engineering. The envelope amplitude EPPA of the physiological pulse wave signal is inputted into the output of the phase-locked amplification module, the envelope peak of the signal sequence is arithmetically averaged in a sliding time window of two seconds, and the output is a real number representing the current pulse intensity. The thermal excitation energy ETIE is inputted into the instantaneous driving current and voltage values applied to the two ends of the thermal excitation unit, and the instantaneous power is time-integrated in the same two-second time window as EPPA, and the output is a real number in joules representing the total energy injected in the window.

[0053] Secondly, the calibration process of the preset optimal interval is regulated, and the lower limit VTRImin of the interval is determined. The engineering goal is to find the lowest physiological response boundary at which the system can stably identify the effective pulse waveform pattern. The determination process is as follows: use a tissue simulation body that can control the perfusion rate through a precision injection pump, start the system and apply a fixed initial thermal stimulation energy; linearly reduce the perfusion rate from the level of simulating normal circulation by 0.1 milliliter per minute; at each step, the signal processing unit continuously executes a pulse wave peak value detection algorithm and calculates the success rate of successfully identifying the pulse wave main peak in the past thirty seconds; when the success rate is first continuously lower than 95%, the system records the vascular thermal response index value at this moment and defines it as VTRImin.

[0054] Furthermore, the equivalent time constant of the digital low-pass filter in the phase-locked amplification module is set. The fundamental technical consideration is to achieve a technical optimization balance between the smoothness of the demodulation signal and the response speed of the system to physiological state changes. If the time constant is set too short, too much high-frequency noise will be left in the demodulated signal envelope, which will excessively sacrifice the stability of subsequent blood oxygen calculation and VTRI evaluation. Conversely, if it is set too long, it will excessively smooth out the real microcirculation state fluctuations at the minute level, which will excessively sacrifice the timeliness of system monitoring. Therefore, in specific engineering practice, the determination of the time constant is not an isolated absolute value, but needs to be set according to the square wave driving frequency of the thermal stimulation unit. Preferably, it is set to five to ten times the square wave driving period. For example, for a driving frequency of 0.2 Hz, i.e. a period of five seconds, the equivalent time constant can be set in the interval of 25 to 50 seconds, ensuring stable demodulation of the modulation signal and effective tracking of physiological changes.

[0055] Meanwhile, the calculation process of blood oxygen saturation is deepened to eliminate the potential influence of thermal modulation, and the input of the process is the mixed signal of the red light and infrared light original photoelectric pulse wave signals output by the photoelectric pulse wave signal acquisition unit; the processing process firstly inputs the two signals into the phase-locked amplification module in parallel, and uses the reference signal synchronized with the thermal excitation signal to demodulate the physiological pulse wave signals of the red light and infrared light respectively to obtain the envelope amplitude of the alternating component; at the same time, the two original signals are processed through a digital low-pass filter with a much lower cutoff frequency than the heart rate frequency to extract the direct current component; finally, the output is obtained by calculating the ratio of the envelope amplitude of the alternating component and the direct current component of the two signals, comparing the two ratios to obtain a ratio value, and substituting the ratio value into a pre-established empirical formula or lookup table based on the calibration of the standard oximeter to calculate the final blood oxygen saturation value; this process ensures that the influence of thermal modulation on the two signals is systematically offset in the final ratio calculation by performing consistent demodulation operations on the two light signals at the same stage of signal processing, thereby ensuring the accuracy of the blood oxygen calculation result.

[0056] And the setting of the absolute energy saving threshold of the energy management unit, the fundamental technical consideration is to achieve a technical optimization balance between ensuring monitoring performance and meeting the power consumption limit of a specific application scenario, especially when applied to a battery-powered portable or wearable device, the threshold is directly derived from the comprehensive calculation of the device battery capacity, target continuous working time and safe discharge specification, for example, a wristband device with a target of twenty-four hours of continuous work, the absolute energy saving threshold of its energy management unit will be set to ensure that the average value of its total power consumption does not exceed the value obtained by dividing the total energy of the battery by twenty-four hours, ensuring that the present application realizes medical-grade monitoring accuracy while complying with the core design constraints of a specific product.

[0057] It is obvious for those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and the present application can be implemented in other specific forms without departing from the spirit or essential characteristics of the present application.

[0058] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting, although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present application.

Claims

1. A photoplethysmographic-based blood oxygen real-time monitoring system, characterized in that, The system comprises: a photoplethysmographic signal acquisition unit configured to emit light signals to a measured tissue region and receive reflected or transmitted light signals from the measured tissue region to obtain a raw photoplethysmographic mixed signal; a thermal excitation unit disposed adjacent to the photoplethysmographic signal acquisition unit and configured to periodically apply a safe micro-thermal excitation to the measured tissue region according to a preset square wave driving frequency, the safe micro-thermal excitation inducing periodic dilation and contraction of blood vessels in the measured tissue region, so that a physiological pulse wave signal contained in the raw photoplethysmographic mixed signal is amplitude-modulated; a signal processing unit electrically connected to the photoplethysmographic signal acquisition unit and configured to: receive the raw photoplethysmographic mixed signal; input the raw photoplethysmographic mixed signal to a phase-locked amplification module, the phase-locked amplification module being configured to perform phase-locked processing on the raw photoplethysmographic mixed signal and a reference reference signal having a frequency of the square wave driving frequency and being phase-synchronized with a driving signal of the thermal excitation unit, so as to demodulate the physiological pulse wave signal from the raw photoplethysmographic mixed signal and amplify the amplitude thereof; calculate blood oxygen saturation of the measured tissue region based on the demodulated and amplified physiological pulse wave signal; and quantitatively judge a microcirculation perfusion state of the measured tissue region according to a relationship between a real-time driving energy of the thermal excitation unit and an envelope amplitude of the demodulated and amplified physiological pulse wave signal.

2. The real-time blood oxygen monitoring system based on the photoelectric plethysmogram according to claim 1, wherein, The thermal excitation unit comprises a patch resistor configured to receive a periodic electric pulse current from a pulse current driving module to generate the safe micro-thermal excitation. 3.The real-time blood oxygen monitoring system based on photoelectric plethysmogram of claim 1, wherein, The thermal excitation unit is configured to generate the safe micro-thermal excitation by periodically overdriving a light-emitting diode in the photoplethysmographic signal acquisition unit at the square wave driving frequency. 4.The real-time blood oxygen monitoring system based on photoelectric plethysmogram of claim 1, wherein, The phase-locked amplification module is configured to perform the phase-locked processing by: performing point-by-point multiplication operation on the raw photoplethysmographic mixed signal and the reference reference signal, and then performing digital low-pass filtering processing on a result of the multiplication operation.

5. The photoplethysmographic-based real-time blood oxygen monitoring system according to claim 1, wherein, The signal processing unit is further configured to: continuously monitor the envelope amplitude of the demodulated and amplified physiological pulse wave signal to calculate a vascular thermal response index reflecting a response intensity of the measured tissue region to the thermal excitation in real time; and the signal processing unit is configured to adaptively adjust the driving energy of the thermal excitation unit according to a comparison result of the vascular thermal response index and a preset optimal interval, so as to stably maintain the vascular thermal response index within the preset optimal interval.

6. The photoplethysmographic-based real-time blood oxygen monitoring system according to claim 5, wherein, The vascular thermal response index is calculated in the following way: wherein, represents the vascular thermal response index, represents the average value of the envelope amplitude of the demodulated and amplified physiological pulse wave signal over a predetermined duration window, represents the average thermal stimulation energy applied to the thermal stimulation unit over the same predetermined duration window. 7.The real-time blood oxygen monitoring system based on photoelectric plethysmogram of claim 1, wherein, The signal processing unit is further configured to: obtain a residual noise signal by subtracting the demodulated and amplified physiological pulse wave signal from the raw photoplethysmographic mixed signal; calculate a complexity index of the residual noise signal within a continuous time window, the complexity index being obtained by calculating a standard deviation of a first-order difference of the residual noise signal; and trigger a warning signal when the complexity index continuously and unidirectionally deviates from an established individual baseline fluctuation range, the warning signal indicating that the measured tissue region may have a risk of physiological state instability. 8.The real-time blood oxygen monitoring system based on photoelectric plethysmogram of claim 1, wherein, The square wave driving frequency ranges from zero point one hertz to zero point five hertz. 9.The real-time blood oxygen monitoring system based on photoelectric plethysmogram of claim 1, wherein, The safe micro-thermal excitation causes the local temperature fluctuation range of the measured tissue region to be 0.0-0.5 degrees Celsius. 10.The real-time blood oxygen monitoring system based on photoelectric plethysmogram according to claim 1, wherein, The system further comprises an energy management unit configured to dynamically adjust the power supply of the thermal excitation unit according to the real-time instructions of the signal processing unit.

Images