A non-invasive monitoring method for effectively monitoring intrinsic biological rhythms

Abstract

The application relates to the technical field of biological rhythm monitoring methods, and discloses a non-interference monitoring method for effectively monitoring internal biological rhythm. The application realizes the non-interference monitoring method for effectively monitoring internal biological rhythm through the following steps: constructing a high-spectrum imaging biological rhythm monitoring technology based on a wideband NIR-LED light source, constructing an algorithm for matching human rhythm data measurement and recognition, and optimizing an artificial lighting light environment matched with a circadian rhythm. The non-interference rhythm monitoring method combines the NIR light source with the high-spectrum imaging technology, reflects the biological rhythm through the interpretation of the facial near-infrared absorption grayscale image, clearly defines the non-interference rhythm monitoring mode combining the near range and the long range, and finally converts and simulates according to the circadian rhythm to guide the synchronous conversion of the biological rhythm of personnel, so that the non-interference monitoring research on the biological rhythm is completed. The research method is not only scientific, but also more accurate in research conclusion.

Description

Technical Field

[0001] This invention belongs to the technical field of biological rhythm monitoring methods, specifically, it relates to a non-intrusive monitoring method for effectively monitoring intrinsic biological rhythms. Background Technology

[0002] Biological rhythms are recurring characteristics in the physiological and behavioral processes of organisms. For example, the mimosa plant cultivated in darkness still exhibits a leaf movement rhythm corresponding to the alternation of light and darkness during the day, with its leaves opening during the day and closing at night. Endogenous biological rhythms are also influenced by environmental factors such as light and temperature, synchronizing them with changes in the external environment. Changes in human heart rate, core body temperature, respiratory rate, vital capacity, and airway resistance all show distinct diurnal rhythm characteristics, with peaks occurring during the day and minimums at night. In working environments lacking natural light-dark cycles, involving non-24-hour shifts, day-night shift work, and time zone crossings, biological rhythms, represented by core body temperature, become asynchronous with the sleep-wake cycle. In 1957, Lewis and Lobban observed the effects of non-24-hour cycles on human physiological and activity rhythms during the midnight sun in the Arctic region. The results showed that although the subjects' urination cycle could change with the environmental cycle, the rhythmic cycle of changes in urine K+ ion levels remained around 24 hours, indicating a discrepancy between the subjects' physiological and activity rhythms. A similar phenomenon was confirmed by the studies of Aschoff and Wever.

[0003] Disruptions to the body's circadian rhythms can trigger insomnia and various discomforts, negatively impacting cognition and behavior, affecting health, reducing work efficiency, and making it difficult to adapt to normal work and social rhythms. On one hand, the lowest core body temperature (CBT) typically occurs two hours before automatic wakefulness (around 4-5 AM), a time when drowsiness and sleepiness are most likely to occur. Disruptions to the circadian rhythm cause the core body temperature to be at its lowest when alertness is needed, significantly impairing alertness and reaction speed. On the other hand, higher body temperatures at night can make it difficult to fall asleep. Therefore, developing non-intrusive monitoring technologies that can effectively monitor internal circadian rhythms is of paramount importance for improving work efficiency.

[0004] Effective monitoring of circadian rhythm changes provides a foundation for intervening in biological rhythms using environmental factors. Throughout evolution, the diurnal variation of light has been the most significant environmental factor influencing human biological rhythms, with light playing a guiding role. Providing light stimulation at different times leads to different phase changes in the biological rhythm. Light exposure before the lowest body temperature helps delay the phase, while light exposure after the lowest body temperature helps advance the phase. If the biological rhythm and sleep-wake rhythm are out of sync, such as due to prolonged artificial lighting or shift work, sleep quality will be affected. Therefore, light can be used to restore the sleep-wake rhythm to normal, thereby alleviating the impact of sleep disorders on physical and mental health.

[0005] In view of this, the present invention is proposed. Summary of the Invention

[0006] To solve the above-mentioned technical problems, the basic concept of the technical solution adopted by the present invention is as follows:

[0007] An effective non-intrusive monitoring method for intrinsic biological rhythms includes the following steps:

[0008] Step 1: Combine NIR light source with hyperspectral imaging technology to reflect biological rhythms by interpreting near-infrared absorption grayscale images of the face;

[0009] Step 2: Construction of algorithms for measuring and matching human rhythm data;

[0010] Step 3: Optimization of artificial lighting environment to match the daily rhythm; Step 2 specifically includes the measurement and correlation analysis of physiological information data, data matching based on image recognition and machine learning, and statistical methods for physiological indicators;

[0011] The measurement and correlation analysis of the physiological information data specifically involves: conducting experimental research on a group of subjects who meet the physical examination requirements; simulating an indoor living environment in a laboratory setting, focusing on creating highly similar acoustic, optical, electrical, and thermal physical environment characteristics; and focusing on the sleep stage as the key monitoring object. The experiment is conducted in stages and over a long period throughout the sleep-wake cycle. The measurement and correlation analysis of the physiological information data includes three different monitoring methods, which are detailed below:

[0012] A. Physiological data are collected using medical measurement equipment. Real-time monitoring and collection of the subjects' EEG (electroencephalogram), ECG (electrocardiogram), EMG (electromyography), SpO2 (blood oxygen saturation), SKT (skin temperature), and PPG (pulse) indicators are performed and stored to form a physiological measurement dataset.

[0013] B. Data collection was conducted using an integrated smart mattress, focusing on real-time measurement and recording of the subjects' heart rate, respiratory rate, and body movement cycle during sleep on the mattress. A sleep measurement dataset was generated after the experiment.

[0014] C. Infrared temperature measurement data acquisition: An infrared thermometer was used to measure the temperature of the human head and face. The surface temperature of several positioning points was recorded in real time during sleep. A facial temperature dataset was generated after the experiment.

[0015] Subjective measurements were taken of the participants before and after the experiment. Questionnaires were used to collect subjective feelings oriented towards environmental perception and cognition. Participants also completed the Pittsburgh Sleep Index, Brief Mood State Scale, and Self-Rating Fatigue Scale to obtain measurement data of personal evaluation before and after the experiment. SPSS software was used to perform statistical analysis on the questionnaire data.

[0016] Establish the correlation between the data of indicators A, B, and C, that is, the correlation between physiological indicators over time under the same experimental conditions and the same physiological indicators.

[0017] Based on the common indicators in the three datasets A, B, and C, and using the physiological measurement dataset as the standard, the measurement validity of the sleep measurement dataset and the facial temperature dataset is verified. By integrating all datasets, the correlation between the physiological measurement dataset, the sleep measurement dataset, and the facial temperature dataset is established, and then a complete dataset integrating all indicators of the three datasets is established.

[0018] As a preferred embodiment of the present invention, the face is irradiated with NIR light. Firstly, the invisible NIR light will not affect normal sleep. Secondly, the absorption of NIR by different areas of the face will fluctuate due to changes in biological rhythms, which will be displayed as differences in grayscale in the photos taken by the NIR camera.

[0019] In a preferred embodiment of the present invention, the first step further includes a packaging method for a broadband NIR-LED light source and an image processing method; the packaging of the broadband NIR-LED light source incorporates mature white LED technology.

[0020] Image processing methods include principal component analysis and band ratio (BR) algorithm. Principal component analysis is a commonly used method for dimensionality reduction and noise reduction, which mainly represents a large amount of information with a small number of independent variables through linear combination.

[0021] The principle is as follows: ;

[0022] In the formula, PC k Represents the image of the k-th principal component; a iThis represents the i-th weight coefficient; p i This represents the image of the i-th band; n represents the number of bands in the original image.

[0023] In a preferred embodiment of the present invention, the band ratio (BR) algorithm obtains a relative band intensity image by dividing two bands.

[0024] The principle is as follows: BW (i,j,r) = BW (i,j,m) / BW (i,j,n)

[0025] In the formula: BW (i,j,r) For the corresponding pixels in the image under different bands ( i , j The ratio of ) BW (i,j,m) and BW (i,j,n) For pixels at the same position in bands m and n ( i , j The grayscale value of ).

[0026] As a preferred embodiment of the present invention, the second step specifically involves: constructing a comprehensive collection of human biorhythm data under laboratory conditions, using medical physiological data collection, sleep monitoring, infrared reflection images, and subjective questionnaire measurements to obtain the effective correlation between human biorhythm data and monitoring methods; systematically calculating and analyzing the characteristics of human biorhythm fluctuations, and then combining hyperspectral facial image recognition and analysis to complete the matching of NIR hyperspectral face image sets and physiological datasets, thereby clarifying a non-intrusive rhythm monitoring method that combines near and long-range measurements.

[0027] In a preferred embodiment of the present invention, during the above process, the infrared reflectance of the face at different wavelengths is used to obtain a series of face image sets under infrared conditions; SPA (Successive Projection Algorithm) is used.

[0028] This method selects a wavelength in the initial state, and then proceeds forward in a cyclical manner. It selects the maximum wavelength of the projection vector by calculating the projection on the unselected wavelengths, and then introduces the vector into the wavelength combination until the loop ends.

[0029] The prediction results of optimal wavelength modeling were tested using GA-BP (Genetic Algorithm-Based Neural Network). Based on the traditional BP which only has an input layer, hidden layer and output layer, GA was used to optimize the initial weights and thresholds of the BP neural network, enabling the model to obtain a higher prediction correlation coefficient.

[0030] Finally, the matching of the NIR hyperspectral face image set and the physiological dataset was completed, thereby obtaining a model algorithm for estimating biological rhythms by non-interference continuous monitoring parameters.

[0031] In a preferred embodiment of the present invention, the statistical method for the physiological indicators includes analyzing all data using SPSS 19.0;

[0032] Normally distributed measurement data are described using the mean and standard deviation. Comparisons between two groups are performed using the t-test; comparisons among multiple groups are performed using one-way ANOVA; pairwise comparisons are performed using the SNK method; and comparisons between groups are performed using... χ 2 test;

[0033] Logistic regression was used for multivariate analysis, and P < 0.05 was considered statistically significant.

[0034] As a preferred embodiment of the present invention, the third step includes the packaging of sunlight-like, ultra-high color index LED white light devices and the optimization of artificial lighting environment to match the circadian rhythm;

[0035] Recently, artificial lighting environment optimization that matches the circadian rhythm specifically includes:

[0036] Three white LEDs of three different color temperatures were selected to form a three-channel light source. Based on the principle of pulse width modulation (PM), the three channels were dimmed in a time-division manner. At a certain lighting test space distance, the spectral distribution of each channel working independently was first measured. The theoretical illuminance, correlated color temperature, and rhythm factor of the three-channel mixed light source at any ratio were obtained through dimming calculation methods. Then, the actual illuminance, correlated color temperature, and rhythm factor of the three-channel light source at different mixing ratios were measured. The theoretical and actual calculated illuminance, correlated color temperature, and rhythm factor were then compared, and the error was analyzed.

[0037] In the laboratory, lighting environments for different behaviors were constructed, and simulation and comparative experiments were conducted on the lighting environments for work, dining, and sleep behavior modules to form scenario lighting; and conversion simulations were carried out according to the daily rhythm to guide the synchronous conversion of people's biological rhythms.

[0038] Compared with the prior art, the present invention has the following advantages:

[0039] This invention achieves a non-intrusive monitoring method for effectively monitoring intrinsic biological rhythms by employing hyperspectral imaging biorhythm monitoring technology based on broadband NIR-LED light sources, constructing algorithms for human rhythm data measurement and identification matching, and optimizing the artificial lighting environment to match the circadian rhythm. It combines NIR light sources with hyperspectral imaging technology, reflecting biological rhythms through the interpretation of near-infrared absorption grayscale images of the face. The method utilizes medical physiological data acquisition, sleep monitoring, infrared reflectance images, and subjective questionnaire measurements to obtain an effective correlation between human rhythm data and monitoring methods. The system calculates and analyzes the characteristics of human biological rhythm fluctuations, then combines hyperspectral facial image recognition and analysis to match NIR hyperspectral facial image sets with physiological datasets, clarifying a non-intrusive rhythm monitoring method combining near and long-range approaches. Finally, it performs conversion simulations based on the circadian rhythm to guide the synchronous conversion of human biological rhythms, completing the non-intrusive monitoring research of biological rhythms. The research method of this invention is not only scientific but also yields more accurate research conclusions.

[0040] The specific embodiments of the present invention will now be described in further detail with reference to the accompanying drawings. Attached Figure Description

[0041] In the attached diagram:

[0042] Figure 1 (a) is a schematic diagram of the interaction between NIR light and body tissues and blood components;

[0043] Figure 1 (b) is a schematic diagram of NIR light source used to display capillaries in the palm;

[0044] Figure 1 (c) is a schematic diagram of the principle of measuring hyperspectral images of the face based on NIR light. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments will be clearly and completely described below with reference to the accompanying drawings. The following embodiments are used to illustrate the present invention.

[0046] Discovery and understanding of human biological rhythms

[0047] The 2017 Nobel Prize in Physiology or Medicine was awarded to three American geneticists for their contributions to the field of circadian rhythms. Circadian rhythms are essentially an evolutionary timing mechanism that enables organisms to predict environmental changes and adjust their behavior, physiological, and biochemical responses to adapt. The molecular mechanism is the transcriptional-translational feedback loop (TTFL), primarily composed of positive regulatory elements such as CLOCK and BMAL1, and negative regulatory elements such as PER (period) and CRY (cryptochrome) transcription factors. The suprachiasmatic nucleus (SCN) in the mammalian hypothalamus acts as the control center, regulating the circadian rhythms of the entire organism. Synchronization between the SCN and the peripheral biological clock system is achieved through more direct pathways such as the central nervous system and hormone secretion, or more indirect methods such as regulating body temperature and feeding. Li Wenqi's research indicates that endogenous hydrogen peroxide in mammals exhibits diurnal oscillation characteristics. Jian Kunlin et al. found that under normal work-rest schedules, tunnel workers exhibited significant rhythmic changes in body temperature, serum IgA, IgG, IgM, complement C3, and C4 levels (P<0.05). Biological rhythms can also cause changes in psychological state. Ma Wenjuan's research showed that compared to artificially created dark environments during the day, subjects experienced stronger fear at night, indicating that brain regions associated with fear exhibit rhythmic changes with the day-night cycle.

[0048] Disruptions to the biological rhythms of the human body can lead to disturbances in the sleep-wake cycle, such as insomnia, delayed / advanced sleep phases, and irregular sleep-wake syndrome, resulting in physical reactions such as sluggishness, poor mental performance, gastrointestinal discomfort, mania, and depression. Disruptions to the rhythms caused by non-24-hour circadian rhythms and shift work are important physiological mechanisms contributing to poor mental and physical performance during wakefulness. The polar day and polar night environment at my country's Antarctic research stations can cause REM sleep shifts in team members, leading to decreased attention, reduced self-control, decreased sensitivity, and mood swings during wakefulness.

[0049] Common methods for measuring biological rhythms

[0050] Circadian rhythms can be characterized by a series of physiological indicators, such as body temperature, cortisol, melatonin, five immune system indicators, basal metabolic rate, heart rate, blood pressure, blood cell counts, and biochemical parameters. Ren Huifeng et al. used a polysomnography system to measure overnight polysomnography records of shift workers, including electroencephalograms (EEG), electrocardiograms (ECG), eye movements (EOG), and myocardiograms (EMG), to study the impact of shift work on circadian rhythms. Liu Dandan used a vital signs monitor to monitor human physiological parameters and found that the correlation coefficients between heart rate variability and rhythm factors were all greater than 0.77, and the correlation coefficients between systolic blood pressure variability and rhythm factors were all greater than 0.86, indicating a good correlation. However, diastolic blood pressure did not correlate with circadian rhythm factors. Wang Shaoping et al. used a 12-lead Holter monitor to find that QRS complex amplitude exhibited diurnal rhythm variations and was correlated with heart rate. Zhou Liangliang et al. used a Holter monitor to monitor 24-hour average systolic blood pressure and explored the relationship between epicardial fat volume and blood pressure levels and diurnal blood pressure rhythm variations. Zhou Peng et al. recorded the brain activity of subjects using an electroencephalogram (EEG) recording system to explore the impact of sleep restriction-induced circadian rhythm disruption on brain alertness. Foreign researchers assessed gene activity by measuring changes in blood ribonucleic acid (RNA), then used machine learning algorithms to calculate which genes better reveal human circadian rhythms; the difference between the test results and those measured using melatonin was within half an hour. Human body temperature changes have a diurnal (24-hour) periodicity, typically reaching its lowest point in the early morning and its highest in the evening, with a difference of approximately 1°C between the peak and trough values ​​during the day and night. Within the normal range of body temperature variation, when body temperature is higher, especially close to the peak of the day, cognitive ability and work efficiency increase accordingly; conversely, when body temperature is lower, cognition and work efficiency decrease. Zhou Xiaoming et al. used an infrared thermometer to measure forehead temperature and fitted the forehead-temperature difference curve of photobiological rhythm factors, finding that photobiological rhythm factors increased linearly with changes in physiological signs, with a linear correlation of 0.95. Furthermore, by incorporating flexible pressure sensors into smart sheets, mattresses, pillows, or other wearable electronic devices, it is possible to non-invasively monitor physiological signals such as the human body's pulse and heartbeat, thereby obtaining changes in circadian rhythms. For example, Lin et al. designed a sheet that can detect changes in a user's sleeping posture. It can achieve real-time pressure signal acquisition through frictional electrification. By analyzing the collected electrical signals, sleep quality and health status can be assessed.

[0051] The non-visual effects of light on guiding biological rhythms

[0052] Modern medical research shows that the visual system not only transmits light signals to the brain but also transmits information about the brightness of the external light environment to the pineal gland. The pineal gland is a type of indole substance mainly synthesized from tryptophan. The pineal gland produces melatonin, which enters the blood, urine, cerebrospinal fluid, and intracellular and extracellular fluids, thereby controlling a person's sleep levels. Melatonin biosynthesis is regulated by the light-dark cycle. Changes in light are the most significant factor influencing melatonin synthesis in the external environment. For example, in summer mornings, when natural light shines on the human body earlier, melatonin in the body breaks down rapidly, allowing people to wake up earlier and feel more energetic. In winter, when it gets light later, people who get up early for school or work often feel less energetic upon waking up. This is because the body lacks light exposure in winter mornings, and the concentration of melatonin in the body cannot decrease rapidly, resulting in a lack of energy upon waking. Therefore, the light-dark cycle adjusts the body's internal biological rhythms (including diurnal, seasonal, and annual rhythms) by influencing the diurnal rhythm of melatonin synthesis. Nighttime light exposure can inhibit melatonin synthesis, and the spectral distribution and intensity of light have different effects on melatonin synthesis. Studies have shown that red light has the weakest inhibitory effect on melatonin, while blue-green light has the strongest inhibitory effect. Strong light can inhibit melatonin synthesis in the pineal gland instantly, while weak light requires a relatively long time to produce an inhibitory effect. Souman et al. studied the inhibitory effect of light-emitting diode (LED) light sources with different short wavelength contents on melatonin under the same illuminance (175 lux) and color temperature, and found that white light with reduced wavelength components of 450-500 nm and increased wavelength components of 400-430 nm had a significantly weaker inhibitory effect on melatonin.

[0053] The Earth's rotation causes changes in the light-dark cycle of the environment, which is an important factor in regulating circadian rhythms. The retina receives light stimulation, which is transmitted to the circadian rhythm center (SCN) via the retinal-hypothalamic tract. It is not the cone and rod cells of the imaging system that receive light stimulation, but rather the retinal ganglion cells containing melanopsin located deep within the retina. Light, as a timing factor, can strengthen and guide circadian rhythms. Patients with shift work disorder (SWD) can reduce light stimulation by wearing sunglasses after work, keeping their bedroom environment dark during their actual sleep time, and providing sufficient light during night shifts, thereby guiding their biological clock and extending their sleep time the following morning.

[0054] Adjusting the light spectrum can alter the effects of light on cognitive function, and clinical studies have found that blue light appears to enhance human cognitive function. Working under blue-spectrum light at night can reduce error rates and accelerate reaction speed. Blue light deprivation in gerbils leads to depressive-like behavior. The alertness effect of blue light is independent of its intensity. Short-wavelength blue light can significantly enhance the activity of brain regions related to cognitive and emotional processing. Subjects under blue light exposure were more sensitive to attentional processing of emotional stimuli, and the activity of brain regions related to emotional processing, such as the hippocampus, amygdala, and hypothalamus, was significantly enhanced. Lu Yuhong et al. used a dosage-based occupational assay, physiological parameter testing, and fatigue evaluation method to study the effects of narrow-band blue light with peak wavelengths of 468, 457, and 453 nm on human photobiological effects. The results showed that subjects under blue light with a peak wavelength of 468 nm had the highest mental work capacity index, the fastest working speed, the most comfortable working experience, and were less prone to fatigue.

[0055] Compared to a fixed color temperature light source, when the human body is exposed to a variable light source like sunlight, drowsiness and fatigue during mental activity are lower, calculation accuracy is higher and time is shorter, and the alpha wave component on the electroencephalogram (EEG) is significantly higher while the beta wave component is significantly lower. Therefore, it can be concluded that if the color temperature of indoor lighting sources could be varied throughout the day, similar to sunlight, it could make people feel relaxed, comfortable, and mentally alert, and could also promote and improve mental activity and increase work efficiency.

[0056] The non-visual effects of light are also manifested in the organic effects of artificial / natural light environments on the eyes. For example, rats are typically nocturnal animals, with their intraocular pressure peaking at night and significantly decreasing during the day. Experiments have shown that the peak intraocular pressure in rats in the artificial light group occurred significantly earlier than in rats in the natural light group, and this was statistically significant. Artificial light can cause the axial length of the developing rabbit eye to increase, thereby altering its refractive power. The uncorrected visual acuity of Kazakhs in pastoral areas, whose primary lighting is outdoor natural light, is significantly better than that of Kazakhs in urban areas, whose primary lighting is indoor artificial light.

[0057] Effective monitoring of circadian rhythm changes provides a foundation for intervening in biological rhythms using environmental factors. Throughout evolution, the diurnal variation of light has been the most significant environmental factor influencing human biological rhythms, with light playing a guiding role. Providing light stimulation at different times leads to different phase changes in the biological rhythm. Light exposure before the lowest body temperature helps delay the phase, while light exposure after the lowest body temperature helps advance the phase. If the biological rhythm and sleep-wake rhythm are out of sync, such as due to prolonged artificial lighting or shift work, sleep quality will be affected. Therefore, light can be used to restore the sleep-wake rhythm to normal, thereby alleviating the impact of sleep disorders on physical and mental health.

[0058] An effective non-intrusive monitoring method for monitoring intrinsic biological rhythms, such as Figure 1 As shown, the work steps include the following:

[0059] Step 1: Hyperspectral Imaging Biorhythm Monitoring Technology Based on Broadband NIR-LED Light Source

[0060] Near-infrared (NIR) light, with wavelengths between 700 and 2500 nm, has wide applications in agriculture, food, and biomedicine. Considering the high overlap between the three optical transmission windows of biological tissues (first window 650-950 nm, second window 1000-1350 nm, and third window 1500-1800 nm) and the NIR band, the different absorption coefficients of NIR light by different tissues can be utilized to achieve medical applications such as in vivo imaging and non-invasive blood oxygenation detection. Figure 1 (a) and Figure 1(b) are shown.

[0061] Biorhythms cause periodic changes in a series of physiological indicators, such as core body temperature, pulse, blood oxygen content, blood flow velocity, muscle tone, and skin resistance. Measuring a single indicator may not fully reflect the regularity of biorhythms. Considering that the most important part of biorhythm monitoring is the nighttime sleep stage, during which the face is exposed for easy measurement, irradiating the face with NIR light is advantageous. Firstly, the invisible NIR light will not affect normal sleep; secondly, the absorption of NIR by different areas of the face fluctuates due to changes in biorhythms, appearing as differences in grayscale in NIR camera images. Hyperspectral imaging uses a series of narrow-band images to form a three-dimensional image dataset of the object under test, simultaneously obtaining spectral and spatial resolution. Hyperspectral technology has been widely used in food safety, medical diagnostics, and aerospace fields. Therefore, this project aims to combine NIR light sources with hyperspectral imaging technology to reflect biorhythms by interpreting the near-infrared absorption grayscale images of the face. The principle is as follows: Figure 1 As shown in (c).

[0062] Packaging of broadband NIR-LED light sources

[0063] NIR light sources are crucial for hyperspectral imaging systems. Drawing inspiration from mature white LED (Light Emitting Diode) technology, near-infrared light can be generated by exciting specific phosphors with the visible blue light emitted from an LED chip. Compared to traditional near-infrared light sources such as halogen lamps, continuously tunable lasers, and near-infrared LEDs based on narrowband semiconductor materials, phosphor-converted NIR light-emitting devices offer advantages such as small size, all-solid-state operation, high stability, high energy conversion efficiency, long lifespan, and broadband emission.

[0064] Image processing methods

[0065] Principal Component Analysis (PCA) is a commonly used method for dimensionality reduction and noise reduction. It primarily represents a large amount of information using a smaller number of independent variables through linear combinations. Its principle is as follows:

[0066]

[0067] In the formula, PC k Represents the image of the k-th principal component; a i This represents the i-th weight coefficient; p i This represents the image of the i-th band; n represents the number of bands in the original image.

[0068] The band ratio (BR) algorithm can effectively reduce the impact of uneven light reflection caused by non-uniform surfaces, and also enhance the spectral differences between bands, providing unique information that cannot be obtained from a single band. Its principle is to obtain a relative band intensity image by dividing two bands.

[0069] BW (i,j,r) = BW (i,j,m) / BW (i,j,n)

[0070] BW (i,j,r) For the corresponding pixels in the image under different bands ( i , j The ratio of ) BW (i,j,m) and BW (i,j,n) For pixels at the same position in bands m and n ( i , j The grayscale value of ).

[0071] Step 2: Construction of algorithms for measuring and matching human rhythm data

[0072] This study aimed to comprehensively collect human biorhythm data under laboratory conditions, employing medical physiological data acquisition, sleep monitoring, infrared reflectance imaging, and subjective questionnaire measurements to establish an effective correlation between human biorhythm data and monitoring methods. The system calculated and analyzed the characteristics of human biorhythm fluctuations, and then, combined with hyperspectral facial image recognition and analysis, matched the NIR hyperspectral face image set with the physiological dataset, clarifying a non-intrusive biorhythm monitoring method that combines short-range and long-range approaches. Specific research content includes:

[0073] Measurement and Correlation Analysis of Physiological Information Data

[0074] Experimental research was conducted on a group of subjects who met the physical examination requirements. The laboratory was used to simulate an indoor living environment, focusing on creating highly similar acoustic, optical, electrical, and thermal physical environmental characteristics. The sleep stage was the primary monitoring target, and a phased, long-term experiment was carried out throughout the entire sleep-wake cycle. Three different monitoring methods were mainly employed.

[0075] A. Physiological data collection using medical measurement equipment. Real-time monitoring and collection of indicators such as EEG (electroencephalogram), ECG (electrocardiogram), EMG (electromyography), SpO2 (oxygen saturation), SKT (skin temperature), and PPG (pulse) of the subjects were conducted, and the indicator data throughout the entire experimental process were stored to form a physiological measurement dataset;

[0076] B. Data collection was conducted using an integrated smart mattress. The focus was on real-time measurement and recording of the subjects' heart rate, respiratory rate, and body movement during sleep on the mattress. A sleep measurement dataset was generated after the experiment.

[0077] C. Infrared Thermometry Data Acquisition. A low-frame-rate, high-sensitivity (<30mK) infrared thermometer was used to measure the temperature of the human head and face. The surface temperature of several positioning points was recorded in real time during sleep, and a facial temperature dataset was generated after the experiment.

[0078] Subjective measurements were conducted on the participants before and after the experiment. Questionnaires were used to collect subjective feelings oriented towards environmental perception and cognition. Participants also completed the Pittsburgh Sleep Quality Index (PSQI), the Brief Mood State Scale (POMS), and the Self-Rating Fatigue Scale (PRFS) (these scales were combined into a single comprehensive information collection form) to obtain pre- and post-experimental personal evaluation data. SPSS software was then used to statistically analyze the questionnaire data.

[0079] Establish the correlation between the data of indicators A, B, and C, that is, the correlation between physiological indicators over time under the same experimental conditions.

[0080] Based on the common indicators in the three datasets A, B, and C, and using the physiological measurement dataset as the standard, the measurement validity of the sleep measurement dataset and the facial temperature dataset is verified. By integrating all datasets, the correlation between the physiological measurement dataset, the sleep measurement dataset, and the facial temperature dataset is established, and then a complete dataset integrating all indicators of the three datasets is established.

[0081] Data matching based on image recognition and machine learning

[0082] In the above process, infrared reflectance of faces at different wavelengths (hyperspectral data) is used to obtain a series of face image sets under infrared conditions. The SPA (Successive Projection Algorithm) is used to select the feature wavelengths of the images. SPA not only eliminates the collinearity between wavelength variables but also extracts feature wavelengths with minimal collinearity and redundancy, representing the spectral information of most images with less information. This method initially selects a wavelength and then iteratively selects the wavelength by calculating the projection onto unselected wavelengths, choosing the wavelength with the largest projection vector. This vector is then incorporated into the wavelength combination until the loop ends. GA-BP (Genetic Algorithm-Based Backpropagation) is used to verify the prediction results of the optimal wavelength modeling. Based on the traditional BP which only has an input layer, hidden layer, and output layer, GA optimizes the initial weights and thresholds of the BP neural network, enabling the model to obtain a higher prediction correlation coefficient. Finally, the NIR hyperspectral face image set is matched with the physiological dataset to obtain a model algorithm for estimating biological rhythms using non-interference continuous monitoring parameters.

[0083] Statistical methods for physiological indicators

[0084] All data were analyzed using SPSS 19.0. Normally distributed continuous data were described using the mean and standard deviation. Comparisons between two groups were performed using t-tests, comparisons among multiple groups were performed using one-way ANOVA, pairwise comparisons were performed using the SNK method, and comparisons between groups were performed using the χ² test. Logistic regression was used for multivariate analysis. A p-value < 0.05 was considered statistically significant.

[0085] Step 3: Optimizing the artificial lighting environment to match the daily rhythm

[0086] Light is the most important timing factor in biological rhythms, guiding them. The cycle of light variation on Earth is 24 hours, causing organisms to exhibit 24-hour rhythms in their physiology and behavior. At different times of day, the human eye perceives sunlight with varying intensities and colors. Therefore, optimizing artificial lighting in enclosed environments to achieve automatic changes in light intensity and color at different times of day helps maintain stable biological rhythms. Specific research content includes:

[0087] Packaging of solar-like, ultra-high color index LED white light devices

[0088] Compared to traditional light sources, white LED semiconductor light sources have advantages such as being all-solid-state, impact-resistant, having a long lifespan, small size, and fast response speed. The color rendering index (CRI) is one of the key performance parameters of LED lighting fixtures, used to characterize the light source's ability to render the colors of objects. Ultra-high CRI generally refers to an LED lamp with a CRI Ra > 95, while full-spectrum white LED lighting requires all special CRIs R1~R15 to be greater than 90m. Higher CRI allows for the representation of the true colors of objects, making the human eye perceive light more closely like natural light. A proposed solution is to use a blue LED chip to excite aluminate yellow-green phosphor + nitride red phosphor + oxynitride blue-green phosphor encapsulation scheme to achieve an ultra-high CRI. In addition to visible light, sunlight also includes near-infrared light. A broadband emitting near-infrared phosphor with a center wavelength of 900 nm is proposed for encapsulation to achieve a relatively complete solar-like spectrum.

[0089] Artificial lighting environment optimization to match daily rhythms

[0090] Simulating ambient lighting variations to reflect sunlight and optimizing ambient lighting quality for various work tasks are effective means of regulating circadian rhythms. The construction of an indoor lighting environment mainly includes several aspects: illuminance level, layout requirements (uniformity, projection direction, glare constraint), color temperature and light emission characteristics (color temperature, color rendering index), power consumption and efficiency requirements, and lighting system control requirements. Among these indicators, illuminance level, light source layout, and color temperature and light emission characteristics have a significant impact on human senses. Light source layout is closely integrated with the specific size, shape, and posture of the indoor space, and the installation position of the luminaires is fixed. Illuminance level and color temperature and light emission characteristics, on the other hand, are highly controllable. Three white LEDs with three color temperatures were selected to form a three-channel light source, and time-division dimming of the three channels was performed based on the pulse width modulation (PM) principle. At a certain lighting test space distance, the spectral distribution of each channel operating independently was first measured. Then, the theoretical illuminance, correlated color temperature, and circadian rhythm factor of the three-channel mixed light source at any ratio were obtained through dimming calculation methods. Finally, the actual illuminance, correlated color temperature, and circadian rhythm factor of the three-channel light source at different mixing ratios were measured. Then, the theoretically and practically calculated illuminance, correlated color temperature, and circadian rhythm factors are compared, and the errors are analyzed.

[0091] While illuminance is related to the lighting method, the size and shape of the indoor space, and the material properties of the illuminated surface, it mainly depends on the total lighting power. Therefore, experimental studies should be conducted to determine the appropriate illuminance levels for behaviors such as working, eating, maintenance, reading, and sleeping, based on the needs of different functional areas or work tasks, in order to explore illuminance indices that are conducive to the stability of biological rhythms in each behavioral pattern.

[0092] Color temperature and color rendering index (CRI) refer primarily to the apparent color of a light source. The choice of color temperature depends mainly on the desired environment: generally, warm-colored, low-color-temperature lighting with a higher red light component creates a more relaxed atmosphere, suitable for rest and living spaces; higher color temperatures can alleviate tension and uplift spirits, suitable for work lighting. The general CRI (Ra) of a light source reflects its color rendering characteristics; the lower the Ra, the lower the color rendering quality. For long-term work and living spaces, Ra is generally required to be no less than 80. With the development of LED lighting, the current CRI standard has reached over 95. Therefore, even with a high CRI, changes in color temperature can still affect people's physical and mental well-being, thus influencing their circadian rhythms.

[0093] Therefore, it is proposed to construct lighting environments for different behaviors in the laboratory, and to conduct comparative simulation experiments on lighting environments for behavioral modules such as work, dining, and sleep to form scenario lighting. Furthermore, simulations of transitions based on circadian rhythms will be conducted to guide the synchronous transition of people's biological rhythms.

[0094] It is understood that the present invention has been described through some embodiments, and those skilled in the art will recognize that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the invention. Furthermore, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of the present invention.

Claims

1. A non-intrusive monitoring method for effectively monitoring intrinsic biological rhythms, characterized in that, The work includes the following steps: Step 1: Combine NIR light source with hyperspectral imaging technology to reflect biological rhythms by interpreting near-infrared absorption grayscale images of the face; Step 2: Construction of algorithms for measuring and matching human rhythm data; Step 3: Optimization of artificial lighting environment to match the daily rhythm; Step 2 specifically includes the measurement and correlation analysis of physiological information data, data matching based on image recognition and machine learning, and statistical methods for physiological indicators; The measurement and correlation analysis of the physiological information data specifically involves: conducting experimental research on a group of subjects who meet the physical examination requirements; simulating an indoor living environment in a laboratory setting, focusing on creating highly similar acoustic, optical, electrical, and thermal physical environment characteristics; and focusing on the sleep stage as the key monitoring object. The experiment is conducted in stages and over a long period throughout the sleep-wake cycle. The measurement and correlation analysis of the physiological information data includes three different monitoring methods, which are detailed below: A. Physiological data are collected using medical measurement equipment. Real-time monitoring and collection of the subjects' EEG (electroencephalogram), ECG (electrocardiogram), EMG (electromyography), SpO2 (blood oxygen saturation), SKT (skin temperature), and PPG (pulse) indicators are performed and stored to form a physiological measurement dataset. B. Data collection was conducted using an integrated smart mattress, focusing on real-time measurement and recording of the subjects' heart rate, respiratory rate, and body movement cycle during sleep on the mattress. A sleep measurement dataset was generated after the experiment. C. Infrared temperature measurement data acquisition: An infrared thermometer was used to measure the temperature of the human head and face. The surface temperature of several positioning points was recorded in real time during sleep. A facial temperature dataset was generated after the experiment. Subjective measurements were taken of the participants before and after the experiment. Questionnaires were used to collect subjective feelings oriented towards environmental perception and cognition. Participants also completed the Pittsburgh Sleep Index, Brief Mood State Scale, and Self-Rating Fatigue Scale to obtain measurement data of personal evaluation before and after the experiment. SPSS software was used to perform statistical analysis on the questionnaire data. Establish the correlation between the data of indicators A, B, and C, that is, the correlation between physiological indicators over time under the same experimental conditions and the same physiological indicators. Based on the common indicators in the three datasets A, B, and C, and using the physiological measurement dataset as the standard, the measurement validity of the sleep measurement dataset and the facial temperature dataset is verified. By integrating all datasets, the correlation between the physiological measurement dataset, the sleep measurement dataset, and the facial temperature dataset is established, and then a complete dataset integrating all indicators of the three datasets is established.

2. The non-intrusive monitoring method for effectively monitoring intrinsic biological rhythms according to claim 1, characterized in that, When NIR light is shone on the face, firstly, the invisible NIR light will not affect normal sleep; secondly, the absorption of NIR by different areas of the face will fluctuate due to changes in biological rhythms, which will be displayed as differences in grayscale in photos taken with an NIR camera.

3. The non-intrusive monitoring method for effectively monitoring intrinsic biological rhythms according to claim 1, characterized in that, The first step also includes a packaging method and an image processing method for a broadband NIR-LED light source; the packaging of the broadband NIR-LED light source draws on mature white LED technology. Image processing methods include principal component analysis and band ratio (BR) algorithm. Principal component analysis is a commonly used method for dimensionality reduction and noise reduction, which mainly represents a large amount of information with a small number of independent variables through linear combination. The principle is as follows: ; In the formula, PC k Represents the image of the k-th principal component; a i This represents the i-th weight coefficient; p i This represents the image of the i-th band; n represents the number of bands in the original image.

4. The non-intrusive monitoring method for effectively monitoring intrinsic biological rhythms according to claim 3, characterized in that, The band ratio (BR) algorithm obtains a relative band intensity image by dividing two bands. The principle is as follows: BW (i,j,r) = BW (i,j,m) / BW (i,j,n) In the formula: BW (i,j,r) For the corresponding pixels in the image under different bands ( i , j The ratio of ) BW (i,j,m) and BW (i,j,n) For pixels at the same position in bands m and n ( i , j The grayscale value of ).

5. The non-intrusive monitoring method for effectively monitoring intrinsic biological rhythms according to claim 1, characterized in that, The second step involves: constructing a comprehensive collection of human biorhythm data under laboratory conditions, using medical physiological data collection, sleep monitoring, infrared reflectance images, and subjective questionnaire measurements to obtain the effective correlation between human biorhythm data and monitoring methods; systematically calculating and analyzing the characteristics of human biorhythm fluctuations, and then combining hyperspectral facial image recognition and analysis to complete the matching of NIR hyperspectral face image sets and physiological datasets, clarifying a non-intrusive rhythm monitoring method that combines near and long range.

6. The non-intrusive monitoring method for effectively monitoring intrinsic biological rhythms according to claim 1, characterized in that, During the above process, the infrared reflectance of the face at different wavelengths is used to obtain a series of face image sets under infrared conditions. The SPA (Successive Projection Algorithm) is used. This method selects a wavelength in the initial state and then proceeds forward in a cyclical manner. The maximum wavelength of the projection vector is selected by calculating the projection on the unselected wavelengths, and then the vector is introduced into the wavelength combination until the loop ends. The prediction results of optimal wavelength modeling were tested using GA-BP (Genetic Algorithm-Based Neural Network). Based on the traditional BP which only has an input layer, hidden layer and output layer, GA was used to optimize the initial weights and thresholds of the BP neural network, enabling the model to obtain a higher prediction correlation coefficient. Finally, the matching of the NIR hyperspectral face image set and the physiological dataset was completed, thereby obtaining a model algorithm for estimating biological rhythms by non-interference continuous monitoring parameters.

7. The non-intrusive monitoring method for effectively monitoring intrinsic biological rhythms according to claim 6, characterized in that, The statistical methods for the physiological indicators included the analysis of all data using SPSS 19.0; Normally distributed measurement data are described using the mean and standard deviation. Comparisons between two groups are performed using the t-test; comparisons among multiple groups are performed using one-way ANOVA; pairwise comparisons are performed using the SNK method; and comparisons between groups are performed using... χ 2 test; Logistic regression was used for multivariate analysis, and P < 0.05 was considered statistically significant.

8. The non-intrusive monitoring method for effectively monitoring intrinsic biological rhythms according to claim 1, characterized in that, The third step includes the packaging of solar-like, ultra-high color rendering index (CRI) LED white light devices and the optimization of artificial lighting environment to match the circadian rhythm; Recently, artificial lighting environment optimization that matches the circadian rhythm specifically includes: Three white LEDs of three different color temperatures were selected to form a three-channel light source. Based on the principle of pulse width modulation (PM), the three channels were dimmed in a time-division manner. At a certain lighting test space distance, the spectral distribution of each channel working independently was first measured. The theoretical illuminance, correlated color temperature, and rhythm factor of the three-channel mixed light source at any ratio were obtained through dimming calculation methods. Then, the actual illuminance, correlated color temperature, and rhythm factor of the three-channel light source at different mixing ratios were measured. The theoretical and actual calculated illuminance, correlated color temperature, and rhythm factor were then compared, and the error was analyzed. In the laboratory, lighting environments for different behaviors were constructed, and simulation and comparative experiments were conducted on the lighting environments for work, dining, and sleep behavior modules to form scenario lighting; and conversion simulations were carried out according to the daily rhythm to guide the synchronous conversion of people's biological rhythms.

Images