A dual focus speckle composite physiological monitoring system, method and storage medium

Abstract

The application provides a physiological monitoring system and method based on double-focus speckle combination and a storage medium. The system comprises a structured light source for outputting monitoring structured light of multiple wavebands, an axial dispersion lens group comprising multiple dispersion components, a video acquisition element and a signal processing element. At least one dispersion component focuses first waveband monitoring structured light to a first target area of a target monitoring object to form a defocus speckle image, while at least one dispersion component focuses second waveband monitoring structured light to a second target area of the target monitoring object to form an in-focus speckle image. The video acquisition element acquires video information containing the defocus speckle image and the in-focus speckle image, and the video information is processed to obtain physiological monitoring data of the target monitoring object. The system and method provided by the application use a single imaging optical system to simultaneously obtain two types of speckle modulation information in one imaging process, thereby improving the efficiency of multi-modal physiological information acquisition.

Description

Technical Field

[0001] This invention relates to the field of physiological monitoring technology, and in particular to a dual-focus speckle composite physiological monitoring system, method, and storage medium. Background Technology

[0002] In existing technologies, when using imaging technology to perform physiological monitoring of target objects, it is often limited to a single sensing mechanism or detection mode.

[0003] Specifically, traditional imaging systems are typically designed with their light source wavelengths, detector types, and signal processing chains tailored to specific physiological parameters. This results in the system primarily capturing and extracting a single dominant physiological signal in a single measurement or imaging modality. This limitation of single measurement leads to a relatively limited dimension of physiological information obtained, resulting in low efficiency in acquiring physiological information. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the purpose of this invention is to provide users with a dual-focus speckle composite physiological monitoring system, method and storage medium, overcoming the defects of the single physiological information dimension in the physiological monitoring methods of the prior art, which leads to low information acquisition efficiency.

[0005] The technical solution adopted by this invention to solve the technical problem is as follows: In a first aspect, the present invention provides a dual-focus speckle composite physiological monitoring system, comprising: Structured light source, used to output multi-band monitoring structured light; An axial dispersion lens group includes multiple dispersive elements disposed in the optical path of the monitoring structured light. At least one dispersive element defocuses the first-band monitoring structured light onto a first target area of ​​the target monitored object, forming a defocused speckle image input to the imaging plane. Simultaneously, at least one dispersive element focuses the second-band monitoring structured light onto a second target area of ​​the target monitored object, forming a focused speckle image input to the imaging plane. The dispersive elements are lenses with axial dispersion characteristics. Each lens is arranged side-by-side in a vertical direction, with its optical axes parallel to each other and perpendicular to the principal optical axis of the physiological monitoring system. The first target area of ​​the target monitored object is the heart region, and the second target area is the palm skin region. The video acquisition element is used to acquire video information corresponding to the defocus speckle image and the focus speckle image on the imaging plane. A signal processing element, connected to the video acquisition element, is used to process the video information to obtain physiological monitoring data of the target monitoring object.

[0006] Optionally, the video acquisition element is a monochrome camera, used to capture video information containing both the focused speckle image and the defocused speckle image within a preset time period during a single imaging process.

[0007] Optionally, the bands corresponding to the first band monitoring structured light and the second band monitoring structured light are two different bands in the near-infrared band.

[0008] Optionally, when the camera defocuses, the formula for calculating the displacement of the interference speckle on the imaging plane is:

[0009] Where d represents the displacement of the interference speckle on the imaging plane. This represents the distance between the plane to which the focal plane has moved and the surface of the thoracic cavity. This represents the distance from the focal plane to the lens plane of the video acquisition element. These represent the distances from the imaging plane to the lens plane of the video acquisition element, respectively. This is a preset fixed value; It indicates the angle at which the chest cavity rotates due to the heartbeat.

[0010] Secondly, the present invention also provides a physiological monitoring method using bifocal speckle composite imaging, wherein the physiological monitoring method utilizes the bifocal speckle composite physiological imaging system as described above, and the physiological monitoring method includes: The surface vibration signal of the heart region of the target monitoring object and the microcirculation perfusion signal of the palm skin region of the target monitoring object were extracted from the video information. Physiological monitoring data of the target monitoring object are determined based on the surface vibration signal of the heart region and the microcirculation perfusion signal of the palm skin region.

[0011] Optionally, the step of extracting the surface vibration signal of the heart region of the target monitoring object from the video information includes: Each video frame in the video information is binarized, and circular structured light spots are detected in each binary image obtained by binarization, forming a circular structured light spot sequence. Extract the average displacement within the region of interest of each structured light spot in the circular structured light spot sequence; The surface vibration signal of the heart region of the target monitoring object is obtained by converting the average displacement of each region of interest.

[0012] Optionally, the video information includes multiple image sequences of structured light on the skin region of the target monitoring object; the step of extracting the microcirculation perfusion signal of the palm skin region of the target monitoring object from the video information includes: The region of interest is manually selected in the image sequence, and the pixels within the region of interest are spatially averaged to obtain the original time-series signal. The original time-series signal is digitally filtered. The DC component is extracted by low-pass filtering to characterize the baseline strength, and the AC component is separated by band-pass filtering to characterize the pulsation information. The ratio of the AC component to the DC component is calculated to obtain the microcirculation perfusion signal of the palm skin area of ​​the target monitoring object.

[0013] Optionally, the step of determining the physiological monitoring data of the target monitoring object based on the surface vibration signal of the heart region and the microcirculation perfusion signal of the palm skin region includes: Based on the surface vibration signal of the heart region, the heart rate value of the target monitoring object is extracted; Based on the microcirculation perfusion signal of the palm skin area, the blood flow velocity and perfusion volume of the target monitoring object are determined.

[0014] Thirdly, the present invention also provides a computer storage medium, which is a computer-readable storage medium, wherein a computer program is stored in the computer storage medium, and when the computer program is executed by a computer, the computer is used to perform the physiological monitoring method of bifocal speckle composite as described above.

[0015] Beneficial effects: This invention provides a dual-focus speckle composite physiological monitoring system, method, and storage medium. The system includes: a structured light source outputting multi-band monitoring structured light, an axial dispersive lens group, a video acquisition element, and a signal processing element. At least one dispersive element defocuses the first-band monitoring structured light onto a first target region of the monitored object, forming a defocused speckle image that is input to the imaging plane; simultaneously, at least one dispersive element focuses the second-band monitoring structured light onto a second target region of the monitored object, forming a focused speckle image that is input to the imaging plane; the video acquisition element acquires video information corresponding to the defocused and focused speckle images on the imaging plane, processes the video information to obtain physiological monitoring data of the monitored object. The system and method provided by this invention utilize a single imaging optical system to simultaneously acquire two types of speckle modulation information in a single imaging process, improving the efficiency of multimodal physiological information acquisition. Attached Figure Description

[0016] Figure 1 A schematic diagram of the structure of the dual-focus speckle composite physiological monitoring system provided by the present invention; Figure 2 This is a schematic diagram illustrating the principle of an application embodiment in this invention; Figure 3This is a schematic diagram illustrating the principle of monitoring cardiac region vibration using structural monitoring optical defocus imaging in the system provided by this invention. Figure 4 This is a schematic diagram illustrating the principle of monitoring skin pulse wave signals using speckle patterns in an embodiment of the present invention. Figure 5 The following is a flowchart of the physiological monitoring method using dual-focus speckle composite technique of the present invention; Figure 6a This is an example diagram of the structured light SCG and single-point laser SCG signal waveforms under experimental conditions of no coil coverage and soft coil coverage in the embodiments of the present invention; Figure 6b This is an example diagram of the spectrum of structured light SCG and single-point laser SCG signals under experimental conditions of no coil coverage and soft coil coverage in an embodiment of the present invention; Figure 7 This is a graph showing the average PI change of multiple target monitoring objects at different experimental stages in an embodiment of the present invention; Figure 8 In this embodiment of the invention, personalized and generalized models are used to estimate reference PI curves and calibration PI curves based on SL (Structured Light) and RGB cameras. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0018] Continuous, accurate, and long-term synchronous monitoring of circulatory perfusion and cardiac function is of great significance for early assessment and clinical intervention of cardiovascular health. Current physiological monitoring methods mainly rely on invasive or contact techniques, such as arterial catheters, echocardiography, and magnetic resonance imaging. While these methods can provide relatively accurate physiological parameters, they typically rely on large, specialized equipment and require professional operation, making long-term continuous monitoring difficult. Wearable devices can continuously collect parameters such as electrocardiogram (ECG), electromyography (EMG), respiration, and heart rate, but they usually require electrodes or sensors to be attached to multiple sites on the body, limiting their application in newborns, especially premature infants, or individuals with skin damage. In recent years, non-contact physiological monitoring technologies have gradually gained attention. For example, pulse wave imaging (rPPG) based on changes in reflected light intensity can reflect the perfusion status of peripheral circulation, but it is difficult to directly characterize cardiac mechanical activity; radio frequency radar or ultrasound-based methods can detect vibrations on the surface of the chest cavity, but their sensitivity to the direction of motion is limited (usually only sensitive to vertical motion) and they are easily affected by environmental interference; defocus speckle imaging technology has high sensitivity to minute mechanical vibrations on the body surface (sensitive to both vertical and rotational motion) and can be used to detect cardiac and respiratory micro-vibration signals in the chest wall.

[0019] However, existing technologies still have significant limitations in simultaneously acquiring cardiac mechanical motion and circulatory perfusion information. Defocused speckle imaging technology can be highly sensitive to minute mechanical vibrations on the body surface, and can be used to detect cardiac and respiratory micro-vibration signals in the chest wall. However, defocusing causes the imaging system to deviate from its focused state, thus weakening its ability to characterize changes in speckle intensity and making it difficult to accurately reflect changes in speckle intensity caused by changes in blood perfusion. Conversely, while a focused imaging system can capture speckle intensity changes better, it has lower sensitivity to micro-vibration displacements on the body surface.

[0020] Therefore, existing physiological monitoring methods cannot simultaneously achieve high-sensitivity detection of both mechanical motion signals and circulatory perfusion signals in the same imaging system, which limits the ability to synchronously acquire and couple multidimensional physiological information such as circulatory, respiratory, and cardiac mechanical activities.

[0021] To overcome the problems in the prior art, this application proposes a dual-focus speckle composite physiological monitoring system, method, and storage medium. This system and method utilize the optical chromatic aberration generated by the imaging optical system at different wavelengths. By adjusting the focal plane of the lens, speckle signals of different wavelengths are respectively placed in defocus and focused states within the same monocular camera imaging system, thereby simultaneously acquiring two types of speckle modulation information in a single imaging process. Specifically, the speckle channel in the defocus state has an optically amplified response to minute mechanical movements on the body surface, and can be used for highly sensitive extraction of chest wall vibration signals caused by the heart and respiration; the speckle channel in the focused state can maintain high speckle contrast and intensity stability, and can be used to characterize changes in skin microcirculation perfusion. By constructing a mapping relationship between defocused speckle displacement and surface mechanical motion, and combining the response characteristics of speckle intensity changes to blood perfusion, joint analysis and fusion modeling of signals from different channels can be performed. This enables the synchronous acquisition and decoupled analysis of multidimensional physiological information such as circulatory perfusion, respiratory activity, and cardiac mechanical motion, thereby achieving continuous, accurate, and non-contact imaging monitoring of the circulatory-respiratory-motor system and improving the system's comprehensive capabilities in acquiring multimodal physiological information.

[0022] The following description, in conjunction with the accompanying drawings, provides a more detailed account of the bifocal speckle composite physiological monitoring system, method, and storage medium disclosed in this embodiment.

[0023] This invention provides a dual-focus speckle composite physiological monitoring system, such as... Figure 1 As shown, it includes: The structured light source 100 is used to output multi-band monitoring structured light. The structured light source provided in this embodiment is a light source device capable of outputting structured light in multiple bands. The structured light emitted by this light source device can be visible light structured light, infrared structured light, short-wave infrared and ultraviolet structured light, or multi / hyperspectral structured light. In specific implementations, it can be a combination of multi-wavelength lasers, that is, combining multiple laser diodes of different wavelengths into the same optical path through a beam combiner; or it can be a tunable laser, which outputs laser light of different wavelengths at different times to achieve the output of multi-band monitoring structured light.

[0024] The axial dispersive lens group 200 includes multiple dispersive elements disposed in the optical path of the monitoring structure light. At least one dispersive element defocuses the first band monitoring structure light to the first target area of ​​the target monitoring object, forming a defocused speckle image that is input to the imaging plane 300. At the same time, at least one dispersive element focuses the second band monitoring structure light to the second target area of ​​the target monitoring object, forming a focused speckle image that is input to the imaging plane 300.

[0025] Because the axial dispersive lens group contains multiple vertically arranged dispersive elements, it can focus light of different wavelengths at different positions on the optical axis (i.e., the axial direction). Therefore, when monitoring structured light of different bands output by the structured light source is input to different dispersive elements, the light of different wavelengths has different effective focal lengths in the same optical system. This allows a certain wavelength to be in focus and a certain band to be out of focus on the imaging plane. As a result, two speckle imaging modes are formed simultaneously in the same optical system, and different processing information is obtained based on the different speckle imaging modes.

[0026] The video acquisition element 400 is used to acquire video information corresponding to the defocus speckle image and the focus speckle image on the imaging plane.

[0027] When structured light, modulated by the axial dispersive lens group, is input to the imaging plane, it forms both a focused speckle pattern and a defocused speckle pattern on the imaging plane. To simultaneously acquire these two patterns, this embodiment utilizes a video acquisition element to simultaneously capture both patterns, achieving synchronous acquisition of the focused and defocused speckle signals in a single acquisition. The video acquisition element includes a monochrome camera, an RGB camera, or a black-and-white camera, etc. Further, the video acquisition element is a monochrome camera, used to capture video information containing both the focused and defocused speckle images within a preset time period during a single imaging process.

[0028] The signal processing element 500 is connected to the video acquisition element and is used to process the video information to obtain the physiological monitoring data of the target monitoring object.

[0029] Once video information containing both focus speckle and defocus speckle signals is acquired, these two types of signals are processed by signal processing components to obtain analysis information for each type of signal.

[0030] In one embodiment, the dispersive element is a lens with axial dispersion characteristics; the lenses are arranged side by side in the vertical direction, the optical axes of the lenses are parallel to each other, and all are perpendicular to the main optical axis of the physiological monitoring system.

[0031] Due to the axial dispersion characteristics of lens materials, light of different wavelengths has different effective focal lengths in the same optical system. Therefore, on a fixed imaging sensor plane, a state of simultaneous focusing and defocusing naturally forms. For example... Figure 2 As shown, when one band is in focus, the other band is usually out of focus, deviating from the focal plane. Based on this physical characteristic, two near-infrared bands, 940 nm and 800 nm, can be selected for dual-band imaging, thus constructing a theoretical model for bifocal speckle composite physiological imaging, providing a mathematical basis for integrating two speckle imaging mechanisms in a single optical system.

[0032] Furthermore, the first band monitoring structured light and the second band monitoring structured light correspond to two different bands in the near-infrared band.

[0033] In specific implementation, the first target area of ​​the target monitoring object is the heart region, and the second target area is the palm skin region. This embodiment constructs a mapping relationship between the defocus speckle displacement and the surface micro-vibration displacement generated in the heart region, as well as the response characteristics of blood perfusion in the palm skin region to changes in speckle intensity. This enables the analysis of two different signals, simultaneously acquiring multi-dimensional physiological information such as circulatory perfusion, respiratory activity, and cardiac mechanical motion. This achieves continuous, accurate, and non-contact imaging monitoring of the circulatory, respiratory, and musculoskeletal systems, improving the performance of multimodal physiological information acquisition.

[0034] Combination Figure 2 and Figure 3 As shown, defocused speckle interferometry is used to monitor subtle vibrations on the chest surface caused by heartbeats. When a laser beam illuminates a rough surface (such as skin or fabric), diffuse reflection occurs. Interference between reflected beams of the same wavelength generates a speckle field in the space surrounding the target object. Any minute movement on the chest surface causes a corresponding movement of the speckle at the focal plane; therefore, the speckle displacement directly characterizes the chest surface movement caused by heartbeats. When the camera defocuses, the focal plane moves from the chest surface to a distance from the chest cavity. On the plane, the displacement (d) of the interference speckle along the imaging plane (which can be the imaging sensor plane in specific implementations) can directly reflect the vibration of the chest surface caused by the heartbeat, and is described by equation (1): (1) in, and These represent the distances from the focal plane and the imaging plane to the lens plane, respectively. It is usually a preset fixed value; This represents the angle of rotation of the chest cavity due to the heartbeat. Therefore, the speckle displacement d is proportional to the degree of defocus and can be expressed using a ratio. / To quantify.

[0035] On the other hand, skin speckle patterns acquired using the camera under focus adjustment, such as... Figure 4As shown, compared to non-skin areas, speckle patterns in skin areas exhibit a hazy halo that gradually transitions with radius outside the central high-contrast structure. This halo primarily originates from the diffuse reflection of deep tissue light formed after incident light enters the subcutaneous tissue and undergoes multiple scattering and refractions in different tissue layers. Non-skin areas, lacking this complex tissue scattering structure, do not exhibit similar spatial distribution characteristics. Therefore, in vivo skin speckle patterns possess a natural layered information characteristic in their spatial distribution, enabling the separation of optical information from different tissue layers: the inner ring region mainly reflects the scattering characteristics of superficial skin tissue, while the outer ring region corresponds to the diffuse reflection signal of deep tissue. Further analysis shows that the photon propagation path corresponding to the outer ring of the speckle pattern exhibits a "banana-shaped" subcutaneous propagation trajectory similar to contact pulse wave detection, possessing a greater tissue penetration depth. Therefore, it is highly sensitive to perfusion changes caused by arterial blood pulsation and can effectively preserve perfusion information related to arterial blood flow.

[0036] In detail, in this embodiment, based on the acquired video information, a region of interest (ROI) is first manually selected in the image sequence, and the pixels within the ROI are spatially averaged to obtain the original temporal signal (rPPG signal); the structured light data is converted into a grayscale signal. Subsequently, the original signal undergoes digital filtering. A low-pass filter (cutoff frequency 5Hz) is used to extract the direct current (DC) component to characterize the baseline intensity, and a band-pass filter (40-150bpm) is used to separate the alternating current (AC) component to characterize pulsation information. Based on this, a sliding time window (window length 30s, step size 1s) is used to calculate characteristic parameters, where the AC component is represented by the standard deviation of the band-pass signal, and the DC component is represented by the mean of the low-pass signal. The AC / DC ratio is then calculated to obtain the initial perfusion index (PI). Finally, a linear regression model was established using the least squares method, and the estimated PI was calibrated with the measurement value of the reference device. The model evaluation adopted two strategies: individual modeling (90% of the training set and 10% of the test set) and leave-one-out cross-validation (LOOCV), and the Pearson correlation coefficient and mean absolute error (MAE) were used as performance indicators.

[0037] To verify the feasibility of the device in this embodiment acquiring two different light spot patterns simultaneously based on the same optical path system, the imaging principle of the imaging system corresponding to the monitoring device in this embodiment will be further explained below.

[0038] In this embodiment, the imaging system is equivalent to a thin lens, and when the wavelength is λ, its focal length satisfies: (2) wherein, n(λ) is the refractive index of the lens material at wavelength λ, and are the curvature radii of the lens (constants). Dual-band imaging is performed in the near-infrared normal dispersion region (940nm and 800nm) to be adopted in this project: n( )>n( ), =800nm, =940nm; (3) Therefore, f(800)<f(940). A first-order expansion is performed on the focal length difference caused by the small dispersion difference Δn=n(800)-n(940): ; (4) This shows that in a single optical system without chromatic aberration correction, there must be differences in focal lengths corresponding to different wavelength bands, which is an objective result determined by the dispersion property of the material, and also constitutes the physical basis for the coexistence of two speckle imaging modes in the same imaging system.

[0039] Set the transverse magnification M (M=image distance / object distance)>0. For the wavelength band of λ, the focal length f, object distance u and image distance z satisfy 1 / f=1 / u+1 / z. Therefore, when , the image distance z(λ) satisfies: ; (5) Therefore, the image plane defocus difference between the two wavelength bands is: ; (6) This result shows that the focal length change caused by wavelength difference will be further converted into image distance difference during imaging, so that different wavelength bands correspond to different defocus amounts on the fixed sensor plane, thereby forming differentiated imaging states. Based on this characteristic, when the 940nm band is adjusted to the in-focus state (i.e., δ(940)≈0), the 800nm band will naturally be in a certain degree of defocus state.

[0040] Take dense flint optical glass SF11 as an example, the refractive index is calculated according to the Sellmeier equation: n(800)≈1.7847,n(940)≈1.7764,Δn=n(800)-n(940)≈0.0083; (7) Since the 800nm band has a higher refractive index, light has a stronger refraction degree in the lens, thereby generating a certain dual-band focal length difference. At this time, set the system parameters: focal length f=10mm, object distance 1000mm (i.e., 1.0 m), the reference focal length difference of the two bands is: ; (8) In the imaging space, due to the vertical magnification effect of the telephoto lens (magnification factor approximately...), Image plane defocus difference between the two bands It will be stretched further: (9) With the same system configuration, when the camera defocuses, the focal plane moves away from the object surface. At this point, the amplification of the speckle motion depends on the distance between the focal plane and the object. The distance between the focal plane and the lens Decision. When the magnification ratio ( / When the value is greater than 1, the minimum image plane defocus difference required between focusing and defocusing is: (10) at this time, Therefore, under these conditions, the image plane defocus difference between the two bands meets the defocus imaging requirement of magnification ratio > 1, thus forming a stable, controllable and engineering-feasible speckle defocus magnification effect.

[0041] This embodiment discloses how the focal length difference caused by optical dispersion is transformed into the image plane defocus difference under the imaging geometry relationship, and further forms the defocus condition that satisfies speckle motion amplification, providing a theoretical basis for the realization of a physiological monitoring device with dual-focus speckle composite.

[0042] Secondly, the present invention also provides a physiological monitoring method for bifocal speckle composite, such as... Figure 5 The physiological monitoring method, utilizing the bifocal speckle composite physiological imaging system as described above, includes: Step S1: Extract the surface vibration signal of the heart region of the target monitoring object and the microcirculation perfusion signal of the palm skin region of the target monitoring object from the video information.

[0043] In this step, the video information acquired by the camera of the monitoring device is used to extract signals, and the surface vibration signal of the heart region and the microcirculation perfusion signal of the palm skin region of the target monitoring object are extracted from the video information respectively.

[0044] On the one hand, the surface vibration signal of the heart region is extracted based on the video information acquired by the camera. The steps include: Step S111: Binarize each video frame in the video information, and detect circular structured light spots in each binary image obtained by binarization to form a circular structured light spot sequence.

[0045] After binarization, a black and white pattern is obtained. Circular structured light spots are detected in the binarized black and white pattern. The circular structured light spots are the defocused speckle patterns corresponding to the surface vibration signals. The circular structured light spots are then detected sequentially in each video frame in the video information according to the time sequence to obtain the graphic structured light spot sequence.

[0046] Step S112: Extract the average displacement of each structured light spot within its region of interest in the circular structured light spot sequence.

[0047] For each pixel within the Region of Interest (ROI), a displacement vector is calculated, and the average displacement magnitude within the ROI is calculated based on the displacement vector. This method of calculating the displacement vector for each pixel can utilize dense optical flow to obtain the displacement of each pixel within the ROI.

[0048] Step S113: Based on the average displacement of each region of interest, the surface vibration signal of the heart region of the target monitoring object is obtained.

[0049] On the other hand, the video information includes multiple image sequences of structured light on the skin region of the target monitoring object; the step of extracting the microcirculation perfusion signal of the palm skin region of the target monitoring object from the video information includes: Step S121: Manually select the region of interest in the image sequence and perform spatial averaging on the pixels within the region of interest to obtain the original time-series signal.

[0050] Step S122: Perform digital filtering on the original time-series signal, extract the DC component through low-pass filtering to characterize the baseline strength, and use band-pass filtering to separate the AC component to characterize the pulsation information.

[0051] Step S123: Calculate the ratio of the AC component to the DC component to obtain the microcirculation perfusion signal of the palm skin area of ​​the target monitoring object.

[0052] Step S2: Determine the physiological monitoring data of the target monitoring object based on the surface vibration signal of the heart region and the microcirculation perfusion signal of the palm skin region.

[0053] Specifically, the step of determining the physiological monitoring data of the target monitoring object based on the surface vibration signal of the heart region and the microcirculation perfusion signal of the palm skin region includes: Step S21: Extract the heart rate value of the target monitoring object based on the surface vibration signal of the heart region.

[0054] Step S22: Based on the microcirculation perfusion signal of the palm skin area, determine the blood flow velocity and perfusion volume of the target monitoring object.

[0055] This invention discloses a monitoring device and method that addresses the problem of existing non-contact physiological monitoring technologies' inability to simultaneously acquire cardiac mechanical motion and circulatory perfusion information. It proposes a bifocal speckle composite physiological monitoring device and method based on optical dispersion modulation. This device and method utilize the normal dispersion characteristics of lens materials to allow different wavelengths of light to correspond to different effective focal lengths on the same fixed imaging plane, thereby constructing a bifocal imaging structure with both focused and defocused states in a monocular imaging system. Through multi-band structured light illumination and speckle imaging mechanisms, the speckle in the focused state primarily characterizes the intensity changes caused by skin microcirculation perfusion, while the speckle in the defocused state amplifies the response to minute mechanical vibrations on the body surface. This allows for the simultaneous acquisition of cardiac and respiratory mechanical motion and circulatory perfusion information during the same imaging process.

[0056] Furthermore, this embodiment also discloses experimental verification of the feasibility of using defocused structured light imaging to monitor cardiac vibration and focused structured light imaging to characterize blood perfusion. Experimental results show that defocused structured light can effectively acquire cardiac vibration signals and accurately reflect the temporal characteristics of key cardiac events; while focused structured light can stably characterize the changing trend of skin microcirculation perfusion, thus preliminarily verifying the feasibility of the dual-focus speckle composite physiological imaging method.

[0057] Based on this, the present invention further establishes a complete theoretical model from dual-band focal length difference—camera image plane defocus difference—defocus speckle motion amplification condition, systematically revealing how the focal length difference caused by optical dispersion is transformed into image plane defocus difference under imaging geometry, and further forming defocus conditions that satisfy speckle motion amplification, providing a clear physical basis and engineering implementation path for the realization of dual-focus speckle composite physiological imaging system. Thirdly, the present invention also provides a computer storage medium, which is a computer-readable storage medium, wherein a computer program is stored in the computer storage medium, and when the computer program is executed by a computer, the computer is used to perform the physiological monitoring method of bifocal speckle composite as described above.

[0058] This invention provides a dual-focus speckle composite physiological monitoring system, method, and storage medium. The system includes: a structured light source outputting multi-band monitoring structured light, an axial dispersive lens group, a video acquisition element, and a signal processing element. At least one dispersive element defocuses the first-band monitoring structured light onto a first target region of the monitored object, forming a defocused speckle image that is input to the imaging plane; simultaneously, at least one dispersive element focuses the second-band monitoring structured light onto a second target region of the monitored object, forming a focused speckle image that is input to the imaging plane; the video acquisition element acquires video information corresponding to the defocused and focused speckle images on the imaging plane, processes the video information to obtain physiological monitoring data of the monitored object. The system and method provided by this invention utilize a single imaging optical system to simultaneously acquire two types of speckle modulation information in a single imaging process, improving the efficiency of multimodal physiological information acquisition.

[0059] To verify the performance of structured light SCG (SL-SCG (Seismocardiography)) and single-point laser SCG (Laser-SCG) in cardiac parameter measurement and key cardiac time detection, and to evaluate the feasibility of SL-SCG as an alternative to Laser-SCG for multi-point cardiac motion assessment, the experiment used 850nm structured light and 650nm single-point laser to illuminate the chest surface, and recorded video sequences at a sampling rate of 200Hz using an IDS high-speed camera (UI-3860CP-C-HQ).

[0060] For signal extraction from single-point laser video, a region of interest (ROI) is manually selected in the first frame and applied to all subsequent frames. Then, the Farneback optical flow algorithm is used to extract the raw SCG signal (Seismocardiography, SCG) from the camera video, denoted as Laser-SCG. For extracting SCG signals from structured light video (SL-SCG), each video frame is first binarized (threshold set to 0.4 times the maximum pixel intensity of the current frame). Then, a circular Hough transform (radius constraint: 25–80 pixels; detection sensitivity = 0.95; edge intensity threshold = 0.05) is used to detect circular structured light spots in the binary image. Simultaneously, circles located at image boundaries (whose center position plus radius exceeds the image resolution) and overlapping circles (the distance between the centers is less than the sum of their radii) with smaller radii are detected and removed. Finally, optical flow signals in the X and Y directions are extracted from each remaining circular region for subsequent automatic selection of the optimal SL-SCG signal. To suppress breathing components and high-frequency noise, Laser-SCG and SL-SCG signals were filtered using a fourth-order Butterworth bandpass filter with a passband of 0.5-40Hz. Subsequently, the quality of the SL-SCG signals at multiple points was quantized using a weighted average of peak prominence and waveform periodicity to select the signal with the best overall quality. For peak positions P={ , , ..., The prominence of the i-th peak (P(n)) of the signal x[n] (n=1,2,……,N) is given by the signal x[n]. The salience Q of the entire signal segment is defined as: (11) (12) in, and Each represents the search from the peak value to the left / right, up to the interval before the signal first exceeds x[Pk]. denoted as the standard deviation of the signal.

[0061] The waveform periodicity P is quantified by the maximum value of the normalized autocorrelation function within the heart rate range over a sliding window (window size 2s, step size 50ms), and the cross-window median is used as the global periodicity index to enhance robustness to noise and transient interference. The form is shown below: (13) Where median() represents the median, max() represents the maximum value, M represents the number of windows, and f represents the sampling rate of the signal. This represents the heart rate range within each window, ranging from 0.5f to 1.5f, corresponding to 40-120 BPM. Finally, the Matlab function (findpeaks) is used to identify the AO peaks in the Laser-SCG and SL-SCG signals, calculate the interval between consecutive peaks, and exclude erroneous intervals outside the heart rate range (waveform distortion caused by exercise).

[0062] like Figure 6a and Figure 6b The waveforms and spectral examples of SL-SCG and Laser-SCG signals obtained under conditions of no coil coverage and soft coil coverage are presented. Visualization results show that the two methods exhibit high consistency in the temporal location and frequency domain distribution of key feature points.

[0063] Furthermore, the heart rate intervals of 10 subjects under two experimental conditions were calculated. Under the no-coil coverage condition, the mean heart rate intervals obtained by SL-SCG and Laser-SCG were 898.95 ms and 895.37 ms, respectively; under the soft coil coverage condition, they were 887.64 ms and 888.33 ms, respectively. The difference between the two conditions was less than one sampling point (i.e., 4 ms), further validating the potential of SL-SCG as an alternative to Laser-SCG for multi-point cardiac motion assessment.

[0064] In addition, the AO peak positions were manually labeled and used as a reference to evaluate the localization accuracy of key biomarkers in the SL-SCG signal. A 50ms tolerance was set, meaning that AO feature points automatically detected within 50ms of the reference peak position were considered correctly detected. Table 1 shows the automatic AO event detection results for the two signals under the 50ms tolerance condition. The results show that under both experimental conditions, the precision, recall, and F1 score of SL-SCG are superior to Laser-SCG, and the advantage is more significant under the condition of no coil coverage. This indicates that SL-SCG can capture AO events more robustly and exhibit stronger robustness to motion interference (see Table 1). Figure 6a This advantage primarily stems from the use of structured light illumination and spatial information integration in SL-SCG, which allows for the selection of measurement points with optimal signal quality from multiple candidate locations, thereby reducing reliance on precise localization and system alignment. In summary, the results demonstrate that SL-SCG exhibits high consistency with Laser-SCG in the temporal representation of key cardiac events, validating the good fidelity of structured light SCG signals acquired in defocused states, and further proving its feasibility and application potential as an alternative to Laser-SCG for multi-point cardiac motion assessment.

[0065] Table 1 shows the AO event detection results within a 50-millisecond error tolerance range;

[0066] On the other hand, the skin speckle image acquired by the camera under focus adjustment, such as Figure 4 As shown, compared to non-skin areas, speckle patterns in skin areas exhibit a hazy halo that gradually transitions with radius outside the central high-contrast structure. This halo primarily originates from the diffuse reflection of deep tissue light formed after incident light enters the subcutaneous tissue and undergoes multiple scattering and refractions in different tissue layers. Non-skin areas, lacking this complex tissue scattering structure, do not exhibit similar spatial distribution characteristics. Therefore, in vivo skin speckle patterns possess a natural layered information characteristic in their spatial distribution, enabling the separation of optical information from different tissue layers: the inner ring region mainly reflects the scattering characteristics of superficial skin tissue, while the outer ring region corresponds to the diffuse reflection signal of deep tissue. Further analysis shows that the photon propagation path corresponding to the outer ring of the speckle pattern exhibits a "banana-shaped" subcutaneous propagation trajectory similar to contact pulse wave detection, possessing a greater tissue penetration depth. Therefore, it is highly sensitive to perfusion changes caused by arterial blood pulsation and can effectively preserve perfusion information related to arterial blood flow.

[0067] The feasibility of using structured light imaging (SL) to monitor blood perfusion was verified in a laboratory setting (compared to visible light imaging (RGB)). The experimental setup consisted of a dual-camera system mounted on a stable iron stand for simultaneous imaging of the subject's palm region. This dual-camera system included an RGB camera and a monochrome camera. The RGB camera (UI-3860CP-C-HQ) recorded color changes on the palm surface, while the monochrome camera (UI-3860CP-M-GL), incorporating a near-infrared bandpass filter to suppress visible light interference, acquired structured light speckle images at the 850 nm wavelength. During the experiment, the subject wore an inflatable cuff; periodic pressure and release altered the blood perfusion level in the palm region, inducing changes in the perfusion index (PI) to assess the responsiveness of structured light imaging to changes in blood perfusion.

[0068] Based on structured light and RGB video data, a region of interest (ROI) is first manually selected in the image sequence, and the pixels within the ROI are spatially averaged to obtain the original temporal signal (rPPG signal). The green channel is extracted from the RGB data, and the structured light data is converted to a grayscale signal. The original signal is then digitally filtered. A low-pass filter (cutoff frequency 5Hz) is used to extract the direct current (DC) component to characterize the baseline intensity, and a band-pass filter (40–150 bpm) is used to separate the alternating current (AC) component to characterize pulsation information. Based on this, a sliding time window (window length 30s, step size 1s) is used to calculate characteristic parameters. The AC component is represented by the standard deviation of the band-pass signal, and the DC component is represented by the mean of the low-pass signal. The AC / DC ratio is then calculated to obtain the initial perfusion index (PI). Finally, a linear regression model was established using the least squares method, and the estimated PI was calibrated with the measurement value of the reference device. The model evaluation adopted two strategies: individual modeling (90% of the training set and 10% of the test set) and leave-one-out cross-validation (LOOCV), and the Pearson correlation coefficient and mean absolute error (MAE) were used as performance indicators.

[0069] Figure 7 The average PI variation trend of 20 subjects across three experimental phases (baseline, compression, and recovery) is presented. The results show that, consistent with the reference data, both the RGB and SL imaging modalities can clearly capture the typical PI variation characteristics across the three phases. Specifically, PI significantly decreases during the compression phase and gradually recovers to baseline levels during the recovery phase, indicating that the applied cuff compression stimulation effectively modulates peripheral blood perfusion, thereby significantly expanding the dynamic range of PI variation. Furthermore, the PI curves estimated by both SL and RGB show a consistent trend with the reference curve, demonstrating that both visual imaging modalities can timely and stably track the dynamic changes in peripheral perfusion.

[0070] like Figure 8The paper further presents the comparison results between the PI curves estimated by SL and RGB cameras and the reference PI curves for 20 subjects under personalized and generalized model conditions. It can be observed that the PI curves obtained by different methods have a high degree of consistency with the reference curve in terms of overall trend. In each stage from ID01 to ID20, all methods can track the periodic fluctuations and overall trend of PI well, indicating that the PI estimation method based on structured light (SL) can effectively reflect the dynamic changes in blood perfusion. Quantitative analysis results further show that the PI values ​​obtained by SL and RGB modalities exhibit a strong linear correlation with the reference PI (Pearson correlation coefficients are both greater than 0.9). Under personalized calibration conditions, the mean absolute errors between the PI values ​​obtained by SL and RGB modalities and the reference values ​​are 0.53% and 0.59%, respectively; under generalized modeling conditions, the mean absolute errors are 1.98% and 2.00%, respectively. These results indicate that the PI estimation method based on structured light can achieve a performance level comparable to the traditional RGB method. In summary, the high consistency between the SL and RGB modes in PI estimation results further validates the feasibility and application potential of structured light imaging in non-contact peripheral perfusion monitoring.

[0071] Other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of the invention are indicated by the following claims.

[0072] The embodiments described above are merely examples of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application.

Claims

1. A dual-focus speckle composite physiological monitoring system, characterized in that, include: Structured light source, used to output monitoring structured light containing multiple wavelengths; An axial dispersion lens group includes multiple dispersive elements disposed in the optical path of the monitoring structured light. At least one dispersive element defocuses the first-band monitoring structured light onto a first target area of ​​the target monitoring object, forming a defocused speckle image input to the imaging plane. Simultaneously, at least one dispersive element focuses the second-band monitoring structured light onto a second target area of ​​the target monitoring object, forming a focused speckle image input to the imaging plane. The dispersive elements are lenses with axial dispersion characteristics. Each lens is arranged side-by-side in a vertical direction, with its optical axes parallel to each other and perpendicular to the principal optical axis of the physiological monitoring system. Due to the axial dispersion characteristics of the lens material, light of different wavelengths has different effective focal lengths in the same optical system, thus naturally forming an imaging state where focusing and defocusing coexist on the fixed imaging sensor plane. The first target area of ​​the target monitoring object is the heart area, and the second target area is the palm skin area; The video acquisition element is used to acquire video information corresponding to the defocus speckle image and the focus speckle image on the imaging plane. A signal processing element, connected to the video acquisition element, is used to process the video information to obtain physiological monitoring data of the target monitoring object; The video acquisition element is a monochrome camera, used to capture video information containing both the focused speckle image and the defocused speckle image within a preset time period during a single imaging process. The first band monitoring structured light and the second band monitoring structured light correspond to two different bands in the near-infrared band; the 940nm and 800nm ​​two near-infrared bands are selected for dual-band imaging, and then a theoretical model of dual-focus speckle composite physiological imaging is constructed. The formula for calculating the displacement of the interference speckle on the imaging plane is: Where d represents the displacement of the interference speckle on the imaging plane. This represents the distance between the plane to which the focal plane has moved and the surface of the thoracic cavity. This represents the distance from the focal plane to the lens plane of the video acquisition element. These represent the distances from the imaging plane to the lens plane of the video acquisition element, respectively. This is a preset fixed value; It indicates the angle at which the chest cavity rotates due to the heartbeat.

2. The dual-focus speckle composite physiological monitoring system according to claim 1, characterized in that, The structured light source is a combination of multi-wavelength lasers or a tunable laser.

3. A physiological monitoring method using a dual-focus speckle composite technique, characterized in that, The physiological monitoring method using the bifocal speckle composite physiological monitoring system as described in any one of claims 1-2 includes: The surface vibration signal of the heart region of the target monitoring object and the microcirculation perfusion signal of the palm skin region of the target monitoring object were extracted from the video information. Physiological monitoring data of the target monitoring object are determined based on the surface vibration signal of the heart region and the microcirculation perfusion signal of the palm skin region.

4. The physiological monitoring method using bifocal speckle composite according to claim 3, characterized in that, The step of extracting the surface vibration signal of the heart region of the target monitoring object from video information includes: Each video frame in the video information is binarized, and circular structured light spots are detected in each binary image obtained by binarization, forming a circular structured light spot sequence. Extract the average displacement within the region of interest of each structured light spot in the circular structured light spot sequence; The surface vibration signal of the heart region of the target monitoring object is obtained by converting the average displacement of each region of interest.

5. The physiological monitoring method using bifocal speckle composite according to claim 3, characterized in that, The video information includes multiple image sequences of structured light on the skin region of the target monitoring object; The steps for extracting the microcirculation perfusion signal of the palm skin region of the target monitoring object from video information include: The region of interest is manually selected in the image sequence, and the pixels within the region of interest are spatially averaged to obtain the original time-series signal. The original time-series signal is digitally filtered. The DC component is extracted by low-pass filtering to characterize the baseline strength, and the AC component is separated by band-pass filtering to characterize the pulsation information. The ratio of the AC component to the DC component is calculated to obtain the microcirculation perfusion signal of the palm skin area of ​​the target monitoring object.

6. The physiological monitoring method of dual-focus speckle composite according to claim 3, characterized in that, The step of determining the physiological monitoring data of the target monitoring object based on the surface vibration signal of the heart region and the microcirculation perfusion signal of the palm skin region includes: Based on the surface vibration signal of the heart region, the heart rate value of the target monitoring object is extracted; Based on the microcirculation perfusion signal of the palm skin area, the blood flow velocity and perfusion volume of the target monitoring object are determined.

7. A computer storage medium, wherein the storage medium is a computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a computer, enables the computer to perform the physiological monitoring method of bifocal speckle composite as described in any one of claims 3-6.

Images