Venipuncture augmented reality navigation system based on near-infrared imaging

Abstract

The invention discloses a venipuncture augmented reality navigation system based on near-infrared imaging, and relates to the technical field of medical diagnosis, and the venipuncture augmented reality navigation system comprises the following steps: collecting a light intensity time sequence signal, and collecting a motion signal representing the physical motion of the venipuncture augmented reality navigation system; emitting a light pulse to the target blood vessel region; and performing frequency domain transformation on the light intensity time sequence signal, and completing display in a target blood vessel region through an augmented reality display module. According to the method, the multi-modal quantitative indexes including physiological, physical and spatial dimensions are constructed and visually presented through augmented reality, the objectivity and success rate of venipuncture decision making are remarkably improved, and hardware cost does not need to be increased.

Description

Technical Field

[0001] This invention relates to the field of medical diagnostic technology, and in particular to an augmented reality navigation system for vein puncture based on near-infrared imaging. Background Technology

[0002] In current venous puncture-assisted techniques, near-infrared imaging is widely used to visualize subcutaneous blood vessels. By detecting the difference in absorption of near-infrared light by deoxyhemoglobin, a clear image of blood vessel distribution is generated, providing anatomical location reference for medical staff. However, in order to obtain stable, noise-free, high-quality images, existing technologies generally use signal processing methods such as smoothing, filtering, and frame averaging to actively suppress various fluctuations in the light signal. As a result, the weak light intensity changes caused by the cardiac cycle, which reflect vascular physiological activity, are completely filtered out as noise.

[0003] Currently, while this technology improves the static quality of images, it sacrifices the ability to assess the functional status of blood vessels, resulting in a single dimension of information. The system can only show where the blood vessels are, but cannot determine whether the blood vessels are usable. Due to the lack of objective quantification of key physiological and physical properties such as vascular elasticity, blood flow filling, and mechanical stability, clinical decisions still rely on the operator's subjective experience. Therefore, how to change the signal processing approach and extract and decode vascular function information from the traditionally discarded dynamic signals without increasing hardware costs has become a core problem that urgently needs to be solved. Summary of the Invention

[0004] In view of the problems existing in the current augmented reality navigation system for vein puncture based on near-infrared imaging, this invention is proposed.

[0005] Therefore, the problem to be solved by this invention is: how to extract and quantify multidimensional information characterizing vascular physiological function and physical properties from dynamic signals in near-infrared imaging that are filtered out as noise by traditional methods without increasing hardware complexity, so as to overcome the limitations of existing technologies that only provide anatomical location and lack objective diagnosis of vascular availability.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, embodiments of the present invention provide an augmented reality navigation system for vein puncture based on near-infrared imaging, comprising: an image acquisition module for acquiring time-series light intensity signals; a motion sensor for synchronously acquiring physical motion signals, acquiring motion signals characterizing the physical motion of the augmented reality navigation system for vein puncture; a near-infrared light source module for emitting light pulses toward a target blood vessel region; the image acquisition module, the motion sensor, and the near-infrared light source module constituting the target blood vessel region; a processor for performing frequency domain transformation on the time-series light intensity signals; the processor being connected to the image acquisition module and the motion sensor; and an augmented reality display module for displaying the image in the target blood vessel region.

[0007] As a preferred embodiment of the near-infrared imaging-based augmented reality navigation system for vein puncture described in this invention, the image acquisition module further includes switching the operating mode of the augmented reality navigation system for vein puncture in response to the selection of the target blood vessel region. The emission of light pulses toward the target blood vessel region includes the processor controlling the near-infrared light source module to emit a light pulse toward the target blood vessel region in depth detection mode; After the light pulse ends, with the light source turned off, the image acquisition module is controlled to acquire the near-infrared radiation signal that changes over time due to heat dissipation in the target blood vessel area. Based on the rate at which the intensity of near-infrared radiation signals decays over time, a subcutaneous depth index characterizing the target blood vessel is determined.

[0008] As a preferred embodiment of the near-infrared imaging-based augmented reality navigation system for vein puncture described in this invention, the processor includes a pulse analysis mode, a stability test mode, and a depth detection mode. The pulsation analysis mode is used to assess the intrinsic physiological function of blood vessels; The stability test mode is used to evaluate vascular external mechanical coupling; The depth detection mode is used to determine the spatial depth information of blood vessels; The frequency domain transformation of the light intensity time series signal includes identifying pulsating signal components and respiratory signal components based on the frequency domain transformation, and generating quantitative diagnostic indicators characterizing the functional state of the target blood vessel based on the presence and state of the pulsating signal components and respiratory signal components.

[0009] As a preferred embodiment of the near-infrared imaging-based augmented reality navigation system for vein puncture described in this invention, the pulsation analysis mode is used to determine a first parameter characterizing the intensity of vascular pulsation based on the normalized amplitude of the energy peak of the pulsation signal component in the first frequency band. The second parameter characterizing the stability of vascular pulsation is determined based on the full width at half maximum (FWHM) of the energy peak. Based on a mathematical model that takes the first and second parameters as input, quantitative diagnostic indicators are calculated and generated.

[0010] As a preferred embodiment of the near-infrared imaging-based augmented reality navigation system for venipuncture described in this invention, the stability test mode is used to take the physical displacement of the venipuncture augmented reality navigation system caused by the operator's physiological micro-vibrations, collected by the motion sensor, as an excitation signal. The image acquisition module tracks the positional changes of the target blood vessel on the sensor, using the image's tracking of the target blood vessel as a response signal; By calculating the cross-correlation coefficient between the excitation signal and the response signal, an index characterizing the mechanical coupling between the target blood vessel and the surrounding tissue is generated.

[0011] As a preferred embodiment of the near-infrared imaging-based augmented reality navigation system for vein puncture described in this invention, the processor is further configured to control the image acquisition module to acquire a baseline signal characterizing the baseline thermal radiation state of the target blood vessel region before controlling the near-infrared light source module to emit light pulses, and to subtract the baseline signal from the near-infrared radiation signal to perform calibration on the near-infrared radiation signal when determining the depth index. Determining a depth index characterizing the subcutaneous depth of a target blood vessel involves using a processor to calculate and solve for the depth index. The specific calculation formula is as follows: ; in, As a depth indicator, The intensity decay time constant is calculated based on the calibrated near-infrared radiation signal. The baseline signal strength value. The heat dissipation coefficient is... This is a pre-stored calibration function used to compensate for measurement offsets caused by different baseline thermal radiation states.

[0012] As a preferred embodiment of the near-infrared imaging-based augmented reality navigation system for vein puncture described in this invention, the augmented reality display module includes a processor-controlled augmented reality display module that overlays and displays the values ​​of quantitative diagnostic indicators onto a real-time image of the target blood vessel region. Based on the values ​​of quantitative diagnostic indicators, the image of the target vascular region is rendered as one of the gradient colors from green to red; If a stability quantification diagnostic index is generated, the augmented reality display module will display an anchor icon next to the image of the target blood vessel region.

[0013] Secondly, embodiments of the present invention provide an augmented reality navigation method for vein puncture based on near-infrared imaging, comprising: acquiring a time-series signal of light intensity; simultaneously acquiring a physical motion signal, acquiring a motion signal characterizing the physical motion of the augmented reality navigation system for vein puncture; emitting a light pulse toward the target blood vessel region; performing a frequency domain transformation on the time-series signal of light intensity; and displaying the signal in the target blood vessel region through an augmented reality display module.

[0014] Thirdly, embodiments of the present invention provide a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement any of the steps of the above-described near-infrared imaging-based augmented reality navigation system for vein puncture.

[0015] Fourthly, embodiments of the present invention provide a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any of the steps of the above-described near-infrared imaging-based augmented reality navigation system for vein puncture.

[0016] The beneficial effects of this invention are as follows: This invention uses an innovative signal processing method to convert the dynamic fluctuations that are filtered out in traditional near-infrared imaging into physiological information. It uses synchronous acquisition and frequency domain analysis to extract vascular pulsation features, and reuses hardware to achieve non-invasive assessment of mechanical stability and subcutaneous depth. Finally, it constructs a multimodal quantitative index that includes physiological, physical and spatial dimensions, which is presented intuitively through augmented reality. This significantly improves the objectivity and success rate of venipuncture decisions without increasing hardware costs. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0018] Figure 1 This is a schematic diagram of the multimodal collaborative analysis functional architecture of a near-infrared imaging-based augmented reality navigation system for vein puncture, provided as an embodiment of the present invention.

[0019] Figure 2 This invention provides a schematic diagram illustrating the relationship between blood vessel depth and thermal radiation signal attenuation rate in a near-infrared imaging-based augmented reality navigation system for vein puncture.

[0020] Figure 3 This is a flowchart illustrating the calculation of vascular vitality index in the pulsation analysis mode of a near-infrared imaging-based augmented reality navigation system for vein puncture, as provided in an embodiment of the present invention.

[0021] Figure 4 This is a schematic diagram of the core hardware and data storage structure of a near-infrared imaging-based augmented reality navigation system for vein puncture, provided as an embodiment of the present invention. Detailed Implementation

[0022] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] This invention is described in detail with reference to the schematic diagrams. When describing the embodiments of this invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not according to the usual scale. Furthermore, the schematic diagrams are merely examples and should not be construed as limiting the scope of protection of this invention. In actual fabrication, the three-dimensional spatial dimensions of length, width, and depth should be included.

[0026] Furthermore, in the description of this invention, it should be noted that the terms "upper," "lower," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used solely for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. In addition, the terms "first," "second," or "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0027] Unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" in this invention should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; similarly, they can refer to mechanical connections, electrical connections, or direct connections, or indirect connections through an intermediate medium, or internal connections between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0028] Example 1 This is the first embodiment of the present invention, which provides an augmented reality navigation system for vein puncture based on near-infrared imaging, comprising: The image acquisition module is used to acquire time-series signals of light intensity; Motion sensors are used to synchronously acquire physical motion signals, and to acquire motion signals that characterize the physical motion of the augmented reality navigation system for venipuncture. Near-infrared light source module, used to emit light pulses towards the target blood vessel area; A processor used to perform frequency domain transformation on time-series light intensity signals; Augmented reality display module, used to display the target blood vessel area.

[0029] Furthermore, the image acquisition module is used to acquire light intensity time-series signals. The image acquisition module also includes switching the operating mode of the vein puncture augmented reality navigation system in response to the selection of the target blood vessel region.

[0030] Further, a motion sensor is used to synchronously acquire physical motion signals, acquiring motion signals that characterize the physical motion of the intravenous puncture augmented reality navigation system. The motion sensor is a three-axis accelerometer, and the processor is used to vector sum the acceleration signals output by the three-axis accelerometer on the three orthogonal axes to generate motion signals.

[0031] Further, a near-infrared light source module is used to emit light pulses towards the target blood vessel region. The near-infrared light source module emits near-infrared light with a wavelength in the range of 750 to 950 nanometers in a continuous low-power manner.

[0032] Further, the processor is used to perform frequency domain transformation on the light intensity time series signal. The frequency domain transformation of the light intensity time series signal includes identifying pulsation signal components and respiratory signal components based on the frequency domain transformation, and generating quantitative diagnostic indicators characterizing the functional state of the target blood vessel based on the presence and state of the pulsation signal components and respiratory signal components.

[0033] Furthermore, the motion signal is transformed in the frequency domain to identify motion artifact components in a second frequency band outside the first frequency band, and a correction operation is performed on the pulsating signal components based on the motion artifact components. When the signal-to-noise ratio of the corrected pulsating signal components is lower than a signal-to-noise ratio threshold, respiratory signal components related to the respiratory cycle are identified in a third frequency band different from the first frequency band in the frequency domain transformation result of the light intensity time series signal, and quantitative diagnostic indicators characterizing the functional state of the target blood vessels are generated based on the presence and state of the pulsating signal components and respiratory signal components.

[0034] Preferably, the first frequency band is 1 Hz to 2 Hz, the second frequency band includes a frequency band below 1 Hz and a frequency band of 6 Hz to 12 Hz, and the third frequency band is 0.1 Hz to 0.5 Hz.

[0035] Furthermore, in pulsation analysis mode, the system's primary task is to ensure the faithful acquisition of weak physiological signals. Given that the light intensity fluctuations caused by changes in vascular volume driven by the cardiac cycle are extremely weak and have a high frequency, the system will stop all frame averaging operations aimed at smoothing the signal and instruct the image acquisition module to continuously acquire high-speed data only for the user-selected target vascular region at a preset high time resolution, i.e., a sampling rate of no less than 100 Hz, forming a raw light intensity time-series signal I(t). Simultaneously, to address the engineering problem of interference to weak pulsation signals caused by unavoidable physical movements (such as body swaying or physiological micro-vibrations) when the operator holds the device, a built-in triaxial accelerometer, acting as a motion sensor, will acquire a motion signal characterizing the device's physical movement in three-dimensional space, completely synchronously with the image acquisition module. Specifically, the processor will record the acceleration signals output by the triaxial accelerometer along the three orthogonal axes at the same sampling rate. , , And through vector summation operations: ; In this process, a total motion signal time series is generated. Thus, the processor obtains two parallel raw data streams that are precisely aligned in time: the light intensity time series signal and the motion signal. This dual-channel synchronous acquisition mechanism provides a data foundation for subsequent separation of physiological signals and motion artifacts through cross-validation.

[0036] After obtaining the two synchronized time-series signals, the processor does not directly analyze these seemingly noisy waveforms in the time domain. Instead, it performs a crucial perspective transformation: using the Fast Fourier Transform (FFT) algorithm, it transforms the light intensity time-series signal and the motion signal to the frequency domain to obtain their spectrograms. In the frequency domain, signal components from different sources converge into energy peaks in different frequency bands according to their frequency characteristics, thus becoming clearly distinguishable. According to physiological knowledge, the fundamental frequency of the pulsating signal caused by heartbeats usually falls within the first frequency band of 1 Hz to 2 Hz, while the frequency of the signal caused by respiratory movements is even lower, located at 0 Hz. Within the third frequency band, from 0.1 Hz to 0.5 Hz, artifacts introduced by operator body movement are typically low-frequency components below 1 Hz, while physiological micro-vibrations of the hand manifest as energy peaks from 6 Hz to 12 Hz. These two components together constitute the second frequency band. To accommodate differences in physiological frequencies among individuals, the definition of the first frequency band in the signal processing flow of the pulsation analysis mode includes a dynamic adaptive step. Specifically, the processor first performs a broadband pre-scan on the initial spectrum of the light intensity time series signal, identifying the center frequency corresponding to the main energy peak within the extended frequency range of 0.7 Hz to 2.8 Hz. Subsequently, the first frequency band used for refined analysis and feature parameter extraction is defined as having this center frequency. A dynamic interval with a center of symmetry and a width of 1 Hz, namely .

[0037] The definition of the third frequency band is also performed in this way, enabling the signal component identification process to automatically align with the target's actual physiological frequency. Therefore, the processor first searches the first frequency band (1 Hz to 2 Hz) in the light intensity signal spectrum to identify pulsating signal components related to the cardiac cycle. Simultaneously, it identifies motion artifact components in the second frequency band within the motion signal spectrum. A correction operation is then performed. A specific correction procedure is as follows: the processor subtracts the spectral energy of the motion signal at the corresponding frequency point from the spectral energy of the light intensity signal point by point, thus obtaining a relatively pure pulsating signal spectrum corrected for motion artifacts. Furthermore, the system sets a motion artifact energy threshold. If the energy peak detected by the processor in the motion signal spectrum exceeds this threshold, it is determined that the current shaking is too severe and the data is unreliable. At this point, the generation of subsequent indicators is paused, and a prompt signal to remain stable is output to the user through the augmented reality display module, thereby ensuring the reliability of the final diagnostic indicators.

[0038] Based on the corrected pulsation signal components, the system further extracts and quantifies their characteristic parameters to generate indicators characterizing vascular function. To transform abstract spectral features into clinically interpretable parameters, the system defines two core dimensions: pulsation intensity and pulsation stability. Pulsation intensity, designed to quantify the amplitude of vascular contraction and relaxation during the cardiac cycle, is engineeringly defined as the normalized amplitude of the main energy peak within the first frequency band, i.e., the first parameter. A well-elastic and well-flowing vessel has a higher pulsation intensity parameter value, and vice versa. Pulsation stability, designed to quantify the regularity of the pulsation rhythm, is defined as the full width at half maximum (FWHM) of the energy peak. The second parameter is that a healthy blood vessel under a stable heart rhythm will produce a sharp peak with a small full width at half maximum (FWHM), while an irregular heart rhythm or poor vascular response will cause the peak to widen. Finally, the processor calculates and generates a dimensionless quantitative diagnostic index, namely the vascular vitality index (VVI), based on a mathematical model stored in memory that takes the first and second parameters as inputs, such as a weighted summation function VVI=w1*Parameter1+w2*(1 / Parameter2). The generation of this index transforms the weak signal fluctuations that are filtered out as noise in traditional imaging into an objective dimension of physiological information.

[0039] Furthermore, in the stability test mode, to ensure that the calculation of the stability index converges to the time-series correlation characterizing the mechanical coupling degree rather than the energy amplitude of the operator's physiological micro-tremors, the processor performs a time-series analysis on the synchronously acquired excitation signal, i.e., the physical displacement of the system, before calculating the cross-correlation coefficient. The response signal, i.e., the time series of image position changes of the target blood vessel Perform interval mapping normalization once for each of the two time series, that is, perform interval mapping normalization on the amplitude of each of the two time series using the function: ; Linear mapping is performed to the normalized interval [-1, 1]. Subsequently, the processor uses only the two signal sequences after this normalization process as inputs to the cross-correlation function, so that the final stability index mainly reflects the synchronization degree of the two sequences in phase. In addition to physiological and mechanical properties, the subcutaneous depth of blood vessels is another important spatial dimension that determines the puncture strategy. To achieve non-invasive depth detection, the system also includes a depth detection mode, which again reflects the idea of ​​reusing the functions of existing hardware. In this mode, the working mode of the near-infrared light source connected to the processor changes. It no longer provides continuous illumination, but instead, under the precise control of the processor, directs light towards the target blood vessel. The region emits a light pulse with a defined energy and duration (e.g., 100 milliseconds, 50 milliwatts). Immediately after the light pulse ends, the light source is completely switched off. At this point, the role of the image acquisition module changes from a reflected light imager to a passive micro-radiation thermal imager. For the next 1-2 seconds, it continuously acquires near-infrared radiation signals from the target area at a high frame rate, generated by heat dissipation due to the slight heating. Since blood is an excellent heat conductor, the presence of subcutaneous blood vessels significantly accelerates surface heat dissipation. Therefore, the processor analyzes the rate of decay of the acquired near-infrared radiation signal over time (e.g., calculating its decay time constant). A depth index is determined to characterize the subcutaneous depth of the target blood vessel. Specifically, a fast decay rate corresponds to superficial blood vessels, while a slow rate corresponds to deep blood vessels.

[0040] To further improve the accuracy of depth detection in variable clinical environments, the system integrates a baseline thermal radiation self-calibration process when executing the aforementioned heat dissipation-based detection principle. This process aims to address measurement drift caused by factors such as individual patient body temperature or differences in ambient temperature. The specific procedure is as follows: In the instant before the near-infrared light source module emits a light pulse, the processor first controls the image acquisition module to acquire a baseline signal that is extremely short in duration (e.g., 30 milliseconds) representing the current baseline thermal radiation state of the target blood vessel region. This baseline signal serves two purposes: firstly, it calibrates subsequently acquired near-infrared radiation signals. For example, when determining depth parameters, the processor subtracts the intensity value of this baseline signal from the original near-infrared radiation signal to eliminate biases in the background thermal radiation. Secondly, this baseline signal can also be used for feedforward control of the detection process; that is, the processor can adjust the baseline signal based on the following parameters: The intensity of the light pulse is dynamically adjusted to control its energy or duration, ensuring a consistent thermal response signal with a uniform signal-to-noise ratio at different base skin temperatures. In the final depth calculation stage, the processor determines the depth index D by solving a pre-calibrated relational expression, for example: ; Where D represents the depth indicator. The intensity decay time constant is calculated based on the calibrated near-infrared radiation signal. The baseline signal strength value. The heat dissipation coefficient is calibrated using the standard organization model, and A pre-stored calibration function is used to compensate for measurement offsets caused by different baseline thermal radiation states. This pre-detection calibration and post-detection correction design ensures that depth detection results maintain consistency and reliability in diverse real-world application environments. Finally, all quantitative diagnostic indicators (such as vascular vitality index), stability indicators, and depth indicators generated through the above pattern analysis are transmitted by the processor to the augmented reality display module for integrated presentation. The processor controls the display module to overlay the values ​​of these indicators onto the image of the target vascular region in real time. Furthermore, based on the values ​​of the quantitative diagnostic indicators, the image of the target vascular region is rendered as one of a gradient colors from green (high vitality) to red (low vitality).

[0041] If a stability index is generated simultaneously, an anchor icon that is positively correlated with the index value will be displayed next to the vascular image. A clear icon represents high stability, while a blurry one represents low stability. In this way, the system transforms abstract diagnostic information from multiple dimensions into an intuitive and easy-to-understand visual navigation interface, providing comprehensive decision support for clinical operators.

[0042] Example 2 This embodiment, based on Embodiment 1, provides an example of applying the system of the present invention for multi-dimensional vascular assessment and puncture decision support in a clinical setting, including: In a geriatric oncology ward, a patient undergoing periodic chemotherapy needed to establish an intravenous infusion line. Due to long-term medication use, the patient's peripheral blood vessels were in a complex condition, exhibiting hardening and decreased elasticity. The operator used the navigation system of this invention for pre-puncture assessment. In conventional imaging mode, the system clearly displayed a cephalic vein (A) that appeared robust in shape and superficially located on the patient's forearm, and another, visually thinner, cephalic vein (B) nearby. Faced with this situation, the operator first locked the system's region of interest (ROI) onto the most clearly visible vessel, A, and triggered the system to switch to pulsation analysis mode. The system then acquired the time-series light intensity signal of this region at a sampling rate of 120 Hz, and simultaneously recorded the motion signal from a triaxial accelerometer. After performing a Fast Fourier Transform on the two signals, the processor first corrected the light intensity signal spectrum for motion artifacts based on the energy distribution of the motion signal spectrum within the 6 Hz to 12 Hz frequency band. In the corrected spectrum, the processor... No effective energy peak with a signal-to-noise ratio higher than 3dB was found in the first frequency band from 1 Hz to 2 Hz. This result indicates that there is no regular pulsation related to the cardiac cycle in this vascular segment. Based on this, the system generated a vascular vitality index (VVI) of only 12, and the image of blood vessel A was rendered as red in the augmented reality display interface. Subsequently, the operator moved the region of interest to blood vessel B, which seemed to have slightly worse conditions, and repeated the pulsation analysis process. After performing the same acquisition and correction processing on the signal of blood vessel B, the system identified a sharp energy peak with a center frequency of 1.2 Hz and a signal-to-noise ratio of 8dB in the first frequency band of its spectrum. Based on the normalized amplitude and full width at half maximum (FWHM) of this energy peak, the processor calculated that the vascular vitality index (VVI) of blood vessel B was 87, and its image was rendered as green accordingly. This result provides physiological functional information contrary to the initial visual image, that is, the visually thinner blood vessel B has a much better internal hemodynamic state than the morphologically robust blood vessel A. This solves the limitation that the value of vascular puncture cannot be evaluated based solely on static anatomical images.

[0043] After confirming that vessel B possesses excellent physiological function, to further assess the physical feasibility of its puncture, the operator switched the system to stability test mode. In this mode, the processor uses the physiological micro-vibrations of the operator's hand (frequency band concentrated around 9 Hz) recorded by the triaxial accelerometer as the excitation signal, and simultaneously records the positional changes of the vessel B image on the sensor as the response signal using a centroid tracking algorithm. By calculating the cross-correlation coefficient of these two time-series signals, the processor obtains a value of 0.92, which is converted into a stability index (SI) of 92. Simultaneously, a clearly defined anchor icon is overlaid next to the vessel B image in the augmented reality display interface. This stability index, combined with the vascular vitality index obtained in the previous step, provides synergistic verification, confirming from an orthogonal physical dimension—namely, mechanical coupling—that this physiologically active vessel is also... Firmly attached to surrounding tissues and resistant to rolling, this system provides a puncture decision-making basis that incorporates both physiological and physical attributes. Ultimately, based on the system's multi-dimensional navigation information—a vascular vitality index of 87 (shown in green) and a stability index of 92 (clear anchor icon)—the operator abandoned vessel A, which initially appeared visually superior but had low measured vitality, and instead chose to puncture vessel B. The catheter was successfully placed on the first attempt, and the subsequent infusion process proceeded smoothly. This application example demonstrates that by multiplexing motion signals used to suppress artifacts in different modes as excitation signals for assessing mechanical stability, and by converting the weak light intensity fluctuations filtered out in traditional imaging into pulsating signals for assessing physiological function, this system can resolve the information asymmetry between what is seen in images and what is physiological within a unified hardware architecture, providing decision support for clinical operations that transcends the anatomical dimension.

[0044] Example 3 This embodiment, based on Embodiment 1, provides a comparative experiment that objectively verifies the blood vessel recognition capability of the system under both physiological function and mechanical stability dimensions by constructing a standardized tissue model matrix, thus providing a decision-making basis. The experiment includes: To assess the effectiveness of differentiating blood vessels with varying physiological and physical characteristics, a comparative experiment based on tissue models was conducted. The experimental platform employed a test matrix consisting of two types of tissue models, A and B. Type A tissue models simulated differences in vascular physiological function, while Type B models simulated differences in vascular mechanical stability. All tissue models were made of silicone gel with translucency similar to human tissue, and each model contained pre-fabricated channels with a depth of 3 mm and an inner diameter of 2.5 mm to simulate subcutaneous blood vessels. Specifically, Type A tissue models included type A1 (high-vitality simulation) and type A2 (low-vitality simulation), with type A1... The channels of the model are connected to a pulsed fluid pump that generates a frequency of 1.2 Hz and a pressure fluctuation of 15% to simulate the hemodynamic pulsation of healthy blood vessels. The channels of the A2 model are connected to a pressure-stabilized pump that provides only continuous laminar flow to simulate hardened or non-pulsating blood vessels. The B-type tissue models include B1 (high stability simulation) and B2 (low stability simulation). The channels of the B1 model are tightly solidified in a silica gel matrix, while the channels of the B2 model are encapsulated in a layer of low-viscosity gel, allowing them to undergo lateral displacement when subjected to minor external perturbations to simulate easily rolling blood vessels.

[0045] The experiment included two control groups and one experimental group based on this invention. Control group one used equipment with only conventional near-infrared imaging capabilities, performing only multi-frame averaging for noise reduction without any dynamic signal analysis. Control group two used equipment with conventional imaging capabilities, but only enabled the pulsation analysis mode to generate the vascular vitality index (VVI), while its stability test mode was disabled. The experimental group used a system with all the functions of this invention, simultaneously enabling both the pulsation analysis mode and the stability test mode, generating both the VVI and the stability index (SI). During the experiment, the equipment from the three groups sequentially measured four tissue models (A1B1, A1B2, A2B1, and A2B2), with each model measured 10 times. The system output indicators were recorded, and the operator provided guidance based on navigation information. A puncture feasibility assessment: Experimental data showed that the device in control group 1 presented clear and stable vascular images for all four tissue models, making it impossible to visually distinguish their physiological or physical characteristics; therefore, the puncture feasibility assessment for all four models was deemed feasible. In pulsation analysis mode, the device in control group 2 showed an average vascular viability index (VVI) of 85.4 with a standard deviation of 3.1 for type A1 models (A1B1 and A1B2) and an average VVI of 11.8 with a standard deviation of 2.5 for type A2 models (A2B1 and A2B2), demonstrating the ability to distinguish physiological functional states. However, it also failed to differentiate the mechanical stability differences between type B1 and type B2 models. The test results of the experimental group of this invention are shown in Table 1, which presents the test data of the system of this invention for different tissue models. Table 1. Test data of the system of the present invention on different tissue models.

[0046] As shown in Table 1, the experimental group of this invention not only distinguished between high-vitality models (A1B1, A1B2) and low-vitality models (A2B1, A2B2) using the vascular viability index (VVI) values ​​(the distinction logic is consistent with control group 2), but also effectively differentiated between high-stability models (A1B1, A2B1) and low-stability models (A1B2, A2B2) by introducing the stability index (SI). Specifically, for the high-stability model, the average value of the stability index (SI) measured by the system is greater than 90, while for the low-stability model, the average value of the SI is less than 25. This difference in values ​​is directly due to the system's built-in cross-correlation analysis algorithm. In the high-stability model, the design stability caused by operator micro-vibrations is significantly improved. The displacement (excitation signal) and the vascular image displacement (response signal) are highly synchronized, with a cross-correlation coefficient approaching 1. However, in the low-stability model, the motion of the two is out of phase, resulting in a significant decrease in the cross-correlation coefficient. Experimental results show that the system of this invention, through the coordinated operation of the pulsation analysis mode and the stability test mode, can simultaneously quantify the target vessel from both physiological function and mechanical stability dimensions. Compared with technical solutions that only provide static anatomical images or single functional indicators, this system can provide a decision-making basis, thereby effectively identifying complex situations where, although physiologically sound, puncture is not suitable due to mechanical instability, or although mechanically stable, puncture is not suitable due to insufficient physiological vitality, thus improving the reliability of the evaluation results.

[0047] Example 4 This embodiment is based on embodiment 1, and incorporates the appendix. Figures 1 to 4 Examples illustrating the overall system architecture, functional connections between modules, and multimodal analysis process of this invention include: A description of an augmented reality navigation system for vein puncture based on near-infrared imaging, such as... Figure 1 As shown, it takes the target vascular region as the processing object and uses a parallel image acquisition module, motion sensor and near-infrared light source module to acquire light intensity time series signals, physical motion signals and provide the light or light pulses required for detection. All the acquired raw data are fed into a core processor, which performs multimodal collaborative analysis by switching between multiple logical modes. Among them, the pulsation analysis mode is used to evaluate the intrinsic physiological function of the blood vessel, the stability test mode is used to evaluate the extrinsic mechanical coupling of the blood vessel, and the depth detection mode is used to determine the spatial depth information of the blood vessel. Finally, the processor integrates the multidimensional diagnostic information generated after analysis and submits it to the augmented reality display module for integrated visualization.

[0048] like Figure 2As shown, the horizontal axis represents time in milliseconds (ms), and the vertical axis represents normalized radiation intensity. The figure shows four signal attenuation curves corresponding to depths of 2.0 mm, 3.0 mm, 4.0 mm, and 5.0 mm, respectively. It can be clearly observed from the figure that the subcutaneous depth of the blood vessel is inversely proportional to the attenuation rate of its thermal radiation signal intensity. That is, the shallower the blood vessel, such as the 2.0 mm depth curve, the faster its signal attenuation rate, while the deeper the blood vessel, such as the 5.0 mm depth curve, the slower its signal attenuation rate.

[0049] like Figure 3 As shown, the light intensity time-series signal from the image acquisition module and the motion signal from the motion sensor are subjected to frequency domain transformation on the basis of time synchronization to obtain the light intensity spectrum and motion artifact spectrum. Subsequently, the system performs motion artifact correction and uses the motion artifact spectrum to correct the light intensity spectrum, thereby obtaining a corrected pulsation signal spectrum. Based on this pure spectrum, the system further extracts the first parameter characterizing the pulsation intensity and the second parameter characterizing the pulsation stability, and inputs these two parameters into a pre-stored mathematical model to finally calculate and generate the quantified vascular vitality index VVI.

[0050] like Figure 4 As shown, the host contains a central processing unit (CPU) as the computing core, on which system firmware and analysis software run. It is connected to a non-volatile memory, which stores a calibration reference database for depth calibration and a mathematical model of vascular vitality index for vitality assessment. The processor is connected to various functional components for data acquisition and human-computer interaction, including a CMOS sensor image acquisition module for capturing vascular images, a three-axis accelerometer motion sensor for synchronously recording physical motion, a near-infrared light source module for active detection, and an augmented reality display module for information output.

[0051] Example 5 This embodiment, based on Embodiment 1, provides an example of the calibration mechanism of the system of the present invention in depth detection mode, ensuring the measurement accuracy of spatial depth information, consistency between devices, and mass production repeatability in the multimodal evaluation system, including: Before leaving the factory, the system needs to undergo a standardized offline calibration procedure to determine the key parameters in its depth calculation model. This procedure establishes a reference standard that includes known physical depth and known baseline thermal radiation state to inversely solve and solidify the unique thermal response conversion function of each device, thereby eliminating measurement deviations introduced by hardware differences and ambient temperature variations. This calibration procedure is carried out in a temperature and humidity controlled optical calibration darkroom, and its core equipment is a tissue model array and the navigation system to be calibrated in this invention. The tissue model array consists of 8 independent silicone gel modules with the same optical and thermal properties. The structure consists of a pre-fabricated channel with an inner diameter of 3 mm in each module. The center depth of this channel is precisely set from 2.0 mm to 5.5 mm in 0.5 mm increments, thus forming a depth gradient reference system. The entire tissue model array is placed on a semiconductor cooling heating plate with adjustable surface temperature. The temperature control accuracy of this heating plate is 0.1 degrees Celsius, which can stably set the surface temperature of the tissue model at a series of preset calibration points, namely 30 degrees Celsius, 33 degrees Celsius, 36 degrees Celsius, and 39 degrees Celsius, to simulate different baseline thermal radiation states.

[0052] At the start of the calibration process, the surface temperature of the semiconductor cooling heating plate is first set to 30 degrees Celsius, the first calibration point. After the surface temperatures of all tissue models stabilize, the navigation system to be calibrated is placed sequentially on tissue models with a depth of 2.0 mm, and its depth detection mode is triggered. Following a predetermined procedure, the system first acquires a baseline signal representing the current baseline thermal radiation state using the image acquisition module, with the light source off, to obtain the intensity value. Subsequently, the near-infrared light source module was controlled to emit a light pulse with an energy of 50 milliwatts and a duration of 100 milliseconds. After the light pulse ended, the attenuation curve of the thermal radiation signal in the following 1.5 seconds was recorded. By performing exponential fitting on the curve, the intensity attenuation time constant was calculated. This measurement process was repeated 20 times on the same tissue model to obtain ( , The data pairs were then used; subsequently, while keeping the heating plate temperature constant, the navigation system was moved sequentially to seven other tissue models at different depths, and the above measurement process was repeated to obtain eight sets of data corresponding to eight known depth values ​​at a baseline temperature of 30 degrees Celsius. , The dataset; after completing a full-depth scan of a temperature point, the surface temperature of the semiconductor cooling heating plate is sequentially set to 33, 36, and 39 degrees Celsius, and the aforementioned complete measurement process is repeated. Ultimately, the processor will obtain a raw calibration database containing 640 data points across 4 temperature levels, 8 depths per temperature level, and 20 repeated measurements per depth. Each record in the database contains data at a known physical depth. Measured decay time constant and measured baseline signal strength The database contains 3D data points; based on this database, the processor solves for the model parameters and calculates the model according to the depth: ; First, at each constant baseline temperature, for and A linear regression analysis was performed on the relationship, and the average slope obtained was determined as the heat dissipation coefficient k of the device. Subsequently, the obtained coefficient k was substituted into the model to calculate the thermal depth component of each data point. and from the corresponding true depth Subtracting this component from the middle yields a residual term. Finally, the residual terms for all data points Its corresponding baseline signal strength By performing polynomial fitting, a calibration function is obtained to compensate for measurement offsets caused by different baseline thermal radiation states. .

[0053] The final output of this calibration process is a heat dissipation coefficient k specific to the device hardware and a calibration function stored in the form of a lookup table or polynomial coefficients. These two parameters will be permanently written into the device's non-volatile memory; after this calibration, the device will be able to perform baseline signal measurements in real-time during practical applications. and decay time constant By using a defined internal model, the system can output a depth index D that is calibrated for both hardware individual differences and ambient temperature, thus ensuring the reproducibility of the entire technical solution in engineering and its consistency after mass production.

[0054] Example 6 This embodiment, based on Embodiment 1, provides an example of constructing and validating a mathematical model based on clinical data to convert parameters extracted under pulsation analysis mode into a standardized vascular vitality index (VVI), including: To establish a mathematical model that converts the first and second parameters extracted from the pulsation analysis mode into a standardized vascular vitality index (VVI), the system underwent a model construction and calibration procedure based on clinical data during the research and development phase. This procedure aims to determine a statistically significant and universally applicable quantitative conversion relationship by correlating the objective physical signal characteristics measured by the instrument with the vascular function status assessed clinically, thereby ensuring that the final output index has clear clinical relevance. The procedure first constructed an initial database by collecting data from 500 volunteers of different ages and health conditions. During the data collection process, the system of this invention was used to measure the heads of both arms of each volunteer. Measurements were taken of the cephalic vein, basilic vein, and median cubital vein. The first parameter, the normalized amplitude of the pulsation signal component, and the second parameter, the full width at half maximum (FWHM), were recorded for each vessel under motion correction in pulsation analysis mode. Simultaneously, without knowledge of the system's measurement results, the filling, elasticity, and blood flow signal of the same vessel segment were independently assessed using traditional palpation combined with a portable Doppler ultrasound, providing a comprehensive functional score from 1 to 10 as a reference standard for the vessel's functional status. Thus, each valid record in the database contains a set of input features, namely the first and second parameters, and a corresponding output target, namely the comprehensive clinical functional score.

[0055] Based on this database, a mathematical model was constructed using multiple linear regression analysis. The reciprocals of the first and second parameters were used as independent variables, and the clinical comprehensive functional score was used as the dependent variable. The least squares method was used for fitting, and the weight coefficients of each independent variable and the constant term of the model were determined, thus establishing a model of the form... The initial regression model was then used; subsequently, to map the model output values ​​to the interval of 0 to 100, the values ​​calculated using this model were applied to all samples in the database. The values ​​were statistically analyzed to determine the maximum and minimum values ​​of their distribution, and a linear normalization function was established based on this. The initial regression model was combined with this normalization function to form the final mathematical model of the Vascular Vitality Index (VVI) embedded in the system processor. Before being put into use, the model was validated using a reserved test dataset that was not used for model training. The correlation coefficient between the output VVI value and the clinical comprehensive functional score reached more than 0.9, which met the preset engineering requirements.

[0056] Example 7 This embodiment, based on Embodiment 1, provides an example of the system of the present invention recognizing vascular function status under complex physiological and noisy environments, including: To determine the optimal signal-to-noise ratio (SNR) threshold for the system when switching between cardiac and respiratory cycle signals, an offline optimization parameter-finding procedure was executed during the system's development phase. The objective function of this procedure is to maximize the Youden exponent in distinguishing valid cardiac signals from noise; that is, to seek a decision threshold that achieves the best balance between sensitivity and specificity, thereby maximizing the reliability of the diagnostic logic switching when facing critical signals. This procedure first constructs a standard signal library for model training and validation, which includes multiple 5-second near-infrared light intensity time series signals. The samples were divided into four categories: the first category consisted of positive cardiac signals, collected from healthy volunteers, whose cardiac cycle signal-to-noise ratios covered different levels from high to low; the second category consisted of positive respiratory signals, collected from volunteers who underwent temporary compression of the upper arm to simulate peripheral hypoperfusion, whose cardiac cycle spectral peaks were missing, but had clear harmonics caused by respiration in the 0.1 Hz to 0.5 Hz frequency band; the third category consisted of negative signals, collected from tissue models without avascular distribution, which did not contain any physiological rhythm information; and the fourth category consisted of noise signals, which introduced different types of physical motion artifacts during the acquisition process.

[0057] After the parameter search process is initiated, the processor sequentially executes the Fast Fourier Transform algorithm on each sample in the signal library and calculates the signal-to-noise ratio (SNR) of the maximum energy peak in the cardiac cycle frequency band, i.e., from 1 Hz to 2 Hz. Subsequently, within a preset SNR threshold scanning range, i.e., from 0.5 dB to 10 dB, the system traverses and searches in steps of 0.1 dB. At each candidate threshold point, the system uses this value as a decision threshold to perform a classification test on all samples in the signal library. Based on the test results and the true labels of the samples, the system calculates the true positive rate (sensitivity) and false positive rate (1 - specificity) at the current threshold. Then, the system calculates the objective function value corresponding to the threshold using the formula (Yorden index = true positive rate + specificity - 1). By traversing all candidate threshold points, the SNR value that makes the Youden index reach its maximum value is finally determined and fixed as the optimal SNR threshold in the system software, serving as the basis for switching diagnostic logic in subsequent actual operation.

[0058] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An augmented reality navigation system for vein puncture based on near-infrared imaging, characterized in that: include, The image acquisition module is used to acquire time-series signals of light intensity; Motion sensors are used to synchronously acquire physical motion signals, and to acquire motion signals that characterize the physical motion of the augmented reality navigation system for venipuncture. Near-infrared light source module, used to emit light pulses towards the target blood vessel area; The image acquisition module, the motion sensor, and the near-infrared light source module constitute the target blood vessel region; A processor used to perform frequency domain transformation on time-series light intensity signals; The processor is connected to the image acquisition module and the motion sensor; Augmented reality display module, used to display the target blood vessel area.

2. The augmented reality navigation system for vein puncture based on near-infrared imaging as described in claim 1, characterized in that: The image acquisition module also includes switching the operating mode of the vein puncture augmented reality navigation system in response to the selection of the target blood vessel region; The emission of light pulses toward the target blood vessel region includes the processor controlling the near-infrared light source module to emit a light pulse toward the target blood vessel region in depth detection mode; After the light pulse ends, with the light source turned off, the image acquisition module is controlled to acquire the near-infrared radiation signal that changes over time due to heat dissipation in the target blood vessel area. Based on the rate at which the intensity of near-infrared radiation signals decays over time, a subcutaneous depth index characterizing the target blood vessel is determined.

3. The augmented reality navigation system for vein puncture based on near-infrared imaging as described in claim 2, characterized in that: The processor includes a pulse analysis mode, a stability test mode, and a depth detection mode; The pulsation analysis mode is used to assess the intrinsic physiological function of blood vessels; The stability test mode is used to evaluate vascular external mechanical coupling; The depth detection mode is used to determine the spatial depth information of blood vessels; The frequency domain transformation of the light intensity time series signal includes identifying pulsating signal components and respiratory signal components based on the frequency domain transformation, and generating quantitative diagnostic indicators characterizing the functional state of the target blood vessel based on the presence and state of the pulsating signal components and respiratory signal components.

4. The augmented reality navigation system for vein puncture based on near-infrared imaging as described in claim 3, characterized in that: The pulsation analysis mode is used to determine a first parameter characterizing the intensity of vascular pulsation based on the normalized amplitude of the energy peak of the pulsation signal component in the first frequency band. The second parameter characterizing the stability of vascular pulsation is determined based on the full width at half maximum (FWHM) of the energy peak. Based on a mathematical model that takes the first and second parameters as input, quantitative diagnostic indicators are calculated and generated.

5. The augmented reality navigation system for vein puncture based on near-infrared imaging as described in claim 4, characterized in that: The stability test mode is used to take the physical displacement of the venous puncture augmented reality navigation system caused by the operator's physiological micro-vibrations, collected by the motion sensor, as an excitation signal. The image acquisition module tracks the positional changes of the target blood vessel on the sensor, using the image's tracking of the target blood vessel as a response signal; By calculating the cross-correlation coefficient between the excitation signal and the response signal, an index characterizing the mechanical coupling between the target blood vessel and the surrounding tissue is generated.

6. The augmented reality navigation system for vein puncture based on near-infrared imaging as described in claim 5, characterized in that: The processor is also used to control the image acquisition module to acquire a baseline signal characterizing the baseline thermal radiation state of the target blood vessel region before controlling the near-infrared light source module to emit light pulses, and to subtract the baseline signal from the near-infrared radiation signal when determining the depth index, and to perform calibration on the near-infrared radiation signal. Determining a depth index characterizing the subcutaneous depth of a target blood vessel involves using a processor to calculate and solve for the depth index. The specific calculation formula is as follows: ； in, As a depth indicator, The intensity decay time constant is calculated based on the calibrated near-infrared radiation signal. The baseline signal strength value. The heat dissipation coefficient is... This is a pre-stored calibration function used to compensate for measurement offsets caused by different baseline thermal radiation states.

7. The augmented reality navigation system for vein puncture based on near-infrared imaging as described in claim 6, characterized in that: The augmented reality display module includes a processor-controlled augmented reality display module that overlays the values ​​of quantitative diagnostic indicators onto a real-time image of the target vascular region. Based on the values ​​of quantitative diagnostic indicators, the image of the target vascular region is rendered as one of the gradient colors from green to red; If a stability quantification diagnostic index is generated, the augmented reality display module will display an anchor icon next to the image of the target blood vessel region.

8. A near-infrared imaging-based augmented reality navigation method for vein puncture, based on the near-infrared imaging-based augmented reality navigation system for vein puncture according to any one of claims 1 to 7, characterized in that: include, Acquire time-series signals of light intensity; Simultaneously acquire physical motion signals, and acquire motion signals that characterize the physical motion of the augmented reality navigation system for venipuncture; Emit light pulses towards the target blood vessel region; Perform frequency domain transformation on the time series signal of light intensity; The target blood vessel area is displayed using an augmented reality display module.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the near-infrared imaging-based augmented reality navigation system for vein puncture as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the near-infrared imaging-based augmented reality navigation system for vein puncture as described in any one of claims 1 to 7.

Images