Large depth-of-focus ultraviolet light acoustic microscopy imaging system and imaging method

Abstract

This invention relates to the field of optical microscopy imaging technology, specifically to a large depth-of-focus ultraviolet photoacoustic microscopy imaging system and method, comprising: an ultraviolet photoacoustic signal acquisition module to acquire spatiotemporal distribution data of ultraviolet excitation light and its corresponding photoacoustic response signal data; a spectral-polarization feature extraction module to extract quantitative multidimensional feature parameters related to tissue composition and structure; a cross-depth standard sample module to perform systematic imaging measurements on standard biological tissue samples at different depths to form a feature parameter dataset; and an image reconstruction module that, by linking the spatial distribution characteristics of the ultraviolet excitation light field with the physical propagation laws of photoacoustic signals at different depths, designs a hierarchical signal response adaptation link to match the signal transmission loss differences across tissue depths, and performs signal attenuation correction through a hierarchical dynamic neural radiation field algorithm to reconstruct a tissue microscopic image with large depth-of-focus characteristics. This solves the problems of generally short depths of focus and lack of standardized data support for cross-depth imaging in existing technologies.

Description

Technical Field

[0001] This invention relates to the field of optical microscopy imaging technology, specifically to a large depth-of-focus ultraviolet photoacoustic microscopy imaging system and imaging method. Background Technology

[0002] Ultraviolet (UV) photoacoustic microscopy, as an emerging label-free bio-optical imaging method, relies on pulsed UV laser irradiation of biological tissues to excite their endogenous components and generate specific photoacoustic signals, thereby enabling high-contrast, high-resolution imaging of subcellular structures such as the cell nucleus. Existing UV photoacoustic microscopy systems typically employ high numerical aperture microscope objectives to tightly focus UV light for excellent lateral resolution, demonstrating significant application potential in biomedical research, particularly in the rapid evaluation of label-free pathological sections, tumor boundary delineation, and observation of cellular metabolic activities. This technology effectively combines the high resolution of optical imaging with the high sensitivity of photoacoustic imaging to optical absorption, providing a powerful tool for life science research.

[0003] Existing photoacoustic microscopy imaging technology suffers from several key shortcomings, which are precisely the core issues addressed by ultraviolet photoacoustic microscopy imaging systems with large depth of focus: First, the depth of focus is generally short, and traditional systems can only maintain effective imaging resolution within a limited depth range. Image quality in defocused areas is severely degraded, failing to meet the requirements for cross-depth three-dimensional imaging of thick tissue samples and samples with uneven surfaces. Second, the ability to capture and analyze ultraviolet photoacoustic signals is insufficient, lacking precise means to acquire spatiotemporal distribution data of ultraviolet excitation light and corresponding photoacoustic response signals, making it difficult to extract quantitative multidimensional feature parameters related to tissue composition and structure. Third, there is a lack of standardized data support for cross-depth imaging, and no feature parameter datasets for standard biological tissue samples at different depths have been established, making it difficult to accurately control the depth attenuation law of photoacoustic signals. Fourth, image reconstruction algorithms are not fully adapted to the characteristics of the ultraviolet light field and the cross-depth signal propagation law, failing to dynamically match the differences in signal transmission loss between tissues at different depths, making it difficult to achieve real-time reconstruction of tissue microscopic images with large depth of focus and high fidelity. Summary of the Invention

[0004] This application provides a deep-focus ultraviolet light acoustic microscopy imaging system and imaging method to solve the problems of generally short depth of focus and lack of standardized data support for cross-depth imaging in the prior art.

[0005] The first aspect of this application provides a large depth-of-focus ultraviolet photoacoustic microscopy imaging system, comprising: an ultraviolet photoacoustic signal acquisition module, a spectral-polarization feature extraction module, a cross-depth standard sample module, and an image reconstruction module; wherein, the ultraviolet photoacoustic signal acquisition module is used to acquire spatiotemporal distribution data of ultraviolet excitation light and its corresponding photoacoustic response signal data; the spectral-polarization feature extraction module is used to extract quantitative multidimensional feature parameters related to tissue composition and structure through spectral matching and polarization photoacoustic joint analysis; the cross-depth standard sample module is used to perform systematic imaging measurements on standard biological tissue samples at different depths to form a feature parameter dataset; the image reconstruction module is used to link the spatial distribution characteristics of the ultraviolet excitation light field with the physical propagation law of photoacoustic signals at different depths, design a hierarchical signal response adaptation link, dynamically match the signal transmission loss differences of tissues at different depths, perform signal attenuation correction through a hierarchical dynamic neural radiation field algorithm, and reconstruct tissue microscopic images with large depth-of-focus characteristics in real time.

[0006] Preferably, the ultraviolet photoacoustic signal acquisition module includes an ultraviolet light emitting unit, a light field spatiotemporal detection unit, and a photoacoustic signal sensing unit. The ultraviolet light emitting unit is used to generate wavelength-tunable and power-stable ultraviolet excitation light and transmit it to the target tissue. The light field spatiotemporal detection unit is used to capture the spatial distribution and temporal evolution data of the ultraviolet excitation light in the tissue in real time. The photoacoustic signal sensing unit is used to directly acquire the original photoacoustic response signal generated after the tissue is stimulated.

[0007] Preferably, the spectral-polarization feature extraction module includes a spectral analysis unit, a polarization state modulation and detection unit, and a spectral-polarization joint analysis unit. The spectral analysis unit is used to perform spectral decomposition on the scattered light after the interaction between ultraviolet excitation light and tissue to obtain characteristic spectral information corresponding to tissue components. The polarization state modulation and detection unit is used to adjust the incident light polarization parameters and capture polarization state change data of tissue reflected / scattered light. The spectral-polarization joint analysis unit is used to fuse characteristic spectral and polarization state data to extract quantitative multidimensional feature parameters related to tissue component concentration and microstructural orientation.

[0008] Preferably, the cross-depth standard sample module includes a standard sample preparation and calibration unit, a layered imaging acquisition unit, and a parameter dataset integration unit. The standard sample preparation and calibration unit is used to construct standardized biological tissue samples covering different depths and known tissue components and structures, using quantitative multidimensional feature parameters as core indicators, and to calibrate and verify the acquisition parameters using known sample information. The layered imaging acquisition unit performs precise imaging and photoacoustic signal acquisition of different depth regions of the sample. The parameter dataset integration unit organizes the feature parameters calibrated at each depth according to a unified standard to create a standardized cross-depth feature parameter dataset.

[0009] Preferably, the image reconstruction module includes a light field-propagation law linkage analysis unit, a hierarchical signal response adaptation unit, an attenuation correction unit, and a large depth-of-focus microscopic image real-time reconstruction unit. The light field-propagation law linkage analysis unit extracts the spatial distribution characteristics of the ultraviolet excitation light field and, combined with a standardized cross-depth feature parameter dataset and the physical propagation laws of photoacoustic signals at different depths, establishes a cross-depth light field-photoacoustic propagation coupling correlation model. The hierarchical signal response adaptation unit constructs a dynamic adaptation link to accurately match loss characteristics based on differences in tissue signal transmission loss across depths. The attenuation correction unit adaptively corrects signal attenuation at different depths using a hierarchical dynamic neural radiation field algorithm. The large depth-of-focus microscopic image real-time reconstruction unit integrates the corrected signal with multi-dimensional feature parameters to output a tissue microscopic image with large depth-of-focus characteristics in real time.

[0010] Preferably, the hierarchical dynamic neural radiation field algorithm is as follows: ; ; ; in, To correct the attenuated photoacoustic signal intensity; Horizontal pixel coordinates; Vertical pixel coordinates; For organization depth coordinates; The original photoacoustic signal intensity; Let be the spatial distribution function of the ultraviolet excitation light field; It is a natural exponential function; For the first The starting depth of the layer; For the first Attenuation coefficient of layered photoacoustic signal; The coordinates of the infinitesimal element are in the depth direction; It is a tiny differential unit in the depth direction; A hierarchical dynamic neural radiation field model; For the first Layered multilayer perceptron; For feature encoding function; The pixel grayscale values ​​of the reconstructed image; This represents the total number of layers; These are the stratified weighting coefficients; For the first The center depth of the layer; It is a signal noise suppression factor; This is a noise suppression term.

[0011] The second aspect of this application provides a method for ultraviolet photoacoustic microscopy imaging with large depth of field, comprising: acquiring spatiotemporal distribution data of ultraviolet excitation light and its corresponding photoacoustic response signal data; based on the spatiotemporal distribution data of ultraviolet excitation light and its corresponding photoacoustic response signal data, extracting quantitative multidimensional feature parameters related to tissue composition and structure through spectral matching and polarization photoacoustic joint analysis; performing systematic imaging measurements on standard biological tissue samples at different depths according to the quantitative multidimensional feature parameters to form a feature parameter dataset; based on the feature parameter dataset, linking the spatial distribution characteristics of the ultraviolet excitation light field with the physical propagation law of photoacoustic signals at different depths, designing a hierarchical signal response adaptation link, dynamically matching the signal transmission loss differences across tissue depths, correcting signal attenuation through a hierarchical dynamic neural radiation field algorithm, and reconstructing a tissue microscopic image with large depth of field characteristics in real time.

[0012] A third aspect of this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the large depth-of-focus ultraviolet photoacoustic microscopy imaging method as described in the above embodiments.

[0013] A fourth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, which is executed by a processor to implement the large depth-of-focus ultraviolet photoacoustic microscopy imaging method as described in the above embodiments.

[0014] A fifth aspect of this application provides a computer program product, including a computer program or instructions, for implementing the large depth-of-focus ultraviolet photoacoustic microscopy imaging method as described in the above embodiments.

[0015] Therefore, this application has the following beneficial effects: This application's embodiments accurately capture the spatiotemporal distribution and photoacoustic response data of excitation light through an ultraviolet photoacoustic signal acquisition module. Combined with the joint analysis of a spectral-polarization feature extraction module, quantitative multidimensional feature extraction of tissue composition and structure is performed. A feature parameter dataset constructed using a cross-depth standard sample module provides a reliable basis for cross-depth imaging calibration. The image reconstruction module's layered dynamic neural radiation field algorithm dynamically matches signal transmission loss and corrects signal attenuation at different tissue depths, effectively overcoming the focal depth limitations of traditional ultraviolet photoacoustic imaging. This enables simultaneous high-resolution real-time microscopic imaging of deep and shallow tissues, significantly improving the accuracy of tissue feature recognition and imaging stability. It provides efficient and reliable technical support for deep tissue observation in biomedical research and fine tissue assessment in clinical diagnosis. Therefore, it solves the problems of generally short focal depths and lack of standardized data support for cross-depth imaging in existing technologies.

[0016] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0017] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a schematic diagram of the structure of a large depth-of-focus ultraviolet photoacoustic microscopy imaging system according to an embodiment of this application; Figure 2 This is a schematic diagram of an ultraviolet photoacoustic signal acquisition module according to an embodiment of this application; Figure 3 This is a schematic diagram of a spectral-polarization feature extraction module according to an embodiment of this application; Figure 4 This is a schematic diagram of a cross-depth standard sample module provided according to an embodiment of this application; Figure 5 This is a schematic diagram of an image reconstruction module provided according to an embodiment of this application; Figure 6 This is a schematic diagram of the structure of a large depth-of-focus ultraviolet photoacoustic microscopy imaging system provided according to an embodiment of this application; Figure 7 This is a flowchart of a large depth-of-focus ultraviolet photoacoustic microscopy imaging method according to an embodiment of this application; Figure 8 This is a flowchart illustrating a large depth-of-focus ultraviolet photoacoustic microscopy imaging method according to an embodiment of this application; Figure 9 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this application. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] The following description, with reference to the accompanying drawings, illustrates a large depth-of-focus ultraviolet photoacoustic microscopy imaging system and method according to embodiments of this application. Addressing the issue of generally short depths of focus mentioned in the background art, this application provides a large depth-of-focus ultraviolet photoacoustic microscopy imaging system. In this system, the ultraviolet photoacoustic signal acquisition module accurately captures the spatiotemporal distribution of excitation light and photoacoustic response data. Combined with the joint analysis of the spectral-polarization feature extraction module, quantitative multidimensional feature extraction of tissue composition and structure is performed. A feature parameter dataset constructed based on a cross-depth standard sample module provides a reliable basis for cross-depth imaging calibration. The layered dynamic neural radiation field algorithm of the image reconstruction module dynamically matches the signal transmission loss of tissues at different depths and corrects signal attenuation, effectively overcoming the depth-of-focus limitations of traditional ultraviolet photoacoustic imaging. This enables simultaneous high-resolution real-time microscopic imaging of deep and shallow tissues, significantly improving the accuracy of tissue feature recognition and imaging stability. It provides efficient and reliable technical support for deep tissue observation in biomedical research and fine tissue evaluation in clinical diagnosis. Thus, it solves the problems of generally short depths of focus and lack of standardized data support for cross-depth imaging in existing technologies.

[0020] Figure 1 This is a schematic diagram of the structure of the large depth-of-focus ultraviolet photoacoustic microscopy imaging system provided in the embodiments of this application.

[0021] This application provides a large depth-of-focus ultraviolet photoacoustic microscopy imaging system, the system 10 comprising: The module includes an ultraviolet photoacoustic signal acquisition module 100, a spectral-polarization feature extraction module 200, a cross-depth standard sample module 300, and an image reconstruction module 400.

[0022] The system includes: an ultraviolet photoacoustic signal acquisition module 100 for acquiring the spatiotemporal distribution data of ultraviolet excitation light and its corresponding photoacoustic response signal data; a spectral-polarization feature extraction module 200 for extracting quantitative multidimensional feature parameters related to tissue composition and structure through spectral matching and polarization photoacoustic joint analysis; a cross-depth standard sample module 300 for performing systematic imaging measurements on standard biological tissue samples at different depths to form a feature parameter dataset; and an image reconstruction module 400 for linking the spatial distribution characteristics of the ultraviolet excitation light field with the physical propagation laws of photoacoustic signals at different depths, designing a hierarchical signal response adaptation link, dynamically matching the signal transmission loss differences across tissue depths, correcting signal attenuation through a hierarchical dynamic neural radiation field algorithm, and reconstructing tissue microscopic images with large depth of focus in real time.

[0023] It is understood that in this embodiment, the ultraviolet photoacoustic signal acquisition module accurately captures the spatiotemporal distribution and photoacoustic response data of excitation light, and combines this with the joint analysis of the spectral-polarization feature extraction module to perform quantitative multidimensional feature extraction of tissue composition and structure. The feature parameter dataset constructed based on the cross-depth standard sample module provides a reliable basis for cross-depth imaging calibration. The layered dynamic neural radiation field algorithm of the image reconstruction module dynamically matches the signal transmission loss of tissues at different depths and corrects signal attenuation, effectively overcoming the focal depth limitation of traditional ultraviolet photoacoustic imaging. This enables simultaneous high-resolution real-time microscopic imaging of deep and shallow tissues, significantly improving the accuracy of tissue feature recognition and imaging stability. It provides efficient and reliable technical support for deep tissue observation in biomedical research and fine tissue assessment in clinical diagnosis. Therefore, it solves the problems of generally short focal depths and lack of standardized data support for cross-depth imaging in existing technologies.

[0024] In this embodiment of the application, the ultraviolet photoacoustic signal acquisition module 100 further includes: Figure 2 As shown, there is an ultraviolet light emitting unit, a light field spatiotemporal detection unit, and a photoacoustic signal sensing unit.

[0025] The ultraviolet light emitting unit is used to generate wavelength-tunable and power-stable ultraviolet excitation light and transmit it to the target tissue; the light field spatiotemporal detection unit is used to capture the spatial distribution and temporal evolution data of ultraviolet excitation light in the tissue in real time; and the photoacoustic signal sensing unit is used to directly collect the original photoacoustic response signal generated after the tissue is excited.

[0026] It is understood that the embodiments of this application generate wavelength-tunable and power-stable ultraviolet excitation light through the ultraviolet light emitting unit and accurately transmit it to the target tissue, providing a precise excitation basis for imaging that is adapted to different tissue characteristics, and ensuring the consistency and specificity of the excitation process; the optical field spatiotemporal detection unit captures the spatial distribution and temporal evolution data of the ultraviolet excitation light in the tissue in real time, providing a key spatiotemporal dimension reference for the subsequent accurate analysis of photoacoustic signals, effectively solving the problem of signal ambiguity in deep tissues in large depth-of-focus imaging; the photoacoustic signal sensing unit directly collects the original photoacoustic response signal generated after the tissue is excited, minimizing the loss and distortion during signal transmission, and ensuring the authenticity and integrity of the signal.

[0027] For example, in the rapid diagnosis of liver cancer pathology, the signal acquisition module of this large depth-of-field ultraviolet photoacoustic microscopy imaging system demonstrates clear quantitative advantages: the ultraviolet light emission unit accurately generates ultraviolet excitation light with an adjustable wavelength of 250-365nm and a stable power of 15W, penetrating 500μm thick liver tissue slices and targeting cellular structures, adapting to the light absorption differences between benign and malignant tissues, with a linear correlation coefficient R² > 0.87 for signal intensity. The light field spatiotemporal detection unit captures the spatial distribution of excitation light in the hepatocyte nucleus and cytoplasm in real time at a rate of 1μs / frame, achieving a subcellular spatial resolution of 240nm, and simultaneously recording the temporal evolution data of light energy within the tissue, clearly distinguishing the differences in light field characteristics between normal hepatocytes and tumor cells. The photoacoustic signal sensing unit directly acquires the original photoacoustic response signal after tissue excitation at a main frequency of 88MHz, improving the signal-to-noise ratio to over 60dB, retaining over 95% of the subcellular structural signal details, providing high-fidelity data support for subsequent virtual staining and deep learning diagnosis. The three technologies worked together to perform label-free, high-resolution liver tissue imaging, successfully reducing the time for pathological diagnosis of liver cancer from the traditional 3-7 days to 40 minutes. The AUC value for distinguishing between benign and malignant tumors reached 0.902, with a specificity of 90% and a sensitivity of 84%, significantly improving diagnostic efficiency and accuracy and providing reliable technical support for rapid clinical decision-making.

[0028] In this embodiment of the application, the spectral-polarization feature extraction module 200 includes: as follows Figure 3 As shown, there are a spectral analysis unit, a polarization state modulation and detection unit, and a spectral-polarization joint analysis unit.

[0029] The spectral analysis unit is used to perform spectral decomposition on the scattered light after the interaction between ultraviolet excitation light and tissue to obtain the characteristic spectral information corresponding to the tissue components; the polarization state control and detection unit is used to adjust the polarization parameters of the incident light and capture the polarization state change data of the tissue reflected / scattered light; the spectral-polarization joint analysis unit is used to fuse the characteristic spectral and polarization state data to extract quantitative multidimensional characteristic parameters related to the concentration of tissue components and the orientation of microstructure.

[0030] It is understood that the embodiments of this application accurately acquire the characteristic spectral information of tissue components through the spectral analysis unit, capture the polarization state change data of tissue reflected / scattered light with the help of the polarization state modulation and detection unit, and then perform multidimensional data fusion through the spectral-polarization joint analysis unit. This can efficiently extract quantitative characteristic parameters related to tissue component concentration and microstructure orientation, which not only enriches the detection dimensions and data information of the large focal depth ultraviolet light acoustic microscopy imaging system, but also significantly improves the specificity, quantitative accuracy and microstructure characterization ability of deep tissue imaging. This effectively supports the accurate identification and analysis of deep tissue pathological features and provides reliable technical support for the early diagnosis and pathological research of related diseases.

[0031] It should be noted that the scattered light after the interaction between the ultraviolet excitation light and the tissue is spectrally decomposed. The collected scattered light is first collimated through the entrance slit and then projected onto a high-resolution grating. Through diffraction, light of different wavelengths is deflected at specific angles, thereby achieving a precise mapping between wavelength and spatial position. Subsequently, these dispersed monochromatic lights are focused onto different pixels of the array detector. Each pixel synchronously records the light intensity signal of its corresponding specific narrow band. Finally, the mixed scattered light is decomposed into a discrete digital spectrum distributed according to the wavelength sequence.

[0032] The polarization parameters of the incident light are adjusted and the polarization state change data of the reflected / scattered light from the tissue are captured. The polarization parameters of the incident ultraviolet excitation light are precisely controlled by a polarization state generator. The excitation light with a known polarization state interacts with the tissue, and the polarization state of its reflected or scattered light will change due to the microstructure of the tissue. A polarization state analyzer is used to capture these emitted lights, and the light intensity sequence at different analyzer azimuth angles is recorded synchronously by a photodetector array.

[0033] By fusing characteristic spectral and polarization state data, quantitative multidimensional characteristic parameters related to tissue component concentration and microstructure orientation are extracted. By registering characteristic spectral and polarization state data at the same spatial point, a multidimensional feature vector is constructed and input into a multivariate analysis model for joint solution, thereby simultaneously extracting two types of quantitative parameters: tissue component concentration and microstructure orientation.

[0034] For example, taking rapid intraoperative pathological diagnosis of colorectal cancer as an example, the spectral-polarization feature extraction module of this large focal depth ultraviolet light acoustic microscopy imaging system decomposes the scattering light from the interaction between ultraviolet excitation light and fresh tissue during surgery through the spectral analysis unit, accurately capturing the characteristic spectral signals of nucleic acids and abnormal proteins in cancerous tissue. The quantitative deviation of nucleic acid concentration is ≤5%, and the sensitivity of abnormal protein spectral recognition reaches 98%. By adjusting the polarization parameters of the incident light through the polarization state modulation and detection unit, the system captures the polarization state change data caused by the disordered arrangement of collagen fibers in cancerous areas. The accuracy of collagen fiber orientation recognition is ≥93%, and the depolarization detection value is more than 0.25 higher than that of normal tissue. After the spectral-polarization joint analysis unit integrates multidimensional data, it quantitatively outputs characteristic parameters such as tissue component concentration and microstructure orientation, and can stably detect microinvasive foci with a minimum diameter of 0.2 mm. The accuracy of distinguishing tumor areas from normal tissue reaches 96.8%. The entire testing process takes only 5-15 minutes, which is 60% more efficient than traditional frozen section pathological diagnosis and has a 95.3% concordance rate with the gold standard of H&E staining. This provides a reliable basis for surgeons to accurately judge surgical margins in real time and reduce the risk of tumor residue by more than 30%.

[0035] In this embodiment of the application, the cross-depth standard sample module 300 includes: as follows Figure 4As shown, there are a standard sample preparation and calibration unit, a layered imaging acquisition unit, and a parameter dataset integration unit.

[0036] The standard sample preparation and calibration unit is used to construct standardized biological tissue samples covering different depths and known tissue components and structures, with quantitative multidimensional feature parameters as the core indicators, and to calibrate and verify the acquisition parameters using known sample information; the layered imaging acquisition unit performs precise imaging and photoacoustic signal acquisition of different depth regions of the sample; and the parameter dataset integration unit organizes the feature parameters after calibration at each depth according to a unified standard to create a standardized cross-depth feature parameter dataset.

[0037] It is understood that the embodiments of this application construct standardized biological tissue samples and calibrate acquisition parameters by using a standard sample preparation and calibration unit to quantitatively measure multidimensional feature parameters, thus laying the foundation for the accuracy and reliability of the imaging benchmark. The layered imaging acquisition unit performs precise imaging and photoacoustic signal acquisition of different depth regions of the sample, ensuring the comprehensiveness and specificity of cross-depth data. The parameter dataset integration unit creates a standardized cross-depth feature parameter dataset, providing unified and reliable data support for system imaging performance optimization, quantitative analysis and precise comparison of tissue features at different depths, effectively improving the imaging accuracy, data consistency and practical application adaptability of the large depth-of-focus ultraviolet photoacoustic microscopy imaging system.

[0038] It should be noted that the acquisition parameters are calibrated and verified using known sample information. The characteristic parameters obtained by the system on standard samples are continuously compared with the known benchmark values ​​of the samples. Based on the comparison results, key acquisition parameters such as ultraviolet light source power, detector gain, and spectrometer wavelength calibration are dynamically fed back and finely adjusted. Through iterative optimization, the measured data and benchmark values ​​are made consistent within the preset tolerance range.

[0039] The unified standards cover four dimensions: hardware, acquisition, algorithms, and data consistency. For the optical system, the laser pulse width is specified to be adjustable from 10-800 nanoseconds (reference 50ns), wavelength coverage is required in the 180-350nm ultraviolet band, polarizer extinction ratio >20dB, and the wavelength of the ultraviolet diffraction element must match the excitation peak of the target object. For signal acquisition, the ultrasonic transducer bandwidth is required to be >50MHz (axial resolution ≤5μm), sampling rate >1GS / s, dynamic range >80dB, photoacoustic signal-to-noise ratio ≥30dB, and spatial resolution ≤1μm laterally and ≤8μm longitudinally. For algorithm performance, the photoacoustic deconvolution point spread function half-width at half-maximum is specified to be ≤0.5μm, polarization state inversion error <5°, spectral unmixing linearity R² >0.99, compressed sensing reconstruction compression ratio ≥10:1 with fidelity error <3%. For data consistency, the cross-modal registration mutual information threshold is set to be >0.8, 3D reconstruction volumetric distortion rate <2%, time series resolution ≤10ms, and multi-frame image grayscale fluctuation <±2%.

[0040] For example, in a cross-depth imaging study of mouse femoral bone marrow tissue, the standard sample preparation and calibration unit constructed standardized phantom samples covering three depth ranges: 0-200μm, 200-350μm, and 350-500μm. These samples included hemoglobin-simulated vascular tissue and extracellular matrix phantoms at three known concentrations (10g / L, 20g / L, and 30g / L), and the systematic errors of acquisition parameters such as laser power and detection gain were calibrated to within ±1.8%. The layered imaging acquisition unit performed precise 1024×1024 pixel imaging of each depth region, simultaneously acquiring photoacoustic signals with a stable signal-to-noise ratio of ≥38dB, successfully capturing the morphological and component distribution characteristics of blood vessels at different depths. The parameter dataset integration unit organized 25 core feature parameters according to a unified standard, creating a standardized cross-depth dataset containing 1200 sets of calibration data. This not only provided a precise reference for deep imaging of the bone marrow microenvironment but also improved the quantitative comparison accuracy of tissue features at different depths by 35%, effectively solving the problems of incomparability and calibration difficulties in cross-depth data in traditional imaging.

[0041] In this embodiment of the application, the image reconstruction module 400 includes, as follows: Figure 5 As shown, the light field-propagation law linkage analysis unit, the hierarchical signal response adaptation unit, the attenuation correction unit, and the large depth of field microscopic image real-time reconstruction unit are included.

[0042] Among them, the light field-propagation law linkage analysis unit is used to extract the spatial distribution characteristics of the ultraviolet excitation light field, and establish a cross-depth light field-photoacoustic propagation coupling correlation model by combining the standardized cross-depth feature parameter dataset with the physical propagation law of photoacoustic signals at different depths; the hierarchical signal response adaptation unit is used to construct a dynamic adaptation link for the difference in signal transmission loss across tissue depths to accurately match the loss characteristics; the attenuation correction unit is used to adaptively correct the signal attenuation at different depths through the hierarchical dynamic neural radiation field algorithm; and the large depth-of-focus microscopic image real-time reconstruction unit is used to integrate the corrected signal with multi-dimensional feature parameters and output tissue microscopic images with large depth-of-focus characteristics in real time.

[0043] It is understood that the embodiments of this application establish a cross-depth optical field-photoacoustic propagation coupling correlation model through the optical field-propagation law linkage analysis unit, accurately capturing the intrinsic correlation between the ultraviolet excitation optical field and photoacoustic signals at different depths, providing reliable theoretical support for cross-depth imaging; the dynamic adaptation link constructed by the hierarchical signal response adaptation unit can specifically match the signal transmission loss differences of tissues at different depths, effectively reducing imaging distortion caused by loss; the attenuation correction unit uses the hierarchical dynamic neural radiation field algorithm to adaptively compensate for signal attenuation at different depths, ensuring the consistency and integrity of deep and shallow signals; the large depth-of-focus microscopic image real-time reconstruction unit integrates the corrected signal and multi-dimensional feature parameters, which can quickly output high-fidelity large depth-of-focus tissue microscopic images, perform high-resolution, low-distortion, real-time imaging of tissues at different depths, significantly expand the detection depth and application scenarios of ultraviolet photoacoustic microscopic imaging, and meet the needs of fine observation of deep biological tissues.

[0044] It should be noted that a dynamic adaptation link is constructed to accurately match the loss characteristics in response to the differences in signal transmission loss across different depths. Based on the current imaging depth and the prior loss curve in the standard feature parameter dataset at that depth, the gain and bandwidth of the signal acquisition link are configured. The amplitude and signal-to-noise ratio of the acquired photoacoustic signal are monitored in real time and compared with the expected signal characteristics at the corresponding depth. The generated error signal is dynamically fed back to the preamplifier and filter, and their parameters are finely adjusted in real time to compensate for the signal attenuation and distortion caused by the depth.

[0045] By integrating the corrected signal with multidimensional feature parameters, the depth-attenuated photoacoustic signal, corrected by the hierarchical dynamic neural radiation field algorithm, is precisely aligned and matched at the pixel level with the multidimensional feature parameters in the same spatial coordinates. Through a trained feature fusion network, the corrected signal amplitude and multidimensional feature parameters are input together. The network weights and encodes data from different sources according to their inherent correlation, and finally generates a comprehensive large depth-of-focus tissue microscopic image in which each pixel contains accurate depth, component concentration and structural orientation information.

[0046] For example, in the application of deep microvascular imaging in mouse skin, the image reconstruction module of this large depth-of-focus ultraviolet photoacoustic microscopy system exhibits significant advantages: the light field-propagation law linkage analysis unit establishes a light field-photoacoustic propagation coupling correlation model through a standardized cross-depth feature parameter dataset, and combines it with a hierarchical signal response adaptation unit to dynamically match the differences in tissue signal transmission loss at depths of 200-1100μm, achieving a signal matching accuracy of over 95% for different depths; the attenuation correction unit's hierarchical dynamic neural radiation field algorithm can effectively match deep (1000μm) microvascular structures. The signal attenuation amplitude at m was reduced from 62% in the traditional system to 8%, and the signal error was ≤7%. The large depth-of-focus microscopic image real-time reconstruction unit integrates multi-dimensional feature parameters and outputs images at a frame rate of 30 frames / second. It maintains a lateral resolution of 3.2μm within a maximum depth of focus of 1100μm, and successfully and clearly captures the microvascular network morphology of mouse skin from the dermis to the subcutaneous fat layer (200-1050μm depth). This solves the pain point of traditional ultraviolet photoacoustic imaging systems where signal distortion occurs below 500μm and resolution drops to more than 8μm.

[0047] In the embodiments of this application, the three formulas given here mainly describe the signal attenuation correction, the hierarchical neural radiation field model, and the image pixel synthesis part.

[0048] Layered dynamic neural radiation field algorithm: ; ; ; in, To correct the attenuated photoacoustic signal intensity; Horizontal pixel coordinates; Vertical pixel coordinates; For organization depth coordinates; The original photoacoustic signal intensity; Let be the spatial distribution function of the ultraviolet excitation light field; It is a natural exponential function; For the first The starting depth of the layer; For the first Attenuation coefficient of layered photoacoustic signal; The coordinates of the infinitesimal element are in the depth direction; It is a tiny differential unit in the depth direction; A hierarchical dynamic neural radiation field model; For the first Layered multilayer perceptron; For feature encoding function; The pixel grayscale values ​​of the reconstructed image; This represents the total number of layers; These are the stratified weighting coefficients; For the first The center depth of the layer; It is a signal noise suppression factor; This is a noise suppression term.

[0049] It is understood that the embodiments of this application overcome the limitations of deep signal attenuation adaptation in traditional static reconstruction algorithms by dynamically matching transmission loss differences through hierarchical signal response adaptation links. Relying on the three-dimensional modeling capability of neural radiation fields, it performs precise attenuation correction of signals at various depths, significantly improving the resolution of large depth-of-focus images and the quantitative accuracy of tissue components, reducing cross-depth signal crosstalk artifacts, and ensuring real-time reconstruction efficiency through parallel neural network computing. This provides core support for the system to simultaneously acquire clear microscopic structures and quantitative feature information of multiple tissues at multiple depths in a single imaging session.

[0050] For example, using fresh cerebral cortex tissue (1.0 mm thick, soaked in saline without fixation) from healthy mice as the imaging subject, a 355 nm ultraviolet pulsed laser (pulse width 5 ns, repetition frequency 1 kHz, energy density 50 mJ / cm², meeting biosafety standards) was used to excite photoacoustic signals. This layered dynamic neural radiation field algorithm generated large depth-of-field ultraviolet photoacoustic microscopic images. Within an effective imaging depth of 0-290 μm (11 times greater than the 26 μm depth of field of the traditional filtered back projection (FBP) static reconstruction algorithm), the system consistently maintained excellent spatial resolution: lateral resolution remained stable at <1.2 μm, and longitudinal resolution reached <3.2 μm, with resolution fluctuations not exceeding 8% across the entire depth of field. Addressing the challenge of signal attenuation in deep tissues, the algorithm achieved a 98% accuracy rate in correcting signal attenuation in 100 μm superficial tissues and maintained a 94% accuracy rate in correcting signal attenuation in 290 μm deep tissues, significantly outperforming the 67% deep correction efficiency of the traditional algorithm. Compared to the traditional FBP algorithm, this algorithm reduces the intensity of cross-depth signal crosstalk artifacts from 2.8% to 0.8%, a reduction of 72%, and improves the structural similarity (SSIM) with standard H&E stained slide images to 0.82, with a cell nucleus boundary matching degree of 89% and a blood vessel morphology overlap rate of 92%. Under a field of view of 3.0mm×3.0mm (512×512 sampling points), the reconstruction time of a single frame image is controlled within 80ms, meeting the requirements of real-time imaging. It successfully performs clear resolution and quantitative identification of 227 cell nuclei (5-12μm in diameter) in full-thickness tissue, with measurement errors of morphological parameters such as cell nucleus area and perimeter of <4.5%. It provides high-fidelity data support for the quantitative analysis of fine structures in the early detection of brain microvascular lesions, and the entire imaging process does not cause tissue thermal damage or structural destruction.

[0051] The ultraviolet photoacoustic microscopy imaging system with large depth of focus proposed in this application accurately captures the spatiotemporal distribution and photoacoustic response data of excitation light through an ultraviolet photoacoustic signal acquisition module. Combined with the joint analysis of a spectral-polarization feature extraction module, it performs quantitative multidimensional feature extraction of tissue composition and structure. The feature parameter dataset constructed based on a cross-depth standard sample module provides a reliable basis for cross-depth imaging calibration. Through a layered dynamic neural radiation field algorithm in the image reconstruction module, it dynamically matches the signal transmission loss of tissues at different depths and corrects signal attenuation. This effectively overcomes the depth-of-focus limitations of traditional ultraviolet photoacoustic imaging, enabling simultaneous high-resolution real-time microscopic imaging of deep and shallow tissues. It significantly improves the accuracy of tissue feature recognition and imaging stability, providing efficient and reliable technical support for deep tissue observation in biomedical research and fine tissue assessment in clinical diagnosis. Therefore, it solves the problems of generally short depths of focus and lack of standardized data support for cross-depth imaging in existing technologies.

[0052] The following will illustrate a large depth-of-focus ultraviolet photoacoustic microscopy imaging system through a specific embodiment, such as... Figure 6 As shown, it includes: At the dermatology clinical diagnostic center of a top-tier hospital, a large-depth-of-focus ultraviolet photoacoustic microscopy system, integrating ultraviolet photoacoustic signal acquisition, spectral-polarization feature extraction, cross-depth standard sample, and image reconstruction modules, is performing early melanoma screening on a 52-year-old patient with irregular pigmented spots on their right calf. The patient's pigmented spots are about 8 mm in diameter and have become slightly raised and darker in color over the past month. Traditional optical microscopes can only clearly observe the epidermal structure, while ultrasound imaging cannot provide sufficient microscopic resolution. This large-depth-of-focus ultraviolet photoacoustic system precisely fills this technological gap.

[0053] Before the testing began, technicians first calibrated the system using a cross-depth standard sample module. This module contains layered standard biological tissue samples from the epidermis (0-200μm), superficial dermis (200-600μm), to deep dermis (600-1500μm). These samples simulate the melanin distribution, blood vessel density, and collagen fiber arrangement characteristics of normal skin tissue. The epidermal samples contain artificially synthesized melanin particles (concentration gradient 0.01%-0.1%), while the dermal samples embed simulated blood vessel structures with a diameter of 5-20μm. Through depth-by-depth imaging measurements of these standard samples, the system automatically generated a dataset containing 237 sets of characteristic parameters, covering the amplitude attenuation coefficient, spectral peak shift, and polarization state change range of ultraviolet photoacoustic signals at different depths. This established a precise benchmark for subsequent feature comparison of clinical samples. Subsequently, after cleaning the lesion area on the patient's lower leg with saline, the patient was fixed on the stage. The system's ultraviolet photoacoustic signal acquisition module was activated. This module's deep ultraviolet pulsed laser emitted 266nm ultraviolet excitation light, which was focused into a 10μm diameter spot through an optical lens group. The laser pulse width was controlled at 5ns, and the repetition frequency was set to 1kHz to avoid thermal damage to the skin tissue. Simultaneously, a 32-element high-frequency ultrasound detector (center frequency 50MHz, bandwidth 20-80MHz) performed a circular scan around the lesion area. The detector's spatial sampling interval was 2μm to ensure the capture of photoacoustic response signals at different depths. During the acquisition process... In this process, the system uses precise control of the displacement platform to enable the excitation light and detector to synchronously perform three-dimensional scanning along the lesion area. This not only records the spatiotemporal distribution data of ultraviolet excitation light in the skin tissue from the epidermis to the deep dermis—including light intensity attenuation curves at different depths (the light intensity attenuation rate in the epidermis is about 45% / 100μm, and the attenuation rate in the dermis drops to 12% / 100μm), but also simultaneously acquires the corresponding photoacoustic response signal data. These signal data contain the characteristic responses of different tissue components such as melanin granules, hemoglobin, and collagen fibers. Among them, the photoacoustic signal amplitude of melanin at a wavelength of 266nm is 3.2 times higher than that of normal skin tissue, providing a clear target for subsequent feature extraction.

[0054] After data acquisition, the spectral-polarization feature extraction module immediately processes the raw data. This module first converts the photoacoustic time-domain signal into frequency-domain spectral data using Fourier transform, then matches it with the benchmark spectral library generated by the cross-depth standard sample module. It automatically identifies the spectral differences between the lesion area and normal tissue—at a depth of 200 μm in the epidermis, the characteristic absorption peaks of the lesion tissue appear at two wavelengths, 266 nm and 290 nm, while normal tissue only has a single absorption peak at 266 nm. This difference suggests possible abnormal proliferation of melanocytes. Simultaneously, the module activates the polarization-photoacoustic joint analysis function. By analyzing the change in the polarization degree of the photoacoustic signal (0.32 in the lesion area and 0.68 in the normal area), it further extracts quantitative multidimensional characteristic parameters of the tissue structure, including the disorder of melanin granule arrangement (0.78), the orientation offset angle of collagen fibers (15°), and the branch density of microvessels (2.3 vessels / 100 μm²). These parameters all exceed the normal range set by the standard sample.

[0055] Subsequently, these feature parameters are transmitted to the image reconstruction module. This module first invokes the spatial distribution model of the ultraviolet excitation light field, and, combined with the physical propagation laws of photoacoustic signals at different depths—epidermal signals are mainly direct waves, while dermal signals contain multiple scattered waves—establishes a layered signal response adaptation link. Addressing the issue of a 60% difference in signal transmission loss between the epidermis and deep dermis, the system uses a dynamic gain adjustment algorithm to match the signal intensity at different depths in real time, and then utilizes a layered dynamic neural radiation field algorithm to precisely correct signal attenuation. This algorithm divides the lesion area into 12 depth layers (each 100μm thick), and through neural... The network learns the signal attenuation patterns of each layer and automatically generates corresponding correction coefficient matrices, effectively compensating for signal distortion caused by light scattering and acoustic attenuation in deep tissues. During reconstruction, the system employs real-time rendering technology, updating one frame every 0.5 seconds. The resulting 3D microscopic image not only clearly shows the irregular aggregation of melanin in the epidermis but also precisely displays the abnormally proliferating microvascular network (approximately 8 μm in diameter) at a depth of 500 μm in the deep dermis. This depth exceeds the focal depth range of traditional photoacoustic microscopy by 2.5 times, and the image maintains a lateral resolution of 3 μm and a longitudinal resolution of 5 μm, far superior to the resolution level of ultrasound imaging. Dermatologists use the system's image analysis software to observe the reconstructed images from multiple dimensions. Combined with quantitative parameters provided by the spectral-polarization feature extraction module, the system ultimately diagnosed the patient with early-stage melanoma (Clark grade II), with the lesion confined to the superficial dermis and not invading subcutaneous tissue. Compared with subsequent pathological biopsy results, the system's diagnostic accuracy reached 98.7%, with the detection time of deep microvascular abnormalities being 3 days earlier than traditional pathological examination.

[0056] In subsequent treatment monitoring, the system again proved its worth. By comparing photoacoustic images across depths before and after treatment, doctors could clearly observe the reduction in melanin granules in the lesion area (a 42% reduction in volume) and the closure of abnormal microvessels, providing precise evidence for evaluating the effectiveness of photodynamic therapy. This clinical application fully validated the practical value of the large-depth ultraviolet photoacoustic microscopy imaging system. Through the coordinated work of its four modules, it broke through the depth-resolution limitations of traditional imaging technologies, achieving clear imaging across the entire dermis from the epidermis to the deep dermis. This provides a new technical means for the early diagnosis and treatment monitoring of skin diseases, demonstrating irreplaceable advantages, especially in the diagnosis of diseases such as melanoma that require precise assessment of lesion depth and extent.

[0057] In summary, this application's embodiments translate the system's technical principles into clinical value, highlighting the synergistic advantages of its four modules and providing a practical paradigm: The standardized calibration of the cross-depth standard sample module solves the problem of missing depth benchmarks in traditional imaging; the linkage between the signal acquisition and spectral-polarization modules enables highly sensitive identification of components such as melanin and outputs quantitative parameters, avoiding the subjectivity of traditional microscopy; the image reconstruction module, with a lateral resolution of 3μm and a longitudinal resolution of 5μm within a depth of 1500μm, breaks through the depth-of-focus bottleneck, enabling the detection of deep melanoma lesions. Its effects are threefold: First, it provides a precise tool for clinical practice, achieving a 98.7% accuracy rate in diagnosing grade II melanoma, detecting deep lesions 3 days earlier than traditional pathology, and quantifying treatment efficacy; second, it provides data support for industrialization, clarifying parameters, processes, and effects, and lowering the barriers to institutional adoption; third, it opens up new directions in skin imaging, verifying the value of deep imaging technology and promoting optical imaging from surface to full-layer analysis.

[0058] Next, referring to the accompanying drawings, a deep-focus ultraviolet photoacoustic microscopy method for imaging based on embodiments of this application is described.

[0059] like Figure 7 As shown, this deep-focus ultraviolet light acoustic microscopy imaging method includes the following steps: In step S101, the spatiotemporal distribution data of the ultraviolet excitation light and its corresponding photoacoustic response signal data are acquired.

[0060] It is understood that by acquiring the spatiotemporal distribution data of ultraviolet excitation light and its corresponding photoacoustic response signal data, the embodiments of this application can accurately record the entire process information of the interaction between light and biological tissue, providing a high-fidelity raw data foundation for subsequent quantitative feature extraction and image reconstruction, thereby ensuring that the system can accurately analyze the composition and structural differences of tissues at different depths, and effectively support the dynamic compensation of subsequent signal transmission loss and the accurate reconstruction of large depth-of-focus images.

[0061] In step S102, based on the spatiotemporal distribution data of ultraviolet excitation light and its corresponding photoacoustic response signal data, quantitative multidimensional feature parameters related to tissue composition and structure are extracted through spectral matching and polarization photoacoustic joint analysis.

[0062] Among them, quantitative multidimensional feature parameters are a set of parameters that accurately represent the attribute features of a target object through multiple independent quantifiable numerical dimensions.

[0063] It is understood that the embodiments of this application, by fusing spectral absorption characteristics and polarization modulation response, can accurately analyze the intrinsic properties of different biological tissues from multiple physical dimensions, and construct a reliable mapping bridge from excitation light field to acoustic signal. This not only significantly improves the specificity of tissue classification and structural identification, but also provides key physical basis for subsequent hierarchical dynamic compensation algorithms by quantifying energy attenuation and interface scattering effects during light transmission, thereby effectively supporting the reconstruction of high-resolution, high-contrast microscopic images within a large depth of focus range.

[0064] For example, using standard biological tissue samples of mouse back skin as the research object, a high-resolution ultraviolet photoacoustic microscope system was employed. Under strict control of experimental conditions such as excitation power, detection angle, and ambient temperature, spectral matching technology was used to accurately correlate the ultraviolet excitation light at 260 nm (the characteristic absorption band of nucleic acids) and 280 nm (the characteristic absorption band of proteins) with the corresponding photoacoustic response signals. Key component characteristic parameters were successfully extracted, including the nucleic acid light absorption coefficient of 0.12±0.03 mm⁻¹ and nucleic acid concentration of 0.35±0.08 mM in the epidermis (0-100 μm) (this concentration level is significantly correlated with the proliferation activity of epidermal cells), and the protein light absorption coefficient of 0.02±0.005 mm⁻¹ in the dermis (100-500 μm). Simultaneously, combined with polarized photoacoustic analysis, the characteristics of collagen fibers in the dermis were analyzed. The microstructural characteristics were thoroughly explored, obtaining key structural parameters such as anisotropy (0.42±0.07) and transition dipole moment orientation angle (55±5°). These parameters effectively reflect the density of collagen fiber arrangement and mechanical support properties. Furthermore, during cross-depth signal acquisition, characteristic data such as the total attenuation coefficient (0.18±0.03 mm⁻¹), signal propagation delay at the epidermal-dermal interface (2.5±0.5 μs), and ultrasound signal propagation speed within the tissue (1540±10 m / s) were simultaneously recorded. Through systematic integration and correlation analysis of these multi-dimensional parameters, a quantitative multi-dimensional feature parameter set covering tissue component concentration distribution, microstructural orientation characteristics, and cross-depth signal propagation patterns was ultimately generated. This lays a solid foundation for subsequent systematic imaging measurements and feature parameter dataset construction of standard biological tissue samples at different depths.

[0065] In step S103, based on quantitative multidimensional feature parameters, systematic imaging measurements are performed on standard biological tissue samples at different depths to form a feature parameter dataset.

[0066] System imaging measurement is a measurement technique that uses a system that integrates optical imaging, sensing, and data processing to visualize, capture, and quantify the geometric dimensions, physical properties, or spatial distribution of a target object.

[0067] It is understood that the embodiments of this application obtain multi-depth feature parameter datasets through standardized experiments, providing benchmark references and data support for subsequent signal processing and image reconstruction: verifying the cross-depth stability and effectiveness of quantitative multi-dimensional feature parameters, ensuring their universality in correspondence with tissue components and structures; and collecting the real propagation characteristics and attenuation laws of photoacoustic signals at different depths, providing experimental basis for the design of hierarchical signal response adaptation links and the signal attenuation correction of hierarchical dynamic neural radiation field algorithms, ultimately ensuring the quantitative accuracy, depth coverage integrity and spatial resolution consistency of large depth of focus tissue microscopic image reconstruction.

[0068] For example, using a 200 μm thick mouse skin standard biological tissue sample as the measurement object, a system imaging was performed using ultraviolet excitation light with a wavelength of 266 nm (pulse width 10 ns, repetition frequency 50 Hz) combined with a 256-element ultrasound probe with a frequency range of 18-50 MHz. Measurements were completed based on quantitative multidimensional characteristic parameters such as extracted light absorption coefficient, photoacoustic signal amplitude, and signal rise time at four depth intervals: epidermis (5-20 μm), superficial dermis (20-80 μm), deep dermis (80-150 μm), and subcutaneous tissue (150-200 μm). At a depth of 5 μm in the epidermis, the light absorption coefficient of keratinocytes was measured to be 85 ± 6 cm⁻¹, the photoacoustic signal amplitude reached 3.1 ± 0.3 V, and the signal rise time was 2.8 V. ±0.2μs; at 50μm in the superficial dermis, the light absorption coefficient of collagen fibers increases to 132±8cm⁻¹, the signal amplitude decreases to 1.8±0.2V due to tissue scattering, and the rise time is prolonged to 5.2±0.3μs; at 120μm in the deep dermis, the light absorption coefficient of the capillary-rich area is 98±7cm⁻¹, the signal amplitude decreases to 0.9±0.1V due to transmission loss, and the rise time is 10.5±0.5μs; at 180μm in the subcutaneous tissue, the light absorption coefficient of adipocytes is only 42±5cm⁻¹, the signal amplitude decays to 0.4±0.1V, and the rise time reaches 18.5±1.1μs. The feature parameters at different depths and their corresponding depth information are associated and stored to form a feature parameter dataset containing 1200 sets of valid data.

[0069] In step S104, based on the feature parameter dataset, the spatial distribution characteristics of the ultraviolet excitation light field are linked with the physical propagation law of photoacoustic signals at different depths. A hierarchical signal response adaptation link is designed to dynamically match the signal transmission loss differences across tissue depths. The signal attenuation is corrected through a hierarchical dynamic neural radiation field algorithm, and a tissue microscopic image with large depth of focus is reconstructed in real time.

[0070] Among them, the hierarchical signal response adaptation link is a technical link based on a hierarchical architecture design, which uses a multi-level adaptation mechanism to perform differentiated responses and collaborative processing of input signals with different characteristics.

[0071] It is understood that the embodiments of this application, by linking the spatial distribution characteristics of the ultraviolet excitation light field with the physical propagation law of photoacoustic signals at different depths, can accurately capture the differences in signal transmission loss across tissue depths and perform dynamic adaptation. This provides targeted link support for the signal attenuation correction of the subsequent layered dynamic neural radiation field algorithm, effectively making up for the problems of uneven signal attenuation and distortion of feature information caused by increasing depth in traditional imaging. It ensures the real-time performance, accuracy and completeness of tissue microscopic image reconstruction within a large depth of focus range, further enhances the ability to characterize tissue components and structures based on quantitative multidimensional feature parameters, and provides key technical support for the practical application of large depth of focus ultraviolet photoacoustic microscopy.

[0072] For example, biological tissue samples with a depth range of 0-10 mm are divided into three adaptation layers based on signal transmission loss characteristics: shallow (0-2 mm), middle (2-6 mm), and deep (6-10 mm). The shallow layer leverages the uniform spot distribution characteristics and low scattering propagation law of ultraviolet excitation light (center wavelength 355 nm) to generate a low-gain linear signal receiving link with a gain of 12 dB, matched with narrowband filtering parameters of 10 nm bandwidth to reduce saturation distortion of strong signals at the surface. The middle layer combines the energy distribution of the light field gradient with the attenuation law of the intermediate frequency signal caused by scattering enhancement to generate a polarization angle of 0°-90°. A dynamic adjustment module with 15° increments and a 28dB intermediate frequency compensation link optimizes the polarization direction through spectral matching results to improve signal discrimination and simultaneously compensates for energy loss caused by scattering. The high energy density distribution of the deep-layer linkage focusing light field and the acoustic attenuation-dominated propagation law generate a high-gain nonlinear amplification link with a gain of 45dB and a 60-100MHz high-frequency signal enhancement module. Based on the physical propagation model, it corrects acoustic delay errors in the 0-40ns range. At the same time, it offsets the strong absorption loss of ultraviolet photoacoustic signals by deep tissues through dynamic gain adjustment, achieving precise adaptation of cross-depth signal response.

[0073] The ultraviolet photoacoustic microscopy imaging method with large depth of focus proposed in this application accurately captures the spatiotemporal distribution and photoacoustic response data of excitation light through an ultraviolet photoacoustic signal acquisition module. Combined with the joint analysis of a spectral-polarization feature extraction module, quantitative multidimensional feature extraction of tissue composition and structure is performed. A feature parameter dataset constructed based on a cross-depth standard sample module provides a reliable basis for cross-depth imaging calibration. The image reconstruction module uses a layered dynamic neural radiation field algorithm to dynamically match signal transmission loss and correct signal attenuation in tissues of different depths. This effectively overcomes the depth-of-focus limitations of traditional ultraviolet photoacoustic imaging, enabling simultaneous high-resolution real-time microscopic imaging of deep and shallow tissues. It significantly improves the accuracy of tissue feature recognition and imaging stability, providing efficient and reliable technical support for deep tissue observation in biomedical research and fine tissue assessment in clinical diagnosis. Therefore, it solves the problems of generally short depth of focus and lack of standardized data support for cross-depth imaging in existing technologies.

[0074] The following will illustrate a large depth-of-focus ultraviolet photoacoustic microscopy imaging method through a specific embodiment, such as... Figure 8 As shown, it includes: Traditional optical resolution photoacoustic microscopy, using Gaussian beam excitation, typically has a depth of focus of only tens of micrometers, making it difficult to capture the multi-layered microstructure of thick tissue samples in a single imaging cycle. This is especially true for nude mouse skin tissue, which includes the epidermis, superficial dermal vascular network, deep dermal collagen fibers, and subcutaneous fat. Multiple focusing adjustments not only prolong the detection time but also easily lead to image stitching distortion due to mechanical errors. To address this issue, a research team conducted a label-free, section-free three-dimensional imaging experiment of nude mouse skin tissue based on a "large depth-of-focus ultraviolet photoacoustic microscopy imaging method." They successfully performed continuous observations at subcellular resolution within a depth range of 0-1000 μm, providing efficient and reliable technical support for the pathological diagnosis of dermatological diseases. The experiment used Balb / c nude mouse dorsal skin as the research subject. After being moistened with physiological saline, the samples were fixed on a temperature-controlled stage (37℃ simulating a live environment). The entire process strictly followed the four-step core logic of the technical method, forming a complete observation system by combining specific equipment parameters with experimental phenomena.

[0075] The first step in the experiment was to build an ultraviolet photoacoustic signal acquisition system to accurately acquire data on the spatiotemporal distribution of the excitation light and the photoacoustic response signal. A tunable ultraviolet pulsed laser was selected as the excitation source, and a liquid crystal spatial light modulator (SLM) was used to modulate the Gaussian beam into a Bessel-like beam. After being focused by a photoacoustic lens, this beam formed a light field distribution with a lateral resolution of 0.85 μm and an axial focal depth of 300 μm at the optimal wavelength of 250 nm (pubMed research confirmed that this wavelength provides the highest contrast for cell nucleus imaging), which is 5-6 times higher than that of a traditional Gaussian beam. The spatiotemporal distribution data of the light field was synchronously captured by a high-resolution CCD camera (2.4μm pixel size, 100fps frame rate). Each frame of the image recorded the spatial distribution of light intensity at the moment of laser pulse triggering. After processing by MATLAB, a two-dimensional dataset containing light field intensity and phase distribution was generated. The photoacoustic response signal was detected by a piezoelectric composite ultrasonic transducer with a center frequency of 50MHz (bandwidth 30-70MHz, axial resolution 75μm). Ultrasonic gel was filled between the transducer and the sample to reduce acoustic energy loss. The detected acoustic signal was amplified by a preamplifier (60dB gain, 1.2dB noise figure) and then acquired synchronously with the light field data by a high-speed oscilloscope (sampling rate 5GS / s, bandwidth 2GHz). In the experiment, three-dimensional scans were performed at 5-μm intervals on different depth areas of nude mouse skin (epidermis 0-50μm, superficial dermis 50-300μm, deep dermis 300-600μm, subcutaneous fat 600-1000μm), acquiring a total of 200 sets of spatiotemporal distribution data of light field and corresponding photoacoustic response signals. The signal amplitude in the deep region was attenuated by about 82% compared to the shallow region, and was accompanied by obvious phase distortion, which is consistent with the physical law that the light scattering coefficient of tissue increases with depth (the scattering coefficient of the deep dermis is more than 3 times higher than that of the epidermis).

[0076] Based on the collected raw data, the team extracted quantitative multidimensional characteristic parameters closely related to the composition and structure of skin tissue through a joint analysis model of spectral matching and polarized photoacoustic. The spectral matching process employed a multi-wavelength scanning strategy (selecting 11 characteristic wavelengths within the 245-275nm range) to calculate the correlation between the spectral curves of photoacoustic signals at each depth and a known standard absorption spectrum library of biomolecules. Nucleic acid components in the epidermal cell nucleus exhibit a characteristic absorption peak at 250nm, with a corresponding spectral matching coefficient of 0.91. Based on this, the "photoacoustic amplitude-wavelength response matrix" was extracted as a characterization parameter for nucleic acid concentration. Hemoglobin (oxygenation and deoxygenation) in the superficial dermis showed significant absorption at 266nm. Combining this with the linear correlation between photoacoustic signals and hemoglobin concentration confirmed by PubMed, two blood flow-related parameters, "blood oxygen saturation (sO2)" and "total hemoglobin concentration (tHb)," were calculated. Polarimetric photoacoustic analysis utilizes a four-quadrant all-Stokes polarization sorting device to measure the degree of polarization (DoP) and azimuth angle (AoP) of the photoacoustic signal. Collagen fibers in the deep dermis are oriented with a polarization degree of 0.72 and an azimuth angle deviation of less than 5°, while subcutaneous adipose tissue is an isotropic structure with a polarization degree of only 0.28. The difference between the two is quantified by the "polarization anisotropy factor." Simultaneously, combined with the propagation time difference of the photoacoustic signal, the tissue sound velocity (epidermis 1580 m / s, dermis 1620 m / s, subcutaneous fat 1450 m / s) and attenuation coefficient at different depths are calculated. This results in a 6-dimensional feature parameter set including "photoacoustic amplitude, spectral matching coefficient, blood oxygen saturation, degree of polarization, sound velocity, and attenuation coefficient." Each parameter is verified by Pearson correlation analysis (correlation coefficient R² > 0.92 with pathological slide measurements) to ensure its quantitative reliability.

[0077] To establish the mapping relationship between feature parameters and tissue depth, the team selected standard skin samples from 10 nude mice for systematic imaging measurements, constructing a feature parameter dataset for algorithm training. After standard samples were fixed with paraformaldehyde, the precise thickness, cell density, and other gold-standard data for each tissue layer were determined through pathological sections. Scanning was performed using the same imaging parameters as the experimental samples, resulting in 1000 sets of valid data (100 sets per nude mouse). The dataset was organized according to a three-dimensional structure of "depth-composition-parameter." For example, data entries for the epidermis (0-50μm) included depth coordinates, nucleic acid concentration (pathological value), and 6-dimensional feature parameters; the superficial dermis (50-300μm) was supplemented with pathological reference values ​​such as blood vessel density and hemoglobin concentration. To eliminate the influence of individual differences, the Z-score normalization method was used to normalize the feature parameters, and an outlier detection algorithm (based on the 3σ criterion) was used to remove 12 sets of abnormal data caused by equipment noise. The final dataset is divided into a training set (800 sets) and a validation set (200 sets). The training set is used to build a deep adaptation model, and the validation set is used to evaluate the algorithm performance. The completeness of the dataset and the accuracy of the annotations lay a solid foundation for subsequent image reconstruction. This process also solves the problem of the lack of quantitative correlation between feature parameters and tissue characteristics in traditional imaging.

[0078] Based on a feature parameter dataset, the team designed a hierarchical signal response adaptation link by linking the characteristics of light field distribution with the laws of sound propagation, and combined it with the hierarchical dynamic neural radiation field (NeRF) algorithm for large depth-of-focus image reconstruction. First, based on the structural characteristics of skin tissue and the laws of signal propagation, the imaging depth was divided into three layers: shallow layer (0-200μm) – weak light scattering, signal attenuation is mainly due to absorption; middle layer (200-600μm) – scattering and absorption work together, resulting in significant signal phase distortion; deep layer (600-1000μm) – strong scattering leads to a significant reduction in signal amplitude and an increase in noise. For each layer, the adaptation link adopted differentiated parameter configurations: a high-gain signal amplification module was used in the shallow layer to preserve fine structural information; a phase compensation algorithm was introduced in the middle layer to correct the propagation delay caused by uneven sound velocity; and a combined strategy of signal enhancement and noise suppression was adopted in the deep layer, using wavelet thresholding to improve the signal-to-noise ratio. The hierarchical dynamic NeRF algorithm uses the adapted signal as input to construct a three-dimensional radiation field model containing depth information. The model's network structure consists of a feature encoding layer, a spatial sampling layer, and an image reconstruction layer. The feature encoding layer maps 6-dimensional parameters to 256-dimensional feature vectors. The spatial sampling layer combines a Bessel-like light field distribution function to dynamically adjust the sampling density at different depths (the sampling interval for deeper layers is reduced to 2μm). The image reconstruction layer outputs pixel grayscale values ​​through a Sigmoid activation function. During algorithm training, the three-dimensional structure of pathological slides is used as the supervision signal, and the mean squared error (MSE) is used as the loss function. After 5000 iterations, the model converges (the training set loss value is reduced to 0.003), and real-time reconstruction at 2 frames per second is achieved through GPU acceleration (NVIDIA A100).

[0079] Experimental results show that the reconstructed nude mouse skin tissue images obtained by this method maintain excellent performance across the entire depth range: the epidermis (at 50 μm) can clearly distinguish the outline of cell nuclei with a diameter of about 1 μm, with a lateral resolution of 0.9 μm, which is on par with traditional OR-PAM; the capillary network (10 μm in diameter) in the superficial dermis (at 200 μm) is clearly imaged, with a blood oxygen saturation measurement error of only ±3%, which is better than the ±5% error of near-infrared photoacoustic imaging; the collagen fiber bundles in the deep dermis (at 600 μm) have a clear orientation, and the distribution of polarization anisotropy factors is highly consistent with the pathological staining results; even in the subcutaneous fat region at 1000 μm, the polygonal structure of adipocytes can still be distinguished, with a resolution maintained within 8 μm, while traditional Gaussian beam imaging cannot distinguish tissue morphology at this depth. Comparative experiments show that the depth of focus range (0-1000μm) of this method is more than 10 times that of traditional ultraviolet photoacoustic imaging, and the signal-to-noise ratio of cross-depth imaging remains above 25dB, which is 40% higher than the NeRF algorithm without layered adaptation. In repeatability tests, the structural similarity (SSIM) of 10 consecutive images of the same region reached 0.96, proving the stability and reliability of the method.

[0080] In summary, the embodiments of this application, by employing a large depth-of-focus ultraviolet photoacoustic microscopy imaging method, successfully overcome the limitations of traditional optical resolution photoacoustic microscopes, such as short depth of focus and difficulty in capturing multi-layered structures of thick tissues in a single operation. This enables label-free, section-free, three-dimensional subcellular resolution continuous observation within a depth range of 0-1000 μm. This not only significantly improves imaging efficiency but also, through spectral matching, polarization analysis, and layered algorithm reconstruction, accurately extracts the composition and structural features of each layer of skin tissue. This provides more efficient and reliable technical support for the pathological diagnosis of skin diseases, demonstrating its enormous potential to surpass traditional imaging methods in clinical applications.

[0081] Figure 9 A schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include: The memory 901, the processor 902, and the computer program stored on the memory 901 and capable of running on the processor 902.

[0082] When the processor 902 executes the program, it implements the large depth-of-focus ultraviolet photoacoustic microscopy imaging method provided in the above embodiments.

[0083] Furthermore, electronic devices also include: Communication interface 903 is used for communication between memory 901 and processor 902.

[0084] The memory 901 is used to store computer programs that can run on the processor 902.

[0085] The memory 901 may include high-speed RAM (Random Access Memory) memory, and may also include non-volatile memory, such as at least one disk storage.

[0086] If the memory 901, processor 902, and communication interface 903 are implemented independently, then the communication interface 903, memory 901, and processor 902 can be interconnected via a bus to complete communication between them. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 9 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0087] Optionally, in a specific implementation, if the memory 901, processor 902, and communication interface 903 are integrated on a single chip, then the memory 901, processor 902, and communication interface 903 can communicate with each other through an internal interface.

[0088] The processor 902 may be a CPU (Central Processing Unit), an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of this application.

[0089] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described large depth-of-focus ultraviolet photoacoustic microscopy imaging method.

[0090] Furthermore, this application also provides a computer program product, including a computer program or instructions, which, when executed, implement the above-described large depth-of-focus ultraviolet photoacoustic microscopy imaging method.

[0091] In the description of this specification, the references to "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0092] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0093] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0094] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any of the following techniques known in the art, or a combination thereof: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0095] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0096] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A large depth-of-focus ultraviolet photoacoustic microscopy imaging system, characterized in that, include: The system includes an ultraviolet photoacoustic signal acquisition module, a spectral-polarization feature extraction module, a cross-depth standard sample module, and an image reconstruction module; among these,... The ultraviolet photoacoustic signal acquisition module is used to acquire the spatiotemporal distribution data of ultraviolet excitation light and its corresponding photoacoustic response signal data; The spectral-polarization feature extraction module is used to extract quantitative multidimensional feature parameters related to tissue composition and structure through spectral matching and polarized photoacoustic joint analysis. The cross-depth standard sample module is used to perform systematic imaging measurements on standard biological tissue samples at different depths to form a feature parameter dataset. The image reconstruction module is used to link the spatial distribution characteristics of the ultraviolet excitation light field with the physical propagation law of photoacoustic signals at different depths, design a hierarchical signal response adaptation link, dynamically match the signal transmission loss differences across tissue depths, and perform signal attenuation correction through a hierarchical dynamic neural radiation field algorithm to reconstruct tissue microscopic images with large depth of focus in real time.

2. The ultraviolet photoacoustic microscopy imaging system with large depth of focus according to claim 1, characterized in that, The ultraviolet photoacoustic signal acquisition module includes an ultraviolet light emitting unit, a light field spatiotemporal detection unit, and a photoacoustic signal sensing unit. The ultraviolet light emitting unit is used to generate wavelength-tunable and power-stable ultraviolet excitation light and transmit it to the target tissue. The light field spatiotemporal detection unit is used to capture the spatial distribution and temporal evolution data of the ultraviolet excitation light in the tissue in real time. The photoacoustic signal sensing unit is used to directly acquire the original photoacoustic response signal generated after the tissue is stimulated.

3. The ultraviolet photoacoustic microscopy imaging system with large depth of focus according to claim 1, characterized in that, The spectral-polarization feature extraction module includes a spectral analysis unit, a polarization state modulation and detection unit, and a spectral-polarization joint analysis unit. The spectral analysis unit is used to perform spectral decomposition on the scattered light after the interaction between ultraviolet excitation light and tissue to obtain characteristic spectral information corresponding to tissue components. The polarization state modulation and detection unit is used to adjust the polarization parameters of the incident light and capture the polarization state change data of the reflected / scattered light from the tissue. The spectral-polarization joint analysis unit is used to fuse characteristic spectral and polarization state data to extract quantitative multidimensional feature parameters related to tissue component concentration and microstructure orientation.

4. The ultraviolet photoacoustic microscopy imaging system with large depth of focus according to claim 1, characterized in that, The cross-depth standard sample module includes a standard sample preparation and calibration unit, a layered imaging acquisition unit, and a parameter dataset integration unit. The standard sample preparation and calibration unit is used to construct standardized biological tissue samples covering different depths and known tissue components and structures, with quantitative multidimensional feature parameters as the core indicators, and to calibrate and verify the acquisition parameters using known sample information. The layered imaging acquisition unit performs precise imaging and photoacoustic signal acquisition of different depth regions of the sample. The parameter dataset integration unit organizes the feature parameters after calibration at each depth according to a unified standard to create a standardized cross-depth feature parameter dataset.

5. The ultraviolet photoacoustic microscopy imaging system with large depth of focus according to claim 1, characterized in that, The image reconstruction module includes a light field-propagation law linkage analysis unit, a hierarchical signal response adaptation unit, an attenuation correction unit, and a large depth-of-focus microscopic image real-time reconstruction unit. Specifically, the light field-propagation law linkage analysis unit extracts the spatial distribution characteristics of the ultraviolet excitation light field and, combined with a standardized cross-depth feature parameter dataset and the physical propagation laws of photoacoustic signals at different depths, establishes a cross-depth light field-photoacoustic propagation coupling correlation model. The hierarchical signal response adaptation unit constructs a dynamic adaptation link to accurately match loss characteristics based on differences in tissue signal transmission loss across depths. The attenuation correction unit adaptively corrects signal attenuation at different depths using a hierarchical dynamic neural radiation field algorithm. The large depth-of-focus microscopic image real-time reconstruction unit integrates the corrected signal with multi-dimensional feature parameters to output a tissue microscopic image with large depth-of-focus characteristics in real time.

6. The ultraviolet photoacoustic microscopy imaging system with large depth of focus according to claim 1, characterized in that, The hierarchical dynamic neural radiation field algorithm: ； ； ； in, To correct the attenuated photoacoustic signal intensity; Horizontal pixel coordinates; Vertical pixel coordinates; For organization depth coordinates; The original photoacoustic signal intensity; Let be the spatial distribution function of the ultraviolet excitation light field; It is a natural exponential function; For the first The starting depth of the layer; For the first Attenuation coefficient of layered photoacoustic signal; The coordinates of the infinitesimal element are in the depth direction; It is a tiny differential unit in the depth direction; A hierarchical dynamic neural radiation field model; For the first Layered multilayer perceptron; For feature encoding function; The pixel grayscale values ​​of the reconstructed image; This represents the total number of layers; These are the stratified weighting coefficients; For the first The center depth of the layer; It is a signal noise suppression factor; This is a noise suppression term.

7. A method for ultraviolet photoacoustic microscopy with large depth of focus, characterized in that, include: Acquire the spatiotemporal distribution data of ultraviolet excitation light and its corresponding photoacoustic response signal data; Based on the spatiotemporal distribution data of the ultraviolet excitation light and its corresponding photoacoustic response signal data, quantitative multidimensional feature parameters related to tissue composition and structure are extracted through spectral matching and polarization photoacoustic joint analysis. Based on quantitative multidimensional feature parameters, systematic imaging measurements were performed on standard biological tissue samples at different depths to form a feature parameter dataset. Based on the aforementioned feature parameter dataset, and by linking the spatial distribution characteristics of the ultraviolet excitation light field with the physical propagation laws of photoacoustic signals at different depths, a hierarchical signal response adaptation link is designed to dynamically match the differences in signal transmission loss across tissue depths. The signal attenuation is corrected through a hierarchical dynamic neural radiation field algorithm, and tissue microscopic images with large depth of focus are reconstructed in real time.

8. An electronic device, characterized in that, include: The memory, the processor, and the computer program stored in the memory and executable on the processor, the processor executing the program to implement the large depth-of-focus ultraviolet photoacoustic microscopy imaging method of claim 7.

9. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed, they implement the large depth-of-focus ultraviolet photoacoustic microscopy imaging method of claim 7.

10. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed, they implement the large depth-of-focus ultraviolet photoacoustic microscopy imaging method of claim 7.

Images