Method for detecting hypoxia condition in noninvasive solid tumor

Abstract

The invention relates to a noninvasive solid tumor internal hypoxia condition detection method, which comprises: chemically modifying a near-infrared two-region fluorophore, an oxygen sensitive group and a tumor targeting group to construct a composite probe, collecting fluorescence and photoacoustic signals of a tumor region after intravenous injection, carrying out de-noising analysis, and combining with multiple parameters to solve hypoxia concentration change, a four-dimensional space-time distribution map is constructed by fusing vascular topological data, and individualized calibration is realized through microfluidic simulation and Bayesian optimization. Compared with a traditional invasive detection and single-mode imaging technology, the advantages of bimodal signals are integrated in a breakthrough mode, non-invasive, real-time and high-resolution dynamic hypoxia monitoring is achieved, acute and chronic hypoxia areas are accurately distinguished, tumor blood vessel heterogeneity interference is reduced through personalized parameter calibration, and the accuracy of hypoxia monitoring is improved. High-precision spatio-temporal data support based on hypoxia microenvironment monitoring is provided for solid tumor radiotherapy target planning, chemotherapy drug delivery optimization, chemotherapy drug resistance and curative effect prediction and evaluation and tumor metastasis progress risk analysis.

Description

Technical Field

[0001] The present invention relates to the field of hypoxia detection, and particularly to a method for detecting the internal hypoxia status of non-invasive solid tumors. Background Art

[0002] The hypoxic microenvironment inside solid tumors is a key factor affecting cancer treatment resistance, metastasis, and prognosis. Accurately and dynamically detecting the spatio-temporal distribution of hypoxic regions within tumors is of great significance for tumor radiotherapy target area planning, optimization of chemotherapy drug delivery, prediction and evaluation of chemotherapy drug resistance and efficacy, and analysis of the risk of tumor metastasis progression. Currently, the commonly used clinical hypoxia detection methods mainly include invasive oxygen electrode measurement, ex vivo tissue immunohistochemical staining, and PET imaging technology based on nitroimidazole tracers. However, these methods have significant limitations in terms of real-time performance, spatial resolution, dynamic monitoring ability, and operability: both oxygen electrode measurement and immunohistochemical techniques require puncturing or excising tumor tissues. Since the tumor tissue specimens are detached from the body, they cannot truly reflect the actual status and dynamic changes of hypoxia in in vivo solid tumors. In addition, these methods are invasive examinations and have a relatively high surgical risk. Although PET imaging can non-invasively evaluate the degree of hypoxia, its spatial resolution is low (it is difficult to distinguish hypoxic regions within tumors from adjacent blood vessels or inflammatory infiltrations, and it is easy to miss early or scattered hypoxic regions), the tracer uptake rate is low, resulting in low accuracy in evaluating hypoxia inside solid tumors, it is difficult to distinguish the spatio-temporal heterogeneity between acute hypoxia (short-term blood flow interruption) and chronic hypoxia (long-term hypoxic microenvironment), and the time resolution is low (usually several hours to several days).

[0003] In the prior art, although non-invasive detection methods based on optical or photoacoustic imaging can partially overcome the above defects, they still face two major bottlenecks: firstly, traditional single-modal probes (such as those relying only on fluorescence or photoacoustic signals) are difficult to synchronously quantify hypoxia concentration and hemodynamic parameters, resulting in insufficient accuracy of dynamic calculation models; secondly, the lack of a calibration mechanism for patient individual characteristics makes the detection results vulnerable to tumor vascular heterogeneity (such as vascular density, permeability) and probe metabolism differences, and it is difficult to achieve high-confidence spatio-temporal mapping. Summary of the Invention

[0004] Based on this, the purpose of the present invention is to provide a method for detecting the internal hypoxia status of non-invasive solid tumors that can monitor the tumor hypoxia state in real time and dynamically, and integrates multi-modal signals and individual calibration.

[0005] The purpose of the present invention is achieved by the following solutions:

[0006] In the first aspect, the present invention provides a method for detecting the internal hypoxia status of non-invasive solid tumors, including:

[0007] S1: Chemically modify the near-infrared second-region fluorescent group, oxygen-sensitive nitroimidazole group, and tumor-targeting group to generate a composite probe;

[0008] Among them, the near-infrared second-region fluorescent group is a fluorescent molecular structure with a wavelength of 1000 - 1700 nm, the oxygen-sensitive nitroimidazole group is a chemical group that is reduced under hypoxic conditions, and the tumor-targeting group is an RGD polypeptide molecule connected by a covalent bond; the composite probe is used to obtain the fluorescent signal and photoacoustic signal in the near-infrared second region of the tumor area;

[0009] S2: Intravenously inject the composite probe, and synchronously collect the fluorescent signal and photoacoustic signal using the composite probe, preprocess the fluorescent signal to obtain the denoised fluorescent signal, and preprocess the photoacoustic signal to obtain the resolved optical signal;

[0010] S3: Generate the real-time hypoxia concentration change based on the time derivative of the denoised fluorescent signal, the blood flow parameters of the resolved optical signal, and the pre-calibration coefficient;

[0011] S4: Perform spatial superposition processing on the real-time hypoxia concentration change and the vascular topology data of the resolved optical signal to generate a four-dimensional hypoxia map, and the four-dimensional hypoxia map includes a dynamic distribution map of three-dimensional spatial coordinates and a time axis;

[0012] S5: Process the four-dimensional hypoxia map through in vitro microfluidic simulation and individualized Bayesian optimization to generate a corrected hypoxia concentration distribution.

[0013] In one embodiment, step S1 includes:

[0014] Chemically modify the iridium complex to generate a probe core structure. The iridium complex is a luminescent molecular structure centered on iridium metal, and the oxygen-sensitive nitroimidazole group and the tumor-targeting group are connected by a covalent bond;

[0015] Perform polyethylene glycol-liposome encapsulation treatment on the probe core structure, and modify the tumor-targeting group on the surface of the polyethylene glycol-liposome to generate a nano-probe with a long circulation time and active targeting function. The polyethylene glycol-liposome is a nano-carrier for extending the blood circulation time, and the nano-probe is used to indicate the intermediate product of the uncalibrated parameter;

[0016] Perform an in vitro hypoxia gradient experiment on the nano-probe to obtain the fluorescence activation threshold and the photoacoustic oxygen saturation sensitivity. The in vitro hypoxia gradient experiment is to test the probe performance in a controllable oxygen concentration chamber. The fluorescence activation threshold is the speed parameter of the probe's response to hypoxia, and the photoacoustic oxygen saturation sensitivity is the parameter of the influence of oxygen concentration on the photoacoustic signal;

[0017] Based on the fluorescence activation threshold and the photoacoustic oxygen saturation sensitivity, obtain the composite probe, and the composite probe includes the nano-probe, the fluorescence activation threshold, and the photoacoustic oxygen saturation sensitivity.

[0018] In one embodiment, an in vitro hypoxia gradient experiment is performed on the nanoprobe to obtain the fluorescence activation threshold and the photoacoustic oxygen saturation sensitivity, including:

[0019] Performing oxygen concentration gradient setting treatment on the chamber of the in vitro hypoxia gradient experiment to generate a continuous oxygen concentration gradient environment from a preset normoxia to a preset hypoxia;

[0020] Performing signal acquisition processing on the nanoprobe in a multi-oxygen concentration environment to generate a fluorescence-photoacoustic response curve;

[0021] Processing the fluorescence-photoacoustic response curve through first derivative analysis and exponential fitting to generate the fluorescence activation threshold and the photoacoustic oxygen saturation sensitivity;

[0022] Among them, the first derivative analysis is used to indicate the oxygen concentration corresponding to the maximum value of the derivative of the fluorescence response curve, and the exponential fitting is used to indicate the fitting of the relationship between the photoacoustic signal intensity and the oxygen concentration.

[0023] In one embodiment, processing the fluorescence-photoacoustic response curve through first derivative analysis and exponential fitting to generate the fluorescence activation threshold and the photoacoustic oxygen saturation sensitivity, including:

[0024] Performing first derivative calculation processing on the fluorescence response curve, extracting the maximum value of the fluorescence signal change rate, and calibrating the oxygen concentration corresponding to the maximum value as the fluorescence activation threshold;

[0025] Processing the photoacoustic response curve to generate the photoacoustic signal intensity and the oxygen concentration;

[0026] Using the following formula, extracting the fitting coefficient in the formula and calibrating it as the photoacoustic oxygen saturation sensitivity:

[0027]

[0028] Wherein, I PA is the photoacoustic signal intensity, I0 is the baseline intensity under normoxia conditions, [O2] is the oxygen concentration, and k2 is the fitting coefficient, that is, the photoacoustic oxygen saturation sensitivity.

[0029] In one embodiment, step S2 includes:

[0030] Injecting the composite probe into the human body through intravenous injection technology;

[0031] Collecting and processing the near-infrared second-region fluorescence signal of the tumor area, and using an InGaAs detector to obtain the fluorescence intensity at a preset high-frequency frame rate. The InGaAs detector is a semiconductor sensor for detecting near-infrared light;

[0032] Denoise the fluorescence intensity based on the background scattering correction technique to generate a denoised fluorescence signal. The background scattering correction technique is a mathematical filtering method used to indicate the removal of interference from tissue reflected light;

[0033] Generate a photoacoustic signal through laser excitation and ultrasonic detector reception and processing, and perform analysis processing on the photoacoustic signal to generate an analyzed photoacoustic signal. The analyzed photoacoustic signal includes vascular oxygen saturation and blood flow velocity. Vascular oxygen saturation is the proportion of oxyhemoglobin in the blood, and blood flow velocity is the rate of blood flow.

[0034] In one embodiment, step S3 includes:

[0035] Perform time derivative calculation processing on the denoised fluorescence signal to generate a signal change rate;

[0036] Use the following formula to calculate the signal change rate, the analyzed photoacoustic signal, and the pre-calibrated coefficient to generate a change amount of hypoxia concentration:

[0037]

[0038] Where Δ[O2](t) is the change amount of real-time hypoxia concentration, k1 is the fluorescence activation threshold of the composite probe, k2 is the photoacoustic oxygen saturation sensitivity of the composite probe, and α is the blood flow-hypoxia coupling coefficient, which is used to represent the interference weight of blood flow velocity change on hypoxia detection.

[0039] In one embodiment, step S4 includes:

[0040] Perform threshold segmentation processing on the change amount of real-time hypoxia concentration, define a preset acute hypoxia determination threshold to generate spatio-temporal marker points. The acute hypoxia determination threshold is a dynamic threshold set based on the hypoxia concentration change rate and the physiological steady state range, and is used to distinguish acute hypoxia and chronic hypoxia regions;

[0041] Perform three-dimensional reconstruction processing on the vascular topology data of the analyzed optical signal to generate a tumor vascular network model. The vascular network model is used to indicate a 3D image showing the vascular branch structure;

[0042] Overlay the spatio-temporal marker points and the tumor vascular network model to generate a four-dimensional hypoxia map.

[0043] In one embodiment, step S5 includes:

[0044] Perform microfluidic chip simulation processing on the analyzed optical signal and the denoised fluorescence signal to generate an in vitro four-dimensional hypoxia map. The microfluidic chip is a transparent plastic device that simulates the biological fluid environment;

[0045] Perform spatio-temporal consistency comparison processing on the four-dimensional hypoxia map and the in vitro four-dimensional hypoxia map to generate an in vitro-in vivo error rate;

[0046] Identify whether the in vitro-in vivo error rate exceeds a preset threshold. If it exceeds, trigger the Bayesian optimization algorithm to process the in vitro-in vivo error rate and generate optimized model parameters. The Bayesian optimization algorithm is an intelligent algorithm that adjusts parameters based on probability distributions;

[0047] Perform iterative correction processing on the dynamic calculation model based on the optimized model parameters to generate a corrected hypoxic concentration distribution. The iterative correction processing is a calculation process that gradually approaches the optimal value.

[0048] In one embodiment, the optimized model parameters are generated through the following steps:

[0049] Construct an objective function based on the in vitro-in vivo error rate. The objective function is the mapping relationship between the error rate, model parameters, and the acute hypoxia determination threshold;

[0050] Define the prior probability distributions of the model parameters and thresholds according to historical data or a preset range. The prior distributions include normal distributions or uniform distributions;

[0051] Perform iterative sampling on the objective function through the Bayesian optimization algorithm to generate new parameter combinations, and update the posterior probability distribution based on the sampling results to gradually narrow the parameter search space;

[0052] When the number of iterations reaches the preset upper limit or the error rate converges within the allowable range, output the optimized model parameters.

[0053] In a second aspect, the present invention provides a detection system for the internal hypoxia condition of a non-invasive solid tumor, including:

[0054] A probe construction module for chemically modifying a near-infrared second-region fluorescent group, an oxygen-sensitive nitroimidazole group, and a tumor-targeting group to generate a composite probe;

[0055] Among them, the near-infrared second-region fluorescent group is a fluorescent molecular structure with a wavelength of 1000 - 1700 nm, the oxygen-sensitive nitroimidazole group is a chemical group that is reduced under hypoxic conditions, and the tumor-targeting group is an RGD polypeptide molecule connected by a covalent bond; the composite probe is used to obtain the fluorescent signal and photoacoustic signal in the near-infrared second region of the tumor area;

[0056] A probe injection and imaging module for intravenously injecting the composite probe, synchronously collecting the fluorescent signal and photoacoustic signal using the composite probe, preprocessing the fluorescent signal to obtain a denoised fluorescent signal, and preprocessing the photoacoustic signal to obtain an analyzed optical signal;

[0057] A dynamic calculation module for generating a real-time hypoxic concentration change amount based on the time derivative of the denoised fluorescent signal, the blood flow parameters of the analyzed optical signal, and a pre-calibrated coefficient;

[0058] A spatio-temporal mapping module, which is used to perform spatial superposition processing on the real-time change amount of hypoxia concentration and the vascular topology data of the optical signal after analysis, to generate a four-dimensional hypoxia map, and the four-dimensional hypoxia map includes a dynamic distribution map of three-dimensional spatial coordinates and a time axis;

[0059] A verification and calibration module, which is used to process the four-dimensional hypoxia map through in vitro microfluidic simulation and individualized Bayesian optimization to generate a corrected hypoxia concentration distribution.

[0060] In summary, the above-provided method for detecting the internal hypoxia condition of a non-invasive solid tumor aims at the influence of the internal hypoxia microenvironment of the solid tumor on cancer treatment resistance and prognosis. By chemical modification, a composite probe is generated, and it is used to synchronously collect the near-infrared second-region fluorescence signal and the photoacoustic signal in the tumor area. After preprocessing, the real-time change amount of hypoxia concentration is generated by combining the blood flow parameter and the pre-calibrated coefficient, and is superimposed with the vascular topology data to form a four-dimensional hypoxia map. Then, through in vitro microfluidic simulation and individualized Bayesian optimization and correction, the high-sensitivity, high-resolution, and real-time dynamic monitoring of the tumor hypoxia state is realized, overcoming the limitations of the existing detection means in terms of real-time performance, spatial resolution, and dynamic monitoring ability, and providing high-precision spatio-temporal data support for radiotherapy target area planning, chemotherapy drug delivery optimization, chemotherapy drug resistance and efficacy prediction and evaluation, and tumor metastasis progression risk analysis based on hypoxia microenvironment monitoring. At the same time, the radiation dose of fluorescence detection using the composite probe is much smaller than the radiation dose brought by PET / CT imaging detection in the prior art, which can significantly reduce the radiation risk to patients and medical staff and improve the safety of cancer treatment.

[0061] For better understanding and implementation, the present invention will be described in detail below with reference to the accompanying drawings. Description of the Drawings

[0062] Figure 1 It is a schematic flowchart of a method for detecting the internal hypoxia condition of a non-invasive solid tumor provided by an embodiment of the present application;

[0063] Figure 2 It is a schematic flowchart of obtaining the fluorescence activation threshold and the photoacoustic oxygen saturation sensitivity provided by an embodiment of the present application;

[0064] Figure 3 It is a schematic structural diagram of a system for detecting the internal hypoxia condition of a non-invasive solid tumor provided by an embodiment of the present application. Detailed Embodiment

[0065] To facilitate the understanding of the present invention, the present invention will be described more comprehensively below with reference to the relevant drawings. Preferred embodiments of the present invention are shown in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, these embodiments are provided to make the understanding of the disclosure of the invention more thorough and comprehensive.

[0066] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which the present invention belongs. The terms used in the specification of the present invention herein are for the purpose of describing specific embodiments only and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0067] In one embodiment, as Figure 1 shown, a method for detecting the internal hypoxia condition of a non-invasive solid tumor is provided. In this embodiment, the application of this method to a terminal is taken as an example for illustration. It can be understood that this method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is realized through the interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0068] S1: Chemically modify a near-infrared second-region fluorescent group, an oxygen-sensitive nitroimidazole group, and a tumor-targeting group to generate a composite probe.

[0069] Among them, the near-infrared second-region fluorescent group is a fluorescent molecular structure with a wavelength of 1000 - 1700 nm, and the oxygen-sensitive nitroimidazole group is a chemical group that is reduced under hypoxic conditions; the tumor-targeting group is an RGD polypeptide molecule connected by a covalent bond, and the composite probe is used to obtain the fluorescent signal and photoacoustic signal in the near-infrared second region of the tumor region.

[0070] Specifically, in the process of generating the composite probe, it is first necessary to carefully select appropriate near-infrared second-region fluorescent groups and oxygen-sensitive nitroimidazole groups. The wavelength range of the near-infrared second-region fluorescent group is 1000 - 1700 nm. The fluorescent molecular structure in this wavelength band has good tissue penetration ability and low biological autofluorescence background, making it have unique advantages in in-vivo imaging and being able to obtain more clear fluorescent signals in the internal region of the tumor. The oxygen-sensitive nitroimidazole group is a chemical group that can be reduced under hypoxic conditions, and its reduction degree is closely related to the oxygen concentration in the surrounding environment, and can be used to monitor the hypoxia state inside the solid tumor tissue; the tumor-targeting group is an RGD polypeptide molecule connected by a covalent bond. The RGD polypeptide can specifically recognize and bind to the integrin receptor on the surface of tumor cells, improving the enrichment efficiency of the probe in the tumor region. This targeting property of the RGD polypeptide enables the composite probe to accurately locate tumor cells and enhance the imaging and treatment effects.

[0071] These three groups are combined together through chemical modification to generate a composite probe. Preferably, the methods of chemical modification may include covalent bonding, embedding reaction, etc., to ensure the stability and activity of each group in the probe. The generated composite probe not only has the ability of near-infrared second-window fluorescence imaging, but also can specifically respond to the hypoxia situation in the internal area of the tumor, and achieve precise localization of the tumor area through the tumor-targeting group. Thus, synchronous acquisition of near-infrared second-window fluorescence signals and photoacoustic signals in the internal area of the tumor is realized, providing a basis for subsequent determination of hypoxia concentration and generation of four-dimensional hypoxia maps.

[0072] In addition, since the fluorescence detection of the composite probe is based on Fluorescence Resonance Energy Transfer (FRET) technology, its radiation dose is almost negligible, significantly lower than that of PET / CT imaging technology. Specifically, the radiation dose of PET / CT examination mainly comes from the CT part, and the radiation dose of a single examination is about 10 - 15 mSv, while the radiation source of the fluorescence detection of the composite probe is a low-intensity excitation light source, and its radiation dose is much smaller than the large radiation dose brought by CT detection. In contrast, the radiation dose of the composite probe is negligible. When using the composite probe for detection, patients do not have to worry about the increased cancer risk caused by the radiation from the detection imaging, nor do they have to consider the radiation impact on the surrounding people after the examination. For patients who need to frequently monitor tumors, the low-radiation characteristic of the composite probe more prominently shows its advantages, which can reduce the cumulative radiation dose caused by multiple examinations and further protect the health of patients. At the same time, medical staff do not need to take complex radiation protection measures during the operation, reducing the risk of occupational exposure.

[0073] S2: Inject the composite probe intravenously, and use the composite probe to synchronously collect fluorescence signals and photoacoustic signals, and preprocess the fluorescence signals to obtain denoised fluorescence signals, and preprocess the photoacoustic signals to obtain resolved optical signals.

[0074] Specifically, when implementing this process, first prepare the composite probe into a suitable solution and inject it into the organism by intravenous injection. It should be noted that the injection dose and rate need to be precisely controlled to ensure that the composite probe can be fully distributed to the tumor area and reach an effective concentration. After the composite probe enters the body, it will reach the tumor site through blood circulation, accurately locate tumor cells under the action of the tumor-targeting group, and play the roles of fluorescence imaging and photoacoustic imaging through its near-infrared second-window fluorescence group and oxygen-sensitive nitroimidazole group respectively.

[0075] Preferably, when collecting signals, a highly sensitive fluorescence detector and a photoacoustic imaging system can be used to synchronously obtain the fluorescence signal and the photoacoustic signal in the tumor region. Due to various interference factors in the living body, such as background noise, scattered light, etc., the collected signals need to be preprocessed. For the fluorescence signal, denoising algorithms such as digital filtering and wavelet transform can be used to remove noise interference and obtain a clear and accurate denoised fluorescence signal; for the photoacoustic signal, the information related to optical absorption can be extracted through a signal analysis algorithm to obtain the analyzed optical signal, so as to more accurately reflect the optical characteristics of the tumor region and provide reliable data support for subsequent hypoxia concentration calculation and blood vessel topology data analysis.

[0076] S3: Generate the real-time hypoxia concentration change based on the time derivative of the denoised fluorescence signal, the blood flow parameters of the analyzed optical signal, and the pre-calibration coefficient.

[0077] Specifically, this process involves the comprehensive application of multiple key parameters. The time derivative of the denoised fluorescence signal reflects the change rate of the fluorescence signal over time, and the change of the fluorescence signal is closely related to the change of the oxygen concentration in the tumor region because the reduction degree of the oxygen-sensitive nitroimidazole group affects the fluorescence intensity of the fluorophore. The blood flow parameters in the analyzed optical signal provide information related to the blood flow in the tumor blood vessels, such as blood flow velocity, blood flow volume, etc., and these parameters are closely related to the delivery and consumption of oxygen. The pre-calibration coefficient is a probe characteristic parameter pre-calibrated through in vitro experiments and is used to establish a quantitative relationship between the fluorescence signal, the blood flow parameters, and the hypoxia concentration.

[0078] Specifically, by experimentally measuring the fluorescence signal intensity of the composite probe and the corresponding blood flow parameter values at different oxygen concentrations, a series of data points are obtained, and then methods such as linear regression and non-linear fitting are used to determine the pre-calibration coefficient. In practical applications, the time derivative of the real-time obtained denoised fluorescence signal and the blood flow parameters of the analyzed optical signal are substituted into the mathematical model determined by the pre-calibration coefficient, and the real-time hypoxia concentration change can be generated through calculation, so as to realize the dynamic monitoring of the hypoxia state in the tumor region and provide an important basis for subsequent hypoxia map generation and treatment strategy adjustment.

[0079] S4: Perform spatial superposition processing on the real-time hypoxia concentration change and the blood vessel topology data of the analyzed optical signal to generate a four-dimensional hypoxia map, and the four-dimensional hypoxia map includes a dynamic distribution map of three-dimensional spatial coordinates and a time axis.

[0080] Specifically, the spatial superposition processing is a key step in fusing the real-time hypoxia concentration change and the blood vessel topology data in the analyzed optical signal. The blood vessel topology data in the analyzed optical signal contains spatial information such as the distribution, morphology, and orientation of blood vessels in the tumor region, and these information are crucial for accurately understanding the specific location and scope of the hypoxia region in the tumor.

[0081] By adopting algorithms such as image registration and three-dimensional reconstruction, the real-time change amount of hypoxia concentration is accurately aligned and superimposed with the vascular topology data in space, so that each hypoxia concentration value corresponds to a specific vascular position and the surrounding tissue structure. The generated four-dimensional hypoxia map not only contains three-dimensional spatial coordinate information, but also adds the dynamic distribution of the time axis, and can intuitively display the evolution process of the hypoxia region in the tumor over time, including dynamic changes such as the expansion, contraction, and migration of the hypoxia region, providing a comprehensive and intuitive visualization tool for studying the hypoxia physiological mechanism of the tumor, evaluating the treatment effect, and formulating personalized treatment plans.

[0082] S5: Process the four-dimensional hypoxia map through in vitro microfluidic simulation and individualized Bayesian optimization to generate a corrected hypoxia concentration distribution.

[0083] Specifically, in vitro microfluidic simulation is a technique that uses a microfluidic device to simulate the tumor blood flow environment. In this process, a microfluidic chip similar to the in vivo tumor vascular structure and blood flow characteristics is constructed, and a composite probe is injected into the microfluidic channel in the chip to simulate the in vivo blood circulation and oxygen delivery process. By monitoring the changes in fluorescence signals and photoacoustic signals in the chip, hypoxia concentration data in the in vitro simulation environment is obtained and used as a reference standard.

[0084] Individualized Bayesian optimization is a parameter adjustment algorithm based on a probability model. Its core idea is to continuously update the estimation of model parameters based on existing data to achieve the optimization of the model. When processing the four-dimensional hypoxia map, the hypoxia concentration data obtained from in vitro microfluidic simulation is used as prior knowledge, combined with the four-dimensional hypoxia map data measured in vivo, and the hypoxia concentration distribution parameters in the map are adjusted and corrected through the Bayesian optimization algorithm. This algorithm can fully consider individual differences and measurement errors, and continuously iterate and optimize, so that the finally generated corrected hypoxia concentration distribution can more accurately and reliably reflect the real hypoxia state in the tumor, providing a more scientific basis for subsequent precision medical decisions.

[0085] In summary, the above-provided method for detecting the internal hypoxia status of non-invasive solid tumors aims at the impact of the internal hypoxic microenvironment of solid tumors on cancer treatment resistance and prognosis. By chemically modifying to generate a composite probe, it synchronously collects the second near-infrared fluorescence signal and the photoacoustic signal in the tumor region, generates the real-time hypoxia concentration change amount by combining the blood flow parameters and the pre-calibrated coefficient after preprocessing, and forms a four-dimensional hypoxia map by superimposing with the vascular topology data. Then, through in vitro microfluidic simulation and individualized Bayesian optimization and correction, it realizes the high-sensitivity, high-resolution, and real-time dynamic monitoring of the tumor hypoxia status, overcomes the limitations of existing detection means in terms of real-time performance, spatial resolution, and dynamic monitoring ability, and provides high-precision spatio-temporal data support for radiotherapy target area planning of solid tumors, optimization of chemotherapy drug delivery, prediction and evaluation of chemotherapy drug resistance and efficacy, and analysis of the risk of tumor metastasis progression based on hypoxia microenvironment monitoring. At the same time, the radiation dose of fluorescence detection using the composite probe is much smaller than the radiation dose brought by PET / CT imaging detection in the prior art, which can significantly reduce the radiation risk to patients and medical staff and improve the safety of cancer treatment.

[0086] In one embodiment, as Figure 2 shown, step S1 includes:

[0087] S110. Chemically modify the iridium complex to generate the probe core structure. The iridium complex is a luminescent molecular structure centered on iridium metal, and an oxygen-sensitive nitroimidazole group and a tumor-targeting group are connected by covalent bonds.

[0088] Specifically, in this process, first, a luminescent molecular structure centered on iridium metal is selected as the base material. This iridium complex has good luminescent properties and can emit fluorescence in the second near-infrared range, providing a clear signal basis for subsequent imaging monitoring. Then, the nitroimidazole group is connected to the iridium complex by covalent bonds. The nitroimidazole group is a chemical group that can be reduced under hypoxic conditions, and its reduction degree is closely related to the oxygen concentration in the surrounding environment. When in a hypoxic state, the nitroimidazole group will undergo chemical changes, which in turn affect the fluorescence and photoacoustic properties of the entire probe, enabling the probe to specifically respond to the hypoxic environment and thus monitor the hypoxia status in the tumor region; the tumor-targeting group can specifically recognize and bind to specific markers on the surface of tumor cells, improving the enrichment efficiency of the probe in the tumor region. By chemically modifying to combine these three groups together, the probe core structure is generated, providing a basis for the subsequent preparation of the nanoprobe.

[0089] S120. The core structure of the probe is encapsulated with polyethylene glycol-liposomes, and tumor-targeting groups are modified on the surface of the polyethylene glycol-liposomes to generate a nano-probe with a long circulation time and active targeting function. The polyethylene glycol-liposomes are nano-carriers for extending the blood circulation time, and the nano-probe is used to indicate the intermediate product of the uncalibrated parameter.

[0090] Specifically, in order to improve the stability and circulation time of the probe in vivo, the generated probe core structure is encapsulated with polyethylene glycol-liposomes. Polyethylene glycol-liposomes are a commonly used nano-carrier material with good biocompatibility and long-circulation characteristics. Through a specific encapsulation process, the probe core structure is uniformly encapsulated inside the polyethylene glycol-liposomes to form nano-probes with uniform particle sizes. At the same time, tumor-targeting groups are modified again on the surface of the polyethylene glycol-liposomes to further enhance the active targeting function of the probe.

[0091] This kind of nano-probe can not only effectively protect the probe core structure from the influence of the complex in vivo environment, such as enzyme degradation, oxidation, etc., extend its time in blood circulation, but also actively recognize and bind to tumor cells, improving the accumulation and imaging effect of the probe in the tumor region. Thereby improving the enrichment efficiency of the composite probe in the tumor region, reducing the non-specific binding of the probe to non-target tissues, reducing background interference, and improving the signal-to-noise ratio of imaging.

[0092] S130. An in vitro hypoxia gradient experiment is carried out on the nano-probe to obtain the fluorescence activation threshold and the photoacoustic oxygen saturation sensitivity. The in vitro hypoxia gradient experiment is to test the performance of the probe in a controllable oxygen concentration chamber. The fluorescence activation threshold is the speed parameter for the probe to respond to hypoxia, and the photoacoustic oxygen saturation sensitivity is the influence parameter of oxygen concentration on the photoacoustic signal.

[0093] Specifically, in order to accurately calibrate the performance parameters of the nano-probe so that it can accurately reflect the hypoxia state in the tumor region in subsequent in vivo experiments, the prepared nano-probe is subjected to an in vitro hypoxia gradient experiment. Specifically, the nano-probe is placed in a series of controllable oxygen concentration chambers with different oxygen concentrations to simulate the environments with different degrees of hypoxia in vivo. Under these different oxygen concentration conditions, the fluorescence characteristics and photoacoustic characteristics of the nano-probe are tested and recorded in detail.

[0094] The system determines the fluorescence activation threshold of the nanoprobe by analyzing the trend and rate of change of the fluorescence signal intensity with oxygen concentration, that is, the minimum oxygen concentration required for the probe to start producing significant changes in the fluorescence signal in a hypoxic environment. This parameter reflects the sensitivity and response speed of the probe to the hypoxic environment. At the same time, by measuring the relationship between the photoacoustic signal intensity and the oxygen concentration, the photoacoustic oxygen saturation sensitivity is obtained, that is, the degree of influence of the oxygen concentration change on the photoacoustic signal intensity. This parameter is used to quantitatively describe the response characteristics of the nanoprobe to the photoacoustic signal at different oxygen concentrations, providing a basis for subsequent inversion of the oxygen concentration using the photoacoustic signal. These in vitro experimental data not only help optimize the preparation process and usage conditions of the nanoprobe, but also provide important reference for subsequent in vivo hypoxia monitoring, ensuring that the nanoprobe can accurately and reliably reflect the hypoxic state of the tumor area in practical applications.

[0095] Preferably, the fluorescence activation threshold and the photoacoustic oxygen saturation sensitivity are obtained by the following method:

[0096] S131. Perform oxygen concentration gradient setting on the chamber of the in vitro hypoxia gradient experiment to generate a continuous oxygen concentration gradient environment from the preset normoxia to the preset hypoxia.

[0097] Specifically, before conducting the in vitro hypoxia gradient experiment, it is necessary to carefully design and set the oxygen concentration gradient of the experimental chamber. First, determine the preset normoxia and hypoxia oxygen concentration ranges. Normoxia usually refers to an environment close to the atmospheric oxygen concentration, while hypoxia is set at a lower oxygen concentration level according to research needs. Then, use professional gas control systems and oxygen concentration adjustment equipment, such as gas mixers, flow controllers, etc., to accurately introduce different proportions of oxygen and other gases (such as nitrogen, carbon dioxide, etc.) into the experimental chamber to generate a continuous oxygen concentration gradient environment that gradually transitions from the preset normoxia to the preset hypoxia.

[0098] During this process, it is necessary to monitor the oxygen concentration in the chamber in real time to ensure that it changes accurately according to the set gradient and remains stable. At the same time, in order to simulate the complex hypoxic microenvironment in vivo, some matrix components similar to tumor tissues, such as extracellular matrix mimics, specific nutrients, etc., can also be added to the chamber to improve the physiological relevance of the experiment.

[0099] S132. Perform signal acquisition and processing of the nanoprobe in an environment with multiple oxygen concentrations to generate a fluorescence-photoacoustic response curve.

[0100] Specifically, connect the experimental chamber that has undergone oxygen concentration gradient setting treatment to a highly sensitive fluorescence and photoacoustic signal acquisition system. Uniformly disperse the prepared nanoprobes in the experimental chamber to ensure that they are fully exposed to different oxygen concentration environments. Then, at each set oxygen concentration point, use an appropriate excitation light source to excite the fluorescence emission of the nanoprobes, and use a photoacoustic imaging system to detect the generated photoacoustic signals. When collecting signals, it is necessary to strictly control experimental conditions such as temperature, pH value, etc. to ensure the stable performance of the nanoprobes. For each oxygen concentration point, conduct multiple repeated measurements to obtain reliable mean values and standard deviations. Finally, correlate the collected fluorescence signal intensities and photoacoustic signal intensities with the corresponding oxygen concentration values respectively, and plot a fluorescence-photoacoustic response curve. This curve intuitively shows the change trends of the fluorescence and photoacoustic signals of the nanoprobes at different oxygen concentrations, providing a basis for subsequent parameter extraction and performance evaluation.

[0101] S133. Process the fluorescence-photoacoustic response curve through first derivative analysis and exponential fitting to generate a fluorescence activation threshold and a photoacoustic oxygen saturation sensitivity.

[0102] Specifically, among them, first derivative analysis is used to indicate the oxygen concentration corresponding to the maximum value of the derivative of the fluorescence response curve, and exponential fitting is used to indicate the fitting of the relationship between the photoacoustic signal intensity and the oxygen concentration. In order to extract the key performance parameters, namely the fluorescence activation threshold and the photoacoustic oxygen saturation sensitivity, from the fluorescence-photoacoustic response curve, the methods of first derivative analysis and exponential fitting are used to process the curve. Preferably, the specific steps of processing the fluorescence-photoacoustic response curve through first derivative analysis and exponential fitting to generate a fluorescence activation threshold and a photoacoustic oxygen saturation sensitivity are as follows:

[0103] S1331. Conduct a first derivative calculation and processing on the fluorescence response curve, extract the maximum value of the fluorescence signal change rate, and calibrate the oxygen concentration corresponding to the maximum value as the fluorescence activation threshold.

[0104] Specifically, first, conduct a first derivative calculation on the fluorescence response curve, that is, obtain the derivative curve of the fluorescence signal intensity with respect to the change in oxygen concentration. On this derivative curve, find the point where the derivative value reaches the maximum, and the oxygen concentration corresponding to this point is the fluorescence activation threshold. This is because at this oxygen concentration, the fluorescence signal of the nanoprobe is most sensitive to the change in oxygen concentration, and the signal change rate is the highest, marking the critical oxygen concentration at which the nanoprobe begins to respond significantly to the hypoxic environment. The fluorescence activation threshold extracted by this method can accurately reflect the response characteristics of the nanoprobe to the hypoxic environment, providing an important reference index for subsequent hypoxic monitoring.

[0105] S1332. Process the photoacoustic response curve to generate the photoacoustic signal intensity and the oxygen concentration.

[0106] Specifically, when processing the photoacoustic response curve, first, the collected photoacoustic signal intensity data is sorted and matched with the corresponding oxygen concentration values to generate data pairs of photoacoustic signal intensity and oxygen concentration. The purpose of this step is to establish a direct correlation between the photoacoustic signal intensity and oxygen concentration, providing a data basis for subsequent exponential fitting. During the process of generating data pairs, preliminary screening and calibration of the data are required to remove possible outliers and data points with large errors, so as to improve the quality and reliability of the data.

[0107] S1333. Use the following formula, and calibrate the fitting coefficient extracted from the formula as the photoacoustic oxygen saturation sensitivity:

[0108]

[0109] where I PA is the photoacoustic signal intensity, I0 is the baseline intensity under normoxic conditions, [O2] is the oxygen concentration, and k2 is the fitting coefficient, that is, the photoacoustic oxygen saturation sensitivity.

[0110] S140. Based on the fluorescence activation threshold and the photoacoustic oxygen saturation sensitivity, a composite probe is obtained. The composite probe includes a nanoprobe, a fluorescence activation threshold, and a photoacoustic oxygen saturation sensitivity.

[0111] Specifically, after completing the above in vitro hypoxia gradient experiment and data processing, the nanoprobe is integrated with the fluorescence activation threshold and photoacoustic oxygen saturation sensitivity obtained through experiments to form a complete composite probe. Specifically, the composite probe not only includes the nanoprobe encapsulated by polyethylene glycol-liposome, but also includes two key parameters, the fluorescence activation threshold and the photoacoustic oxygen saturation sensitivity, calibrated through in vitro experiments.

[0112] The fluorescence activation threshold is used to indicate the critical oxygen concentration at which the nanoprobe begins to produce significant fluorescence signal changes in a hypoxic environment, while the photoacoustic oxygen saturation sensitivity is used to quantitatively describe the degree of influence of oxygen concentration changes on the photoacoustic signal intensity. The addition of these two parameters enables the composite probe to more accurately reflect the hypoxic state of the tumor region in practical applications, providing a necessary basis for subsequent determination of hypoxic concentration and generation of four-dimensional hypoxia maps. This integrated characteristic of the composite probe not only improves its accuracy and reliability in tumor hypoxia monitoring, but also provides a more scientific basis for the formulation and optimization of clinical treatment plans.

[0113] In summary, the above-provided method for detecting the internal hypoxia status of non-invasive solid tumors constructs an oxygen concentration gradient environment, systematically studies the fluorescence and photoacoustic signals of the nanosensor at different oxygen concentrations, determines its fluorescence activation threshold and photoacoustic oxygen saturation sensitivity, and finally obtains a composite sensor with clear response characteristics. This series of steps lays a solid foundation for the application of the composite sensor in the biomedical field, enabling it to more accurately reflect the changes in oxygen content in biological tissues, and providing high-precision spatio-temporal data support for radiotherapy target area planning of solid tumors, optimization of chemotherapy drug delivery, prediction and evaluation of chemotherapy drug resistance and efficacy, and risk analysis of tumor metastasis progression based on the monitoring of the hypoxic microenvironment.

[0114] In one embodiment, step S2 includes:

[0115] S210. Inject the composite sensor into the human body through intravenous injection technology.

[0116] Specifically, in this process, first, the composite sensor is formulated into a suitable injection solution to ensure that its concentration and dosage meet clinical requirements. Through intravenous injection technology, the composite sensor is quickly and evenly injected into the human blood circulation. During the injection process, it is necessary to strictly control the injection rate to ensure that the composite sensor can be fully distributed throughout the body, especially in the tumor area. At the same time, the injection operation needs to be carried out under aseptic conditions to avoid the occurrence of complications such as infection.

[0117] S220. Collect and process the fluorescence signal in the second near-infrared region of the tumor area, and use an InGaAs detector to obtain the fluorescence intensity at a preset high-frequency frame rate. The InGaAs detector is a semiconductor sensor for detecting near-infrared light.

[0118] Specifically, after the composite sensor is injected into the human body, the fluorescence signal in the second near-infrared region of the tumor area is collected by a highly sensitive InGaAs detector. The InGaAs detector is a semiconductor sensor specifically designed for detecting near-infrared light, with the characteristics of high quantum efficiency and low noise, and can accurately capture the weak fluorescence signal emitted by the composite sensor. To monitor the changes in the fluorescence signal in real time, the detector obtains the signal at a preset high-frequency frame rate, usually with the frame rate set above 30 frames per second to ensure that rapid physiological changes and dynamic changes in the fluorescence signal can be captured.

[0119] S230. Denoise the fluorescence intensity based on background scattering correction technology to generate a denoised fluorescence signal. The background scattering correction technology is a mathematical filtering method used to indicate the removal of tissue reflected light interference.

[0120] Specifically, due to the scattering and absorption of light by biological tissues, the collected fluorescence signals will contain certain background noise and scattered light interference. To improve the purity and accuracy of the signals, background scattering correction technology can be used to denoise the fluorescence intensity.

[0121] Background scattering correction technology is a mathematical filtering method. By establishing a background scattering model, the collected fluorescence signals are filtered to remove the background noise and scattered light components. Commonly used correction methods include wavelet transform filtering, polynomial fitting correction, etc. After denoising, the obtained denoised fluorescence signals can more truly reflect the fluorescence intensity changes of the composite probe in the tumor region, providing a reliable data basis for subsequent hypoxia concentration calculation.

[0122] S240. Generate photoacoustic signals through laser excitation and ultrasonic detector reception and processing, and perform analytical processing on the photoacoustic signals to generate the analyzed photoacoustic signals. The analyzed photoacoustic signals include vascular oxygen saturation and blood flow velocity. The vascular oxygen saturation is the proportion of oxyhemoglobin in the blood, and the blood flow velocity is the rate of blood flow.

[0123] Specifically, to obtain the photoacoustic signals of the tumor region, first irradiate the tumor region with a laser of a specific wavelength. After the laser energy is absorbed by the optical absorption substances (such as hemoglobin) in the tissue, it causes local temperature increase and volume expansion, thus generating ultrasonic signals. These photoacoustic signals are received and converted by a high-sensitivity ultrasonic detector. The ultrasonic detector converts the received photoacoustic signals into electrical signals and performs preliminary amplification and filtering processing.

[0124] Next, perform analytical processing on the collected photoacoustic signals to extract the useful information therein. The analytical processing includes steps such as signal filtering, amplification, and analog-to-digital conversion to improve the signal quality and resolution. Then, analyze the processed photoacoustic signals through specific algorithms to calculate parameters such as vascular oxygen saturation and blood flow velocity. The vascular oxygen saturation reflects the proportion of oxyhemoglobin in the blood and is an important indicator for measuring tissue oxygen supply; the blood flow velocity reflects the rate of blood flow in the blood vessels and is closely related to the tissue blood perfusion situation. The acquisition of these parameters provides important data support for subsequent hypoxia concentration calculation and tumor hemodynamics analysis.

[0125] In summary, the method for detecting the internal hypoxia condition of non-invasive solid tumors provided by the present invention injects a composite probe intravenously, and uses an InGaAs detector and an ultrasonic detector to collect the near-infrared second-region fluorescence signal and the photoacoustic signal of the tumor region respectively, and through background scattering correction technology and signal analysis and processing, parameters such as denoised fluorescence signal, vascular oxygen saturation and blood flow velocity are generated. This series of steps realizes high-sensitivity, high-resolution, real-time dynamic monitoring of the tumor region, overcomes the limitations of existing clinical detection means in terms of real-time performance, spatial resolution and dynamic monitoring ability, and provides high-precision spatio-temporal data support based on hypoxia microenvironment monitoring for radiotherapy target area planning of solid tumors, optimization of chemotherapy drug delivery, prediction and evaluation of chemotherapy drug resistance and efficacy, and analysis of the risk of tumor metastasis progression.

[0126] In one embodiment, step S3 includes:

[0127] S310. Perform time derivative calculation processing on the denoised fluorescence signal to generate a signal change rate.

[0128] Specifically, after the system completes the acquisition and denoising processing of the near-infrared second-region fluorescence signal of the tumor region, the denoised fluorescence signal is obtained. In order to obtain the change of the fluorescence signal over time, it is necessary to perform time derivative calculation on the denoised fluorescence signal. Specifically, the time derivative calculation can be realized by numerical differentiation method. Preferably, the difference method can be adopted:

[0129]

[0130] where Δt is the time step, usually determined according to the frame rate of signal acquisition, is the change rate of the fluorescence signal, that is, the signal change rate, which reflects the change speed of the fluorescence signal intensity over time. This signal change rate is crucial for the subsequent calculation of hypoxia concentration because it is directly related to the response dynamics of the probe in the hypoxic environment.

[0131] S320. Use the following formula to calculate the signal change rate, the resolved photoacoustic signal and the pre-calibrated coefficient to generate the change amount of hypoxia concentration:

[0132]

[0133] where Δ[O2](t) is the change amount of real-time hypoxia concentration, k1 is the fluorescence activation threshold of the composite probe, k2 is the photoacoustic oxygen saturation sensitivity of the composite probe, α is the blood flow-hypoxia coupling coefficient, used to represent the interference weight of blood flow velocity change on hypoxia detection, S PA (t) is the vascular oxygen saturation, and v(t) is the blood flow velocity.

[0134] This embodiment concisely and precisely describes the relationship between the signal change rate, the resolved photoacoustic signal, the pre-calibration coefficient, and the change in hypoxia concentration using this calculation formula. The following are the explanations of the relevant parameters:

[0135] is the fluorescence signal change rate term. This term converts the change rate of the fluorescence signal into a quantity related to the change in oxygen concentration. Since the change in the fluorescence signal is related to the activation state of the probe (i.e., the degree of hypoxia), by dividing by the fluorescence activation threshold k1 of the composite probe, the change rate of the fluorescence signal can be normalized to the change in oxygen concentration.

[0136] is the oxygen saturation correction term. This term is used to correct the influence of vascular oxygen saturation on the calculation of hypoxia concentration. The higher the oxygen saturation, the more oxygen available in the blood and the relatively lower the degree of hypoxia. By correcting the oxygen saturation through an exponential function, the actual hypoxia state can be more accurately reflected.

[0137] is the blood flow velocity change correction term. This term is used to correct the interference of blood flow velocity change on hypoxia detection. The change in blood flow velocity affects the oxygen delivery and consumption rates, thereby indirectly affecting the measurement of the hypoxia state. By introducing the blood flow-hypoxia coupling coefficient α, this interference can be quantitatively corrected to improve the accuracy of hypoxia concentration calculation.

[0138] In one embodiment, step S4 includes:

[0139] S410. Perform threshold segmentation processing on the real-time change in hypoxia concentration, define preset acute hypoxia determination thresholds to generate spatio-temporal marker points. The acute hypoxia determination threshold is a dynamic threshold set based on the hypoxia concentration change rate and the physiological steady-state range, and is used to distinguish acute hypoxia and chronic hypoxia regions.

[0140] Specifically, after obtaining the real-time change in hypoxia concentration Δ[O2](t), in order to distinguish acute hypoxia and chronic hypoxia regions, threshold segmentation processing needs to be performed on it. First, a dynamic acute hypoxia determination threshold θ(t) is set according to the hypoxia concentration change rate and the physiological steady-state range. This threshold is not fixed but dynamically adjusted according to the physiological state of the tissue and the hypoxia change rate. Specifically, a basic threshold range can be determined through statistical analysis and clinical experience, and then adjusted according to the real-time hypoxia concentration change rate and the tissue's oxygen metabolism situation. For each time point t, the real-time change in hypoxia concentration Δ[O2](t) is compared with the acute hypoxia determination threshold θ(t). If Δ[O2](t) exceeds θ(t), then mark the time point and the corresponding spatial position as the spatio-temporal marker points of the acute hypoxia region; otherwise, mark it as the chronic hypoxia region. In this way, the acute hypoxia and chronic hypoxia regions can be accurately distinguished and marked in time and space.

[0141] S420. Perform three-dimensional reconstruction processing on the vascular topology data of the analyzed optical signal to generate a tumor vascular network model, and the vascular network model is used to indicate a 3D image showing the vascular branch structure.

[0142] Specifically, the analyzed optical signal contains rich vascular topology data, which record information such as the distribution, direction, and branch structure of blood vessels in the tumor region. In order to visually display these vascular structures, three-dimensional reconstruction processing needs to be performed on it. Three-dimensional reconstruction usually uses algorithms such as voxelization or surface rendering to convert two-dimensional vascular data into a three-dimensional vascular network model.

[0143] During the reconstruction process, first, the vascular data is segmented and extracted to separate the blood vessels from the surrounding tissues; then, through methods such as interpolation and surface fitting, the gaps and discontinuities in the data are filled to generate a smooth and continuous vascular surface; finally, using three-dimensional graphics rendering technology, the vascular network is displayed in the form of a 3D image. The generated tumor vascular network model can not only clearly display the branch structure and spatial distribution of blood vessels but also highlight specific vascular features or hypoxia regions through different visual effects such as colors and transparencies, providing important vascular structure information for the subsequent generation of four-dimensional hypoxia maps.

[0144] S430. Perform superposition processing on the spatio-temporal marker points and the tumor vascular network model to generate a four-dimensional hypoxia map.

[0145] Specifically, after the generation of spatio-temporal marker points and the three-dimensional reconstruction of the tumor vascular network model are completed, the system superimposes and processes these two sets of data to generate a four-dimensional hypoxia map. Specifically, the system spatially aligns and fuses each spatio-temporal marker point with the corresponding position in the tumor vascular network model. Based on the three-dimensional vascular network model, according to the hypoxia type (acute or chronic) and degree of hypoxia of the spatio-temporal marker points, they are labeled and displayed with different colors, symbols or intensities. At the same time, taking time as the fourth-dimensional axis, the hypoxia distribution at different time points is dynamically connected to form a dynamic hypoxia map containing three-dimensional spatial coordinates and a time axis.

[0146] Through this four-dimensional map, the evolution process of the hypoxic region in the tumor over time can be intuitively observed, including the expansion, contraction, migration of the hypoxic region and its interaction with the vascular network, etc. This map not only provides a powerful tool for studying the hypoxic physiological mechanism of tumors, but also can provide important visual evidence for the formulation and optimization of clinical treatment plans, helping doctors to more accurately plan radiotherapy target areas, optimize chemotherapy drug delivery and evaluate treatment efficacy.

[0147] The above method for detecting the hypoxia status in a non-invasive solid tumor performs threshold segmentation processing on the real-time hypoxia concentration change amount to accurately identify acute and chronic hypoxia regions; uses three-dimensional reconstruction technology to convert vascular topology data into an intuitive three-dimensional image; and finally generates a four-dimensional hypoxia map by superimposing spatio-temporal marker points and the tumor vascular network model. These steps provide important technical support for comprehensively evaluating the hypoxia status and its microenvironment of tumors, helping to improve the accuracy of tumor diagnosis, guiding the formulation of personalized treatment plans, and providing dynamic evidence for monitoring treatment effects.

[0148] In one embodiment, step S5 includes:

[0149] S510. Perform microfluidic chip simulation processing on the analyzed optical signal and the denoised fluorescence signal to generate an in vitro four-dimensional hypoxia map. The microfluidic chip is a transparent plastic device that simulates the biological fluid environment.

[0150] Specifically, in this process, first input the analyzed optical signal and the denoised fluorescence signal into the microfluidic chip system. The microfluidic chip is a transparent plastic device that simulates the biological fluid environment and can precisely control the flow and distribution of fluids to simulate in vivo physiological conditions. By setting microchannel structures similar to in vivo tumor blood vessels in the chip and injecting composite probes into the chip, the blood flow and hypoxia environment in the tumor area can be simulated. Using high-resolution imaging technologies such as fluorescence microscopes and photoacoustic imaging systems, the signals in the chip are monitored and collected in real time to generate an in vitro four-dimensional hypoxia map. This in vitro map can provide hypoxia distribution information similar to that in vivo and provide a reference for subsequent model calibration and optimization.

[0151] 520. Perform spatio-temporal consistency comparison processing on the four-dimensional hypoxia map and the in-vitro four-dimensional hypoxia map to generate the in-vitro - in-vivo error rate.

[0152] Specifically, the system performs spatio-temporal consistency comparison between the four-dimensional hypoxia map generated by spatial superposition processing and the in-vitro four-dimensional hypoxia map generated by the in-vitro microfluidic chip. By comparing the hypoxia concentration distributions of the two maps at the same time points and spatial positions, the in-vitro - in-vivo error rate is calculated. Preferably, various methods can be used to calculate the error rate, such as mean square error, correlation coefficient, etc., to quantify the difference between the two. Specifically, for each time point and spatial position, the difference in hypoxia concentration between in-vivo and in-vitro is calculated, and then statistical analysis is performed on the differences of all data points to obtain the overall error rate. This error rate reflects the degree of agreement between in-vitro simulation and in-vivo actual situation, providing an important evaluation index for subsequent model optimization.

[0153] S530. Identify whether the in-vitro - in-vivo error rate exceeds a preset threshold. If it exceeds, trigger the Bayesian optimization algorithm to process the in-vitro - in-vivo error rate and generate optimized model parameters. The Bayesian optimization algorithm is an intelligent algorithm that adjusts parameters based on probability distribution.

[0154] Specifically, after obtaining the in-vitro - in-vivo error rate, the system compares it with the preset threshold. If the identified error rate exceeds the preset threshold, it means that the difference between in-vitro simulation and in-vivo actual situation is large, and the model parameters need to be optimized, thus triggering the Bayesian optimization algorithm. This algorithm adjusts parameters based on probability distribution and is an efficient intelligent optimization method. The Bayesian optimization algorithm constructs a probabilistic surrogate model of the objective function, such as a Gaussian process, and uses an acquisition function to guide the selection of sampling points, gradually exploring the parameter space to find the parameter combination that minimizes the objective function (i.e., the in-vitro - in-vivo error rate). During the optimization process, the algorithm continuously updates the probability distribution of the parameters according to historical data, gradually narrowing the parameter search range to improve the optimization efficiency and accuracy.

[0155] S540. Perform iterative correction processing on the dynamic solution model based on the optimized model parameters to generate the corrected hypoxia concentration distribution. The iterative correction processing is a calculation process that gradually approaches the optimal value.

[0156] Specifically, the system iteratively corrects the dynamic calculation model using the optimized model parameters obtained by the Bayesian optimization algorithm. Specifically, the optimized parameters are substituted into the dynamic calculation model to recalculate the hypoxic concentration distribution. In each iteration, the calculation process of the model is updated according to the new parameter values, and the predicted results of the hypoxic concentration are gradually adjusted to be closer to the actual in-vivo situation. The iterative correction process usually adopts numerical methods such as gradient descent and Newton iteration to gradually approach the optimal value. Through multiple iterations, the in-vitro - in-vivo error rate is continuously reduced, and finally the corrected hypoxic concentration distribution is obtained, improving the accuracy and reliability of the model.

[0157] Preferably, the optimized model parameters are generated through the following steps:

[0158] S541. Construct an objective function based on the in-vitro - in-vivo error rate. The objective function is the mapping relationship between the error rate, model parameters, and the acute hypoxia determination threshold.

[0159] Specifically, take the in-vitro - in-vivo error rate as the output of the objective function, and model parameters (such as the fluorescence activation threshold k1 of the composite probe, the photoacoustic oxygen saturation sensitivity k2 of the composite probe, the blood flow - hypoxia coupling coefficient α, etc.) and the acute hypoxia determination threshold as input variables to establish the mapping relationship between them. Specifically, the objective function can be expressed as:

[0160] f(γ) = η(γ)

[0161] where γ is the set of model parameters and the acute hypoxia determination threshold, η is the error rate, and the value of the objective function reflects the degree of difference between in-vitro simulation and the actual in-vivo situation under the given parameters. The goal of the Bayesian optimization algorithm is to find the parameter combination θ that minimizes f(γ). * .

[0162] S542. Define the prior probability distribution of model parameters and thresholds according to historical data or a preset range. The prior distribution includes normal distribution or uniform distribution.

[0163] Specifically, in the Bayesian optimization algorithm, the definition of the prior probability distribution has an important impact on the optimization result. According to historical data or a preset range, define the prior probability distribution for model parameters and the acute hypoxia determination threshold. Common prior distributions include normal distribution and uniform distribution. For example, if the approximate range of a certain parameter is known, a uniform distribution can be adopted; if there is some knowledge about the mean and variance of the parameter, a normal distribution can be adopted. The prior distribution reflects the beliefs and knowledge about the possible values of the parameters before the optimization starts, providing a basis for subsequent sampling and probability update.

[0164] S543. Iteratively sample the objective function through the Bayesian optimization algorithm to generate new parameter combinations, and update the posterior probability distribution based on the sampling results, gradually narrowing the parameter search space.

[0165] Specifically, the Bayesian optimization algorithm gradually optimizes the objective function through iterative sampling. In each iteration, according to the current posterior probability distribution, the next sampling point (i.e., parameter combination) is selected. The selection of the sampling point is usually based on acquisition functions such as the upper confidence bound (UCB), expected improvement (EI), etc. These acquisition functions can achieve a balance between exploration and exploitation, guiding the algorithm to effectively explore the parameter space. After obtaining the new sampling point, calculate the corresponding objective function value and use this information to update the posterior probability distribution. As the iteration progresses, the posterior probability distribution gradually concentrates on the parameter region with a smaller objective function value, thereby gradually narrowing the parameter search space and improving the efficiency and accuracy of the optimization.

[0166] S544. When the number of iterations reaches the preset upper limit or the error rate converges within the allowable range, output the optimized model parameters.

[0167] Specifically, set a preset upper limit on the number of iterations or an error rate convergence condition. When either condition is met, stop the optimization process and output the current optimal model parameters. Setting the upper limit on the number of iterations can prevent the algorithm from running excessively and save computing resources; the error rate convergence condition ensures that the optimization result reaches the required accuracy. For example, when the change in the error rate for several consecutive iterations is less than a preset threshold, it is considered that the error rate has converged and the optimization can be stopped. The output optimized model parameters are the best parameter combinations finally used to correct the dynamic solution model, which can effectively improve the prediction accuracy of the model for the in-vivo hypoxia concentration distribution.

[0168] The above-provided method for detecting the internal hypoxia condition of a non-invasive solid tumor generates an in-vitro four-dimensional hypoxia map through the microfluidic chip simulation processing of the analyzed optical signal and the denoised fluorescence signal, performs a spatio-temporal consistency comparison with the in-vivo four-dimensional hypoxia map, and calculates the in-vitro - in-vivo error rate. When the error rate exceeds the preset threshold, the Bayesian optimization algorithm is used to optimize the model parameters, and the dynamic solution model is iteratively corrected to finally generate the corrected hypoxia concentration distribution. This series of steps effectively improves the accuracy and reliability of the model, enabling it to more realistically reflect the hypoxia state of the in-vivo tumor region, providing high-precision spatio-temporal data support for radiotherapy target area planning of solid tumors, optimization of chemotherapy drug delivery, prediction and evaluation of chemotherapy drug resistance and efficacy, and risk analysis of tumor metastasis progression based on hypoxia microenvironment monitoring.

[0169] In a second aspect, as Figure 3 shown, the present invention provides a detection system 600 for the internal hypoxia condition of a non-invasive solid tumor, including:

[0170] A probe construction module 610 for chemically modifying a near-infrared second-region fluorescent group, an oxygen-sensitive nitroimidazole group, and a tumor-targeting group to generate a composite probe;

[0171] Among them, the near-infrared second-region fluorescent group is a fluorescent molecular structure with a wavelength of 1000 - 1700 nm, the oxygen-sensitive nitroimidazole group is a chemical group that is reduced under hypoxic conditions, and the tumor-targeting group is an RGD polypeptide molecule connected by a covalent bond; the composite probe is used to obtain the fluorescent signal and photoacoustic signal in the near-infrared second region of the tumor area;

[0172] A probe injection and imaging module 620 for intravenously injecting the composite probe, synchronously collecting the fluorescent signal and photoacoustic signal using the composite probe, preprocessing the fluorescent signal to obtain a denoised fluorescent signal, and preprocessing the photoacoustic signal to obtain an analyzed optical signal;

[0173] A dynamic calculation module 630 for generating a real-time hypoxia concentration change based on the time derivative of the denoised fluorescent signal, the blood flow parameter of the analyzed optical signal, and a pre-calibration coefficient;

[0174] A spatio-temporal mapping module 640 for performing spatial superposition processing on the real-time hypoxia concentration change and the vascular topology data of the analyzed optical signal to generate a four-dimensional hypoxia map, and the four-dimensional hypoxia map includes a dynamic distribution map of three-dimensional spatial coordinates and a time axis;

[0175] A verification and calibration module 650 for processing the four-dimensional hypoxia map through in vitro microfluidic simulation and individualized Bayesian optimization to generate a corrected hypoxia concentration distribution.

[0176] In summary, a non-invasive detection system for hypoxia in solid tumors provided by the present invention generates a specific composite probe through chemical modification, utilizes its enrichment and response characteristics in the tumor area, and combines multimodal imaging technology to achieve real-time, dynamic, and high-resolution imaging of the tumor hypoxia state. Through signal acquisition, processing, and analysis, a four-dimensional hypoxia map is generated and further corrected, providing high-precision spatio-temporal data support based on hypoxia microenvironment monitoring for radiotherapy target area planning of solid tumors, optimization of chemotherapy drug delivery, prediction and evaluation of chemotherapy drug resistance and efficacy, and risk analysis of tumor metastasis progression.

[0177] Preferably, the probe construction module 610 is configured with the following components:

[0178] A probe core construction sub-module 611 for chemically modifying an iridium complex to generate a probe core structure, and the iridium complex is a luminescent molecular structure centered on iridium metal, and an oxygen-sensitive nitroimidazole group and a tumor-targeting group are connected by a covalent bond;

[0179] The nano - probe preparation sub - module 612 is used to wrap the probe core structure with polyethylene glycol - liposome and modify the tumor - targeting group on the surface of the polyethylene glycol - liposome to generate a nano - probe with a long circulation time and active targeting function. The polyethylene glycol - liposome is a nano - carrier for extending the blood circulation time, and the nano - probe is used to indicate the intermediate product of the uncalibrated parameter;

[0180] The probe parameter determination sub - module 613 is used to conduct an in - vitro hypoxia gradient experiment on the nano - probe to obtain the fluorescence activation threshold and the photoacoustic oxygen saturation sensitivity. The in - vitro hypoxia gradient experiment is to test the probe performance in a controllable oxygen - concentration chamber. The fluorescence activation threshold is the speed parameter of the probe's response to hypoxia, and the photoacoustic oxygen saturation sensitivity is the influence parameter of oxygen concentration on the photoacoustic signal;

[0181] The composite - probe fusion sub - module 614 is used to obtain a composite probe based on the fluorescence activation threshold and the photoacoustic oxygen saturation sensitivity. The composite probe includes the nano - probe, the fluorescence activation threshold, and the photoacoustic oxygen saturation sensitivity;

[0182] The oxygen - concentration gradient setting unit 6131 is used to perform oxygen - concentration gradient setting on the chamber of the in - vitro hypoxia gradient experiment to generate a continuous oxygen - concentration gradient environment from the preset normoxia to the preset hypoxia;

[0183] The multi - oxygen - response acquisition unit 6132 is used to perform signal acquisition processing on the nano - probe in a multi - oxygen - concentration environment to generate a fluorescence - photoacoustic response curve;

[0184] The parameter fitting and processing unit 6133 is used to process the fluorescence - photoacoustic response curve through first - derivative analysis and exponential fitting to generate the fluorescence activation threshold and the photoacoustic oxygen saturation sensitivity;

[0185] The fluorescence - threshold calibration sub - unit 61331 is used to perform first - derivative calculation processing on the fluorescence response curve, extract the maximum value of the fluorescence signal change rate, and calibrate the oxygen concentration corresponding to the maximum value as the fluorescence activation threshold;

[0186] The photoacoustic data generation sub - unit 61332 is used to process the photoacoustic response curve to generate the photoacoustic signal intensity and the oxygen concentration;

[0187] The photoacoustic sensitivity extraction sub - unit 61333 is used to use the following formula to extract the fitting coefficient in the formula and calibrate it as the photoacoustic oxygen saturation sensitivity:

[0188]

[0189] Where, I PA is the photoacoustic signal intensity, I0 is the baseline intensity under normoxia conditions, [O2] is the oxygen concentration, and k2 is the fitting coefficient, that is, the photoacoustic oxygen saturation sensitivity.

[0190] Preferably, the probe injection and imaging module 620 is configured with the following components:

[0191] A probe injection sub-module 621 for injecting a composite probe into the human body through intravenous injection technology;

[0192] A fluorescence signal acquisition sub-module 622 for collecting and processing near-infrared second-region fluorescence signals in the tumor area, using an InGaAs detector to obtain fluorescence intensity at a preset high-frequency frame rate. The InGaAs detector is a semiconductor sensor for detecting near-infrared light;

[0193] A fluorescence signal denoising sub-module 623 for denoising the fluorescence intensity based on background scattering correction technology to generate a denoised fluorescence signal. The background scattering correction technology is a mathematical filtering method for indicating the removal of tissue reflected light interference;

[0194] An optoacoustic signal analysis sub-module 624 for generating an optoacoustic signal through laser excitation and ultrasonic detector reception and processing, and analyzing and processing the optoacoustic signal to generate an analyzed optoacoustic signal. The analyzed optoacoustic signal includes vascular oxygen saturation and blood flow velocity. Vascular oxygen saturation is the proportion of oxyhemoglobin in the blood, and blood flow velocity is the rate of blood flow.

[0195] Preferably, the dynamic calculation module 630 is configured with the following components:

[0196] A fluorescence signal calculation sub-module 631 for performing time derivative calculation processing on the denoised fluorescence signal to generate a signal change rate;

[0197] An oxygen deficiency concentration calculation sub-module 632 for using the following formula to calculate the signal change rate, the analyzed optoacoustic signal, and a pre-calibrated coefficient to generate an oxygen deficiency concentration change amount:

[0198]

[0199] where Δ[O2](t) is the real-time oxygen deficiency concentration change amount, k1 is the fluorescence activation threshold of the composite probe, k2 is the optoacoustic oxygen saturation sensitivity of the composite probe, and α is the blood flow-oxygen deficiency coupling coefficient, which is used to represent the interference weight of blood flow velocity change on oxygen deficiency detection.

[0200] Preferably, the spatio-temporal mapping module 640 is configured with the following components:

[0201] A spatio-temporal marker point generation sub-module 641 for performing threshold segmentation processing on the real-time oxygen deficiency concentration change amount, defining a preset acute oxygen deficiency determination threshold to generate spatio-temporal marker points. The acute oxygen deficiency determination threshold is a dynamic threshold set based on the oxygen deficiency concentration change rate and the physiological steady state range, and is used to distinguish acute oxygen deficiency and chronic oxygen deficiency regions;

[0202] The vascular network model reconstruction sub-module 642 is used to perform three-dimensional reconstruction processing on the vascular topology data of the parsed optical signals to generate a tumor vascular network model, and the vascular network model is used to indicate a 3D image showing the vascular branch structure;

[0203] The four-dimensional atlas generation sub-module 643 is used to perform superposition processing on the spatio-temporal marker points and the tumor vascular network model to generate a four-dimensional hypoxia atlas.

[0204] Preferably, the verification and calibration module 650 is configured with the following components:

[0205] The in vitro atlas generation sub-module 651 is used to perform microfluidic chip simulation processing on the parsed optical signals and the denoised fluorescence signals to generate an in vitro four-dimensional hypoxia atlas, and the microfluidic chip is a transparent plastic device simulating a biological fluid environment;

[0206] The error rate calculation sub-module 652 is used to perform spatio-temporal consistency comparison processing on the four-dimensional hypoxia atlas and the in vitro four-dimensional hypoxia atlas to generate an in vitro-in vivo error rate;

[0207] The parameter optimization sub-module 653 is used to identify whether the in vitro-in vivo error rate exceeds a preset threshold. If it exceeds, it triggers the Bayesian optimization algorithm to process the in vitro-in vivo error rate to generate optimized model parameters, and the Bayesian optimization algorithm is an intelligent algorithm based on probability distribution to adjust parameters;

[0208] The model correction sub-module 654 is used to perform iterative correction processing on the dynamic solution model based on the optimized model parameters to generate a corrected hypoxia concentration distribution, and the iterative correction processing is a calculation process of gradually approaching the optimal value;

[0209] The objective function construction unit 6541 is used to construct an objective function based on the in vitro-in vivo error rate, and the objective function is a mapping relationship between the error rate, the model parameters, and the acute hypoxia determination threshold;

[0210] The prior distribution definition unit 6542 is used to define the prior probability distribution of the model parameters and the threshold according to historical data or a preset range, and the prior distribution includes a normal distribution or a uniform distribution;

[0211] The parameter sampling and updating unit 6543 is used to perform iterative sampling on the objective function through the Bayesian optimization algorithm to generate new parameter combinations, and update the posterior probability distribution based on the sampling results to gradually narrow the parameter search space;

[0212] The optimized parameter output unit 6544 is used to output the optimized model parameters when the number of iterations reaches a preset upper limit or the error rate converges within an allowable range.

[0213] In one embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of a method for detecting the internal hypoxia condition of a non-invasive solid tumor as described above are implemented.

[0214] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of a method for detecting the internal hypoxia condition of a non-invasive solid tumor as described above are implemented.

[0215] In the description of this specification, the description with reference to terms such as "one embodiment", "some embodiments", "example", "specific example" or "some examples" means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0216] In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present application, "a plurality" means two or more, unless otherwise specifically defined.

[0217] As described above, the above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of various changes or substitutions within the technical scope disclosed in the present application, and these should all be covered by the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A method for detecting the internal hypoxia condition of non-invasive solid tumors, characterized in that, Including the following steps: S1: Chemically modify a near-infrared II region fluorescent group, an oxygen-sensitive nitroimidazole group, and a tumor-targeting group to generate a composite probe; Among them, the near-infrared II region fluorescent group is a fluorescent molecular structure with a wavelength of 1000 - 1700 nm, the oxygen-sensitive nitroimidazole group is a chemical group that is reduced under hypoxic conditions; the tumor-targeting group is an RGD polypeptide molecule connected by a covalent bond, and the composite probe is used to obtain the fluorescent signal and photoacoustic signal in the near-infrared II region of the tumor area; S2: Intravenously inject the composite probe, and synchronously collect the fluorescent signal and photoacoustic signal using the composite probe, preprocess the fluorescent signal to obtain a denoised fluorescent signal, and preprocess the photoacoustic signal to obtain an analyzed optical signal; S3: Generate a real-time hypoxia concentration change based on the time derivative of the denoised fluorescent signal, the blood flow parameter of the analyzed optical signal, and a pre-calibrated coefficient; S4: Perform a spatial superposition process on the real-time hypoxia concentration change and the vascular topology data of the analyzed optical signal to generate a four-dimensional hypoxia map, and the four-dimensional hypoxia map includes a dynamic distribution map of three-dimensional spatial coordinates and a time axis; S5: Process the four-dimensional hypoxia map through in vitro microfluidic simulation and individualized Bayesian optimization to generate a corrected hypoxia concentration distribution.

2. The detection method according to claim 1, wherein The step S1 includes: Chemically modify an iridium complex to generate a probe core structure, the iridium complex is a luminescent molecular structure centered on iridium metal, and connect the oxygen-sensitive nitroimidazole group and the tumor-targeting group through a covalent bond; Perform polyethylene glycol-liposome coating treatment on the probe core structure, and modify the tumor-targeting group on the surface of the polyethylene glycol-liposome to generate a nanoscale probe with a long circulation time and active targeting function, the polyethylene glycol-liposome is a nanoscale carrier for extending the blood circulation time, and the nanoscale probe is used to indicate an intermediate product of an uncalibrated parameter; Perform an in vitro hypoxia gradient experiment on the nanoscale probe to obtain a fluorescence activation threshold and a photoacoustic oxygen saturation sensitivity, the in vitro hypoxia gradient experiment is to test the probe performance in a controllable oxygen concentration chamber, the fluorescence activation threshold is a speed parameter for the probe to respond to hypoxia, and the photoacoustic oxygen saturation sensitivity is an influence parameter of oxygen concentration on the photoacoustic signal; Based on the fluorescence activation threshold and the photoacoustic oxygen saturation sensitivity, obtain a composite probe, and the composite probe includes the nanoscale probe, the fluorescence activation threshold, and the photoacoustic oxygen saturation sensitivity.

3. The detection method according to claim 2, wherein The performing an in vitro hypoxia gradient experiment on the nanoscale probe to obtain a fluorescence activation threshold and a photoacoustic oxygen saturation sensitivity includes: Perform an oxygen concentration gradient setting process on the chamber of the in vitro hypoxia gradient experiment to generate a continuous oxygen concentration gradient environment from a preset normoxia to a preset hypoxia; Perform signal acquisition processing on the nanoscale probe in a multi-oxygen concentration environment to generate a fluorescence-photoacoustic response curve; Process the fluorescence-photoacoustic response curve through first derivative analysis and exponential fitting to generate the fluorescence activation threshold and the photoacoustic oxygen saturation sensitivity; Among them, the first derivative analysis is used to indicate the oxygen concentration corresponding to the maximum value of the derivative of the fluorescence response curve, and the exponential fitting is used to indicate the fitting of the relationship between the photoacoustic signal intensity and the oxygen concentration.

4. The detection method according to claim 3, characterized in that, Processing the fluorescence-photoacoustic response curve through the first derivative analysis and the exponential fitting to generate the fluorescence activation threshold and the photoacoustic oxygen saturation sensitivity includes: Performing a first derivative calculation process on the fluorescence response curve, extracting the maximum value of the change rate of the fluorescence signal, and calibrating the oxygen concentration corresponding to the maximum value as the fluorescence activation threshold; Processing the photoacoustic response curve to generate the photoacoustic signal intensity and the oxygen concentration; Using the following formula, extracting the fitting coefficient in the formula and calibrating it as the photoacoustic oxygen saturation sensitivity: Among them, I PA is the photoacoustic signal intensity, I0 is the baseline intensity under normoxic conditions, [O2] is the oxygen concentration, and k2 is the fitting coefficient, that is, the photoacoustic oxygen saturation sensitivity.

5. The detection method according to claim 1, wherein The step S2 includes: Injecting the composite probe into the human body through the intravenous injection technique; Collecting and processing the near-infrared second-region fluorescence signal of the tumor region, using an InGaAs detector to obtain the fluorescence intensity at a preset high-frequency frame rate, and the InGaAs detector is a semiconductor sensor for detecting near-infrared light; Performing denoising processing on the fluorescence intensity based on the background scattering correction technique to generate a denoised fluorescence signal, and the background scattering correction technique is a mathematical filtering method for indicating the removal of tissue reflected light interference; Generating a photoacoustic signal through laser excitation and ultrasonic detector reception and processing, and performing analysis processing on the photoacoustic signal to generate an analyzed photoacoustic signal, and the analyzed photoacoustic signal includes vascular oxygen saturation and blood flow velocity, the vascular oxygen saturation is the proportion of oxyhemoglobin in the blood, and the blood flow velocity is the rate of blood flow.

6. The detection method according to claim 1, wherein The step S3 includes: Performing a time derivative calculation process on the denoised fluorescence signal to generate a signal change rate; Using the following formula, calculating the signal change rate, the analyzed photoacoustic signal and a pre-calibrated coefficient to generate a change amount of hypoxic concentration: where, Δ[O2](t) is the real-time change in hypoxia concentration, k1 is the fluorescence activation threshold of the composite probe, k2 is the photoacoustic oxygen saturation sensitivity of the composite probe, α is the blood flow-hypoxia coupling coefficient, which is used to represent the interference weight of blood flow velocity change on hypoxia detection, S PA (t) is the vascular oxygen saturation, and v(t) is the blood flow velocity.

7. The detection method according to claim 1, wherein The step S4 includes: Performing threshold segmentation processing on the real-time change amount of hypoxic concentration, defining a preset acute hypoxia determination threshold to generate spatio-temporal marker points, and the acute hypoxia determination threshold is a dynamic threshold set based on the hypoxic concentration change rate and the physiological steady state range, and is used to distinguish acute hypoxia and chronic hypoxia regions; Performing three-dimensional reconstruction processing on the vascular topology data of the analyzed optical signal to generate a tumor vascular network model, and the vascular network model is a 3D image for indicating the display of the vascular branch structure; Performing superposition processing on the spatio-temporal marker points and the tumor vascular network model to generate the four-dimensional hypoxia map.

8. The detection method according to claim 1, wherein The step S5 includes: Performing microfluidic chip simulation processing on the analyzed optical signal and the denoised fluorescence signal to generate an in vitro four-dimensional hypoxia map, and the microfluidic chip is a transparent plastic device for simulating a biological fluid environment; Performing spatio-temporal consistency comparison processing on the four-dimensional hypoxia map and the in vitro four-dimensional hypoxia map to generate an in vitro-in vivo error rate; Identify whether the in vitro-in vivo error rate exceeds a preset threshold. If it exceeds, trigger the Bayesian optimization algorithm to process the in vitro-in vivo error rate and generate optimized model parameters. The Bayesian optimization algorithm is an intelligent algorithm that adjusts parameters based on probability distributions; Perform iterative correction processing on the dynamic calculation model based on the optimized model parameters to generate a corrected hypoxic concentration distribution. The iterative correction processing is a calculation process that gradually approaches the optimal value.

9. The detection method according to claim 8, wherein The optimized model parameters are generated through the following steps: Construct an objective function based on the in vitro-in vivo error rate. The objective function is a mapping relationship between the error rate, model parameters, and the acute hypoxia determination threshold; Define the prior probability distributions of the model parameters and thresholds according to historical data or a preset range. The prior distributions include normal distributions or uniform distributions; Perform iterative sampling on the objective function through the Bayesian optimization algorithm to generate new parameter combinations, and update the posterior probability distribution based on the sampling results to gradually narrow the parameter search space; When the number of iterations reaches a preset upper limit or the error rate converges within an allowable range, output the optimized model parameters.

10. A detection system for the internal hypoxia condition of non-invasive solid tumors, characterized in that, Include: A probe construction module for chemically modifying a near-infrared second-region fluorescent group, an oxygen-sensitive nitroimidazole group, and a tumor-targeting group to generate a composite probe; Among them, the near-infrared second-region fluorescent group is a fluorescent molecular structure with a wavelength of 1000 - 1700 nm, the oxygen-sensitive nitroimidazole group is a chemical group that is reduced under hypoxic conditions, and the tumor-targeting group is an RGD polypeptide molecule connected by a covalent bond. The composite probe is used to obtain the fluorescent signal and photoacoustic signal in the near-infrared second region of the tumor area; A probe injection and imaging module for intravenously injecting the composite probe, synchronously collecting the fluorescent signal and photoacoustic signal using the composite probe, preprocessing the fluorescent signal to obtain a denoised fluorescent signal, and preprocessing the photoacoustic signal to obtain an analyzed optical signal; A dynamic calculation module for generating a real-time hypoxic concentration change based on the time derivative of the denoised fluorescent signal, the blood flow parameters of the analyzed optical signal, and a pre-calibrated coefficient; A spatio-temporal mapping module for performing spatial superposition processing on the real-time hypoxic concentration change and the vascular topology data of the analyzed optical signal to generate a four-dimensional hypoxia map. The four-dimensional hypoxia map contains a dynamic distribution diagram of three-dimensional spatial coordinates and a time axis; A verification and calibration module for processing the four-dimensional hypoxia map through in vitro microfluidic simulation and individualized Bayesian optimization to generate a corrected hypoxic concentration distribution.

Images