Detection method for diagnosing health condition of ovarian cancer

Abstract

The invention discloses a detection method for diagnosing the health condition of ovarian cancer, and relates to the technical field of biomedical detection.The detection method comprises the steps that a venous blood sample of a patient is collected and subpackaged into a centrifugal tube, an exosome precipitation reagent is added, a blood cell layer is separated, and sterile exosome suspension and a CTC sample are obtained; capturing ovarian cancer specific exosomes based on the sterile exosome suspension, measuring the pressed sinking depth, and calculating the viscoelastic modulus; identifying a star-shaped structure of the ovarian cancer specific exosome, carrying out proportion quantification, and calculating a topological malignancy index; based on the customized scanning instruction, space-time modulation Raman scanning is carried out on the CTC sample, an original Raman spectrum space matrix is generated, spectral feature purification is carried out on the original Raman spectrum space matrix, and a metabolic disorder cell coordinate atlas is generated. According to the method, the star-shaped structure of the ovarian cancer specific exosome is recognized, the proportion is quantified, the topological malignancy index is calculated, the limitation of binary interpretation of a traditional marker is broken through, and dynamic gradient monitoring of the canceration process is achieved.

Description

Technical Field

[0001] This invention relates to the field of biomedical detection technology, and in particular to a method for diagnosing ovarian cancer health status. Background Technology

[0002] In the field of ovarian cancer diagnosis, routine testing methods primarily rely on the integration of serum biomarker analysis and imaging assessment. Serological analysis uses chemiluminescent immunoassay to determine glycoprotein concentrations, while imaging employs transvaginal ultrasound to observe ovarian morphological abnormalities. Furthermore, liquid biopsy techniques are increasingly being applied clinically, including the detection of circulating tumor cells (CTCs) enrichment and exosome isolation and analysis. CTC detection relies on microfluidic chip antibody capture, combined with immunofluorescence staining to achieve counting and phenotypic analysis. These technologies collectively constitute the current standard procedure for the auxiliary diagnosis of ovarian cancer, providing crucial evidence for clinical staging.

[0003] However, existing methods still have limitations: First, traditional exosome analysis focuses on qualitative detection of biomarkers and is insufficient in quantifying cell mechanical properties (such as membrane viscoelasticity), which cannot reflect the dynamic evolution of cell physical properties during the carcinogenesis process; Second, although CTC detection obtains spatial distribution, it lacks multimodal data fusion analysis, making it difficult to establish a quantitative correlation mechanism between metabolic activity and topological malignancy, resulting in limited sensitivity and specificity of early micrometastases risk assessment. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a diagnostic method for ovarian cancer health status detection that addresses the problems of insufficient quantification of cell mechanical properties and lack of multimodal data fusion.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: This invention provides a method for diagnosing ovarian cancer health status, comprising: collecting venous blood samples from patients, aliquoting them into centrifuge tubes, adding exosome precipitation reagent and separating the blood cell layer to obtain a sterile exosome suspension and CTC samples; capturing ovarian cancer-specific exosomes based on the sterile exosome suspension, measuring the indentation depth under pressure, and calculating the viscoelastic modulus; identifying the stellate structure of ovarian cancer-specific exosomes, quantifying their proportion, and calculating the topological malignancy index; dynamically configuring time-controlled Raman scanning parameters based on the topological malignancy index, calculating the metabolic flux amplitude warning threshold based on a preset scanning configuration database, and generating customized scanning instructions; performing time-controlled Raman scanning on the CTC samples based on the customized scanning instructions, generating the original Raman spectral spatial matrix, and performing spectral feature purification on the original Raman spectral spatial matrix to generate a metabolic disorder cell coordinate map; mapping the spatial distribution of cancer cells using a three-dimensional heatmap algorithm based on the metabolic disorder cell coordinate map and the topological malignancy index to generate a three-dimensional risk heatmap of ovarian cancer, determining the ovarian cancer risk based on the three-dimensional risk heatmap of ovarian cancer, and generating an ovarian cancer health status report.

[0007] As a preferred embodiment of the diagnostic method for ovarian cancer health status detection according to the present invention, the steps for obtaining sterile exosome suspension and CTC samples are as follows: Peripheral venous blood was collected from the patient and aliquoted into two centrifuge tubes to obtain whole blood vessel A and whole blood vessel B. The whole blood vessel A was centrifuged and resuspended in buffer solution with exosome precipitation reagent to generate a sterile exosome suspension. The blood cell layer of the whole blood vessel was separated by differential centrifugation to obtain a leukocyte-rich precipitate; The leukocyte-rich precipitate was diluted and injected into a microfluidic chip to obtain CTC samples.

[0008] In a preferred embodiment of the method for diagnosing ovarian cancer health status detection according to the present invention, the calculation of the viscoelastic modulus includes the following steps. A sterile exosome suspension was injected into a silicon nanoneedle array chip, and surface acoustic waves were applied to generate chip-fixed exosomes. A constant pressure is applied to the chip-fixed exosomes, and the pressure indentation depth of the chip-fixed exosomes is recorded. The viscoelastic modulus is then calculated.

[0009] In a preferred embodiment of the ovarian cancer health status detection method of the present invention, the calculation of the topological malignancy index includes the following steps. The viscoelastic modulus is screened based on a preset modulus threshold, and the chip space coordinates are marked to obtain a set of coordinates for high-risk exosomes. Based on the coordinate set of high-risk exosomes, a high-resolution AFM probe was used to scan the surface topology, identify triangular star-shaped protrusions, generate positive signals of star-shaped structures, and calculate the proportion of positive signals of star-shaped structures to determine the topological malignancy index.

[0010] As a preferred embodiment of the ovarian cancer health status detection method of the present invention, the dynamic configuration of time-controlled Raman scanning parameters includes the following steps. Based on predefined interval rules, the topological malignancy index is divided into three levels, generating a topological malignancy index classification label; Based on the topological malignancy index classification label, the preset scan configuration database is invoked to obtain the time-controlled Raman scan parameters.

[0011] In a preferred embodiment of the ovarian cancer health status detection method of the present invention, the steps for generating customized scanning instructions are as follows: Based on the topological malignancy index, the warning threshold for metabolic flux amplitude is calculated using a linear decay formula. The metabolic flux amplitude warning threshold and time-controlled Raman scan parameters are integrated and converted into binary code to generate customized scan instructions.

[0012] As a preferred embodiment of the diagnostic method for ovarian cancer health status detection described in this invention, the generation of the original Raman spectral spatial matrix refers to driving the stage to perform time-controlled Raman scanning on the CTC sample based on customized scanning instructions, simultaneously acquiring lipid and Raman shift intensities, and generating the original Raman spectral spatial matrix.

[0013] As a preferred embodiment of the ovarian cancer health status detection method of the present invention, the steps for generating a metabolic disorder cell coordinate map are as follows: Wavelet transform is applied to the original Raman spectrum spatial matrix for noise reduction and background signal removal to generate a net metabolic intensity distribution map. Based on the net metabolic intensity distribution map, the metabolic flux amplitude is calculated and compared with the metabolic flux amplitude warning threshold to generate an excess coordinate marker queue. Based on the out-of-range coordinates, a two-dimensional thermal layer is rendered in real time to generate a coordinate map of metabolically disordered cells.

[0014] As a preferred embodiment of the method for diagnosing ovarian cancer health status detection according to the present invention, the steps for generating a three-dimensional risk heat map of ovarian cancer are as follows: Based on the coordinate map of metabolic disordered cells and the topological malignancy index, the coordinate system of the potassium bromide slide was transformed to three-dimensional space through affine transformation to generate a spatial registration dataset. Based on the spatial registration dataset, coordinate mapping was performed, and the topological malignancy index was used to drive the color gradation changes to generate a three-dimensional risk heat map of ovarian cancer.

[0015] In a preferred embodiment of the ovarian cancer health status detection method of the present invention, the steps for generating the ovarian cancer health status detection report are as follows: Based on the three-dimensional risk heat map of ovarian cancer, the weighted score of malignant clusters is calculated, and a list of spatial distribution of malignant clusters is generated. Based on the list of spatial distribution of malignant clusters, a dual-threshold risk classification is performed to generate a risk classification code. By integrating a three-dimensional risk heat map of ovarian cancer, a list of spatial distributions of malignant clusters, and risk classification codes, an ovarian cancer health status monitoring report is generated.

[0016] The beneficial effects of this invention are as follows: by identifying the stellate structure of ovarian cancer-specific exosomes and quantifying their proportion to calculate the topological malignancy index, the microscopic morphological characteristics are transformed into continuous risk variables, breaking through the limitations of traditional binary interpretation of biomarkers and realizing dynamic gradient monitoring of the carcinogenesis process; based on the topological malignancy index, the time-controlled Raman scanning parameters are dynamically configured and customized scanning instructions are generated, and the laser power and step size accuracy are dynamically optimized according to the risk level. Simultaneously combined with the metabolic flux amplitude warning threshold, targeted scanning of high-risk areas is realized, achieving precise spatiotemporal analysis of single-cell metabolic disorders. Attached Figure Description

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

[0018] Figure 1 A flowchart for a health status testing method for diagnosing ovarian cancer.

[0019] Figure 2 This is a flowchart for calculating the viscoelastic modulus and topological malignancy index of exosomes.

[0020] Figure 3 A flowchart generated for customized scan instructions.

[0021] Figure 4 This is a flowchart for generating a cellular coordinate atlas of metabolic disorders. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

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

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

[0025] Reference Figures 1-4 As one embodiment of the present invention, this embodiment provides a method for diagnosing ovarian cancer health status detection, including the following steps: S1. Collect venous blood samples from patients, aliquot them into centrifuge tubes, add exosome precipitation reagent and separate the blood cell layer to obtain sterile exosome suspension and CTC samples; Peripheral venous blood was collected from the patient and aliquoted into two centrifuge tubes to obtain whole blood vessel A and whole blood vessel B. Further, peripheral venous blood from the patient was collected into an anticoagulation vacuum tube. After gentle inversion and mixing, an anticoagulated whole blood sample was obtained. The anticoagulated whole blood sample was then placed in an intelligent mixer for optical homogenization detection. When the ready indicator light illuminated, a mixing target signal was generated. Based on the mixing target signal, the anticoagulated whole blood sample was aliquoted into two centrifuge tubes with different codes using a sterile transfer gun to obtain whole blood vessel A and whole blood vessel B. The whole blood vessel A was centrifuged and resuspended in buffer solution with exosome precipitation reagent to generate a sterile exosome suspension. Furthermore, firstly, the whole blood vessel A was centrifuged for a fixed time (exemplary: 15 minutes) to separate the plasma layer sample, and a precipitation reagent was added to generate a plasma-reagent mixture; then, the plasma-reagent mixture was centrifuged again to obtain coarse exosome precipitate clumps, which were washed with PBS (phosphate-buffered saline) buffer and centrifuged repeatedly to generate purified exosome precipitate; then, the purified exosome precipitate was resuspended in sterile PBS buffer to form an exosome suspension; finally, the exosome suspension was filtered through a filter membrane to generate a sterile exosome suspension; The blood cell layer of the whole blood vessel was separated by differential centrifugation to obtain a leukocyte-rich precipitate; Furthermore, the whole blood vessel B was first centrifuged for a fixed time to separate the upper plasma layer, the middle leukocyte layer, and the lower erythrocyte sedimentation layer, generating a three-component stratified sample. Then, the middle leukocyte layer was aspirated from the three-component stratified sample using a sterile pipette, avoiding contamination with the upper plasma layer or the lower erythrocyte layer. Next, erythrocyte lysis buffer was added to the middle leukocyte layer, vortexed to mix, and then allowed to stand at room temperature to generate a lysis mixture. Finally, the lysis mixture was centrifuged repeatedly to obtain a leukocyte-rich sedimentation. The leukocyte-rich precipitate was diluted and injected into a microfluidic chip to obtain CTC samples; Furthermore, firstly, leukocyte-rich precipitates were resuspended in sterile PBS buffer to generate leukocyte resuspension. Then, the microfluidic chip modified with EpCAM (epithelial cell adhesion molecule) antibody was washed with PBS to generate an antibody-activated chip. Next, the leukocyte resuspension was injected into the antibody-activated chip at a uniform rate to form a chip that captures CTC (circulating tumor cell) complexes. Then, the chip was washed with PBS buffer to capture CTC complexes, and unbound cells were removed before the purified CTC chip was output. Finally, paraformaldehyde solution was injected into the purified CTC chip for fixation, and the cells were transferred to a potassium bromide slide. The CTC sample fixed on the potassium bromide slide was then output.

[0026] S2. Based on sterile exosome suspension, capture ovarian cancer-specific exosomes, measure the depth of indentation under pressure, and calculate the viscoelastic modulus; identify the stellate structure of ovarian cancer-specific exosomes, quantify their proportion, and calculate the topological malignancy index. A sterile exosome suspension was injected into a silicon nanoneedle array chip, and surface acoustic waves were applied to generate chip-fixed exosomes. Furthermore, firstly, the silicon nanoneedle array chip was immersed in EpCAM monoclonal antibody solution for incubation to generate an antibody-modified chip. The antibody-modified chip was then rinsed with PBS buffer to remove unbound antibody molecules, resulting in a clean antibody chip. Next, a sterile exosome suspension was dropped onto the surface of the clean antibody chip to form a chip-suspension complex. Surface acoustic waves were applied to the chip-suspension complex to drive exosome migration, generating an acoustic wave processing chip. Finally, the acoustic wave processing chip was rinsed with PBS buffer to remove unbound exosomes, resulting in a captured exosome chip. The captured exosome chip was then validated against CD63 (lysosome-associated membrane glycoprotein 3) antibody. If the validation was negative, the capture was considered a failure, and ovarian cancer-specific exosomes were recaptured. If the validation result was positive, the chip was considered successful, and the exosomes were immobilized. A constant pressure was applied to the chip-fixed exosomes, and the indentation depth of the chip-fixed exosomes was recorded. The viscoelastic modulus was then calculated. The expression for calculating the viscoelastic modulus is: ; in, It is the viscoelastic modulus; It refers to the lateral deformation capability of the material (the lipid bilayer membrane for chip-fixed exosomes); It is the vertical force applied by the probe to fix the exosomes on the chip; It is the radius of curvature of the silicon probe tip in an atomic force microscope; It is the depth of exosome indentation under pressure when the chip is fixed; Furthermore, an atomic force microscope (AFM) silicon probe positioning chip is used to fix the target area on the surface of the exosome and a constant pressure (e.g., 0.5 nN) is applied to generate a pressure loading state signal. Based on the pressure loading state signal, the probe offset is monitored in real time by a laser interferometer, and the offset is converted into the pressure indentation depth using the principle of spring elastic deformation to obtain the deformation depth data stream. According to the deformation depth data stream, numerical calculation is performed using the Hertzian contact formula to output the viscoelastic modulus. It should be noted that the principle of spring elastic deformation refers to the linear deformation of the probe cantilever when it is subjected to force. Its offset is proportional to the applied force, and the offset measured by the laser is directly converted into the depression depth. The viscoelastic modulus is screened based on a preset modulus threshold, and the chip space coordinates are marked to obtain a set of coordinates for high-risk exosomes. Furthermore, the viscoelastic modulus is first screened based on a preset modulus threshold: values ​​with viscoelastic modulus greater than the preset threshold are extracted to generate an index set of excessive viscoelastic modulus, and each index number in the index set of excessive viscoelastic modulus is mapped to the corresponding coordinates in the AFM scan spatial coordinate set to generate a temporary set of high-risk coordinates; then, density clustering is performed on the temporary set of high-risk coordinates to merge adjacent points with a spacing <2μm (based on the spatial distribution characteristics of exosomes and clinical data validation settings; if the spacing >2μm, merging will lead to misfusion of different clusters), generating a cluster-optimized coordinate set; finally, based on the cluster-optimized coordinate set, the morphology database is called to filter artifacts to obtain the high-risk exosome coordinate set; It should be noted that the preset modulus threshold is set based on the analysis of ROC curves (Receiver Operating Characteristic curves) of a large number of clinical samples (including healthy / benign / malignant samples), and the exemplary value range is (1.5-3.0); the AFM scanning spatial coordinate set is a set of physical coordinates automatically recorded by the atomic force microscope when scanning the chip surface, which is acquired in real time through a piezoelectric ceramic displacement sensor; the morphology database is a standardized feature library established by scanning a large number of clinical samples with an atomic force microscope, which includes exosome curvature radius, height and artifact features; Based on the coordinate set of high-risk exosomes, a high-resolution AFM probe was used to scan the surface topology, identify triangular star-shaped protrusions, generate positive signals of star-shaped structures, and calculate the proportion of positive signals of star-shaped structures to calculate the topological malignancy index. The expression for calculating the topological malice index is: ; in, It is the topological malignancy index; These are the weighting coefficients for the star-shaped structure; It is the weighting coefficient for the viscoelastic modulus; It is the Bayesian prior success number; It is the number of Bayesian prior failures; It is the modulus normalization reference value; It is the number of stellate-positive exosomes; It is the total number of exosomes in the effective region; It is the median value of the viscoelastic modulus; Furthermore, firstly, the target area for scanning is located based on the high-risk exosome coordinate set, and a high-resolution AFM probe is used to perform a tapping scan on the target area with a fixed step, simultaneously recording height and phase signals, and outputting the original morphology data stream. Next, Gaussian filtering is applied to the original morphology data stream for noise reduction, and a convolutional neural network is used to identify triangular star-shaped protrusions, generating a candidate set of star-shaped structures. Then, the candidate set of star-shaped structures is compared with the morphology database to eliminate artifacts. When a protrusion structure is confirmed for 3 consecutive frames (based on the matching of the AFM probe scanning frequency and the relaxation time of the exosome surface structure, and the validation setting of a large number of clinical samples), it is marked as positive, and a positive star-shaped structure signal is output. The number of positive points is counted based on the positive star-shaped structure signal set, and combined with the total number of exosomes in the effective region of the high-risk exosome coordinate set, the observation ratio is corrected by the Bayesian posterior probability estimation method to generate the star proportion value. Finally, the median value of viscoelastic modulus is extracted from the high-risk exosome coordinate set, and the topological malignancy index is calculated by a linear weighted fusion formula. It should be noted that a convolutional neural network is a deep learning architecture that automatically extracts local features of an image through multiple layers of convolutional kernels.

[0027] S3. Based on the topological malignancy index, dynamically configure the time-controlled Raman scanning parameters, calculate the metabolic flux amplitude warning threshold based on the preset scanning configuration database, and generate customized scanning instructions. Based on predefined interval rules, the topological malignancy index is divided into three levels, generating a topological malignancy index classification label; Furthermore, predefined interval rules are loaded to generate a tiered threshold parameter set (including low-risk threshold θ1, medium-risk threshold θ2, and high-risk threshold θ3), dividing the topological malignancy index into three levels: when <θ1 indicates low risk; θ1≤ When <θ2, it is marked as intermediate risk; High risk is marked when ≥θ2, generating an original grading label set; the original grading label set undergoes clinical boundary verification. ∈[θ1-0.01,θ1+0.01] are forcibly classified as low risk), and the verified classification label is output; the verified classification label is converted into a standard medical code (low risk: L1, medium risk: M2, high risk: H3), and finally the topological malignancy index classification label is generated; It should be noted that the predefined interval rules are a set of topological malignancy index grading thresholds based on clinical validation data, including a low-risk threshold θ1 (default 2.5), an intermediate-risk threshold θ2 (default 4.0), and a high-risk threshold θ3 (default ∞), along with boundary tolerance rules (such as...). (∈[θ1-0.01,θ1+0.01] are forcibly classified as low risk). Based on the topological malignancy index classification label, the preset scan configuration database is called to obtain the time-controlled Raman scan parameters; Furthermore, the topological malignancy index classification label is read and the risk level code is extracted. The risk level identifier is output, and the original scan parameter set is obtained by querying the preset scan configuration database based on the risk level identifier. Then, it is checked whether the laser wavelength in the original scan parameter set matches the current detector (for example, if the laser wavelength is 785nm, it is verified whether the detector is an InGaAs model; if the laser wavelength is 532nm, it is verified whether the detector is an EMCCD model). The verified time-controlled Raman scan parameters are output. It should be noted that the preset scan configuration database is a time-controlled Raman scan parameter group classified and stored according to the topological malignancy index classification label (L1 / M2 / H3), specifically including scan step size, laser power, wavelength combination and single-point dwell time; Based on the topological malignancy index, the warning threshold for metabolic flux amplitude is calculated using a linear decay formula. The expression for calculating the original baseline threshold is: ; in, It is the original baseline threshold; It is the maximum baseline threshold; It is the decay slope; It is an effective topological malignancy index; It is the inflection point of decay; The expression for calculating the warning threshold for metabolic flux amplitude is: ; in, It is the warning threshold for metabolic flux amplitude; It is the lower limit of the warning threshold for metabolic flux amplitude; It is the upper limit of the warning threshold for metabolic flux amplitude; It is the signal-to-noise ratio calibration value; Furthermore, the effectiveness of the topological malignancy index (range 0 ≤ ≤10, truncation if exceeding the limit), obtain the effective topology malice index; match the effective topology malice index with the preset attenuation coefficient database, output the attenuation slope parameter, and calculate the original baseline threshold using the linear attenuation formula; then apply range constraints to the original baseline threshold (e.g., lower limit 3.0, upper limit 5.0), when When <3.0, take 3.0; when If the value is greater than 5.0, the value is set to 5.0, and the constrained metabolic threshold is output. Finally, the constrained metabolic threshold is subjected to real-time signal-to-noise ratio calibration and constraint, and the metabolic flux amplitude warning threshold is output. It should be noted that the validity range of the Topological Malignancy Index is set based on the distribution analysis of a large number of clinical samples, and the range is extended to 10 to cover extreme cases, while negative values ​​and values ​​exceeding the limit are truncated to avoid calculation overflow; the constraint range of the original benchmark threshold is set based on the distribution analysis of a large number of clinical samples, with a lower limit of 3.0 to ensure the detection of high-risk samples and an upper limit of 5.0 to prevent misjudgment of healthy samples. It should be noted that the preset attenuation coefficient database is constructed through regression analysis of the topological malignancy index and metabolic flux amplitude of thousands of multi-center clinical samples (including healthy / benign / malignant) to store the attenuation slope and inflection point value corresponding to different risk levels (low / medium / high); the maximum benchmark threshold is determined by the 95th percentile of the metabolic flux amplitude distribution of a large number of clinical samples, and is set based on the balance point between the upper limit of metabolic intensity (specificity > 95%) of healthy people and the lower limit of detection (sensitivity > 92%) of malignant samples. The metabolic flux amplitude warning threshold and time-controlled Raman scan parameters are integrated and converted into binary code to generate customized scan instructions; Furthermore, firstly, the metabolic flux amplitude warning threshold and the time-controlled Raman scan parameters are aligned by timestamps to generate a spatiotemporally aligned dataset. Based on the spatiotemporally aligned dataset, dynamic laser power correction is performed to output a fused scan parameter set. Next, the fused scan parameter set is truncated within a safe range according to hard boundary reset rules to generate safe and compliant scan parameters. The safe and compliant scan parameters are then converted into 32-bit floating-point binary code according to the IEEE 754 standard to obtain the original binary instruction block. Then, a CRC-32 checksum is added to the original binary instruction block to generate a binary instruction stream with checksum. Finally, the physical location of the high-risk exosome coordinate set is converted into a 16-bit integer and written into the binary instruction stream with checksum to generate customized scan instructions. The expression for dynamic correction of laser power is: ; in, This is the laser power value after dynamic correction; It is the initial power; It is the power regulation coefficient; It is the risk neutrality threshold (determined by Raman signal-to-noise ratio testing of thousands of clinical samples, with an exemplary range of 3.8-4.2). It should be noted that the hard boundary reset rule refers to the rule that forces the parameter value to be reset to the nearest boundary value of the preset safety range when the parameter value exceeds the preset safety range (set according to medical device safety standards and clinical validation data). It includes boundary control in three dimensions: laser power, scanning step size and duration.

[0028] S4. Based on customized scanning instructions, perform time-controlled Raman scanning on CTC samples to generate the original Raman spectral spatial matrix, and perform spectral feature purification on the original Raman spectral spatial matrix to generate a coordinate map of metabolically disordered cells. Based on customized scanning commands, the stage is driven to perform time-controlled Raman scanning on CTC samples, simultaneously acquiring lipid and Raman shift intensities to generate the original Raman spectral spatial matrix. Furthermore, based on the customized scanning command, the laser power, step size, and scanning area coordinates are analyzed to drive the stage to position itself at the starting point and generate a ready scanning signal. The ready scanning signal triggers a piezoelectric ceramic stepper motor to move at a set step size, simultaneously activating the laser to output a step trigger pulse. Then, based on the step trigger pulse, the stage stops at the current position, and an InGaAs detector is used to acquire the lipid characteristic peak intensity, obtaining the lipid intensity value. Simultaneously, the system switches to a 785nm laser to acquire the nucleic acid characteristic peak intensity, obtaining the nucleic acid intensity value. Finally, the current scanning area coordinates are bound to the lipid and nucleic acid intensity values, and a CRC-16 checksum is added to generate a single-point spectral data packet. All coordinate points in the scanning area are traversed, and the single-point spectral data packet is integrated to construct a quadruple matrix, generating the original Raman spectral spatial matrix. Wavelet transform is applied to the original Raman spectrum spatial matrix for noise reduction and background signal removal to generate a net metabolic intensity distribution map. Furthermore, firstly, wavelet transform processing is performed on the original Raman spectral spatial matrix to generate a wavelet decomposition coefficient set, and the noise standard deviation of each sub-band is calculated based on the wavelet decomposition coefficient set. At the same time, a statistical threshold is calculated based on the noise standard deviation. Then, hard threshold filtering is performed on the high-frequency sub-band coefficients based on the statistical threshold to generate filtered wavelet coefficients. Simultaneously, low-frequency approximate components are extracted from the wavelet decomposition coefficient set to generate a low-frequency background spectrum. The filtered wavelet coefficients and the low-frequency background spectrum are reconstructed by inverse wavelet transform to generate a denoised Raman spectral matrix. Then, the low-frequency background spectrum is matched with a preset background database to generate the optimal matching background spectrum. The optimal matching background spectrum is removed from the denoised Raman spectral matrix to generate a background subtraction spectral matrix. At the same time, negative value zeroing and baseline shift processing are performed on the background subtraction spectral matrix to finally generate a net metabolic intensity distribution map. The expression for calculating the statistical threshold is: ; in, It is a statistical threshold; These are the original wavelet coefficients; It is the Gaussian correction factor; It is the total number of high-frequency subband coefficients; Based on the net metabolic intensity distribution map, the metabolic flux amplitude is calculated and compared with the metabolic flux amplitude warning threshold to generate an excess coordinate marker queue. The expression for calculating the amplitude of metabolic flux is: ; in, It is the amplitude of metabolic flux; It is the lipid strength value; It is the nucleic acid strength value; Furthermore, based on the net metabolic intensity distribution map, the metabolic flux amplitude is calculated pixel by pixel to generate a metabolic flux amplitude distribution map. According to the metabolic flux amplitude distribution map, all coordinate points where the metabolic flux amplitude is greater than the metabolic flux amplitude warning threshold are marked to generate an original set of out-of-range coordinates. Then, DBSCAN spatial clustering is performed on the original set of out-of-range coordinates to generate clustered out-of-range coordinate clusters. Based on the clustered out-of-range coordinate clusters, the morphology database is called to verify the cell curvature radius. At the same time, debris artifacts are filtered out to generate a queue of true positive out-of-range coordinate markers. It should be noted that DBSCAN spatial clustering is a density-based unsupervised learning algorithm that automatically identifies dense regions by scanning the local density of coordinate points, clusters high-density connected points into clusters and removes noise points; the preset background database is constructed by collecting Raman spectra of standardized healthy volunteer samples, filtering characteristic peaks and reducing dimensionality through principal component analysis, and includes normalized background spectrum mean and variance data. Based on the queue of out-of-standard coordinates, a two-dimensional thermal layer is rendered in real time to generate a coordinate map of metabolically disordered cells. Furthermore, a 1000×1000 pixel two-dimensional zero-value grid is initialized based on the out-of-standard coordinate marker queue to generate an initial thermal grid. Then, each coordinate point in the out-of-standard coordinate marker queue is traversed, and a Gaussian density distribution is superimposed on the grid with each coordinate point as the center to generate a density thermal grid. Then, linear normalization is performed on the density thermal grid to output a normalized thermal grid, and the normalized thermal grid values ​​are mapped to rainbow color levels to generate a pseudo-color thermal layer. Finally, each pixel position in the pseudo-color thermal layer is converted into physical coordinates and bound to the pixel RGB (three primary color light mode) values ​​to generate a coordinate map of metabolically disordered cells. S5. Based on the coordinate map of metabolic disorder cells and the topological malignancy index, the spatial distribution of cancer cells is mapped through a three-dimensional heat map algorithm to generate a three-dimensional risk heat map of ovarian cancer. Based on the three-dimensional risk heat map of ovarian cancer, the risk of ovarian cancer is determined and an ovarian cancer health status test report is generated. Based on the coordinate map of metabolic disordered cells and the topological malignancy index, the coordinate system of the potassium bromide slide was transformed to three-dimensional space through affine transformation to generate a spatial registration dataset. Furthermore, a three-dimensional spatial framework is first defined based on the coordinate atlas of metabolic disordered cells and the topological malignancy index. Then, the actual positions of the four corner cross-shaped markers on the potassium bromide slide are directly measured using high-resolution microscopy, and a set of actual marker coordinates is generated using sub-pixel fitting. At the same time, the actual marker coordinate set is compared with a preset standard coordinate template to measure the rotation angle and translation, generating slide calibration parameters. Then, an affine transformation matrix is ​​constructed based on the slide calibration parameters, and each point in the coordinate atlas of metabolic disordered cells and its associated topological malignancy index are input into the affine transformation matrix to obtain three-dimensional coordinates and generate a spatial point cloud. Finally, the spatial point cloud is encoded into an IEEE 754 single-precision floating-point array using quadruples, and a slide ID and timestamp are appended to generate a spatial registration dataset. It should be noted that the coordinate system of the potassium bromide slide is a two-dimensional rectangular coordinate system physically located by the crosshair marks etched on the surface of the slide. The origin is the center of the lower left corner mark, the X-axis extends along the long side, and the Y-axis extends along the short side. It is generated by sub-pixel fitting of the microscopic image. The preset standard coordinate template is the theoretical position template of the crosshair marks at the four corners of the potassium bromide slide, which is pre-determined by high-precision optical calibration equipment. Based on the spatial registration dataset, coordinate mapping was performed, and the topological malignancy index was used to drive color gradation changes to generate a three-dimensional risk heat map of ovarian cancer. Furthermore, firstly, inverse distance weighted interpolation is performed based on the spatial registration dataset to map discrete points to a 1000×1000×1000 voxel grid with a resolution of 1μm, generating three-dimensional meshed metabolic data. Dynamic color level rules are then loaded according to the topological malignancy index, and dynamic color level parameters are output. Next, the metabolic intensity value of each voxel in the three-dimensional meshed metabolic data is bound to the dynamic color level parameters to generate the original RGB voxel matrix. Finally, a vascular morphology database is used to perform tubular structure filtering on the original RGB voxel matrix, outputting the filtered RGB voxels. The matrix is ​​then processed; each voxel value in the filtered RGB voxel matrix is ​​extracted, and the transparency is dynamically adjusted according to the voxel value (for example, when the voxel value is > 5000 photon count, the transparency is adjusted to 0.2 for low transparency and highlighting; when the voxel value is ≤ 3000, the transparency is adjusted to 0.8 for high transparency and weakening the background), generating an RGBA voxel matrix. Ray casting rendering is then performed on the RGBA voxel matrix (light source direction (0,1,0), resolution 4096×4096), finally generating a three-dimensional risk heat map of ovarian cancer. It should be noted that voxel values ​​are the quantized values ​​of each volume unit (voxel) in a three-dimensional digital image, used to characterize the intensity of features (such as metabolic activity, density, or optical signal intensity) at that spatial location; the dynamic color gradation rule is based on the topological malignancy index (…). The set of color gradient modes (value) toggled between, with the specific rules as follows: when When the value is ≤4.0, a blue-green gradient is used to determine it as a low-risk range; When 4.0 < A score of ≤7.0 is determined to be in the medium-risk range using a green-yellow gradient. when A reading of >7.0 indicates a high-risk zone, indicated by a red-yellow gradient. It should be noted that the vascular morphology database is a standardized dataset that stores vascular structural feature parameters, including parameters such as vascular radius of curvature, tubular structure length, branching angle, wall thickness, and spatial coordinates. It is generated by fusing microvascular casting imaging and clinical angiography data. Based on the three-dimensional risk heat map of ovarian cancer, the weighted score of malignant clusters is calculated, and a list of spatial distribution of malignant clusters is generated. The expression for calculating the weighted score of malignant clusters is: ; in, It is a cluster-weighted score for malignant clusters. It is the weight of the metabolic amplitude in the malignancy score; yes The contribution weight of the peak value to the malignancy score; It is the peak value of the topological malignancy index; It is the median of metabolic flux amplitude; Furthermore, based on the three-dimensional risk heatmap of ovarian cancer, DBSCAN spatial clustering was performed to identify high-density regions and generate malignant clusters. Each cluster within the malignant clusters was traversed, and the median metabolic flux amplitude of all voxels within the cluster was extracted, outputting the median cluster metabolic flux amplitude column. Simultaneously, the maximum topological malignancy index of each voxel within the cluster was extracted, and the cluster index was output. Peak value column; then the median column of cluster metabolic amplitude and cluster The peak column undergoes weighted calculation to generate a malignant cluster score column, and the malignant cluster score column is scanned when... >5.2 (based on ROC curve analysis of a large number of ovarian cancer samples) marked as high-risk malignant clusters, outputting a high-risk malignant cluster label set; finally, the center coordinates of the high-risk malignant cluster label set, the median column of cluster metabolic amplitude, and The peak value column and the malignant cluster score column are encapsulated in JSON-LD format to generate a list of spatial distributions of malignant clusters. Based on the list of spatial distribution of malignant clusters, a dual-threshold risk classification is performed to generate a risk classification code. Furthermore, a malignant cluster scoring column is extracted based on the spatial distribution list of malignant clusters, and the malignant cluster scoring column is scanned: when Clusters with a value less than 4.0 are marked as low-risk clusters, and the low-risk cluster label set is output; clusters with a value less than or equal to 4.0 are marked as low-risk clusters. When <5.2, it is marked as a medium-risk cluster, and the medium-risk cluster label set is output; when If the cluster value is ≥5.2, it is marked as a high-risk cluster, and the high-risk cluster label set is output. The low-risk cluster label set, medium-risk cluster label set and high-risk cluster label set are merged, sorted in ascending order by cluster ID, and a risk classification coding table is generated. It should be noted that the dual thresholds refer to the health threshold (exemplary: 4.0) and the metastasis threshold (exemplary: 5.2), which were set based on ROC curve analysis of a large number of ovarian cancer clinical samples; Integrate the three-dimensional risk heat map of ovarian cancer, the spatial distribution list of malignant clusters, and the risk classification code to generate an ovarian cancer health status monitoring report; Furthermore, firstly, the 3D risk heatmap of ovarian cancer, the list of spatial distributions of malignant clusters, and the risk grading codes are aligned according to timestamps to generate a time-stamped multi-source dataset. Based on the time-stamped multi-source data, a report framework is constructed to generate a structured report framework. Next, the 3D risk heatmap of ovarian cancer is downsampled to 512×512 pixels and encapsulated as a DICOM (Digital Imaging and Communication in Medicine) object, outputting a standard heatmap DICOM object. The list of spatial distributions of malignant clusters is converted into an XML (Extensible Markup Language) table, outputting a structured cluster distribution table. The risk grading code table is mapped to clinical terms, outputting a clinical grading terminology table. Finally, the standard heatmap DICOM object, the structured cluster distribution table, and the clinical grading terminology table are filled into the corresponding chapters of the structured report framework, and format conversion and electronic signature are performed to generate an ovarian cancer health status monitoring report.

[0029] This embodiment also provides a computer device applicable to the diagnostic method for ovarian cancer health status detection, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the diagnostic method for ovarian cancer health status detection as proposed in the above embodiment.

[0030] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0031] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the method for diagnosing ovarian cancer health status detection as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0032] In summary, this invention achieves dynamic gradient monitoring of carcinogenesis by: identifying the stellate structure of ovarian cancer-specific exosomes and quantifying their proportion to calculate the topological malignancy index, transforming microscopic morphological characteristics into continuous risk variables, thus overcoming the limitations of traditional binary interpretation of biomarkers; dynamically configuring time-controlled Raman scanning parameters based on the topological malignancy index and generating customized scanning instructions, dynamically optimizing laser power and step size accuracy according to risk level, and simultaneously combining metabolic flux amplitude warning thresholds to achieve targeted scanning of high-risk areas, thus achieving precise spatiotemporal analysis of single-cell metabolic disorders.

[0033] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for diagnosing ovarian cancer health status, characterized in that: include, Collect venous blood samples from patients, aliquot them into centrifuge tubes, add exosome precipitation reagent and separate the blood cell layer to obtain sterile exosome suspension and CTC samples; Based on sterile exosome suspension, ovarian cancer-specific exosomes were captured, and the depth of indentation under pressure was measured to calculate the viscoelastic modulus; the stellate structure of ovarian cancer-specific exosomes was identified, and the proportion was quantified to calculate the topological malignancy index. Based on the topological malignancy index, the time-controlled Raman scanning parameters are dynamically configured, and the metabolic flux amplitude warning threshold is calculated based on the preset scanning configuration database to generate customized scanning instructions. Based on customized scanning instructions, a time-controlled Raman scan is performed on the CTC sample to generate the original Raman spectral spatial matrix. The original Raman spectral spatial matrix is ​​then purified to generate a coordinate map of metabolically disordered cells. Based on the coordinate map of metabolic disorder cells and the topological malignancy index, a three-dimensional heat map algorithm is used to map the spatial distribution of cancer cells, generate a three-dimensional risk heat map of ovarian cancer, and determine the risk of ovarian cancer based on the three-dimensional risk heat map, generating an ovarian cancer health status monitoring report.

2. The method for diagnosing ovarian cancer health status as described in claim 1, characterized in that: The steps for obtaining sterile exosome suspensions and CTC samples are as follows. Peripheral venous blood was collected from the patient and aliquoted into two centrifuge tubes to obtain whole blood vessel A and whole blood vessel B. The whole blood vessel A was centrifuged and resuspended in buffer solution with exosome precipitation reagent to generate a sterile exosome suspension. The blood cell layer of the whole blood vessel was separated by differential centrifugation to obtain a leukocyte-rich precipitate; The leukocyte-rich precipitate was diluted and injected into a microfluidic chip to obtain CTC samples.

3. The method for diagnosing ovarian cancer health status detection as described in claim 1, characterized in that: The steps for calculating the viscoelastic modulus are as follows: A sterile exosome suspension was injected into a silicon nanoneedle array chip, and surface acoustic waves were applied to generate chip-fixed exosomes. A constant pressure is applied to the chip-fixed exosomes, and the pressure indentation depth of the chip-fixed exosomes is recorded. The viscoelastic modulus is then calculated.

4. The method for diagnosing ovarian cancer health status as described in claim 3, characterized in that: The steps for calculating the topological malice index are as follows: The viscoelastic modulus is screened based on a preset modulus threshold, and the chip space coordinates are marked to obtain a set of coordinates for high-risk exosomes. Based on the coordinate set of high-risk exosomes, a high-resolution AFM probe was used to scan the surface topology, identify triangular star-shaped protrusions, generate positive signals of star-shaped structures, and calculate the proportion of positive signals of star-shaped structures to determine the topological malignancy index.

5. The method for diagnosing ovarian cancer health status detection as described in claim 1, characterized in that: The steps for dynamically configuring the air conditioning Raman scanning parameters are as follows: Based on predefined interval rules, the topological malignancy index is divided into three levels, generating a topological malignancy index classification label; Based on the topological malignancy index classification label, the preset scan configuration database is invoked to obtain the time-controlled Raman scan parameters.

6. The method for diagnosing ovarian cancer health status detection as described in claim 5, characterized in that: The steps for generating customized scanning instructions are as follows: Based on the topological malignancy index, the warning threshold for metabolic flux amplitude is calculated using a linear decay formula. The metabolic flux amplitude warning threshold and time-controlled Raman scan parameters are integrated and converted into binary code to generate customized scan instructions.

7. The method for diagnosing ovarian cancer health status as described in claim 1, characterized in that: The generation of the original Raman spectral spatial matrix refers to the process of driving the stage to perform time-controlled Raman scanning on the CTC sample based on customized scanning instructions, simultaneously acquiring lipid and Raman shift intensities, and generating the original Raman spectral spatial matrix.

8. The method for diagnosing ovarian cancer health status as described in claim 7, characterized in that: The steps for generating a coordinate map of metabolically disordered cells are as follows. Wavelet transform is applied to the original Raman spectrum spatial matrix for noise reduction and background signal removal to generate a net metabolic intensity distribution map. Based on the net metabolic intensity distribution map, the metabolic flux amplitude is calculated and compared with the metabolic flux amplitude warning threshold to generate an excess coordinate marker queue. Based on the out-of-range coordinates, a two-dimensional thermal layer is rendered in real time to generate a coordinate map of metabolically disordered cells.

9. The method for diagnosing ovarian cancer health status as described in claim 1, characterized in that: The steps for generating the three-dimensional risk heat map of ovarian cancer are as follows. Based on the coordinate map of metabolic disordered cells and the topological malignancy index, the coordinate system of the potassium bromide slide was transformed to three-dimensional space through affine transformation to generate a spatial registration dataset. Based on the spatial registration dataset, coordinate mapping was performed, and the topological malignancy index was used to drive the color gradation changes to generate a three-dimensional risk heat map of ovarian cancer.

10. The method for diagnosing ovarian cancer health status as described in claim 9, characterized in that: The steps for generating an ovarian cancer health status report are as follows: Based on the three-dimensional risk heat map of ovarian cancer, the weighted score of malignant clusters is calculated, and a list of spatial distribution of malignant clusters is generated. Based on the list of spatial distribution of malignant clusters, a dual-threshold risk classification is performed to generate a risk classification code. By integrating a three-dimensional risk heat map of ovarian cancer, a list of spatial distributions of malignant clusters, and risk classification codes, an ovarian cancer health status monitoring report is generated.

Images