Gastric cancer evaluation system and device based on platelet subcellular structure

Abstract

The invention relates to a gastric cancer evaluation system based on a platelet subcellular structure, comprising: a data acquisition module for acquiring a carcino-embryonic antigen level in a blood sample of a detected subject and a platelet proportion in which alpha particles are regularly distributed; the data analysis and judgment module is used for jointly calculating the risk score of the subject suffering from the gastric cancer according to the platelet proportion of'regular distribution 'of alpha particles and the carcino-embryonic antigen level; and the result output module is used for outputting a corresponding gastric cancer assessment result according to the risk score. According to the platelet proportion of'regular distribution 'of alpha particles in a blood sample of a subject and the carcino-embryonic antigen level, the risk score of the subject suffering from the gastric cancer is jointly calculated, the risk of the subject suffering from the gastric cancer is judged, the specificity is 87.5%, and the sensitivity is 87.5%.

Description

Technical Field

[0001] This invention relates to the field of disease assessment system technology, and in particular to a gastric cancer assessment system and device based on platelet subcellular structure. Background Technology

[0002] Gastric cancer is a highly prevalent and deadly malignant tumor. Despite significant advancements in diagnostic and treatment techniques, the prognosis for patients with advanced-stage gastric cancer remains poor, with an overall survival rate of less than 40%. However, early detection can increase the 5-year survival rate to over 90%. Therefore, accurate diagnosis is crucial for prolonging survival. While endoscopic biopsy is considered the gold standard for diagnosis, its invasiveness and dependence on equipment limit its widespread application. Upper gastrointestinal endoscopy can cause patient discomfort, posing numerous challenges to early screening. Liquid biopsy based on bio-venous blood, with its non-invasive nature and good patient compliance, has become a powerful complement to endoscopy and holds promise for promoting the large-scale application of early gastric cancer screening and diagnosis.

[0003] Existing non-invasive biomarkers for gastric cancer diagnosis include traditional serum tumor markers such as AFP and CEA. These markers are primarily used for screening and monitoring, rather than early detection. Other markers, such as pepsinogen and anti-Helicobacter pylori IgG antibodies, exhibit moderate sensitivity for precancerous lesions of the stomach. Circulating tumor cells (CTCs) may play a crucial role in monitoring gastric cancer spread and assessing treatment response, but their levels are extremely low. Exosomal circRNAs play an important role in the development and progression of gastric cancer, but isolation techniques are complex. ctDNA, as an early cancer detection tool, is low in quantity and unevenly released, posing numerous challenges to detection. Currently, there is a lack of non-invasive, convenient, and intelligent analytical biomarkers for gastric cancer liquid biopsy. Developing such biomarkers is crucial for tumor screening, diagnosis, prognostic assessment, and monitoring of recurrence after treatment.

[0004] Platelets play a crucial role not only in hemostasis and thrombosis but also in interacting with tumor cells, thus influencing cancer progression. Alpha granules, organelles within platelets, contain contents such as RNA and proteins, which have shown potential value in cancer diagnosis. However, the concentrations of certain alpha granule contents (such as endothelin-1) are extremely low, making accurate quantification difficult with conventional ELISA or Western blot, requiring highly sensitive techniques. Furthermore, the specificity of detection indicators is insufficient. For example, while indicators such as GMP-140 (P-selectin) and β-TG released from alpha granules are significantly elevated in gastric cancer patients, these markers are also widely present in platelet activation processes in other diseases (such as atherosclerosis and inflammatory diseases), and cannot specifically point to gastric cancer. Pro-angiogenic factors such as PDGF-A can be released from platelet alpha granules or secreted by tumor cells themselves; existing detection methods (such as ELISA) cannot clearly distinguish the contributions of either, leading to biased interpretation of results. Summary of the Invention

[0005] The inventors discovered through statistical research that there are significant differences in platelet counts between gastric cancer patients and healthy individuals. Specifically, gastric cancer patients show a significantly higher proportion of platelets with a "regularly distributed" alpha particle pattern (regular distribution is defined as the alpha particle fluorescence signal in a single platelet forming a regular ring). This difference stems from the specific regulation of platelet activation by the gastric cancer microenvironment, making platelets with a "regularly distributed" alpha particle pattern a potential biomarker for gastric cancer.

[0006] This invention provides a gastric cancer assessment system and device based on platelet subcellular structure. The system calculates the risk score of a subject for gastric cancer based on the proportion of platelets with "regularly distributed" α particles and the level of carcinoembryonic antigen (CEA, a common clinical indicator). This solves the problems of insufficient specificity, limited sensitivity, and cumbersome detection process in the existing technology of using α particle contents to diagnose gastric cancer.

[0007] The technical solution provided by this invention is as follows: In a first aspect, the present invention provides a gastric cancer assessment system based on platelet subcellular structure, comprising: The data acquisition module is used to acquire the level of carcinoembryonic antigen and the percentage of platelets with "regular distribution" of alpha particles in the blood samples of the test subjects; The data analysis and judgment module is used to calculate the risk score of gastric cancer in subjects based on the combined proportion of platelets with "regular distribution" of alpha particles and the level of carcinoembryonic antigen. The results output module is used to output the corresponding gastric cancer assessment results based on the risk score.

[0008] In conjunction with the first aspect of the invention, some embodiments include: calculating a risk score for gastric cancer in a subject based on a combination of the percentage of platelets with a “regularly distributed” α-particle pattern and the level of carcinoembryonic antigen (CEA), including: The risk score for gastric cancer in the subjects was calculated using the following mathematical formula: logit[p] =-13.910 + 0.167·X1 + 1.349·X2; Where p is the risk score, X1 is the percentage of platelets with "regularly distributed" α particles, and X2 is the carcinoembryonic antigen level.

[0009] In conjunction with the first aspect of the present invention, some embodiments include: outputting a corresponding gastric cancer assessment result based on the risk score, including: when the risk score p is greater than 0.339, the gastric cancer assessment result is high risk.

[0010] In conjunction with the first aspect of the invention, in some embodiments: the gastric cancer assessment system is used to distinguish between patients with benign gastric diseases and patients with cancer.

[0011] In conjunction with the first aspect of the present invention, in some embodiments: obtaining the level of carcinoembryonic antigen and the percentage of platelets with "regularly distributed" α-granules in the blood sample of the test subject includes: The received super-resolution fluorescence images of platelets were analyzed; Obtain the total number of platelets in a super-resolution fluorescence image of platelets; For platelets with a "regular distribution" of α-particles, count the platelets and calculate the percentage of platelets with a "regular distribution" of α-particles.

[0012] In conjunction with the first aspect of the present invention, in some embodiments: the platelets in which the α particles are "regularly distributed" are platelets in which the particle signals in the fluorescence image of α particles in a single platelet are wrapped in a regular ring.

[0013] In conjunction with the first aspect of the present invention, in some embodiments: the blood sample is obtained from venous blood through platelet separation, platelet fixation, and platelet immunostaining steps.

[0014] In conjunction with the first aspect of the present invention, in some embodiments: the gastric cancer assessment system further includes: The sample collection module is used to collect venous blood from the subject. Platelet separation and fixation module for separating platelets from collected venous blood; Platelet fixation module, used to fix the separated platelets; The platelet immunostaining module is used for alpha particle fluorescence staining of fixed platelets.

[0015] In conjunction with the first aspect of the present invention, in some embodiments: the gastric cancer assessment system further includes a super-resolution imaging module for acquiring super-resolution fluorescence images of platelet α particles in the blood sample.

[0016] Secondly, the present invention provides a gastric cancer assessment device based on platelet subcellular structure, including the above-mentioned gastric cancer assessment system based on platelet subcellular structure.

[0017] Compared with the prior art, the present invention has the following beneficial effects: This invention calculates the risk score of gastric cancer in subjects by combining the proportion of platelets with "regular distribution" of α particles in their blood samples and the level of carcinoembryonic antigen, and determines the risk of gastric cancer in subjects. The training set sensitivity is 87.5% and the specificity is 87.5%; the validation set sensitivity is 85.7% and the specificity is 57.1%. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0019] Figure 1 : A schematic diagram of the gastric cancer assessment system based on platelet subcellular structure provided by the present invention.

[0020] Figure 2 A schematic diagram illustrating the classification of platelets based on the distribution pattern of α-granules within a single platelet.

[0021] Figure 3 Platelet imaging images of healthy individuals, patients with benign gastritis, and patients with gastric cancer.

[0022] Figure 4 Platelet classification statistics of healthy individuals, patients with benign gastritis, and patients with gastric cancer.

[0023] Figure 5 The ROC curve for distinguishing between patients with benign gastric disease and gastric cancer based on the proportion of platelets with "regularly distributed" alpha granules (AUC value: 0.859, P=0.016). The ROC curve for distinguishing between patients with benign gastric disease and gastric cancer based on the combined proportion of platelets with "regularly distributed" alpha granules and CEA (AUC value: 0.906, P-0.006).

[0024] Figure 6The ROC curve of the validation set for distinguishing between patients with benign gastric disease and gastric cancer was obtained by combining the proportion of platelets with "regularly distributed" alpha particles and CEA (AUC value: 0.878, P=0.018). Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with the embodiments of this invention. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0026] Existing methods for detecting alpha granules mainly include flow cytometry, enzyme-linked immunosorbent assay (ELISA), super-resolution fluorescence imaging, chemiluminescence immunoassay (CLIA), and electron microscopy. Among these, super-resolution fluorescence imaging is used to detect the distribution and quantity of alpha granules. For example, imaging platelets using structured illumination super-resolution microscopy (SIM) can clearly observe the subcellular distribution pattern of alpha granules, providing detailed information about alpha granules within individual platelets.

[0027] Chinese patent document CN115661074 A (published on January 31, 2023) discloses a platelet classification method and system based on alpha particle super-resolution images. This classification method classifies platelets according to the distribution pattern of alpha particles, specifically into three categories: "regular distribution", "aggregated distribution" and "scattered distribution".

[0028] Chinese patent document CN 116958694 A (published on October 27, 2023) Example 2 discloses a platelet classification system based on α particles. After extracting images of individual platelet cells using deep learning methods, the system classifies, detects, counts, and statistically analyzes the submicroscopic structure of platelets.

[0029] This invention, referencing CN115661074 A and CN 116958694 A, classifies platelets according to the distribution pattern of α-granules within a single platelet. Statistical analysis of platelet classification in venous blood from gastric cancer patients and healthy individuals revealed a significant difference between the classification results of platelets from gastric cancer patients and healthy individuals. However, when classifying platelets according to the distribution pattern of platelet microtubules, no significant difference was found between the classification results of platelets from gastric cancer patients and healthy individuals. Based on this, this invention provides a gastric cancer assessment system and device based on the subcellular structure of platelets, addressing the problems of poor specificity and low sensitivity in existing technologies that determine whether a subject has gastric cancer based on the content level of α-granules.

[0030] The gastric cancer assessment system based on platelet subcellular structure provided by this invention includes: The data acquisition module is used to acquire the level of carcinoembryonic antigen and the percentage of platelets with "regular distribution" of alpha particles in the blood samples of the test subjects; The data analysis and judgment module is used to calculate the risk score of gastric cancer in subjects based on the combined proportion of platelets with "regular distribution" of alpha particles and the level of carcinoembryonic antigen. The results output module is used to output the corresponding gastric cancer assessment results based on the risk score.

[0031] The percentage of platelets with a "regularly distributed" alpha particle count refers to the proportion of platelets with a "regularly distributed" alpha particle count in a blood sample out of all platelets. Carcinoembryonic antigen (CEA) level, i.e., the concentration of CEA in serum, is derived from hospital clinical data.

[0032] This gastric cancer assessment system achieves accurate risk assessment by combining the proportion of platelets with a "regularly distributed" alpha particle pattern in the blood sample of the subject with the level of carcinoembryonic antigen (CEA). The system includes a data acquisition module to obtain the proportion of regularly distributed alpha particles, a data analysis and judgment module to calculate a risk score based on a joint model of the proportion of regularly distributed alpha particles and CEA levels, and a results output module to directly output the gastric cancer assessment result. This effectively solves the problems of insufficient specificity, limited sensitivity, and cumbersome detection procedures in existing technologies for gastric cancer diagnosis.

[0033] In some embodiments of the present invention: the risk score of a subject for gastric cancer is calculated based on the combined proportion of platelets with "regularly distributed" α-particles and the level of carcinoembryonic antigen, including: calculating the risk score of a subject for gastric cancer using the following mathematical formula: logit[p] = -13.910 + 0.167·X1 + 1.349·X2; where p is the risk score, X1 is the percentage of platelets with "regular distribution" of α particles, and X2 is the carcinoembryonic antigen level.

[0034] The inventors collected blood samples from patients with gastric cancer, benign gastric cancer, and healthy individuals. They measured the proportion of platelets with a "regularly distributed" alpha particle pattern (X1) and the level of carcinoembryonic antigen (X2). Optimal coefficients were determined through statistical modeling (using maximum likelihood estimation), ensuring a high degree of match between the model's predicted gastric cancer risk probability and actual clinical diagnostic results. The coefficients in the formula (0.167, 1.349) were optimized through ROC curve analysis (the maximum value of the Youden's index corresponds to the threshold), ensuring optimal model performance in distinguishing between patients with gastric cancer and those with benign gastric diseases. Clinical validation showed that the system maintained a sensitivity of 87.5% while improving specificity to 87.5%, achieving an accuracy of 90.6%, significantly outperforming traditional detection methods.

[0035] In some embodiments of the present invention: A gastric cancer assessment result is output based on the risk score, including: when the risk score p is greater than 0.339, the gastric cancer assessment result is high risk. The Youden index = sensitivity + specificity - 1. This index is used to measure the ability of a diagnostic test to distinguish between patients and healthy individuals; the maximum value corresponds to the optimal threshold. Calculations show that when p = 0.339, the Youden index reaches its maximum (approximately 0.750), at which point sensitivity and specificity achieve the best balance.

[0036] In some embodiments of the present invention, the gastric cancer assessment system is used to distinguish between patients with benign gastric diseases and those with cancer. For example, in cases where a patient is diagnosed with gastric disease and suspected of having gastric cancer, fasting peripheral venous blood is collected, and the subcellular structural distribution characteristics of platelet α-granules are analyzed using super-resolution microscopy. The proportion of platelets with "regularly distributed" α-granules and the CEA level (derived from hospital clinical data) are input into a risk scoring model to calculate a comprehensive risk score (p-value). When the risk score p > 0.339, the system determines a high risk of gastric cancer, effectively distinguishing between patients with benign gastric diseases (such as gastric polyps and gastric ulcers) and those with gastric cancer.

[0037] In some embodiments of the present invention: obtaining the level of carcinoembryonic antigen and the percentage of platelets with "regularly distributed" α-granules in the blood sample of the test subject includes: The received super-resolution fluorescence images of platelets were analyzed; Obtain the total number of platelets in a super-resolution fluorescence image of platelets; For platelets with a "regular distribution" of α-particles, count the platelets and calculate the percentage of platelets with a "regular distribution" of α-particles.

[0038] This invention significantly improves the performance of gastric cancer screening through an automated process of "super-resolution fluorescence image analysis → total platelet count → calculation of the proportion of regular distribution": It utilizes super-resolution imaging technology (resolution ~100nm) to accurately capture the microscopic features of the "regular distribution" of alpha particles, overcoming the bottleneck of traditional detection methods in identifying extremely low concentrations of alpha particle contents, achieving a sensitivity of 87.5%; it eliminates human observation errors through automatic statistical calculation of proportions using algorithms, and after combined analysis with carcinoembryonic antigen levels, the specificity is increased to 87.5%, achieving high sensitivity, high specificity, and high efficiency in early non-invasive gastric cancer screening, effectively solving the core problems of insufficient specificity, high false negative rate, and cumbersome operation of existing technologies.

[0039] In some embodiments of the present invention: the platelets with "regularly distributed" α particles are those in which the particle signals in the fluorescence image of a single platelet α particles are wrapped in regular rings. After collecting blood samples, the inventors used super-resolution fluorescence imaging technology to observe the microscopic morphology of platelet α particles and found that some platelets in the blood have "regularly distributed" α particles, that is, the particle signals are wrapped in regular rings. Figure 2 ).

[0040] In some embodiments of the present invention: the blood sample is obtained from venous blood through platelet separation, platelet fixation, and platelet immunostaining. Platelet separation precisely removes interfering components such as red blood cells and white blood cells, ensuring that subsequent analysis focuses on the target platelets; platelet fixation stabilizes the subcellular structure of α-granules, preventing morphological distortion during imaging; platelet immunostaining enables fluorescent specific labeling of extremely low concentrations of α-granule contents, providing clear microscopic features for super-resolution imaging. In the embodiments, super-resolution fluorescence images of platelet α-granules and serum CEA levels of each subject are detected based on blood samples from training and validation sets, wherein the CEA data of these subjects are derived from clinical data from hospitals.

[0041] To construct a standardized and highly accurate sample processing workflow, in some embodiments of the present invention, the gastric cancer assessment system further includes: The sample collection module is used to collect venous blood from the subject. Platelet separation and fixation module for separating platelets from collected venous blood; Platelet fixation module, used to fix the separated platelets; The platelet immunostaining module is used for alpha particle fluorescence staining of fixed platelets.

[0042] Furthermore, this invention integrates a super-resolution imaging module (with a resolution of up to 100 nm) into the gastric cancer assessment system. This module can accurately capture the distribution of α-particles on platelets. In some embodiments of this invention, the gastric cancer assessment system further includes a super-resolution imaging module for acquiring super-resolution fluorescence images of platelet α-particles in the blood sample.

[0043] The present invention provides a gastric cancer assessment device based on platelet subcellular structure, which includes the above-mentioned gastric cancer assessment system based on platelet subcellular structure.

[0044] The technical solution of the present invention will be described in detail below through specific embodiments.

[0045] The ACDT solution used in the following examples was: 10% citrate glucose and 90% Tyrode buffer, with Tyrode buffer consisting of 138 mM NaCl, 2.9 mM KCl, 2 mM MgCl, 4 mM NaH2PO4, 12 mM NaHCO3, 5.5 mM glucose, and 10 mM HEPES; the fixative was 8% paraformaldehyde aqueous solution; the primary antibody was anti-VWF primary antibody; and the secondary antibody was goat anti-rabbit IgG H&L (Alexa Fluor). ® 488); PFA refers to paraformaldehyde.

[0046] Example 1: Using the distribution pattern of platelet α-granules to differentiate between patients with benign gastric diseases (i.e., patients with gastric polyps and chronic gastritis) and patients with gastric cancer. This embodiment involves a sample collection module, a platelet separation and fixation module, an immunostaining module, a super-resolution imaging module, a data acquisition module, a data analysis and judgment module, and a result output module, as detailed below: 1.1 Sample Acquisition Module Peripheral venous blood of 2 mL was collected from 13 healthy individuals, 15 patients with benign gastritis, and 15 patients with gastric cancer. The blood was placed in EDTA anticoagulant blood collection tubes and kept at room temperature. Platelet fixation was completed within 24 hours.

[0047] 1.2 Platelet separation and fixation module (1) Platelet separation module: Collect 2 mL of whole blood into an EDTA anticoagulated blood collection tube, then centrifuge at 200 g for 12 min to separate the platelet-rich supernatant into a 10 mL centrifuge tube, add ACDT solution, and then incubate the platelets in a 37 ℃ incubator containing 5% CO2 for 2 h.

[0048] (2) Platelet fixation module: Remove the platelets from the incubator, add an equal volume of fixative to the centrifuge tube, and let stand for 30 min. After fixation, place the centrifuge tube in a centrifuge and centrifuge at 1500 g for 3 min.

[0049] 1.3 Platelet Immunostaining Module ① After diluting the fixed platelets, spread them in a confocal dish pre-coated with poly-L-lysine, let them stand for 30 min to 1 h, and observe the platelet density under a microscope. If the density is appropriate, wash once with PBS and air dry.

[0050] ② Add 0.2 mL of 0.2% Triton X-100 solution to the confocal dish, let it stand for 10 min, then remove the liquid from the confocal dish, add 0.2 mL of blocking solution, and let it stand at room temperature for 20~30 min.

[0051] ③ Dilute the primary antibody in blocking buffer to obtain a primary antibody solution, and add 0.2 mL of the primary antibody solution to the confocal dish. Place the confocal dish in a 4 ℃ refrigerator overnight for primary antibody incubation. After the primary antibody incubation is complete, aspirate the primary antibody dilution buffer. Add 0.2 mL of washing buffer, and place the confocal dish on a shaker to rinse for 5 min. Finally, aspirate the washing buffer, and repeat 5 times.

[0052] ④ Dilute the secondary antibody in platelet blocking solution, then add 0.2 mL of the secondary antibody dilution solution to the confocal dish, protect from light, and incubate at room temperature for 1-2 hours. After incubation, wash 5 times using the same method as for washing the primary antibody.

[0053] ⑤ Add 0.2 mL of 4% PFA to the confocal dish in step ④, let it stand at room temperature for 10 min, then remove the fixative and wash twice with PBS.

[0054] 1.4 Super-resolution imaging module The fluorescence intensity and exposure time were adjusted for each sample to ensure that each reconstructed image had a high signal-to-noise ratio. Then, super-resolution microscopy and structured illumination microscopy (SIM) were used to image each stained confocal dish, and imaging data of 500 platelets were obtained from each confocal dish.

[0055] 1.5 Data Acquisition Module Super-resolution fluorescence images were input into the ResUNet and ResNet-50 algorithm models. Following the methods described in Chinese patent documents CN115661074 A and CN 116958694 A, platelets were classified according to the distribution pattern of α-particles, specifically into three categories: "regular distribution," "aggregated distribution," and "scattered distribution." The "scattered distribution" category was further divided into two types based on the number of α-particles within the platelet: "N < 30" and "N ≥ 30." The final platelet count was as follows: Figure 2 Platelets are categorized into four types: "N < 30", "N ≥ 30", "Regular Distribution", and "Aggregated Distribution". "Regular Distribution" α-particles represent a regular distribution of α-particles within a single platelet; "Aggregated Distribution" α-particles represent aggregated α-particles within a single platelet; "N < 30" platelets represent fewer than 30 scattered α-particles within a single platelet; and "N ≥ 30" platelets represent 30 or more scattered α-particles within a single platelet. Statistical analysis was conducted to examine the percentage differences in each type of platelet among healthy individuals, patients with benign gastric diseases, and patients with gastric cancer. The results are as follows: Figure 4 As shown, there are significant differences in the proportion of platelets with "regularly distributed" alpha particles between healthy individuals and patients with gastric cancer, benign gastritis, and gastric cancer.

[0056] 1.6 Analysis and Judgment Module Furthermore, receiver operating characteristic curve (ROC) analysis is mainly used to evaluate the effectiveness of a certain indicator in classifying or differentiating two types of test subjects (such as benign or malignant), and to find the optimal critical value of the indicator, thereby determining the critical value of this evaluation indicator.

[0057] Sensitivity (True Positive Rate = TPR) refers to the probability of a test result being positive, assuming it is actually positive.

[0058] Specificity (True negative rate, TNR) refers to the probability that a test result is negative given a true negative result.

[0059] The Youden Index, also known as the correctness index, is a method commonly used when it is assumed that the harm of false negatives (missed diagnoses) and false positives (misdiagnoses) is equally significant. It reflects the overall ability of true gastric cancer patients versus non-patients.

[0060] The Youden index is the sum of sensitivity and specificity minus 1; a higher Youden index indicates greater validity. Furthermore, the value of the test variable corresponding to the maximum Youden index is the diagnostic critical value for this method.

[0061] Yoden Index = Sensitivity + Specificity - 1 like Figure 5 As shown, ROC curve analysis between patients with benign gastric disease and gastric cancer indicates that the proportion of platelets with "regularly distributed" α-particles can distinguish between benign and malignant gastric cancer. The area under the ROC curve was 0.859 (95% significance C1 0.672 to 1.000), with a significance level of P=0.016. The proportion of platelets with "regularly distributed" α-particles distinguishing between patients with benign gastric disease and gastric cancer was 66.70%, which can be used as a cutoff value for distinguishing between benign gastric disease and gastric cancer. When the proportion of platelets with "regularly distributed" α-particles is greater than or equal to 66.70%, it can be identified as gastric cancer, with a sensitivity of 87.5% and a specificity of 75.0%. This indicates that the proportion of platelets with "regularly distributed" α-particles can effectively distinguish between patients with benign gastric disease and gastric cancer.

[0062] Example 2: Further assessment of gastric cancer risk using the proportion of platelets with "regularly distributed" α-particles and the clinical tumor marker CEA. This embodiment combines two indicators—the proportion of platelets with "regularly distributed" α-particles and the clinical indicator CEA level—to assess the risk of gastric cancer. The combined ROC curve of the two indicators is shown in the figure below. Figure 6 As shown, the area under the ROC curve (AUC) of the training set for the combined detection of the two methods was 0.906, P=0.006. The sensitivity and specificity of the detection were 87.5% and 87.5%, respectively. Compared with the method in Example 1 that uses the proportion of platelets with "regularly distributed" α particles to assess the risk of gastric cancer alone, this method can significantly improve the specificity of gastric cancer risk assessment.

[0063] The technical solution of this embodiment will be described in detail below: This study aims to evaluate the diagnostic efficacy of a combined indicator of two factors for gastric cancer: the proportion of platelets with a "regularly distributed" alpha particle pattern and the clinical indicator CEA level. The disease state diagnosed by the gold standard was used as the dependent variable Y, where Y=0 represents benign gastric disease and Y=1 represents gastric cancer. The proportion of platelets with a "regularly distributed" alpha particle pattern was defined as the independent variable X1, and the CEA level was defined as the independent variable X2. A joint predictive model was constructed using binary logistic regression to estimate the conditional probability p=P(Y=1|X1,X2) under the combined effects of X1 and X2. The two independent variables were simultaneously included in SPSS using the ENTER method, yielding the following model and statistical results.

[0064] The binary logistic joint model is statistically significant overall (χ²). 2 =9.93, P=0.007), good fit (HLP=0.462), explanatory power R 2The Logit equation obtained from this is approximately 0.46–0.62: logit[p] = -13.910 + 0.167·X1 + 1.349·X2. Based on this, the individualized probability of gastric cancer can be calculated as p = 1 / {1 + exp[-(-13.910 + 0.167·X1 + 1.349·X2)]}.

[0065] Following the established procedure, the predicted probability p calculated by the above regression model was used as the test variable for the ROC curve. Using the gold standard diagnosis as the state variable, the ROC curve was further plotted, and the area under the curve (AUC) and its 95% confidence interval were calculated. The optimal cutoff point was determined when the Youden index was at its maximum, and the corresponding sensitivity and specificity were reported to quantify the discriminative efficacy of the combined indicators. In summary, the regular distribution of α-particles was an independent and significant predictor. The binary logistic model formed by combining CEA levels showed good fit and strong discriminative ability, providing a reliable statistical basis and implementation path for the auxiliary screening and early identification of gastric cancer.

[0066] Based on the construction and testing of the joint model, ROC curve analysis was further conducted. The individual predicted probability p calculated by the binary logistic regression equation was used as the test variable, and the gold standard diagnosis was used as the state variable to plot the ROC curve. According to the ROC coordinate table, sensitivity and specificity were calculated at each threshold, and the optimal cutoff point was determined using the Youden index (J = Sensitivity + Specificity - 1). The results showed that the optimal threshold appeared at a predicted probability of approximately p = 0.339, at which point the sensitivity and specificity were both 87.5%, and the Youden index was 0.750, indicating that this cutoff point achieved a good balance between sensitivity and specificity.

[0067] The area under the curve (AUC) is used to evaluate the overall discriminative power of the joint model.

[0068] The AUC of the single CEA level was 0.750 (SE = 0.134, 95% CI 0.457–0.981, P = 0.141).

[0069] The percentage of platelets with "regularly distributed" alpha particles had an AUC of 0.859 (SE = 0.095, 95% CI 0.672–1.000, P = 0.016).

[0070] The AUC value of the ROC curve plotted based on the combined prediction probability of the two obtained from regression was 0.906, which is higher than the AUC value of the single indicator of CEA level and the AUC value of the single indicator of the proportion of platelets with "regular distribution" of α particles. Moreover, it achieved high sensitivity and specificity at the optimal threshold, suggesting that the combined model has good discrimination and clinical discriminative value.

[0071] In summary, the combined indicator model showed a good fit (Hosmer–Lemeshow P=0.462), was statistically significant (Omnibus test χ²=9.93, P=0.007), and demonstrated high discriminative power in ROC analysis (sensitivity 87.5%, specificity 87.5%, J=0.750 at the optimal cutoff point p≈0.339). These results support the use of α-granule regular distribution combined with CEA for the auxiliary diagnosis and risk stratification of gastric cancer.

[0072] Subsequently, seven patients with benign gastric diseases and seven patients with gastric cancer were selected as clinical samples to conduct external validation of the combined indicator model. The ROC curve of this validation set is shown in [Figure number missing]. Figure 6 The sensitivity was 85.7%, the specificity was 57.1%, and the AUC was 0.878.

[0073] In the description of this specification, the references to terms such as "one embodiment / mode," "some embodiments / modes," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment / mode or example is included in at least one embodiment / mode or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment / mode or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments / modes or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments / modes or examples described in this specification, as well as the features of different embodiments / modes or examples.

[0074] It should be noted that in this invention, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element. In this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise expressly specified.

[0075] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments described herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A gastric cancer assessment system based on platelet subcellular structure, characterized in that, include: The data acquisition module is used to acquire the level of carcinoembryonic antigen and the percentage of platelets with "regular distribution" of alpha particles in the blood samples of the test subjects; The data analysis and judgment module is used to calculate the risk score of gastric cancer in subjects based on the combined proportion of platelets with "regular distribution" of alpha particles and the level of carcinoembryonic antigen. The results output module is used to output the corresponding gastric cancer assessment results based on the risk score.

2. The gastric cancer assessment system based on platelet subcellular structure as described in claim 1, characterized in that: The risk score for gastric cancer in subjects was calculated based on the combined percentage of platelets with a "regular distribution" of alpha granules and the level of carcinoembryonic antigen (CEA). This score included: The risk score for gastric cancer in the subjects was calculated using the following mathematical formula: logit[p] =-13.910 + 0.167·X1 + 1.349·X2; Where p is the risk score, X1 is the percentage of platelets with "regularly distributed" α particles, and X2 is the carcinoembryonic antigen level.

3. The gastric cancer assessment system based on platelet subcellular structure as described in claim 2, characterized in that: Based on the aforementioned risk score, the corresponding gastric cancer assessment results are output, including: When the risk score p is greater than 0.339, the gastric cancer assessment result is high risk.

4. The gastric cancer assessment system based on platelet subcellular structure as described in claim 1, characterized in that: The gastric cancer assessment system is used to distinguish between patients with benign gastric diseases and those with cancer.

5. The gastric cancer assessment system based on platelet subcellular structure as described in claim 1, characterized in that: The acquisition of carcinoembryonic antigen levels and the percentage of platelets with a "regular distribution" of alpha granules in the blood samples of the test subjects includes: The received super-resolution fluorescence images of platelets were analyzed; Obtain the total number of platelets in a super-resolution fluorescence image of platelets; For platelets with a "regular distribution" of α-particles, calculate the percentage of platelets with a "regular distribution" of α-particles.

6. The gastric cancer assessment system based on platelet subcellular structure as described in claim 1, characterized in that: The α-particles of platelets are "regularly distributed" in the fluorescence image of a single platelet, where the particle signals are wrapped in regular rings.

7. The gastric cancer assessment system based on platelet subcellular structure as described in claim 1, characterized in that: The blood sample was obtained from venous blood through platelet separation, platelet fixation, and platelet immunostaining.

8. The gastric cancer assessment system based on platelet subcellular structure as described in claim 1 or 7, characterized in that: The gastric cancer assessment system also includes: The sample collection module is used to collect venous blood from the subject. Platelet separation and fixation module for separating platelets from collected venous blood; Platelet fixation module, used to fix the separated platelets; The platelet immunostaining module is used for alpha particle fluorescence staining of fixed platelets.

9. The gastric cancer assessment system based on platelet subcellular structure as described in claim 1 or 7, characterized in that: The gastric cancer assessment system also includes a super-resolution imaging module, which is used to acquire super-resolution fluorescence images of platelet α particles in the blood sample.

10. A gastric cancer assessment device based on platelet subcellular structure, characterized in that: The gastric cancer assessment system based on platelet subcellular structure as described in any one of claims 1 to 9.

Images