Brain glioma diagnosis system based on platelet alpha particle super-resolution imaging
Through the super-resolution imaging technology of platelet α particles, the area value and quantity ratio of platelet α particles were calculated, and a diagnostic model was established, which solved the problem of insufficient sensitivity and specificity of existing brain glioma diagnosis methods, and achieved high-accuracy non-invasive diagnosis.
Patent Information
- Application Number
- CN202310643454.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-06-01
- Publication Date
- 2024-12-06
AI Technical Summary
Existing brain glioma diagnosis methods have problems with low sensitivity and poor specificity, especially the liquid biopsy-based method is insufficient in identifying brain glioma patients, and the invasive biomarker detection methods cause physiological and Psychological burden.
Using a diagnostic system based on platelet α particles super-resolution imaging, the average α particle area value and the proportion of α particles in each platelet 'N<30' is calculated by detecting the super-resolution image of platelet α particles in blood samples. Establish ROC curve evaluation and regression model for diagnosis.
The accuracy of brain glioma diagnosis is achieved to reach 82% to 90%, sensitivity is 86.70%, specificity is 80.00%, and non-invasive, which can quickly and accurately identify patients with brain glioma.
Smart Images

Figure CN119092090A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of clinical testing and diagnosis, and more specifically, relates to a brain glioma diagnosis system based on platelet alpha particle super-resolution imaging. Background Art
[0002] Gliomas are tumors that originate from brain glial cells. They are the most common primary intracranial tumors, accounting for about 75% of adult malignant primary brain tumors. They can be divided into grades 1 to 4, of which grades 1 and 2 are low-grade gliomas, and grades 3 and 4 are high-grade gliomas. Some high-grade gliomas have become the most difficult tumors to treat. In my country, the annual incidence of gliomas is 5 to 8 per 100,000, and the 5-year mortality rate is second only to pancreatic cancer and lung cancer among systemic tumors. Therefore, early screening and accurate diagnosis are conducive to prolonging the survival of patients with gliomas.
[0003] At present, magnetic resonance imaging is usually used to diagnose brain gliomas, but this imaging examination cannot provide accurate pathological information of the tumor before postoperative histological examination and has insufficient correlation with the actual tumor burden. In addition, biomarkers such as isocitrate dehydrogenase gene (IDH), O 6 -Methylguanine-DNA methyltransferase gene (MGMT) methylation, etc., can also be used for the screening and diagnosis of brain gliomas, but the tumor-derived DNA of these biomarkers is extremely low in the patient's blood, and requires craniotomy or lumbar puncture to obtain cerebrospinal fluid for testing. This invasive diagnostic method will bring great physical and psychological burden to patients. According to statistics, only 49.4% of brain glioma patients have these biomarkers of tumor-derived DNA isolated from cerebrospinal fluid, and their diagnostic accuracy is poor.
[0004] Existing studies have demonstrated the applicability and potential of liquid biopsy in cancer diagnosis, and this liquid biopsy is less invasive to patients. In cancer diagnosis research, a variety of blood-based biomarkers such as circulating tumor cells and circulating tumor DNA have been found. This is mainly because there are a small number of free circulating tumor cells (CTC) in the blood of cancer patients and necrotic cancer cells also release a small amount of circulating tumor DNA (ctDNA) into the blood. Therefore, cancer can be detected by detecting CTC and ctDNA in the blood. However, these blood biomarkers also generally have extremely low content in patients with gliomas, and generally 20% to 73% of glioma patients can be identified through circulating tumor cells, while about 10% of glioma patients can be identified through plasma ctDNA, and the diagnostic accuracy is poor. Therefore, it is of great clinical and socioeconomic significance to find highly accurate biomarkers based on liquid biopsy and develop a non-invasive glioma diagnosis system with high accuracy and sensitivity. Summary of the invention
[0005] In view of the above defects or improvement needs of the prior art, the present invention provides a glioma diagnosis system based on platelet α granule super-resolution imaging, the purpose of which is to detect the subcellular structure of platelets in sample blood, find that the average area value of α granules per platelet in the obtained platelet α granule super-resolution image has significant differences between healthy people and glioma patients, and the average area value of α granules per platelet can better distinguish healthy people from glioma patients through ROC curve evaluation, and can be used as a biomarker; the glioma diagnosis system established based on the biomarker, according to which the larger the average area value of α granules per platelet, the greater the possibility of diagnosing glioma, the accuracy of the diagnosis system can reach 82%, especially when the average area value of α granules per platelet and the proportion of platelets with α granules "N<30" are used as joint indicators for diagnosis, the system diagnosis accuracy can reach 90% with high sensitivity and good specificity, thereby solving the technical problem of low sensitivity and poor specificity in judging glioma based on liquid biopsy results.
[0006] To achieve the above-mentioned object, according to one aspect of the present invention, a brain glioma diagnosis system based on platelet alpha granule super-resolution imaging is provided, which comprises a platelet alpha granule super-resolution image acquisition module, an image analysis module, and a brain glioma diagnosis module;
[0007] The platelet alpha granule super-resolution image acquisition module is used to perform binarization to obtain a plurality of platelet alpha granule super-resolution images consisting of fluorescent signal pixels and non-fluorescent signal pixels, and submit them to the image analysis module;
[0008] The image analysis module is used to calculate the average area value of α particles in each platelet in the obtained α particle super-resolution image, and submit it to the brain glioma diagnosis module;
[0009] The brain glioma diagnosis module is based on the obtained average α-particle area value in each platelet, and the larger the average α-particle area value in each platelet, the greater the possibility of brain glioma.
[0010] Preferably, the glioma diagnosis system based on platelet α-particle super-resolution imaging, wherein the glioma diagnosis module compares the obtained average α-particle area value in each platelet with the threshold value A and makes the specific judgment as follows:
[0011] If the average α-particle area value of each platelet in the sample is greater than the threshold A, the patient is judged to be a glioma patient;
[0012] If the average area of α particles per platelet in the sample is less than or equal to the threshold value A, the sample is judged to be healthy.
[0013] Preferably, in the glioma diagnosis system based on platelet α-granule super-resolution imaging, the threshold value A is the area value corresponding to the Youden maximum value corresponding to the ROC curve between healthy people and glioma patients.
[0014] Preferably, the threshold value A of the brain glioma diagnosis system based on platelet α particle super-resolution imaging is 0.75 μm. 2 .
[0015] Preferably, the brain glioma diagnosis system based on platelet α-particle super-resolution imaging further comprises a platelet classification and statistics module;
[0016] The platelet classification and statistics module, based on all acquired α-particle super-resolution images, divides platelets into four categories according to the distribution form of α-particles in a single platelet, namely, "regular distribution", "aggregated distribution", α-particle number "N<30" and α-particle number "N≥30", calculates the percentage of platelets with "N<30" in the sample in the total amount of these four categories of platelets, and submits them to the brain glioma diagnosis module.
[0017] Preferably, the glioma diagnosis system based on platelet α particle super-resolution imaging, the glioma diagnosis module thereof, takes the proportion of platelets with α particle number "N<30" in the sample and the average α particle area value in each platelet of the sample as a joint index, and calculates the joint index value by fitting regression, and makes a judgment based on the comparison between the calculated joint index value and the threshold value B, and the specific judgment is as follows:
[0018] If the combined index value in the sample is less than or equal to the threshold B, the sample is diagnosed as a healthy person;
[0019] If the combined index value in the sample is greater than the threshold B, the sample is diagnosed as a brain glioma.
[0020] Preferably, in the glioma diagnosis system based on platelet α-granule super-resolution imaging, the threshold value B is the ratio of the Youden maximum value corresponding to the ROC curve between healthy people and glioma patients.
[0021] Preferably, the threshold value B of the brain glioma diagnosis system based on platelet α-granule super-resolution imaging is 55.38%.
[0022] In general, compared with the prior art, the above technical solution conceived by the present invention can achieve the following beneficial effects because the present invention provides a brain glioma diagnosis system based on platelet α granule super-resolution imaging:
[0023] The brain glioma diagnosis system provided by the present invention is based on liquid biopsy. It obtains an alpha particle super-resolution image by detecting a blood sample, and calculates the average alpha particle area value in each platelet of the sample. The larger the average alpha particle area value in each platelet, the greater the possibility of judging a brain glioma. In particular, if the average alpha particle area value in each platelet of the sample is greater than the critical value obtained by the ROC curve, the sample is judged to be a brain glioma patient. The diagnostic accuracy of the system can reach 82%, and the required detection is less invasive.
[0024] Furthermore, the average area of α particles in each platelet and the proportion of platelets with α particles "N<30" are used as joint indicators, and the predicted probability of the joint indicator is calculated by the regression model, that is, the joint indicator value is obtained. According to the joint indicator value being greater than the threshold value B, the patient is judged to be a glioma patient. The diagnostic accuracy of the system can reach 90%, the sensitivity is 86.70%, and the specificity is 80.00%. The glioma diagnosis system provided by the present invention requires less invasiveness, high sensitivity, and good specificity, and can quickly and accurately identify glioma patients, which is conducive to the early treatment of glioma patients. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] Figure 1 It is the difference analysis result of the average α-particle area value in each platelet between healthy people and patients with brain glioma;
[0026] Figure 2 It is a ROC curve analysis based on the average α-particle area value per platelet to identify healthy people and brain gliomas;
[0027] Figure 3 It is the result of classifying platelets according to the distribution pattern of α particles;
[0028] Figure 4 It is the result of differential analysis of the proportion of different types of platelets between healthy people and glioma patients based on the distribution of α particles;
[0029] Figure 5 It is a ROC curve analysis based on the proportion of platelets with α-particle number "N < 30" to identify healthy people and brain gliomas;
[0030] Figure 6 It is a ROC curve analysis based on the proportion of platelets with α-particle number "N ≥ 30" to identify healthy people and brain gliomas;
[0031] Figure 7 It is an ROC curve analysis based on the proportion of platelets with "regular distribution" of α particles to identify healthy people and brain gliomas;
[0032] Figure 8 It is an ROC curve analysis based on combined indicators to identify healthy people and brain gliomas. DETAILED DESCRIPTION
[0033] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.
[0034] Blood is one of the commonly used test samples in liquid biopsy, and platelets exist in large quantities in human blood. It is speculated that platelets may be a good biomarker for cancer diagnosis. However, in the previous studies, the present invention found that it is difficult to use platelet count and activation for screening and diagnosis of brain gliomas, mainly because the specificity and sensitivity of platelet count and activation are poor; while the detection of platelet RNA after tumor education is limited to screening glioblastoma and it is difficult to avoid interference from cell fragments of similar size, residual nuclear blood cells in platelet separation, and platelet RNA after tumor education, and the accuracy is poor.
[0035] Existing studies have used electron microscopy to detect platelet subcellular structures and found that ovarian cancer can cause changes in the morphology and quantity of platelet subcellular structures (microtubules, mitochondria), but whether these changes can be used to differentiate cancer requires further research. In addition, platelet subcellular structures include organelles such as tubules, mitochondria, microfilaments and microtubules, as well as platelet granules and glycogen granules, among which platelet granules include α granules and dense granules. Whether brain gliomas can cause changes in platelet subcellular structures and how they change is still unknown.
[0036] The present invention detects the subcellular structure of platelets based on super-resolution images and studies the differences between healthy people and patients with brain glioma. The results show that in this experiment, the average area of α particles in each platelet of healthy people is 0.5649 μm 2 The average area of α particles in each platelet of patients with brain glioma is 0.9304 μm 2 The significance analysis showed that the average area of α particles in each platelet was significantly different between healthy people and patients with brain glioma. Further ROC curve analysis showed that the ROC curve area AUC was 0.820, with a high diagnostic accuracy, indicating that the average area of α particles in each platelet can be used as a biomarker, especially the average area of α particles in each platelet can better distinguish between healthy people and patients with brain glioma.
[0037] In addition, the difference in the number of α particles between healthy people and patients with glioma was compared. The results showed that the number of α particles in healthy platelets was mainly concentrated in the range of 5 to 20, with 11 particles accounting for the largest proportion, while the number of α particles in platelets of patients with glioma was mainly concentrated in the range of 5 to 40, with 18 particles accounting for the largest proportion. Based on these differences, combined with the platelet classification method and system based on platelet α particle super-resolution images of patent CN115661074A, further analysis found that the proportion of platelets with α particle number "N<30", α particle number "N≥30" and "regular distribution" were significantly different between healthy people and patients with glioma. ROC curve analysis showed that the proportion of platelets with α particle number "N<30" as a biomarker can better distinguish healthy people from patients with glioma. The ROC curve area AUC is 0.867, which is higher in diagnostic accuracy than the proportion of other types of platelets in distinguishing healthy people from patients with glioma.
[0038] In order to further improve the sensitivity and specificity of clinical diagnosis, the present invention also studied the effect of multi-indicator combined diagnosis. The results showed that the combined α-particle number "N<30" platelet ratio and the average α-particle area value in each platelet can further improve the accuracy of the diagnostic system, and its ROC curve area AUC is 0.900.
[0039] Based on the above findings, the present invention provides a brain glioma diagnosis system based on platelet alpha granule super-resolution images, which includes a platelet alpha granule super-resolution image acquisition module, an image analysis module, and a brain glioma diagnosis module;
[0040] The platelet alpha granule super-resolution image acquisition module is used to perform binarization to obtain a plurality of platelet alpha granule super-resolution images consisting of fluorescent signal pixels and non-fluorescent signal pixels, and submit them to the image analysis module;
[0041] The image analysis module is used to calculate the average area value of α particles in each platelet in the obtained α particle super-resolution image and submit it to the glioma diagnosis module; the average area value of α particles in each platelet refers to the ratio of the sum of the fluorescence areas in the α particle super-resolution image to the total number of platelets in the sample super-resolution image.
[0042] In some embodiments, based on all the acquired α-particle super-resolution images, the fluorescent signal pixels in the acquired super-resolution images are segmented with a single platelet as the segmentation unit, and the fluorescent area of the α-particles in the segmentation unit is calculated, and the total fluorescent area of the α-particles and the number of segmentation units, i.e. the total number of platelets, are counted. The ratio of the two is the average α-particle area value in each platelet.
[0043] The brain glioma diagnosis module is based on the obtained average α-particle area value in each platelet, and the larger the average α-particle area value in each platelet, the greater the possibility of brain glioma.
[0044] Preferably, the brain glioma diagnosis module compares the obtained average area value of α particles in each platelet with the threshold value A to make a judgment, and the specific judgment is as follows:
[0045] If the average α-particle area value of each platelet in the sample is greater than the threshold A, the patient is judged to be a glioma patient;
[0046] If the average α-particle area value of each platelet in the sample is less than or equal to the threshold A, the sample is judged to be healthy;
[0047] The threshold value A is preferably the area value corresponding to the Youden maximum value corresponding to the ROC curve between healthy people and patients with brain glioma.
[0048] In some embodiments, the average area of α particles in each platelet in the sample is used as a marker to establish a ROC curve. The area value corresponding to the Youden maximum value of the ROC curve between healthy people and patients with brain glioma is 0.75 μm 2 , which can be used as the critical value to distinguish healthy people from patients with brain glioma, that is, the average area of α particles in a single platelet in the sample is greater than 0.75 μm 2 , judging that the sample comes from a patient with brain glioma.
[0049] Furthermore, the brain glioma diagnosis system also includes a platelet classification and statistics module;
[0050] The platelet classification and statistics module, based on all acquired α particle super-resolution images, classifies platelets into four categories according to the distribution form of α particles in a single platelet, namely, "regular distribution", "aggregated distribution", α particle number "N<30" and α particle number "N≥30", and calculates the percentage of platelets with α particle number "N<30" in the sample to the total number of platelets in the four categories, and submits the results to the brain glioma diagnosis module;
[0051] The brain glioma diagnosis module uses the proportion of platelets with the number of α particles "N<30" in the sample and the average α particle area value in each platelet of the sample as a joint index, and fits the regression to calculate the predicted probability of the joint index, that is, the joint index value. The judgment is made based on the comparison between the calculated joint index value and the threshold value B. The specific judgment is as follows:
[0052] If the combined index value in the sample is less than or equal to the threshold B, the sample is diagnosed as a healthy person;
[0053] If the combined index value in the sample is greater than the threshold B, the sample is diagnosed as a brain glioma;
[0054] The threshold B is used to establish a ROC curve with the predicted probability of the combined index as the independent variable, and the predicted probability corresponding to the Youden maximum value corresponding to the ROC curve between healthy people and patients with brain glioma is preferably the threshold B.
[0055] In the present invention, the smaller the proportion of platelets with α granules "N<30", the greater the possibility of suffering from glioma. In some embodiments, the proportion of platelets with α granules "N<30" is less than 82.19%, which is judged as glioma.
[0056] In some embodiments, when the proportion of platelets with an alpha particle number "N<30" in a sample and the average alpha particle area value in each platelet of the sample are used as combined indicators for diagnostic screening, the ratio of the Youden maximum value corresponding to the ROC curve between healthy people and glioma patients is 55.38%, which can be used as the critical value for distinguishing healthy people from glioma patients, that is, the calculated combined indicator value is greater than 55.38%, and it is judged that the sample comes from a glioma patient.
[0057] The following are examples:
[0058] Example 1 Diagnosis of brain glioma based on the average α-particle area value per platelet
[0059] In this embodiment, blood samples of healthy people and patients with brain glioma are collected respectively, and super-resolution imaging of platelet α particles is obtained respectively, as follows:
[0060] 1-1 Obtaining super-resolution imaging of platelet α particles
[0061] (1) Centrifugation of blood samples: centrifuge the EDTA-K2 anticoagulant blood collection tube containing the blood sample at room temperature and 200 g for 12 min. After centrifugation, collect the supernatant (clear, transparent orange-yellow color rich in platelets) into a centrifuge tube.
[0062] (2) Platelet sample recovery: Add ACDT solution to the centrifuge tube containing the platelet supernatant according to the proportion, and place it in a 37°C incubator with 5% CO2 to recover for 2 hours.
[0063] (3) Fixation of platelet samples: Add a fixative solution of equal volume to the solution in the centrifuge tube (2) and let it stand at room temperature for 30 minutes.
[0064] (4) Washing of platelet samples: Place the centrifuge tube containing the fixed platelets in step (3) in a medical centrifuge and centrifuge at room temperature and 1500 g for 3 min. After centrifugation, remove the centrifuge tube, aspirate and discard the supernatant, add 3 ml of PBS, pipette the solution to resuspend the platelets, and repeat twice.
[0065] (5) Platelet sample plating: Resuspend the platelet solution after washing in (4), pipette an appropriate amount of the diluted platelet suspension into a confocal dish treated with poly-L-lysine, let stand for 1 hour, and observe the platelet density under a microscope. If the density is appropriate, wash once with PBS, blow dry, and store at 4°C.
[0066] (6) Punching: Pipette 0.2% Triton X-100 solution into the confocal dish containing platelets, let it stand at room temperature for 10 min, then aspirate and discard the solution.
[0067] (7) Blocking: Pipette the platelet blocking solution and add it to the co-focusing dish in (6). After standing at room temperature for 20 minutes, the solution is aspirated and discarded.
[0068] (8) Primary antibody labeling: dilute the primary antibody in platelet blocking solution, then add the platelet blocking solution diluted with the primary antibody into the confocal dish and incubate at 4°C overnight.
[0069] (9) Primary antibody rinsing: First, aspirate and discard the primary antibody dilution in the confocal dish, then add washing buffer and rinse on a shaker for 5 minutes. Finally, aspirate and discard the washing buffer. Repeat 5 times.
[0070] (10) Secondary antibody labeling: First dilute the secondary antibody in platelet blocking solution, then add the platelet blocking solution diluted with the secondary antibody into the confocal dish and incubate at room temperature for 2 h.
[0071] (11) Secondary antibody rinsing: First, aspirate and discard the secondary antibody dilution in the confocal dish, then add washing buffer and rinse on a shaker for 5 minutes. Finally, aspirate and discard the washing buffer. Repeat 5 times.
[0072] (12) Re-fixation: Pipette 4% PFA into the confocal dish in (11), let stand at room temperature for 10 min, then aspirate and discard the solution.
[0073] (13) Platelet α granule super-resolution imaging: Each sample is tested using super-resolution imaging techniques such as SIM, STED, and STORM to obtain the corresponding platelet α granule super-resolution images. 500 platelet α granule super-resolution images are obtained for each sample.
[0074] 1-2 Calculate the average area of α particles in a single platelet
[0075] All the platelet α granule super-resolution images obtained by 1-1 were binarized, and the number of pixels with a value of 255 in the sample image was calculated. Then, the number was multiplied by the square of the image scale and divided by the total number of each sample image to obtain the average value of the α granule fluorescence area in each sample platelet. The average α granule area value in each platelet of healthy people and patients with brain glioma was calculated. The results are shown in Figure 1 shown.
[0076] Depend on Figure 1 It can be seen that there is a significant difference in the average α-particle area value per platelet between healthy people and glioma patients.
[0077] The average area of α particles in each platelet was used as a diagnostic indicator to analyze the ROC curve between healthy people and patients with glioma. It was found that the average area of α particles in each platelet can better distinguish healthy people from patients with glioma. Figure 2 As shown in Figure 2, the AUC area of the ROC curve is 0.820 (95% Cl 0.653-0.987), the significant p value is <0.05, and the area corresponding to the Youden maximum value of the average α-particle area value in each platelet to distinguish between healthy people and brain gliomas is 0.7459 μm 2 , which can be used as a critical value to distinguish healthy people from patients with brain glioma. When the average area of α particles in each platelet is greater than 0.7459 μm 2 , diagnosed as brain glioma, the corresponding sensitivity was 73.30% and the specificity was 80.00%, indicating that the average α-particle area value in each platelet can better distinguish healthy people from brain glioma patients.
[0078] Example 2 Differentiation of healthy subjects and brain gliomas based on platelet type distribution of platelet α particles
[0079] In this study, blood samples from 10 healthy subjects and 15 patients with glioma were collected to detect platelet α-granule super-resolution imaging, as follows:
[0080] 1-1 Super-resolution imaging of platelet α particles
[0081] Platelet α-granule super-resolution imaging: same as in Example 1;
[0082] 1-2 Classify platelets according to the distribution of α particles and analyze the proportion of each type
[0083] All the platelet α granule super-resolution images of each patient were obtained to classify individual platelets according to the distribution form of α granules, and the proportion of each category was counted. Among them, α granules were classified into four categories: "regular distribution", "aggregated distribution", "N<30" and "N≥30" according to the platelet classification method based on platelet α granule super-resolution images and the classification method in the system of patent CN115661074A, such as Figure 3 As shown; the average values of the proportions of various types of platelets between healthy people and patients with brain glioma were calculated, and the significant differences between the two groups were compared using independent sample t-test. The results are shown in Figure 4 shown.
[0084] Figure 3 In order to classify platelets according to the distribution of α particles, they are divided into four categories: α particle number "N < 30", α particle number "N ≥ 30", "regular distribution" and "aggregated distribution".
[0085] Depend on Figure 4 It can be seen that there are significant differences in the percentages of platelets with α granules "N<30", platelets with α granules "N≥30" and platelets with "regular distribution" of α granules between healthy people and glioma patients, but there is no significant difference in the percentage of platelets with "aggregated distribution" of α granules between the two.
[0086] Receiver operating characteristic curve analysis (ROC curve analysis) is mainly used to evaluate the effect of a certain indicator on the classification or diagnosis of two types of testers (such as patients and normal people), to find the best indicator critical value, and then to determine the critical value of this evaluation indicator.
[0087] The true positive rate (TPR) is also called sensitivity: the ratio of the number of samples that are actually positive and are correctly judged as positive to the number of samples that are actually positive.
[0088] Specificity refers to the proportion of people who are actually disease-free that can be correctly identified as non-patients by the screening method, where specificity is equal to 1-false positive probability;
[0089] False positive rate (FPR): The ratio of the number of samples that are actually negative but are mistakenly judged as positive to the number of samples that are actually negative.
[0090] Furthermore, the ROC curves between healthy people and glioma patients were analyzed by taking the proportion of platelets with α-particle number "N<30" as the diagnostic index, as shown in Figure 2. Figure 5 As shown, the results show that the proportion of platelets with α granules "N<30" can better distinguish healthy people from patients with glioma, the AUC area of the ROC curve is 0.867 (95% Cl 0.719-1.00), the significant p value is <0.05, and the proportion of platelets with α granules "N<30" that distinguishes healthy people from patients with glioma corresponds to the Youden maximum of 82.19%, which can be used as the critical value for distinguishing healthy people from patients with glioma, that is, when the proportion of platelets with α granules "N<30" is less than 82.19%, it is diagnosed as glioma, with a corresponding sensitivity of 80.00% and a specificity of 86.70%; it shows that the proportion of platelets with α granules "N<30" can better distinguish healthy people from patients with glioma.
[0091] In addition, the proportion of platelets with α-particle number "N ≥ 30" was used as a diagnostic indicator to analyze the ROC curve between healthy people and glioma patients, such as Figure 6 As shown;
[0092] The ROC curves between healthy people and patients with brain glioma were analyzed by taking the proportion of platelets with "regular distribution" of α particles as the diagnostic index. Figure 7 shown.
[0093] Depend on Figure 6 and Figure 7 It can be seen that when the number of α particles "N ≥ 30" and the proportion of platelets with "regular distribution" of α particles are used to distinguish healthy people and glioma patients, the AUC areas of the corresponding ROC curves are 0.760 and 0.713, respectively. The diagnostic accuracy is low and it is difficult to distinguish healthy people and glioma patients well.
[0094] Example 3 Using the ratio of platelets with α-particle number "N<30" and the average α-particle area per platelet as a combined index for diagnosing brain glioma
[0095] In order to improve the sensitivity and specificity of clinical diagnosis, ROC curve analysis often uses multi-index combined diagnosis. This embodiment further uses combined indicators for analysis, taking the proportion of platelets with α particle number "N < 30" and the average α particle area value in each platelet as combined indicators, where the gold standard result is the dependent variable y, y = 0 for healthy people, and y = 1 for patients with glioma; and taking the proportion of platelets with α particle number "N < 30" and the average α particle area value in each platelet as independent variables x1 and x2, respectively, fitting a binary logistic regression model, and obtaining the conditional probability p = p(y = 1|x1, x2) of a positive result under the action of independent variables x1 and x2, taking the predicted probability as the test variable of the ROC curve, and taking the gold standard result as the state variable, drawing the ROC curve, and analyzing the diagnostic value of the combined indicator; the details are as follows:
[0096] In this example, blood samples from 10 healthy subjects and 15 glioma patients were collected to detect platelet α particle super-resolution imaging, as follows:
[0097] 1-1 Platelet α-granule super-resolution imaging: same as Example 1;
[0098] 1-2 Calculate the joint index value
[0099] In this embodiment, the proportion of platelets with an α-particle number "N<30" and the average α-particle area value in each platelet are used as a combined index, and the predicted probability of the combined index obtained by binary logistic regression is the combined index value.
[0100] The combined index was used as an independent diagnostic index to analyze the ROC curve between healthy people and patients with glioma, e.g. Figure 8 As shown, the results show that the combined index can well distinguish healthy people from glioma patients, the AUC area of the ROC curve is 0.900 (95% Cl 0.779-1.000), and the value corresponding to the Youden maximum value for distinguishing healthy people from glioma patients using the combined index is 55.38%, which can be used as the critical value for distinguishing healthy people from glioma patients, that is, when the combined index value is greater than 55.38%, glioma is diagnosed, and the corresponding sensitivity is 86.70%, and the specificity is 80.00%; it shows that the proportion of platelets with α-particle number "N<30" and the average α-particle area value in each platelet combined as a diagnostic indicator can well distinguish healthy people from glioma patients.
[0101] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A brain glioma diagnosis system based on platelet α-granule super-resolution imaging, It is characterized in that It includes a super-resolution image acquisition module of platelet α particles, an image analysis module, and a brain glioma diagnosis module; The platelet alpha granule super-resolution image acquisition module is used to perform binarization to obtain a plurality of platelet alpha granule super-resolution images consisting of fluorescent signal pixels and non-fluorescent signal pixels, and submit them to the image analysis module; The image analysis module is used to calculate the average area value of α particles in each platelet in the obtained α particle super-resolution image, and submit it to the brain glioma diagnosis module; The brain glioma diagnosis module is based on the obtained average α-particle area value in each platelet, and the larger the average α-particle area value in each platelet, the greater the possibility of suffering from brain glioma.
2. The brain glioma diagnosis system based on platelet α-granule super-resolution imaging as claimed in claim 1, It is characterized in that The brain glioma diagnosis module compares the obtained average area of α particles in each platelet with the threshold A and makes the following judgment: If the average α-particle area value of each platelet in the sample is greater than the threshold A, the patient is judged to be a glioma patient; If the average area of α particles per platelet in the sample is less than or equal to the threshold value A, the sample is judged to be healthy.
3. The brain glioma diagnosis system based on platelet α-granule super-resolution imaging as claimed in claim 2, It is characterized in that The threshold value A is the area value corresponding to the Youden maximum value corresponding to the ROC curve between healthy people and patients with brain glioma.
4. The brain glioma diagnosis system based on platelet α granule super-resolution imaging as claimed in claim 3, It is characterized in that The threshold value A is 0.75 μm 2 .
5. The brain glioma diagnosis system based on platelet α granule super-resolution imaging according to any one of claims 1 to 4, It is characterized in that The brain glioma diagnosis system also includes a platelet classification and statistics module; The platelet classification and statistics module, based on all acquired α-particle super-resolution images, divides platelets into four categories according to the distribution form of α-particles in a single platelet, namely "regular distribution", "aggregated distribution", α-particle number "N<30" and α-particle number "N≥30", calculates the percentage of platelets with α-particle number "N<30" in the sample to the total number of platelets in these four categories, and submits the results to the brain glioma diagnosis module.
6. The brain glioma diagnosis system based on platelet α-granule super-resolution imaging as claimed in claim 5, It is characterized in that The brain glioma diagnosis module uses the proportion of platelets with the number of α particles "N<30" in the sample and the average area of α particles in each platelet in the sample as a joint index, and calculates the joint index value by fitting regression. The judgment is made based on the comparison between the calculated joint index value and the threshold value B. The specific judgment is as follows: If the combined index value in the sample is less than or equal to the threshold B, the sample is diagnosed as a healthy person; If the combined index value in the sample is greater than the threshold B, the sample is diagnosed as a brain glioma.
7. The brain glioma diagnosis system based on platelet α-granule super-resolution imaging as claimed in claim 6, It is characterized in that The threshold value B is the ratio of the Youden maximum value corresponding to the ROC curve between healthy people and patients with brain glioma.
8. The brain glioma diagnosis system based on platelet α-granule super-resolution imaging as claimed in claim 7, It is characterized in that The threshold value B is 55.38%.
Citation Information
Patent Citations
Platelet classification method and system based on platelet alpha particle super-resolution image
CN115661074A
Cited By
Lung cancer early screening system based on platelet alpha particle super-resolution imaging
CN122199404A
