Biological sample pathology detection method and system based on characteristic signal threshold

Abstract

The invention is applicable to the technical field of pathological analysis, and provides a biological sample pathological detection method and system based on a characteristic signal threshold, and the detection system comprises: a molecular probe labeling module for performing molecular probe labeling on a standard sample and a biological sample to be detected; the three-dimensional gel electrophoresis device is used for carrying out three-dimensional gel electrophoresis separation on the marked sample; the three-dimensional gel electrophoresis signal collection module is used for extracting and collecting sample signals after electrophoretic separation; the tissue characteristic signal threshold establishing module is used for establishing a characteristic signal threshold of the standard sample; and the automatic diagnosis module is used for detecting the biological sample to be detected by utilizing the established characteristic signal threshold. The biological sample pathology detection method based on the characteristic signal threshold value shows excellent performance in clinical pathology detection, the clinical pathology work burden can be relieved, the clinical pathology work accuracy can be improved, and the biological sample pathology detection method has wide application prospects and huge market potential.

Description

Technical Field

[0001] The present invention belongs to the technical field of pathological analysis, and in particular relates to a biological sample pathology detection method and system based on characteristic signal thresholds. Background Art

[0002] In tumor surgery, rapid and accurate histopathological diagnosis is essential for clinical decision making. However, traditional intraoperative pathology consultation methods are time-consuming and labor-intensive, and are particularly dependent on the expertise of experienced pathologists. Deep learning tools have demonstrated performance levels comparable to those of pathologists in certain diagnostic tasks, but these models are usually trained for specific diagnostic tasks, which means that a single model cannot handle complex clinical pathology work scenarios, especially in intraoperative environments where multiple cancer biological samples need to be analyzed, and its application is limited. In contrast, fluorescent probes have achieved relatively ideal results in distinguishing tumors from normal tissues in vivo or in vitro. In vivo, fluorescent probes are used to guide surgical resection of solid tumors in real time, among which the near-infrared dye indocyanine green approved by the US Food and Drug Administration has been widely evaluated in fluorescence-guided surgery. Because tumor tissue has a higher uptake capacity for fluorescent probes, it is able to produce a higher signal-to-noise ratio between tumors and normal tissues.

[0003] However, unlike in vivo diagnosis, which relies on the differential uptake of fluorescent probes by tumors and normal tissues, ex vivo diagnosis of surgically removed tissues using fluorescent probes relies on the overall differences in biomarkers between malignant and normal tissues. Taking proteins as an example, studies have shown that there are significant differences in protein types and expression levels between tumor tissues and healthy tissues. Fluorescent probes can identify protein differences between tumors and normal tissues without bias. However, in actual operation, taking the detection of tissue section biological samples as an example, directly using fluorescent probes to label tissue section proteins will be interfered by a large amount of free dyes, which seriously reduces the reliability and accuracy of the probe molecules in distinguishing tumors from normal tissues through signals inside the tissue sections. Therefore, there is an urgent need for a new strategy that can spatially identify target signals and allow rapid, accurate and reliable analysis of biological samples. To this end, the present invention proposes a biological sample pathology detection method and system based on characteristic signal thresholds. Summary of the invention

[0004] The purpose of the present invention is to provide a biological sample pathology detection method and system based on characteristic signal thresholds, aiming to solve the problems raised in the above background technology.

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

[0006] A method for detecting pathology of biological samples based on characteristic signal thresholds comprises the following steps:

[0007] Step 1: labeling the standard sample with molecular probe;

[0008] Step 2: Perform three-dimensional gel electrophoresis separation on the labeled standard samples, use a camera or scanner with the corresponding wavelength to image the gel after electrophoresis, collect and analyze the signals, and form a signal matrix with clear position information;

[0009] Step 3: Based on the signal distribution curve of at least one cohort of standard samples, determine the characteristic signal threshold of the standard sample by analyzing the signal distribution of the standard sample, and use the characteristic signal threshold of the standard sample as the standard detection threshold of such samples;

[0010] Step 4: labeling the biological sample to be tested with a molecular probe;

[0011] Step 5: Perform three-dimensional gel electrophoresis separation on the labeled biological sample to be tested, use a camera or scanner with a corresponding wavelength to image the gel after electrophoresis, collect and analyze the signal, and form a signal matrix with clear position information;

[0012] Step 6: Perform distribution analysis on the signal matrix of the biological sample to be tested;

[0013] Step 7: Using the signal peak value of the distribution analysis as the detection background value, perform background correction on the signal matrix;

[0014] Step 8: Automatically analyzing the signal of the biological sample to be tested based on the characteristic signal threshold of the standard sample to diagnose the tissue type contained in the biological sample to be tested;

[0015] Step 9: Set the threshold of the signal matrix to different characteristic signal thresholds, and realize spatial analysis of the characteristics by displaying the spatial distribution position of a specific tissue type in the biological sample to be tested.

[0016] Further, the molecular probe includes a class of molecules covalently bound to the biological sample to be tested, including IR-780, IR-3B3, IR-4B3, IR-820, IR-808 and IR-806 as shown in Formula 1;

[0017] Formula 1:

[0018]

[0019] Furthermore, in the step of collecting and analyzing signals, the signals come from gel layer 1 to gel layer 8. The range of the gel layer to be detected can also be expanded according to the specific biological sample.

[0020] Furthermore, in step 7, the background correction method is: setting the minimum value of the signal matrix as the background value to eliminate the interference of non-target signals.

[0021] A biological sample pathology detection system based on characteristic signal threshold, comprising:

[0022] A molecular probe labeling module is used to label standard samples and biological samples to be tested with molecular probes;

[0023] A three-dimensional gel electrophoresis device is used to perform three-dimensional gel electrophoresis separation on the labeled standard sample and the biological sample to be tested;

[0024] A three-dimensional gel electrophoresis signal collection module is connected to the three-dimensional gel electrophoresis device and is used to image the gel layers 1 to 8 after electrophoresis using a camera or scanner of a corresponding wavelength, collect and analyze signals, and form a signal matrix with clear position information;

[0025] A tissue characteristic signal threshold establishment module, which is used to determine the characteristic signal threshold of the standard sample by analyzing the signal distribution of the standard sample based on the signal distribution curve of at least one cohort of standard samples, and use the characteristic signal threshold of the standard sample as the standard detection threshold of such samples;

[0026] The automatic diagnosis module is used to perform distribution analysis on the signal matrix of the biological sample to be tested; use the signal peak value of the distribution analysis as the detection background value to perform background correction on the signal matrix; automatically analyze the signal of the biological sample to be tested based on the characteristic signal threshold of the standard sample to diagnose the tissue type contained in the biological sample to be tested; set the threshold of the signal matrix to different characteristic signal thresholds, and realize spatial analysis of the characteristics by displaying the spatial distribution position of a specific tissue type in the biological sample to be tested.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] 1. The biological sample pathology detection method based on characteristic signal threshold proposed in the present invention shows excellent performance in clinical pathology detection, which can not only reduce the workload of clinical pathology, but also improve the accuracy of clinical pathology work. It has broad application prospects and huge market potential.

[0029] 2. The present invention is compatible with a variety of biological sample types (such as tissue sections, cells, body fluids, and exosomes, etc.), and provides automatic diagnostic analysis methods based on characteristic signal thresholds for different types of biological samples. It can also test protein molecules, nucleic acid molecules, etc. derived from biological samples, showing a very high prospect for clinical transformation. At the same time, different probe molecules can be selected for different analytical samples, and cameras or scanners with corresponding wavelengths can be used for signal collection. This flexible and diverse signal collection method further enhances the applicability and practicality of the method.

[0030] 3. The present invention allows the analysis and automatic detection of biological samples to be completed within 5 to 45 minutes. Combined with the high-throughput analysis tool of three-dimensional gel electrophoresis, the method has extremely high efficiency and greatly improves the speed of clinical pathology detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 The invention is to apply the method of the present invention to the workflow of tissue section diagnosis during tumor surgery.

[0032] Figure 2 The figure is an effect diagram of setting a standard characteristic signal threshold and applying the method of the present invention to the intraoperative diagnosis of liver cancer tissue sections; wherein A is the process of setting the standard characteristic signal threshold, and B is the automatic diagnosis of liver nodule areas and liver cancer areas in the same tissue section sample based on the characteristic signal threshold.

[0033] Figure 3 Schematic diagram of the principle of eliminating interference from free dyes in the diagnosis of intraoperative tissue sections.

[0034] Figure 4 The figures are the effect of eliminating the interference of free dyes in the diagnosis of intraoperative tissue sections; A is the test result of liver cancer tissue, B is the test result of bile duct cancer tissue, and C is the test result of pancreatic cancer tissue.

[0035] Figure 5 The figures are effect diagrams of applying the method of the present invention to intraoperative diagnosis of different tissue sections; A is the detection result of bile duct cancer tissue section, and B is the detection result of pancreatic cancer tissue section.

[0036] Figure 6 These are the receiver operating characteristic curves of different cohorts; A is the receiver operating characteristic curve of the liver cancer cohort, B is the receiver operating characteristic curve of the bile duct cancer cohort, and C is the receiver operating characteristic curve of the pancreatic cancer cohort.

[0037] Figure 7The comparison is between the detection results of different surgical tissues by the method of the present invention and the detection results based on the clinical hematoxylin-eosin tissue section staining method; wherein A is the detection result of liver cancer tissue, B is the detection result of bile duct cancer tissue, and C is the detection result of pancreatic cancer tissue. DETAILED DESCRIPTION

[0038] In order to have a clearer understanding of the technical features, purposes and beneficial effects of the present invention, the technical solution of the present invention is now described in detail below, but it should not be construed as limiting the applicable scope of the present invention.

[0039] The specific implementation of the present invention is described in detail below in conjunction with specific embodiments.

[0040] An embodiment of the present invention provides a method for detecting pathology of a biological sample based on a characteristic signal threshold, the method comprising the following steps:

[0041] Step 1, labeling the standard sample with a molecular probe;

[0042] Step 2, performing three-dimensional gel electrophoresis separation on the standard sample labeled in step 1, and collecting the analysis signal in the gel to obtain a signal matrix;

[0043] Step 3, performing distribution analysis on the signal points in the signal matrix obtained in step 2, and establishing a characteristic signal threshold of the standard sample;

[0044] Step 4, labeling the biological sample to be tested with a molecular probe;

[0045] Step 5, performing three-dimensional gel electrophoresis separation on the biological sample labeled in step 4, and collecting the analysis signal in the gel to obtain a signal matrix;

[0046] Step 6, performing distribution analysis on the signal points in the signal matrix obtained in step 5;

[0047] Step 7: Using the signal peak value of the distribution analysis as the detection background value, the signal matrix is ​​subjected to background correction;

[0048] Step 8: Automatically analyze the signal according to the characteristic signal threshold of the standard sample to realize automatic diagnosis of unknown biological samples;

[0049] Step 9: Set the threshold of the signal matrix to different characteristic signal thresholds to analyze the spatial distribution of the feature.

[0050] Example 1: This example demonstrates the application of the method of the present invention in intraoperative tissue diagnosis of liver cancer tumor resection surgery. The specific method is to quickly detect the slices of surgically removed tissue based on the method to determine whether there is a malignant lesion and distinguish between malignant areas and benign lesion areas. The steps of intraoperative application of the method are as follows: Figure 1 As shown, the details are as follows:

[0051] 1) After surgical resection of the tissue, quickly freeze the tissue at -80°C (about 5 minutes), then freeze-section the frozen tissue (about 5 minutes), with a slice thickness of 60μm (ranging from 3 to 180μm). Transfer the tissue slices to the surface of a 20×20 microwell array plate, and lyse and label the tissue slices simultaneously. Lysis is performed at room temperature to 60°C using a lysis buffer, which includes but is not limited to RIPA and other mixed solutions that can be used for tissue lysis. 5mL of lysis buffer is used for each slice. Labeling uses probe molecules that can covalently bind to protein molecules in tissue slices, such as IR-780. The throughput of the microwell array plate is 4cm per 4cm 2 There are 100 to 250,000 micropores inside.

[0052] 2) After the lysis and labeling are completed, a three-dimensional gel electrophoresis device is used to extract and separate the protein molecules in the tissue slice in situ. The electrophoresis process is constant voltage, the electric field direction is in the z direction, and it takes about 15 minutes. The gel layer 1 to gel layer 8 after electrophoresis is imaged by an imaging device (including but not limited to an 800nm ​​channel scanner or camera), and the protein molecular weight range corresponding to the collected signal is 10 to 180kDa, which takes about 10 minutes. The entire gel medium containing tissue protein information is frozen and cut into gel layers that are perpendicular to the electric field direction and have uniform thickness, and the signal is read. The signal has both protein signal intensity value, molecular weight range (ie, gel layer), and xy direction position information (one-to-one correspondence with the microwell coordinates of the microwell array sample plate). The signal matrices of gel layer 1 to gel layer 8 read are added to obtain the signal matrix of the tissue to be tested.

[0053] This example establishes surgical tissue sections from 9 liver cancer patients with clear histological annotations as standard samples, and establishes a signal distribution curve based on the analysis results. According to the signal distribution curve of the standard sample, the standard characteristic signal thresholds of liver cancer tissue, liver nodule tissue, normal liver tissue and necrotic tissue are set. Under the experimental conditions of this example, these thresholds are: liver cancer tissue>76000, liver nodule tissue 61200~76000, normal liver tissue 12000~61200, necrotic tissue 0~12000 ( Figure 2 By using the established characteristic signal threshold, this embodiment automatically diagnoses the liver nodule area and liver cancer area in the liver cancer tissue sample to be tested ( Figure 2 (middle B).

[0054] Example 2: This example demonstrates the effectiveness of the three-dimensional gel electrophoresis signal collection module in eliminating interference from free dyes. The three-dimensional gel electrophoresis method can remove free dyes that are not effectively covalently bound to protein molecules in tissue sections to prevent their signals from affecting detection accuracy. The schematic diagram of removing free dyes is shown in FIG. Figure 3 As shown. Compared with protein molecules, free dyes have a smaller molecular weight (<1000Da) and move faster during electrophoresis. Therefore, the molecular weight range (10-180kDa) covered by the collected gel layers 1 to 8 can exclude free dye molecules. The results show that in liver cancer ( Figure 4 Middle A), bile duct cancer ( Figure 4 Middle B) and pancreatic cancer ( Figure 4 In the surgical tissue test in C), the signal of the tumor area (T) in the tissue section after staining analysis is not always higher than that of the adjacent cancer area (P) and normal tissue area (N), which is caused by nonspecific dye adsorption. After three-dimensional gel electrophoresis analysis, the free dye is effectively removed, and the tumor can be identified through the specific signal of covalent binding between the dye and the protein molecules in the section. In other words, the signal-to-noise ratio between tumors and healthy tissues is effectively enhanced by eliminating the interference of free dyes through the three-dimensional gel electrophoresis signal collection module.

[0055] Example 3: This example demonstrates the application of the method of the present invention in intraoperative tissue diagnosis of bile duct and pancreatic tumor resection surgery. The specific method is to quickly detect the slices of surgically removed tissue based on the method to determine whether there is a malignant lesion and distinguish between malignant areas and benign lesion areas. The specific steps are the same as those in Example 1.

[0056] Based on the method of the present invention, the resected tissues of two cases of bile duct tumor resection (numbered G0065 and G0193) were analyzed and processed, and the diagnosis results showed that ( Figure 5 In the figure (A), the G0065 sample contained malignant lesions of bile duct cancer and was diagnosed as a malignant tumor. The spatial distribution of the lesions in the tissue section is shown in the figure; while the G0193 sample did not contain malignant lesions and was diagnosed as a benign cyst.

[0057] Based on the method of the present invention, the resected tissues of two pancreatic tumor resection surgeries (numbered G1631 and G0534) were analyzed and processed, and the diagnosis results showed that ( Figure 5 In the figure (B), the G1631 sample contained pancreatic cancer malignant lesions and was diagnosed as a malignant tumor. The spatial distribution of the lesions in the tissue section is shown in the figure; while the G0534 sample did not contain malignant lesions and was diagnosed as a benign cyst.

[0058] It is worth noting that the benign tumor lesions involved in this embodiment are difficult to diagnose in existing clinical pathological examinations, which is time-consuming and labor-intensive. However, the method of the present invention can achieve automatic, rapid and accurate diagnosis of such benign tumor lesions.

[0059] Example 4: This example demonstrates the detection performance of the method of the present invention in a liver cancer cohort (9 patients), a bile duct cancer cohort (6 patients), and a pancreatic cancer cohort (8 patients). These cohorts were tested and the test results were plotted into a receiver operating characteristic curve (ROC curve) to evaluate the detection performance of the method of the present invention.

[0060] The results showed that the area under the curve (AUC) of the receiver operating characteristic curve for the liver cancer cohort was 0.92, while the AUC of the traditional staining method was only 0.55 ( Figure 6 A), indicating that the method of the present invention has significant advantages in liver cancer detection. Similarly, the AUC of the cholangiocarcinoma cohort is 0.98 ( Figure 6 Middle B), the AUC for the pancreatic cancer cohort was 0.90 ( Figure 6 The AUC values ​​of the three-dimensional gel electrophoresis method were significantly higher than those of the conventional staining method (0.69 and 0.68, respectively). This result indicates that the strategy of removing free dyes by three-dimensional gel electrophoresis effectively improves the accuracy of detection, further verifying that the method of the present invention has excellent detection performance.

[0061] Example 5: This example compares in detail the detection results of the method of the present invention on liver cancer surgical tissue, bile duct cancer surgical tissue and pancreatic cancer surgical tissue with the detection results of the clinical hematoxylin-eosin tissue section staining method. We used the same method steps as Example 1 to detect the malignant area in the surgical tissue, and quantified the contour similarity between the area reported by the method of the present invention and the area reported by the clinical method. The contour comparison algorithm script was written in MATLAB.

[0062] The results showed that in the comparison of liver cancer surgical tissues, for samples G0375 and G1952, the contour similarities between the reported areas detected by the method of the present invention and the reported areas by the clinical method were 94.44% and 94.01%, respectively ( Figure 7 In the comparison of bile duct cancer surgical tissues, the contour similarities of samples G0235 and G0329 were 95.29% and 94.03%, respectively ( Figure 7 In the comparison of pancreatic cancer surgical tissues, the contour similarities of samples G1172 and G0241 were 91.32% and 80.71%, respectively ( Figure 7The above results show that the average quantitative similarity between the detection report area of ​​various clinical tumor tissues and the clinical method report area of ​​the method of the present invention is greater than 90%, which fully demonstrates the excellent performance of the method of the present invention in clinical pathology detection. This method can not only reduce the workload of clinical pathology, but also improve the accuracy of clinical pathology work, and has broad application prospects and huge market potential.

[0063] The above are only preferred embodiments of the present invention. It should be pointed out that, for those skilled in the art, several modifications and improvements can be made without departing from the concept of the present invention. These should also be regarded as the protection scope of the present invention. These will not affect the effect of the implementation of the present invention and the practicality of the patent.

Claims

1. A biological sample pathology detection method based on characteristic signal threshold, characterized in that: The following steps are involved: Step 1: labeling the standard sample with molecular probe; Step 2: Perform three-dimensional gel electrophoresis separation on the labeled standard samples, use a camera or scanner with the corresponding wavelength to image the gel after electrophoresis, collect and analyze the signals, and form a signal matrix with clear position information; Step 3: Based on the signal distribution curve of at least one cohort of standard samples, determine the characteristic signal threshold of the standard sample by analyzing the signal distribution of the standard sample, and use the characteristic signal threshold of the standard sample as the standard detection threshold of such samples; Step 4: labeling the biological sample to be tested with a molecular probe; Step 5: Perform three-dimensional gel electrophoresis separation on the labeled biological sample to be tested, use a camera or scanner with a corresponding wavelength to image the gel after electrophoresis, collect and analyze the signal, and form a signal matrix with clear position information; Step 6: Perform distribution analysis on the signal matrix of the biological sample to be tested; Step 7: Using the signal peak value of the distribution analysis as the detection background value, perform background correction on the signal matrix; Step 8: Automatically analyzing the signal of the biological sample to be tested based on the characteristic signal threshold of the standard sample to diagnose the tissue type contained in the biological sample to be tested; Step 9: Set the threshold of the signal matrix to different characteristic signal thresholds, and realize spatial analysis of the characteristics by displaying the spatial distribution position of a specific tissue type in the biological sample to be tested.

2. The biological sample pathology detection method based on characteristic signal threshold according to claim 1, characterized in that: The molecular probes include a class of molecules covalently bound to the biological sample to be tested, including IR-780, IR-3B3, IR-4B3, IR-820, IR-808 and IR-806 as shown in Formula 1; Formula 1:

3. The biological sample pathology detection method based on characteristic signal threshold according to claim 1, characterized in that: In the step of collecting and analyzing signals, the signals come from gel layer 1 to gel layer 8.

4. The biological sample pathology detection method based on characteristic signal threshold according to claim 1, characterized in that: In step 7, the background correction method is: setting the minimum value of the signal matrix as the background value to eliminate the interference of non-target signals.

5. A biological sample pathology detection system based on characteristic signal threshold, characterized in that: include: A molecular probe labeling module is used to label standard samples and biological samples to be tested with molecular probes; A three-dimensional gel electrophoresis device is used to perform three-dimensional gel electrophoresis separation on the labeled standard sample and the biological sample to be tested; A three-dimensional gel electrophoresis signal collection module, located downstream of the three-dimensional gel electrophoresis device, is used to image gel layers 1 to 8 after electrophoresis using a camera or scanner of a corresponding wavelength, collect and analyze signals, and form a signal matrix with clear position information; A tissue characteristic signal threshold establishment module, which is used to determine the characteristic signal threshold of the standard sample by analyzing the signal distribution of the standard sample based on the signal distribution curve of at least one cohort of standard samples, and use the characteristic signal threshold of the standard sample as the standard detection threshold of such samples; The automatic diagnosis module is used to perform distribution analysis on the signal matrix of the biological sample to be tested; use the signal peak value of the distribution analysis as the detection background value to perform background correction on the signal matrix; automatically analyze the signal of the biological sample to be tested based on the characteristic signal threshold of the standard sample to diagnose the tissue type contained in the biological sample to be tested; set the threshold of the signal matrix to different characteristic signal thresholds, and realize spatial analysis of the characteristics by displaying the spatial distribution position of a specific tissue type in the biological sample to be tested.

Images