Construction method and application of tumor cell drug resistance level detection model based on surface enhanced Raman scattering
The tumor cell drug resistance level detection model trained by surface-enhanced Raman scattering technology and machine learning algorithm solves the problems of low efficiency and poor accuracy of existing detection methods, and realizes rapid and accurate drug resistance level detection, which is applicable to a variety of tumor types and drug resistance mechanisms.
Patent Information
- Application Number
- CN202511076670.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-01
- Publication Date
- 2025-11-21
AI Technical Summary
Existing methods for detecting drug resistance levels in tumor cells are inefficient and have poor accuracy, failing to accurately reflect changes in drug resistance levels and being susceptible to heterogeneity in tumor sample collection sites.
A tumor cell drug resistance level detection model was trained by combining surface-enhanced Raman scattering (SERS) with recursive feature elimination, radial basis function (RBF) kernel function, and support vector machine. By enhancing the Raman signal with silver nanoparticles and screening key spectral features, efficient and accurate drug resistance level detection was achieved.
It achieves rapid and accurate detection of tumor cell drug resistance levels, and can predict drug resistance levels within 10 minutes. It avoids the complex steps of traditional methods, improves detection consistency and applicability, and is applicable to a variety of tumor types and drug resistance mechanisms.
Smart Images
Figure CN120992578A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of tumor cell drug resistance detection technology, and in particular to a method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering (SERS) and its application. Background Technology
[0002] Currently, chemotherapy is the primary treatment for cancer, but its efficacy is often affected by tumor drug resistance, which can further develop into multidrug resistance, becoming the main cause of death in over 90% of patients. To effectively treat drug-resistant tumors, early determination of resistance levels, and subsequent targeted selection of chemotherapy drugs and dosages, has become a crucial issue that urgently needs to be addressed in current clinical cancer treatment.
[0003] Since drug-resistant tumor cells are not visually distinct from conventional tumor cells, clinical practice primarily involves isolating and culturing tumor cells from patients, followed by high-throughput drug sensitivity testing to determine the presence of drug resistance, screen drugs, and determine drug concentrations. While this method is relatively accurate, it is complex, time-consuming, and inefficient, significantly reducing clinical treatment effectiveness. Cancer biomarkers, such as the expression level of drug resistance proteins, are another important traditional method for detecting drug-resistant tumors and have been widely used in clinical practice. They can accurately identify the occurrence of drug-resistant tumors, but they cannot reflect changes in the drug resistance level of tumor cells and may lead to inaccurate detection due to heterogeneity in tumor sample locations.
[0004] Therefore, there is an urgent need for an accurate and efficient method for detecting drug resistance levels in tumor cells. Summary of the Invention
[0005] This invention provides a method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering and its application, in order to solve the shortcomings of existing tumor cell drug resistance level detection methods, such as low efficiency and poor accuracy.
[0006] On one hand, the present invention provides a method for obtaining training data for a tumor cell drug resistance level detection model, comprising: Prepare tumor cell test samples with different drug resistance indices, wherein the different drug resistance indices include any one of the following drug resistance indices or any combination thereof: 1x, 3x, 5x, 10x, 15x; A silver nanoparticle solution was prepared, wherein the diameter of the silver nanoparticles in the solution was between 70 and 80 nm, and the concentration of the silver nanoparticle solution was between 9 and 10 mM. Using tumor cell samples with different drug resistance indices, cell homogenates with different drug resistance indices were prepared. Silver nanoparticle solutions were mixed with cell homogenates of different drug resistance indices according to a preset ratio to obtain mixed solutions with different drug resistance indices, wherein the preset ratio was between 1:2 and 1:4. Surface-enhanced Raman scattering data of tumor cell samples with different drug resistance indices were obtained by using a Raman spectrometer to detect mixed solutions with different drug resistance indices.
[0007] According to a method for obtaining training data for a tumor cell drug resistance level detection model provided by the present invention, the step of mixing silver nanoparticle solution with cell homogenates of different drug resistance indices in a preset ratio to obtain mixed solutions with different drug resistance indices includes: For a mixed solution with a certain drug resistance index, the mixed solution with the drug resistance index is dropped onto the surface of a glass slide uniformly wrapped with aluminum foil, and the droplet is spread into a circular area with a diameter of about 0.5 cm using a pipette.
[0008] According to the method for obtaining training data for a tumor cell drug resistance level detection model provided by the present invention, the optimal diameter of silver nanoparticles in the silver nanoparticle solution is 73 nm, the optimal concentration of the silver nanoparticle solution is 9.13 nM, and the optimal concentration of cell homogenates with different drug resistance indices is 2.5 × 10⁻⁶. 6 -10 7 per mL.
[0009] On the other hand, the present invention provides a method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering, comprising: According to the method for obtaining training data of the tumor cell drug resistance level detection model described above, surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices are obtained, wherein the different drug resistance indices include any one of the following drug resistance indices or any combination thereof: 1x, 3x, 5x, 10x, 15x. Based on surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices, and combined with recursive feature elimination (RFE), radial basis function (RBF), and support vector machine (SVM), the model is trained to learn the features of surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices, thus obtaining a tumor cell drug resistance level detection model.
[0010] According to the present invention, a method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering (SERS) is provided. The method involves training a model based on SERS data of target tumor cells with different drug resistance indices, combined with recursive feature elimination (RFE), radial basis function (RBF), and support vector machine (SVM), to learn the features of SERS data of target tumor cells with different drug resistance indices, thereby obtaining a tumor cell drug resistance level detection model. The method includes: Based on the surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices, a recursive feature elimination method was used to screen out several characteristic peaks related to the drug resistance index, and the characteristic peak data related to the drug resistance index were obtained from the surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices. By using radial basis function (RBF) to train a support vector machine (SVM), the features of characteristic peak data related to drug resistance index in the surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices are learned, thus obtaining a tumor cell drug resistance level detection model.
[0011] According to the present invention, a method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering (SERS) is provided. The method involves training a model based on SERS data of target tumor cells with different drug resistance indices, combined with recursive feature elimination (RFE), radial basis function (RBF), and support vector machine (SVM), to learn the features of SERS data of target tumor cells with different drug resistance indices, thereby obtaining a tumor cell drug resistance level detection model. The method includes: Bar charts are used to show the weights of several characteristic peaks related to the drug resistance index in the tumor cell drug resistance level detection model, so as to reflect the importance of several characteristic peaks related to the drug resistance index.
[0012] According to the present invention, a method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering (SERS) is provided. The method involves training a model based on SERS data of target tumor cells with different drug resistance indices, combined with recursive feature elimination (RFE), radial basis function (RBF), and support vector machine (SVM), to learn the features of SERS data of target tumor cells with different drug resistance indices, thereby obtaining a tumor cell drug resistance level detection model. The method includes: Principal component analysis (PCA) was used to reduce the dimensionality of surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices and several feature peaks related to drug resistance indices, in order to visualize the effect of cell classification while demonstrating the impact of feature selection on sample separability.
[0013] In another aspect, the present invention provides a method for detecting the drug resistance level of tumor cells, comprising: Receive surface-enhanced Raman scattering data from tumor cells to be tested; Based on the surface-enhanced Raman scattering data of the tumor cells to be tested, the tumor cell drug resistance level detection model obtained by constructing the tumor cell drug resistance level detection model based on surface-enhanced Raman scattering as described in any of the above-mentioned methods is used to obtain the tumor cell drug resistance level detection result of the tumor cells to be tested.
[0014] Furthermore, the present invention also provides a tumor cell drug resistance level detection system, comprising: Raman spectrometer is used to detect surface-enhanced Raman scattering data of tumor cells under test. The data receiving module is used to: receive surface-enhanced Raman scattering data of tumor cells under test from a Raman spectrometer; The detection module is used to: obtain the drug resistance level detection result of the tumor cells to be tested based on the surface-enhanced Raman scattering data of the tumor cells to be tested, and obtain the tumor cell drug resistance level detection model obtained by the construction method of the tumor cell drug resistance level detection model based on surface-enhanced Raman scattering described above; The data output module is used to output the drug resistance level detection results of the tumor cells to be tested to at least one terminal.
[0015] It should be noted that a terminal refers to an input / output device connected to a computer system. Depending on the function, terminals can be divided into various types: smart terminals or intelligent terminals, dumb terminals, interactive terminals or online terminals. Specifically, a terminal can be various mobile communication devices, such as mobile phones and tablets. This article aims to provide users with the function of inputting data and outputting data.
[0016] The present invention also provides an electronic device, including a processor and a memory storing a computer program, wherein the processor executes the computer program to implement the method for acquiring training data of the tumor cell drug resistance level detection model described above, the method for constructing the tumor cell drug resistance level detection model based on surface-enhanced Raman scattering, and the method for detecting tumor cell drug resistance level.
[0017] The present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the method for acquiring training data of the tumor cell drug resistance level detection model, the method for constructing the tumor cell drug resistance level detection model based on surface-enhanced Raman scattering, and the method for detecting tumor cell drug resistance level as described above.
[0018] The present invention also provides a computer program product, the computer program product comprising a computer program that can be stored on a non-transitory computer-readable storage medium, wherein when the computer program is executed by a processor, the computer is capable of executing any of the above-described methods for obtaining training data of a tumor cell drug resistance level detection model, constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering, and detecting tumor cell drug resistance levels.
[0019] This invention provides a method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering and its application. Based on surface-enhanced Raman scattering technology combined with machine learning algorithms, a high-precision tumor cell drug resistance level detection model is trained, which can achieve efficient and accurate detection of tumor cell drug resistance levels.
[0020] The present invention has at least the following beneficial effects: 1) Raman spectral data of tumor cell homogenates can be directly obtained using SERS technology. Combined with a pre-trained machine learning model (RFE-RBF-SVM), drug resistance level prediction can be completed within 10 minutes, which is significantly better than traditional drug sensitivity tests (which usually take several days to several weeks). This effectively avoids complex steps such as cell culture and high-throughput screening, achieving "one-stop" rapid detection from sample to result, and greatly improving the efficiency of clinical treatment decision-making.
[0021] 2) Utilizing silver nanoparticles (70-80 nm, 9-10 mM) as the SERS substrate, the Raman signal is amplified millions of times through electromagnetic enhancement, enabling sensitive capture of minute molecular-level differences (such as resistance proteins and metabolites) in drug-resistant cells. Combined with recursive feature elimination (RFE) to screen key spectral features, and learning the nonlinear mapping relationship between different resistance indices (1 to 15 times) through the RBF-SVM model, quantitative assessment of resistance levels can be achieved, overcoming the limitation of traditional biomarker detection (such as protein expression) which can only qualitatively determine resistance.
[0022] 3) Using cell homogenate as the test sample effectively avoids errors caused by heterogeneity in tumor tissue sampling, improving test consistency. The machine learning model processes high-dimensional spectral data through the RBF kernel function, exhibiting robustness against noise and baseline drift, ensuring the comparability of experimental data from different batches.
[0023] 4) It is compatible with various tumor types (such as breast cancer, lung cancer, etc.) and different drug resistance mechanisms (such as P-gp overexpression, metabolic pathway alterations, etc.), and the application scenarios can be expanded simply by adjusting the training data. The detection process is highly standardized and easy to integrate with existing large-scale instruments (Raman spectrometers) or portable devices (portable Raman spectrometers), making it feasible for large-scale promotion.
[0024] In summary, this invention deeply integrates SERS technology with intelligent algorithms, which can solve the core pain points of traditional methods such as long cycle, complex operation and insufficient quantitative ability. It provides an efficient tool for the early detection of tumor drug resistance and personalized medicine, and has important clinical value and scientific research significance. Attached Figure Description
[0025] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0026] Figure 1This invention provides a method for obtaining training data for a tumor cell drug resistance level detection model, a method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering, and a schematic diagram of the overall process of the tumor cell drug resistance level detection method.
[0027] Figure 2 The second part of the overall flowchart of the tumor cell drug resistance level detection method is shown below, which describes the method for obtaining training data for the tumor cell drug resistance level detection model provided by the present invention, the method for constructing the tumor cell drug resistance level detection model based on surface-enhanced Raman scattering.
[0028] Figure 3 Raman fingerprints of silver nanoparticles of different sizes mixed with cell homogenates are shown.
[0029] Figure 4 Raman fingerprints of different concentrations of silver nanoparticles mixed with cell homogenate are shown.
[0030] Figure 5 Normalized Raman spectra of 73 nm silver nanoparticles mixed with cell homogenates at different concentrations are shown, specifically (a) 0.935 nM, (b) 1.87 nM, (c) 3.74 nM, (d) 7.48 nM, (e) 9.13 nM and (f) 14.96 nM.
[0031] Figure 6 Raman fingerprints of different cell densities are shown, specifically (a) for primitive lung cancer cells (A549). PC (b) are doxorubicin-resistant lung cancer cells (A549). DOX The normalized Raman spectra of these cells after homogenization and mixing with 9.13 nM 73 nm silver nanoparticles.
[0032] Figure 7 An example of the characteristic peak of the drug resistance index of glioma (HS683) cells is shown, which exhibits a regular change with increasing drug resistance.
[0033] Figure 8 The main characteristic peaks of drug-resistant cells obtained based on machine learning algorithms are shown.
[0034] Figure 9 The results of principal component analysis (PCA) dimensionality reduction visualization of Raman spectral data of primitive glioma (HS683) cells and Hs683 cells with different drug resistance indices are shown after feature selection.
[0035] Figure 10 This is a schematic diagram of the structure of a system for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering, as provided by the present invention.
[0036] Figure 11 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation
[0037] 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 with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, embodiments of this invention, and should not be construed as limiting the invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention. In the description of this invention, it should be understood that the terminology used is for descriptive purposes only and should not be construed as indicating or implying relative importance.
[0038] Figure 1 The present invention provides a method for acquiring training data for a tumor cell drug resistance level detection model, a method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering, and an overall flowchart of the tumor cell drug resistance level detection method. Figure 2 The overall flowchart is provided for further detail. The execution entity of the various methods provided by this invention can be any applicable terminal-side device or network-side device, such as a device for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering.
[0039] On one hand, the method for obtaining training data for a tumor cell drug resistance level detection model provided by the present invention may include: S110. Prepare tumor cell samples with different drug resistance indices. Specifically, obtain original tumor cells and cell lines with different drug resistance states as cell samples, and sonicate them into cell homogenates. The tumor cells involved can cover major cancer types, including glioma, liver cancer, lung cancer, and breast cancer. The drugs involved may include temozolomide (glioma), doxorubicin, vincristine, cisplatin (the latter being broad-spectrum chemotherapy drugs), etc. For tumor cells, their drug resistance status is characterized by the drug resistance index, which is calculated using the IC50 formula. 50 (Drug-resistant cells) / IC 50 (Primitive cells), IC 50 Gradient toxicity assays derived from tumor cells represent the median lethal concentration (LD50) obtainable from the fitted curve. Different resistance indices include any one of the following resistance indices or any combination thereof: 1x, 3x, 5x, 10x, and 15x.
[0040] S120. Silver nanoparticles were synthesized using a seed-mediated citrate and ascorbic acid reduction method. Specifically, 9 mg of silver nitrate was added to 50 mL of a 40% (w / v) glycerol-water solution, and the solution was heated to 95°C with stirring. Then, 1 mL of a 3% (w / v) trisodium citrate solution was rapidly added, and stirring continued for 1 hour until heating was stopped. The solution was then cooled to room temperature to complete the synthesis of the silver seed solution (23 nm). 5 mL of the silver seed solution was mixed with 138 mL of water, 23 mL of glycerol, and 0.58 g of polyvinylpyrrolidone. After the reaction started, 1.15 mL of diamine silver solution (composed of 20 mg silver nitrate, 1 mL of water, and 220 μL of ammonia) and 92 mL of L-ascorbic acid (36.8 mg dissolved in 92 mL of water) were added after 20 seconds. After stirring for 1 hour, 6 g of polyvinylpyrrolidone was added to stabilize the synthesized silver nanoparticles. Ultimately, the silver nanoparticles in the solution had a diameter between 70 and 80 nm and a concentration between 9 and 10 mM.
[0041] S130. Using tumor cell samples with different drug resistance indices, prepare cell homogenates with different drug resistance indices. Specifically, when testing the cells, suspend them in PBS at a density of 5 × 10⁻⁶. 6 -10 7 Cells per mL were processed using a cell disruptor (65 W ultrasonic power). During the process, the cells were ultrasonicated for 3 seconds at a time with a 5-second interval, for a total of 3 cycles, to prepare cell homogenates with different drug resistance indices.
[0042] S140. The silver nanoparticle solution is mixed with cell homogenates of different drug resistance indices according to a preset ratio to obtain mixed solutions with different drug resistance indices. The preset ratio is between 1:2 and 1:4. For each mixed solution with a drug resistance index, the mixed solution is dropped onto the surface of a glass slide uniformly wrapped with aluminum foil, and the droplet is spread into a circular area with a diameter of about 0.5 cm using a pipette.
[0043] S150. Using a Raman spectrometer, the mixed solutions with different drug resistance indices were detected to obtain surface-enhanced Raman scattering data of tumor cell samples with different drug resistance indices. The parameters of the Raman spectrometer were set as follows: 532 nm laser wavelength, 0.5 mW laser power, 20× objective lens, 10 s integration time, 1 accumulation, and 15 μm laser spot.
[0044] The following specific embodiment 1 describes a method for obtaining training data for a tumor cell drug resistance level detection model provided by the present invention.
[0045] Example 1 The target tumor cells in this embodiment are lung cancer cells (A549), and the optimization mainly focuses on the steps in the method for acquiring training data for the doxorubicin (DOX) resistance level detection model of tumor cells. This includes the following steps: 1) Synthesize silver nanoparticles of different sizes, including 23 nm, 30 nm, 40 nm, 55 nm, 73 nm and 88 nm, using the same synthesis method as above.
[0046] 2) Silver nanoparticle size optimization: At the same concentration of 9.13 nM, silver nanoparticles of different sizes were mixed with A549 cells, and SERS spectroscopy was performed under the same conditions as above. Figure 3 As the size of silver nanoparticles increases, the overall signal-to-noise ratio of the cell Raman spectrum gradually increases, the signal intensity gradually improves, and the spectral peaks become more abundant. Furthermore, at a 660 cm⁻¹... -1 Feature intensity comparison at ( Figure 3 (c) It was found that as the size of silver nanoparticles increases, 660 cm⁻¹ -1 The peak value gradually increases at a size of 73 nm, reaching its maximum value before decreasing. Therefore, this is considered the optimal size.
[0047] 3) Silver nanoparticle concentration optimization: The concentration gradient was set to 0.935 nM, 1.87 nM, 3.74 nM, 7.48 nM, 9.13 nM, and 14.96 nM. According to... Figure 4 and Figure 5 The Raman signal in cell homogenates increased with increasing silver nanoparticle concentration, with better enhancement observed at concentrations of 7.48 nM and above. However, at concentrations of 0.935 nM, 1.87 nM, and 3.74 nM, the original lung cancer cells (A549) showed significantly reduced Raman signal. PC ) and doxorubicin-resistant lung cancer cells (A549) DOX The Raman spectra of the cells were highly overlapping, making it impossible to successfully distinguish between the two cell types. However, at concentrations of 7.48 nM, 9.13 nM, and 14.96 nM... Figure 5 (df)), the two cell types at 600-800 cm -1 Significant differences began to appear within the Raman shift range, and the Raman characteristics of the two cells could be distinguished visually at 9.13 nM, resulting in the best differentiation effect. Therefore, considering both signal intensity and cell differentiation, 9.13 nM was selected as the optimal concentration for this experiment.
[0048] 4) Cell concentration optimization: Using previously optimized 9.13 nM, 73 nm silver nanoparticles as the SERS enhancement substrate, the effect of cell density on Raman spectroscopy was further observed. Figure 6As cell density decreased, the cell pattern became clearer, and the signal intensity gradually increased. However, when the cell density decreased to 2.5 × 10⁻⁶, the signal intensity decreased further. 6 When the number of cells / mL is at or below, the characteristic peak of the cell (e.g., 748 cm⁻¹) is... -1 The concentration of the two cell types was weakened, and their Raman spectra highly overlapped. Considering the actual sample requirements, the optimal cell concentration was determined to be 2.5 × 10⁻⁶. 6 -10 7 per mL.
[0049] On the other hand, the method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering provided by the present invention may include: S160. Based on the above method for obtaining training data of the tumor cell drug resistance level detection model, obtain surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices, wherein the different drug resistance indices include any one of the following drug resistance indices or any combination thereof: 1x, 3x, 5x, 10x, 15x. S170. Background correction and normalization processing are performed on the surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices. The recursive feature elimination method is used to screen out several feature peaks related to the drug resistance index, and the feature peak data related to the drug resistance index are obtained from the surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices. S180. Using the radial basis function (RBF) to train a support vector machine (SVM) to learn the characteristics of the characteristic peak data related to the drug resistance index in the surface enhanced Raman scattering data of target tumor cells with different drug resistance indices, a tumor cell drug resistance level detection model is obtained. The model classification effect is evaluated on the test set, and the evaluation index of model performance (including classification report, confusion matrix, receiver operating characteristic curve ROC and area under the curve AUC) is calculated and output. S190. Use bar charts to show the weights of several characteristic peaks related to the drug resistance index in the tumor cell drug resistance level detection model, so as to reflect the importance of several characteristic peaks related to the drug resistance index. S1100 performs dimensionality reduction on the surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices and several feature peaks related to the drug resistance index, so as to visualize the effect of cell classification while demonstrating the impact of feature selection on sample separability.
[0050] The following specific embodiment 2 describes a method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering provided by the present invention.
[0051] Example 2 Based on the optimal scheme in Example 1, a model for predicting the drug resistance level of glioma cells (HS683) was constructed.
[0052] 1) Using silver nanoparticles with a size of 73 nm and a concentration of 9.13 nM as the enhanced substrate, the cell density during detection was 5 × 10⁻⁶. 6 Cells / mL or higher. Raman spectra of Hs683 cells under these conditions after different temozolomide stimulation times (resistance indices corresponding to 1, 3, 5, and 10, respectively) are shown below. Figure 7 As shown.
[0053] 2) Silver nanoparticles were synthesized using a seed-mediated citrate and ascorbic acid reduction method. 9 mg of silver nitrate was added to 50 mL of a 40% v / v glycerol-water solution, and the solution was heated to 95 °C with stirring. Then, 1 mL of a 3% (w / v) trisodium citrate solution was rapidly added, and stirring continued for 1 hour until heating was stopped. The solution was then cooled to room temperature to complete the synthesis of the silver seed solution (23 nm). 5 mL of the silver seed solution was mixed with 138 mL of water, 23 mL of glycerol, and 0.58 g of polyvinylpyrrolidone. After the reaction started, 1.15 mL of diamine silver solution (composed of 20 mg silver nitrate, 1 mL of water, and 220 μL of ammonia) and 92 mL of L-ascorbic acid (L-AA, 36.8 mg dissolved in 92 mL of water) were added over 20 seconds. After stirring for 1 hour, 6 g of polyvinylpyrrolidone was added to stabilize the synthesized silver nanoparticles. Ultimately, the diameter of the silver nanoparticles was between 70-80 nm, and the concentration was between 9-10 mM.
[0054] 3) Preferably, when the cells are used for detection, they are suspended in PBS at a density of 5 × 10⁻⁶. 6 -10 7 Cells / mL were processed using a cell disruptor (65 W ultrasonic power). During the process, the cells were sonicated for 3 seconds, with a 5-second interval, for a total of 3 cycles to prepare cell homogenate.
[0055] 4) Mix the silver nanoparticle solution and cell homogenate in a ratio of 1:2 to 1:4, drop the mixture onto the surface of a glass slide uniformly wrapped with aluminum foil, and use a pipette to spread the droplet into a circular area with a diameter of about 0.5 cm.
[0056] 5) Raman spectroscopy was used for detection, with the following parameters: 532 nm laser wavelength, 0.5 mW laser power, 20× objective lens, 10 s integration time, 1 accumulation, and 15 μm laser spot.
[0057] 6) Raman fingerprints of cell homogenates were collected to obtain the Raman spectra R1 of normal tumor cells and Rn of cell lines with different drug resistance indices. Background correction and normalization were performed on the collected spectra. Feature selection was conducted based on the recursive feature elimination (RFE) method to identify key feature peaks highly correlated with drug resistance phenotypes. Simultaneously, cell classification was achieved using a machine learning-based Raman spectroscopy data classification and analysis method (SVM-RFE method). Finally, k optimal features were selected, corresponding to the feature peaks R1-Rn associated with tumor drug resistance. k ,like Figure 8 As shown. After feature selection, the SVM model is trained using the feature subset selected based on RFE and the radial basis function (RBF) kernel. The model's classification performance is evaluated on the test set, and performance metrics (including classification report, confusion matrix, receiver operating characteristic (ROC) curve, and area under the curve (AUC)) are calculated and output. A bar chart is used to display the weights (coefficients) of the k features selected by RFE in the support vector machine (SVM), visually reflecting the importance of each feature. Principal component analysis (PCA) is applied to both the original data and the data after RFE feature selection for dimensionality reduction, demonstrating the impact of feature selection on sample separability while visualizing the cell classification effect. The classification results for HS683 cells are shown below. Figure 9 As shown, it can be completely distinguished, with a distinction accuracy of 100%.
[0058] In another aspect, the method for detecting drug resistance levels in tumor cells provided by the present invention may include: S210, Receive surface-enhanced Raman scattering data of the tumor cells to be tested; S220. Based on the surface-enhanced Raman scattering data of the tumor cells to be tested, the tumor cell drug resistance level detection model obtained by the above-mentioned method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering is used to obtain the tumor cell drug resistance level detection result of the tumor cells to be tested.
[0059] The following is a specific embodiment 3 describing a method for detecting drug resistance levels in tumor cells provided by the present invention. Example 3 Based on the trained tumor cell drug resistance level detection model, the drug resistance index of the sample to be tested is predicted.
[0060] Unknown cell and tissue samples were selected, and their Raman spectra were detected. The Raman spectra of the cells to be tested were input into a tumor cell drug resistance level detection model. Principal component analysis (PCA) was used to reduce the dimensionality of the Raman data after RFE feature selection and then visualized. By comparing the distribution of the cells with those of cells with known different drug resistance states in the 2D or 3D PCA plots, the drug resistance stage of the cells to be tested was determined. For the cells obtained so far, the drug resistance index was determined to be 3 ( Figure 9 The results (*) were compared with those from cell experiments (the measured drug resistance index was 3.25), and the results were basically consistent.
[0061] This invention provides a method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering and its application. Based on surface-enhanced Raman scattering technology combined with machine learning algorithms, a high-precision tumor cell drug resistance level detection model is trained, which can achieve efficient and accurate detection of tumor cell drug resistance levels.
[0062] The present invention has at least the following beneficial effects: 1) Fast, efficient, and real-time: Raman spectral data of tumor cell homogenates are directly acquired using SERS technology. Combined with a pre-trained machine learning model (RFE-RBF-SVM), drug resistance level prediction can be completed within 10 minutes, significantly outperforming traditional drug sensitivity tests (which typically take several days to weeks). This effectively avoids complex steps such as cell culture and high-throughput screening, achieving a "one-stop" rapid detection from sample to result, greatly improving the efficiency of clinical treatment decision-making.
[0063] 2) Precise quantification and high sensitivity: Utilizing silver nanoparticles (70-80 nm, 9-10 mM) as the SERS substrate, the Raman signal is amplified millions of times through electromagnetic enhancement, enabling sensitive capture of minute differences at the molecular level in drug-resistant cells (such as resistance proteins and metabolites). Combined with recursive feature elimination (RFE) to screen key spectral features, and learning the nonlinear mapping relationship between different resistance indices (1 to 15 times) through the RBF-SVM model, continuous quantitative assessment of drug resistance levels can be achieved, overcoming the limitation of traditional biomarker detection (such as protein expression) which can only make qualitative judgments.
[0064] 3) Strong anti-interference capability and stable and reliable results: Using cell homogenate as the detection sample effectively avoids errors caused by the heterogeneity of tumor tissue sampling, improving detection consistency. The machine learning model processes high-dimensional spectral data through the RBF kernel function, which is robust to noise and baseline drift, ensuring the comparability of experimental data from different batches.
[0065] 4) Wide applicability and great potential for clinical translation: It can be adapted to various tumor types (such as breast cancer, lung cancer, etc.) and different drug resistance mechanisms (such as P-gp overexpression, metabolic pathway alterations, etc.), and the application scenarios can be expanded simply by adjusting the training data. The detection process has a high degree of standardization and is easy to integrate with existing detection equipment (Raman spectrometer or portable Raman spectrometer), making it feasible for large-scale promotion.
[0066] In summary, this invention deeply integrates SERS technology with intelligent algorithms, which can solve the core pain points of traditional methods such as long cycle, complex operation and insufficient quantitative ability. It provides an efficient tool for the early detection of tumor drug resistance and personalized medicine, and has important clinical value and scientific research significance.
[0067] The tumor cell drug resistance level detection system provided by the present invention is described below. The tumor cell drug resistance level detection system described below can be referred to in correspondence with the tumor cell drug resistance level detection method described above.
[0068] See Figure 10 The present invention provides a tumor cell drug resistance level detection system, which may include: Raman spectrometer is used to detect surface-enhanced Raman scattering data of tumor cells under test. The data receiving module is used to: receive surface-enhanced Raman scattering data of tumor cells under test from a Raman spectrometer; The detection module is used to: obtain the detection result of the tumor cell drug resistance level of the tumor cells to be tested based on the surface-enhanced Raman scattering data of the tumor cells to be tested, and obtain the tumor cell drug resistance level detection model obtained by the construction method of the tumor cell drug resistance level detection model based on surface-enhanced Raman scattering described above; The data output module is used to output the results of the tumor cell drug resistance level detection of the tumor cells to be tested to at least one terminal.
[0069] Figure 11 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 11 As shown, the electronic device may include a processor 810, a communication interface 820, a memory 830, and a communication bus 840, wherein the processor 810, the communication interface 820, and the memory 830 communicate with each other via the communication bus 840. The processor 810 can call logical instructions in the memory 830 to execute any of the above-described methods for acquiring training data for a tumor cell drug resistance level detection model, constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering, and detecting tumor cell drug resistance levels.
[0070] Furthermore, the logical instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0071] On the other hand, the present invention also provides a computer program product, the computer program product including a computer program, the computer program being able to be stored on a non-transitory computer-readable storage medium, and when the computer program is executed by a processor, the computer is able to execute any of the above-described methods for obtaining training data of the tumor cell drug resistance level detection model, the method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering, and the method for detecting tumor cell drug resistance levels.
[0072] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a method for acquiring training data for a tumor cell drug resistance level detection model as described above, a method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering, and a method for detecting tumor cell drug resistance levels.
[0073] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.
[0074] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.
[0075] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for acquiring training data for a tumor cell drug resistance level detection model, characterized in that, include: Prepare tumor cell test samples with different drug resistance indices, wherein the different drug resistance indices include any one of the following drug resistance indices or any combination thereof: 1x, 3x, 5x, 10x, 15x; A silver nanoparticle solution was prepared, wherein the diameter of the silver nanoparticles in the solution was between 70 and 80 nm, and the concentration of the silver nanoparticle solution was between 9 and 10 mM. Using tumor cell samples with different drug resistance indices, cell homogenates with different drug resistance indices were prepared. Silver nanoparticle solutions were mixed with cell homogenates of different drug resistance indices according to a preset ratio to obtain mixed solutions with different drug resistance indices, wherein the preset ratio was between 1:2 and 1:
4. Surface-enhanced Raman scattering data of tumor cell samples with different drug resistance indices were obtained by using a Raman spectrometer to detect mixed solutions with different drug resistance indices.
2. The method for acquiring training data for the tumor cell drug resistance level detection model according to claim 1, characterized in that, The step of mixing the silver nanoparticle solution with cell homogenates of different drug resistance indices according to a preset ratio to obtain mixed solutions with different drug resistance indices includes: For a mixed solution with a certain drug resistance index, the mixed solution with the drug resistance index is dropped onto the surface of a glass slide uniformly wrapped with aluminum foil, and the droplet is spread into a circular area with a diameter of 0.5 cm using a pipette.
3. The method for acquiring training data for the tumor cell drug resistance level detection model according to claim 1, characterized in that, The optimal diameter of silver nanoparticles in the silver nanoparticle solution was 73 nm, the optimal concentration of the silver nanoparticle solution was 9.13 nM, and the optimal concentration of cell homogenates with different drug resistance indices was 2.5 × 10⁻⁶. 6 -10 7 per mL.
4. A method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering, characterized in that, include: The method for obtaining training data of the tumor cell drug resistance level detection model according to any one of claims 1-3 obtains surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices; Based on surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices, and combined with recursive feature elimination, radial basis function, and support vector machine, the model is trained to learn the features of surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices, thus obtaining a tumor cell drug resistance level detection model.
5. The method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering according to claim 4, characterized in that, The surface-enhanced Raman scattering (SERS) data of target tumor cells with different drug resistance indices, combined with recursive feature elimination, radial basis function (RBF) kernel function, and support vector machine (SVM), are used to train a model to learn the features of SERS data of target tumor cells with different drug resistance indices, resulting in a tumor cell drug resistance level detection model, including: Based on the surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices, a recursive feature elimination method was used to screen out several characteristic peaks related to the drug resistance index, and the characteristic peak data related to the drug resistance index were obtained from the surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices. By using radial basis function kernels to train a support vector machine, the characteristics of the feature peak data related to the drug resistance index in the surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices are learned, and a tumor cell drug resistance level detection model is obtained.
6. The method for constructing a tumor cell drug resistance level detection model based on surface-enhanced Raman scattering according to claim 5, characterized in that, The surface-enhanced Raman scattering (SERS) data of target tumor cells with different drug resistance indices, combined with recursive feature elimination, radial basis function (RBF) kernel function, and support vector machine (SVM), are used to train a model to learn the features of SERS data of target tumor cells with different drug resistance indices, resulting in a tumor cell drug resistance level detection model, including: Bar charts are used to display the weights of several characteristic peaks related to the drug resistance index in the tumor cell drug resistance level detection model, reflecting the importance of these characteristic peaks; and / or, Principal component analysis was used to reduce the dimensionality of surface-enhanced Raman scattering data of target tumor cells with different drug resistance indices and several feature peaks related to drug resistance indices, in order to visualize the effect of cell classification while demonstrating the impact of feature selection on sample separability.
7. A method for detecting drug resistance levels in tumor cells, characterized in that, Receive surface-enhanced Raman scattering data from tumor cells to be tested; Based on the surface-enhanced Raman scattering data of the tumor cells to be tested, the drug resistance level detection model of the tumor cells is obtained by constructing the tumor cell drug resistance level detection model based on surface-enhanced Raman scattering as described in any one of claims 4-6, and the drug resistance level detection result of the tumor cells to be tested is obtained.
8. A system for detecting drug resistance levels in tumor cells, characterized in that, include: Raman spectrometer is used to detect surface-enhanced Raman scattering data of tumor cells under test. The data receiving module is used to: receive surface-enhanced Raman scattering data of tumor cells under test from a Raman spectrometer; The detection module is used to: obtain the drug resistance level detection result of the tumor cells to be tested based on the surface-enhanced Raman scattering data of the tumor cells to be tested, using the tumor cell drug resistance level detection model obtained by the method of constructing the tumor cell drug resistance level detection model based on surface-enhanced Raman scattering as described in any one of claims 4-6; The data output module is used to output the results of the tumor cell drug resistance level detection of the tumor cells to be tested to at least one terminal.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method for acquiring training data of the tumor cell drug resistance level detection model as described in any one of claims 1 to 3, the method for constructing the tumor cell drug resistance level detection model based on surface-enhanced Raman scattering as described in any one of claims 4 to 6, and the tumor cell drug resistance level detection method as described in claim 7.
10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method for acquiring training data of the tumor cell drug resistance level detection model as described in any one of claims 1 to 3, the method for constructing the tumor cell drug resistance level detection model based on surface-enhanced Raman scattering as described in any one of claims 4 to 6, and the tumor cell drug resistance level detection method as described in claim 7.