Brain glioma EGFR amplification state prediction method based on terahertz time-domain spectroscopy

Abstract

The invention discloses a brain glioma EGFR amplification state prediction method based on terahertz time-domain spectroscopy, and belongs to the technical field of machine learning, the brain glioma EGFR amplification state prediction method comprises the following steps: collecting terahertz time-domain spectroscopy data of a brain glioma tissue frozen section, the method comprises the following steps: preprocessing terahertz time-domain spectroscopic data of brain glioma tissue frozen sections, and constructing a sample data set; constructing an EGFR amplification state prediction model; training the constructed EGFR amplification state prediction model by using the sample data set; and realizing brain glioma EGFR amplification state prediction by using the trained EGFR amplification state prediction model. According to the method, the brain glioma EGFR amplification state is predicted by combining the terahertz time-domain spectroscopy with the convolutional neural network, and the method has the advantages of being lossless and rapid in diagnosis. The method is expected to be put into practical application to assist doctors in predicting the brain glioma EGFR amplification state in an operation.

Description

Technical Field

[0001] The present invention relates to the technical field of machine learning, and in particular to a method for predicting the EGFR amplification state of brain glioma based on terahertz time-domain spectroscopy. Background Art

[0002] Gliomas are the most common malignant tumors of the central nervous system in adults, among which glioblastoma is the most malignant and harmful. The epidermal growth factor receptor (EGFR) is composed of a family of tyrosine kinase receptors and directly regulates glioma angiogenesis, which plays an important role in the growth, invasion and recurrence of gliomas. As more and more studies have shown, EGFR amplification and mutation have been identified as driver events in a variety of cancers, especially non-small cell lung cancer, breast cancer and glioblastoma.

[0003] At present, molecular biology techniques are mainly used to detect the amplification status of EGFR. The main methods used are fluorescent in situ hybridization, chromogenic in situ hybridization, real-time quantitative PCR and next-generation sequencing. These methods can accurately diagnose the amplification status of EGFR, but the waiting time for obtaining the results is long. Summary of the invention

[0004] The present invention provides a method for predicting the EGFR amplification state of brain glioma based on terahertz time-domain spectroscopy, so as to solve the current technical problem that the detection time of EGFR amplification state is long.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] On the one hand, the present invention provides a method for predicting the EGFR amplification state of glioma based on terahertz time-domain spectroscopy, and the method for predicting the EGFR amplification state of glioma based on terahertz time-domain spectroscopy comprises:

[0007] The terahertz time-domain spectroscopy data of the frozen sections of brain glioma tissue are collected, and the terahertz time-domain spectroscopy data of the frozen sections of brain glioma tissue are preprocessed to construct a sample data set;

[0008] Construct a prediction model for EGFR amplification status;

[0009] The constructed EGFR amplification status prediction model was trained using the sample data set;

[0010] The trained EGFR amplification status prediction model is used to predict the EGFR amplification status of brain glioma.

[0011] Furthermore, terahertz time-domain spectroscopy data of frozen sections of brain glioma tissue are collected, including:

[0012] The glioma tissue removed during surgery was placed in a pre-cooled phosphate buffered saline in an ice box;

[0013] After the brain glioma tissue is taken out from the phosphate buffer and the surface water is dried, it is placed in a freezing embedding box containing a freezing tissue embedding agent, and the freezing embedding box is placed in a -80°C refrigerator. After the brain glioma tissue is completely solidified, the brain glioma tissue is cut into 100-micron thick slices using a freezing tissue slicer;

[0014] The slices of brain glioma tissue were attached to polyethylene PE sheets as samples;

[0015] The sample is irradiated using a transmission terahertz time-domain spectroscopy system to obtain its terahertz time-domain spectroscopy data.

[0016] Furthermore, when the transmission terahertz time-domain spectroscopy system is used to irradiate the sample, the terahertz optical path of the transmission terahertz time-domain spectroscopy system is placed as a whole in a plexiglass box filled with nitrogen. During the terahertz time-domain spectroscopy data collection process, the relative humidity is maintained below 5% and the temperature is maintained at 20°C±0.2°C.

[0017] Furthermore, when using a transmission-type terahertz time-domain spectroscopy system to irradiate the sample, the sample is placed in a six-hole sample rack, which includes six receiving holes, in which a PE sheet is placed as a reference sample, and one sample is placed in each of the other five receiving holes; before collecting terahertz time-domain spectroscopy data, the six-hole sample rack is placed in an experimental box filled with flowing dry nitrogen for a preset time, and the six-hole sample rack is kept rotating at a preset speed to ensure that the surface of each sample is evenly dried; when starting to collect terahertz time-domain spectroscopy data, the six-hole sample rack is first rotated so that the center of the reference sample is at the focus of the terahertz light, and a reference signal is measured, and then the six-hole sample rack is rotated to ensure that each sample is exactly located at the focus of the terahertz light in turn, so as to obtain terahertz time-domain spectroscopy data of five samples; wherein, when collecting terahertz time-domain spectroscopy data, each sample is collected multiple times and the average of the multiple collection results is taken as the final terahertz time-domain spectroscopy data of the sample.

[0018] Furthermore, the terahertz time-domain spectroscopy data of frozen sections of brain glioma tissues were preprocessed to construct a sample data set, including:

[0019] The reference signal and the terahertz time-domain spectrum data of the frozen section of the glioma tissue are respectively converted into frequency domain spectra by Fourier transform, and the absorption coefficient of the frozen section of the glioma tissue is calculated based on the frequency domain spectrum of the reference signal and the frequency domain spectrum of the terahertz time-domain spectrum data of the frozen section of the glioma tissue;

[0020] The absorption coefficient is smoothed using a Savitzky-Golay filter; wherein the window size of the Savitzky-Golay filter is set to 15 and the polynomial order is set to 3;

[0021] Convert the smoothed absorption coefficient into GASF image and GADF image;

[0022] Convert the generated GASF image and GADF image into grayscale images respectively;

[0023] A sample data set is constructed using the grayscale images of the GASF image and the GADF image corresponding to the frozen section of the brain glioma tissue, as well as the labeling results corresponding to the frozen section of the brain glioma tissue; wherein the labeling results corresponding to the frozen section of the brain glioma tissue are whether the corresponding frozen section of the brain glioma tissue has EGFR amplification.

[0024] Furthermore, when the reference signal and the terahertz time-domain spectrum data of the frozen section of the brain glioma tissue are respectively converted into frequency domain spectra using Fourier transform, for the data to be converted, 11ps before the first echo appears is used as the interception point, the data before the interception point is retained, the remaining data is padded with zeros to 2048 points, and then a radix-2 fast Fourier transform is performed to transform the data into the frequency domain.

[0025] Furthermore, the EGFR amplification status prediction model includes two convolutional neural networks with the same structure, which are used to extract features from the input image, wherein the input of one convolutional neural network is a grayscale image of the GASF image, and the input of the other convolutional neural network is a grayscale image of the GADF image; the features extracted by the two convolutional neural networks are input into the mid-end fusion layer, and the mid-end fusion layer splices the features extracted by the two convolutional neural networks along one dimension into a one-dimensional feature to obtain a fused feature; three fully connected layers are connected to the mid-end fusion layer, and the last fully connected layer outputs the prediction result, that is, whether EGFR amplification exists.

[0026] Furthermore, the convolutional neural network consists of two convolutional layers with a convolution kernel size of 5×5 and two maximum pooling layers of 2×2; the number of neurons in the three fully connected layers is 512, 256, and 256, respectively.

[0027] Furthermore, the RELU function is added after the convolutional neural network and the fully connected layer, and the Sigmoid function is used for the binary classification output. The Adam optimizer is selected to compile the model, and the binary cross entropy is used as the loss function to calculate the loss value in the model optimization process.

[0028] Furthermore, when training the constructed EGFR amplification status prediction model, a five-fold cross-validation training model was used, and the number of training rounds was set to 100;

[0029] The area under the receiver operating characteristic curve was used as the performance indicator for model evaluation.

[0030] On the other hand, the present invention further provides an electronic device, comprising a processor and a memory; wherein the memory stores at least one instruction, and the instruction is loaded and executed by the processor to implement the above method.

[0031] In yet another aspect, the present invention further provides a computer-readable storage medium, wherein at least one instruction is stored in the storage medium, and the instruction is loaded and executed by a processor to implement the above method.

[0032] The beneficial effects brought about by the technical solution provided by the present invention include at least:

[0033] 1. The present invention uses terahertz time-domain spectroscopy combined with convolutional neural network to predict the EGFR amplification status of brain glioma, which has the advantages of non-destructive and rapid diagnosis.

[0034] 2. The present invention is the first to convert the absorption coefficient into two-dimensional images of GASF and GADF as a data set to predict the EGFR amplification status of brain glioma, which can effectively improve the accuracy of EGFR amplification status prediction.

[0035] 3. Compared with conventional machine learning methods, the present invention has higher classification accuracy and generalization ability, and is expected to be put into practical application to assist doctors in predicting the EGFR amplification status of brain glioma during surgery. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0037] Figure 1 It is a schematic diagram of the execution flow of a method for predicting the EGFR amplification state of brain glioma based on terahertz time-domain spectroscopy provided in an embodiment of the present invention;

[0038] Figure 2 is a schematic diagram of the optical path of a terahertz time-domain spectrometer provided in an embodiment of the present invention;

[0039] Figure 3 is a time domain spectrum curve diagram provided by an embodiment of the present invention;

[0040] Figure 4 is a schematic diagram of sample data processing provided by an embodiment of the present invention;

[0041] Figure 5 It is a structural diagram of the EGFR amplification state prediction model provided by an embodiment of the present invention;

[0042] Figure 6 It is a system block diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0043] To make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0044] First of all, it should be noted that in the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "exemplary" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the word "exemplarily" is intended to present the concept in a concrete way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or it can be either of the two.

[0045] First embodiment

[0046] This embodiment provides a method for predicting the EGFR amplification state of brain glioma based on terahertz time-domain spectroscopy. The method can be implemented by an electronic device. The execution flow of the method is as follows: Figure 1 As shown, the following steps are included:

[0047] S1, collecting terahertz time-domain spectroscopy data of frozen sections of brain glioma tissue, preprocessing the terahertz time-domain spectroscopy data of frozen sections of brain glioma tissue, and constructing a sample data set;

[0048] Specifically, in this embodiment, the implementation process of the above S1 is as follows:

[0049] S11, the surgically removed tumor and peritumoral tissue were immediately placed in pre-cooled phosphate buffered saline in an ice box and brought back to the laboratory for frozen section processing;

[0050] S12, after taking the specimen out of the phosphate buffer and drying the surface water, place it in a freezing embedding box containing frozen tissue embedding agent, and place it in a -80℃ refrigerator. After it is completely solidified, use a frozen tissue slicer to cut the tissue into 100 micron thick slices, and then use a customized polyethylene PE sheet with a thickness of 2mm as a carrier to evenly attach the frozen slices to the PE sheet for terahertz measurement experiments. Among them, all slices are immediately stored at -80℃ until the terahertz measurement experiment begins.

[0051] S13, using a transmission terahertz time-domain spectroscopy system (THz-TDS) to illuminate the slice and obtain the sample signal;

[0052] It should be noted that the optical path of the transmission terahertz time-domain spectroscopy system is as follows: Figure 2 As shown in the figure, the generation and detection of terahertz waves uses a photoconductive antenna device based on low-temperature gallium arsenide. The infrared light output by the femtosecond laser (Toptica Photonics Femtofiber Smart 780, central wavelength 780nm, pulse width <90fs) is beam-split and used as pump light and detection light to illuminate the center of the photoconductive antenna, with an intensity of 15mW. A stepper motor delay line is added to the pump optical path to realize the waveform scanning of the terahertz pulse. The terahertz pulse is beam-converted through four off-axis parabolic mirrors and focused in the sample area, with a beam waist diameter of 5mm. The scanning step length of the time domain signal is 33.3fs, and a total of 1024 points are collected. When there are biological samples in the terahertz wave propagation path, the available frequency band is selected to be 0.2-1.4THz.

[0053] In addition, it should be noted that water vapor in the terahertz wave propagation path has a strong absorption on the signal. In order to eliminate the characterization effect caused by humidity changes, the terahertz optical path is placed in a plexiglass box filled with nitrogen. During the test, the relative humidity is kept below 5% and the temperature is kept at 20℃±0.2℃. In order to improve the test efficiency and make the test environment in the same control group the same, this experiment uses a self-made six-hole sample rack, which is used to place 1 reference sample (pure PE sheet) and 5 experimental samples (PE sheets with tissue samples). During the experiment, the samples were installed and placed on a porous rotating sample rack in sequence. Before measurement, in order to remove the condensed water generated on the surface of the samples after transferring from a frozen environment to a room temperature environment, they needed to be placed in an experimental box filled with flowing dry nitrogen for a period of time. During the process, the sample rack was kept rotating slowly to ensure that the surface of each sample was evenly dried. When the measurement started, the sample position was first rotated and adjusted so that the center of the pure PE sheet used as a reference signal was at the focus of the THz light, and the reference signal was measured. Thereafter, the sample rack was rotated 60 degrees each time to ensure that the subsequent sample to be measured was exactly at the focused light. In this way, the subsequent 5 groups of sample signals were obtained in sequence. The data for each measurement was collected twice and averaged to reduce the error.

[0054] S14, using Fourier transform to convert the reference signal and the terahertz time-domain spectrum data of the frozen section of the brain glioma tissue into frequency domain spectra respectively, and based on the frequency domain spectrum of the reference signal and the frequency domain spectrum of the terahertz time-domain spectrum data of the frozen section of the brain glioma tissue, calculate the absorption coefficient of the frozen section of the brain glioma tissue;

[0055] It should be noted that when the collected raw data is plotted in the time domain, a curve with echoes will be obtained. At this time, direct Fourier transform will result in the instrument effect caused by multiple reflections inside the sample. In order to eliminate the oscillation in the spectrum signal, Figure 3 As shown, in this embodiment, a certain point (11ps) before the first echo appears is used as the interception point, the data before the interception point is retained, and the remaining time domain data is padded with zeros to 2048 (2 11 ) point and then perform a radix-2 fast Fourier transform to the frequency domain. Figure 4 As shown, the reference signal spectrum is recorded as E ref (ω), the sample signal is recorded as E sam (ω), and the transfer function T(ω) inside the sample is obtained by using the ratio of the two. Since the data was intercepted before, only the main peak signal was retained after preprocessing, and multiple reflections do not need to be considered when calculating the terahertz parameters. Therefore, the sample refractive index n can be obtained by calculation. s (ω), extinction coefficient κ s (ω), and the absorption coefficient α s (ω), the specific calculation formula is as follows:

[0056]

[0057] The absorption coefficient of the sample can be calculated from the extinction coefficient:

[0058]

[0059] Where d is the sample thickness, T(ω) is the sample transfer function, c is the speed of light in vacuum, and n PE is the refractive index of PE, and n is measured experimentally. PE =1.50.

[0060] S15, using a Savitzky-Golay filter to smooth the absorption coefficient; wherein the window size of the Savitzky-Golay filter is set to 15, and the polynomial order is set to 3;

[0061] S16, converting the smoothed absorption coefficient into a two-dimensional image of a Gram angle sum field (GASF) image and a Gram angle difference field (GADF), as follows:

[0062]

[0063]

[0064] Where GASF is the cosine of the angle sum, GADF is the cosine of the angle difference, n is the number of frequency points, φ n is the angle value at the nth frequency point.

[0065] S17, converting the GASF image and the GADF image into grayscale images with a resolution of 256×256 respectively, so the size of a single image is 256×256×1 (pixel×pixel×channel);

[0066] S18, constructing a sample data set using the grayscale images of the GASF image and the GADF image corresponding to the frozen section of the brain glioma tissue, and the labeling result corresponding to the frozen section of the brain glioma tissue; wherein the labeling result corresponding to the frozen section is whether the corresponding frozen section of the brain glioma tissue has EGFR amplification.

[0067] S2, construct a prediction model for EGFR amplification status;

[0068] Among them, the EGFR amplification status prediction model structure constructed in this embodiment is as follows Figure 5 As shown in the figure, the network mainly consists of two convolutional neural networks with the same structure, which are fused and fully connected after feature extraction. The convolutional neural network model mainly consists of two convolutional layers with a convolution kernel size of 5×5 and two maximum pooling layers of 2×2. The mid-end fusion splices the extracted features into a one-dimensional feature along one dimension. At the end of the model, three fully connected layers are constructed. The number of neurons in the fully connected layers is 512, 256, and 256 respectively. In addition, the RELU function is added after the convolutional layer and the fully connected layer, and the binary classification output uses the Sigmoid function. The Adam optimizer is selected to compile the model, and the binary cross entropy is used as the loss function as the loss value in the model optimization process. The model is trained using five-fold cross validation, and the training round is set to 100.

[0069] S3, using the sample data set to train the constructed EGFR amplification status prediction model;

[0070] Among them, the number of samples collected in this embodiment is 440, of which 120 are EGFR amplification and the remaining 320 are EGFR non-amplification. After smoothing by Savitzky-Golay filter and generating corresponding two-dimensional images using GASF and GADF, there are 440 groups of image data, each group has corresponding GASF and GADF images, and these data are divided into training set and test set at a ratio of 4:1, wherein the training set uses five-fold cross validation to train and verify the model.

[0071] Further, this embodiment uses the area under the receiver operating characteristic curve (Area Under the Receiver Operating Characteristic Curve, ROC AUC) value as the main performance indicator of model evaluation. The ROC curve is a graphical tool for representing the performance of a classification model. It plots the relationship between the true positive rate (True Positive Rate, TPR) and the false positive rate (False Positive Rate, FPR) under different threshold settings. AUC represents the area under the ROC curve, which is used to measure the performance of the classifier. The closer the AUC value is to 1, the better the classifier performance is, while the AUC value close to 0 indicates poor performance.

[0072] After experiments, the final prediction AUC value of the model of this embodiment on the test set was 94.74%, which proves that the model of this embodiment has a high classification accuracy and is expected to be put into practical application to assist doctors in predicting the EGFR amplification status of brain glioma during surgery.

[0073] S4, using the trained EGFR amplification status prediction model to predict the EGFR amplification status of brain glioma.

[0074] In summary, this embodiment provides a method for predicting the EGFR amplification state of glioma based on terahertz time-domain spectroscopy, which uses terahertz time-domain spectroscopy combined with convolutional neural networks to predict the EGFR amplification state of glioma, and has the advantages of non-destructive and rapid diagnosis. Compared with conventional machine learning methods, this method is the first to convert the absorption coefficient into a two-dimensional image of GASF and GADF as a data set to predict the EGFR amplification state of glioma, which can effectively improve the accuracy of EGFR amplification state prediction. It has high classification accuracy and generalization ability, and can be used to assist doctors in predicting the EGFR amplification state of glioma during surgery.

[0075] Second embodiment

[0076] This embodiment provides an electronic device, such as Figure 6 As shown, the electronic device includes: a processor and a memory; wherein the processor and the memory can be connected via a communication bus; the memory stores at least one instruction, and the instruction is loaded and executed by the processor to implement the method of the first embodiment. In addition, the electronic device may also include a transceiver, the processor and the transceiver can be connected via a communication bus, and the transceiver is used to communicate with other devices.

[0077] Next, combine Figure 6 The following is a detailed introduction to the various components of the electronic device:

[0078] Among them, the processor is the control center of the electronic device, and the electronic device may include multiple processors, each of which may be a single-core processor (single-CPU) or a multi-core processor (multi-CPU). The processor here may be a processor or a general term for multiple processing elements. For example, the processor is one or more central processing units (CPUs), or other general-purpose processors, application specific integrated circuits (ASICs), or one or more integrated circuits configured to implement an embodiment of the present invention, such as one or more microprocessors (digital signal processors, DSPs), or one or more field programmable gate arrays (field programmable gate arrays, FPGAs), or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor may execute various functions of the electronic device by running or executing software programs stored in the memory and calling data stored in the memory.

[0079] In a specific implementation, as an embodiment, the processor may include one or more CPUs, such as Figure 6 The CPU0 and CPU1 shown in the figure are, of course, only exemplary.

[0080] The memory is used to store the software program for executing the solution of the present invention, and the execution is controlled by the processor. The specific implementation method can refer to the above method embodiment, which will not be repeated here.

[0081] Optionally, the memory may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory may be integrated with the processor or exist independently and accessed through the interface circuit ( Figure 6 (not shown) is coupled to the processor, which is not specifically limited in this embodiment of the present invention.

[0082] The transceiver may include a receiver and a transmitter ( Figure 6 The receiver is used to implement the receiving function, and the transmitter is used to implement the sending function. The transceiver can be integrated with the processor or exist independently and communicate with the electronic device through the interface circuit ( Figure 6 (not shown) is coupled to the processor, which is not specifically limited in this embodiment of the present invention.

[0083] In addition, it should be noted that Figure 6 The structure of the electronic device shown in the figure does not constitute a limitation on the device, and the actual device may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently. In addition, the technical effects achieved by the electronic device when executing the method of the first embodiment above can refer to the technical effects described in the first embodiment above, so they are not repeated here.

[0084] Third embodiment

[0085] This embodiment provides a computer-readable storage medium, which stores at least one instruction, and the instruction is loaded and executed by a processor to implement the method of the first embodiment. The computer-readable storage medium may be a ROM, a random access memory, a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc. The instructions stored therein may be loaded by a processor in a terminal to execute the method.

[0086] In addition, it should be noted that the present invention can be provided as a method, an apparatus or a computer program product. Therefore, the embodiment of the present invention can be in the form of a full or partial hardware embodiment, a full or partial software embodiment or an embodiment combining software and hardware. Moreover, when implemented using software, the embodiment of the present invention can be in the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program codes. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center containing one or more available media sets. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a DVD), or a semiconductor medium. The semiconductor medium may be a solid state hard disk.

[0087] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0088] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device, so that a series of operation steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable terminal device provide for implementing the process in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0089] It should also be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. The terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of more restrictions, the elements defined by the sentence "including one..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements. In addition, the term "and / or" is only an association relationship describing the associated objects, indicating that there can be three relationships, for example, A and / or B, which can represent: A exists alone, A and B exist at the same time, and B exists alone, wherein A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding. "At least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can be represented by: a, b, c, ab, ac, bc or abc, where a, b, c can be single or plural.

[0090] In addition, it can be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0091] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0092] In several embodiments provided by the present invention, it should be understood that the disclosed equipment, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of functional modules / units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point, the coupling or direct coupling or communication connection between each other shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms. The unit described as a separate component may or may not be physically separated, and the component displayed as a unit may or may not be a physical unit, that is, it may be located in one place, or it may be distributed on multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, each functional unit in each embodiment of the present invention can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0093] If the method is implemented in the form of a software functional unit and sold or used as an independent product, it 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 the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0094] Finally, it should be noted that the above is only a preferred embodiment of the present invention. It should be pointed out that although the preferred embodiment of the present invention has been described, for ordinary technicians in this technical field, once the basic creative concept of the present invention is known, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the protection scope of the present invention. Therefore, the attached claims are intended to be interpreted as including the preferred embodiment and all changes and modifications that fall within the scope of the embodiments of the present invention.

Claims

1. A method for predicting EGFR amplification status of brain glioma based on terahertz time-domain spectroscopy, characterized in that: The method for predicting the EGFR amplification state of brain glioma based on terahertz time-domain spectroscopy includes: The terahertz time-domain spectroscopy data of the frozen sections of brain glioma tissue are collected, the terahertz time-domain spectroscopy data of the frozen sections of brain glioma tissue are preprocessed, and a sample data set is constructed; Construct a prediction model for EGFR amplification status; The constructed EGFR amplification status prediction model was trained using the sample data set; The trained EGFR amplification status prediction model is used to predict the EGFR amplification status of brain glioma.

2. The method for predicting EGFR amplification status of glioma based on terahertz time-domain spectroscopy according to claim 1, characterized in that: Acquire terahertz time-domain spectroscopy data of frozen sections of brain glioma tissue, including: The glioma tissue removed during surgery was placed in a pre-cooled phosphate buffered saline in an ice box; After the brain glioma tissue is taken out from the phosphate buffer and the surface water is dried, it is placed in a freezing embedding box containing a freezing tissue embedding agent, and the freezing embedding box is placed in a -80°C refrigerator. After the brain glioma tissue is completely solidified, the brain glioma tissue is cut into 100-micron thick slices using a freezing tissue slicer; The slices of brain glioma tissue were attached to polyethylene PE sheets as samples; The sample is irradiated with a transmission terahertz time-domain spectroscopy system to obtain its terahertz time-domain spectroscopy data.

3. The method for predicting EGFR amplification status of glioma based on terahertz time-domain spectroscopy according to claim 2, characterized in that: When using the transmission terahertz time-domain spectroscopy system to irradiate the sample, the terahertz optical path of the transmission terahertz time-domain spectroscopy system is placed in a plexiglass box filled with nitrogen. During the terahertz time-domain spectroscopy data collection process, the relative humidity is maintained below 5% and the temperature is maintained at 20°C ± 0.2°C.

4. The method for predicting EGFR amplification status of glioma based on terahertz time-domain spectroscopy according to claim 2, characterized in that: When using a transmission-type terahertz time-domain spectroscopy system to irradiate a sample, the sample is placed in a six-hole sample rack, which includes six receiving holes, in which a PE sheet is placed as a reference sample, and one sample is placed in each of the other five receiving holes; before collecting terahertz time-domain spectroscopy data, the six-hole sample rack is placed in an experimental box filled with flowing dry nitrogen for a preset time, and the six-hole sample rack is kept rotating at a preset speed to ensure that the surface of each sample is evenly dried; when starting to collect terahertz time-domain spectroscopy data, the six-hole sample rack is first rotated so that the center of the reference sample is at the focus of the terahertz light, and a reference signal is measured, and then the six-hole sample rack is rotated to ensure that each sample is exactly located at the focus of the terahertz light in turn, so as to obtain terahertz time-domain spectroscopy data of five samples; wherein, when collecting terahertz time-domain spectroscopy data, each sample is collected multiple times and the average of the multiple collection results is taken as the final terahertz time-domain spectroscopy data of the sample.

5. The method for predicting EGFR amplification status of brain glioma based on terahertz time-domain spectroscopy according to claim 4, characterized in that: The terahertz time-domain spectroscopy data of frozen sections of brain glioma tissue were preprocessed to construct a sample data set, including: The reference signal and the terahertz time-domain spectrum data of the frozen section of the glioma tissue are respectively converted into frequency domain spectra by Fourier transform, and the absorption coefficient of the frozen section of the glioma tissue is calculated based on the frequency domain spectrum of the reference signal and the frequency domain spectrum of the terahertz time-domain spectrum data of the frozen section of the glioma tissue; The absorption coefficient is smoothed using a Savitzky-Golay filter; wherein the window size of the Savitzky-Golay filter is set to 15 and the polynomial order is set to 3; Convert the smoothed absorption coefficient into GASF image and GADF image; Convert the generated GASF image and GADF image into grayscale images respectively; A sample data set is constructed using the grayscale images of the GASF image and the GADF image corresponding to the frozen section of the brain glioma tissue, as well as the labeling results corresponding to the frozen section of the brain glioma tissue; wherein the labeling results corresponding to the frozen section of the brain glioma tissue are whether the corresponding frozen section of the brain glioma tissue has EGFR amplification.

6. The method for predicting EGFR amplification status of glioma based on terahertz time-domain spectroscopy according to claim 5, characterized in that: When the reference signal and the terahertz time-domain spectrum data of the frozen section of the brain glioma tissue are respectively converted into frequency domain spectra using Fourier transform, for the data to be converted, 11ps before the first echo appears is used as the interception point, the data before the interception point is retained, the remaining data is padded with zeros to 2048 points, and then a radix-2 fast Fourier transform is performed to transform the data into the frequency domain.

7. The method for predicting EGFR amplification status of glioma based on terahertz time-domain spectroscopy according to claim 5, characterized in that: The EGFR amplification status prediction model includes two convolutional neural networks with the same structure, which are used to extract features from the input image, wherein the input of one convolutional neural network is the grayscale image of the GASF image, and the input of the other convolutional neural network is the grayscale image of the GADF image; the features extracted by the two convolutional neural networks are input into the mid-end fusion layer, and the mid-end fusion layer splices the features extracted by the two convolutional neural networks along one dimension into a one-dimensional feature to obtain a fused feature; three fully connected layers are connected to the mid-end fusion layer, and the last fully connected layer outputs the prediction result, that is, whether EGFR amplification exists.

8. The method for predicting EGFR amplification status of glioma based on terahertz time-domain spectroscopy according to claim 7, characterized in that: The convolutional neural network consists of two convolutional layers with a convolution kernel size of 5×5 and two maximum pooling layers of 2×2; the number of neurons in the three fully connected layers is 512, 256, and 256, respectively.

9. The method for predicting EGFR amplification status of glioma based on terahertz time-domain spectroscopy according to claim 8, characterized in that: The RELU function is added after the convolutional neural network and the fully connected layer, and the Sigmoid function is used for the binary classification output. The Adam optimizer is selected to compile the model, and the binary cross entropy is used as the loss function to calculate the loss value in the model optimization process.

10. The method for predicting EGFR amplification status of glioma based on terahertz time-domain spectroscopy according to claim 9, characterized in that: When training the constructed EGFR amplification status prediction model, a five-fold cross-validation training model was used, and the training rounds were set to 100; The area under the receiver operating characteristic curve was used as the performance indicator for model evaluation.

Images