CN110874548A

CN110874548A - Lung cancer cell and normal cell recognition method based on combination of Raman spectrum and SVM

Info

Publication number: CN110874548A
Application number: CN201811011200.1A
Authority: CN
Inventors: 赵萌; 闫静; 石凡; 贾晨; 陈胜勇
Original assignee: Tianjin University of Technology
Current assignee: Tianjin University of Technology
Priority date: 2018-08-31
Filing date: 2018-08-31
Publication date: 2020-03-10

Abstract

A method for identifying lung cancer cells and normal cells based on a Raman spectrum and an SVM comprises the following steps of ① culturing two cells into cell clusters, ② directly placing the cells on a glass slide after the cells are cultured into the cell clusters, measuring by using a laser confocal Raman spectrometer to obtain a Raman spectrum, ③ obtaining the Raman spectrum, then combining Project FOUR 4.1 software to remove cosmic rays from some spectrums, ④ extracting characteristics of the spectrums after the preprocessing of ③, wherein the extracted characteristics are the positions of characteristic peaks and the intensity ratio of the characteristic peaks, ⑤ classifying and identifying the characteristics extracted from ④ by combining a machine learning method to SVM spectrum data, and ⑥ selecting the rest samples to test to obtain the accuracy of cell identification.

Description

Lung cancer cell and normal cell recognition method based on combination of Raman spectrum and SVM

Technical Field

The invention belongs to the field of computer vision and pattern recognition, and particularly relates to a method for recognizing lung cancer cells and normal cells based on a Raman spectrum combined SVM.

Background

Identifying and locating objects is an important research content in the field of computer vision and pattern recognition, and as a branch of object detection, classification of cancer cells is object detection in a special case. Cells are a special class of substances that are not only ubiquitous but also diverse in specificity. Therefore, the biological identification detection has wide scientific research value and application prospect and has important research significance in the aspect of medicine.

Currently, fluorescent labeling is mainly used to identify the cell type due to its specificity. Fluorescent labeling is based on the specific binding of antigen and antibody, which is a method that is prone to damage to the original physiological activity of the cell and to produce false positive results for specific binding proteins. In addition, there are many drawbacks in clinical applications due to the complexity, cost and inefficiency of sample processing.

The Raman spectrum technology is a molecular inelastic scattering fingerprint spectrum technology, is a non-contact technology, can specifically identify cancer cells on a physical layer, can keep integrity and cell activity, and can effectively solve the complexity and efficiency of pretreatment efficiency of biological samples. The Raman spectrum has strong specificity, reflects the change of biochemical components of living cells in aqueous solution, and has no any mark or fixation. Therefore, the raman spectroscopy has been widely used in the fields of clinical diagnosis, toxicological detection, tissue engineering, and the like. In addition, compared with other technologies, the Raman spectrum technology has the characteristics of rapidness, simplicity, convenience, good repeatability and more importance, and is a nondestructive qualitative and quantitative analysis method. It does not require sample preparation, and the sample can be measured directly by a fiber optic probe or by glass or quartz and fiber optic measurements.

Laser confocal raman spectroscopy is an effective spectroscopic method for analyzing the composition and structure of a substance, and its principle is that incident laser light causes molecules (or crystal lattices) to vibrate. Losing (or gaining) part of the energy, shifting the frequency of the scattered light to scattered light analysis, i.e., raman spectroscopy, can explore the composition, structure, and relative content of known molecules. The micro-Raman technology can focus the light spot of the exciting light to the micron order, so as to accurately analyze the micro-area of the sample, and the exact position of the laser acting on the sample can be clearly displayed by a CCD identifier and a TV monitor. The confocal micro-Raman spectrum can select any part of any sample interested in analysis, and the whole analysis and identification process is very visual and easy to observe and control.

Disclosure of Invention

The invention aims to provide a lung cancer cell and normal cell recognition method based on combination of Raman spectrum and SVM, which utilizes the characteristic that a laser confocal Raman spectrometer can acquire the spectrum and combines with the subsequent characteristic extraction to eliminate the low recognition rate caused by errors generated in the experiment or sample culture process.

In order to achieve the purpose, the scheme of the invention is as follows: a lung cancer cell and normal cell recognition method based on Raman spectrum combined with SVM is characterized in that: the method comprises the following steps:

① culturing the two cells into cell mass;

② culturing the cells into cell clusters, directly placing on a glass slide, and measuring with a laser confocal Raman spectrometer to obtain Raman spectrum;

③ obtaining Raman spectra, and removing cosmic rays from some spectra by combining Project FOUR 4.1 software;

④, extracting the characteristics of the spectrum pretreated in step ③, wherein the extracted characteristics are the position of a characteristic peak and the intensity ratio of the characteristic peak;

⑤, classifying and identifying the spectral data by combining the features extracted from ④ with a machine learning method SVM;

⑥ the remaining samples are selected for testing and the accuracy of cell identification is obtained.

The two cells adopt an adenocarcinoma cell line A549 and a pleural mesothelial cell line Met-5A.

The SVM classifier adopted by the machine learning method SVM is LIBSVM.

The laser confocal raman spectrometer described above is a WITec spectrometer produced in germany.

The invention has the following advantages and positive effects:

1. the invention uses the laser confocal Raman spectrometer with the micro-Raman technology as the background as the experimental equipment, can obtain the real-time Raman spectrum to determine whether the molecular components of the biological substances show the characteristic peak or not through continuous exposure for several times, sampling for several times and proper sampling time, and then obtain the Raman spectrum with grade;

2. the laser confocal Raman spectrometer adopted by the invention is a WITec spectrometer produced by Germany. Meanwhile, by using Project FOUR 4.1 software matched with the spectrometer, the spectrum can be preprocessed to remove interfering cosmic rays, and a Raman spectrum before feature extraction is obtained; and the Raman spectra obtained by experiments can be processed in a large amount in a short time by using Project FOUR 4.1 software;

3. the SVM classifier used after feature extraction is LIBSVM, and the SVM classifier is a simple, easy-to-use, quick and effective SVM pattern recognition and regression software package developed and designed by professor Lingzren of Taiwan university, and the software has relatively less adjustment on related parameters, provides a plurality of default parameters and has obvious classification effect. And provides better performance of interactive tests whether large samples or small samples.

4. The method combines the traditional Raman spectrum feature extraction (the position and the intensity ratio of the feature peak) and the machine learning method SVM to classify and identify the preprocessed Raman spectrum, and for the preprocessing process of the Raman spectrum, the obtained Raman spectrum feature peak has biological meaning and the experimentally obtained spectral line is a very clear visible feature peak, only individual Raman spectrum has cosmic rays, and the cosmic rays are removed by using Project FOUR 4.1 software without smoothing.

5. The data set used as the experimental sample is established by using a WITec spectrometer, and the intensity ratio in the characteristics is obtained by normalizing one peak;

drawings

FIG. 1 is an experimental flow chart of the present invention.

Detailed Description

A lung cancer cell and normal cell recognition method based on Raman spectrum combined with SVM comprises the following steps:

1. two cells were cultured to form a cell pellet: the two cells were: the lung adenocarcinoma cell line A549 and the pleural mesothelial cell line Met-5A.

The culture conditions were that the lung adenocarcinoma cell line A549 was cultured in a DMEM basal medium (containing 10% fetal bovine serum and 1% penicillin-streptomycin double antibody) and the pleural mesothelial cell line Met-5A was cultured in a DMEM high-glucose medium (containing 10% fetal bovine serum) at 37 ℃ in the presence of 5% CO²Culturing in an incubator, collecting into a 15ml sterile centrifuge tube after culturing, washing with Phosphate Buffer Solution (PBS) for 2 times, centrifuging to precipitate cells, centrifuging at 4500rpm for 10min, discarding all supernatant, collecting cells, and facilitating observation;

2. after all the supernatant liquid is discarded, the collected cells are uniformly laid on a glass slide, and a laser confocal Raman spectrometer is used for measuring a Raman spectrum;

3. after obtaining the raman spectra, some of the spectra were processed to remove cosmic rays in conjunction with Project FOUR 4.1 software.

4. Extracting the features of the spectrum after pretreatment, wherein the extracted features are the positions of characteristic peaks, and the intensity ratio in the features is obtained by normalizing one of the peaks; the position of the characteristic peak comprises 1080cm^-1、1128cm^-1、1258cm^-1、1301cm^-1、1342cm^-1、1449cm^-1、1578cm^-1、1617cm^-1、1659cm^-1And the intensity ratio of these 9 position characteristic peaks, where i take the intensity value of the 7 th peak as the normalized value, and these 8 characteristic values are (for example, the first sample) 1.077, 1.047, 1.180, 1.263, 1.214, 1.468, 1.036, and 1.414, respectively.

5. And training the selected characteristics by adopting a machine learning method SVM to obtain an SVM model, wherein the SVM model is obtained by selecting two cell samples in a certain proportion for training, and the lung cancer cell samples are 242, and the normal cell samples are 231. The SVM classifier employs LIBSVM.

6. And selecting the rest samples for testing to obtain the accuracy of cell identification.

The method can effectively identify the cancer cells and the normal cells, the data set serving as the experimental sample is the cell Raman spectrum data set constructed by using the WITec spectrometer, the operation is simple, the size of the constructed SVM model can be determined according to the size of the corresponding data set, the requirement on machine hardware is not high, and the method can be applied to pattern identification in other fields.

It should be noted that the above-mentioned embodiments are only examples of the present invention, and are only illustrative of the present invention, and therefore do not limit the scope of the present invention. The technical idea of the invention is that only obvious changes are needed and still fall within the scope of the invention.

Claims

1. A lung cancer cell and normal cell recognition method based on Raman spectrum combined with SVM is characterized in that: the method comprises the following steps:

① culturing the two cells into cell mass;

2. The method for identifying lung cancer cells and normal cells based on the combination of Raman spectroscopy and SVM of claim 1, wherein: the two cells adopt an adenocarcinoma cell line A549 and a pleural mesothelial cell line Met-5A.

3. The method for identifying lung cancer cells and normal cells based on the combination of Raman spectroscopy and SVM of claim 1, wherein: the SVM classifier adopted by the machine learning method SVM is LIBSVM.

4. The method for identifying lung cancer cells and normal cells based on the combination of Raman spectroscopy and SVM of claim 1, wherein: the laser confocal raman spectrometer described above is a WITec spectrometer produced in germany.