Terahertz near-field imaging quality analysis system and method

Abstract

The invention relates to the technical field of biomedical imaging and terahertz detection, in particular to a terahertz near-field imaging quality analysis system and method. According to the technical scheme, the system comprises an atomic force microscope system, a terahertz source, an image acquisition module, an image scoring module and a data processing module, and the atomic force microscope system adopts a tapping mode based on the terahertz scattering type scanning near-field optical microscope technology and comprises a Z voltage control unit used for adjusting the probe insertion depth; the terahertz source is used for providing terahertz waves required by imaging, and the terahertz source and the probe module of the atomic force microscope system are coaxially arranged. By quantifying the association rule of the probe insertion depth and the imaging quality, the core pain points of blind parameter adjustment and unstable quality in terahertz near-field imaging are solved at one stroke, the imaging consistency is greatly improved, the detection process and the quality evaluation system are exclusive, the method is particularly suitable for undyed paraffin sections, and the detection accuracy is improved. And the pathological characteristics of papillary carcinoma cells can be clearly presented.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of biomedical imaging and terahertz detection, and particularly relates to a terahertz near-field imaging quality analysis system and method. BACKGROUND

[0002] Papillary thyroid carcinoma is the most common pathological type of thyroid cancer, accounting for about 80-90% of all thyroid cancers. The gold standard for its pathological diagnosis relies on high-resolution morphological observation of cell nucleus characteristics (such as ground glass nuclei, nuclear grooves, and intranuclear pseudoinclusions) and papillary structures. Currently, conventional pathological diagnosis mainly relies on observation of stained paraffin sections under an optical microscope, but the staining process may introduce chemical interference, and cannot provide inherent physical information related to biological macromolecular vibration, hydration state, etc.

[0003] Terahertz waves (frequency range 0.3-3 THz) have great application potential in the field of label-free biomedical detection due to their low photon energy, non-ionizing damage characteristics, and high sensitivity to water content, macromolecular conformation, and vibration modes in biological tissues. In particular, terahertz near-field imaging technology, combined with atomic force microscopy (AFM), achieves nanoscale spatial resolution, breaking through the diffraction limit of traditional terahertz imaging, and providing a powerful tool for observing the microstructure of biological samples at the subcellular scale.

[0004] However, when applying terahertz near-field imaging technology to the detection of unstained paraffin sections of papillary thyroid carcinoma, the existing technology has the following outstanding problems and limitations:

[0005] Imaging quality is unstable, and there is a lack of core parameter control basis: the needle penetration depth of the atomic force microscope probe is a key physical parameter that affects the coupling efficiency of the terahertz near-field signal and the imaging quality. The existing technology usually takes the needle penetration depth as a rough operating variable, and fails to establish a clear correlation between the quantitative control parameter (such as the Z-axis voltage controlling the probe position) and the final imaging quality. This leads to large fluctuations in the imaging clarity of the same cell sample under different experimental batches or different operators, with differences of more than 30%, seriously affecting the reliability and comparability of the results.

[0006] Parameter optimization process is blind and has poor repeatability: due to the lack of the above quantitative relationship as a guide, the parameter optimization in the current technology relies on the personal experience of the operator for trial-and-error adjustment, which is highly blind. This non-standardized operating procedure results in very low repeatability of the imaging results, and the quality consistency of repeated experiments is usually not more than 60%, which is difficult to meet the strict requirements of preclinical research on data repeatability.

[0007] Insufficient adaptability to unstained pathological samples: existing terahertz imaging researches mostly focus on stained samples or cultured cell lines, and there is a lack of research on unstained paraffin sections which are more common in clinics and can better preserve the original information of biological molecules. Unstained samples lack color contrast, making it difficult to quickly and accurately locate blank areas, normal cell areas and cancer cell areas for targeted detection. At the same time, existing technologies fail to systematically quantify the response differences of these different areas in the terahertz band.

[0008] Lack of targeted image quality evaluation criteria: general image quality evaluation software does not combine with the unique pathological morphological features of papillary thyroid cancer cells (such as papillary structure and high nuclear-cytoplasmic ratio), and the correlation between its evaluation indicators and diagnostic value is not strong, which cannot provide specific and authoritative feedback for imaging quality aimed at pathological identification.

[0009] The "terahertz near-field imaging system and method" disclosed in Chinese patent document CN106442394B mentions the basic architecture of combining an atomic force microscope with a terahertz source for near-field imaging, but does not address the problem of probe penetration parameter optimization for biological sample (especially thyroid cancer cell) imaging, thus failing to solve the key technical problems of unstable imaging quality and poor repeatability caused by inaccurate parameter regulation. Therefore, the present application proposes a terahertz near-field imaging quality analysis system and method. SUMMARY

[0010] The purpose of the present application is to address the problem of the terahertz near-field imaging system not involving probe penetration parameter optimization for biological sample imaging in the background art, and to propose a terahertz near-field imaging quality analysis system and method.

[0011] In a first aspect, the present application provides a terahertz near-field imaging quality analysis system, comprising:

[0012] An atomic force microscope system adopts a tapping mode based on terahertz scattering scanning near-field optical microscope (THz s-SNOM) technology and includes a Z voltage control unit for adjusting the probe penetration depth;

[0013] A terahertz source is used to provide terahertz waves required for imaging and is coaxially arranged with the probe module of the atomic force microscope system;

[0014] An image acquisition module is electrically connected to the atomic force microscope system and is used to acquire terahertz near-field images of thyroid papillary carcinoma cell samples;

[0015] An image scoring module is used to evaluate the quality of the acquired terahertz near-field images and output a quality score, and the evaluation indicators include at least image clarity, signal-to-noise ratio and edge sharpness;

[0016] A data processing module is configured to analyze the image quality scores obtained at different Z voltages and establish a quantitative correlation between the Z voltage and the imaging quality.

[0017] Optionally, the Z voltage control unit has an adjustment range of 20V to 180V, wherein 180V corresponds to the zero position of the contact between the probe and the sample surface, and a decrease in voltage indicates an increase in needle depth.

[0018] Optionally, in the image scoring module, the weights of the image definition, signal-to-noise ratio, and edge sharpness are 40%, 30%, and 30%, respectively, and the weighted sum gives a comprehensive quality score of 0 to 100.

[0019] Optionally, the evaluation core of the image definition index is the recognition of the papillary structure and the nuclear-cytoplasmic ratio characteristics of the papillary thyroid cancer cells.

[0020] Optionally, the terahertz source has an operating frequency range of 0.3THz to 3THz and an output power of 1W.

[0021] In a second aspect, a method for analyzing the quality of terahertz near-field imaging is applied to the system for analyzing the quality of terahertz near-field imaging in the first aspect. The method comprises the following steps:

[0022] Sample preparation: prepare an undyed paraffin section of papillary thyroid cancer cells as a test sample, and determine the blank area, normal cell area, and cancer cell area thereon;

[0023] Parameter setting and imaging: set an initial Z voltage on the atomic force microscope system, and perform terahertz near-field imaging on a specific area of the sample at a series of different Z voltage values, and collect multiple groups of images;

[0024] Image quality scoring: automatically score all the images collected in the step by the image scoring module to obtain the quality score of each image;

[0025] Correlation relationship establishment step: statistically analyze the image quality scores obtained at each Z voltage by the data processing module, establish a quantitative relationship model between the Z voltage and the imaging quality score, and determine the optimal imaging Z voltage range based on the model.

[0026] Optionally, in the parameter setting and imaging step, the series of different Z voltage values at least include 180V, 140V, 100V, 60V, and 20V.

[0027] Optionally, in the correlation relationship establishment step, a linear fitting algorithm is used to establish the quantitative relationship model, and the goodness of fit R 2 is not less than 0.9.

[0028] Optionally, based on the established quantitative relationship model, it is determined that the optimal imaging Z voltage range is 20V to 60V, and the image quality score obtained in this range is not less than 80.

[0029] Compared with the prior art, the present application includes at least one of the following beneficial technical effects:

[0030] The present application first quantifies the negative correlation between the probe penetration depth (characterized by Z voltage) of atomic force microscope and the quality of terahertz near-field imaging through a large number of repeated experiments, that is, the lower the Z voltage (the deeper the penetration), the higher the imaging quality score. This provides a clear, data-driven scientific basis for the optimization of key parameters of terahertz near-field imaging technology in biomedical applications, and solves the core pain point of long-term dependence on experience and lack of theoretical guidance in parameter adjustment in this field.

[0031] By using the above quantitative relationship to guide parameter setting, the present application can stabilize the optimal imaging parameters in the range of Z voltage 20V-60V. This greatly improves the imaging quality consistency of repeated experiments from not more than 60% in the prior art to more than 90%. At the same time, the images obtained based on this optimized parameter can clearly present the microscopic structure of cells, and the matching degree of the imaging results with pathological diagnosis reaches more than 85%, greatly enhancing the accuracy and credibility of the technology as an auxiliary diagnostic tool.

[0032] The present application is specially designed for unstained papillary thyroid carcinoma cell paraffin sections, and performs comparative imaging and quantitative analysis on different regions (blank, normal cells, cancer cells) through a standardized process. This method effectively solves the problems of difficulty in distinguishing detection regions and poor imaging contrast caused by the lack of color markers in unstained samples, and provides a special solution for high-quality, marker-free detection of such samples that are closer to the original state of the organism.

[0033] The image scoring software used in the present application emphasizes the recognition ability of key pathological features of papillary thyroid carcinoma cells (such as papillary structure, nuclear-cytoplasmic ratio) through its evaluation indicators (sharpness, signal-to-noise ratio, edge sharpness) and weight settings, overcoming the disadvantage of the disconnection between general scoring software and clinical diagnostic needs, so that the quality score results can directly reflect the potential value of images for pathological identification.

[0034] In summary, the present application solves the core pain points of blind parameter adjustment and unstable quality in terahertz near-field imaging by quantifying the correlation between probe penetration depth (Z voltage) and imaging quality, greatly improves the imaging consistency, and provides a special detection process and quality evaluation system that is particularly suitable for unstained paraffin sections and can clearly present the pathological features of papillary carcinoma cells. BRIEF DESCRIPTION OF DRAWINGS

[0035] Figure 1: Analysis of images collected at different needle voltages based on the analysis system of the present application, wherein:

[0036] Figure 1 -A Analysis of 3 images collected at the same position for a needle voltage of 180v;

[0037] Figure 1 -B Analysis of 3 images collected at the same position for a needle voltage of 140v;

[0038] Figure 1 -C Analysis of 3 images collected at the same position for a needle voltage of 100v;

[0039] Figure 1 -D Analysis of 3 images collected at the same position for a needle voltage of 60v;

[0040] Figure 1 -E Analysis of 3 images collected at the same position for a needle voltage of 20v;

[0041] Figure 2 is a score data table of the collected images of Example 2;

[0042] Figure 3 is a curve graph of the relationship between the needle Z voltage and the imaging quality score of Example 2. DETAILED DESCRIPTION

[0043] The embodiments of the present application will be described in detail with specific reference being made to certain particular embodiments of the application. The advantages of the present application will be apparent from the following description with reference to the figures. Other advantages of the present application will be learned from the practice of the application. The description herein is presented to enable any person skilled in the art to make and use the present application. The description includes specific details for the purpose of providing a thorough understanding of the present application. However, it will be apparent to those skilled in the art that the present application can be practiced without these specific details. In some instances, well-known structures and functions have not been described in detail in order to avoid obscuring the concept of the present application. The following detailed description is, therefore, not to be taken in a limiting sense, and the scope of the present application is indicated by the appended claims.

[0044] Example 1, the present application proposes a terahertz near-field imaging quality analysis system, comprising an atomic force microscope system, a 1W terahertz source, an image collection module, an image scoring software and a data processing module; the components will be described in detail below.

[0045] In this embodiment, the atomic force microscope system adopts THz s-SNOM technology and tapping mode AFM, and contains a control unit for adjusting the needle Z voltage, with a spatial resolution of ≤50nm; the adjustment range of the needle Z voltage of the atomic force microscope system is 20V-180V, and the adjustment accuracy is ±1V;

[0046] The 1W terahertz source is coaxially arranged with the probe module of the atomic force microscope, and is used to provide a terahertz imaging light source; the working frequency of the 1W terahertz source is 0.3THz-3THz, and the power stability is ≤±5%.

[0047] The image acquisition module is electrically connected with the atomic force microscope, and is used for acquiring a terahertz near-field image of the thyroid papillary carcinoma cell;

[0048] The image scoring software is used for scoring the quality of the acquired image, and the data processing module is used for analyzing the correlation between the image quality score and the needle-in Z voltage. The scoring indicators of the image scoring software include image definition (weight 40%), signal-to-noise ratio (weight 30%), and edge sharpness (weight 30%), and the scoring range is 0-100 points, and the higher the score, the better the imaging quality; the definition indicator takes the papillary structure and the nuclear-cytoplasmic ratio of the thyroid papillary carcinoma cell as the core evaluation basis.

[0049] The embodiment also provides a terahertz near-field imaging quality analysis method, which comprises the following steps:

[0050] (1) Sample preparation: fix the undyed thyroid papillary carcinoma cell paraffin section on the sample table of the atomic force microscope, observe under the optical microscope, determine the blank area, the normal cell area (normal follicle periphery) and the cancer cell area (differentiated follicle periphery) as the to-be-measured area, and ensure that the sample surface is flat;

[0051] (2) System debugging: start the atomic force microscope system and the 1W terahertz source, set the working frequency of the terahertz source to 1THz, the initial needle-in Z voltage of the atomic force microscope probe to 180V (needle-in distance 0), the scanning range to 60μm, the scanning frequency to 0.4Hz, and the resolution to 256x256 pixels;

[0052] (3) Multi-parameter imaging: adjust the needle-in Z voltage to a preset value (20V-180V, set a group of parameters at an interval of 40V), acquire 3-5 terahertz near-field images of the same to-be-measured area under each group of voltage through the image acquisition module; the preset Z voltage value includes 180V, 140V, 100V, 60V and 20V; the correlation analysis in step (5) adopts a linear fitting algorithm, and the goodness of fit R² is greater than or equal to 0.9; when the needle-in Z voltage is 20V-60V, the imaging quality is optimal (the score is greater than or equal to 80 points);

[0053] (4) Quality scoring: import the images acquired in step (3) into the image scoring software, and calculate the quality score (take the weighted average value of definition, signal-to-noise ratio and edge sharpness) of each image;

[0054] (5) Correlation analysis: statistically calculate the average quality score corresponding to each group of Z voltage through the data processing module, draw a “Z voltage-imaging quality score” relationship curve, and determine the correlation law.

[0055] In embodiment 2, a terahertz near-field imaging quality analysis system is provided, and the system is described in detail as follows.

[0056] 1. System components

[0057] Atomic force microscope system: light tapping AFM based on THz s-SNOM technology (model: Neaspec neaSNOM), Z voltage regulation range 20V-180V, regulation accuracy ±5V, spatial resolution ≤50nm, probe type Tap300Al-G (resonant frequency 300kHz, force constant 40N / m);

[0058] 1W terahertz source: TDS8000 terahertz source from Virginia Diodes, USA, working frequency 1THz, power stability ±3%, beam diameter 5mm, coaxiality error with probe module ≤0.05mm;

[0059] Image acquisition module: resolution 256x256 pixels, acquisition rate 1 frame / s, supports TIFF format uncompressed storage;

[0060] Image scoring software: developed based on Python, with three major index algorithms built-in:

[0061] o Clarity: calculate the high-frequency component of the image through the Laplacian operator, focus on the identification of the papillary structure of cancer cells (threshold ≥0.8 for clear);

[0062] o Signal-to-noise ratio: gray value ratio of target area (cell) to background area (≥15 for excellent);

[0063] o Edge sharpness: gradient change rate of cell contour, focusing on the clarity of the nuclear-cytoplasmic boundary (≥0.7 for sharp);

[0064] Three indicators weighted calculation of total score (weight 4:3:3);

[0065] Data processing module: equipped with Origin data statistical software, supporting linear fitting, relationship curve drawing and other functions, fitting goodness R²≥0.9.

[0066] 2. Specific operation steps of the analysis method

[0067] (1) Sample preparation:

[0068] Take the undyed papillary thyroid carcinoma cell paraffin section, place it under the optical microscope and observe it at 500 times magnification, and determine three types of test areas: blank area (with conductive film and no cells), normal cell area (normal follicle periphery), and cancer cell area (papillary structure area around differentiated follicle);

[0069] Fix the section on the atomic force microscope sample table, ensure that the sample surface flatness error is ≤2μm.

[0070] (2) System initialization:

[0071] Turn on the atomic force microscope and the terahertz source, and preheat for 5 minutes to stabilize the power;

[0072] Set the terahertz source output power to 1 W and the frequency to 1 THz; the atomic force microscope scanning range to 60 μm, the scanning frequency to 0.4 Hz, the resolution to 256 x 256 pixels, the initial position of the probe to 10 μm from the sample surface, and the initial needle insertion Z voltage to 180 V.

[0073] (3) Multi-parameter imaging:

[0074] Adjust the Z voltage to 180 V (needle insertion distance 0), focus on the cancer cell area, and collect 3 terahertz near-field images;

[0075] Adjust the Z voltage to 140 V, 100 V, 60 V, and 20 V in turn (stabilize for 5 seconds after each adjustment), keep the scanning area, rate, and resolution unchanged, and collect 3 images for each voltage;

[0076] Repeat the above steps to image the normal cell area and the blank area, respectively.

[0077] (4) Quality score:

[0078] Import 45 images (5 groups of voltage x 3 images x 3 repeated experiments) into the scoring software, automatically calculate the clarity, signal-to-noise ratio, and edge sharpness scores of each image, and output the weighted total score.

[0079] (5) Data statistics and correlation analysis:

[0080] Calculate the average quality score for each voltage group (180 V: 62 ± 3.2 points; 140 V: 69 ± 2.8 points; 100 V: 76 ± 2.5 points; 60 V: 83 ± 2.1 points; 20 V: 91 ± 1.8 points);

[0081] Draw the "Z voltage-imaging quality score" relationship curve using Origin software, and obtain the fitting equation: quality score = -0.29 x Z voltage + 117.2 (R² = 0.9873);

[0082] Conclusion: When the needle insertion Z voltage is 20 V-60 V, the imaging quality is optimal (score ≥ 80 points), which can clearly present the papillary structure and nuclear-cytoplasmic ratio characteristics of thyroid papillary carcinoma cells.

[0083] The parameter-mass correlation law of the method has a reproducibility of 95% verified by three repeated experiments, which provides a reliable parameter optimization and quality control basis for terahertz near-field imaging of unstained papillary thyroid carcinoma cell paraffin sections. By observing the imaging graphs under different voltages through three repeated experiments, it can be seen that Figure 1 -D and Figure 1 -E the images have high definition and clear cell outlines, and by observing Figure 1 -A , Figure 1 -B and Figure 1 -C , it can be seen that in the case of shallow needle insertion, the images have different degrees of scratches, affecting the quality of the pictures, and the imaging effect under 180v is the worst, and the imaging effect under 20v is the best.

[0084] As Figure 2 , by analyzing the score data of repeated experiments, it is found that the error between groups is small, and the score error of the images under the same voltage is small, proving the feasibility of the data.

[0085] Figure 3 The relationship between the Z voltage of the needle insertion and the imaging quality score of Example 2 is shown in the graph. By analyzing the relationship between the Z voltage and the imaging score, the relationship between the needle insertion depth and the image quality can be obtained. Through our fitting of the data, we found that the needle insertion depth and the imaging quality can be linearly fitted, and the R square value of the fitted curve is about 0.9873, which is greater than 0.9, further confirming that the needle insertion depth will affect the quality of the images, and the lower the voltage, the better the quality. When the needle insertion voltage is 0V, i.e. the probe is in contact with the sample surface, the best effect can be achieved. However, due to the uneven surface of biological samples, the probe is easily damaged, so in actual operation, we use the system with a needle insertion voltage of about 20V, so under the needle insertion voltage of 20-60V, better imaging pictures can be obtained.

[0086] As Figure 3 shown, the relationship between the needle insertion Z voltage and the imaging quality score is shown in the graph. The horizontal axis is the needle insertion Z voltage (unit: V), taking values of 20, 60, 100, 140, and 180; the vertical axis is the imaging quality score (unit: points); the curve is the linear fitting result of the mean of the three groups of data (R²=0.9873, detailed formula and data are shown in the table at the top right corner of the graph), which shows that as the Z voltage decreases (the needle insertion depth increases), the quality score increases from 62 points (180V) to 91 points (20V), so we can conclude that:

[0087] The needle insertion Z voltage and the imaging quality score are significantly linearly negatively correlated (R²=0.9873, the fitting degree is very high), indicating that the correlation between the two has strong statistical reliability, and the imaging quality can be precisely controlled by the Z voltage;

[0088] When the needle Z voltage decreases from 180V to 20V (corresponding to the gradual deepening of the needle depth), the imaging quality score increases from 62 to 91, with an increase of 46.8%, indicating that deepening the needle depth can effectively reduce imaging interference (such as surface scattering, signal attenuation, etc.), significantly optimizing the imaging clarity, contrast and detail restoration;

[0089] Based on the linear fitting law, a quantitative relationship model between the imaging quality score and the needle Z voltage can be derived (combined with the formula in the upper right corner), providing a precise theoretical basis for the preset of the needle parameters of the detection instrument;

[0090] In actual detection scenarios, if high imaging quality is required (such as precision part defect detection, microstructure analysis, etc.), the needle Z voltage in the range of 20-60V can be preferentially selected; if the balance between imaging efficiency and quality is required (such as rapid batch detection), the voltage parameter can be flexibly adjusted according to the relationship curve to achieve the optimal matching of detection performance and efficiency.

[0091] The present application establishes a clear quantitative relationship between the probe needle depth and the terahertz near-field imaging quality, and by systematically adjusting the Z-axis voltage of the atomic force microscope, the optimization process of the imaging parameters is changed from the trial-and-error mode relying on the personal experience of the operator to the standardized process based on experimental data. The determination of this correlation law provides a key operating basis for obtaining stable and reproducible terahertz images of cells at the nanoscale.

[0092] It is worth noting that the present application improves the repeatability and reliability of terahertz imaging of papillary thyroid carcinoma cells. By setting a specific needle parameter range according to the above quantitative relationship, the system can output images of consistent quality for the same type of sample in different experimental batches. This improvement in stability makes the cell region recognition and differentiation results based on image features more reliable, enhancing the practical value of the imaging technology for pathological analysis. In addition, the present application designs a special detection and evaluation scheme for undyed papillary thyroid carcinoma paraffin sections. The system integrates image scoring indicators for the pathological characteristics of this cell type, achieving a more effective evaluation of the correlation between imaging quality and diagnosis. This solves the problems of difficult region positioning and unclear image interpretation significance faced by general imaging techniques when dealing with such unlabeled biological samples.

[0093] The above specific embodiments are only a few optional embodiments of the present application, and based on the technical solutions of the present application and the related inspiration of the above embodiments, those skilled in the art can make various alternative improvements and combinations to the above specific embodiments.

Claims

1. A terahertz near-field imaging quality analysis system, characterized in that, include: The atomic force microscope system employs a tapping mode based on terahertz scattering scanning near-field optical microscopy technology and includes a Z-voltage control unit for adjusting the probe insertion depth. A terahertz source is used to provide the terahertz waves required for imaging and is coaxially arranged with the probe module of the atomic force microscope system. The image acquisition module is electrically connected to the atomic force microscope system and is used to acquire terahertz near-field images of thyroid papillary carcinoma cell samples. The image scoring module is used to evaluate the quality of the acquired terahertz near-field images and output a quality score. Its evaluation indicators include at least image sharpness, signal-to-noise ratio and edge sharpness. The data processing module is used to analyze the image quality scores obtained under different Z voltages and establish a quantitative correlation between Z voltage and imaging quality.

2. The terahertz near-field imaging quality analysis system according to claim 1, characterized in that, The Z voltage control unit has an adjustment range of 20V to 180V, where 180V corresponds to the zero point position where the probe contacts the sample surface, and a decrease in voltage indicates an increase in the insertion depth.

3. The terahertz near-field imaging quality analysis system according to claim 1, characterized in that, In the image scoring module, the weights of the three indicators—image sharpness, signal-to-noise ratio, and edge sharpness—are 40%, 30%, and 30%, respectively. After weighted summation, a comprehensive quality score ranging from 0 to 100 is obtained.

4. The terahertz near-field imaging quality analysis system according to claim 1, characterized in that, The core of the image sharpness index evaluation is the ability to distinguish the papillary structure and nucleocytoplasmic ratio characteristics of thyroid papillary cancer cells.

5. The terahertz near-field imaging quality analysis system according to claim 1, characterized in that, The terahertz source operates in the frequency range of 0.3THz to 3THz and has an output power of 1W.

6. A terahertz near-field imaging quality analysis method, applied to the terahertz near-field imaging quality analysis system according to any one of claims 1-5, characterized in that, The method includes the following steps: Sample preparation: Unstained paraffin sections of papillary thyroid carcinoma cells were prepared as test samples, and blank areas, normal cell areas and cancer cell areas were identified. Parameter setting and imaging: An initial Z voltage is set on the atomic force microscope system, and terahertz near-field imaging is performed on a specific area of ​​the sample under a series of different Z voltage values ​​to acquire multiple sets of images; Image quality scoring: The image scoring module automatically scores all images acquired in the steps to obtain a quality score for each image; The steps for establishing the correlation are as follows: Through the data processing module, the image quality score obtained under each Z voltage is statistically analyzed to establish a quantitative relationship model between Z voltage and imaging quality score, and the optimal imaging Z voltage range is determined based on the model.

7. The terahertz near-field imaging quality analysis method according to claim 6, characterized in that, In the parameter setting and imaging steps, the series of different Z voltage values ​​include at least 180V, 140V, 100V, 60V and 20V.

8. The terahertz near-field imaging quality analysis method according to claim 6, characterized in that, In the step of establishing the association relationship, a linear fitting algorithm is used to establish the quantitative relationship model, and the goodness of fit R is... 2 Not less than 0.

9.

9. A terahertz near-field imaging quality analysis method according to claim 6, characterized in that, Based on the established quantization relationship model, the optimal imaging Z voltage range is determined to be 20V to 60V, and the image quality score obtained within this range is no less than 80 points.

Images