Multilayer imaging method for terahertz near-field system tomography
By adjusting the signal bias value of the terahertz near-field system to form an arithmetic sequence, the problem of unclear depth in terahertz near-field imaging is solved, enabling multi-layer three-dimensional imaging of the internal structure of samples and improving the accuracy of biomedical and materials testing.
Patent Information
- Application Number
- CN202511629112.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-07
- Publication Date
- 2026-02-03
AI Technical Summary
Existing terahertz near-field imaging technology cannot accurately determine the specific internal depth corresponding to each order of signal, making it difficult to achieve three-dimensional tomographic reconstruction and limiting its application in biomedical diagnosis and materials testing.
By adjusting the signal bias value through multiple scans, an arithmetic sequence is formed, and signal data representing different linear depths within the interior are collected to construct a multi-layer three-dimensional tomographic image.
It enables in-depth quantitative imaging of the internal structure of samples, improves the information dimension and practicality of three-dimensional tomographic images, and can accurately locate the depth of the internal structure of biological cells and micro-defects in materials.
Smart Images
Figure CN121453716A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of terahertz near-field imaging technology, specifically relating to a multi-layer imaging method for terahertz near-field system tomography. Background Technology
[0002] Terahertz near-field imaging is an advanced technique capable of high-resolution imaging of sample surfaces and internal structures, with imaging precision down to the nanometer level. This system utilizes the penetrating power of terahertz waves into biological tissues and semiconductor materials, enabling non-destructive probing of internal information without damaging the sample. During imaging, the system simultaneously acquires information about the sample's surface morphology and internal structure: surface information is typically represented by height data, while internal information is decomposed into signal components of different orders, such as first-, second-, and third-order signals, through a Fourier transform of the probe signal. Theoretically, each order corresponds to a response at different depths within the sample.
[0003] However, existing technologies have significant limitations. Although different order signals reflect information about different internal depths to some extent, these depth values lack clear quantitative definitions. There is no definite linear relationship between the depths of different signal orders, making it impossible to accurately determine the specific depth of the microstructure corresponding to a particular signal, such as the precise location of the cell nucleus in organelles or the depth of internal defects in materials. This lack of depth information makes it difficult for traditional terahertz near-field imaging techniques to achieve true three-dimensional tomographic reconstruction, limiting their application in fields requiring precise localization and quantitative analysis of internal structures, such as the fine discrimination of tumor cells or the depth assessment of micro-defects within materials.
[0004] Therefore, the field of terahertz near-field imaging technology urgently needs to develop a multi-layer imaging method that can achieve quantitative calibration of internal depth in order to solve the key bottleneck in current three-dimensional tomographic imaging, namely, how to establish a linear correspondence between signal and internal depth, so as to provide a more reliable and comprehensive three-dimensional imaging tool for biomedical diagnosis and non-destructive testing of materials. Summary of the Invention
[0005] To address the aforementioned technical problems, this invention provides a multi-layer imaging method for tomography of terahertz near-field systems. Based on the characteristic that the imaging data inside the terahertz near-field system is sampled synchronously during scanning, the signal with the deepest imaging depth inside the terahertz near-field system is processed multiple times. Different signal biases are adjusted at the same magnification, so that the signal bias values are arranged in an arithmetic sequence, representing the imaging data at different linear depths inside, thereby achieving the purpose of multi-layer imaging.
[0006] To achieve the above objectives, the present invention adopts the following technical solution:
[0007] A multi-slice imaging method for terahertz near-field system tomography, the method comprising:
[0008] Step 1, Initial Scan: Use a terahertz near-field system to scan the sample, find and determine the signal with the deepest internal linear imaging depth;
[0009] Step 2, Rescan: Based on the signal with the deepest internal linear imaging depth, perform a rescan. While ensuring that the signal has the same amplification factor for each scan, adjust the bias voltage of different scan signals so that the bias values of the scan signals are arranged in an arithmetic sequence to collect signal data representing different internal linear depths.
[0010] Step 3, Imaging: Imaging is performed based on the signal data of different linear depths to form a multi-layer three-dimensional tomographic image with internal depth calibration.
[0011] Furthermore, in step 1, the signal with the deepest internal linear imaging depth is a first-order or second-order signal after Fourier transform.
[0012] Furthermore, when the sample is a semiconductor material, the signal with the deepest internal linear imaging depth is a second-order signal.
[0013] Furthermore, in step 1, the signal with the deepest internal linear imaging depth is the signal with the largest signal amplitude.
[0014] Furthermore, the rescanning in step 2 is achieved through a secondary scanning mode, which involves performing a single scan and simultaneously acquiring signal data under different bias voltages using multiple channels.
[0015] Furthermore, the rescanning in step 2 is achieved through a multi-scanning mode, which involves performing multiple scans, in which signal data under a specific bias voltage is acquired through a single channel in each scan.
[0016] Furthermore, in step 3, the two-dimensional imaging images corresponding to different bias voltages are stacked and fused in the depth direction to construct the multi-layer three-dimensional tomographic image.
[0017] Furthermore, the method is applicable to the observation and localization of organelles at different depths inside biological cells, or to the non-destructive identification and depth localization of micro-defects inside materials.
[0018] In a second aspect, the present invention provides an electronic device comprising: one or more processors; and a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned multi-layer imaging method for terahertz near-field system tomography.
[0019] Thirdly, the present invention provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enable the processor to implement the aforementioned multi-layer imaging method for terahertz near-field system tomography.
[0020] The beneficial effects of this invention are as follows:
[0021] This method enables deep quantitative tomographic imaging of the internal structure of samples. By setting a signal bias that is distributed in an arithmetic sequence, imaging signals at different depths possess linearly corresponding depth information, overcoming the shortcomings of traditional methods that suffer from unclear depth at each order and difficulty in performing three-dimensional quantitative analysis.
[0022] This enhances the information dimensionality and practicality of 3D tomographic images. The reconstructed multi-layer 3D images not only clearly display the internal structure but also accurately annotate the imaging depth of each layer, providing crucial depth coordinates for subsequent analysis.
[0023] In the biomedical field, this method enables direct observation and localization of organelles at different depths within cells, and provides a new means for the precise identification of the nuclear structure inside cancer cells, which has important promotional value for tumor identification and research.
[0024] In the field of materials science, this method can non-destructively and accurately identify and locate micro-defects and their depth inside materials, which helps to analyze the causes of defects and thus improve the testing level and pass rate of material products. Attached Figure Description
[0025] Figure 1 This is a flowchart of a multi-layer imaging method for terahertz near-field system tomography according to the present invention;
[0026] Figure 2 This is a schematic diagram of the signal after a single scan;
[0027] Figure 3 It shows the internal imaging diagram and a single-line scan schematic;
[0028] Figure 4 This is a schematic diagram of the cross-section line of a single-line scan;
[0029] Figure 5 These are internal imaging images and a schematic diagram of a single-line scan after adjusting the bias value;
[0030] Figure 6 This is a schematic diagram of a single-line scan profile after adjusting the offset value. Detailed Implementation
[0031] The present invention will be further described below with reference to the accompanying drawings and embodiments.
[0032] like Figure 1 The diagram shown is a flowchart of a multi-slice imaging method for terahertz near-field system tomography according to the present invention. The method includes:
[0033] Step 1, Initial Scan: Use a terahertz near-field system to scan the sample, find and determine the signal with the deepest internal linear imaging depth;
[0034] Step 2, Rescan: Based on the signal with the deepest internal linear imaging depth, perform a rescan to execute multiple data processing steps. While ensuring that the amplification factor of the scan signal is the same for each scan, adjust the bias voltage of different scan signals so that the bias values of the scan signals are arranged in an arithmetic sequence to collect signal data representing different internal linear depths.
[0035] Step 3, Imaging: Imaging is performed based on the signal data of different linear depths to form a multi-layer three-dimensional tomographic image with internal depth calibration.
[0036] Specifically, in step 1:
[0037] First, the terahertz near-field system is activated, and the sample is initially scanned using a point-by-point scanning method. During this process, the system simultaneously acquires surface data (usually height values) and internal data (usually voltage values) at each scan point. The internal data is then decomposed into signal components of different orders after Fourier transform, such as... Figure 2 As shown, these order signals (such as first order, second order, third order, etc.) roughly correspond to information at different depths inside the sample.
[0038] The key to this step is identifying the signal from these order signals that best characterizes the deepest internal imaging. Generally, a larger signal amplitude indicates a deeper detection depth. In most cases, the first-order signal is considered the deepest signal due to its largest amplitude. However, this rule is not absolute. For example, when the material under test is a semiconductor, the smaller transmission loss of terahertz waves may result in a second-order signal with a larger amplitude than the first-order signal. Therefore, the system operator needs to refer to the spectrum obtained from the initial scan (such as...) Figure 2 The signal with the largest effective order, determined manually or automatically by algorithm, is used as the benchmark for subsequent processing. Generally, this largest signal is a first-order signal, but for semiconductor materials, it may be a second-order signal.
[0039] In step 2:
[0040] After determining the deepest signal in the internal imaging (e.g., selecting a first-order signal), a second scan is required, and the specific signal is processed multiple times to extract information at different linear depths.
[0041] Before performing a second scan, key settings need to be configured for the system scan parameters:
[0042] Fixed magnification: Ensure that all subsequent data acquisitions are performed at the same scanning signal magnification to guarantee data consistency and accuracy of depth calibration.
[0043] Adjusting the bias voltage: For each selected deepest signal, set different bias voltages. These bias voltage values should be arranged in an arithmetic sequence. For example, the bias voltages can be set sequentially as V, V+ΔV, V+2ΔV..., where ΔV is a fixed voltage step value.
[0044] After the parameters are set, perform a second scan. This process can be implemented in two modes:
[0045] Mode A (Secondary Scan - Multi-channel Acquisition): Performs a single scan, but the system simultaneously activates multiple acquisition channels, with each channel synchronously acquiring data of the selected signal according to different preset bias voltage values.
[0046] Mode B (Multiple Scans - Single Channel Acquisition): Performs multiple independent scans. Before each scan, the bias voltage of a single acquisition channel is sequentially adjusted to a preset arithmetic sequence value, thereby acquiring data in stages.
[0047] By using any of the above modes, it is possible to obtain multiple datasets of signals of the same order with different biases.
[0048] In step 3:
[0049] Multiple sets of data obtained from the rescanning were processed for imaging. Each set of data acquired under a specific bias voltage was processed to generate a two-dimensional image. Since the bias voltage varied in an arithmetic sequence, these two-dimensional images represented slice information at different linear depths within the sample.
[0050] Figure 3 and Figure 4 The internal imaging image before bias adjustment and its single-line scan profile are shown. Figure 5 and Figure 6 This shows the changes in the internal imaging and corresponding profile lines after adjusting the bias voltage to a specific value. Figure 3 The medium and light colors represent images of different depths. A dashed line is located at the 10000nm position on the vertical axis, which can be considered a single scan line. The cross-sectional analysis of this scan line is as follows... Figure 4 As shown, the voltage values of different images within the same image differ after a single-line scan. A center line is temporarily drawn at the 0 point of the vertical axis as a reference. After modifying the voltage polarization values and scanning again, the resulting image is shown in Figure 5. It can be seen that... Figure 3There has been a change; the content in the image has significantly decreased, because... Figure 3 The image shows a transverse cut-off portion inside. Similarly, a dashed line is located at the 10000nm position on the vertical axis. Analysis of the scanning profile corresponding to this dashed line is as follows... Figure 6 As shown, Figure 4 The center line, which serves as a reference, has shifted to a position above the 0-point vertical coordinate line. This demonstrates that the shift of the reference line (reflecting a change in the bias voltage) directly leads to changes in the internal imaging information, verifying that internal images of different depths can be "cut out" by controlling the bias.
[0051] Finally, by stacking and fusing all the two-dimensional slice images with different biases (i.e., different depths) in the depth direction, a three-dimensional tomographic image with quantitative depth annotation can be constructed. Each layer in the image corresponds to a specific depth value calculated using an arithmetic progression of bias voltages.
[0052] Experiments have verified the effectiveness of this method in scanning cancer cell samples in the biomedical field. The three-dimensional tomographic images obtained through the above steps can clearly distinguish organelles at different depths within the cell and accurately label the depth of the cell nucleus, providing intuitive and quantitative three-dimensional data for tumor cell identification and biological cell research.
[0053] In the field of materials science, this method is applied to non-destructive testing of semiconductor devices. The generated three-dimensional tomographic images can accurately reveal micron- or nanometer-scale defects (such as pores and cracks) inside the material and directly provide the depth of these defects, greatly improving the accuracy of material quality analysis and product yield.
[0054] In a second aspect, the present invention provides an electronic device comprising: one or more processors; and a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned multi-layer imaging method for terahertz near-field system tomography.
[0055] Thirdly, the present invention provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enable the processor to implement the aforementioned multi-layer imaging method for terahertz near-field system tomography.
[0056] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A multi-slice imaging method for terahertz near-field system tomography, characterized in that, The method includes: Step 1, Initial Scan: Use a terahertz near-field system to scan the sample, find and determine the signal with the deepest internal linear imaging depth; Step 2, Rescan: Based on the signal with the deepest internal linear imaging depth, perform a rescan. While ensuring that the signal has the same amplification factor for each scan, adjust the bias voltage of different scan signals so that the bias values of the scan signals are arranged in an arithmetic sequence to collect signal data representing different internal linear depths. Step 3, Imaging: Imaging is performed based on the signal data of different linear depths to form a multi-layer three-dimensional tomographic image with internal depth calibration.
2. The multi-slice imaging method for terahertz near-field system tomography according to claim 1, characterized in that, In step 1, the signal with the deepest internal linear imaging depth is either a first-order or second-order signal after Fourier transform.
3. The multi-slice imaging method for terahertz near-field system tomography according to claim 2, characterized in that, When the sample is a semiconductor material, the signal with the deepest internal linear imaging depth is a second-order signal.
4. The multi-slice imaging method for terahertz near-field system tomography according to claim 1, characterized in that, In step 1, the signal with the deepest internal linear imaging depth is the signal with the largest signal amplitude.
5. The multi-slice imaging method for terahertz near-field system tomography according to claim 1, characterized in that, The second scan in step 2 is achieved through a secondary scan mode, which involves performing a single scan and simultaneously acquiring signal data under different bias voltages using multiple channels.
6. The multi-slice imaging method for terahertz near-field system tomography according to claim 1, characterized in that, The rescanning in step 2 is achieved through a multi-scanning mode, which involves performing multiple scans, in which signal data under a specific bias voltage is acquired through a single channel in each scan.
7. The multi-slice imaging method for terahertz near-field system tomography according to claim 1, characterized in that, In step 3, the two-dimensional imaging images corresponding to different bias voltages are stacked and fused in the depth direction to construct the multi-layer three-dimensional tomographic image.
8. The multi-slice imaging method for terahertz near-field system tomography according to claim 1, characterized in that, The method is applicable to the observation and localization of organelles at different depths inside biological cells, or to the non-destructive identification and depth localization of micro-defects inside materials.
9. An electronic device, characterized in that, include: One or more processors; Memory, used to store one or more programs; When one or more programs are executed by the one or more processors, the one or more processors implement a multi-slice imaging method for terahertz near-field system tomography as described in any one of claims 1-8.
10. A computer-readable storage medium, characterized in that, It stores executable instructions that, when executed by a processor, enable the processor to implement a multi-slice imaging method for terahertz near-field system tomography as described in any one of claims 1-8.