High-resolution cell tomography method based on transmission matrix

Abstract

The invention belongs to the technical field of cell imaging, and particularly relates to a high-resolution cell tomography method based on a transmission matrix. According to the method, the limitation of a traditional optical microscopy technology is broken through by combining a transmission matrix theory, and accurate imaging of a cell structure is realized under the condition that a sample is not damaged. Through accurate reconstruction and processing of a light field in a scattering medium, challenges caused by light diffraction and multiple scattering can be overcome, and clearer cell structure information is provided. Compared with a traditional imaging method, dyeing or repeated scanning is not needed, and the complexity of sample treatment is reduced. Therefore, the invention provides an innovative solution for the development of the cell tomography technology, and opens up a new direction for the application in the fields of biomedical research, pathological analysis, disease diagnosis and the like.

Description

Technical Field

[0001] The invention belongs to the technical field of cell imaging, and in particular relates to a high-resolution cell tomography method based on a transmission matrix. Background Art

[0002] In the field of biomedical imaging, obtaining cell tomography information without damaging the sample has always been a challenge. Traditional optical microscopy technology is limited by light diffraction and scattering effects in deep imaging. In particular, when faced with multiple scattering of photons, the imaging quality will deteriorate, resulting in image blur and reduced contrast, thus limiting the resolution and imaging depth of traditional microscopes. In addition, traditional methods usually require multiple scans or staining, which not only increases the complexity of the experiment, but may also cause damage to the sample. At the same time, high-resolution imaging systems have high requirements for equipment, which also increases costs and time investment.

[0003] Phase microscopy, as an imaging technique that does not rely on staining, is widely used in living cell research by directly capturing the phase information of the sample. It can observe living cells non-invasively by detecting the phase changes caused by the sample. However, traditional phase microscopy still has limitations in tomographic imaging, especially in the analysis of the internal structure of cells, which is limited by scattering and diffraction effects, resulting in signal loss and reduced resolution. Therefore, there is an urgent need for a new imaging method that breaks through the limitations of traditional microscopy and achieves higher-resolution tomographic imaging.

[0004] In summary, traditional optical microscopy has many limitations in depth imaging, cell tomography, and high-resolution imaging, especially in imaging effects in scattering media and avoiding sample damage. Traditional techniques often require multiple scans or staining, which increases the complexity of the experiment and may change the biological properties of the sample or even cause sample damage. Therefore, how to develop a high-resolution, deep imaging technology that does not damage the sample has become an important issue that needs to be solved in the current biomedical imaging field. Summary of the invention

[0005] In view of the above shortcomings, the present invention provides a high-resolution cell tomography technology based on a transmission matrix. This technology uses the transmission matrix theory to accurately measure and process light field information, which can break through the limitations of traditional microscopy technology and achieve higher-resolution cell tomography. Unlike traditional imaging methods, this technology can achieve non-invasive tomography without damaging the sample, avoiding the risk of damage to the sample due to staining or scanning.

[0006] The technical solution adopted by the present invention is:

[0007] A high-resolution cell tomography technique based on a transmission matrix, comprising:

[0008] The measured light field of the cell is resampled through a Gaussian aperture. Specifically, the light field contribution rate is a method to quantify the influence of different propagation distances of the light field in a scattering medium, and is evaluated using a Gaussian aperture. By measuring the light field and establishing a transmission matrix, and then performing singular value decomposition, the cell tomography image of each layer is constructed based on the singular values.

[0009] First, the light field contribution is evaluated using a Gaussian aperture:

[0010]

[0011] Where G(i,j) is the equation of the Gaussian aperture, (i,j) is the center point of the Gaussian aperture, σ is the standard deviation that determines the size of the Gaussian aperture, x∈[1,n],y∈[1,n],i∈[1,n],j∈[1,n], and n represents the pixel size of the light field;

[0012] Define f to represent the overlapping light field. By performing Hadamard product operation on G and f, the sampled light field is obtained, which is recorded as f s (i,j):

[0013] f s (i,j)=G(i,j)⊙f

[0014] The transmission matrix (TM) of the cell is obtained by reconstructing the sampled light field.

[0015] After reconstructing the transmission matrix (TM), singular value decomposition (SVD) is performed. By performing SVD on the TM, the statistical characteristics of the scattering pattern can be obtained. The singular value distribution of the TM can be expressed as:

[0016] TM=UΣV *

[0017] Among them, U and V are orthogonal matrices, * represents the transpose of the matrix, and Σ is a diagonal matrix consisting of real positive singular values ​​arranged in descending order ∑1>∑2>…>∑ N , the size of Σ is N. Larger singular values ​​correspond to the main transmitted photons, while smaller singular values ​​correspond to scattered photons. Therefore, in the transmission matrix, larger singular values ​​correspond to single scattered photons, providing shallow information, while smaller singular values ​​correspond to multiple scattered photons, providing deep information.

[0018]

[0019] in, and is from the vector U m and V m The transformed wavefront, I layer-m Provides information from layer m.

[0020] The cell tomography method provided by the present invention is simple and efficient. In the experiment, only the light field information of the cell needs to be obtained to achieve the acquisition of the transmission matrix, avoiding multiple scanning or staining steps in the traditional method, and significantly improving the imaging efficiency. The technology can provide high-resolution tomography, providing a new tool for cell research, pathological analysis and medical diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of high-resolution cell tomography based on transmission matrix adopted in the present invention.

[0022] Figure 2 Tomographic results of red blood cells. DETAILED DESCRIPTION

[0023] The present invention is described in detail below in conjunction with the accompanying drawings:

[0024] Based on the method of the present invention, high-resolution cell tomography is constructed, such as Figure 1 As shown. The present invention proposes a method for obtaining high-resolution cell tomography from a cell light field. In order to construct the transmission matrix of the cell, we introduce the concept of light field contribution rate and use Gaussian aperture to evaluate it. Here, G(i,j) represents the equation of the Gaussian aperture in equation (2), where (i,j) is the center point of the Gaussian aperture and σ is the standard deviation that determines the size of the Gaussian aperture. Where x∈[1,n],y∈[1,n],i∈[1,n],j∈[1,n], and n represents the pixel size of the light field.

[0025]

[0026] By applying the light field contribution principle, the overlapping light fields can be effectively separated. Let f represent the overlapping light fields. By performing the Hadamard product operation on G and f, the sampled light field can be obtained, which is recorded as fs(i,j).

[0027] f s (i,j)=G(i,j)⊙f (3)

[0028] In this case, adjusting the standard deviation of the Gaussian aperture can control the contribution of the overlapping light fields. Finally, the transmission matrix of the cell can be reconstructed by reconstructing the sampled light fields.

[0029] After reconstructing the transmission matrix (TM), singular value decomposition (SVD) is performed. By performing SVD on the TM, the statistical characteristics of the scattering pattern can be obtained. The singular value distribution of the TM can be expressed as:

[0030] TM=U∑V *

[0031] Among them, U and V are orthogonal matrices, * represents the transpose of the matrix, and Σ is a diagonal matrix consisting of real positive singular values ​​arranged in descending order ∑1>∑2>…>∑ N , the size of Σ. is N. Larger singular values ​​correspond to the main transmitted photons, while smaller singular values ​​correspond to scattered photons. Therefore, in the transmission matrix, larger singular values ​​correspond to single scattered photons, providing shallow information, while smaller singular values ​​correspond to multiple scattered photons, providing deep information.

[0032]

[0033] in, and is from the vector U m and V m The transformed wavefront, I layer-m Provides information from layer m.

[0034] Figure 2 The experimental process of realizing cell tomography is shown. The experiment uses red blood cells as the imaging target, as shown in (a). First, the phase microscope obtains the original light field image of the red blood cells. Then, the phase extraction technology is applied to reconstruct the original light field to obtain the phase light field shown in (b). (c) shows the phase distribution characteristics at the position marked by the white dotted line in the phase light field. Subsequently, based on (b), the phase transfer matrix is ​​constructed, as shown in (d). After singular value decomposition, the probability density distribution of the singular value is shown in (f). In (g)-(j), the tomographic phase image is reconstructed using different singular values, showing the reconstruction results at different depths, revealing the internal structure of the red blood cells. The experimental results show that selecting appropriate singular values ​​for light field reconstruction can effectively extract cell information at different depths. In addition, (k) shows the overall three-dimensional structure of the red blood cells through ImageJ 3D reconstruction technology. Through these experimental results, we can observe the morphological changes of red blood cells at different depths in detail, which helps to identify structural features such as membrane undulations and internal component distribution. By adjusting the interval of singular values, the tomographic thickness can be controlled, with the minimum tomographic thickness being 8.33 nanometers, while the overall cell thickness is about 1 micron. The experimental results show that this technology can achieve high-resolution cell tomography and provides an effective method for high-resolution cell imaging. This technology demonstrates its great potential in cell imaging and can be widely used in biomedical research and cell structure analysis.

[0035] The present invention is not only applicable to the tomographic measurement of cells, but also can be widely used in the tomographic measurement of biological tissues. By utilizing the principle of transmission matrix, the technology can efficiently obtain the light field information in the scattering medium, thereby realizing high-resolution imaging of complex biological structures. This provides a powerful tool for the biomedical field, especially in the observation and analysis of larger-scale and more complex biological tissues, and has great application potential. By performing tomographic imaging on biological tissues, the fine distribution of cell structure, tissue characteristics and internal components can be deeply revealed, thereby providing more details for basic biological research, and providing technical support for early diagnosis of diseases, lesion monitoring and treatment effect evaluation. In addition, the technology can also be applied to fields such as high-resolution microscopy. In microscopic imaging, the measurement of the transmission matrix is ​​crucial for obtaining high-resolution images. The rapid measurement method provided by the present invention can effectively improve the resolution and imaging quality of the microscope. This provides a more sophisticated observation tool for research in the fields of biology and medicine, helping researchers to obtain more accurate image information and further promote the development of related disciplines.

Claims

1. A high-resolution cell tomography method based on a transmission matrix, characterized in that: include: After measuring the cell light field, the measured light field is resampled through the Gaussian aperture. Specifically, the light field contribution rate is defined to quantify the impact of different propagation distances of the light field in the medium, and the Gaussian aperture is used to evaluate the light field contribution rate: Where G(i,j) is the equation of the Gaussian aperture, (i,j) is the center point of the Gaussian aperture, σ is the standard deviation that determines the size of the Gaussian aperture, x∈[1,n],y∈[1,n],i∈[1,n],j∈[1,n], and n represents the pixel size of the light field; Define f to represent the overlapping light field. By performing Hadamard product operation on G and f, the sampled light field is obtained, which is recorded as f s (i,j): f s (i,j)=G(i,j)⊙f The transmission matrix of the cell is obtained by reconstructing the sampled light field; After reconstructing the transmission matrix TM, singular value decomposition is performed to obtain the statistical characteristics of the scattering pattern. The singular value distribution of TM is expressed as: TM=UΣV * Among them, U and V are orthogonal matrices, * represents the transpose of the matrix, and Σ is a diagonal matrix consisting of real positive singular values ​​arranged in descending order Σ1>Σ2>…>Σ N ; Larger singular values ​​correspond to transmitted photons, while smaller singular values ​​correspond to scattered photons. Thus, in the transmission matrix, larger singular values ​​correspond to single scattered photons, providing shallow information, and smaller singular values ​​correspond to multiple scattered photons, providing deep information. The information of the mth layer is expressed as: in, and is from the vector U m and V m The transformed wavefront, 1≤m≤N.

Images