A fast projection image reconstruction method based on optical coherence tomography system

By replacing traditional steps with one-dimensional digital filters in optical coherence tomography systems, quickly extracting A-line depth information, solving the problem of long imaging in the prior art, real-time reconstruction of en face projected images is achieved, significantly improving Imaging efficiency of OCT system.

CN115644811BActive Publication Date: 2024-11-22UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF 8 Cites 0 Cited by

Patent Information

Application Number
CN202211287849.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-10-20
Publication Date
2024-11-22
Estimated Expiration
2042-10-20

AI Technical Summary

Technical Problem

The existing optical coherence tomography system takes a long time during the imaging process, making it difficult to reconstruct the en-face projected image in real time, affecting the judgment of the position and state of the imaging sample.

Method used

A fast projection image reconstruction method based on optical coherence tomography system is adopted. By acquiring the spectral interference signal, a one-dimensional digital filter is used for filtering, the average or maximum value of the A-line depth information is extracted, and the en face frontal projection is generated. Images, replace traditional linear wavenumber interpolation, spectral shaping, dispersion compensation and Fourier transform steps.

Benefits of technology

It significantly shortens image reconstruction time and improves imaging efficiency, allowing the OCT system to quickly process frontal projected images in clinical applications, suitable for time window sensitive applications, and can be integrated in FPGA hardware.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115644811B_ABST
    Figure CN115644811B_ABST
Patent Text Reader

Abstract

The present invention belongs to the field of image processing technology, and specifically relates to a fast projection image reconstruction method based on an optical coherence tomography system. It includes obtaining a spectral interference signal, filtering the obtained spectral interference signal with a one-dimensional digital filter to obtain A-line depth information; obtaining the average value or maximum value of the A-line depth information; synthesizing according to the average value or maximum value of the obtained A-line depth information to obtain an en face frontal projection image. Compared with the prior art, the present invention omits time-consuming operations such as linear wave number interpolation, spectral shaping, Fourier transform, and modulus extraction performed on data in traditional OCT systems, and its image processing and reconstruction speeds are significantly improved, enhancing the real-time performance of Fourier domain optical coherence tomography technology, which is beneficial to the clinical transformation and application of OCT systems, especially Fourier domain OCT. This method can be integrated in software or FPGA hardware.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of image processing, and in particular relates to a fast projection image reconstruction method based on an optical coherence tomography system. Background Art

[0002] Optical coherence tomography (OCT) is a tomographic imaging technology based on the principle of optical interference. It uses the interference signal between the backscattered light of the sample arm imaging object at different depths and the reflected light of the reference arm plane mirror to obtain the tomographic structure information of the object. OCT can perform non-invasive, minimally invasive, high-resolution three-dimensional imaging of living biological tissues. If it achieves sufficient sensitivity and specificity, it can also be used for real-time diagnosis of pathology. It has broad application prospects in many fields such as ophthalmology, skin, and blood vessels. The current mainstream OCT system is based on the Fourier Domain OCT (Fourier Domain OCT, FDOCT) imaging principle. FDOCT is mainly divided into two types. One is spectral domain OCT (spectral-domain OCT, SDOCT), that is, a spectrometer is used at the detection end to detect the intensity of interference signals of different wavelengths, and then the sample tomographic image is obtained through Fourier transform. The other is swept-source OCT (SSOCT), which uses a fast swept-source light source to change the input wavelength, thereby detecting the intensity of interference signals of different wavelengths or wavenumbers, and then obtaining a sample tomographic image through Fourier transform. During the imaging process, in order to timely observe the position of the sample in the imaging field of view and the target state, such as when imaging the retina, it is necessary to quickly find the macula area or the optic nerve head area for subsequent imaging operations, and it is often necessary to generate an en face frontal projection image in real time, that is, the frontal projection of the sample. The steps of traditional FDOCT processing the collected spectral interference signal: SDOCT first needs to remove DC and perform linear wavenumber interpolation; then perform spectral shaping and dispersion compensation on the signal; further perform Fourier transform and take the modulus value to obtain the depth information of the sample; after logarithmic operation (optional), extract the maximum value or average value of each A-line depth information; finally obtain the en face frontal projection image. SSOCT based on optical clock sampling does not require linear wavenumber interpolation, but is generally similar to the SDOCT data post-processing method. Since the traditional method of obtaining an en face frontal projection image requires scanning the entire sample X and Y planes to obtain a three-dimensional image and then reconstructing it, the above operation is very time-consuming during the imaging process, making it difficult to reconstruct the en face in real time, affecting the judgment of the position and state of the imaging sample. Therefore, there is an urgent need for a new technology that can reconstruct and obtain an equivalent image at a faster speed when processing en face projection images, thereby significantly improving the efficiency of OCT imaging and further assisting clinical applications. Summary of the invention

[0003] The purpose of the present invention is to provide a fast projection image reconstruction method based on an optical coherence tomography system to solve the problem that the imaging process of the existing OCT system takes a long time, resulting in difficulty in real-time reconstruction of the en face and affecting the judgment of the position and state of the imaging sample.

[0004] To achieve the above object, the present invention adopts the following technical solutions:

[0005] A fast projection image reconstruction method based on an optical coherence tomography system, applicable to a spectral domain OCT system or a swept frequency OCT system, comprises the following steps:

[0006] Step 1, obtaining a spectral interference signal;

[0007] Step 2: Using a one-dimensional digital filter to filter the acquired spectral interference signal to obtain A-line depth information;

[0008] Step 3, obtaining the average or maximum value of the A-line depth information;

[0009] Step 4: synthesize the A-line depth information obtained in step 3 according to its average value or maximum value to obtain the enface front projection image.

[0010] Furthermore, when the above-mentioned fast projection image reconstruction method based on the optical coherence tomography system is used in a spectral domain OCT system, step 1 also includes removing DC from the acquired spectral interference signal, and the DC removal method is to take the average value of the spectrum in the fast axis x direction.

[0011] Furthermore, the step 2 also includes performing a logarithmic operation on the obtained A-line depth information to improve image quality.

[0012] Furthermore, the one-dimensional digital filter in step 2 is a high-pass or band-pass filter. When used, the filter coefficient is specified according to the requirements, and the high-pass or band-pass filter is specifically selected to achieve the extraction of different depth information.

[0013] The present invention provides a fast projection image reconstruction method based on an optical coherence tomography system. Based on the law of conservation of energy, the linear wave number interpolation, spectral shaping, dispersion compensation, Fourier transform, and modulus value operations commonly used in traditional FDOCT image processing are replaced with a one-dimensional digital filter; by converting the coefficient vector of the specified digital filter into a high pass or band pass, the information of a specific depth can be extracted, and finally the en face frontal projection of the required depth can be synthesized more quickly. When the processing time of the traditional method is T, the processing time of the fast projection image reconstruction method of the present invention is only T / p (p=2.0~10.0), which greatly improves the image reconstruction efficiency and provides a practical and efficient method for fast processing of frontal projection images for clinical applications sensitive to time windows. In addition, the method of the present invention can also be integrated in FPGA hardware to realize pure hardware real-time frontal projection image reconstruction, which is conducive to the large-scale clinical application of OCT, especially Fourier domain OCT. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 A flowchart of a fast projection image reconstruction method applied to spectral domain OCT (SDOCT) of an embodiment;

[0015] Figure 2 A flowchart of a fast projection image reconstruction method of an embodiment applied to swept-spectrum OCT (SSOCT);

[0016] Figure 3 A method and process for generating en face projection images in a conventional OCT system;

[0017] Figure 4 Flowchart of the reconstruction process of traditional spectral domain OCT (SDOCT) en face projection image;

[0018] Figure 5 Flowchart of the reconstruction process of en face projection images for conventional swept-spectrum OCT (SSOCT);

[0019] Figure 6 A schematic diagram of using a filter to extract specific depth information of a spectral signal in a fast projection image reconstruction method of an embodiment;

[0020] Figure 7 Comparison of the en face projection images and duration of the reconstruction of mouse colon crypts by the traditional SDOCT method and the fast method proposed in this invention;

[0021] Figure 8 Comparison of the en face projection image and time length of reconstructing human finger fingerprint by the traditional SSOCT method and the fast method proposed in this invention. DETAILED DESCRIPTION

[0022] The technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings and embodiments.

[0023] The present embodiment provides a fast projection image reconstruction method based on an optical coherence tomography system, which is applicable to spectral interference signals collected by a SDOCT system and a SSOCT system.

[0024] Figure 1 FIG. 4 is a flowchart of a fast projection image reconstruction method applied to spectral domain OCT (SDOCT) in an embodiment. Figure 1 As shown, the implementation steps are:

[0025] Step 1: Obtain the spectral interference signal, and then take the average value of the spectrum in the fast axis x direction. The spectral interference signal is collected by the line scanning camera in the detection module of the SDOCT system.

[0026] Step 2: Use a one-dimensional digital filter to filter the acquired spectral interference signal to obtain A-line depth information. The filter coefficient of the one-dimensional digital filter is specified by actual needs to achieve the extraction of different depth information. Perform logarithmic operation on the obtained A-line depth information to further improve the image quality.

[0027] Step 3: Calculate the average or maximum value of the A-line depth information.

[0028] Step 4: synthesize the A-line depth information obtained in step 3 according to its average value or maximum value to obtain the enface front projection image.

[0029] The above method can also be integrated into FPGA hardware. The line scan camera in the detection module of the SDOCT system collects the spectral interference signal and transmits it to the computer for storage. The computer transmits the spectral interference signal to the FPGA, which performs subsequent operations such as one-dimensional digital filtering, and finally transmits the result to the computer to synthesize the en face frontal projection image.

[0030] Figure 2 FIG. 4 is a flow chart of a fast projection image reconstruction method applied to swept-spectrum OCT (SSOCT) in an embodiment. Figure 2 As shown, the implementation steps are:

[0031] Step 1: Acquire the spectral interference signal. The spectral interference signal is acquired by a data acquisition card in the detection module of the SSOCT system.

[0032] Step 2: Use a one-dimensional digital filter to filter the acquired spectral interference signal to obtain A-line depth information. The filter coefficient of the one-dimensional digital filter is specified by actual needs to achieve the extraction of different depth information. Perform logarithmic operation on the obtained A-line depth information to further improve the image quality.

[0033] Step 3: Calculate the average or maximum value of the A-line depth information.

[0034] Step 4: synthesize the A-line depth information obtained in step 3 according to its average value or maximum value to obtain the enface front projection image.

[0035] The above method can also be integrated into FPGA hardware. The data acquisition card in the detection module of the SSOCT system collects the spectral interference signal and transmits it to the computer for storage. The computer transmits the spectral interference signal to the FPGA, which performs subsequent operations such as one-dimensional digital filtering, and finally transmits the result to the computer to synthesize the en face frontal projection image.

[0036] Figure 3 The reconstruction method of en face frontal projection images in conventional OCT systems is explained. Figure 3 As shown in the figure, the data obtained by the OCT system is the Bscan longitudinal section data of the xz plane. After scanning in the y direction, a dense scanning sequence Bscan image of the imaging object is obtained, that is, the entire volume data of the object is obtained. The required depth information in the volume data is subsequently processed and projected along the z direction, and the image obtained is the en face frontal projection image.

[0037] Figure 4 The figure shows the SDOCT en face projection image reconstruction process in the traditional OCT system: the spectral interference signal is firstly removed of DC; because the spectral signal collected by the system camera is linearly distributed according to the wavelength, and the Fourier transform needs to operate on the wavenumber, further linear wavenumber interpolation is required; the interference signal is modulated by the light source envelope, and Fourier transform based on the non-Gaussian envelope will cause signal broadening and sidelobes, so further spectral shaping is performed to perform toe-cutting on the signal; the appropriate dispersion coefficient is selected to correct the phase, that is, dispersion compensation is performed; then Fourier transform is performed and the modulus value is taken to obtain the depth information; logarithmic Log operation is performed to improve the image quality, and the maximum or average value information of the A-line depth is extracted according to the actual situation; finally, the information extracted from each A-line is synthesized into an en face front projection image.

[0038] Figure 5The figure shows the SSOCT en face projection image reconstruction process based on optical clock in the traditional OCT system, which is similar to SDOCT: the spectral interference signal is directly spectrally shaped and dispersion compensated, then Fourier transformed and modulo-valued, and logarithmic operation is performed to improve image quality, and the maximum or average value of the A-line depth is extracted according to the actual situation; finally, the information extracted from each A-line is synthesized into an en face frontal projection image.

[0039] By comparing the reconstruction method of this embodiment with the traditional reconstruction method, it can be seen that the en face frontal projection image reconstruction process in the traditional OCT system includes more operations such as linear wavenumber interpolation, spectral shaping, Fourier transform, and modulo taking than the fast reconstruction method proposed in this embodiment. The reconstruction time required is longer and it is difficult to implement and apply.

[0040] Figure 6 FIG. 1 is a schematic diagram of using a filter to extract specific depth information of a spectral signal in a fast projection image reconstruction method of an embodiment. Figure 6 As shown, in this embodiment, in the process of fast en face frontal projection image reconstruction, a one-dimensional digital filter is used to extract specific depth information. After removing the DC from the spectral interference signal modulated by the light source waveform, a new interference signal is obtained. Then, the filter type is determined by specifying a suitable coefficient vector. Using a high-pass filter on it can extract z 0 All the depth information after the depth is processed to synthesize the en face front projection image of the focal plane of the imaging object. The information of a specific depth range (z1-z2) can be extracted by using a bandpass filter, and the en face front projection image of a specific depth can be obtained by subsequent processing.

[0041] Figure 7 The comparison of the traditional and the fast en face projection image reconstruction of the mouse colon crypts using the SDOCT system is shown. The images are collected by the SDOCT system. The left image is the en face projection image processed by the traditional SDOCT method, and the processing time is T; the right image is the image processed by the fast en face method, and the processing time is T / p, where p is usually 2.0-10.0. It can be seen that for the SDOCT system, under the premise of reconstructing an equivalent en face projection image, the reconstruction time is significantly shortened and the imaging efficiency is significantly improved, which is suitable for clinical applications with strong real-time performance and the need to quickly identify the imaging area.

[0042] Figure 8The comparison of traditional and fast en face projection image reconstruction of human finger fingerprints by SSOCT system is shown. The image is collected by SSOCT system. The left picture is the en face image processed by traditional SSOCT method, and the processing time is T; the right picture is the image processed by fast en face method, and the processing time is T / p, where p is usually 2.0-10.0. It can be seen that for SSOCT system, under the premise of reconstructing equivalent en face projection image, the reconstruction time is significantly shortened by the method proposed by the present invention, and the imaging efficiency is significantly improved. The human finger fingerprint image is clearly visible, and its clinical application possibility is greatly improved.

[0043] In summary, the fast projection image reconstruction method based on the optical coherence tomography system proposed in the present invention reduces the processing steps of the en face projection image, reduces its reconstruction and processing time while obtaining an equivalent image, and significantly improves the efficiency of real-time imaging. In addition, the method of the present invention can be integrated into software or FPGA hardware, which is conducive to the clinical transformation and application of OCT systems, especially Fourier domain OCT.

[0044] The above embodiments only illustrate the principles and effects of the present invention, and are not intended to limit the present invention. They are only intended to help understand the principles of the present invention. The protection scope of the present invention is not limited to the above configurations and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the disclosed technology without departing from the essence of the present invention, but they are still within the protection scope of the present invention.

Claims

1. A fast projection image reconstruction method based on an optical coherence tomography system, characterized in that: The following steps are involved: Step 1, obtaining a spectral interference signal; Step 2: Using a one-dimensional digital filter to filter the acquired spectral interference signal to obtain A-line depth information; the one-dimensional digital filter is a high-pass or band-pass filter; Step 3, obtaining the average or maximum value of the A-line depth information; Step 4: Synthesize the average or maximum value of the A-line depth information obtained in step 3 to obtain an en face frontal projection image.

2. The method for rapid projection image reconstruction based on an optical coherence tomography system according to claim 1, characterized in that: The step 1 also includes performing DC removal processing on the acquired spectral interference signal, and the DC removal method is to take the average value of the spectrum in the fast axis x direction.

3. The method for rapid projection image reconstruction based on an optical coherence tomography system according to claim 1, characterized in that: The step 2 also includes performing a logarithmic operation on the obtained A-line depth information to improve image quality.

Citation Information

Patent Citations

  • Optical coherence tomography method

    CN110881947A

  • Endoscopic OCT-Raman dual-mode imaging device and imaging method

    CN112089404A

  • Dispersion compensation method based on Fourier domain optical coherence tomography technology

    CN112597947A

  • Optical coherence tomography dispersion compensation method based on fractional domain parameter detection

    CN112630188A

  • Dispersion compensation method for optical coherence tomography system

    CN112683848A