阅读说明：本技术 体内全场干涉显微成像的方法和系统 (Method and system for in vivo full-field interference microscopic imaging ) 是由 V·马兹林 肖鹏 M·芬克 A·C·博卡拉 于 2019-09-27 设计创作，主要内容包括：根据一个方面,本发明涉及一种用于对散射三维样品进行体内全场干涉显微成像的系统(101)。该系统包括：用于提供样品的正面图像的全场OCT成像系统(130),其中,所述全场OCT系统包括干涉设备(145)以及采集设备(138),所述干涉设备(145)具有旨在接收样品的物臂(147)以及包括光学透镜(134)和第一反射表面(133)的参考臂(146),所述采集设备(138)被配置为采集由在成像场的每个点处产生的干涉引起的二维干涉信号(I-1、I-2)的时间序列；OCT成像系统(110),该OCT成像系统(110)用于在所述二维干涉信号的相同采集时间处提供样品以及所述全场OCT成像系统(130)的第一反射表面(133)的横截面图像；处理单元(160),该处理单元(160)被配置为确定样品的多个切片的多个正面图像(X-Y),每个正面图像从具有给定相移的至少两个二维干涉信号(I-1、I-2)确定,该处理单元(160)被配置为从OCT成像系统(110)在所述两个二维干涉信号(I-1、I-2)中的每一个的采集时间处提供的横截面图像,确定所述多个切片的每个正面图像(X-Y)的深度(z),并且被配置为从样品的所述多个切片的所述多个正面图像和深度确定样品的3D图像。(According to one aspect, the invention relates to a system (101) for in vivo full field interferometric microscopy imaging of a scattering three-dimensional sample. The system comprises: a full-field OCT imaging system (130) for providing a frontal image of a sample, wherein the full-field OCT system comprises an interference device (145) and an acquisition device (138), the interference device (145) having an object arm (147) intended to receive the sample and a reference arm (146) comprising an optical lens (134) and a first reflective surface (133), the acquisition device (138) being configured to acquire a two-dimensional interference signal (I) caused by interference generated at each point of the imaging field 1 、I 2 ) A time series of (a); an OCT imaging system (110), the OCT imaging system (110) being forProviding cross-sectional images of a sample and a first reflective surface (133) of the full-field OCT imaging system (130) at a same acquisition time of the two-dimensional interference signal; a processing unit (160), the processing unit (160) being configured to determine a plurality of frontal images (X-Y) of a plurality of slices of the sample, each frontal image being derived from at least two-dimensional interference signals (I) with a given phase shift 1 、I 2 ) Determining, the processing unit (160) is configured to determine, from the OCT imaging system (110), at the two-dimensional interference signals (I) 1 、I 2 ) Determining a depth (z) of each frontal image (X-Y) of the plurality of slices, and configured to determine a 3D image of the sample from the plurality of frontal images and depths of the plurality of slices of the sample.)

1. A method for in vivo full field interferometric microscopy imaging of a scattering three-dimensional sample, the method comprising:

-placing the sample in an object arm (147) of an interference apparatus of a full-field OCT imaging system (130), wherein the interference apparatus further comprises a reference arm (146) having an optical lens (134) and a first reflective surface (133);

-generating, at each point of the imaging field, an interference between a reference wave obtained from the reflection of an incident light wave on a cell surface of the first reflecting surface (133) corresponding to said point of the imaging field and an object wave obtained from the backscatter of an incident light wave through a voxel of the sample at a given depth, said voxel corresponding to said point of the imaging field;

-acquiring a time series of two-dimensional interference signals caused by the interference generated at each point of the imaging field by using an acquisition device (138) of the full-field OCT imaging system (130);

-storing the acquisition time for each two-dimensional interference signal;

-providing a cross-sectional image (X-Z) of the sample and the first reflective surface (133) of the full-field OCT imaging system (130) using an OCT imaging system (110) at each acquisition time of the two-dimensional interference signal;

-determining a plurality of frontal images (X-Y) of a plurality of slices of said sample, each frontal image being derived from at least two-dimensional interference signals (I) with a given phase shift₁、I₂) Determining;

-from the OCT imaging system (110) at the two-dimensional interference signals (I)₁、I₂) Determining a depth (z) of each frontal image (X-Y) of the plurality of slices;

-determining a 3D image of the sample from the plurality of frontal images and depths of the plurality of slices of the sample.

2. The method of claim 1, wherein the full-field OCT imaging system (130) and the OCT imaging system (110) are mounted on a moving platform (150), the method further comprising moving the platform (150) at least along an optical axis (Z) of the object arm to determine the plurality of frontal images (X-Y).

3. The method of claim 1, further comprising moving the stage (150) at least along a direction (X, Y) perpendicular to the optical axis of the object arm.

4. The method of any of claims 2 or 3, wherein the reference arm (146) is mounted on a moving platform (131), the method further comprising moving the platform (131) to compensate for defocus.

5. The method of claim 1, wherein the object arm (147) is mounted on a moving platform (143), the method further comprising moving the platform (143) along an optical axis (Z) of the object arm to determine the plurality of frontal images (X-Y).

6. The method of claim 5, wherein the reference arm (146) is mounted on a moving platform (131), the method further comprising moving the platform (131) to compensate for defocus.

7. The method of claim 1, wherein the reference arm (146) is mounted on a moving platform (131), the method further comprising moving the platform (131) along an optical axis (X) of the reference arm to compensate for defocus to determine the plurality of frontal images (X-Y).

8. The method of any of the preceding claims, further comprising positionally offsetting the first reflective surface (133) of the reference arm of the full-field OCT imaging system (130) to interfere signals (I) in the at least two dimensions₁、I₂) Providing said phase shift.

9. Method according to any one of the preceding claims, further comprising selecting the at least two-dimensional interference signals (I) with the phase shift in the time sequence of the two-dimensional interference signals acquired by the acquisition device₁、I₂) Wherein the phase shift is caused by in vivo motion of the sample.

10. A system (101, 102) for in vivo full field interferometric microscopy imaging of a scattering three-dimensional sample, the system comprising:

-a full-field OCT imaging system (130) for providing a frontal image of the sample, wherein the full-field OCT system comprises:

an interference device comprising an object arm (147) intended to receive the sample and a reference arm (146) comprising an optical lens (134) and a first reflective surface (133), wherein

The object arm (147) and the reference arm (146) are separated by a beam splitter (135), and

wherein the interference device is adapted to produce, at each point of the imaging field when the sample is placed on an object arm of the interference device, interference between a reference wave obtained from reflection of incident light waves on a cell surface of the first reflective surface corresponding to the point of the imaging field and an object wave obtained from backscatter of incident light waves by voxels of a slice of the sample at a given depth, the voxels corresponding to the points of the imaging field,

an acquisition device (138), the acquisition device (138) being configured to acquire a two-dimensional interference signal (I) caused by interference generated at each point of the imaging field₁、I₂) The time series of (a) and (b),

-an OCT imaging system (110) for providing cross-sectional images of the sample and the first reflective surface (133) of the full-field OCT imaging system (130) at a same acquisition time of the two-dimensional interference signal;

-a processing unit (160), the processing unit (160) being configured to:

o determining the sampleA plurality of frontal images (X-Y) of a plurality of slices of the artefact, each frontal image being derived from at least two-dimensional interference signals (I) with a given phase shift₁、I₂) Determining;

o from the OCT imaging system (110) on the two-dimensional interference signals (I)₁、I₂) Determining a depth (z) of each frontal image (X-Y) of the plurality of slices;

determining a 3D image of the sample from the plurality of frontal images and depths of the plurality of slices of the sample.

11. The system as recited in claim 10, wherein the first reflective surface (133) of a reference arm of the full-field OCT imaging system (130) is positionally offset to provide the at least two-dimensional interference signal (I |)₁、I₂) The optical path difference therebetween.

12. The system according to either one of claims 10 and 11, wherein the processing unit (160) is further configured to:

o selecting the at least two-dimensional interference signal (I) with the given optical path difference in the time sequence of the two-dimensional interference signals acquired by the acquisition device₁、I₂) Wherein the optical path difference is caused by in vivo movement of the sample.

13. The system of any of claims 10 to 12, wherein the object arm (147) of the full-field OCT imaging system (130) further comprises an optical lens (142).

14. The system of any of claims 10 to 13, wherein the reference arm (146) and/or object arm of the full-field OCT imaging system (130) is movable relative to the beam splitter (135) of the interference device of the full-field OCT imaging system (130).

15. The system of any of claims 10 to 14, further comprising a moving platform (101), wherein the full-field OCT imaging system (130) and the OCT imaging system (110) are mounted on the moving platform (101).

16. The system of any of claims 10 to 15, wherein the OCT imaging system is a spectral domain OCT imaging system, a time domain OCT imaging system, or a swept source OCT imaging system.

Technical Field

The present description relates to methods and systems for in vivo full-field interferometric microscopy. The present description is applicable to in vivo imaging of randomly movable objects, and in particular to in vivo imaging of ocular tissue.

Background

Optical Coherence Tomography (OCT) has become a powerful imaging modality over its 25 years (see, for example, "Optical Coherence Tomography-Technology and Applications" -wolf' gating Drexler-James G. Fujimoto-instruments-Springer 2015). OCT is an interferometric technique that can be viewed as an "optical analogy" to ultrasound imaging. OCT has a wide range of applications, particularly in the biomedical fields of ophthalmology, dermatology, cardiovascular, gastroenterology, and the like.

In vivo tissue can move involuntarily, looking at history, which has been challenging for all OCT techniques. More precisely, motion causes misalignment, shifting and double artifacts in conventional scanning OCT images. The type of artifact is related to the OCT method, according to which not all image pixels are acquired simultaneously, but the sample is scanned point by point.

To avoid these motion artifacts in the image, this has led to the advancement of OCT technology to achieve higher imaging speeds, resulting in spectral domain OCT (SD-OCT) capable of imaging at speeds in excess of 300000A scans/s (one-dimensional distributions) (see, e.g., l.an et al, "High speed spectral domain optical coherence for reliable imaging at 500,000A-line continuous" -biological optical Express 2,2770 (2010)), and more recently swept source OCT (SS-OCT) (see, e.g., b.a. disposed et al, "ultra High speed 1050nm scan source/user domain OCT 2011reliable and anti-section continuous imaging 100,000to 400,000A scans continuous" (Optics 18,20029). However, even at such scanning speeds, OCT images cannot avoid in vivo motion artifacts.

Based on the same goal, i.e. in order to obtain images without Motion artifacts, some publications and patents propose software and hardware based Motion compensation schemes (see, for example, m.kraus et al, "Motion correction in optical coherence tomography volume on a per a-scan basis using orthogonal scan patterns", biological Optics Express 3,1182 (2012)). However, hardware-based solutions make the apparatus more complex and often bulky and expensive, while software-based solutions are sample-specific and motion-specific, which means that software-based solutions can only compensate for a few motions of a specific object.

One special case of OCT, known as full-field OCT (ffoct), uses a camera to acquire all image pixels simultaneously, without requiring point-by-point or line-by-line scanning, and therefore can avoid the above-mentioned artifacts. For example, Full-field OCT imaging techniques are described in the article "Full-field Optical Coherence Tomography" by F.Harms et al in the article "Optical Coherence Tomography-Technology and Applications" -pages 791-. French patent application FR 2817030 also describes a full-field OCT imaging technique.

Full-field OCT imaging techniques are based on the use of light backscattered by the sample when illuminated by a light source having a low coherence length, especially in the case of biological samples, by microscopic cells and tissue structures. This technique exploits the low coherence of the light source to isolate light backscattered by a virtual slice in the depth direction of the sample. The use of an interferometer makes it possible to generate, by interference phenomena, an interference signal representative of light selectively coming from a given slice of the sample and to eliminate light coming from the rest of the sample. More specifically, to obtain a single 2D FFOCT image, several (typically 2 to 5) direct images need to be acquired on the camera. Each of these direct images is acquired with a specific interference phase set by a precisely positioned mirror with a piezoelectric element (PZT) in the reference arm of the interferometer. These direct images with a specific phase are post-processed so that the FFOCT image can be recovered.

In addition to avoiding scanning artifacts as described above, FFOCT also provides higher lateral resolution than OCT by using a high Numerical Aperture (NA) objective since typical OCT uses a relatively low NA objective due to the need for a larger depth of field. Similar axial resolution can be provided by using inexpensive broadband spatially incoherent illumination sources.

However, current FFOCT 3D imaging schemes are feasible for static samples (or for in vivo samples with no or low motion transients), since any motion of the sample may shift the predetermined phase and degrade the FFOCT signal, or even corrupt the FFOCT image. The approach of 3D imaging is not suitable for in vivo imaging, since the position (X, Y, Z) of the captured 2D image becomes unknown, and thus a 3D image cannot be constructed. Thus, to date, the application of FFOCT has been almost entirely limited to static ex vivo samples.

The present description relates to devices and methods that have the advantages of full-field optical coherence tomography and at the same time can image constantly moving in-vivo objects.

Disclosure of Invention

According to a first aspect, the present description relates to a method for in vivo full field interferometric microscopy imaging of a scattering three-dimensional sample, the method comprising:

-placing the sample in an object arm of an interference apparatus of a full-field OCT imaging system, wherein the interference apparatus further comprises a reference arm having an optical lens and a first reflective surface;

-generating, at each point of the imaging field, an interference between a reference wave obtained from the reflection of an incident light wave on a cell surface of said first reflecting surface corresponding to said point of the imaging field and an object wave obtained from the backscatter of an incident light wave through a voxel of a slice of said sample at a given depth, said voxel corresponding to said point of the imaging field;

-acquiring a time series of two-dimensional interference signals caused by the interference generated at each point of the imaging field by using an acquisition device of the full-field OCT imaging system;

-storing the acquisition time for each two-dimensional interference signal;

-providing a cross-sectional image of the sample and the first reflective surface of the full-field OCT imaging system by using an OCT imaging system at each acquisition time of the two-dimensional interference signal;

-determining a plurality of frontal images of a plurality of slices of the sample, each frontal image being determined from at least two-dimensional interference signals with a given phase shift;

-determining a depth of each frontal image of the plurality of slices from a cross-sectional image provided by the OCT imaging system at an acquisition time of each of the two-dimensional interference signals;

-determining a 3D image of the sample from the plurality of frontal images and depths of the plurality of slices of the sample.

In this specification, a "front image" is an image determined in a plane ("X-Y" plane) perpendicular to the optical axis of the object arm (also referred to as a sample arm). In this specification, the "front image" is also referred to as an "X-Y image" or an "FFOCT signal".

The "cross-sectional image" is an image (1D or 2D) determined in a plane containing the optical axis of the object arm. In this specification, the cross-sectional image is also referred to as an "X-Z image". However, the cross-sectional image is not limited to a particular plane, and may be determined in any plane perpendicular to the "X-Y" plane.

The "optical lens" in the present specification refers to any optical device that focuses or disperses light by refraction of the light. Thus, "optical lens" includes both conventional optical lenses (convex, plano-convex, doublet, etc.) and other imaging systems (e.g., microscope objectives).

The imaging method thus described makes it possible to accurately determine the depth of a slice imaged by the FFOCT imaging system even in the case of imaging an in-vivo sample having natural motion. This can be achieved by simultaneously acquiring a two-dimensional interference image obtained using the FFOCT imaging system and a cross-sectional image provided by the OCT imaging system.

Thus, the natural motion of the object in vivo can be used for 3D imaging, which means that we can try to eliminate or overcome the effects with most methods.

In accordance with one or more embodiments, determining the depth of each frontal image of the plurality of slices of the sample comprises determining the relative axial position of the first reflective surface and at least one identified structure of the sample in a cross-sectional image provided by the OCT imaging system.

In practice, a plurality of frontal images of a plurality of slices of the sample may be determined within an investigation volume of the sample. The depth of the frontal image of the slice is determined from the OCT image from the difference between the detected axial position of the reference mirror peak and the axial position of any sample peak. It is not important which peak of the sample is used, but generally the brightest peak can be used. However, the same sample peak will be used throughout one volume acquisition so that the relative depth of the frontal slice is correct and a 3D image can be determined.

According to one or more embodiments, the full-field OCT imaging system and the OCT imaging system are mounted on a moving platform, the method further comprising moving the platform at least along an optical axis (Z) of the object arm to determine the plurality of frontal images.

According to one or more embodiments, the method further comprises moving the platform in at least one of directions (X, Y) perpendicular to an optical axis of the object arm. Thus, cross-sectional images can be stacked axially and laterally and larger 3D volumes can be formed (e.g., by image registration).

In accordance with one or more embodiments, the object arm is mounted on a moving platform, the method further comprising moving the platform along an optical axis of the object arm to determine the plurality of frontal images.

According to one or more embodiments, the natural in vivo movement of the sample is used to determine the plurality of frontal images. There is no need to move the object arm or the full-field OCT imaging system and any stage of the OCT imaging system.

According to one or more embodiments, the object arm further comprises an optical lens, such as a microscope objective, for example for corneal imaging. The depth of focus of such an optical lens is much smaller than the depth of focus of the eye. Thus, when the relative positions of the sample arm and the sample are changed, the method further comprises moving the reference arm along the optical axis of the reference arm to compensate for the defocusing, i.e. to keep the coherence plane within the depth of focus of the sample arm microscope objective. In practice, when moving from one medium to another (e.g. air and eyes), an offset occurs between the focal point and the position where the optical paths in the two arms are equalized. This defocus needs to be compensated.

According to one or more embodiments, the depth of focus is high, for example for retinal imaging, and there is no need to compensate for defocus as the relative position of the sample arm and sample is changed.

According to one or more embodiments, the method further comprises positionally offsetting the first reflective surface of a reference arm of the full-field OCT imaging system to provide the phase shift between the at least two-dimensional interference signals. These embodiments assume that the natural motion of the sample is slow during the acquisition time of the at least two-dimensional interferometric signals. Typical acquisition times are 1-10 ms.

According to one or more embodiments, the method further comprises selecting the at least two-dimensional interference signals with the phase shift in the time sequence of the two-dimensional interference signals acquired by the acquisition device, wherein the phase shift is caused by in vivo motion of the sample.

Again, the natural motion of the in vivo sample is used for frontal imaging, which means that we have utilized most methods to try to eliminate or overcome the effects.

The different embodiments of the imaging method according to the first aspect of the present description may be combined with each other.

According to a second aspect, the present description relates to a system for in vivo full-field interferometric microscopy imaging of a scattering three-dimensional sample, configured for implementing one or more embodiments of the method according to the first aspect.

According to one or more embodiments, the system according to the second aspect comprises:

-a full-field OCT imaging system for providing a frontal image of the sample, wherein the full-field OCT system comprises:

an interference device comprising an object arm intended to receive the sample and a reference arm comprising an optical lens and a first reflective surface, wherein the object arm and the reference arm are separated by a beam splitter, and wherein the interference device is adapted to produce, at each point of an imaging field when the sample is placed on the object arm of the interference device, interference between a reference wave obtained from reflection of an incident light wave on a cell surface of the first reflective surface corresponding to said point of the imaging field and an object wave obtained from backscatter of an incident light wave through a voxel of a slice of the sample at a given depth, said voxel corresponding to said point of the imaging field,

-an acquisition device configured to acquire a time series of two-dimensional interference signals resulting from interference produced at each point of the imaging field,

-an OCT imaging system for providing cross-sectional images of the sample and the first reflective surface of the full-field OCT imaging system at the same acquisition time of the two-dimensional interference signal;

-a processing unit configured to:

determining a plurality of frontal images of a plurality of slices of the sample, each frontal image being determined from at least two-dimensional interference signals with a given phase shift;

determining a depth of each frontal image of the plurality of slices from a cross-sectional image provided by the OCT imaging system at an acquisition time of each of the two-dimensional interference signals;

determine a 3D image of the sample from the plurality of frontal images and depths of the plurality of slices of the sample.

The advantages stated for the imaging method may be transferred to the imaging system according to the second aspect of the present description.

In accordance with one or more embodiments, the first reflective surface of a reference arm of the full-field OCT imaging system is offset in position to provide the optical path difference between the at least two-dimensional interference signals.

According to one or more embodiments, the first reflective surface of the reference arm of the full-field OCT imaging system is fixed and the processing unit is further configured to select the at least two-dimensional interference signals having the given optical path difference in the time sequence of the two-dimensional interference signals acquired by the acquisition device, wherein the optical path difference is caused by in-vivo motion of the sample.

According to one or more embodiments, the object arm of the full-field OCT imaging system further includes an optical lens.

According to one or more embodiments, the optical lens of the reference arm and/or object arm is a microscope objective.

According to one or more embodiments, the reference arm and/or object arm of the full-field OCT imaging system may be moved (along the respective optical axes of the reference arm and object arm) relative to the beam splitter of the interference device of the full-field OCT imaging system.

In accordance with one or more embodiments, the system further comprises a moving platform, wherein the full-field OCT imaging system and the OCT imaging system are mounted on the moving platform.

According to one or more embodiments, the OCT imaging system is a spectral domain OCT imaging system or a swept source OCT imaging system or a time domain OCT imaging system.

The different embodiments of the imaging system according to the present description may be combined with each other.

The different features and embodiments of the various aspects of the present description may also be combined with each other.

Drawings

Other advantages and technical features presented above will become apparent from a reading of the following detailed description and a review of the accompanying drawings.

Figures 1A and 1B are schematic diagrams of a system according to embodiments of the present description;

figures 1C and 1D show exemplary light source spectra of an OCT source and an FFOCT source according to embodiments of the present description and blocking portions of such spectra treated with filters of the system;

fig. 2A, 2B are a flow chart of embodiments of an imaging method according to the present description and images to illustrate the steps of some of these embodiments;

fig. 3A, 3B are a flow chart of further embodiments of the imaging method according to the present description and images to illustrate the steps of some of these embodiments;

figure 4A is an example of a cross-sectional image (without visible sample) of a reference mirror obtained using an exemplary OCT imaging system of an imaging system according to the present description, and figure 4B represents a graph of intensity as a function of vertical line of figure 4A;

figure 5A is an example of a cross-sectional image of a reference mirror (with a visible corneal sample) obtained using an exemplary OCT imaging system of an imaging system according to the present description, and figure 5B represents a plot of intensity as a function of the vertical line of figure 5A;

FIG. 6 is a diagram showing the required phase shift by the in-vivo motion;

fig. 7 shows FFOCT images of the deep layers of the human cornea in vivo acquired using phase shift by natural eye movement (reference mirror motionless).

Fig. 8 shows FFOCT images of the deep layers of the human cornea (stroma) in vivo acquired with phase shift by natural eye movement (reference mirror motionless) at different camera exposure times.

Detailed Description

System for controlling a power supply

Fig. 1A and 1B show two embodiments 101, 102, respectively, of a system for performing in vivo full-field interferometric microscopy according to the present description. The system 101 is adapted to carry out a method for 3D imaging of a sample moving in vivo, in particular, but not limited to, the anterior portion 11 (cornea) of an eye in vivo. The system 102 is suitable for implementing a method for 3D imaging of a sample moving in vivo, in particular, but not limited to, the posterior portion 13 (retina) of an eye in vivo.

The system 101 shown in fig. 1A includes two imaging systems (a full-field OCT ("FFOCT") imaging system 130 and an optical coherence tomography ("OCT") imaging system 110) and at least one processing unit 160. The FFOCT imaging system is capable of acquiring a "frontal" image of a moving in vivo sample 11, i.e., an image of a deep section of the sample, and the optical coherence tomography ("OCT") imaging system 110 provides information about the position of the sample in the axial (Z) direction (e.g., along the optical axis). System 101 may also include a moving platform 150, e.g., having one or several motors, on which moving platform 150 FFOCT imaging system 130 and OCT imaging system 110 are mounted. The moving platform 150 enables the FFOCT imaging system and the OCT imaging system to be translated together in all X, Y and Z orthogonal directions.

The FFOCT imaging system 130 of fig. 1A includes an interference device 145 and an acquisition device 138 connected to the at least one processing unit 160.

According to one embodiment, interference device 145 includes a beam splitter element 135 (e.g., a non-polarizing beam splitter cube) such that two arms can be formed, having an optical axis Δ_RAnd an object arm 147 having an optical axis delta. In FIG. 1A, the optical axis Δ of the object arm defines the Z-axis, while the optical axis Δ of the reference arm_RThe X-axis is defined. The reference arm 146 includes a reflective surface 133. The reflective surface 133 may be flat. The reflective surface 133 is, for example, a metal mirror, Neutral Density (ND) filter glass, or a simple glass plate. The object arm 147 is intended to receive a three-dimensional scattering sample 11 in operation, for which it is desired to generate tomographic images of the volume of the three-dimensional scattering sample 11.

In the embodiment of FIG. 1A, the reflective surface 133 is mounted on a piezoelectric stage (PZT)132 for phase modulation. As will be further described, such phase modulation may be used in one embodiment of a method according to the present description, and may not be used in another embodiment.

The interference device is suitable for generating optical interference between a reference wave, obtained on the one hand by reflecting light emitted by the light source 141 (spatially incoherent or with a low coherence length) by each cell surface of the reflecting surface 133 of the reference arm 146, and an object wave, obtained on the other hand by backscattering light emitted by the same light source by each voxel of a slice of the sample 11 in the depth direction of the sample, the sample 11 being placed on the object arm 147, said voxel and said cell surface corresponding to the same point of the imaging field.

The light source 141 is a light source that is spatially incoherent and has a short temporal coherence length (in practice, in the range of 1 to 20 microns), such as a thermal light source (e.g. a halogen lamp) or an LED. According to one or more exemplary embodiments, as in the example of fig. 1A, the light source 141 may form part of the FFOCT imaging system 130, or may be an element external to the imaging system, with the FFOCT imaging system 130 configured to cooperate with the light waves emitted by the light source. The optical system 140 may be used to achieve kohler-like illumination. In operation, light emitted by light source 141 is reflected by dichroic mirror 139 and reaches beam splitter element 135 of interference device 145.

The acquisition device 138 makes it possible to acquire at least one two-dimensional interference signal resulting from the interference between the reference wave and the object wave.

The acquisition device 138 is for example an image sensor of the CCD (charge coupled device) or CMOS (complementary metal oxide semiconductor) camera type. The acquisition device is capable of acquiring images at high rates, for example, at frequencies between 100Hz and 1000Hz or higher. Depending on the dynamics of the sample under investigation, more specifically the dynamics of the motion within the sample, a camera operating between a few hertz to a few kilohertz may be used.

The processing unit 160 is configured to perform at least one step of processing the at least one two-dimensional interference signal acquired by the acquisition device 138 and/or at least one step of image generation according to at least one imaging method according to the present description, to generate at least one image of the sample slice.

In one embodiment, the processing unit 160 is a computing device comprising a first memory CM1 (not shown) for storing digital images, a second memory CM2 (not shown) for storing program instructions, and a data processor capable of executing the program instructions stored in the second memory CM2, in particular for controlling the execution of at least one step of processing at least one two-dimensional interference signal acquired by the acquisition device 138 and/or at least one step of image calculation according to at least one imaging method according to the present description.

The processing unit may also be produced in the form of an integrated circuit comprising electronic components adapted to perform one or more of the functions described herein for the processing unit. Processing unit 160 may also be implemented by one or more physically distinct devices.

In the example of fig. 1A, the interference device is a linnic interferometer and includes two optical lenses 134, 142 (e.g., microscope objectives) disposed on each of the reference and object arms. MicroscopeThe objective lenses 134, 142 may have a relatively high numerical aperture (typically, a high numerical aperture)) While providing a relatively large field of view (typically of). Thus, the reflective surface 133 is located at the focus of the objective 134 of the reference arm and is intended to position the sample 11 at the focus of the objective 142 of the object arm. More specifically, the layer of interest in the sample is intended to be positioned at the focal point of the objective lens 142. It is contemplated that other types of interferometers, particularly but not limited to michelson interferometers, may be used in connection with practicing methods according to the present description.

In the example of fig. 1A, the microscope objective 142 of the object arm 147 is mounted on a motorized stage 143, which stage 143 is movable in the direction of the optical axis of the object arm (Z axis), i.e. closer to or further away from the sample 11. The reflecting surface 133 of the reference arm 146 and the microscope objective 134 are both mounted on a further motorized stage 131, which stage 131 is movable in the direction of the optical axis (X-axis) of the reference arm.

At the output of interferometer 145, there may be a spectral filter 136 and an optical lens 137 (e.g., an achromatic doublet), the focal length of which optical lens 137 is adapted so that the collection device 138 can properly sample the sample 11, and which optical lens 137 can conjugate the plane located at the focal points of the two objective lenses with the detection surface of the collection device 138. Thus, the acquisition device 138 acquires the interference signal generated by the interference device. In order not to limit the resolution allowed by the microscope objectives 134 and 142, the focal length of the optics 137 will be chosen in accordance with the shannon sampling standard. The focal length of the optics 137 is, for example, a few hundred millimeters, typically 300 mm.

As described further below, advantageously, the spectral filter 136 transmits the wavelength of the light source 141 while blocking the wavelength of the OCT source 112.

A glass plate or a so-called dispersion compensation block (not shown in fig. 1A) may be provided on each arm to compensate for dispersion.

OCT imaging system 110 comprises a spatially coherent light source 112, a detector 113, and an interference device with a beam splitter element 114, which beam splitter element 114 defines a reference arm and an object arm of the interference device of the OCT imaging system. In general, the spatially coherent light source 112 may be a superluminescent diode (SLD) (e.g., in the case of spectral domain OCT or time domain OCT) or a swept laser source. In general, the detector 113 may be a device that converts incident optical power directly into an electrical signal, such as a photodiode in the case of time-domain OCT or swept-source OCT, or a spectrometer in the case of spectral-domain OCT.

Light from the light source 112 is collimated into an optical fiber 118 and split by a beam splitter element 114 into two optical fibers 121 (object arm) and 120 (reference arm). In operation, light, after passing through the optical fiber 120, passes through the lens 115, the rotatable dispersion compensation plate 116, and reaches the reflective surface 117 (e.g., a metalized mirror). After passing through the optical fiber 121, the light reaches the transverse scanning mechanism 111, which can scan the light beam in the 2D (X-Y) direction. The beam then passes through filter 122, through dichroic mirror 139, and split by beam splitter 135 into FFOCT reference arm 146 and FFOCT sample arm 147.

Filter 122 is selected so that the beam emitted from OCT source 112 propagates in all of the OCT reference arm, FFOCT reference arm, and FFOCT sample arm, but blocks light from FFOCT source 141. On the other hand, filter 136 blocks the beam emitted from OCT source 112 and passes light from the FFOCT source.

The function of filters 122 and 136 is further described in conjunction with fig. 1C and 1D. In the example shown in FIG. 1C, filter 122 may block (dashed line) OCT light source 112 below a given wavelength λ_Filter122Of the given wavelength λ_Filter122Greater than the highest wavelength λ used in FFOCT imaging systems_FFOCTmax. On the other hand, as shown in FIG. 1D, a filter 136 in the FFOCT imaging system can block more than a given wavelength λ in the OCT light source_Filter136Of the given wavelength λ_Filter136Greater than the highest wavelength λ used in FFOCT imaging systems_FFOCTmax. Therefore, no OCT light reaches the acquisition device 138.

It is clear that fig. 1C and 1D only represent examples of the function of the filters 122, 136. Many other configurations are also available as long as no OCT light reaches the acquisition device 138.

In a preliminary step, the optical path from beam splitter 114 to the OCT arm (reference arm) of mirror 117 may be matched to the optical path from beam splitter 114 to mirror 133 in FFOCT reference arm 146. The optical path matching of the OCT and FFOCT reference arms can be achieved in a simple manner. And (5) viewing the OCT image in real time. The reference arms of the OCT and FFOCT systems do not match if the mirrors of the FOCCT reference arm are not visible on the OCT image. We extended the reference arm of the OCT imaging system until the mirror of the foct reference arm was visible on the OCT image.

In operation, back-reflected light from the reflective surface 133 in the reference arm 146 of the FFOCT imaging system is combined with back-reflected light from different layers of the sample at the beam splitter 135. The beam splitter 135 splits the light into two parts again: the reflected portion is blocked by the filter 136 (as described in connection with fig. 1D) and the transmitted portion passes through the dichroic mirror 139, the filter 122, the optical fiber 121. This beam is then mixed with the back-reflected light from fiber 120 and collected by detector 113 after passing through fiber 119. The detector 113, e.g. a spectrometer, is configured to record a so-called a-scan (1D profile) containing information about the reflectivity at different depths of the object being imaged. Further, the detector 113 collects information about the position of the reference mirror 133 of the reference arm 146 of the FFOCT imaging system. By scanning the light beam with the scanning mechanism 111, 2D and 3D reflectivity images can be acquired.

The OCT imaging system may be spectral domain OCT (detector 113 is a spectrometer), but may also be time domain OCT or swept source OCT.

OCT imaging systems can also provide information about the velocity of the sample based on several successive positions of the sample and the time intervals between them. Information about the instantaneous speed of the sample may be helpful in predicting its future motion (e.g., if the sample is moving rapidly in the Z direction at a first time, we can expect it to continue moving in the same direction at the next time).

As will be explained further below, embodiments of methods according to the present description use the OCT imaging system described above to obtain information about the location of different layers of interest of the sample 11 and the location of the reference mirror 133 on the reference arm 146 of the FFOCT imaging system.

The system 102 shown in FIG. 1B is similar to the system of FIG. 1A with minor differences. In the embodiment of fig. 1B, the action of the microscope objective in the sample arm 147 of the FFOCT imaging system is performed by the cornea 11 and the crystalline lens 12 of the eye. In addition, an adaptive lens 148, such as a liquid lens, and a rotating glass plate 149 may be inserted into the reference arm to compensate for aberrations and dispersion mismatch introduced by the eye. Since the lens 12 of the eye generally has a large depth of focus, imaging of retinal layers at different depths can be performed without correcting for defocus, and therefore, in contrast to the device shown in fig. 1A, there is no need to move the reference arm 146. In all other respects, the system may be similar to that of the embodiment shown in FIG. 1A.

3D imaging method

Fig. 2A is a flow chart of an embodiment of an imaging method according to the present description. For example, the process may be implemented using a system as shown in FIG. 1A. The described method is suitable for 3D imaging of in vivo moving samples, in particular, but not limited to, the anterior part 11 (cornea) of the eye in vivo.

Figure 2B shows images acquired by both the OCT imaging system and the FFOCT imaging system during the different steps shown in figure 2A.

The steps of fig. 2A, 2B show the acquisition of a 3D image.

In step 201, images from both OCT imaging system and FFOCT imaging system are obtained and displayed. As will be further described, FFOCT images can be obtained using modulated PZT or static PZT. In corresponding step 201 of fig. 2B, OCT image 221 shows mirror 133 of the reference arm of the FFOCT imaging system. On images 222 and 228, there is only camera noise because defocus correction has not been performed and/or the optical paths of the sample and reference arms do not match.

In step 203, it is checked whether the corneal layer is visible in the OCT image.

If NO, as shown in image 226 of fig. 2B, the entire device 150 may be moved along X, Y and the Z-axis (step 204). For example, it may be moved by an operator until the desired layer is observed on the screen.

If YES is shown as image 227 of FIG. 2B, the OCT image of the reference mirror 133 can be superimposed with the corneal layer (reference corneal layer) by moving the entire apparatus 150 along the X and Z axes (step 205). The reference corneal layer may be any layer, although a layer that provides a bright peak may generally be selected. In FIG. 2B, image 229 shows the reference mirror image superimposed with the image of the reference corneal layer. At the same time, a blurred FFOCT image 230 can be seen due to the lack of defocus correction.

The FFOCT image contains only no information about the position in the sample where the image was taken. The OCT imaging system 110 used in conjunction with FFOCT compensates for this lack by providing X, Y, Z coordinates of the captured image. The 2D FFOCT image piles, each with its position, can be grouped to form a 3D image. The method of 3D image acquisition 209 will be described more accurately below.

In the first embodiment (210), the microscope objective 142 is moved only by the motor under the sample arm 147. At the same time, reference arm 146 is moved away from (or closer to) beam splitter 135 to compensate for the optical path mismatch between sample arm 147 and reference arm 146.

In the second embodiment (211), the entire apparatus 150 is moved close to (or away from) the sample 11 by the motor 101. At the same time, reference arm 146 is moved away from (or closer to) beam splitter 135 to compensate for the optical path mismatch between sample arm 147 and reference arm 146.

In the third embodiment (212), only the reference arm 146 is moved away from the beam splitter 135 to compensate for the optical mismatch (defocus) between the sample arm 147 and the reference arm 146. The range of movement of the reference arm depends on the instantaneous sample position (or depth in the sample). Changes in sample position (or depth) are only dominated by sample motion within the body.

In all embodiments, respective frontal images of slices at different depths in the sample are recorded according to the method described below. At the same time, the location (X, Y, Z) of the slice corresponding to each 2D image is recorded by the OCT imaging system 110. With the location information of each 2D image, the 2D image can be repositioned to form a 3D image.

The depth of each slice acquired to the frontal image is determined by storing (213) the times at which those images were acquired. Acquisition is stopped when needed (214). In step 215, we use the stored position information from the OCT images at different times to realign the 2D OCM image (i.e., the image obtained by the FFOCT device) and form a 3D corneal image.

Examples of a 2D cross-sectional image and a 1D image for position detection (a-scan) are shown in fig. 4A, 4B and 5A, 5B.

Fig. 4A, 4B, 5A, 5B illustrate how an OCT imaging system is used to determine the depth of a slice imaged by an FFOCT imaging device. Compensation for defocus is also shown.

Fig. 4A and 4B show a 2D cross-sectional image and a corresponding profile (a-scan), respectively, obtained using an OCT imaging system when no sample is introduced. Fig. 5A and 5B show a 2D cross-sectional image when a sample is introduced into the sample arm and within the field of view of the OCT imaging device and the corresponding profile obtained using the OCT imaging system, respectively. The depth of the slice is measured relative to the reference position that has not been defocus corrected (corresponding to the black vertical line at depth 0 in fig. 4A, 4B, 5A, 5B). For example, the reference layer is the top layer of the cornea.

On image 231, the topmost layer of the cornea (cornea shown in parentheses) overlaps the reference mirror (shown by arrow). This position corresponds to the "0" position in fig. 4A, 5A, i.e. the case where the difference between the position of the top corneal layer and the position of the reference mirror which has not been corrected for defocus is equal to zero. As a result, without applying defocus correction, we obtain an image 232 of the corneal surface.

On image 233, the top layer of the cornea is shifted up (on the image) relative to the reference arm position that is not defocus corrected. Thus, a non-zero depth is measured. Based on this depth, the reference arm is moved down (on the image) from the reference position that is not defocus corrected. As a result, we obtain an image 234 from the corneal layer that overlaps the reference mirror image in the OCT image.

On the image 235, all operations are repeated as in the previous steps. The cornea is again moved up and the reference arm with mirror is again moved down, resulting in us being provided with FFOCT images 236 from the deep layers of the cornea.

The above-described embodiments are presented for imaging samples moving in vivo, and in particular, for imaging the anterior portion of the eye in vivo.

The embodiments of the method described below can also be used to image a variety of in vivo samples, but of particular interest is imaging the posterior portion of the eye in vivo. Fig. 3A is a flow chart of an embodiment of an imaging method according to the present description. For example, the flow may be implemented using a system as shown in FIG. 1B. The described method is suitable for imaging of moving samples in vivo, particularly, but not exclusively, the posterior part 13 (retina) of the eye in vivo.

Figure 3B shows images acquired by both the OCT imaging system and the FFOCT imaging system during the different steps shown in figure 3A.

In step 301, acquisition begins (acquisition includes processing performed to acquire an image) and images are displayed from both the OCT imaging system and the FFOCT imaging system. FFOCT acquisition can be performed using modulated PZT or static PZT, as described below. In step 303, it is checked whether the retinal layer is visible in the OCT image. On images 322 and 326, there is only camera noise because defocus correction has not been performed and/or the optical paths of the sample and reference arms do not match.

If NO is shown as image 324 of FIG. 3B, the entire device 150 may be moved along X, Y and the Z axis (step 304). If YES is shown as image 325 of FIG. 3B, the OCT image of the reference mirror 133 can be superimposed with the retinal layer of interest by moving the entire device 150 along X, Y and the Z axis (step 305).

At that stage, the optical path is matched between the mirror 133 and any retinal layers, as shown in OCT images 327 or 329 and corresponding FFOCT images 328, 330 of figure 3B.

Then, 3D image acquisition 309 is started.

In the first embodiment (310), only the reference arm 146 is moved by the lower motor.

In the second embodiment (311), the entire apparatus 150 is moved close to (or away from) the sample 11 by the motor 101.

In a third embodiment, the motor is not moved and the creation of the 3D stack is achieved by in vivo movement of the sample.

For fig. 2A, 2B, individual 2D cross-sectional images of the in vivo sample while in motion were recorded according to the method described above. If in vivo movement of the sample is used, the device can be adjusted to obtain the maximum FFOCT signal with respect to the frequency of typical sample movement. At the same time, the sample location (X, Y, Z) corresponding to each 2D image is recorded by OCT imaging system 110 (steps 313, 314, 315 and corresponding image 331-336 on FIG. 3B). By having the location information of each 2D image, the 2D image can be repositioned to form a 3D image.

Determination of a front image

In order to extract the FFOCT image from the direct camera image, a phase shift scheme is required.

In a first embodiment of the present description, a standard FFOCT image recovery method is used, according to which a phase shift is provided by modulating a piezoelectric element (PZT) 132. This embodiment is useful for the case of slowly moving samples (their movement during typical image acquisition should be < < pi phase shift) or for fast moving samples at moments of no movement. FFOCT images can be extracted from 2, 4 or 5 direct images according to this scheme.

For example, for 2 direct images:

wherein:

-phi is the phase difference between the sample signal and the reference signal;

phi is the phase shift caused by PZT

-I₀Is the photon flux of the illumination;

-R_ref(x, y) ≈ const is the reference reflectivity, which is spatially uniform;

-R_sam(x, y) is the reflectivity of the sample structure (i.e., the plane of interest) within the coherent volume;

-R_inc(x, y) is the reflectivity of all other structures outside the coherence volume and other spurious reflections.

The two phase shifted images are:

by subtracting the two images and taking the modulus, we get the FFOCT image or "FFOCT signal".

Making the phase shift between two consecutive direct camera frames equal to pi (in a 2-phase shift scheme) allows to acquire as high an FFOCT signal as possible.

In a second embodiment of the present specification, the image recovery method used relies on the natural in vivo movement of the sample.

Applicants have shown that, for example in ophthalmic tissue imaging applications, natural eye movements can introduce phase variations between successive direct images that can be large enough to extract FFOCT images. More precisely, the applicant has measured the motion of the human eye in vivo and has shown that, when the exposure time of the camera is set (for example, ranging from 1ms to 10ms (i.e. two consecutive camera frames are acquired within 2-20 ms respectively)), the phase shift between consecutive camera frames introduced by the eye motion can take any value between 0 and ± 30 radians (or equivalently, ± 10 π). More generally, in vivo motion may cause phase changes between successive direct camera images. These phase changes can be used to extract FFOCT images. According to this method, the FFOCT image can be extracted from 2, 4 or 5 direct images, depending on the scheme, but is not limited to this sequence. In the following, we will give an example of the FFOCT extraction method for 2 direct images, but the present invention is not limited to the 2-image scheme, but is applicable to each FFOCT image restoration scheme.

The interference phases of the sample beam and the reference beam are changed by a random amount ψ when the sample is moved in the Z direction. During the process of capturing an image by the camera, different phase shifts may occur. In the simplest case, it can be said that each camera image has an average phase < ψ >. Thus, the direct image signal recorded on the camera is given by:

wherein:

-phi is the phase difference between the sample signal and the reference signal;

- < ψ > is a random phase shift caused by natural motion of the sample in vivo, and an average value was taken over the acquisition time.

-I₀Is the photon flux of the illumination;

-R_ref(x, y) ≈ const is the reference reflectivity, which is spatially uniform;

-R_sam(x, y) is the reflectivity of the sample structure (i.e., the plane of interest) within the coherent volume;

-R_inc(x, y) is the reflectivity of all other structures outside the coherence volume and other spurious reflections.

Thus, the two direct images are:

by subtracting the two images and simplifying the formula, we get:

as can be seen from the formula, an FFOCT image can be acquired for each average phase difference < ψ > of two consecutive or more distant camera frames, but a maximum FFOCT signal can be obtained for < ψ > -pi (consider Φ -0).

In fig. 6, it is shown that in order to obtain a high FFOCT signal, the phase shift not only needs to be large enough, but should also occur during the acquisition of several frames by the camera (an example of 2 frames is shown in fig. 6). By knowing the instantaneous velocity of the motion in the sample body, the time interval required to achieve an average pi phase shift can be found. A direct image pile can be recorded and two frames corresponding to a pi phase shift can be extracted from the pile and processed to obtain a high FFOCT signal. For example, the method may be used for eye imaging: when the eye's motion is such that the phase shift induced between two consecutive direct camera images is less than 1 radian (usually out of a large spike similar to a heartbeat), the phase shift between the two images is sufficient to obtain an FFOCT image. In addition, to increase the number of useful (phase shifted) direct images in the camera, the acquisition speed of the camera and the wavelength of the light source can be adjusted according to the typical speed of the in vivo sample movement. Sample speed at which maximum FFOCT signal is achieved:

wherein:

λ is the wavelength of the FFOCT light source.

T is the time required for the camera to capture two direct images

From the equation, we can adjust the camera speed and the wavelength of the light source by initially knowing the typical velocity v of the motion in the sample to obtain an average pi phase difference between the direct images (and thus an optimal FFOCT signal) at the typical velocity of the sample. The two-phase image of standard FFOCT can be used when the eye's motion causes a phase shift between two successive images of less than 1 arc (typically arising from a large spike similar to the heartbeat).

Previously, for simplicity, consider that each camera image has an average phase<ψ>. By taking into account the phase ψ (T) at each moment and taking into account the passage of the camera during the exposure time (e.g. from the moment T)₀To time T₁) Integrating the light to acquire an image allows for more comprehensive analysis.

The two successive direct images are therefore:

by subtracting the two images and simplifying the formula, we get:

the applicant has measured the equation ψ (t) of the human eye in vivo and demonstrated that a higher FFOCT signal can be achieved for different camera exposure times (e.g. 1 ms-10 ms).

In the example of fig. 6, to obtain an FFOCT image, we need to acquire at least two direct images with an average phase shift between them equal to pi (Y-axis on the graph). If we know the average moving speed of the sample, we know the time required for the sample to move by π (e.g., 3.4ms, as shown in the figure). Therefore, we can adjust the acquisition speed of the camera to acquire two images in 3.4 ms. As a result, the phase shift can be performed only by the sample.

Fig. 7 shows an image of a deep layer of the cornea acquired using a continuous two-dimensional interference signal that is phase shifted only by natural eye movement (i.e., without moving the reference mirror). The LED emits light with a wavelength of 850 nm. The spectrum was 30nm wide, resulting in an optical slice thickness of 7.8 μm. The camera is set to acquire 550 direct images per second. Each image was acquired by integrating the light over an exposure time of 1.75 ms. During exposure, the sample is moving and changes the optical phase of the interference signal. By subtracting two consecutive direct images from the camera, we subtract the two integrals of the time-varying phase from the above equation and obtain the FFOCT image. The FFOCT signal depends on the slave time T₀To T₂Function ψ (t).

In fig. 7, images 71 to 76 show reflections from the epithelium and tear film (71), the epithelium and sub-basal nerves (72), the anterior, medial and posterior stroma (73-75) and the endothelium (76), respectively, of a human cornea in vivo.

Fig. 8 shows images 81, 82, 83, 84 of the corneal deep stromal layer acquired using different camera exposure times (and hence different camera frame rates, 550 frames/sec, 300 frames/sec, 200 frames/sec and 100 frames/sec, respectively). These images were taken under the same conditions as in fig. 7, i.e. using a continuous two-dimensional interference signal that was phase shifted only by natural eye movement (i.e. without moving the reference mirror).

The applicant has shown that such an embodiment makes it possible to acquire images of very high quality and greatly simplifies the system without requiring camera-piezo synchronization.

While described with several detailed exemplary embodiments, the systems and methods for in vivo full field interferometric microscopy imaging of a three-dimensional sample of scatter according to the present description include variations, modifications, and improvements which will be apparent to those skilled in the art. It is to be understood that such various alterations, modifications and improvements are within the scope of the invention, as defined by the appended claims.