Control system for OCT imaging, OCT imaging system, and method for OCT imaging

Abstract

The present invention relates to a control system (130) for controlling an optical coherence tomography imaging means for imaging an object (190), the control system being configured to perform the following steps: receiving scan data (122) from the object (190) acquired using optical coherence tomography; performing data processing on the scan data (122); and obtaining image data (142) for an image (144) of the object, the processing system (130) being further configured to adapt at least one parameter of the OCT imaging means based on a change in a value characterizing an axial position (z) of the object (190) relative to the OCT imaging means between two sets of image data. The present invention also relates to a processing system, an OCT imaging system (100), and corresponding methods.

Description

Technical Field

[0001] The present invention relates substantially to a control system, a processing system, an OCT imaging system including such a control system, and a method for imaging an object for optical coherence tomography (OCT) imaging means for imaging an object.

Background Art

[0002] Optical coherence tomography (hereinafter also referred to by its typical abbreviation OCT) is an imaging technique that uses low-coherence light to capture high-resolution secondary and three-dimensional images from within a light-scattering medium (e.g., biological tissue), which is used especially for imaging in the medical field. Optical coherence tomography is based on the low-coherence interference method, and typically near-infrared light is used. By using light of a relatively long wavelength, the light can penetrate into the scattering medium. A particular medical field of interest in OCT is ophthalmology, which is a branch of medicine concerned with the (especially human) eye and its disorders and related surgeries.

Summary of the Invention

Means for Solving the Problems

[0003] According to the present invention, a control system, a processing system, an OCT imaging system, and a method for imaging an object are proposed, which have the features of the independent claims. Preferred further developments form the subject matter of the dependent claims and the following description.

[0004] The present invention relates to a control system for optical coherence tomography (OCT) imaging means for imaging an object, especially for real-time imaging of an object. The object preferably includes or is an eye. The type of OCT used is preferably spectral domain OCT (also known as Fourier domain OCT), as will also be explained later.

[0005] Spectral or Fourier-region OCT can be based on broadband light sources and spectrometer systems (e.g., those equipped with diffraction gratings or other dispersion detectors), but wavelength-swept light source OCT (SS-OCT) (i.e., spectral scanning systems), where the frequency of light changes over time, can also be used.

[0006] The control system is configured to control optical coherence tomography (OCT) imaging means that scan an object using optical coherence tomography to acquire scanning data or a scanning dataset. Data processing is performed on the scanning data to acquire image data about the object, and includes, for example, DC or baseline removal, spectral filtering, wavenumber resampling, dispersion correction, Fourier transform, scaling, image filtering, and optionally additional image enhancement steps. Typically, a set or frame of image data (particularly in the sense of a two-dimensional image in live or real-time imaging) is combined into a two-dimensional OCT image, B scan. Thus, the underlying scanning data includes multiple spectral or A scans.

[0007] In OCT, areas of a sample (object) or tissue that reflect a lot of light cause greater interference than areas that do not reflect light. Any light outside a short coherence length does not interfere. This reflectance profile is called the A scan and contains information about the spatial dimensions and location of structural parts within the sample or tissue. A cross-sectional tomography called a B scan can be achieved by a lateral combination of these series of axial depth scans (A scans). This B scan can then be used to create a two-dimensional OCT image to be viewed.

[0008] In particular, Fourier-domain optical coherence tomography (FD-OCT) uses the principle of low-coherence interferometry to generate two-dimensional or three-dimensional images of the object (sample). Light from a light source is split between a reference arm and a sample arm. The signal pattern in the detector is composed of a reference light spectrum and is generated by modulation through the interference of light between the reference arm and the sample arm.

[0009] Typically, a two-dimensional image with x and z dimensions (where x is the transverse direction and z is the axial or depth direction, i.e., the direction of the OCT light beam incident on the object) is displayed to the end user in real time. The detection position along the z dimension (or axial direction) in space depends on the difference in optical path length between the sample (object) and the reference beam. When imaging a living sample, the movement of the target sample along the axis of the probe beam will consequently move the generated OCT image up and down along the z axis (axial direction). This can result in a blurred image with loss of image quality, and may even cause the image to move completely out of the field of view.

[0010] To stabilize images and maintain image quality, the following techniques are proposed within this invention. The control system is configured to adapt at least one parameter of the OCT imaging means based on a change in a value characterizing the axial position of an object relative to the OCT imaging means between two sets of image data. To obtain such a value, the control system may be configured to determine a value (or the magnitude of such a value) characterizing the axial position of an object relative to the OCT imaging means. This value may, in particular, be an absolute or relative measurement of the axial or z-direction position measured at a pixel in the OCT image, for example. This allows tracking of the axial movement of the object.

[0011] Furthermore, the control system may be configured to receive such values ​​from an external processing unit. In this case, such an (external) processing unit is configured to perform the steps of determining a value characterizing the axial position of an object with respect to the OCT imaging means, and providing that value to the control system. This processing unit can be connected to the control system via a communication means such as Ethernet or the Internet. In particular, the processing unit can be formed by a server or a cloud computing system, which is of particular interest in special cases such as those described below. Alternatively, such a processing unit can be formed by a PC or other computer. In that case, the control system and the processing unit can be considered as a single processing system.

[0012] The values ​​can be obtained, in particular, by image processing methods or by image metrics, preferably using image metrics based on maximum and / or weighted average signal intensity. The use of machine learning methods is also preferred, such machine learning methods configured to identify parts of an object within an image dataset. A more detailed description of such methods is provided below. Of particular interest is the use of such external processing units in conjunction with machine learning methods (and also with image processing methods or metrics) due to the high computational power they can thus provide. Nevertheless, such methods can also be implemented on control systems depending on the specific circumstances.

[0013] Another possibility for obtaining such a value characterizing the axial position of an object with respect to OCT imaging is to use an external probe beam from a laser, such as a laser rangefinder (which may be based on the time-of-flight principle).

[0014] As described above, the control system is configured to adapt at least one parameter of the OCT imaging means based on the change in value between two sets of image data. Each set of image data corresponds, in particular, to a frame of an (OCT) image during live or real-time imaging. These parameters are preferably the focal position of the sample arm of the OCT imaging means, the axial position of the reference arm mirror (in the reference arm of the OCT imaging means), and the axial position of the objective lens of the sample arm (of the OCT imaging means). Note that it is usually sufficient to select either the axial position of the reference arm reflector (or mirror) or the axial position of the objective lens of the sample arm, as they typically have the same effect.

[0015] In this way, tracking the axial movement of the sample (object) can be used in combination with dynamically adjusting the optical focus and reference arm position, thereby stabilizing the image and maintaining image quality.

[0016] To adjust at least one parameter of the OCT means, the control unit may be configured to provide a signal for a corresponding adaptation means configured to mechanically and / or digitally adapt at least one parameter of the OCT imaging means upon reception of a signal. Such means may include a reference arm controller configured to change the axial position of a reference arm reflector. Similar means may also be provided on the sample arm, particularly the objective lens, for adjustment of its focus or focal position. In this case, such means may be part of the microscope used for OCT, including the objective lens. Such adaptation means may be provided in the OCT imaging system in any case, or added as needed. Of course, each OCT imaging means (reference arm, objective lens, etc.) must be capable of its respective adaptation.

[0017] By using metrics derived from images in real time, the depth position of the sample can be calculated and tracked frame by frame. These metrics (or generally, image processing methods) can include the z-position of the maximum signal, the weighted average z-position, or any other image-derived quantity relating to the sample position. This information can be fed, for example, to a reference arm controller that mechanically adjusts the position of the reference arm reflector so that the OCT image of the sample is maintained at the same relative depth as before any movement of the sample. The same data can also be used to adjust the focal position of the sample arm, which is also digitally controlled. This should enable optimal positioning of the sample within the imaging window while maintaining high-quality imaging.

[0018] Machine learning algorithms can also be employed to detect specific features of interest within an image (or frame), similar to facial recognition. Machine learning can include, for example, the use of multilayer (artificial) neural networks having convolutional layers, pooling layers, activation layers, fully connected layers, normalization layers, and dropout layers. Each layer can include a sequence of weights and biases that govern their response to a given input image.

[0019] This allows for the creation of complex models in which weights can be iteratively trained on a set of ground truth (or reference) input data until the model can sufficiently predict the number of regions corresponding to the desired features and provide a confidence score for the strength of each prediction. For example, when using ground truth or known input images (images where the locations of relevant features are known), weights and / or biases can be fitted based on the deviation between the neural network's current output and known and / or expected values.

[0020] The results from the model are then preferably filtered to ensure that only an appropriate number of features and feature types are detected. For example, a neural network can be trained to detect a specific center of retinal vision (i.e., the fovea) within an OCT image or volume, and to distinguish this region from the optic disc. Prior knowledge of a target sample can be used to ensure that only a single instance or an appropriate number of instances of each feature are detected. Differences between the cornea, the anterior and / or posterior surfaces of the lens, and the corneal angle can also be determined to enable focusing and centering of the OCT probe beam over a particular region of interest. This allows similar access to z-direction position information for axial tracking, enabling detection and xy tracking toward a target, i.e., bilateral tracking of an image dataset or a corresponding OCT image. It can also be provided that the OCT imaging means (particularly within the sample arm) can be adapted to such lateral positional changes.

[0021] Further aspects of the present invention are directed toward machine learning methods. One aspect is the training of a machine learning method or neural network as described above. In this case, a processing unit can be provided that receives or acquires input data, performs the training steps described above, and provides or generates output data. For example, therein, data is collected, features of interest in the data are manually labeled, the data is split into training / validation sets, augmented into datasets, and then a neural network model is trained using model modification / retraining / deployment.

[0022] A further embodiment involves applying the trained neural network or machine learning method described above to determine a value that characterizes the axial position of an object with respect to the OCT imaging means. In this case, a processing unit (such as the external unit described above) can be provided that receives or acquires input data (image data acquired by OCT), performs the processing steps described above (identification of features in the image data), and provides or generates output data (values).

[0023] Machine learning may be used to detect and segment specific layers of interest within an image (such as the corneal endothelium or retinal lining). Data from different image sources, such as OCT images and parallel video camera images, may be combined to provide complementary information for applications such as tool tracking during surgery to provide feedback for robot-assisted surgery and maintain OCT imaging focused on tool positioning. Models can also be trained to detect anomalies such as fluid pocket accumulation or pore formation or tissue rupture (in the object).

[0024] Previous solutions, such as those described in U.S. Patent No. 7,349,098, apply to time-domain OCT, which is an older generation version of OCT that requires axial scanning of either the sample arm or the reference arm. Current-generation OCT, known as spectral-domain OCT or Fourier-domain OCT, does not have this requirement and allows for much faster image acquisition rates. Furthermore, by adjusting both the reference arm and the focus, imaging quality can be improved, especially for real-time imaging. If the focus is not adjusted in accordance with the adjustment of the reference arm, the sample will no longer be in the optimal lower limit of the imaging plane.

[0025] Furthermore, in the case of a sample having a large structural spread in the depth direction (e.g., anterior eye imaging), it may be difficult to automatically determine and track an appropriate position within the sample. Previous solutions were devised for specific use cases such as retinal tracking, but may not be very effective for tracking the cornea or other large samples. The same applies to machine learning methods for feature tracking. Also, previous solutions applied to post-processing as described in U.S. Patent Application Publication No. 2016 / 0040977 are not useful for improving quality in live imaging or real-time imaging.

[0026] The different suitable features described above enable specific suitable combinations or embodiments, some of which are described below.

[0027] In one embodiment, image tracking is provided via a comparison measurement criterion calculated from the current image frame. For example, this may include a weighted average calculation based on the signal intensity or another numerical value of the image to detect the z-direction position of the sample within one frame. This position is then compared to the next frame to determine the z-direction shift between frames. Information is then supplied to the controllers for the sample arm focus and the reference arm position to optimally adjust the image position.

[0028] Another embodiment provides image recognition via a machine learning method, which can be used to identify a portion of the sample of interest along the axial z-direction position. This will also enable the classification of each image and the potentially automated identification of features of interest (i.e., the macula of the eye, the upper part of the cornea, the tip of the lens, etc.). This may enable a more generalized z-direction tracking algorithm, and as a result, will also be supplied to the controller mechanism for the sample focus and the reference arm position.

[0029] Another embodiment provides z-direction tracking that can be performed using an external probe beam to an OCT system specialized for calculating the depth range of a sample (such as laser ranging). Information is similarly supplied to the optical control mechanism.

[0030] Another embodiment provides z-direction tracking calculated using any of the above methods instead of adjusting the reference arm, provided that the sample arm objective lens can be adjusted to compensate for the target's axial movement. This would allow for complete compensation of any axial movement and would typically require digital control over the sample arm of the microscope or OCT imaging means.

[0031] The present invention relates to an optical coherence tomography (OCT) imaging system for imaging an object such as an eye (particularly in real time), comprising a control system according to the present invention as described above and optical coherence tomography imaging means for performing an OCT scan (for a more detailed description of such OCT imaging means, reference should be made to the drawings and corresponding description). Preferably, the OCT imaging system is configured to display an image of the object on display means. Such display means may be part of the OCT imaging system. Preferably, the OCT imaging system further comprises means configured to mechanically and / or digitally adapt at least one of the focal position of the sample arm, the axial position of the reference arm reflector, and the axial position of the objective lens of the sample arm upon reception of a signal. Additionally, a probe can be provided that is configured to provide a probe beam for determining a value characterizing the axial position of the object with respect to the OCT imaging means.

[0032] The present invention also relates to a method for imaging an object such as an eye using optical coherence tomography (OCT), preferably spectral region OCT (or SS-OCT). The method includes the following steps: acquiring scanning data from an object using optical coherence tomography; performing data processing on the scanning data; obtaining image data of an image of the object; and determining values ​​that characterize the axial position of the object with respect to the OCT imaging means. The method further includes fitting at least one parameter of the OCT imaging means based on the change in values ​​between two sets of image data.

[0033] The present invention also relates, in particular, to a computer program comprising program code for causing a method according to the present invention to be performed when the computer program is executed on a processor, processing system or control system, as described above.

[0034] For further preferred details and advantages of this OCT imaging system and method, please also refer to the notes on the control system described herein, as applicable.

[0035] Further advantages and embodiments of the present invention will become apparent from the description and accompanying drawings.

[0036] It should be noted that the features described above and those further described below can be used not only in the combinations shown, but also in further combinations or individually, without departing from the scope of the present invention. [Brief explanation of the drawing]

[0037] [Figure 1] This is a schematic diagram showing a preferred embodiment of the OCT imaging system according to the present invention. [Figure 2] This is a schematic flowchart illustrating a preferred embodiment of the method according to the present invention. [Figure 3]This is a schematic diagram showing different OCT images of an object acquired by OCT at different points in time. [Figure 4] This diagram shows a time-axis representation of the changes in the axial position of an object. [Figure 5] This figure shows OCT images exhibiting specific features. [Modes for carrying out the invention]

[0038] Figure 1 shows a schematic diagram of a preferred embodiment of the optical coherence tomography (OCT) imaging system 100 according to the present invention. This OCT imaging system 100 includes a light source 102 (e.g., a low-coherence light source), a beam splitter 104, a reference arm 106, a sample arm 112, a diffraction grating 118, a detector 120 (e.g., a camera), a control system 130, and display means 140 (e.g., a display or monitor).

[0039] Light generated from the light source 102 is guided, for example, via an optical fiber cable 150 to a beam splitter 104, and a first portion of the light is transmitted through the beam splitter 104 and then guided to a reference reflector or reference mirror 110 to create a light beam 109 via an optical system 108 (shown schematically only and represented by a lens), where the optical system 108 and the reference mirror 110 are part of a reference arm 106. Adaptation means 160 are provided and configured to adapt (digitally and / or mechanically) the axial position of the mirror 110 (indicated by the left and right arrows of the mirror 110).

[0040] Light reflected from the reference mirror 110 is guided back to the beam splitter 104, transmitted through this beam splitter 104, and then guided to the diffraction grating 118 to create a light beam 117 via the optical system 116 (shown schematically only and represented by lenses).

[0041] A second portion of the light generated from the light source 102 and transmitted through the beam splitter 104 is guided through an optical system 114 (shown schematically only and represented by a lens) to an object 190 to be imaged in order to create a (scanning) light beam 115. This object 190 is, for example, an eye. The optical system 114 is part of the sample arm 112. In particular, the optical system 114 is used to focus the light beam 115 to a desired focal plane, i.e., to set a focal position indicated by reference numeral 113. The optical system 114 also represents the objective lens (e.g., of a microscope) used in the OCT. The adjustment means 162 is configured and provided to adjust the focal position (digitally and / or mechanically), for example, by adjusting the axial position of the optical system 114.

[0042] Light reflected from the object 190 or the tissue material within it is guided back to the beam splitter 104, transmitted through this beam splitter 104, and then guided to the diffraction grating 118 via the optical system 116. Thus, the light reflected from the reference arm 106 and the light reflected from the sample arm 112 are combined using the beam splitter 104 and guided to the diffraction grating 118 in the combined light beam 117, for example via an optical fiber cable 150.

[0043] Light reaching the diffraction grating 118 is diffracted and captured by the detector 120. In this way, the detector 120, which functions as a spectrometer, creates or acquires scanning data or scanning data 122, which is transmitted to the control system 130, for example, via an electrical cable 152. This control system 130 includes processing means (or processor) 132. The scanning data 122 is then processed to obtain an image dataset 142, which is transmitted to the display means 140, for example, via the electrical cable 152, and displayed as a real-time image 144, i.e., an image representing the object 190 being scanned in real time. All components except the control system and the display means can be considered as OCT imaging means.

[0044] Additionally, an external probe 180, such as a laser providing a probe beam 182, is provided and connected to the control unit 130. This may be used to determine the z-direction of the object 190, as will be described later.

[0045] Furthermore, a processing unit 136 is provided, connected to the control system 130 via a communication means 138 such as Ethernet or the Internet (and corresponding cables). As described above, all relevant processing steps can be performed on the control system 130 (or its processing means 132). However, steps that determine values, particularly by machine learning methods or image measurement criteria, can also be performed on an external processing unit 136 which may have high computing power.

[0046] Furthermore, another image source or imaging means 170, such as a video camera, is provided to provide video camera images in parallel. Such images may be combined with the OCT images to provide complementary information for applications such as tool tracking during surgery to provide feedback for robot-assisted surgery and maintain OCT imaging focused on the tool's position. Exemplarily, a tool 174 and robotic means 172 for controlling and moving the tool 174 (e.g., a scalpel or cutting laser) are shown. The robotic means 172 can be controlled, for example, by a control system.

[0047] The process by which the intensity scanning data 122 is processed or converted into an image dataset 142 that allows the scanned object 190 to be displayed on the display means 140 is described in more detail below. In particular, the axial or z-direction in which the object 190 may move during observation (or surgery) is indicated by the reference symbol z.

[0048] Figure 2 schematically shows a flowchart illustrating a preferred embodiment of the method according to the present invention. To provide real-time OCT images, scanning data is acquired and image data is provided, for example, in an image acquisition and processing process 200, so as to be continuously displayed on a display means as an image or OCT image, as described in relation to Figure 1.

[0049] In step 210, scanning data is acquired from the object using optical coherence tomography; in step 212, the scanning data is received by the control system; and in step 214, data processing is performed on the scanning data. The data processing step 214 may include, for example, DC or baseline removal, spectral filtering, wavenumber resampling, dispersion correction, Fourier transform, scaling, image filtering, and optionally additional image enhancement steps to obtain image data of the object to be displayed in step 216. Such image data can be provided to the display means in step 230 to provide real-time imaging, as described in relation to Figure 1.

[0050] Furthermore, in step 218, a value characterizing the axial position of the object with respect to the OCT imaging means is determined. Typically, it should be noted that the OCT imaging means (see above) does not move axially during scanning. The sample arm or optical system 114 (e.g., objective lens) may also be considered components whose axial position is determined or measured with respect to them. Such a determination may be performed on a set of image data representing a frame or two-dimensional OCT image. The value may also be, for example, the relative z-position. Figure 3 shows different OCT images 300, 302, and 304 of an object (in this example, the retina of an eye) acquired by OCT at different time points. In this example, axial motion in the z-direction is shown during imaging of the retina. Each image frame or OCT image was acquired at a different time point, and eye movement is causing the retina image to move up and down within the frame. For example, in the central image 302, out-of-focus movement is causing signal loss.

[0051] As mentioned above, step 218 may be performed on the control system or on an external processing unit. In the latter case, the image data will need to be provided to the external processing unit, which will need to receive values ​​from the control system after step 218.

[0052] Furthermore, the method includes, in step 220, a step of determining a potential change in values ​​between, for example, a subsequent image dataset or frame, such two values ​​are indicated by reference numerals 240 and 242. If the determined change exceeds, for example, a certain threshold, then in step 222, two parameters 244 and 246 of the OCT imaging means are adapted. For example, parameter 244 is the focal position of the sample arm adapted by means 162 shown in Figure 1. Parameter 246 may be the axial position of the reference arm reflector or mirror adapted by means 160 shown in Figure 1.

[0053] In this way, axial movement of the object can be compensated for. The diagram in Figure 4 shows the change in the axial position of the object over time. The z-direction is indicated by the pixels of the image, and the time axis t is indicated by the number of frames. This example specifically shows tracking the z-position of maximum brightness in each image over 100 frames. Frames without depth pixels indicate that the image has moved out of the field of view.

[0054] As described above, the values ​​characterizing the axial position of an object with respect to the OCT imaging means can be determined in different ways (see step 218). For example, as shown in Figure 4, an image measurement criterion 250 can be used to find the maximum brightness in each image. Alternatively, a machine learning method 260 can be used to find specific features within the OCT image.

[0055] Figure 5 shows an OCT image 500 in which specific features are indicated within a frame drawn with a white dotted line. This visualizes the concept of image recognition in OCT using a machine learning algorithm that can be trained to detect different features within the image. This provides information about the location of each feature and can assist in z-direction tracking toward a target as well as oriented scanning of a specific region of interest.

[0056] In Figure 5, the top white frame represents the cornea of ​​the eye, the bottom white frame represents the lens of the eye, and the left white frame represents the angle between the cornea and the lens.

[0057] In addition to such digital methods, in step 270, it is also possible to orient an external probe beam toward the object in order to obtain or determine a value indicating the axial direction of the object.

[0058] As used herein, the term "and / or" includes all possible combinations of one or more of the items listed herein and may be abbreviated as " / ".

[0059] While several embodiments have been described in the context of the apparatus, it is clear that these embodiments also represent descriptions of the corresponding methods, where blocks or apparatus correspond to steps or features of steps. Similarly, embodiments described in the context of steps also represent descriptions of the corresponding blocks, items, or features of the corresponding apparatus.

[0060] Some embodiments relate to OCT imaging systems that include a control system as described in relation to one or more of Figures 1 to 5. Alternatively, the OCT imaging system may be part of a system as described in relation to one or more of Figures 1 to 5, or it may be connected to a system as described in relation to one or more of Figures 1 to 5. Figure 1 shows a schematic diagram of an OCT imaging system 100 configured to perform the methods described herein. The OCT imaging system 100 includes OCT imaging means and a computer or control system 130. The OCT imaging means is configured to perform imaging and is connected to the control system 130. The control system 130 is configured to perform at least a portion of the methods described herein. The control system 130 may be configured to perform machine learning algorithms. The control system 130 and (part of) the OCT imaging means may be separate entities or may be integrated within a single common housing. The control system 130 may be part of the central processing system of the OCT imaging system 100, and / or the control system 130 may be part of the dependent components of the OCT imaging system 100, such as sensors, actuators, cameras, or lighting units.

[0061] The control system 130 may be a local computer device (e.g., a personal computer, laptop, tablet computer, or mobile phone) comprising one or more processors and one or more storage devices, or it may be a distributed computer system (e.g., a cloud computing system comprising one or more processors and one or more storage devices distributed to various locations such as local clients and / or one or more remote server farms and / or data centers). The control system 130 may include any circuit or combination of circuits. In one embodiment, the control system 130 may include one or more processors, which can be of any kind. As used herein, the processor may be intended to be any kind of computing circuit, such as a microprocessor for a microscope or microscopic component (e.g., a camera), a microcontroller, a composite instruction set computing (CISC) microprocessor, a reduced instruction set computing (RISC) microprocessor, a very long instruction word (VLIW) microprocessor, a graphics processor, a digital signal processor (DSP), a multicore processor, a field-programmable gate array (FPGA), or any other kind of processor or processing circuit. Other types of circuits that may be included in the control system 130 may be custom circuits, application-specific integrated circuits (ASICs), etc., such as one or more circuits (communication circuits, etc.) used in wireless devices such as mobile phones, tablet computers, laptop computers, two-way radios, and similar electronic systems. The control system 130 may also include one or more storage devices that may include one or more memory elements suitable for a particular application, such as main memory in the form of random access memory (RAM), one or more hard drives and / or one or more drives that handle removable media such as compact discs (CDs), flash memory cards, digital video discs (DVDs), etc.The control system 130 may include a display device, one or more speakers and a controller which may include a keyboard and / or mouse, trackball, touchscreen, voice recognition device, or any other device which enables the user of the system to input information to and receive information from the control system 130.

[0062] Some or all of the method steps may be performed by a hardware device (or by using a hardware device), such as a processor, microprocessor, programmable computer, or electronic circuit. In some embodiments, one or more of the most important method steps may be performed by such a device.

[0063] Depending on certain implementation requirements, embodiments of the present invention may be implemented in hardware or software. This implementation is feasible using a non-transient recording medium, which is a digital recording medium, etc., that stores electronically readable control signals and cooperates (or can cooperate) with a programmable computer system to carry out each method. Examples include floppy disks, DVDs, Blu-rays, CDs, ROMs, PROMs and EPROMs, EEPROMs, or FLASH memory. Thus, the digital recording medium may be computer-readable.

[0064] Some embodiments of the present invention include a data carrier having electronically readable control signals that can cooperate with a programmable computer system so as to carry out any of the methods described herein.

[0065] Generally, embodiments of the present invention can be implemented as a computer program product comprising program code, which operates to perform one of the methods when the computer program product is executed on a computer. This program code may be stored, for example, on a machine-readable carrier.

[0066] Another embodiment includes a computer program stored in a machine-readable carrier for carrying out any of the methods described herein.

[0067] Therefore, in other words, embodiments of the present invention are computer programs having program code for carrying out any of the methods described herein when the computer program is executed on a computer.

[0068] Accordingly, another embodiment of the present invention is a recording medium (or data carrier or computer-readable medium) containing a stored computer program for carrying out any of the methods described herein when executed by a processor. The data carrier, digital recording medium, or recording medium is typically tangible and / or non-transient. Another embodiment of the present invention is an apparatus, such as those described herein, comprising a processor and a recording medium.

[0069] Therefore, another embodiment of the present invention is a data stream or signal sequence representing a computer program for carrying out any of the methods described herein. The data stream or signal sequence may be configured to be transmitted, for example, over a data communication connection, such as the Internet.

[0070] Another embodiment includes processing means, for example, a computer or programmable logic device configured or adapted to carry out any of the methods described herein.

[0071] Another embodiment includes a computer having an installed computer program for carrying out any of the methods described herein.

[0072] Another embodiment of the present invention includes an apparatus or system configured to transfer (e.g., electronically or optically) a computer program for carrying out any of the methods described herein to a receiver. The receiver may be, for example, a computer, a mobile device, a storage device, etc. The apparatus or system may include, for example, a file server for transferring the computer program to the receiver.

[0073] In some embodiments, a programmable logic device (e.g., a field-programmable gate array) may be used to perform some or all of the functions of the methods described herein. In some embodiments, the field-programmable gate array may cooperate with a microprocessor to carry out any of the methods described herein. Generally, the methods are advantageously carried out by any hardware device.

[0074] Embodiments may be based on the use of machine learning models or machine learning algorithms. Instead of relying on models and inference, machine learning may refer to algorithms and statistical models that a computer system can use to perform a particular task without using explicit instructions. For example, machine learning may use data transformations inferred from the analysis of historical data and / or training data instead of rule-based data transformations. For example, image content may be analyzed using a machine learning model or a machine learning algorithm. For a machine learning model to analyze image content, the machine learning model may be trained with training images as input and training content information as output. By training a machine learning model with a large number of training images and / or training sequences (e.g., words or sentences) and associated training content information (e.g., labels or annotations), the machine learning model “learns” to recognize image content so that image content not included in the training data becomes recognizable using the machine learning model. The same principle may be used in the same way for other types of sensor data; that is, by training a machine learning model with training sensor data and a desired output, the machine learning model “learns” the transformation between sensor data and output, which can then be used to provide output based on non-trained sensor data provided to the machine learning model. The provided data (e.g., sensor data, metadata, and / or image data) may be preprocessed to obtain feature vectors that can be used as input to a machine learning model.

[0075] A machine learning model may be trained using training input data. The example above uses a training method called "supervised learning." In supervised learning, a machine learning model is trained using multiple training samples, each of which may contain multiple input data values ​​and multiple desired output values; that is, each training sample is associated with a desired output value. By specifying both the training samples and the desired output values, the machine learning model "learns" during training which output values ​​to provide based on input samples similar to the provided samples. In addition to supervised learning, semi-supervised learning may be used. In semi-supervised learning, some of the training samples lack corresponding desired output values. Supervised learning may be based on a supervised learning algorithm (e.g., a classification algorithm, a regression algorithm, or a similarity learning algorithm). A classification algorithm may be used if the output is limited to a limited set of values ​​(categorical variables), i.e., the input is classified into one of a limited set of values. A regression algorithm may be used if the output may have any numerical value (within a range). Similarity learning algorithms may be similar to both classification and regression algorithms, but are based on learning from examples using a similarity function that measures how similar or related two objects are. In addition to supervised or semi-supervised learning, unsupervised learning may be used to train machine learning models. In unsupervised learning, input data (only) may be provided, and unsupervised learning algorithms may be used to find structure in the input data (for example, by grouping or clustering the input data, or by finding commonalities in the data). Clustering is the process of assigning input data containing multiple input values ​​into multiple subsets (clusters), so that input values ​​within the same cluster are similar according to one or more (predefined) similarity criteria, but are not similar to input values ​​in another cluster.

[0076] Reinforcement learning is a third group of machine learning algorithms. In other words, reinforcement learning may be used to train machine learning models. In reinforcement learning, one or more software actors (referred to as “software agents”) are trained to take actions in their surroundings. A reward is calculated based on the actions taken. Reinforcement learning is based on training one or more software agents to choose actions that result in software agents that perform better on a given task, with cumulative rewards increasing (as revealed by the increase in rewards).

[0077] Furthermore, several techniques may be applied as part of a machine learning algorithm. For example, feature representation learning may be used. In other words, a machine learning model may be trained at least partially using feature representation learning, and / or a machine learning algorithm may include feature representation learning components. A feature representation learning algorithm, which may be called a representation learning algorithm, may not only store information in its own input but may also transform the information to make it useful, often as a preprocessing step before performing classification or prediction. Feature representation learning may be based, for example, on principal component analysis or cluster analysis.

[0078] In some examples, anomaly detection (i.e., outlier detection) may be used, which aims to provide the identification of input values ​​that raise suspicion by being significantly different from the majority of the input or training data. In other words, a machine learning model may be trained with anomaly detection, at least in part, and / or a machine learning algorithm may include anomaly detection components.

[0079] In some examples, a machine learning algorithm may use a decision tree as its predictive model. In other words, a machine learning model may be based on a decision tree. In a decision tree, observations about an item (e.g., a set of input values) may be represented by branches of the decision tree, and the output values ​​corresponding to these items may be represented by leaves of the decision tree. A decision tree may support both discrete and continuous values ​​as output values. When discrete values ​​are used, the decision tree may be represented as a classification tree, and when continuous values ​​are used, the decision tree may be represented as a regression tree.

[0080] Correlation rules are another technique that can be used in machine learning algorithms. In other words, a machine learning model may be based on one or more correlation rules. Correlation rules are created by identifying relationships between variables in a large amount of data. A machine learning algorithm may identify and / or utilize one or more correlational rules that represent knowledge derived from the data. These rules may be used, for example, to store, manipulate, or apply knowledge.

[0081] Machine learning algorithms are typically based on machine learning models. In other words, the term “machine learning algorithm” may refer to a set of instructions that can be used to create, train, or use a machine learning model. The term “machine learning model” may refer to a set of data structures and / or rules that represent learned knowledge (for example, based on training performed by a machine learning algorithm). In embodiments, usage of a machine learning algorithm may mean usage of one underlying machine learning model (or multiple underlying machine learning models). Usage of a machine learning model may mean that a machine learning model and / or a set of data structures / rules that are a machine learning model are trained by a machine learning algorithm.

[0082] For example, a machine learning model may be an artificial neural network (ANN). An ANN is a system influenced by biological neural networks, such as those found in the retina or brain. An ANN consists of multiple interconnected nodes and multiple junctions, or edges, between the nodes. Typically, there are three types of nodes: input nodes that receive input values, hidden nodes that are (simply) connected to other nodes, and output nodes that provide output values. Each node may represent an artificial neuron. Each edge may transmit information from one node to another. The output of a node may be defined as a (nonlinear) function of its input (e.g., the sum of its inputs). The input of a node may be used in a function based on the "weights" of the edges or nodes that provide the input. The weights of nodes and / or edges may be adjusted during the learning process. In other words, training an artificial neural network may involve adjusting the weights of the nodes and / or edges of the artificial neural network to obtain a desired output for a given input.

[0083] Alternatively, a machine learning model may be a support vector machine, a random forest model, or a gradient boosting model. A support vector machine (i.e., a support vector network) is a supervised learning model with a relevant learning algorithm that can be used to analyze data (e.g., in classification or regression analysis). A support vector machine may be trained by providing inputs with multiple training input values ​​belonging to one of two categories. A support vector machine may be trained to assign new input values ​​to one of two categories. Alternatively, a machine learning model may be a Bayesian network, which is a stochastic directed acyclic graphical model. A Bayesian network may use a directed acyclic graph to represent a set of random variables and their conditional dependencies. Alternatively, a machine learning model may be based on a search algorithm and a genetic algorithm, which is a heuristic method that mimics the process of natural selection. [Explanation of symbols]

[0084] 100 OCT imaging systems 102 Light source 104 Beam Splitter 106 Reference Arm 108,114,116 Optical system 109,115,117 Light beams 110 Reference Mirror 112 Sample Arm 113 Focus position 118 Diffraction grating 120 detectors 122 Intensity scanning data 130 Control Systems 132 Processing means 136 Processing Units 138 Means of communication 140 Display means 142 image dataset 150 fiber optic cables 152 Electrical Cables 160,162 Adaptation means 170 video cameras 172 Robotic means 174 Tools 180 External probe 182 External probe beam 190 Object x,z direction t time 200-230, 270 Method Steps 240,242 values 244,246 parameters 250 Image Measurement Standards 260 Machine Learning Methods 300-304,500 OCT images

Claims

1. A control system (130) for controlling an optical coherence tomography imaging means for imaging an object (190), The control system (130) follows the following steps, namely, Step (212) of receiving scanning data (122) from the object (190) acquired using optical coherence tomography, Step (214) of performing data processing on the scanned data (122), Step (216) to obtain image data (142) of the image (144) of the object, It is configured to perform, The control system (130) is further configured to adapt (222) at least one parameter (244, 246) of the OCT imaging means based on a change in a value (240, 242) that characterizes the axial position (z) of the object (190) with respect to the OCT imaging means between two sets of image data. The control system (130) is further configured to perform a step (218) of determining the values ​​(240, 242), The control system (130) is configured to determine the values ​​(240, 242) for the image data (142) from data acquired by an external probe beam (182). Control system (130).

2. At least one of the parameters (244, 246) of the OCT imaging means is selected from the focal position (113) of the sample arm (112), the axial position of the reference arm reflector (110), and the axial position of the objective lens (114) of the sample arm (112). The control system (130) according to claim 1.

3. The control system (130) is configured to adapt at least one of the parameters (244, 246) of the OCT imaging means by providing a signal for corresponding adaptation means (160, 162) configured to mechanically and / or digitally adapt at least one of the parameters (244, 246) of the OCT imaging means upon receiving the values ​​(240, 242). The control system (130) according to claim 1 or 2.

4. The scanning data (122) from the object (190) is acquired using spectral region optical coherence tomography. A control system (130) according to any one of claims 1 to 3.

5. The control system (130) is further configured to receive the values ​​(240, 242) from an external processing unit (136). A control system (130) according to any one of claims 1 to 4.

6. A processing system for optical coherence tomography imaging means, The processing system includes the control system (130) described in claim 5, The processing system further includes a processing unit (136), The processing unit (136) is configured to perform the steps of: determining a value that characterizes the axial position of an object with respect to the OCT imaging means (218); and providing the value to the control system (130). The processing unit (136) is preferably connected to the control system (130) via communication means (138). Processing system.

7. An optical coherence tomography imaging system (100) for imaging an object (190), wherein the optical coherence tomography imaging system (100) comprises a control system (130) according to any one of claims 1 to 5 or a processing system according to claim 6, an optical coherence tomography imaging means, and preferably a display means (140) configured to display an image (144) of the object (190), An optical coherence tomography imaging system (100) including the above.

8. The optical coherence tomography imaging means includes a sample arm containing an objective lens (114) and a reference arm (106) containing a reference arm reflector (110). The optical coherence tomography imaging system (100) further includes adaptation means (160, 162) configured to mechanically and / or digitally adapt at least one of the focal position (113) of the sample arm (112), the axial position of the reference arm reflector (110), and the axial position of the objective lens (114) of the sample arm (112) upon receiving the values ​​(240, 242). The optical coherence tomography imaging system (100) according to claim 7.

9. The optical coherence tomography imaging system (100) further includes a probe (180) configured to provide a probe beam (182) for determining the values ​​(240, 242). The optical coherence tomography imaging system (100) according to claim 7 or 8.

10. The optical coherence tomography imaging system (100) is configured to be used during a surgical procedure performed on the object (190), The optical coherence tomography imaging system (100) according to any one of claims 7 to 9.

11. A method for imaging an object (190) using optical coherence tomography, the method comprising the following steps: The steps include: acquiring scanning data (122) from the object using optical coherence tomography (210); Step (214) of performing data processing on the scanned data, Step (216) to obtain image data (142) of the image of the object, For the image data (142), a step (218) is taken to determine values ​​(240, 242) that characterize the axial position of the object with respect to the OCT imaging means from data acquired by an external probe beam (182), Includes, The method further includes the step (222) of adapting at least one parameter (244, 246) of the OCT imaging means based on the change in the values ​​(240, 242) between two sets of image data. method.

12. At least one of the parameters of the OCT imaging means is selected from the focal position (113) of the sample arm (112), the axial position of the reference arm reflector (110), and the axial position of the objective lens (114) of the sample arm (112). The method according to claim 11.

13. The above values ​​(240, 242) Image measurement criteria (250), preferably an image measurement criteria (250) based on maximum and / or weighted average signal intensity, A machine learning method (260) configured to identify a portion of the object in an image dataset, The steps include: oriented the external probe beam onto the object (190) (270), The decision is made using at least one of the following: The method according to claim 11 or 12.

14. A computer program comprising program code that, when the computer program is executed on a control system (130), causes the method described in any one of claims 11 to 13 to be executed. Computer program.

Images