Method and system for analyzing Optical Coherence Tomography (OCT) images

Abstract

Disclosed is a method for analyzing Optical Coherence Tomography (OCT) images. The method comprises acquiring an OCT image and detecting one or more of a vessel lumen and a guidewire shadow in the OCT

Description

TECHNICAL FIELD The present disclosure relates to methods for analyzing Optical Coherence Tomography (OCT) images. Moreover, the present disclosure relates to systems for analyzing Optical Coherence Tomography (OCT) images. Furthermore, the present disclosure relates to computer programs comprising computer executable program codes. BACKGROUND Coronary artery disease (CAD) is one of the major causes of death in the world. CAD is characterized by the buildup of plaque inside the coronary arteries, which can lead to blockages and reduced blood flow to the heart. This, in turn, can result in heart attacks and other severe cardiovascular events. Therefore, early and accurate diagnosis, as well as monitoring of CAD, is crucial for effective treatment and prevention of further complications. In recent years, Optical Coherence Tomography (OCT) has emerged as a preferred imaging technique for studying CAD. OCT offers high-resolution cross-sectional images of the coronary arteries, providing clinicians with detailed views of the artery wall, plaque build-up, and other pertinent features. In particular, OCT can both identify different tissues including normal vessel, lipid and calcium, and measure both lipid and calcium arc and depth as well as micron-level features such as fibrous cap thickness and presence of macrophages. Moreover, OCT can be used to guide placement of stents, to test efficacy of drugs and / or devices for treatment of CAD, and to predict risk of future multiple adverse cardiovascular events (MACE) such as cardiovascular death and heart attacks. In addition to structural predictors of events, coronary plaques experience considerable mechanical stress from changes in pressure and artery geometry during systole and diastole. This plaque structural stress (PSS) is associated with presentation with acute coronary syndromes (unstable angina / heart attacks), plaque rupture, and future events, particularly when measured longitudinally in three-dimensions (3D). Therefore, PSS calculations can increase the ability of intracoronary imaging to identify high-risk plaques and predict future events. However, it is often that PSS varies significantly within the plaque, around the lumen, and longitudinally, and is affected by very small components such as calcium. To date, solutions to calculate PSS from intracoronary OCT are limited to gross features (e.g., lumen or plaque curvature, plaque burden) because of major difficulties in understanding plaque composition. Therefore, in light of the foregoing discussion, there exists a need to overcome the aforementioned limitations / drawbacks. SUMMARY The aim of the present disclosure is to provide a method and a system to process OCT images for accurately identifying regions of high stress in the vessel lumen. The aim of the present disclosure is achieved by a method and a system for analyzing Optical Coherence Tomography (OCT) images as defined in the appended independent claims to which reference is made to. Advantageous features are set out in the appended dependent claims. Throughout the description and claims of this specification, the words "comprise", "include", "have", and "contain" and variations of these words, for example "comprising" and "comprises", mean "including but not limited to", and do not exclude other components, items, integers or steps not explicitly disclosed also to be present. Moreover, the singular encompasses the plural unless the context otherwise requires. In particular, where the indefinite article is used, the specification is to be understood as contemplating plurality as well as singularity, unless the context requires otherwise. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a flowchart depicting steps of a method for analyzing Optical Coherence Tomography (OCT) images, in accordance with one or more embodiments of the present disclosure; FIG. 2 is a system for analyzing Optical Coherence Tomography (OCT) images, in accordance with one or more embodiments of the present disclosure; and FIG. 3 is an example of calculation of plaque structural stress (PSS) from an OCT image, in accordance with one or more embodiments of the present disclosure. DETAILED DESCRIPTION OF EMBODIMENTS The following detailed description illustrates embodiments of the present disclosure and ways in which they can be implemented. Although some modes of carrying out the present disclosure have been disclosed, those skilled in the art would recognize that other embodiments for carrying out or practicing the present disclosure are also possible. In a first aspect, the present disclosure provides a method for analyzing Optical Coherence Tomography (OCT) images, the method comprising: - acquiring an OCT image; - detecting one or more of a vessel lumen and a guidewire shadow in the OCT image; - transforming the OCT image into OCT segments; - converting OCT segments into OCT pixels corresponding to a tissue composition of a target site; - performing a non-uniform circumferential vessel shrinkage algorithm to OCT pixels to obtain zero-pressure OCT pixels; - performing a two-dimensional finite element analysis (FEA) simulation on zero-pressure OCT pixels to calculate two-dimensional plaque structural stress distribution band plots of the target site; - converting the two-dimensional plaque structural stress distribution band plots of the target site into a three-dimensional plaque structural stress distribution map of the target site; and - integrating the three-dimensional plaque structural stress distribution map of the target site with the original OCT image with at least one of: the OCT image and OCT segments, to analyze the OCT images. The present method offers a systematic and comprehensive approach to process OCT intracoronary images for biomechanical analysis. By transforming high resolution OCT images into structural stress maps that can be viewed in 3D and integrating them with the original (native) or segmented OCT images for composite analysis, the method targets detecting small differences in tissue composition accurately calculating plaque structural stress (PSS), thereby accurately identifying regions of high stress in the vessel lumen. Moreover, the method effectively segments plaque components into high resolution segments, converts these 2D images into 3D, and integrates them with the original or segmented OCT images for composite analysis ensures preservation of all important details. Overall, this synergistic approach ensures clinically valuable OCT images. In a second aspect, the present disclosure provides a system for analyzing Optical Coherence Tomography (OCT) images, the system comprising: - an input module configured to acquire an OCT image; and - a processing module in signal communication with the input module, the processing module configured to: - detect one or more of a vessel lumen and a guidewire shadow in the OCT image; - transform the OCT image into OCT segments; - convert OCT segments into OCT pixels corresponding to a tissue composition of a target site; - perform a non-uniform circumferential vessel shrinkage algorithm to OCT pixels to obtain zero-pressure OCT pixels; - perform a two-dimensional finite element analysis (FEA) simulation on zero-pressure OCT pixels to calculate two-dimensional plaque structural stress distribution band plots of the target site; - convert the two-dimensional plaque structural stress distribution band plots of the target site into a three-dimensional plaque structural stress distribution map of the target site; and - integrate the three-dimensional plaque structural stress distribution map of the target site with the original OCT image with at least one of: the OCT image and OCT segments, to analyze the OCT images. The present system provides an integrated approach to process OCT intracoronary images for biomechanical analysis. The system comprises the input module to acquire the OCT image, and the processing module for image transformation and composite analysis. The processor employs AutoOCT software that uses deep convolutional neural networks with encoding-decoding architecture to detect the vessel lumen and the guidewire shadow, and segments OCT images into its component tissues for biomechanical analysis. In this regard, the system applies biomechanical constants for different tissues and calculates PSS from the processed OCT images. Moreover, the system integrates the processed OCT image with the original or segmented OCT images for composite analysis, thereby ensuring preservation of all important details. The method comprises acquiring the OCT image. The term "Optical Coherence Tomography (OCT)" refers to a non-invasive imaging technique that captures high-resolution cross-sectional images of biological tissues. OCT, as an imaging modality, uses light waves to capture micrometer-resolution, cross-sectional images of biological tissues. Typically, acquiring an OCT image involves directing a beam of light onto the tissue of interest and then measuring the echo time delay and intensity of the reflected light. A scanning mechanism then captures multiple such measurements across a specified area, and these measurements are algorithmically processed to construct a two-dimensional or three-dimensional representation of the tissue's internal structure. This acquired image serves as the initial input for the subsequent transformation and integration process for composite analysis. In present embodiments, the OCT image is of a coronary artery. The coronary arteries are blood vessels responsible for supplying oxygenated blood to the heart muscle. Given their critical function, any obstructions, anomalies, or diseases affecting these arteries can have severe consequences, leading to conditions such as coronary artery disease (CAD). Typically, a coronary artery is a tubular structure having a vessel lumen on the inner side thereof. The vessel lumen is characterized by at least one property corresponding to the tissue composition thereof. The at least one property comprises a calcium arc; calcium deposits; calcium area or depth; a fibrous cap; fibrous cap thickness or area; lipid deposits; lipid arc, area or depth; stress; plaque structural stress (PSS); wall stress; vessel lumen minimum or maximum diameter or area; vessel or lumen stenosis, vessel, lumen or plaque curvature, irregularity or roughness, plaque burden, plaque area, as known to a person skilled in the art. Notably, a given OCT image, acquired through OCT imaging procedures, captures the coronary artery's detailed view, encompassing aspects like the artery wall (comprising intima (innermost layer), media (middle layer), and adventitia (outer layer)), plaque formations (such as fibrous plaques, lipid-rich plaques, and calcified plaques), and other relevant anatomical features such as coronary artery lumen (the inner open space) dimension, fibrous cap thickness, thrombus detection, stent assessment, tissue perfusion, and so forth. Due to the presence of various features such as plaques, lumen, and arterial walls, the given OCT image aims to capture these details with the highest possible resolution. The specific focus on the coronary arteries emphasizes the applicability of the present disclosure for composite analysis of the OCT images of the coronary arteries. In particular, composite analysis of the intracoronary OCT image (i.e., OCT image of the coronary artery) involves the integration of multiple OCT images to create a comprehensive view of the coronary artery, facilitating the diagnosis and treatment of coronary artery diseases. Beneficially, it provides high-resolution information about the vessel's structure, helping clinicians make informed decisions about patient care. In an embodiment, the method comprises detecting one or more of the vessel lumen and the guidewire shadow in the OCT image. Notably, the vessel lumen and the guidewire shadow analysis are relevant for coronary arteries. The vessel lumen is the innermost open space within a vessel through which blood flows, accurate identification of the vessel lumen is therefore important for various diagnostic and therapeutic procedures. Moreover, the guidewire shadow is an artifact that often appears in OCT images, especially when imaging is performed during certain interventional procedures. Notably, a guidewire, a medical instrument used to guide catheters or other devices within the vascular system, can introduce shadows or dark regions in the OCT images. These shadows, if not identified, can lead to image misinterpretation, and may thus affect the subsequent artifact correction process. Optionally, the OCT image may be preprocessed to remove artifacts, such as the guidewire shadow. Notably, the artifacts may emerge due to a myriad of factors, ranging from the inherent complexities of the coronary artery's anatomy to external interventions and even the imaging process itself. Beneficially, such preprocessing of the OCT image to remove such artifacts ensures preserving essential details in the OCT images, potentially leading to accurate diagnosis and interpretation. In an embodiment, the method further comprises implementing a neural network to detect the one or more of the vessel lumen and the guidewire shadow in the OCT image. Neural networks are a category of algorithms within the broader field of artificial intelligence and machine learning. These algorithms are designed to recognize patterns within data, making them suitable for tasks such as image feature detection. In an embodiment, the neural network is a deep convolutional neural network with encoding-decoding architecture, and wherein the neural network is trained to detect the one or more of the vessel lumen and / or the guidewire shadow in the OCT images. Deep convolutional neural networks, often referred to as Convolutional Neural Networks (CNNs), are a subclass of neural networks particularly tailored for image analysis. These networks utilize convolutional layers to automatically, adaptively learn spatial hierarchies of features, from the input images. Given the spatial nature of OCT images, with intricate details and varying contrasts, CNNs are aptly suited for their analysis. The encoding-decoding architecture, often seen in networks designed for tasks like image (such as OCT images) segmentation, is a two-fold process. The encoding phase involves a series of convolutional layers that progressively reduce the spatial dimensions of the input image, herein the OCT images, while increasing the depth, capturing the essential features and details. Following the encoding phase is the decoding phase, wherein the compressed feature representation is gradually up sampled to restore the image, herein the OCT images, to its original dimensions, but with the desired features, in this case, the vessel lumen and / or the guidewire shadow, distinctly highlighted or segmented. In the context of the present method, the neural network is trained on a dataset of OCT images wherein the vessel lumen and the guidewire shadow have been previously identified and labelled. Through this training, the neural network learns to discern the patterns and characteristics associated with the vessel lumen and the guidewire shadow. Once adequately trained, the neural network can then process new, unlabeled OCT images and accurately detect these features. By integrating this neural network-based analysis step, the method ensures that the given OCT image, which will undergo further transformation and processing, is already optimized in terms of the identification of critical features and potential artifacts. This optimization not only improves the efficiency of the subsequent steps but also enhances the overall accuracy and reliability of the analysis resulting from the OCT images that are of superior quality and clinical relevance. The method comprises transforming the OCT image into OCT segments. It will be appreciated that the OCT image is a high-resolution image, typically in a range of micrometers, and such micrometer-range high-resolution OCT image allows for visualization of fine details within tissues, making it valuable for clinical purpose, such as diagnosis of various medical conditions predictive by a state of the coronary arteries. The term "OCT segments" typically refers to the various layers or regions that are segmented or delineated within an OCT image. For example, the OCT segments involve identifying layers of arterial walls and other biochemical or biomechanical aspects associated therewith. Moreover, the OCT segments provide detailed information about the microstructure, composition, and pathology of the coronary artery wall. The OCT segments are essential for diagnosing coronary artery diseases and guiding interventional procedures. In other words, OCT segments enable accurately quantifying tissue thickness and other morphological measurements. Optionally, transforming the OCT image into OCT segments involves the process of identifying and delineating different anatomical or pathological structures or layers within the tissue being imaged. In particular, transforming the OCT image into OCT segments requires specialized image processing techniques and software, comprising, but not limited to at least one of: a noise reduction step (achieved using median filtering, Gaussian smoothing, or wavelet denoising, for example); image enhancement to improve the visibility of structures in the OCT image by adjusting contrast and brightness, for example; thresholding by setting a threshold value to separate the foreground (structures of interest) having high pixel value from the background having low pixel value compared to the threshold value; edge detection to identify abrupt intensity changes in the image by employing algorithms, like Canny, Sobel, or the Laplacian of Gaussian (LoG), for example, wherein the edges correspond to boundaries between different tissues or structures; region growing for segmenting homogeneous regions within an image by iteratively adding neighboring pixels with similar intensity to the segmented region comprising a seed point ; active contour models or snakes for segmenting irregularly shaped structures, by automatically adapting to the boundaries of structures in the image; watershed transform segments objects with uneven or ambiguous boundaries based on the topography of intensity values; machine learning and deep learning approaches, such as convolutional neural networks (CNNs) for segmenting structures directly from labelled data with high accuracy; post-processing steps for removing small-connected components, filling gaps, smoothing the boundaries, and refining the results; and validation of the segmentation results for assessing the quality of segmentation by using quantitative metrics such as Dice coefficient, Jaccard index, or Hausdorff distance. In an example, a software for segmentation (i.e., transformation of OCT image into OCT segments), such as a proprietary custom software, for example AutoOCT, is used for transforming the OCT image into OCT segments. Such software employs a combination of preprocessing steps and postprocessing steps. The preprocessing steps comprise noise reduction, contrast enhancement, and image registration in case of multiple alignments. Subsequently, a region of interest (R.OI) is selected by any one of a manual selection or automated selection step and the selected segmentation algorithm is applied to the OCT image within the chosen ROI to identify and delineate the boundaries of the structures of interest. The postprocessing steps comprise removing any potential small artifacts, smoothing segmented boundaries, and filling gaps from the OCT segments. The accuracy of the segmentation results is subsequently validated by using quantitative metrics (such as Dice coefficient, Jaccard index) or visual inspection. Optionally, relevant parameters like thickness, area, volume, or other anatomical features are measured from the OCT segments and analyzed for clinical interpretation. The AutoOCT software segments the OCT images into its component tissues and measures a range of tissue characteristics (e.g., lipid and calcium arcs and areas, fibrous cap thickness, and so on). It will be appreciated that one or more steps mentioned above may be performed reiteratively to refine the transformation of the OCT images to OCT segments. The method comprises converting OCT segments into OCT pixels corresponding to the tissue composition of the target site. Herein, the term "converting" may be related to the term "representing" since an image, such as the OCT image, is inherently composed of pixels, wherein each pixel represents a single point in the image and contains information about its color or intensity. For example, in grayscale images, each pixel typically has a single value representing its brightness (e.g., 0 for black and 255 for white in an 8-bit image), and in color images, each pixel usually has multiple values (e.g., Red, Green, Blue, or RGB values, or Cyan, Magenta, Yellow, Black, or CMYK values) to represent its color. Moreover, pixels are arranged in a grid, and each pixel is assigned coordinates within that grid. A top-left pixel is often assigned coordinates (0, 0), and a bottom-right pixel has coordinates corresponding to the image's width and height (e.g., (1919, 1079) for a Full HD image). Furthermore, a pixel depth (bit depth) determines the number of unique values in each pixel. For example, an 8-bit image can have 256 different grayscale values for each pixel, while a 24-bit color image has 8 bits for each of the Red, Green, and Blue channels, allowing for over 16 million colors. Therefore, the arrangement of these pixels, their values, and their color information collectively define the image's visual content. Beneficially, converting (or representing) OCT segments of the coronary artery into corresponding OCT pixels and analyzing each pixel individually allows for accurately identifying and locating features or regions of interest (i.e., tissues) within the image, as well as quantitative measurements of various OCT image properties, that is essential for diagnosis and monitoring. Moreover, each OCT pixels corresponds to the at least one property corresponding to the coronary artery. For example, each OCT pixels corresponds to the plaque component of the coronary artery. Moreover, the identified features may be used in machine learning and pattern recognition algorithms to classify or characterize features within the image. Furthermore, OCT pixels enable enhancing the original (or native) OCT image quality, image compression while preserving important details, and manipulating individual pixels for creative image processing. The method comprises performing the non-uniform circumferential vessel shrinkage algorithm to OCT pixels to obtain zero-pressure OCT pixels. The non-uniform circumferential vessel shrinkage algorithm is used to reduce size of a tubular structure, such as the vessel lumen, in the OCT image of the coronary artery while preserving its shape and important features. In this regard, the non-uniform circumferential vessel shrinkage algorithm employs a pre-defined shrinkage profile that specifies an extent of shrinkage of each of the OCT pixels. Notably, the pre-defined shrinkage profile may be based on a mathematical equation that takes into account one or more physiological parameters associated with vessel morphology. Optionally, the one or more physiological parameters include, but do not limit to, vessel diameter, vessel curvature and radius thereof, vessel winding, vessel branching angle, vessel length, vessel orientation, vessel tapering vessel density with a certain region of a subject's body, vessel lumen diameter, vessel wall elasticity, vessel wall texture, vessel wall composition, hemodynamic parameters of the vessel. Beneficially, such one or more physiological parameters allow tailoring the shrinkage to specific characteristics of the vessels being analyzed. The pre-defined shrinkage profile when applied to each pixel in the OCT pixel under zero pressure conditions moves said pixel inwards to contract the OCT pixel. The term "zero-pressure OCT pixels" as used herein refers to OCT pixels generated after applying the shrinkage algorithm that involves no external force or pressure to be applied to the tissues within the image that might cause distortions or deformations or artifacts to the image content. Such zero-pressure OCT pixels preserve the original proportions, aspect ratio, and spatial relationships of the image content. In this regard, it will be appreciated that shrinking the OCT pixel results in modification of the coordinates or change in the vessel width at the corresponding pixel, however, the modification of the coordinate or change in the vessel width is uniform throughout the zero-pressure OCT pixels, thereby maintaining the original spatial and structural integrity of tissues within the zero-pressure OCT pixels. In this regard, the outer boundary pressure is set at zero, such that the farthest boundary of the simulated model, often considered as an "outlet" boundary, is assumed to have zero pressure. Optionally, shrinking of the OCT pixel may be performed iteratively, and with each shrinkage, the method comprises updating the corresponding pixel in the original OCT pixel to reflect the shrinking vessel. Moreover, the method may employ interpolation or other techniques to fill in the gaps created by the shrinkage. Optionally, the non-uniform circumferential vessel shrinkage algorithm is applied longitudinally to OCT pixels, allowing monitoring changes in vascular structures over time, particularly useful for tracking disease progression or assessing the effects of therapeutic interventions. Beneficially, the non-uniform circumferential vessel shrinkage algorithm allows for standardization of vessel diameters for quantitative analysis, as reducing the variability in vessel diameter makes it easier to measure vessel dimensions accurately, thus making it valuable in studies related to vascular morphology. Moreover, shrinking the vessels can improve the visualization of other structures within or near the vascular bed, thus enabling identification of lesions, tumors, or other abnormalities that may be partially obscured by large vessels in the original image. Furthermore, shrinkage algorithms help in reducing noise and clutter in OCT images by simplifying the appearance of vascular structures, thereby resulting in clearer images, making it easier to identify and analyze important features. Such clearer images may subsequently be used to train machine learning models on standardized data to generalize well to new cases. The method comprises performing the two-dimensional finite element analysis (FEA) simulation on zero-pressure OCT pixels to calculate two-dimensional plaque structural stress distribution band plots of the target site. It will be appreciated that the zero-pressure OCT pixels, that represent the at least one property corresponding to the tissue composition of the vessel lumen, are used as an initial input for 2D FEA simulations. The application of two-dimensional finite element analysis (2D FEA) simulation on zero-pressure OCT pixels comprises converting the zero-pressure OCT pixels into a finite element mesh, defining material properties and boundary conditions. Optionally, the material properties comprise, for example, elastic modulus, thermal conductivity, or other relevant properties based on the application. Optionally, boundary conditions may include constraints, loads, or other external influences, in presence of which the zero-pressure OCT pixels image interacts differently with its surroundings. In this regard, 2D FEA simulation subdivides the zero-pressure OCT pixels into a grid of elements, wherein each element represents a small portion of the zero-pressure OCT pixels, and nodes are placed at the element corners. Moreover, the 2D FEA simulation further comprises applying FEA principles to analyze various zero-pressure OCT pixels image-related properties or phenomena. Herein, the FEA simulations enable calculating two-dimensional plaque structural stress (2D PSS) distribution band plots of the target site. The term "plaque structural stress" or "PSS" refers to the mechanical stress experienced by atherosclerotic plaques in arteries. Typically, PSS is a composite of circumferential wall stress and radial wall stress and PSS values refer to maximum principal stress normalized by coronary pressure. The 2D PSS distribution band plots are graphical representations of distribution of the structural stress within atherosclerotic plaques. Typically, horizontal and vertical components of the 2D PSS distribution band plots represent an angular direction and a longitudinal direction along the 3D vessel, respectively. Atherosclerotic plaques are fatty deposits that build up within the walls of arteries. The atherosclerotic plaques can become unstable and rupture, leading to cardiovascular events like heart attacks and strokes. The term "structural stress" as used herein refers to mechanical stress experienced by the atherosclerotic plaque due to changes in pressure and arterial geometry from systole and diastole, which affects the stability of the atherosclerotic plaque. The 2D PSS distribution band plots, which focus on a cross-sectional view of the artery and the atherosclerotic plaque, provide insights into the mechanical stress experienced by arterial plaques to assess stress distribution along the plaque's length and width (since the plague is also segmented along with the OCT image) and their vulnerability and potential for rupture. Typically, the plaque may be segmented into various regions, such as the fibrous cap, the lipid core, and the calcified regions. Optionally, the method comprises color mapping different stress levels, for example colors like blue or green can be used to indicate lower stress and colors like red or yellow can be used to indicate higher stress. Beneficially, such color mapping annotates high stress concentrations in certain regions that may indicate an increased risk of plaque rupture, and may require consideration of interventions like stent placement or surgery. Beneficially, the FEA simulation allows for enhancing image quality or resolving details that may not be visible in the original image, thereby enabling accurate quantitative analysis of the zero-pressure OCT pixels image data, such as related to stress, strain, temperature, or other relevant parameters. Moreover, the FEA simulation simulations can predict behavior of an object or structure within the image under different conditions. Furthermore, the FEA simulation allows assessing the structural integrity of objects or components within the zero-pressure OCT pixels image, to identify potential risks. Additionally, the FEA simulation allows fluid flow (such as hemodynamic parameters of the vessel) analysis within the zero-pressure OCT pixels. Optionally, each plaque component is modelled as incompressible, piecewise homogeneous, non-linear isotropic and hyper-elastic. In the context of atherosclerotic plaques, the term "incompressible" as used herein implies that the plaque is assumed to have a nearly constant volume or is highly resistant to volume changes when subjected to mechanical loading. The term "piecewise homogeneous" as used herein suggests that the plaque is divided into distinct regions or segments, and each segment is considered to have uniform material properties. Moreover, such regions may correspond to different components within the plaque, such as the fibrous cap, lipid core, and calcified regions. Each region is treated as having uniform material properties for modeling purposes. The term "Non-linear" as used herein indicates that the mechanical behavior of the plaque material does not follow a linear stress-strain relationship, and as and when subjected to mechanical forces, its response is not proportional to the applied load and may exhibit non-linear behavior, which can include material stiffening or softening under different conditions. The term "Isotropic" as used herein means that the material properties of the plaque are assumed to be the same in all directions irrespective of the direction of the applied force. The term "hyper-elastic" as used herein refers to a specific type of material model used to describe the behavior of soft tissues like arterial walls and plaque components. Hyper-elastic materials are characterized by a strain-energy density function that captures the material's response to deformation. Beneficially, such material properties are employed in simulations for analyzing the mechanical behavior of plaques under various loading conditions, allowing for a better understanding of plaque vulnerability and the potential for rupture. It will be appreciated that each plaque component is modelled as incompressible, piecewise homogeneous, non-linear isotropic and hyper-elastic element, as described by the modified Mooney-Rivlin strain energy density function: W = -3) + D^exp^D^ - 3)) - 1] + k( / - 1) where / , is the first invariant of the modified right Cauchy-Green tensor — —2 / 3 C = J C; J is the Jacobian of the deformation gradient tensor, C; k is the Lagrange multiplier; and c , D and D2 are material constants derived from previous experimental work and include; arterial vessel wall, calcification, fibrous tissue, and lipid / necrotic core. The material properties of dense calcification are derived by fitting a Young's modulus based on experimental data. The motion of each plaque component is governed by kinetic equations as: pv = ct (i, j = 2) LJ,J where [v.] and [a ] are the displacement vector and stress tensor, respectively, p = density of each component, and t = time. The entire plaque model is then meshed using quadrilaterals, generating nodes and elements for each plaque model. Herein, both displacement and strain are assumed to be large, with no relative movement at the interface of individual atherosclerotic components and a relative energy tolerance set. Adjacent points are fixed at the outer wall of the model to prevent rigid body movement. The loading conditions for each simulation are taken from coronary or peripheral pressure recordings, with outer boundary pressure set at zero. Optionally, the FEA simulations may be performed using commercial TM source software (e.g., ADINA (ADINA R&D Inc., Watertown, US)). Variation in PSS during one cardiac cycle is calculated, being defined as: Variation of PSS = -fpSS^ i \ i J \ i J wherein the subscript i is the / th integration node and the superscript t = time computed. Optionally, the two-dimensional plaque structural stress distribution band plot of the target site is calculated, for a cardiac cycle, by applying biomechanical constants. Calculating the two-dimensional plaque structural stress distribution band plot for a cardiac cycle requires data corresponding to the plaque's geometry (such as dimensions, thickness, and composition (e.g., fibrous cap, lipid core)), material properties (such as elastic modulus and Poisson's ratio), and dynamic forces acting on it throughout the cardiac cycle, and phase-wise division of the cardiac cycle (i.e., diastole and systole). Optionally, the dynamic forces acting on plaques throughout the cardiac cycle in the arteries may be related to the cyclic changes in blood pressure, flow, and shear stress. Optionally, the dynamic forces include blood pressure (diastolic and systolic), blood flow, pulsatile flow, wall shear stress, plaque deformation, pulse wave reflection, and so on. In this regard, the method comprises applying a biomechanical constant defined by the FEA simulation for solving equations of motion for each element of the zero-pressure OCT pixels image data based on the data corresponding to the plaque's geometry, material properties, and the dynamic forces. Subsequently, calculate the plaque structural stress distribution within the plaque with respect to time steps simulating the plaque structural stress changing at each phase of the cardiac cycle, and plot the 2D PSS distribution band plot. Optionally, the plaque structural stress is sampled at equiangular intervals to generate the two-dimensional plaque structural stress distribution band plot of periluminal plaque structural stress. Notably, sampling PSS involves assessing and measuring the mechanical stress experienced by atherosclerotic plaques within the walls of arteries. Stress values at specific locations or regions of interest within the plaque are sampled to identify vulnerable plaque components like the fibrous cap's thinnest points or regions with high stress concentrations. Such high stress locations or regions are used to generate the 2D PSS distribution band plot of periluminal plaque structural stress. In an embodiment, the sampling of PSS is performed at equiangular intervals by selecting specific points or regions within the atherosclerotic plaque. Optionally, the specific points or regions are selected along the circumference or contour of the atherosclerotic plaque, at regular angular intervals, thereby ensuring that stress values are obtained uniformly across the plaque's surface. Optionally, the equiangular intervals range from 5-45 degrees. The equiangular intervals may for example be from 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35 or 40 up to 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40 or 45. Optionally, preferably, the equiangular intervals range from 5-30 degrees. For example, if the equiangular interval is a 10-degree interval, the sample stress values is recorded at 36 equiangular positions (i.e., 360 degrees / 10 degrees per interval) around the plaque's circumference. The method comprises converting the two-dimensional plaque structural stress distribution band plots of the target site into the three-dimensional plaque structural stress (3D PSS) distribution map of the target site. In this regard, a volumetric reconstruction of the plaque is created by using the segmented plaque components from the segmented OCT pixels and a 2D cross-section of the plaque along its longitudinal axis is performed to create the 3D PSS distribution map of the target site. The above-described FEA simulation may optionally be performed on the 3D PSS distribution map to calculate stress values throughout the 3D structure of the plaque and depict the stress values on the 3D PSS distribution map, for example in a color-mapped manner, as described above. Beneficially, the 3D PSS distribution map provides a more comprehensive understanding of plaque structural stress distribution within the plaque and assists in clinical decision-making for patients at risk of cardiovascular events. The method comprises integrating the three-dimensional plaque structural stress distribution map of the target site with the original OCT image with at least one of: the OCT image and OCT segments, to analyze the OCT images. That is, following the generation of the 3D PSS distribution map, the method further includes integrating of this 3D PSS distribution map with the original, acquired OCT image and / or the OCT segments. It may be appreciated that the OCT image, which serves as the initial input for the method, includes genuine anatomical details and optionally no undesired artifacts, therefore, it is imperative to ensure that the inherent details of the given OCT image are preserved and not overshadowed by the processing steps. Similarly, the OCT segments also include genuine anatomical details which are rather refined by processing steps for isolation and measurement of specific features or regions within the image. To achieve this balance, the 3D PSS distribution map is integrated with the OCT image and / or the OCT segments. It may be appreciated that the integrating process is not a mere overlay (or merging) of the two images; but instead, it is a calculated integration that ensures that the refined details from the 3D PSS distribution map enhance the genuine anatomical features of the OCT image and / or the OCT segments, while simultaneously suppressing or eliminating undesired artifacts or low-resolution pixels. The resultant integrated 3D OCT image retains the details of the OCT image and / or the OCT segments, while simultaneously incorporating the more comprehensive understanding of plaque structural stress distribution within the plaque introduced by the 3D PSS distribution map. The resultant integrated 3D OCT image, thus, offers a clear and unobstructed view of the imaged tissue, facilitating more accurate and informed clinical decision-making. Optionally, the detected one or more of the vessel lumen and the guidewire shadow is merged to generate the integrated 3D OCT image. Specifically, during integration of the 3D PSS distribution map with the OCT image and / or the OCT segments, the detected vessel lumen and guidewire shadow are also integrated. Such integration is done with precision, ensuring that the boundaries and positions of these detected features align accurately with their representations in the OCT image and / or the OCT segments and the 3D PSS distribution map. Beneficially, this ensures that the final integrated 3D OCT image is not only refined in terms of resolution but also includes critical anatomical and procedural details. Optionally, the method comprises integrating the two-dimensional plaque structural stress distribution band plot with the original OCT image to create a composite view of the vessel lumen. Integrating the 2D PSS distribution band plot with the original OCT image calculated integration of the stress distribution information onto the image for a more comprehensive visualization. It will be appreciated that for achieving integration, any image processing softwares or tools known to a person skilled in the art, may be used. Therefore, the present method provides a structured, precise, and comprehensive approach to analyze OCT images. By leveraging advanced mathematical transformations, custom shrinkage algorithms, and precise conversion of the 2D OCT images into 3D maps, the method ensures high-quality OCT images that are important for clinical diagnosis and treatment. The present disclosure also relates to the system for analyzing Optical Coherence Tomography (OCT) images as described above. Various embodiments and variants disclosed above, with respect to the aforementioned method, apply mutatis mutandis to the system. In the system, the input module serves as the primary interface for receiving or acquiring the OCT images. The input module acts as a gateway through which OCT images, generated from various imaging devices or sources, are fed into the system for further processing. The input module may be equipped with capabilities to handle different formats or standards of OCT images, ensuring versatility in accommodating images from various OCT imaging devices. In an example, raw OCT data may be exported in OCT, DICOM, TIF format, and so on, as input using an image loader. Generally, as used herein, the term "processing module" refers to a computational element that is operable to respond to and processes instructions that drive the system. Optionally, the processing module includes, but is not limited to, a microprocessor, a microcontroller, a complex instruction set computing (CISC) microprocessor, a reduced instruction set (RISC) microprocessor, a very long instruction word (VLIW) microprocessor, or any other type of processing circuit. Furthermore, the term "processing module" may refer to one or more individual processors, processing devices and various elements associated with a processing device that may be shared by other processing devices. Additionally, the one or more individual processors, processing devices and elements are arranged in various architectures for responding to and processing the instructions that drive the system. In an embodiment, the system further comprises a neural network implemented by the processing module, wherein the neural network is further configured to detect one or more of the vessel lumen and the guidewire shadow in the OCT image. In an embodiment of the system, the neural network is a deep convolutional network with an encoding-decoding architecture, and wherein the neural network is trained to detect the one or more of the vessel lumen and the guidewire shadow in the OCT image. In an embodiment of the system, the processing module is configured to calculate the two-dimensional plaque structural stress distribution band plots of the target site, for a cardiac cycle, by applying biomechanical constants. In an embodiment of the system, the processing module is configured to sample plaque structural stress, and wherein the plaque structural stress is sampled at equiangular intervals to generate the two-dimensional plaque structural stress distribution band plots of periluminal plaque structural stress. In an embodiment of the system, the processing module is configured to integrate the two-dimensional plaque structural stress distribution band plots with the original OCT image to create a composite view of the vessel lumen. In a third aspect, the present disclosure provides an apparatus comprising a computer program stored in a memory, the computer program being configured to control the apparatus to perform the method of the first aspect as described above. Various embodiments and variants disclosed above, with respect to the aforementioned method and the aforementioned system, apply mutatis mutandis to the apparatus. The apparatus may be implemented as a standalone workstation that integrates specialized hardware and software components to execute the method efficiently. The apparatus may include a central processing unit (CPU), complemented by a graphics processing unit (GPU) tailored for the intensive image processing tasks, especially beneficial for operations like Fourier transformation and neural network computations. Additionally, the apparatus would feature a dedicated input interface, compatible with various OCT imaging devices, allowing for acquisition of OCT images. Furthermore, to facilitate real-time visualization and assessment, the apparatus may incorporate a display which may render the transformed OCT images for clinical analysis; or sometimes both the given OCT images and the transformed OCT images side by side, offering clinicians an immediate comparative view. In a fourth aspect, the present disclosure provides a computer program comprising computer executable program code, when executed the program code controls a computer to perform the method of the first aspect as described above. Various embodiments and variants disclosed above, with respect to the aforementioned method and the aforementioned system, apply mutatis mutandis to the computer program. The computer program may be implemented as Al-based image pre-processing software that processes OCT intracoronary images for biomechanical analysis. Such OCT image processing targets detecting small differences in tissue composition accurately calculating plaque structural stress (PSS), thereby accurately identifying regions of high stress in the vessel lumen. This Al-based processing is non-destructive and retains accuracy of lumen and plaque parameters. These clinically relevant OCT images may assist clinicians and researchers to optimize stent placement and identify higher-risk plaques and the effects of drugs on various medical conditions, such as atherosclerosis. DETAILED DESCRIPTION OF THE DRAWINGS Referring to FIG. 1, illustrated is a flowchart 100 of steps of a method for analyzing Optical Coherence Tomography (OCT) images, in accordance with one or more embodiments of the present disclosure. At step 102, an OCT image is acquired. At step 104, one or more of a vessel lumen and a guidewire shadow is detected in the OCT image. At step 106, the OCT image is transformed into OCT segments. At step 108, OCT segments are converted into OCT pixels corresponding to a tissue composition of a target site. At step 110, a non-uniform circumferential vessel shrinkage algorithm is performed on to OCT pixels to obtain zero-pressure OCT pixels. At step 112, a two-dimensional finite element analysis (FEA) simulation is performed on zero-pressure OCT pixels to calculate two-dimensional plaque structural stress distribution band plots of the target site. At step 114, the two-dimensional plaque structural stress distribution band plots of the target site are converted into a three-dimensional plaque structural stress distribution map of the target site. At step 116, the three-dimensional plaque structural stress distribution map of the target site is integrated with the original OCT image with at least one of: the OCT image and OCT segments, to analyze the OCT images. Referring to FIG. 2, illustrated is a system for analyzing Optical Coherence Tomography (OCT) images, in accordance with one or more embodiments of the present disclosure. As shown the system comprises an input module 202 configured to acquire an OCT image; and a processing module 204 in signal communication with the input module 202. The processing module 204 is configured to: detect one or more of a vessel lumen and a guidewire shadow in the OCT image; transform the OCT image into OCT segments; convert OCT segments into OCT pixels corresponding to a tissue composition of a target site; perform a non-uniform circumferential vessel shrinkage algorithm to OCT pixels to obtain zero-pressure OCT pixels; perform a two-dimensional finite element analysis (FEA) simulation on zero-pressure OCT pixels to calculate two-dimensional plaque structural stress distribution band plots of the target site; convert the two-dimensional plaque structural stress distribution band plots of the target site into a three-dimensional plaque structural stress distribution map of the target site; and integrate the three-dimensional plaque structural stress distribution map of the target site with the original OCT image with at least one of: the OCT image and OCT segments, to analyze the OCT images. Referring to FIG. 3, illustrated is an example of plaque structural stress (PSS) calculation from an OCT image, in accordance with one or more embodiments of the present disclosure. As shown, the OCT image (left panel) depicts the tubular cross-section of a vessel lumen comprising plaque at a distal portion of bottom of the tubular vessel lumen. A corresponding two-dimensional PSS distribution band plot of the OCT image identifies regions of high stress, associated with the identified plaque at a distal portion of bottom of the tubular vessel lumen, color-coded in a manner that the colors like violet to blue represent low stress and the colors like red to pink represent high stress. Moreover, the 2D PSS distribution band plot also shows the varying thickness of the vessel lumen as captured in the OCT image. Modifications to embodiments of the present disclosure described in the foregoing are possible without departing from the scope of the present disclosure as defined by the accompanying claims. Expressions such as "including", "comprising", "incorporating", "have", "is" used to describe and claim the present disclosure are intended to be construed in a non-exclusive manner, namely allowing for items, components or elements not explicitly described also to be present. Reference to the singular is also to be construed to relate to the plural.

Claims

1. A method for analyzing Optical Coherence Tomography (OCT) images, the method comprising:- acquiring an OCT image;- detecting one or more of a vessel lumen and a guidewire shadow in the OCT image;- transforming the OCT image into OCT segments;- converting OCT segments into OCT pixels corresponding to a tissue composition of a target site;- performing a non-uniform circumferential vessel shrinkage algorithm to OCT pixels to obtain zero-pressure OCT pixels;- performing a two-dimensional finite element analysis (FEA) simulation on zero-pressure OCT pixels to calculate two-dimensional plaque structural stress distribution band plots of the target site;- converting the two-dimensional plaque structural stress distribution band plots of the target site into a three-dimensional plaque structural stress distribution map of the target site; and- integrating the three-dimensional plaque structural stress distribution map of the target site with the original OCT image with at least one of: the OCT image and OCT segments, to analyze the OCT images.

2. A method according to claim 1, wherein the OCT image is of a coronary artery.

3. A method according to claim 1, further comprising implementing a neural network to detect the one or more of the vessel lumen and the guidewire shadow in the OCT image.

4. A method according to claim 2, wherein the neural network is a deep convolutional neural network with encoding-decoding architecture, and wherein the neural network is trained to detect the one or more of the vessel lumen and / or the guidewire shadow in the OCT images.

5. A method according to claim 1, wherein the two-dimensional plaque structural stress distribution band plot of the target site is calculated, for a cardiac cycle, by applying biomechanical constants.

6. A method according to claim 1, wherein the plaque structural stress is sampled at equiangular intervals to generate the two-dimensional plaque structural stress distribution band plot of periluminal plaque structural stress.

7. A method according to claims 1, wherein the method comprises integrating the two-dimensional plaque structural stress distribution band plot with the original OCT image to create a composite view of the vessel lumen.

8. A method according to claim 1, wherein each plaque component is modelled as incompressible, piecewise homogeneous, non-linear isotropic and hyper-elastic.

9. A system for analyzing Optical Coherence Tomography (OCT) images, the system comprising:- an input module configured to acquire an OCT image; and- a processing module in signal communication with the input module, the processing module configured to:- detect one or more of a vessel lumen and a guidewire shadow in the OCT image;- transform the OCT image into OCT segments;- convert OCT segments into OCT pixels corresponding to a tissue composition of a target site;- perform a non-uniform circumferential vessel shrinkage algorithm to OCT pixels to obtain zero-pressure OCT pixels;- perform a two-dimensional finite element analysis (FEA) simulation on zero-pressure OCT pixels to calculatetwo-dimensional plaque structural stress distribution band plots of the target site;- convert the two-dimensional plaque structural stress distribution band plots of the target site into a three-dimensional plaque structural stress distribution map of the target site; and- integrate the three-dimensional plaque structural stress distribution map of the target site with the original OCT image with at least one of: the OCT image and OCT segments, to analyze the OCT images.

10. A system according to claim 9 further comprising a neural network implemented by the processing module, wherein the neural network is further configured to detect one or more of the vessel lumen and the guidewire shadow in the OCT image.

11. A system according to claim 10, wherein the neural network is a deep convolutional network with an encoding-decoding architecture, and wherein the neural network is trained to detect the one or more of the vessel lumen and the guidewire shadow in the OCT image.

12. A system according to claim 9, wherein the processing module is configured to calculate the two-dimensional plaque structural stress distribution band plots of the target site, for a cardiac cycle, by applying biomechanical constants.

13. A system according to claim 9, wherein the processing module is configured to sample plaque structural stress, and wherein the plaque structural stress is sampled at equiangular intervals to generate the two-dimensional plaque structural stress distribution band plots of periluminal plaque structural stress.

14. A system according to claim 9, wherein the processing module is configured to integrate the two-dimensional plaque structural stressdistribution band plots with theoriginal OCT image to create acomposite view of the vessel lumen.

15. An apparatus comprising a computer program stored in amemory, the computer program being configured to control theapparatus to perform the method according to any one of claims 1-8.

16. A computer program comprising computer executable program code, when executed the program code controls a computer to perform the method according to any one of claims 1-8.Application No: GB2413596.4 Examiner: Mr Joe McCannClaims searched: 1-16Date of search: 24 January 2025Patents Act 1977: Search Report under Section 17Documents considered to be relevant:Category Relevant to claims Identity of document and passage or figure of particular relevance Y 1,9 CN 118171540 A (BAIYIHUIXIN HANGZHOU NETWORK TECH CO LTD) - See whole document Y 1,9 WO 2006 / 062958 A2 (WORCESTER POLYTECH INST) - See abstract, paragraphs 82-82,115,116 and figures 1 and 4-9 Y 1,9 CN 114841991 A (BEIJING INSTITUTE TECH) - See whole document Y 1,9 US 11948301 B2 (MIN et al.) - See abstract and figure 9F Y 1,9 WO 2023 / 000039 Al (NAVIER MEDICAL LTD) - See whole document Y 1,9 Atherosclerosis, vol. 254, 2016, Kang S -J et al., Plaque structural stress assessed by virtual histology-intravascular ultrasound predicts dynamic changes in phenotype and composition of untreated coronary artery lesions, pages 85-92. Y 1,9 JACC: Cardiovascular Imaging, vol. 10, no. 12, 2017, Samady H et al., The Ongoing Quest to Predict Plaque Rupture, pages 1484-1486.Categories:X Document indicating lack of novelty or inventive step A Document indicating technological background and / or state of the art. Y Document indicating lack of inventive step if p Document published on or after the declared priority date but combined with one or more other documents of same category. before the filing date of this invention. & Member of the same patent family E Patent document published on or after, but with priority date earlier than, the filing date of this application.Field of Search:Intellectual Property Office is an operating name of the Patent Ofncewww.gov.uk / ipoThe following online and other databases have been used in the preparation of this search report SEARCH-PATENT, SEARCH-NPLInternational Classification:Subclass Subgroup Valid From G06T 0007 / 00 01 / 01 / 2017 G06T 0007 / 10 01 / 01 / 2017Intellectual Property Office is an operating name of the Patent Ofncewww.gov.uk / ipo

Images