Method and system for identifying shadows in intravascular images

Abstract

The present invention relates, in part, to shadow detection and shadow verification with respect to data sets obtained from intravascular imaging data collection. The method can use local adaptive thresholding and scan line level analysis with respect to candidate shadow regions to determine a set of candidate shadows for verification or rejection. In one embodiment, the shadows are stent strut shadows, guidewire shadows, collateral vessel shadows, or other shadows.

Description

[0001] This application is a divisional application of Chinese Patent Application No. 201680067856.8, filed on November 22, 2016, entitled "Detection and Validation of Shadows in Intravascular Images", the disclosure of which is incorporated by reference in its entirety. TECHNICAL FIELD

[0002] The present invention relates to systems and methods for feature detection such as shadows and stent struts in intravascular images. BACKGROUND

[0003] Interventional cardiologists incorporate various diagnostic tools in the treatment of interventional procedures to plan, guide, and assess treatment. Angiographic imaging of blood vessels is often performed using x-ray fluoroscopy. Such blood vessel imaging is then used by physicians to diagnose, locate, and treat vascular disease during interventions such as heart bypass surgery or stent implantation. Intravascular imaging techniques such as optical coherence tomography (OCT) are also valuable tools that can be used in place of or in conjunction with x-ray fluoroscopy to obtain high resolution data about the condition of a given subject's blood vessels.

[0004] Intravascular optical coherence tomography is a catheter-based imaging modality that uses light to peer through the walls of coronary arteries and generate images of them for study. Using coherent light, interferometry, and micro-optics, OCT can provide in-vivo tomography at video rates with micron-scale resolution inside diseased blood vessels. The use of a fiber-optic probe to view subsurface structures at high resolution makes OCT particularly useful for minimally invasive imaging of internal tissues and organs and implanted medical devices such as stents.

[0005] Stents are a common intervention for treating stenosis of blood vessels. It is important for clinicians to develop a personalized stent plan that is made according to the patient's vascular anatomy to ensure the best outcome of the intravascular procedure. Stents cast shadows in intravascular images, and detecting existing stent deployments must address various challenges related to shadows in intravascular images.

[0006] The present invention addresses various challenges related to shadow detection and shadow validation. SUMMARY

[0007] Disclosed herein are systems and methods for detecting shadows and enhancements related to shadow detection in the context of intravascular datasets such as vascular images. In one embodiment, the systems and methods use locally adaptive thresholding to detect candidate shadows. Furthermore, in some embodiments, candidate shadows can be validated to reduce false positive shadows.

[0008] The systems and methods disclosed herein detect various shadows associated with stent struts, guidewires, and other intravascular imaging probe components and blood vessel features. In one embodiment, the shadows produced by stent struts during imaging are used to detect the stent struts.

[0009] In part, the present invention relates to a method of detecting shadows in intravascular images. The method includes determining a local estimate of tissue intensity, generating / determining a locally adaptive threshold as a function of scan line, and detecting a shadow associated with an intravascular object based on one or more sets of scan lines in which the tissue projection intensity is below the locally adaptive threshold. In one embodiment, the method includes storing one or more intravascular data sets using an intravascular diagnostic system, each intravascular data set including a plurality of scan lines.

[0010] In one embodiment, the shadow detection is performed using a locally adaptive threshold. In one embodiment, the locally adaptive threshold method is applied on a per-scan line basis with respect to various intensity levels. In one embodiment, the shadow detection method is configured to have a sensitivity level appropriate for finding shadows even if two methods, such as the first and second methods, are used with different shadow search criteria or features. Thus, the methods can also include one or more verification steps to verify the shadows. The use of some verification steps can improve overall performance and accuracy in detecting struts / guidewires based on initially detected and verified shadows.

[0011] In one embodiment, the shadow detection is performed using a locally adaptive threshold. In one embodiment, the locally adaptive threshold method is applied on a per-scan line basis with respect to various intensity levels. In addition, as a subsequent, backup, or alternative shadow detection method, local minima can be searched for and detected based on user-specified or diagnostic intravascular data collection system-specified criteria. In one embodiment, the local minima have an intensity value greater than or equal to the locally adaptive threshold (LAT). In one embodiment, the local minima have an intensity value greater than the LAT.

[0012] In one embodiment, one or more steps of the method are implemented using a diagnostic system including an input for receiving intravascular data, one or more electronic storage devices for storing the data sets, one or more computing devices / data processing apparatus in electrical communication with the input and the one or more electronic storage devices, and instructions, image filters, sampling methods, kernels, operators, and image processing software modules executable by the one or more computing devices to perform one or more steps of the method. Implementations of the described technology can include hardware, a method or process, or computer software on a computer-accessible medium or stored in a computer-readable medium such as a non-transitory computer-readable medium.

[0013] In part, the present invention relates to a system of one or more computing devices configured to perform a particular operation or action by installing a software image processing module and other software, firmware, hardware, or combinations thereof on a system that causes the system to perform the operation or action. One or more computer programs can be configured to perform a particular operation or action by having them include instructions that, when executed by a data processing apparatus, cause the apparatus to perform the action. One general aspect includes a method of detecting shadows in intravascular images. The method includes storing one or more intravascular data sets using an intravascular diagnostic system, each intravascular data set including a plurality of scan lines. The method can further include determining a plurality of line projections on each scan line, each line projection determined using a near-tissue offset and a far-tissue offset.

[0014] In one embodiment, the method further includes determining a local estimate of tissue intensity using the line projections. The method can further include determining a locally adaptive threshold that varies with the scan line. The method can further include identifying a shadow representing a feature of interest in the intravascular data set using a grouping of consecutive scan lines in which the local estimate of intensity is below the locally adaptive threshold. Other embodiments of this aspect include corresponding computer systems, apparatus, and computer programs recorded on one or more computer storage devices, each configured to perform the actions of the methods.

[0015] In one embodiment of the present invention, implementations can include one or more of the following features. The method can further include determining a plurality of near offsets for the plurality of scan lines. The method can further include determining a plurality of far offsets for the plurality of scan lines. The method can further include identifying a candidate shadow based on a presence of a local minimum within the line projection, where the local minimum has an intensity that is less than a given fraction of one or more maximum intensities found within a neighborhood on either side of a scan line of the plurality of scan lines. The method can further include estimating a plurality of slope values related to a search window around each scan line to identify a change in slope that indicates an edge of a shadow region. The method can further include performing one or more shadow verification methods for the detected edge. In one embodiment, the local estimate of tissue intensity is a smoothed projection generated on a per-scan line basis. The method can further include searching for one or more relative extrema along the smoothed projection and identifying a shadow based on a label using the one or more relative extrema. In the method, the label is a trough between two peaks.

[0016] In one embodiment, the method can further include performing a search for shadow regions in the one or more line projections. The method can further include validating the identified shadows. In one embodiment, validating the shadows further includes detecting one or more edges with a kernel. The method can further include displaying one or more objects in the representation of the blood vessel, the objects being associated with the one or more validated shadows. The method can further include identifying shadows of the line projections that are below the locally adaptive threshold. The method can further include generating a locally adaptive threshold based on each scan line using a local average of the tissue.

[0017] In one embodiment, one or more steps of the method are performed using a diagnostic system including an input for receiving one or more intravascular data sets, one or more electronic storage devices for storing the one or more intravascular data sets, one or more computing devices in electrical communication with the input and the one or more electronic storage devices, and instructions, image filters, and image processing software modules executable by the one or more computing devices to perform one or more steps of the method. In one embodiment, the intravascular diagnostic system is an optical coherence tomography system.

[0018] In one embodiment, the method further includes generating a locally adaptive threshold based on each scan line using a local average of the tissue. The method further includes identifying shadows of the line projections that are below the locally adaptive threshold. The method further includes performing a local minimum search to identify additional candidate shadows. The method further includes performing edge refinement on one or more shadow boundary scan lines using the measured slope values of the line projections.

[0019] In one embodiment, one or more steps of the method are performed using a diagnostic system including an input for receiving one or more intravascular data sets, one or more electronic storage devices for storing the one or more intravascular data sets, one or more computing devices in electrical communication with the input and the one or more electronic storage devices, and instructions, image filters, and image processing software modules executable by the one or more computing devices to perform one or more steps of the method. Implementations of the described technology can include hardware, a method or process, or computer software on a computer-accessible medium.

[0020] In one aspect, the present disclosure is directed to a method of detecting shadows in intravascular images, which can include storing one or more intravascular data sets using an intravascular diagnostic system, each intravascular data set comprising a plurality of scan lines. The method can also include determining a first offset and a second offset for the plurality of scan lines. The method can also include determining a line projection for each scan line of the plurality of scan lines by averaging samples between the first offset and the second offset. The method can also include performing a search for shadow regions within the line projections. The method can also include validating identified shadows. The method can also include displaying one or more objects in a representation of the blood vessel, the objects being associated with one or more of the validated shadows. Other implementations of this aspect include corresponding computer systems, apparatus, and computer programs recorded on one or more computer storage devices, each configured to perform the actions of the methods.

[0021] In one implementation, the implementations can include one or more of the following features. In one implementation, the intravascular diagnostic system is an optical coherence tomography system. The method can also include generating a local adaptive threshold using local average tissue values on a per-scan line basis. The method can also include identifying shadows of the line projections that are below the local adaptive threshold. The method can also include performing a local minimum search to identify additional candidate shadows. The method can also include performing edge refinement on one or more shadow boundary scan lines using measured slope values of the line projections.

[0022] In one implementation, one or more steps of the method are implemented using a diagnostic system comprising an input for receiving one or more intravascular data sets, one or more electronic storage devices for storing the one or more intravascular data sets, one or more computing devices in electrical communication with the input and the one or more electronic storage devices, and instructions, image filters, and image processing software modules executable by the one or more computing devices to perform one or more steps of the method. Implementations of the described technology can include hardware, a method or process, or computer software positioned on a computer- accessible medium as well as the other features disclosed herein.

[0023] While the present disclosure relates to different aspects and implementations, it is to be understood that the different aspects and implementations disclosed herein can be combined, as appropriate, in whole or in part, with one another. Thus, each implementation disclosed herein can incorporate each aspect, as appropriate, to varying degrees for a given implementation, and steps from various methods can be combined without limitation.

[0024] Other features and advantages of the disclosed implementations will be apparent from the following description and drawings.

[0025] In one embodiment, stent struts suitable for the detection steps described herein are generally metal stent struts. Any stent struts that create shadows during imaging using an intravascular probe are also suitable for detection using the methods described herein. BRIEF DESCRIPTION OF DRAWINGS

[0026] The drawings are not necessarily to scale and emphasis has usually been placed upon illustrating the principles of the application, the application being fully described in the detailed description supplied herein. These drawings are considered to be exemplary in nature and are not intended to limit the application, the scope of which is defined solely by the claims.

[0027] Figure 1A is an exemplary intravascular data collection system and associated intravascular data collection probe according to an exemplary embodiment of the present application, as well as related image processing, detection and other software components.

[0028] Figure 1B is a process flow diagram for detecting shadows, stent struts and other intravascular features according to an exemplary embodiment of the present application.

[0029] Figure 2 is a process flow diagram for stent detection according to an exemplary embodiment of the present application.

[0030] Figure 3A is an intravascular polarity image in 2-D spatial coordinates according to an exemplary embodiment of the present application, including various shadow regions analyzed and detected using the methods described herein.

[0031] Figure 3B is an intravascular polarity image according to an exemplary embodiment of the present application, representing polar coordinates as a rectangular R-0 image generated with respect to Figure 3A including various shadow regions analyzed and detected using the methods described herein.

[0032] Figure 3C is a mask generated in a 2-D spatial coordinate system and with respect to an image representing polar coordinates as a rectangular R-0 image according to an exemplary embodiment of the present application. Figure 3A

[0033] Figure 3D is a mask generated with respect to an image representing polar coordinates as a rectangular R-0 image according to an exemplary embodiment of the present application. Figure 3B

[0034] Figure 4 is a process flow diagram for various shadow detection and verification steps, as well as other intravascular data processing steps, according to an exemplary embodiment of the present application.

[0035] Figure 5 ​​is an exemplary plot of line projections generated using data from intravascular image frames (such as OCT image frames) and values associated with intensity, projections, relative extrema, local adaptive threshold, shadows (such as curves, lines), or data points of tissue determined from data of intravascular image frames in the case of scan line pair intensity values according to exemplary embodiments of the present application.

[0036] Figure 6 is a process flow chart showing an exemplary shadow search method according to exemplary embodiments of the present application.

[0037] Figures 7A to 7C is an example of an operator such as can be applied to an image to detect features or other values of interest according to exemplary embodiments of the present application. DETAILED DESCRIPTION

[0038] The systems and methods disclosed herein relate to intravascular imaging and shadows that can appear in such images due to stent struts, intravascular imaging probe components, and other factors. The presence of shadows in intravascular regions is problematic because in the diagnostic process they can be incorrectly identified as side branches, stenosis, lipid pools, or otherwise obscure features of interest. In intravascular images such as OCT and IVUS images, dim and fuzzy shadows can cause unwanted image processing errors and interfere with other steps in the image processing thread. Moreover, in one embodiment, accurate shadow detection is a prerequisite step in stent strut detection, guidewire detection, and shadow generating object (such as metallic object) detection.

[0039] In part, the present application relates to methods that enhance shadow detection to be more sensitive to fuzzy shadows. As a competing factor, increasing the sensitivity threshold for detecting fuzzy shadows can result in many false positives being identified. In one embodiment of the present application for candidate shadows, a shadow verification step is performed to reduce or remove the number of false positives. The methods and implementations described herein can be used with a variety of intravascular imaging systems and probes.

[0040] Figure 1Ais a high level schematic depicting a blood vessel 5, such as an artery, a data collection probe 7, and an intravascular data collection and processing system 10. The system 10 can include, for example, an OCT system, an IVUS system, or other intravascular imaging system. A stent 12 is shown in the blood vessel 5. The stent includes a plurality of struts. Some of the struts can create a shadow or shadow region SR as part of the process of imaging the blood vessel with the intravascular probe. The system 10 can include various software modules suitable for performing side branch detection, peak detection, shadow region detection and processing, error correction, model comparison, lumen detection, and various other processes described herein. The system 10 can include suitable light sources to meet the consistency and bandwidth requirements of the applications and data collection described herein. The system 10 can include an ultrasound imaging system. The probe 7 can include a catheter 20 having a catheter portion with one or more optical fibers 15 disposed therein and a probe tip 17. In one embodiment, the probe tip 17 includes a beam director.

[0041] As shown, the catheter 20 is introduced into a lumen 11, such as an arterial lumen. The probe 7 can include a rotating optical fiber 15 or a slidable optical fiber 15 that directs light forward into the lumen 14 or in a direction perpendicular to the longitudinal axis of the optical fiber 15. Thus, with the light directed from the side of the probe as the optical fiber 15 is rotated, OCT data is collected about the wall of the blood vessel 5. The wall of the blood vessel 5 defines a lumen boundary. This lumen boundary can be detected with distance measurements obtained from the optical signals collected at the probe tip 17 using lumen detection software components. Shadow regions and other features can be identified in scan lines generated during pullback of the probe through the artery. The shadow regions can or can not be associated with stent struts. In one embodiment, the probe 7 can include other imaging modalities such as ultrasound in addition to OCT.

[0042] As Figure 1A shown, the probe tip 17 is positioned in the lumen 14 such that the probe tip 17 is distal of the stent region of the blood vessel 5. The probe tip 17 is configured to emit light and receive backscattered light from objects such as the stent 12 and the wall of the blood vessel 5. The probe tip 17 is pulled through the stent region and the stent struts are imaged by pulling the probe tip 17 and the rest of the data collection probe 7 through the lumen 14. These struts can create shadows as they are imaged. The probe 7 is in optical communication with the OCT system 10. The OCT system or subsystem 10 connected to the probe tip 17 via the optical fiber 15 can include a light source such as a laser, an interferometer with a sample arm and a reference arm, various optical paths, a clock generator, a photodiode, and other OCT system components.

[0043] In one embodiment, the light receiver 31, such as a balanced photodiode-based system, can receive light exiting from the probe 7. The computing device 40, such as a computer, processor, application specific integrated circuit (ASIC), or other device, can be part of the OCT system 10 or can be included as a separate subsystem in electrical or optical communication with the OCT system 10. The computing device 40 can include memory, storage devices, buses, and other components suitable for processing data, as well as software 44, such as image data processing stages configured for side branch detection, for candidate stent strut selection or identification, for candidate stent strut shadow region detection, for correlation and comparison for stent image data visualization, and for pullback data collection discussed below. The software modules 44 can include a shadow detection module as described herein, as well as associated processes and steps.

[0044] In one embodiment, the computing device 40 includes or accesses software modules or programs 44, such as a side branch detection module, a lumen detection module, a stent detection module, a stent strut verification module, a candidate stent strut identification module, and other software modules. The software modules or programs 44 can include image data processing threads or component modules thereof, as well as one or more graphical user interfaces (GUIs). These modules can be subsets of one another and arranged and connected by various inputs, outputs, and data categories. In one embodiment, the software modules or programs 44 include a shadow detection module and processes, a line projection determination module and processes, a shadow verification module and processes, as well as other processes and modules depicted and described herein, without limitation.

[0045] The present application can be implemented as one or more computer program products, i.e., one or more modules of computer program instructions encoded on a computer readable medium for execution by, or to control the operation of, data processing apparatus. The computer readable medium can be a machine-readable storage device, a machine-readable storage substrate, a memory device, or a combination of one or more of them. The term "data processing apparatus" or computing device encompasses all apparatus, devices, and machines for processing data, including by way of example a programmable processor, a computer, or multiple processors or other computing devices or data processing devices, or a combination of one or more of them. The apparatus / device can also include, in addition to hardware, code that creates an execution environment for the computer program in question, e.g., code that constitutes processor firmware, a protocol stack, a database management system, an operating system, or a combination of one or more of them.

[0046] A computer program (also known as a program, software, software application, script, or code) can be written in any form of programming language, including compiled or interpreted languages, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. A computer program does not necessarily correspond to a file in a file system. A program can be stored in a portion of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document), in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files that store one or more modules, sub programs, or portions of code). A computer program can be deployed to be executed on one computer or on multiple computers that are located at one site or distributed across multiple sites and are interconnected by a communication network.

[0047] The processes and logic flows described in this specification can be performed by one or more programmable processors executing one or more computer programs to perform functions by operating on input data and generating output. The processes and logic flows can also be performed by special purpose logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application specific integrated circuit), and that processor(s) can be implemented as special purpose logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application specific integrated circuit).

[0048] Processors suitable for the execution of a computer program include, by way of example, both general and special purpose microprocessors, and any one or more processors of any kind of digital computer. Generally, a processor will receive instructions and data from a read-only memory or a random access memory or both. The essential elements of a computer are a processor for performing instructions and one or more memory devices for storing instructions and data. Generally, a computer will also include, or be operatively coupled to receive data from or transfer data to, or both, one or more mass storage devices for storing data, e.g., magnetic, magneto-optical disks, or optical disks. However, a computer need not have such devices.

[0049] A computer or computing device can include a machine-readable medium or other memory including one or more software modules for displaying a graphical user interface, such as an interface. The computing device can exchange data, such as monitoring data or other data, using a network, which can include one or more wired connections, optical connections, wireless connections, or other data exchange connections.

[0050] A computing device or computer can include a server computer, a client user computer, a control system, an intravascular or angiographic diagnostic system, a microprocessor, or any computing device capable of executing a set of instructions (sequential or otherwise) that specifies actions to be taken by that computing device. Further, the term "computing device" should also be understood to include any collection of computing devices that individually or jointly execute a set (or multiple sets) of instructions to perform any one or more of the software features or methods described herein or operate as a system component as described herein.

[0051] Exemplary image processing threads and components thereof can constitute one or more programs in program 44. Software modules or programs 44 receive image data and convert the image data into two-dimensional and three-dimensional views of the blood vessel, and the stent can include a lumen detection software module, a peak detection software module, a stent detection software module, a side branch detection software module, a shadow detection module, a scan line selection module, strut detection within a detected candidate stent strut shadow region module or as a source of the detected candidate stent strut shadow region module, a shadow verification module, image processing kernels and operators, and other software modules to perform the steps described herein. The image data processing threads, component software modules, and related methods, and any methods described herein are stored in memory and executed using one or more computing devices, such as processors, devices, or other integrated circuits.

[0052] As shown, in Figure 1A , display 46 can also be part of system 10 for displaying information 47, such as cross-sectional and longitudinal views of the blood vessel generated using collected image data. A representation of the stent and lumen boundaries, such as OCT or IVUS images of the stent and lumen boundaries, can be displayed to the user through display 46. Side branch detection, shadow detection, and stent detection are performed prior to displaying these features and any coding or markers with identifying markers that can be included in the displayed images. This OCT-based information 47 can be displayed using one or more graphical user interfaces (GUIs). Figure 3A and 3B Images are examples of information 47 that can be displayed and interacted with using a GUI and various input devices.

[0053] Additionally, this information 47 may include, but is not limited to, cross-sectional scan data, longitudinal scans, diameter maps, image masks, stents, poorly adhered areas, luminal boundaries, and other images or representations of the blood vessel, or potential distance measurements obtained using an OCT system and data collection probes. The computing device 40 may also include software or a program 44, which may be stored in one or more storage devices 45 and configured to identify shadows and stent struts (including struts within shadowed areas) and other vascular features using text, arrows, color coding, highlighting, contour lines, or other suitable human or machine-readable markings.

[0054] Once OCT data is obtained using a probe and stored in memory, the OCT data can be processed to generate information 47, such as cross-sectional views, longitudinal views, and / or three-dimensional views of the blood vessel along the length of the traction region, or subsets thereof. These views can be depicted as follows: Figure 3A and Figure 3B This is part of the user interface shown and otherwise described herein.

[0055] Stent testing process and related sub-processes and parallel processes

[0056] Partly, this invention relates to shadow detection methods applicable to the detection of various shadow-generating objects. Therefore, in part, the invention also relates to a method for detecting metal instruments or objects, comprising an automated method for detecting points or elements of such metal objects within each frame of intravascular recording or retraction (such as OCT retraction or IVUS retraction). The metal objects or instruments may include stents and components, stent struts, guidewires, and other metal or shadow-generating elements. In one embodiment, stent detection may include detection of tissue displacement, shadow detection, detection of struts within detected shadows, detection of struts at guidewire boundaries, and detection of struts within collateral branches. Steps for verifying shadows / struts and searches related to shadows / struts can be performed to reduce false positives. The original peak method at the line can be used ( The atPeak Line Method (NPLM) is used to perform strut testing. Figure 1B This includes an overview of these steps.

[0057] Figure 1B This is a process flow diagram for stent testing process 80. Various sources exist for input data (such as guidewire data 204, lumen boundary data 106, and side branch data 122). This can be achieved by using data such as... Figure 1AThe described probe-acquired intravascular data is manipulated and transformed to obtain these and other data sets. In one embodiment, the first step in the stent detection process is shadow detection 101. Next, in one embodiment, the next step in the process is offset detection 110. Stent struts produce shadows, and the shadow is observed in the tissue region between near offset (closer to the lumen / probe) and far offset (within the vessel wall). It is empirically observed that the shadow will appear in the tissue region between near offset and far offset. In one embodiment, the far offset amount is an approximation that separates the tissue region from the noise floor. In one embodiment, false positive analysis is performed using cross-frame validation.

[0058] This region between near offset and far offset defines the region within which to search when detecting shadows. In one embodiment, strut detection candidates are generated from the detected shadows and offsets. In this way, candidate struts are determined (115). The apposition level of the struts is generated (135), and the depicted strut position and apposition level contained in the depicted struts are displayed using color scales or other indicia that indicate the apposition level. Thus, the detected struts are displayed using a 2D or 3D display.

[0059] In one embodiment, the side branch detection process operates in parallel with the stent detection process. The strut detection in side branches is performed using a line at peak method (NPLM) (120), followed by a false positive reduction method. The NPLM process can be used to detect covered or incarcerated side branches in which a stent strut covers at least a portion of the side branch. The final strut determination is updated by the strut detection results in the side branches. Throughout the method of the image processing thread, the guidewire data is used to refine the strut search region of the image. Similarly, the lumen detection data provides information for detecting offsets (110) and calculating apposition values (135). The results (137) can be displayed in 2D or 3D as described herein.

[0060] In part, the present invention relates to embodiments of a shadow detection method or process suitable for various intravascular data analysis and diagnostic display applications. In one embodiment, the shadow detection includes various sub-steps or sub-processes, such as, for example, computing both near offset and far offset, using a locally adaptive threshold (LAT), and performing one or more shadow validation steps to reduce the incidence of false positives (FPs).

[0061] The detection of the tissue region is typically the first step in various detection methods, such as, for example, shadow detection, stent detection, and guidewire detection. Figure 2An exemplary tissue region and steps of a method of near and far offset detection are shown. The tissue region provides a search area for stent strut shadow. The method relies heavily on shadow detection to identify a region of interest for strut search. As used herein, offset refers to the distance from the center of the scan converted image to the location of the tissue region, in addition to its ordinary meaning. Near offset determines the boundary of the lumen wall. Far offset determines the boundary of detectable tissue signal.

[0062] In one embodiment, the method determines vectors for near and far offsets of the lumen. The offsets are used to determine from line projections the region of the blood vessel in which a shadow will be produced, which will be discussed in more detail below. The shadow start scan line and the shadow end scan line across the shadow region can be identified as an output of the operation of the shadow detection method of the present invention. The shadow detection method can be used in various other methods, such as stent detection and guidewire detection. The shadow detection method operates on scan lines, offsets, arrays, and vectors stored in memory of a data processing system such as that of FIG. 1 to identify candidate shadows in the tissue region using the steps and processes described herein.

[0063] For each scan line that is not within the range of a previously detected guidewire, the start-end pairs are compared, and that start-end pair having a thickness corresponding to the target thickness or the maximum thickness is retained and stored in memory as a vector of near offset values or a vector of a set of values corresponding to the lumen. Near offset can be described as the offset from the inside (catheter) to the lumen. Far offset represents the offset at which the data within the blood vessel is no longer imaging tissue, but the data is indicative of the presence of background noise. The ability to process tissue stops as the background noise is approached due to noise and signal attenuation. Figure 3C and 3D Near offset / far offset and binary median mask on a typical image are shown.

[0064] In one embodiment, start-end pairs are generated from the binary mask to obtain an estimate of where tissue backscatter occurs relative to the lumen or other region, such as a shadow region. Because speckles and other artifacts can appear in the mask, a weighting of the start-end pairs is used in the binary mask (205). The weighting step filters out some noise or artifacts to find the main portion of the scan line that corresponds to tissue. In the method 200 of FIG. 2, the start-end pairs are weighted by a function of the distance from the center of the scan converted image to the location of the tissue region. The function is a Gaussian function, but other functions can be used. The function is a function of the distance from the center of the scan converted image to the location of the tissue region. The function is a Gaussian function, but other functions can be used. Figure 2 In one embodiment identified in the method 200 of FIG. 2, the start-end pairs are used to determine near and far offsets. Additional details regarding start-end pairs are described in U.S. Patent No. 9, 138, 147, the details of which are incorporated by reference herein in their entirety.

[0065] As Figure 2As shown, start-stop pairs determined in the binary mask are grouped by assigning weights (105) to each start-stop pair to determine near-tissue offsets and far-tissue offsets. The start of the start-stop pair with the highest weight determines the offset to the tissue mask and is stored in memory as a vector associated with the near offset value. The far offset will be calculated later in the process or in parallel. An ellipse is fitted to the near offset value (210) and the start-stop pairs (220) are trimmed by this method to the following: the termination of the start-stop pair is outside the ellipse and its thickness is less than a percentage of the standard deviation of the thickness of all start-stop pairs. Figure 2 The step or stage 220 shows the start-end pair (SS pair) and the standard deviation (STDEV).

[0066] In one implementation, the shadow detection software module and associated method reweight (225) or refine the start-end pairs by reweighting the remaining start-end pairs and retaining the start-end pairs with the maximum weight (230). A spline is fitted to the filtered list of reweighted near offsets (235). The near offset is calculated using the fitted spline. In one implementation, the far offset is determined to be between the near offset and the background noise (240), and the far offset is located at or above the background noise. In one implementation, the far offset is calculated as the near offset plus the local average thickness of the vessel wall. In one implementation, the local average thickness may be scaled or otherwise adjusted based on the location of the background noise or other factors.

[0067] about Figures 3A-D The intravascular image and its binary mask illustrate a tissue offset detection method. Figure 3A This is an example of a single cross-section of a vascular record with a newly implanted metal stent. Figure 3A and Figure 3B An OCT image generated using a scan line obtained by pulling back along an artery is shown. Figure 3C and Figure 3D The corresponding binary median masks are shown, as well as the corresponding images for the above images. Figure 3A The shadows in the image come from the support struts that obstruct the light signal.

[0068] like Figure 3A Images and their use Figure 3B As shown in the mask, the shadow unfolds into a dark fan shape relative to the stent struts surrounding the lumen boundary. Not all the shadow originates from the stent struts. The largest shadow comes from the guidewire, such as... Figure 3AThe near offset and far offset are shown as curves / histograms and indicated by the white curved arrows shown to illustrate that the shadow detection method can tolerate large mask notches and artifacts. In one embodiment, the offsets can be considered to be representative of the limit of the lumen border near the probe adjacent to the lumen (near offset) and the penetration depth within the vessel wall (far offset).

[0069] For example, large shadows caused by stent struts do not affect the far offset. In one embodiment, the offset is also calculated over the guidewire shadow. As the imaging signal decays, the near offset lies on the vessel lumen border and the far offset delimits the visible tissue region. In one embodiment, the near offset is the closest point of tissue to the center of the intravascular data collection probe. In one embodiment, as the tissue signal decays, the far offset scales slightly and tends to "float". In one embodiment, this is a result of the far offset scaling.

[0070] Figures 3A-D An illustration of the near offset and far offset determined using the tissue offset detection software module is provided. These near and far offsets are inputs that can be manipulated and transformed by the shadow detection software module. The binary image module is used to generate binary images of Figure 3A and 3B Figure 3C and Figure 3D

[0071] The binary images are used as a preprocessing step to determine the near offset and far offset. The near offset and far offset determined for a scan line are then used to determine the values in the line projection. In addition, the line projection is used to generate a locally adaptive threshold that varies over different scan lines. The locally adaptive threshold can be compared to the projection values to determine shadow regions. In one embodiment, as opposed to the LAT, a constant threshold can be used; however, using a constant threshold can find some candidate shadow regions and miss others. Thus, in one embodiment, the locally adaptive threshold is preferred.

[0072] The next step in the stent detection method is to detect the shadow corresponding to a given shadow source, such as a strut point, guidewire, catheter, or other object. Figure 4 The steps involved in shadow detection are summarized. The first step is to calculate the line projection. Each value of the line projection refers to a subset of the scan line that is processed using one or more operations. In one embodiment, the operations can include sorting the components of the scan line, excluding components, and / or selecting the highest intensity value of the scan line.

[0073] ​​In one embodiment, line projections are determined by performing one or more operations on the portion of each scan line between the near offset and the far offset to generate a value indicative of an intensity value of the scan line. The intensity value can correspond to a level of intensity of shadow, tissue, lumen, or non-shadow. The operations can include averaging, summing, sampling, selecting, or other statistical operations such as order statistic operations, median operations, mean operations, mode operations, or other operations performed with respect to the scan lines and their components or values associated therewith. Samples are obtained on any given scan line (or one or more scan lines), and tissue intensity information is extracted from such samples of the given scan line (or one or more scan lines). Intensities that are occluded by objects that produce shadows then have an associated lower intensity with respect to the intensities of tissue that includes the scan line or samples obtained with respect to such scan lines. In one embodiment, the samples are intensity values or another value obtained with respect to the scan lines.

[0074] In one embodiment, line projections are searched to determine whether they include a shadow, tissue, lumen, non-shadow region, or combinations thereof. Values of a locally adaptive threshold are compared to values of the line projections to assist in shadow detection as described herein. In one embodiment, line projections are evaluated with respect to the LAT value or the line projections are searched to determine whether they include a shadow. A final step is to validate detected shadows. Performing validity checks with respect to candidate shadows using software improves the accuracy of shadow detection and other related detection methods that use shadow detection such as stent strut detection and guidewire detection.

[0075] Figure 4 A high level of shadow detection steps or stages 250 are shown that occur between the offset calculation 255 described herein and the addition of detected struts to the intravascular data set that includes information about detected shadows. In one embodiment, these steps include calculating line projections 257, performing shadow searching 260, performing shadow validation 265, and performing shadow refinement processing 270 with respect to initially detected candidate shadows. One shadow related step is complete; the shadow is evaluated to determine the physical object that has been detected 275. Thus, the shadow can be identified as corresponding to a strut, guidewire, other object, or the source of the physical entity that produced the detected shadow can be unknown. Additional details related to these steps are described in more detail below.

[0076] Embodiments of calculating line projection methods

[0077] The near offset and the far offset of the tissue mask are used to calculate the line projection between the near point and the far point of each scan line. In one embodiment, the pixels in the line that are bounded by the near offset and the far offset are arranged and the percentage of lower pixel values is averaged. Thus, for each scan line, if all pixels are considered in the ensemble, the average pixel value can be determined. A low value relative to this average (for all pixels) or relative to another average obtained using a subset of the pixels of the scan line (the average of the pixels below a certain intensity threshold) can be used to identify a candidate shadow. A fraction of the average tissue intensity, such as 50% of the average tissue intensity, can be used as an intensity floor above which shadows are identified using the LAT-based method. This fraction of the average tissue intensity used as the floor can range from about 20% to about 80% to select candidate shadows based on Figure 5 the troughs in the smoothed line projection.

[0078] A ranking process is performed to increase the likelihood that the brightest pixel, which can correspond to strut distraction, does not obscure a shadow on the intensity projection. Once the projection for each line is calculated, a filter, such as a moving average filter, is used to smooth the entire line projection. Figure 5 A typical example of determining the line projection is shown. In Figure 5 the data curve 300 is plotted relative to the intensity axis and the scan line axis as shown. The first horizontal line at an intensity level of about 50 is the average tissue intensity. The second horizontal line at an intensity level of about 25 is about half of the average tissue intensity.

[0079] In Figure 5 the smoothed line projection is plotted along with the original projection and the local adaptive threshold, LAT. The LAT is below the average tissue intensity and above and below half of the average tissue intensity at different points. As shown, various shadow events correspond to the smoothed projection falling below the LAT curve. As shown, the original projection is jagged and above or below or overlapping the smoothed line projection. In Figure 5 the average tissue intensity and half of the average tissue intensity are also shown. As shown, the LAT is above the LAT precursor.

[0080] As Figure 5The candidate shadows are labeled with numbers 1 through 9 as shown. The smoothed line projections are shown relative to the original line projections (unsmoothed data oscillating relative to the smooth data with spikes and jagged points). The asterisk mark at point 9 is a true shadow that is not detected by the initial operation of the LAT method. A secondary or backup detection method using relative extreme data can be used in parallel with the LAT-based detection method to detect shadows such as the one associated with point 9. As shown, point 9 is higher than the LAT, while the other detected shadows 1-8 have projection intensity values lower than the LAT, so shadows 1-8 are indicated as shadows.

[0081] In one embodiment, each shadow has a starting shadow line. For example, shadow 2 has an approximate starting line 150 and shadow 8 has an approximate ending line 455. As shown around scan line 350, the tissue values are lower, so the LAT is lower relative to the scan line intensity at about scan line 75, and as a result the LAT varies based on the intensity changes in the scan line and line projections. In one embodiment, the tissue intensity values local to the scan line are used to calculate the LAT at that scan line.

[0082] Embodiments and features of shadow searching

[0083] In one embodiment, the shadow searching method uses a local adaptive threshold (LAT) on the line projections to determine the shadow regions. The determination of the LAT improves the accuracy of the shadow searching method. The method calculates a LAT for each line, as shown in the line labeled LAT in Figure 5 Figure 6 is a flowchart 350 showing the process of an exemplary shadow searching method according to an embodiment of the application.

[0084] In one embodiment, the method first calculates the overall range of projection intensity values. The next step calculates the mean of the tissue as the mean of all values whose projection values are greater than the middle of the range of projection intensity values. The following steps use the mean of the tissue intensity (MT). For each scan line L, create a value in the line projection. The local mean tissue (LMT) is created by aggregating the projection values within a certain radius about a given scan line and calculating the mean of the projection values from that region. In one embodiment, these projection values are in the upper half of the range.

[0085] In one embodiment, a local projection is generated for each line (305). The list of local projection values is sorted by MT and capped. The local mean tissue (LMT) value is calculated as the mean of the local projection values in the upper half of the range of local projection values. The LAT for line L is calculated as half of the LMT (310). Finally, the LAT is smoothed using a moving average filter or other smoothing operator or filter.​

[0086] The scan lines are searched and if the projection falls below the smoothed (315) LAT of the line, the line is marked as belonging to a shadow. If the previous line was a non-shadow line, the method determines a new shadow. The method also checks for the special case of a shadow that wraps around the edge of the image. The scan lines correspond to the polar representation of the blood vessels.

[0087] As a result, when the scan line zero (or other arbitrary origin) is adjacent to scan line 500 (or other final scan line), the scan lines are wrapped with image data. Thus, when evaluating shadows across the first and last scan lines in a set of data collected within a set of blood vessels, the polarity of the scan lines and their range of wrapping can be considered. In the case of wrapping, shadows on the edge of the image are merged, meaning that the shadow at scan line 1 and the shadow at scan line 500 (or whatever the last scan line is numbered as) are considered to be a single shadow given the adjacent orientation of such scan lines.

[0088] Further, to detect shadows using the LAT, as another parallel shadow detection method or second shadow detection method, the relative extreme / local minimum points on the smoothed line projection are used as an additional detection method to identify other types of shadows. The use of relative extrema represents a method of detecting shadows that is performed in addition to the LAT method to identify shadows that can be missed by the LAT-based method. In one embodiment, the LAT method is the primary or first method and the use of local extrema or minima to detect shadows is the secondary or second method (or vice versa).

[0089] In one embodiment, a separate local minimum search is performed on non-shadow regions to identify shadows that are not dark enough to fall below the LAT. A local minimum or other relative extreme is identified for a line if there is a percentage of the values of the smoothed line's projection that are greater than the value of the trough (or peak, depending on implementation details). In one embodiment, a local minimum or other relative extreme is identified if it exists within a windowed search radius (such as 10 lines, 20 lines, or 30 lines) before and after evaluating each scan line.

[0090] In one embodiment, the windowed search radius is a valley-to-peak search radius. In one embodiment, the search is for a valley bounded by two peaks and uses that as a label to indicate a shadow. The star at point 9, which was not detected using the LAT-based method, can be evaluated by looking at the 20 lines in front of point 9 and the 20 lines behind point 9 to determine if the intensity pattern of the smoothed projection experienced a change including a valley with a peak on either side. In one embodiment, the detection of this feature can be used to find shadows missed by the LAT-based method as part of a secondary shadow detection method. The presence of a valid local minimum within a search window on either side of each scan line searched (which can be all scan lines) can correspond to another detected shadow. In addition, the occurrence of this pattern can be used to identify a shadow if the difference between the minimum of the smoothed projection and the maximum of the projection in the search window exceeds a threshold.

[0091] In Figure 5 In one embodiment, the shadow labeled with a star as #9 is an example of a shadow detected by a shadow search step, such as a fuzzy shadow search step. In one embodiment, for each scan line or subset thereof, a valley-to-peak search radius of approximately 10 scan lines out of 504 scan lines is searched, which represents approximately 7 degrees. In one embodiment, the valley-to-peak search radius can be from about 5 scan lines to about 40 scan lines.

[0092] In one embodiment, the implementation of a shadow search step or process provides additional sensitivity such that the search process detects fuzzy shadows that are too bright to fall below the LAT. Thus, a fuzzy shadow can have an intensity above the LAT threshold, but still constitute a shadow region.

[0093] In one embodiment, the primary or first shadow search process based on LAT and the secondary or second relative extremum / peak valley search can further include a step by which the process refines the start and end positions to the location of the maximum slope on the projection. This refinement can include one or more performed slope-related applications or searches. For example, in one embodiment, slope measurements are used to fine-tune the shadow start / stop lines by identifying the true center of the edge values, such as those corresponding to the shadow start scan line or the shadow stop scan line.

[0094] In one embodiment, no adjustment is made to the position for a shadow consisting of a single scan line. Star 9 is not captured as a shadow because the intensity value is above the LAT. In one embodiment, a slope measurement is used to produce an improved estimate for each starting line and ending line of each shadow. The slope measurement is used to select the edges of the shadow start and end. In one embodiment, the edge selection improves the accuracy of the verification step. As an example, as shown in FIG. 3, shadow 3, which is roughly around scan line 300, has a maximum or steep slope that occurs before the smoothed projection falls below the LAT, and a similar maximum or steep slope that slowly increases as the smoothed projection goes above and through the LAT. Figure 5

[0095] In one embodiment, by scanning using a window of scan lines or other radius of windows, the slope of the projection can be calculated and relative extrema and their changes, such as with respect to shadow 3 and shadow 9, can be used to verify shadows or identify shadows that are not detected by the LAT-based method, such as shadow 9. In addition, the slope can be used in cases where the LAT method only detects a portion of a shadow. Using a slope measurement based on each scan line helps to better estimate the edges of a shadow that span multiple scan lines by detecting the edges corresponding to the shadow start scan line and the shadow end scan line.

[0096] Shadow verification embodiments and features

[0097] In one embodiment, shadow verification follows the shadow search and is used to reduce false positives and to ease the burden on subsequent strut deflection detection methods that operate on the output from the shadow search or shadow detection using the LAT and smoothed line projections. In one embodiment, while the shadow detection attempts to determine that a shadow is present on a scan line, the verification attempts to confirm the presence of a true shadow edge. The shadow start scan line and the shadow end scan line define the edges of the region of interest, such as where the shadow starts and ends from the reference frame of the blood vessel imaging probe.

[0098] In one embodiment, the candidates from the previous software module processing step are initially marked as valid by default, but are marked as invalid if they do not pass verification. In one embodiment, the process of marking the lines relative to the method of Figure 6 The process of marking lines includes a one-dimensional region marking or connected component analysis. The marking process can include searching for regions or groups of scan lines where the smoothed projection falls below the LAT. Each region defines a different shadow. Thus, the LAT can be used to evaluate Figure 5 ​corresponding to shadow 1 through shadow 9. As noted above, while shadow 9 is omitted, a heuristic search method can be used to identify a sub-search looking for relative extrema (where there is a trough where intensity drops and climbs back up) to identify all valid shadows or at least certain classes of shadows that the LAT method cannot identify.

[0099] The first validation test is based on the shadow width determined by the distance between the shadow start line and the shadow end line located at near offset. If the width or other shadow dimension of the shadow is greater than a predetermined dimension, which represents the maximum shadow width (or other shadow dimension) associated with the type of object that produced the shadow, the shadow is flagged as invalid. As a result, the shadow width / dimension of a stent strut, guidewire, or other shadow producing object can be designated as a basis for rejecting shadows that are not associated with one or more of the aforementioned objects. In this case, all subsequent validation steps are skipped.

[0100] In one embodiment, shadows that exceed the width criteria typically correspond to guidewires or side branches. Thus, in one embodiment, the validation process includes a step of excluding guidewire shadows and / or side branch shadows from the set of candidate stent strut shadows. The shadow width can vary depending on what is being searched for or what is excluded from the search. If stent struts are being sought, shadows that exceed the stent strut shadow dimension, for example, can be excluded.

[0101] If the shadow meets the maximum width criteria or other selection threshold or criteria, the candidate shadow in the image is selected for the validation phase. In this phase, the validation method uses the application of an operator, such as an edge detection kernel, to confirm the shadow start-end edges (across the scan line of the shadow). In one embodiment, the application is a convolution application. For shadows with well-defined edges, various kernels or other image processing / edge detection operators are used. In one embodiment, one or more Prewitt kernels, including one or more Prewitt kernel features, are used.

[0102] Figures 7A to 7C is an example of an operator, such as a [1 x N] image processing kernel, that can be applied to an image to detect the start edge or end edge of a shadow, and filtered to detect narrow gap shadows associated with intensity or other value of interest. In one embodiment of the present invention, Figures 7A to 7CThe operators shown can be applied to the 1-D projections derived from the 2-D intravascular images to find the start and end lines of the shadow. In general, the verification can be performed using 2-D image processing of the 2-D image data with 2-D kernels. In one embodiment, initially the projections are generated first along the scan lines, then the system uses 1-D filters as operators instead of 2-D operators such as kernels because it provides computational advantage in terms of speed.

[0103] As an example kernel, Figure 7A A diagram depicting a start edge filter kernel is shown. In one embodiment, the start edge finder kernel can be a vector or matrix of the form [1 1 0 -1 -1]. As an example kernel, Figure 7B A diagram depicting a stop edge filter kernel is shown. Other kernels and operators can be used that are designed to detect or filter edges of the shadow or other parts or features of the shadow. Thus, edge detection can be performed after candidate shadow selection and shadow exclusion (for guidewire and side branch) as a verification step to improve shadow detection accuracy.

[0104] These kernels or other operators are applied to detect edges in the scan lines of the polarity image within the region in the intensity image (ROI) determined by the shadow start / stop lines and the corresponding near / far offsets. The output of the filtering operation is projected along the sample line (in the filtered output image). Thus, the average effect is obtained without the need for a full 2-D kernel, shortening the computation time. In one embodiment, instead of a 2-D kernel, a 1-D kernel is used in a 2-D convolution. However, a full kernel can be used in some embodiments of the invention. The projected signal is searched for a peak to determine if there is a valid edge. The full extent of the ROI or a subset thereof is used for the initial verification attempt. If at least one edge passes the verification, the shadow is considered valid. In one embodiment, a one-dimensional kernel is used to find the edges of the shadow region, such as to identify the shadow start scan line and the shadow stop scan line. A projection is generated along the scan line, and then a one-dimensional edge detection operator is applied to the projection to identify the edges.

[0105] Another scenario that is important for evaluation occurs when the shadow is very thin (1 to 2 scan line width). In one embodiment, these shadows are similarly verified by a notch filter kernel. Figure 7C A diagram showing a notch filter kernel [1 1 -4 11] is shown. The notch filter effectively searches for narrow or thin shadows (which are 1 or 2 scan line width) and helps select them so that they are not ignored or excluded from the process.

[0106] In one embodiment, invalid shadows undergo a second validation step. The second validation step breaks the ROI into equivalent blocks in the sample direction. Then, the first validation technique described earlier is applied to each block again. If an individual block passes validation, the shadow is re-labeled as valid. In this way, no fuzzy shadows are missed by the imaging processing steps described herein for shadow detection and subsequent stent detection processing.

[0107] Additional shadow refinement / validation

[0108] Validated shadows are further distinguished by comparing each shadow to all other shadows on the frame. Overlapping shadows are merged into a single shadow and duplicate shadows are removed. Shadows that failed validation in the previous section will be ignored in this refinement step. This method accounts for shadows that wrap around the image.

[0109] Some portions of the detailed description are presented in terms of algorithms and symbolic representations of operations on data bits within a computer memory. These algorithmic descriptions and representations are the means used by those skilled in the computer and software related arts to most effectively convey the substance of their work to others skilled in the art. In one embodiment, an algorithm is here, and generally, conceived to be a self-consistent sequence of operations leading to a desired result. The operations performed as the steps of an algorithm or otherwise described herein are those requiring physical manipulations of physical quantities. Usually, though not necessarily, these quantities take the form of electrical or magnetic signals capable of being stored, transferred, combined, transformed, compared, and otherwise manipulated.

[0110] The algorithms and displays presented herein are not inherently related to any particular computer or other apparatus. Various general purpose systems can be used with programs in accordance with the teachings herein, or it can prove convenient to construct more specialized apparatus to perform the required method steps. The required structure for a variety of these systems will appear from the description given.

[0111] Embodiments of the present application can be realized in a variety of forms, including, but in no way limited to, computer program logic for use with a processor (e.g., a microprocessor, microcontroller, digital signal processor, or general purpose computer) computer program logic for use with a programmable logic device (e.g., a Field Programmable Gate Array (FPGA) or other PLD), discrete components, an integrated circuit (e.g., an Application Specific Integrated Circuit (ASIC)), or any other device(s) including any combination thereof. In this example, some or all of the processing collected by the OCT probe, IVU probe, and other imaging and target monitoring devices and processor-based systems is implemented as a set of computer program instructions that is converted into a computer executable form, stored in a computer readable medium, and executed by a microprocessor under the control of an operating system. Thus, user interface instructions and triggers to complete a pullback or fusion request are converted, for example, into instructions that are understandable by a processor that generates OCT data, performs image processing using the various and other features and embodiments described above.

[0112] Computer program logic implementing all or part of the functionality previously described herein can be embodied in various forms, including, but in no way limited to, a source code form, a computer executable form, and various intermediate forms (e.g., forms generated by an assembler, compiler, linker, or locator). Source code can include a series of computer program instructions implemented in any of various programming languages (e.g., an object code, an assembly language, or a high-level language such as Fortran, C, C++, JAVA, or HTML) for use with various operating systems or operating environments. Source code can define and use various data structures and communication messages. Source code can be in computer executable form (e.g., via an interpreter), or source code can be converted (e.g., via a translator, assembler, or compiler) into computer executable form.

[0113] A computer program can be fixed in any form (e.g., source code form, computer executable form, or an intermediate form) either permanently or transitorily in a tangible storage medium, such as a semiconductor memory device (e.g., a RAM, ROM, PROM, EEPROM, or Flash-Programmable RAM), a magnetic memory device (e.g., a diskette or fixed disk), an optical memory device (e.g., a CD-ROM), a PC card (e.g., PCMCIA card), or other memory device. The computer program can be fixed in any form in a signal that is transmittable to a computer using any of

[0114] The hardware logic (including programmable logic for use with a programmable logic device) that implements all or part of the functionality previously described herein can be designed using traditional manual methods, or can be designed, captured, simulated, or documented electronically using various tools, such as a computer aided design (CAD) simulation or an hardware description language (e.g., VHDL or AHDL) simulation.

[0115] The hardware logic (including programmable logic for use with a programmable logic device) that implements all or part of the functionality previously described herein can be designed using traditional manual methods, or can be designed, captured, simulated, or documented electronically using various tools, such as a computer aided design (CAD) simulation or an hardware description language (e.g., VHDL or AHDL) simulation.

[0116] Various examples of suitable processing modules are discussed in greater detail below. As used herein, a module refers to software, hardware, or firmware suitable for performing a particular data processing or data transmission task. In one embodiment, a module refers to a software routine, program, or other resident application residing in memory that is adapted to receive, transform, route, and process instructions or various types of data such as angiogram data, OCT scan data, FFR data, IVUS data, registration table data, peaks, offsets, line projections, scan lines, local minima, local maxima, shadows, pixels, intensity patterns, and other information of interest.

[0117] The computers and computer systems described herein can include computer- readable media operatively associated therewith, such as memory for storing software applications used in obtaining, processing, storing, and / or communicating data. It will be appreciated that such memory can be internal, external, remote, or local with respect to its operatively associated computer or computer system.

[0118] Memory can also include any means for storing software or other instructions, such as for example a hard disk, an optical disk, floppy disk, DVD (digital versatile disc), CD (compact disc), memory stick, flash memory, ROM (read only memory), RAM (random access memory), DRAM (dynamic random access memory), PROM (programmable ROM), EEPROM (extended programmable ROM), and / or other like computer-readable media.

[0119] Generally, a computer readable storage medium associated with the embodiments of the application described herein can include any storage medium capable of storing instructions which are executed by a programmable device. Where applicable, the method steps described herein can be implemented or performed in relation to a stored instruction or in relation to an instruction stream. According to embodiments of the application, the instructions can be software implemented in various programming languages such as C++, C, Java, and / or various other types of software programming languages that can be applicable to create instructions.

[0120] Aspects, embodiments, features, and examples of the present application are considered illustrative, and not limiting, of the present application, which is defined only by the claims. Other embodiments, modifications, and uses will be apparent to those skilled in the art.

[0121] The use of headings and sections in the present application is not meant to limit the present application; each section can apply to any aspect, embodiment, or feature of the present application.

[0122] Throughout this application, where compositions are described as having, including, or comprising specific components, or where processes are described as having, including, or comprising specific process steps, it is contemplated that compositions of the present teachings can also consist essentially of, or consist of, the recited components, and that the processes of the present teachings can also consist essentially of, or consist of, the recited process steps.

[0123] In this application, where an element or component is said to be included in the presence of, or selected from a list of recited elements or components, it should be understood that in

[0124] The use of the term "including" or "having" should generally be understood as open-ended and non-limiting, unless otherwise explicitly stated.

[0125] The use of the singular herein includes the plural (and vice versa) unless otherwise expressly stated. Additionally, the singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise. Further, the use of the term "about" in connection with a numerical value throughout this document refers to ±10% variation from the nominal value, unless otherwise expressly stated.

[0126] It should be understood that the order of steps or order of performing certain actions is immaterial so long as the present teachings remain operable. Moreover, two or more steps or actions can be conducted simultaneously.

[0127] Where a range of values or list of values is provided, it is intended to encompass each and every value within the range or list, whether specifically listed or not, and vice versa. Furthermore, a smaller range or list of values is intended to encompass any value or values within that smaller range or list, and vice versa. A list of example values or ranges is not intended to be an exhaustive list of possible values or ranges.

Claims

1. A method for identifying shadows in an intravascular image, the method comprising: Use an intravascular diagnostic system to access one or more intravascular datasets, each of which includes multiple scan lines; Determine the local line projection value for each scan line in each dataset; Determine the local average tissue value, where the local average tissue value is the average of two or more local line projection values ​​within a specific radius of a given scan line; Determine the smooth local adaptive threshold; and Based on the comparison between the local line projection value and the smooth local adaptive threshold, when the local line projection value is less than the smooth local adaptive threshold, one or more shadows representing features of interest in one or more intravascular datasets are identified.

2. The method according to claim 1, wherein, Determining the smooth local adaptive threshold includes: Determine the local adaptive threshold; and Apply smoothing operators or filters.

3. The method according to claim 2, wherein, The smoothing operator or filter is a moving average filter.

4. The method according to claim 2, wherein, The local adaptive threshold is half the local average tissue value.

5. The method according to claim 1, wherein, Each local line projection is determined using near tissue offset and far tissue offset, the near tissue offset corresponding to the offset to the vascular lumen boundary, and the far tissue offset corresponding to the offset at the background noise.

6. The method according to claim 1, further comprising: Based on the comparison between the local minimum or relative extreme value and the smooth local adaptive threshold, one or more shadows are identified when the local minimum or relative extreme value is greater than the smooth local adaptive threshold.

7. The method according to claim 6, further comprising: Identify local minimum or local maximum values ​​within the windowed search radius.

8. The method according to claim 7, wherein, The windowed search radius is the search radius from the trough to the peak.

9. The method according to claim 1, further comprising: Estimate multiple slope values ​​associated with the search window around each scan line to identify changes in slope that indicate the edges of the shaded area.

10. The method of claim 9, further comprising: Identify the edge value of the edge of the shaded region, where the edge value corresponds to the start or end scan line of the shade.

11. The method according to claim 1, further comprising: The verification test is performed based on the shadow width, which is determined by the distance between the shadow start scan line and the shadow end scan line.

12. The method according to claim 11, wherein, The shadow is marked as invalid when the width is greater than a predetermined size representing the maximum shadow width associated with the type of object that produces the shadow.

13. The method according to claim 11, wherein, When the width is less than a predetermined size representing the maximum shadow width associated with the type of object producing the shadow, the method further includes: Apply a kernel or edge detection operator to determine the shadow start edge and shadow end edge, wherein the shadow start edge and shadow end edge span the shadow.

14. The method according to claim 1, wherein, One or more steps of the method are implemented using a diagnostic system, the diagnostic system comprising: an input for receiving one or more intravascular datasets; one or more electronic storage devices for storing the one or more intravascular datasets; one or more computing devices in electrical communication with the input and the one or more electronic storage devices; and instructions, image filters, and image processing software modules executable by the one or more computing devices to perform one or more steps of the method.

15. A system for identifying shadows in an intravascular image, comprising: Memory; as well as One or more processors that communicate with the memory, said one or more processors being configured to: Access one or more intravascular datasets from the memory, each intravascular dataset comprising multiple scan lines; Determine the local line projection value for each scan line in each dataset; Determine the local average tissue value, where the local average tissue value is the average of two or more local line projection values ​​within a specific radius of a given scan line; Determine the smooth local adaptive threshold; and Based on the comparison between the local line projection value and the smooth local adaptive threshold, when the local line projection value is less than the smooth local adaptive threshold, one or more shadows representing features of interest in one or more intravascular datasets are identified.

16. The system according to claim 15, wherein, When determining the smoothing local adaptive threshold, the one or more processors are further configured to: Determine the local adaptive threshold; and Apply smoothing operators or filters.

17. The system according to claim 16, wherein, The smoothing operator or filter is a moving average filter.

18. The system according to claim 16, wherein, The local adaptive threshold is half the local average tissue value.

19. The system according to claim 15, wherein, Each local line projection is determined using near tissue offset and far tissue offset, the near tissue offset corresponding to the offset to the vascular lumen boundary, and the far tissue offset corresponding to the offset at the background noise.

20. The system according to claim 15, wherein, The one or more processors are further configured to: identify one or more shadows when the local minimum or relative extreme value is greater than the smooth local adaptive threshold, based on a comparison between the local minimum or relative extreme value and the smooth local adaptive threshold.

21. The system according to claim 20, wherein, The one or more processors are further configured to: identify local minimum or local maximum values ​​within the windowed search radius.

22. The system according to claim 21, wherein, The windowed search radius is the search radius from the trough to the peak.

23. The system according to claim 15, wherein, The one or more processors are also configured to estimate multiple slope values ​​associated with a search window around each scan line to identify changes in the slope that indicate the edges of the shaded region.

24. The system according to claim 23, wherein, The one or more processors are further configured to: identify edge values ​​of the edges of the shadowed regions, wherein the edge values ​​correspond to the shadow start scan line or the shadow end scan line.

25. The system according to claim 15, wherein, The one or more processors are further configured to perform a verification test based on the shadow width, wherein the shadow width is determined by the distance between the shadow start scan line and the shadow end scan line.

26. The system according to claim 25, wherein, The shadow is marked as invalid when the width is greater than a predetermined size representing the maximum shadow width associated with the type of object that produces the shadow.

27. The system according to claim 25, wherein, When the width is less than a predetermined size representing the maximum shadow width associated with the type of object producing the shadow, the one or more processors are further configured to: Apply a kernel or edge detection operator to determine the shadow start edge and shadow end edge, wherein the shadow start edge and shadow end edge span the shadow.

Images