Method and system for determining local weakening of blood vessels using 2D ultrasonic imaging

Abstract

A method and system for determining a regional weakening (RW) value in a blood vessel using ultrasonic imaging. A set of images of a portion of a given patient's body collected over time is received. A set of measurements representing the movement of the tissue of the blood vessel is received, and the set of measurements is determined based on the set of images. The movement of the wall and lumen at least partially indicates strain. A blood flow velocity value in the blood vessel is determined based on a Doppler image, and an RW parameter indicating the state of weakening in a region of the blood vessel is determined based on at least the set of measurements, thrombus burden, and blood flow velocity value. A map showing the circumference of the blood vessel including the regional weakening value and the strain value may be displayed.

Description

Technical Field

[0001] Cross - reference to Related Applications This application claims the priority of U.S. Provisional Patent Application No. 63 / 369,813, filed on July 29, 2022.

[0002] This technology relates to the field of medical imaging. More precisely, this technology relates to methods and systems for performing regional weakening (RW) analysis of blood vessels using two - dimensional (2D) ultrasound imaging.

Background Art

[0003] Regional weakening (RW) analysis, sometimes referred to as regional rupture potential (RRP), enables the evaluation of blood vessels based on parameters that correlate with regional weakening, dilation, and rupture of blood vessels, and also provides a theoretical basis for clinical decisions by performing calculations based only on images collected by medical imaging devices. For the aorta, regional weakening is called regional aortic weakening (RAW) analysis, and RAW analysis identifies areas of aortic wall weakness, including aortic aneurysms, in order to predict growth and potential rupture and enable physicians to treat the appropriate patients at the appropriate time. RAW analysis is currently performed using dynamic computed tomography (CT) and magnetic resonance imaging (MRI) scans. These scans are segmented to generate accurate three - dimensional (3D) and four - dimensional (4D) models of the aorta that are analyzed to measure regional strain, wall shear stress, and thrombus burden. These three measurements are combined to generate a RAW index. Physicians are presented with detailed aortic maps showing these variables and the RAW index. Regional weakening values for other blood vessels, such as the common iliac artery and visceral arteries, can be determined in a similar manner when the imaging resolution is sufficient.

[0004] RAW analysis using ultrasonic imaging would seem to be more convenient for both patients and physicians, less costly, and not use ionizing radiation or toxic contrast agents, so it would be preferable to either CT or MRI imaging. Ultrasonic imaging has a higher resolution than either CT or potentially toxic MRI, thereby enabling accurate measurement of aortic wall thickness. However, RAW analysis using ultrasonic imaging is technically challenging compared to CT and MRI imaging. Ultrasonic imaging does not generate evenly spaced image slices like CT and MRI, making it difficult to generate accurate 3D and 4D models of the aorta. Since the ultrasonic transducer is handheld, the angle of the image with respect to the body and aorta is variable, again making it difficult to create an accurate model. Additionally, ultrasonic images have numerous artifacts including reflections, shadowing, refraction, etc., and the resolution within the ultrasonic image varies with the depth of the structure.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Patent Document 2

Non-Patent Documents

[0006]

Non-Patent Document 1

Non - Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0007] The objective of this technology is to improve at least some of the inconveniences existing in the prior art. One or more embodiments of this technology can provide and / or expand the scope of methods and / or ways to achieve the goals and objectives of this technology.

[0008] The developers of the present technology understand that while 3D ultrasound imaging is available on many high-end ultrasound scanners, these scanners are extremely expensive, currently have insufficient resolution, and small fields of view. Local weakening analysis requires 4D imaging (three spatial dimensions + time to see movement), and the field of view of 4D imaging using ultrasound is too small, so it may not be useful at a reasonable frame rate in some situations.

[0009] The developers of the present technology theorized that it is possible to measure strain and intraluminal thrombus (ILT) and calculate a local weakening index using a standard 2D ultrasound scanner commonly found in medical centers.

[0010] The developers of the present technology propose to calculate a regional weakening (RW) index without constructing a 3D model of a blood vessel (e.g., the aorta). The RW index will be calculated only for axial images generated via an ultrasound probe and displayed on a screen. For example, a user or operator will move the ultrasound probe up and down the aorta while examining the area of interest, the RW index will be calculated and displayed, the diameter will be automatically measured and reported to the user. Then, the minimum diameter, maximum diameter, and equivalent circular diameter can be calculated, giving a measure of eccentricity.

[0011] The present technology enables vascular surgeons to perform local weakening analysis in their own examination rooms without having to send the patient for dynamic CT or MRI scans, thereby avoiding the radiation associated with CT scans and potentially toxic contrast agents. The present technology is compatible with existing workflows and is much faster and less expensive. The analysis can be performed in the cloud, and the software can potentially be integrated within the ultrasound scanner or the software can be run on a computer connected to the video output of a medical imaging device.

[0012] Ultrasonic imaging has a higher resolution than either CT or MRI, thereby enabling accurate measurement of the aortic wall thickness. By having a local thickness, the surgeon can evaluate the local material stiffness (based on strain and blood pressure). Furthermore, this enables improvement of the local weakening map to compensate for the loss of accuracy and 3D evaluation.

[0013] Ultrasonic imaging also has a higher temporal resolution than either CT or MRI. The higher temporal resolution may enable evaluation of the viscoelasticity in the wall (and in thrombus). The viscoelastic properties, particularly energy loss, are linked to wall degeneration in the thoracic aorta (see the paper by Chung J, Lachapelle K, Wener E, Cartier R, De Varennes B, Fraser R, Leask RL. Energy loss, a novel biomechanical parameter, correlates with aortic aneurysm size and histopathologic findings. J Thorac Cardiovasc Surg. September 2014, 148(3):1082 - 8, discussion 1088 - 9. doi:10.1016 / j.jtcvs.2014 June 21. Epub 2014 June 13. PMID:25129601), and furthermore, in the abdominal aorta, it has been demonstrated that a high RAW corresponds to a higher energy loss (see the paper by Forneris, A., Nightingale, M., Ismaguilova, A., Sigaeva, T., Neave, L., Bromley, A., Moore, R.D., Di Martino, E.S. Heterogeneity of Ex Vivo and In Vivo Properties along the Length of the Abdominal Aortic Aneurysm. Appl. Sci. November 2021, 3485. https: / / doi.org / 10.3390 / app11083485).

[0014] Accordingly, one or more embodiments of the present technology are directed to a method and system for determining local weakening of blood vessels using 2D ultrasound imaging. **Means for Solving the Problem**

[0015] In a broad aspect of the present technology, there is provided a method for determining local weakening parameters of a blood vessel of a given patient, the method being executed by at least one processing device. The method includes receiving a set of images of a part of the body of a given patient collected over time using an ultrasound imaging device, the set of images including Doppler images; receiving a set of measurements representing the movement of the blood vessel wall and lumen, the set of measurements being determined based on the set of images; determining a blood flow velocity value in the blood vessel based on the Doppler images in the set of images; generating a local weakening parameter indicating the state of weakening in the region of the blood vessel based on at least the set of measurements and the blood flow velocity value; and outputting the local weakening parameter.

[0016] In one or more embodiments of the method, the method further includes determining a radial deformation in the blood vessel by analyzing the set of images over time based on the set of measurements, before generating the local weakening parameter indicating the state of weakening in the region of the blood vessel based on at least the set of measurements and the blood flow velocity value, wherein the radial deformation at least partially indicates strain in the blood vessel, and generating the local weakening parameter is based on the radial deformation in the blood vessel.

[0017] In one or more embodiments of the method, after receiving the set of images, the method further includes detecting respective wall positions and lumen positions in the blood vessel for the set of images over time, and calculating the movement of the blood vessel wall and lumen based on the respective wall positions and lumen positions over time, thereby receiving the set of measurements.

[0018] In one or more embodiments of the present method, calculating the movement of the blood vessel wall and lumen based on each wall position and lumen position over time includes determining a thrombus thickness based on the wall position and lumen position, and generating the local weakening parameter is based on the thrombus thickness.

[0019] In one or more embodiments of the present method, a thrombus thickness of 0 indicates that no thrombus is present in the blood vessel.

[0020] In one or more embodiments of the present method, the blood vessel includes at least one of the aorta, the common iliac artery, and the visceral artery.

[0021] In one or more embodiments of the present method, the method includes generating a local weakening map based on a set of local weakening parameters and an image of the blood vessel, and outputting the local weakening map, wherein the local weakening map shows the circumference of the blood vessel including local weakening values.

[0022] In one or more embodiments of the present method, detecting each wall position and lumen position in the blood vessel for a set of images over time includes segmenting the outer wall, inner wall, and lumen of the blood vessel using a segmentation model trained on the set of images.

[0023] In one or more embodiments of the present method, the method further includes determining a minimum diameter, a maximum diameter, an equivalent circular diameter, and an eccentricity for each of the outer wall diameter, inner wall diameter, and lumen diameter.

[0024] In one or more embodiments of the method, the trained segmentation model includes a fully convolutional neural network (FCN).

[0025] In one or more embodiments of the present method, before receiving the set of images of blood vessels, the method comprises receiving a plurality of images of a given patient collected using an ultrasonic imaging device, wherein the plurality of images includes at least one Doppler image; generating a plurality of high-resolution (HR) images of the blood vessels using a resolution improvement machine learning (ML) model trained on the plurality of images, wherein the resolution of the plurality of HR images is higher than that of the plurality of images; and denoising the plurality of HR images using a trained ML denoising model to obtain a set of images.

[0026] In one or more embodiments of the present method, the trained resolution improvement ML model includes a very deep super-resolution (VDSR) neural network, and the trained noise removal ML model includes a cycle-consistent generative adversarial network (CycleGAN).

[0027] In one or more embodiments of the present method, the method further comprises generating a strain map based on the radial deformation value and the set of images of the blood vessels, and outputting the strain map, wherein the strain map shows the circumference of the blood vessel including strain values.

[0028] In one or more embodiments of the present method, the method further comprises determining the viscoelasticity of the blood vessel wall.

[0029] In one or more embodiments of the present method, the method further comprises determining the viscoelasticity of the thrombus.

[0030] In one or more embodiments of the present method, at least one processing device is operably connected to the ultrasonic imaging device.

[0031] In one or more embodiments of the method, the method further includes displaying a regional weakening map on a display operably connected to at least one processing device.

[0032] According to a broad aspect of the technology, a system for determining regional weakening (RW) of a blood vessel of a given patient is provided. The system includes at least one processing device and a non-transitory storage medium operably connected to the at least one processing device, the non-transitory storage medium storing computer-readable instructions. When the processor executes the computer-readable instructions, the processor is configured to receive a set of images of a portion of the body of a given patient collected over time using an ultrasonic imaging device, the set of images including Doppler images; receive a set of measurements representing movement of the blood vessel wall and lumen, the set of measurements determined based on the set of images; determine a blood flow velocity value in the blood vessel based on the Doppler images in the set of images; generate a regional weakening parameter indicative of a state of weakening in a region of the blood vessel based on at least the set of measurements and the blood flow velocity value; and output the regional weakening parameter.

[0033] In one or more embodiments of the system, the at least one processing device is further configured to determine a radial deformation in the blood vessel by analyzing a set of images over time based on the set of measurements before generating the regional weakening parameter indicative of a state of weakening in a region of the blood vessel based on at least the set of measurements and the blood flow velocity value, the radial deformation at least partially indicating strain in the blood vessel, and generating the regional weakening parameter is based on the radial deformation in the blood vessel.

[0034] In one or more embodiments of the present system, after receiving a set of images, at least one processing device detects, for the set of images over time, each wall position and lumen position in a blood vessel, and calculates the movement of the blood vessel wall and lumen based on each wall position and lumen position over time, thereby further configured to calculate the movement to receive a set of measurement values.

[0035] In one or more embodiments of the present system, calculating the movement of the blood vessel wall and lumen based on each wall position and lumen position over time includes determining a thrombus thickness based on the wall and lumen positions, and generating the local weakening parameter is based on the thrombus thickness.

[0036] In one or more embodiments of the present system, a thrombus thickness of 0 indicates that no thrombus is present in the blood vessel.

[0037] In one or more embodiments of the present system, the blood vessel includes at least one of the aorta, the common iliac artery, and the visceral artery.

[0038] In one or more embodiments of the present system, at least one processing device generates a local weakening map based on the local weakening parameter and a set of images of the blood vessel, and outputs the local weakening map, wherein the local weakening map shows the circumference of the blood vessel including local weakening values, and further configured to output the local weakening map.

[0039] In one or more embodiments of the present system, detecting each wall position and lumen position in the blood vessel for the set of images over time includes segmenting the outer wall, inner wall, and lumen of the blood vessel using a segmentation model trained on the set of images.

[0040] In one or more embodiments of the present system, at least one processing device is further configured to determine the minimum diameter, maximum diameter, equivalent circular diameter, and eccentricity of each of the outer wall diameter, inner wall diameter, and lumen diameter.

[0041] In one or more embodiments of the present system, the trained segmentation model includes a fully convolutional neural network (FCN).

[0042] In one or more embodiments of the present system, at least one processing device is to receive a plurality of images of a given patient collected using an ultrasonic imaging device before receiving the set of images of the blood vessel, the plurality of images including at least one Doppler image, receive the plurality of images, generate a plurality of high-resolution (HR) images of the blood vessel using a resolution improvement machine learning (ML) model trained on the plurality of images, the resolution of the plurality of HR images being higher than the resolution of the plurality of images, and further configured to perform noise removal on the plurality of HR images using a trained ML noise removal model to obtain the set of images.

[0043] In one or more embodiments of the present system, the trained resolution improvement ML model includes a very deep super-resolution (VDSR) neural network, and the trained noise removal ML model includes a cycle-consistent generative adversarial network (CycleGAN).

[0044] In one or more embodiments of the present system, at least one processing device is further configured to generate a strain map based on the radial deformation value and the set of images of the blood vessel, and output the strain map, the strain map showing the circumference of the blood vessel including strain values.

[0045] In one or more embodiments of the present system, at least one processing device is further configured to determine the viscoelasticity of the blood vessel wall.

[0046] In one or more embodiments of the present system, at least one processing device is further configured to determine the viscoelasticity of a thrombus.

[0047] In one or more embodiments of the present system, at least one processing device is operably connected to an ultrasonic imaging device.

[0048] In one or more embodiments of the present system, at least one processing device is further configured to display a local weakening map on a display operably connected to the at least one processing device.

[0049] Terms and Definitions

[0050] In the context of this specification, a "server" is a computer program that operates on suitable hardware and is capable of receiving requests (e.g., from an electronic device) via a network (e.g., a communication network) and executing those requests or causing those requests to be executed. The hardware can be a single physical computer or a single physical computer system, but neither is required for the present technology. In this context, the use of the expression "server" does not mean that any task (e.g., received instructions or requests) or any particular task is received, executed, or caused to be executed by the same server (i.e., the same software and / or hardware). That is, the use of the expression "server" means that any number of software elements or hardware devices can be involved in receiving / sending, executing, or causing to be executed any task or request, or the result of any task or request, and all of this software and hardware can be one server or multiple servers, both of which are included within the expressions "at least one server" and "server".

[0051] In the context of this specification, an "electronic device" is any computing device or computer hardware capable of operating appropriate software for the current task at hand. Thus, some (non-limiting) examples of electronic devices include general-purpose personal computers (desktop, laptop, netbook, etc.), mobile computing devices, smartphones, and tablets, as well as network devices such as routers, switches, and gateways. Note that an electronic device in this context is not excluded from acting as a server to other electronic devices. The use of the expression "electronic device" does not exclude multiple electronic devices that receive / transmit, execute, or are caused to execute any task or requirement, or the result of any task or requirement, or any step of any method described herein. In the context of this specification, a "client device" refers to any of various end-user client electronic devices associated with a user, such as a personal computer, tablet, smartphone, etc.

[0052] In the context of this specification, unless otherwise explicitly stated, a computer system may refer to, without limitation, an "electronic device", "computing device", "operating system", "system", "computer-based system", "computer system", "network system", "network device", "controller unit", "monitoring device", "control device", "server", and / or any combination thereof appropriate for the current task at hand.

[0053] In the context of this specification, the expression "computer-readable storage medium" (also referred to as "memory medium" and "storage") includes, without limitation, any nature and any type of non-transitory medium, including but not limited to RAM, ROM, disks (such as CD-ROMs, DVDs, floppy (registered trademark) disks, hard drives, etc.), USB keys, solid state drives, tape drives, etc. To form a computer information storage medium, multiple components can be combined, including two or more media components of the same type and / or two or more media components of different types.

[0054] In the context of this specification, a "database" is any structured collection of data that is independent of its particular structure, database management software, or the computer hardware on which the data is stored, implemented, or otherwise made available for use. The database can be on the same hardware as the process that stores or uses the information stored in the database, or the database can be on separate hardware, such as a dedicated server or multiple servers.

[0055] In the context of this specification, the expression "information" includes any nature or any type of information that can be stored in a database. Thus, information includes, without limitation, audiovisual works (such as images, movies, audio recordings, presentations, etc.), data (such as location data, numerical data, etc.), text (such as opinions, comments, questions, messages, etc.), documents, spreadsheets, lists of words, etc.

[0056] In the context of this specification, unless otherwise explicitly stated, an "instruction" of an information element is either the information element itself or a pointer, reference, link, or other indirect mechanism that enables the recipient of the instruction to identify the location of a network, memory, database, or other computer-readable medium from which the information element can be retrieved. For example, an instruction for a document may include the document itself (i.e., its content), or the instruction for the document may be a unique document descriptor that identifies the file relative to a particular file system, or the instruction may be some other means that leads the recipient to a network location, memory address, database table, or other location where the file can be accessed. As will be appreciated by those skilled in the art, the precision required in such an instruction depends on the degree of prior understanding of the interpretation to be given to the information exchanged between the sender and recipient of the instruction. For example, if prior to communication between the sender and recipient, it is understood that an instruction for an information element takes the form of a database key for an entry in a particular table of a given database that contains the information element, then even if the information element itself is not transmitted between the sender and recipient, all that is required to effectively convey the information element to the recipient is the transmission of the database key.

[0057] In the context of this specification, the expression "communication network" shall include telecommunications networks such as computer networks, the Internet, telephone networks, Telex networks, TCP / IP data networks (e.g., WAN networks, LAN networks, etc.). The term "communication network" includes wired networks or direct wired connections, and wireless media such as acoustic, radio frequency (RF), infrared, and other wireless media, as well as any combination of the foregoing.

[0058] In the context of this specification, words such as "first", "second", "third", etc. are used as adjectives only for the purpose of enabling a distinction between the nouns they modify, and are not used for the purpose of explaining a specific relationship between those nouns. Thus, for example, the use of the terms "first server" and "third server" does not imply a particular order, type, time series, hierarchy or ranking among / of the servers, and their use (by themselves) does not imply that a "second server" must necessarily exist in a given situation. Further, as explained in other contexts in this specification, references to "first" elements and "second" elements do not exclude the two elements from being the same actual real-world element. Thus, for example, in some cases, the "first" server and the "second" server may be the same software and / or hardware, and in other cases, they may be different software and / or hardware.

[0059] Implementation examples of the present technology each have at least one of the above-described objectives and / or aspects, but do not necessarily have all of them. It should be understood that some aspects of the present technology obtained from attempts to achieve the above objectives may not meet this objective and / or may meet other objectives not specifically set forth herein.

[0060] Additional and / or alternative features, aspects, and advantages of implementation examples of the present technology will become apparent from the following description, the accompanying drawings, and the appended claims.

[0061] For a better understanding of the present technology, as well as other aspects and further features thereof, reference should be made to the following description to be used in conjunction with the accompanying drawings.

Brief Description of the Drawings

[0062]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

DETAILED DESCRIPTION OF THE INVENTION

[0063] The examples and conditional language set forth herein are mainly to assist the reader in understanding the principles of the present technology and are not intended to limit the scope of the present technology to such specifically set forth examples and conditions. It will be understood that those skilled in the art can devise various configurations that are not explicitly described or illustrated herein but still implement the principles of the present technology and are within its spirit and scope.

[0064] Furthermore, for the sake of understanding, in the following description, a relatively simplified implementation example of the present technology may be described. As those skilled in the art will understand, various implementation examples of the present technology can be more complex.

[0065] In some cases, useful examples of changes to the present technology may also be described. This is done merely for the sake of understanding, and in this case as well, it is neither for limiting the scope of the present technology nor for describing the limits of the present technology. These changes are not an exhaustive list, and those skilled in the art can still make other changes while remaining within the scope of the present technology. Furthermore, if no examples of changes are described, it should not be construed that changes are not possible and / or that what is described is the only way to implement that element of the present technology.

[0066] Moreover, all statements in this specification that set forth the principles, aspects, and implementation examples of the present technology, as well as specific examples thereof, shall be construed to include both structural and functional equivalents thereof, whether currently known or developed in the future. Thus, for example, any block diagrams in this specification will be understood by those skilled in the art to represent conceptual diagrams of exemplary circuits implementing the principles of the present technology. Similarly, any flowcharts, flow diagrams, state transition diagrams, pseudocode, etc. can be substantially represented in a computer-readable medium and, thus, will be understood to represent various processes that can be executed by such a computer or processor, whether or not the computer or processor is explicitly illustrated.

[0067] The functionality of the various elements shown in the figures, including the functional blocks labeled "processor" or "graphics processing unit", can be provided by dedicated hardware, as well as by hardware capable of executing software in association with appropriate software. When provided by a processor, the functionality can be provided by a single dedicated processor, by a single shared processor, or by a plurality of individual processors some of which may be shared. In some non-limiting embodiments of the present technology, the processor can be a general-purpose processor such as a central processing unit (CPU), or a processor dedicated to a particular purpose such as a graphics processing unit (GPU). Moreover, the explicit use of the terms "processor" or "controller" should not be construed as exclusively referring to hardware capable of executing software, and may implicitly include, but is not limited to, digital signal processor (DSP) hardware, network processors, application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), read-only memory (ROM) for storing software, random access memory (RAM), and non-volatile storage. Conventional and / or other custom hardware may also be included.

[0068] A software module, or simply a module that is implied to be software, can be represented herein as any combination of flowchart elements or other elements that indicate execution of a process step and / or text-based description. Such modules can be executed by hardware that is explicitly or implicitly shown.

[0069] Applying these principles appropriately, the inventors then consider some non-limiting examples to illustrate various implementations of aspects of the present technology.

[0070] Referring to FIG. 1, a schematic diagram of a suitable electronic device 100 for use with some non-limiting embodiments of the present technology is shown.

[0071] Electronic device

[0072] The electronic device 100 includes various hardware components, including one or more single or multi-core processors collectively represented by the processor 110, a graphics processing unit (GPU) 111, a solid-state drive 120, a random access memory 130, a display interface 140, and an input / output interface 150.

[0073] Communication between the various components of the electronic device 100 can be enabled by one or more internal buses and / or external buses 160 (e.g., PCI bus, Universal Serial Bus, IEEE 1394 "FireWire" bus, SCSI bus, Serial ATA bus, etc.), and those various hardware components are electronically coupled to the internal bus and / or external bus 160.

[0074] The input / output interface 150 can be coupled to the touch screen 190 and / or one or more internal and / or external buses 160. The touch screen 190 can be part of the display. In some embodiments, the touch screen 190 is the display. The touch screen 190 may also be equally referred to as the screen 190. In the embodiment shown in FIG. 2, the touch screen 190 includes a touch hardware 194 (e.g., a pressure-sensitive cell embedded in a layer of the display that enables detection of physical interaction between the user and the display), and a touch input / output controller 192 that enables communication with the display interface 140 and / or one or more internal buses and / or external buses 160. In some embodiments, the input / output interface 150 can be connected to a keyboard (not shown), a mouse (not shown), or a trackpad (not shown) that enables the user to interact with the electronic device 100 in addition to or instead of the touch screen 190.

[0075] According to an implementation example of the present technology, the solid state drive 120 stores program instructions suitable for being loaded into the random access memory 130 and executed by the processor 110 and / or the GPU 111 to perform local weakening (RW) analysis of blood vessels using 2D ultrasound imaging. For example, the program instructions can be part of a library or an application.

[0076] As can be understood by those skilled in the art, the electronic device 100 can be implemented in the form of a server, a desktop computer, a laptop computer, a tablet, a smartphone, a personal digital assistant, or any device configured to implement the present technology.

[0077] System

[0078] Referring to FIG. 2, a schematic diagram of a communication system 200, referred to as system 200, is shown, and the communication system 200 is suitable for implementing non-limiting examples of the present technology. It should be clearly understood that the illustrated system 200 is only an exemplary implementation of the present technology. Therefore, the following description of system 200 shall be only an explanation of an exemplary example of the present technology. This description is neither for limiting the scope of the present technology nor for describing the limitations of the present technology. In some cases, examples of changes that may be useful for system 200 may also be described below. This is done merely to assist understanding, and in this case too, it is neither for limiting the scope of the present technology nor for describing the limitations of the present technology. These changes are not an exhaustive list, and other changes are likely possible, as would be understood by those skilled in the art. Further, if this has not been done (i.e., if examples of changes are not described), it should not be construed that the changes are not possible and / or that what is described is the only way to implement that element of the present technology. As would be understood by those skilled in the art, this may not be the case. Further, it should be understood that system 200 may, in some instances, provide a simple implementation example of the present technology, and, if so, they are presented in this manner to assist understanding. As would be understood by those skilled in the art, various implementation examples of the present technology may be more complex.

[0079] System 200 includes, among other things, a medical imaging device 210, a server 230, and a database 235 coupled across a communication network 220 via respective communication links 225 (not separately numbered).

[0080] In one or more embodiments, at least a portion of system 200 implements Picture Archiving and Communication System (PACS) technology.

[0081] The medical imaging device 210 is operated by a user 202 (e.g., a physician, sonographer, or technician) to collect medical images of a body part of a given patient 204.

[0082] Medical imaging device

[0083] The medical imaging device 210 is an ultrasonic imaging device comprising a housing 212 connected to an ultrasonic probe or transducer 214. The medical imaging device 210 is configured to generate an image of the interior of the body by using high-frequency sound waves transmitted and received by the probe 214.

[0084] In one or more embodiments, the housing 212 may include components such as one or more processing devices, a storage medium, a display, a wired and / or wireless communication interface, etc., for generating and displaying an image of the scanned anatomical region, connecting to the probe 214, and processing the reflected ultrasonic energy received by the probe 214. In some embodiments, the housing 212 may include at least a portion of the components of the electronic device 100 of FIG. 1. Thus, images and other types of data may be stored in one or more storage media of the housing 212 and / or transmitted to other electronic devices (e.g., server 230). The display may provide feedback to the user 202 when collecting an image via the probe 214 by displaying the image, collection parameters, patient information, and other additional information known in the art.

[0085] The probe 214 includes one or more ultrasonic transceiver elements and one or more transducer elements for converting an electrical signal, transmitting an ultrasonic signal toward a region of interest of a patient (e.g., a blood vessel, an organ, a joint, etc.), and receiving an acoustic reflection or echo generated by internal structures / tissues within the anatomical part, and the acoustic reflection or echo may be converted by the probe 214 back into an electrical signal and transmitted for conversion into image data.

[0086] In one or more other embodiments, probe 214 integrates at least some of the functions of the components of housing 212 and also includes a processing device (e.g., a microprocessor), firmware, and / or software associated with the processing device for operably controlling probe 214 and processing the ultrasonic energy reflected to generate an ultrasonic image. Probe 214 may include a display (such as a liquid crystal display (LCD), a light emitting diode (LED)-based display, or other type of display that provides text and / or image data to the user), or / and the image may be transmitted for display on a mobile device such as a smartphone or tablet.

[0087] In use, user 202 may place probe 214 on a portion of the body that includes a target vessel, such as the abdominal aorta, to obtain an image of the abdominal aorta, among other things.

[0088] Medical imaging device 210 is configured to, among other things, (i) collect the blood vessels of a given patient 204 over a time period according to respective collection parameters, and (ii) collect a Doppler image of the blood vessels of a given patient 204 according to respective collection parameters.

[0089] In the context of the present technology, medical imaging device 210 is configured to collect images of blood vessels such as the aorta and / or iliac artery and / or visceral artery. Medical imaging device 210 may be configured using specific collection parameters for collecting an image of a patient that includes a target blood vessel. The collection parameters (e.g., the scan mode, amplitude, frequency, and duration of the pulses emitted from probe 214) may be specified by user 202 via an input / output interface of medical imaging device 210 such as a keyboard, cursor, and / or touch screen. By way of non-limiting example, the collection parameters include transducer selection (curvilinear phased array, 2 - 5 MHz), optimization of contrast and brightness (referred to as "window" and "level"), and optimization of ultrasonic gain and focus zone.

[0090] User 202 can further perform additional operations such as measurements in the image, as well as image processing operations (e.g., filtering, enhancement, smoothing, display formatting), calculations, and annotations that can be displayed on the display screen of the medical imaging device 210 using the input / output interface.

[0091] In some embodiments, the medical imaging device 210 is configured to perform Doppler ultrasound examinations. It will be appreciated that Doppler ultrasound examinations use the Doppler effect to image the movement of tissues and body fluids and their relative velocity with respect to the probe.

[0092] The medical imaging device 210 is configured to perform color flow Doppler imaging to determine the blood flow velocity in blood vessels (e.g., the aorta). The medical imaging device 210 may be configured to generate a multicolor signal (e.g., red for flow towards the probe and blue for flow away from the probe) when performing Doppler imaging.

[0093] In one or more alternative embodiments, the medical imaging device 210 may include, or be connectable to, a workstation computer (not shown), particularly for controlling acquisition parameters and image data transmission.

[0094] In one or more embodiments, the workstation computer may be provided together with the medical imaging device 210, i.e., within the housing 212. In one or more other embodiments, the workstation computer may be implemented as a mobile device such as a smartphone or a tablet.

[0095] In one or more embodiments, the medical imaging device 210 is part of an image storage and communication system (PACS) for storing and retrieving medical images together with other electronic devices such as a server 230.

[0096] Server

[0097] Server 230 is configured to perform, among other things, (i) receive an ultrasonic image including a Doppler image, (ii) perform thickness measurement, flow turbulence measurement, and deformation measurement using the ultrasonic image, (iii) generate a radial strain value parameter or index based on the ultrasonic image, thickness measurement value, flow turbulence measurement value, and deformation measurement value, (iv) generate a regional weakness (RW) parameter or index based on the ultrasonic image, thickness measurement value, flow turbulence measurement value, deformation measurement value, and radial strain index, and (v) generate one of a radial strain map based on the radial strain index and / or an RW map based on the RW index of the circumference of the wall of the aorta.

[0098] Below in this specification, it will be described in more detail how Server 230 is configured to do so.

[0099] Server 230 can be implemented as a conventional computer server and can include some or all of the components of the electronic device 100 shown in FIG. 2. In an example of one or more embodiments of the present technology, Server 230 can be implemented as a Dell (trademark) PowerEdge (trademark) server running a Microsoft (trademark) Windows (registered trademark) Server (trademark) operating system. Needless to say, Server 230 can be implemented in any other suitable hardware and / or software and / or firmware, or a combination thereof. In the illustrated non-limiting example of the present technology, Server 230 is a single server. In an alternative non-limiting example of the present technology, the functions of Server 230 can be distributed and implemented via a plurality of servers (not shown).

[0100] The implementation of server 230 is well known to those skilled in the art of this technology. However, briefly speaking, server 230 is structured and configured with a communication interface (not shown) to communicate with various entities via communication network 220 (such as, for example, workstation computer 215 and other devices potentially coupled to network 220). Server 230 further comprises at least one computer processor (such as processor 110 or GPU 111 of electronic device 100) that is operably connected to the communication interface and structured and configured to execute the various processes described herein.

[0101] In one or more embodiments, server 230 may be implemented as electronic device 100 or may comprise components of electronic device 100 such as processor 110, graphics processing unit (GPU) 111, solid state drive 120, random access memory 130, display interface 140, and input / output interface 150.

[0102] It will be appreciated that server 230 may provide the output of one or more processing steps to another electronic device for display, verification, and / or troubleshooting. As a non-limiting example, server 230 may transmit images, calculated values, results, machine learning parameters for display on a client device configured similarly to electronic device 100 such as a smartphone, tablet, etc.

[0103] Server 230 has access to a set of machine learning (ML) models 250.

[0104] Machine learning (ML) model

[0105] The set of ML models 250 includes, among others, a resolution enhancer ML model 260, a noise removal ML model 265, a segmentation ML model 270, a flow turbulence estimation ML model 275, and a strain estimation ML model 280.

[0106] Here, the ML model is referred to as the model.

[0107] Each set of the model 250 is parameterized, among other things, by model parameters and hyperparameters.

[0108] Model parameters are the configuration variables of the model used to make predictions, and the configuration variables are estimated or learned from training data, that is, the coefficients are selected during learning based on an optimization strategy for outputting predictions. Hyperparameters are the configuration variables of the model that determine the structure of the initial model and how the initial model is trained.

[0109] It will be understood that the number of model parameters to be initialized depends, among other things, on the type of the model (i.e., classification or regression), the architecture of the model (e.g., DNN, SVM, etc.), and the model hyperparameters (e.g., the number of layers, the type of layers, the number of neurons in the NN).

[0110] In one or more embodiments, the hyperparameters include one or more of the number of hidden layers and units, the optimization algorithm, the learning rate, the momentum, the activation function, the mini-batch size, the number of epochs, and dropout.

[0111] Resolution enhancer model

[0112] The resolution enhancer model 260 is configured to improve the resolution of ultrasonic images. It will be understood that the resolution enhancer model 260 may not be used in every embodiment of the present technology.

[0113] In some embodiments of the present technology, it is contemplated that a plurality of resolution enhancer models may perform the functions of the resolution enhancer model 260.

[0114] In one or more embodiments, the resolution enhancer model 260 may be implemented as a very deep super-resolution (VDSR) neural network.

[0115] In one or more embodiments, the resolution enhancer model 260 learns to improve the resolution of ultrasonic images to obtain a high-resolution (HR) image given a low-resolution (LR) image. By using the HR image as a target (or ground truth) and the LR image as an input, the resolution enhancer model 260 is trained in a supervised manner to improve ultrasonic images.

[0116] As a non-limiting example, in an embodiment where the resolution enhancer model 260 is implemented as VDSR, the LR image is interpolated to obtain an interpolated low-resolution (ILR) image, which is input into the resolution enhancer model 260 network. The ILR image passes through (D - 1) convolutional and rectified linear unit (ReLU) layers, and then the D-th Conv (Conv.D (Residual)) is performed. The output is added to the ILR image to obtain the HR image.

[0117] In one or more embodiments, the ground truth HR image may include CT images and / or MRI images of the aorta of the same patient. In an alternative embodiment, the training may be performed using unprocessed radio frequency (RF) signals available on some types of ultrasonic devices.

[0118] The resolution enhancer model 260 is configured to output an HR ultrasonic image based on the ultrasonic image received from the medical imaging device 210.

[0119] Noise removal model

[0120] The noise removal model 265 is configured to remove noise from ultrasonic images.

[0121] In some embodiments of the present technology, it is contemplated that multiple ML models or networks can perform the function of the noise removal model 265.

[0122] It will be appreciated that noise is an unwanted signal introduced during the transmission process due to noisy channels or by the image collection process. The various types of noise that may be present in an ultrasound image include one or more of Gaussian noise, salt-and-pepper noise, speckle noise, Poisson noise, and noise resulting from the effects of air in the intestine and scattering from the metal parts of EVAR and other stents.

[0123] In one or more embodiments, the noise removal model 265 can be implemented as a Cycle-Consistent Generative Adversarial Network (CycleGAN).

[0124] The noise removal model 265 can be trained to remove noise from 2D images to improve the quality of the 2D images (i.e., resolution and / or noise filtering). The noise removal model 265 can be trained to change an image from a source domain to a target domain. The source domain includes source noisy and / or low-quality images of interest, and the target domain includes high-quality images of the same interest. The purpose of the noise removal model 265 is to learn how to generate higher-quality target domain images from noisy images in the source domain, i.e., to transfer all image characteristics from one image domain to another.

[0125] In one or more embodiments, the noise removal model 265 includes a generator network (or generator) and a discriminator network (or discriminator). The generator is configured to generate realistic images in the transformed domain (high-quality image domain). The discriminator is configured to evaluate the generated images. The generator and the discriminator can be trained simultaneously.

[0126] In one or more embodiments, the noise removal model 265 may be specifically trained to remove noise from 2D ultrasound images of the aorta.

[0127] In one or more embodiments, the noise removal model 265 is configured to remove noise from the HR ultrasound image generated by the resolution enhancer model 260. In one or more other embodiments, the noise removal model 265 is configured to first remove noise from the ultrasound image, and the ultrasound image may then be sent to the resolution enhancer model 260 to convert the noise-removed image into an HR ultrasound image.

[0128] In one or more alternative embodiments, the noise removal model 265 may be combined with the resolution enhancer model 260.

[0129] Additionally or alternatively, filtering techniques such as order statistic filters, Gaussian filters, bilateral filters, average filters, and Laplacian filters may be used.

[0130] As a non-limiting example, the noise removal model 265 may be configured to remove artifacts including reflections, shadowing, and refraction from the ultrasound image based on CT or MRI images of the same patient.

[0131] The noise removal model 265 is configured to output the noise-removed ultrasound image.

[0132] It will be appreciated that the noise removal model 265 may not be used in every embodiment of the present technology, for example, by determining that the noise level is below a threshold and / or based on device models or acquisition parameters.

[0133] Segmentation model

[0134] The segmentation model 270 is configured to perform segmentation of tissues in an ultrasonic image. In one or more embodiments, the segmentation model 270 is configured to perform semantic segmentation of tissues, where the segmentation model 270 is configured to detect (i.e., define) all boundaries and distinguish (i.e., classify) various tissue types in the ultrasonic image of blood vessels. The segmentation model 270 is a model pre-trained to perform this task.

[0135] In one or more embodiments, the segmentation model 270 is configured to receive, as input, the HR image and / or the noise-removed ultrasonic image output by the resolution enhancer model 260 and the noise removal model 265, respectively.

[0136] The segmentation model 270 is configured to segment the outer wall of a blood vessel, the inner wall of the blood vessel, the lumen, and in some embodiments, an intraluminal thrombus (ILT). Thus, in the case of an aorta, the segmentation model 270 can classify each pixel in the ultrasonic image as being one of the outer wall of the aorta, the inner wall of the aorta, the lumen, an intraluminal thrombus (ILT) (if present), and the background (e.g., pixels outside the outer wall).

[0137] In one or more embodiments, the segmentation model 270 refers to a plurality of segmentation models 270 each configured to perform a specific segmentation task. As a non-limiting example, the segmentation model 270 may include a first segmentation model configured to perform foreground and background segmentation, a second segmentation model configured to perform semantic segmentation of the lumen in the aorta, and a third model configured to perform classification of pathological tissues (e.g., classification of calcified tissue versus non-calcified tissue in the aortic wall and intraluminal thrombus if present). Non-limiting examples of such segmentation models are described in International Patent Application No. PCT / IB2022 / 051558 entitled "METHOD AND SYSTEM FOR SEGMENTING AND CHARACTERIZING AORTIC TISSUES" filed on February 22, 2022 by the same applicant.

[0138] In one or more embodiments, the segmentation model 270 includes a fully convolutional neural network (FCN).

[0139] In one or more embodiments, the segmentation model 270 is trained to perform segmentation of the aorta in ultrasound images. In one or more embodiments, the segmentation model 270 may be trained to perform segmentation based on the segmented ultrasound image and the corresponding segmented CT image or MRI image.

[0140] In some embodiments of the present technology, the segmentation model 270 may further perform segmentation based on Doppler images collected by the medical imaging device 210.

[0141] In one or more embodiments, the segmentation model 270 has a ResNet-based FCN architecture. Non-limiting examples of ResNet include ResNet50 (50 layers), ResNet101 (101 layers), ResNet152 (152 layers), ResNet50V2 (50 layers with batch normalization), ResNet101V2 (101 layers with batch normalization), and ResNet152V2 (152 layers with batch normalization).

[0142] In one or more alternative embodiments, the segmentation model 270 may be implemented based on one of AlexNet, GoogleNet, and VGG.

[0143] Flow turbulence estimation model

[0144] The flow turbulence estimation model 275 is configured to estimate flow turbulence in 2D ultrasound Doppler images of the aorta.

[0145] The flow turbulence estimation model 275 is configured to determine forward and reverse flow values at different locations in the Doppler ultrasound image during the cardiac cycle. Further, the flow turbulence estimation model 275 may estimate pressure.

[0146] In one or more embodiments, the flow turbulence estimation model 275 may not be a machine learning model. It will be appreciated that the Doppler signal is a direct measurement of velocity that enables identification of turbulent flow, low flow, reverse flow, etc. The flow turbulence estimation model 275 may be used to apply a correction factor determined after verification after performing computer fluid dynamics (CFD) simulation procedures and / or MRI images (e.g., by using Ansys® Fluent® sold by Ansys, Inc.).

[0147] In one or more other embodiments, the flow turbulence estimation model 275 can be trained to estimate flow turbulence in 2D ultrasound Doppler images by correlating the measured local blood flow velocity with CFD simulation analysis performed using CT images and / or MRI images. In such embodiments, the flow turbulence estimation model 275 can be a machine learning model based on a deep neural architecture, such as, but not limited to, a transformer and a long short-term memory (LSTM).

[0148] During training, the flow turbulence estimation model 275 receives a training data set that includes, for each given patient, a set of Doppler ultrasound images and blood flow parameters such as velocity and pressure at each of the nodes generated during a computational fluid dynamics (CFD) simulation procedure. The flow turbulence estimation model 275 is then trained to estimate the blood flow values in the Doppler ultrasound images based on the blood flow values calculated during the CFD simulation procedure performed on the CT or MRI images of the same patient.

[0149] It should be understood that the blood flow parameters used to train the flow turbulence estimation model 275 can be generated during a CFD simulation procedure that simulates blood flow in an arterial shape by adopting a finite volume method for the numerical implementation of the Navier-Stokes equations that describe fluid flow. A CFD simulation finite volume method for solving the discretized form of the Navier-Stokes equations over all finite volume elements in the domain. The CFD simulation procedure applies an iterative method for simulating blood flow to obtain a converged numerical solution due to the governing equations being non-linear and coupled. It should be noted that in an alternative embodiment of the present technology, a finite element method or a finite difference method may be used instead of the finite volume method to obtain the same CFD parameters. Such a CFD simulation procedure is described in more detail in International Patent Application Publication WO2021 / 059243A1 entitled "METHOD AND SYSTEM FOR DETERMINING REGIONAL RUPTURE POTENTIAL OF BLOOD VESSEL" by the same applicant.

[0150] The flow turbulence estimation model 275 outputs blood flow parameters or indices including blood flow velocity values and pressure values in an ultrasonic Doppler image.

[0151] Database

[0152] The database 235 is configured to perform, inter alia, (i) storing collection parameters and data related to the medical imaging device 210, (ii) storing medical images including ultrasonic images, (iii) storing model parameters and hyperparameters of a set of ML models 250, (iv) storing data sets for training, testing, and validating a set of ML models 250, (v) storing data output by a set of ML models 250, and (vi) storing strain and / or RW parameters and / or maps.

[0153] The database 235 is configured to store ultrasonic images and videos. In one or more embodiments, the database may store medical digital imaging and communications (DICOM) files, for example, including DCM and DCM30 (DICOM 3.0) file extensions. Additionally or alternatively, the database 235 may store medical image files in a tag image file format (TIFF), a digital storage and retrieval (DSR) TIFF-based format, and a data exchange file format (DEFF) TIFF-based format.

[0154] In one or more embodiments, the database 235 may store ML file formats such as.tfrecords,.csv,.npy, and.petastorm, as well as file formats used to store models such as.pb and.pkl. The database 235 may also store well-known file formats such as, but not limited to, image file formats (e.g.,.png,.jpeg,.exif,.bmp,.tiff), video file formats (e.g.,.mp4,.mkv, etc.), archive file formats (e.g.,.zip,.gz,.tar,.bzip2), document file formats (e.g.,.docx,.pdf,.txt), or web file formats (e.g.,.html).

[0155] It will be appreciated that the database 235 may store other types of data such as a validation dataset (not shown), a test dataset (not shown), etc.

[0156] Communication network

[0157] In some embodiments of the present technology, communication network 220 is the Internet. In alternative non-limiting embodiments, communication network 220 can be implemented as any suitable local area network (LAN), wide area network (WAN), private communication network, etc. It should be clearly understood that the implementation examples for communication network 220 are for illustrative purposes only. How the communication link 225 (not separately numbered) between the medical imaging device 210 and / or the server 230 and / or another electronic device (not shown) and the communication network 220 is implemented depends, inter alia, on how each of the medical imaging device 210 and the server 230 is implemented.

[0158] Communication network 220 can be used to transmit data packets between the medical imaging device 210, the server 230, and the database 235. For example, communication network 220 can be used to transmit requests between the medical imaging device 210 and the server 230.

[0159] Regional Weakening (RW) Map Generation Procedure

[0160] Referring to FIG. 3, a schematic diagram of a regional weakening (RW) map generation procedure 300 according to one or more non-limiting embodiments of the present technology is shown.

[0161] The RW map generation procedure 300 is executed by at least one electronic device, such as the electronic device 100, which can be implemented as a mobile device (e.g., smartphone, tablet), a laptop, a desktop computer, a server (e.g., server 230), or can be incorporated into the medical imaging device 210.

[0162] The RW map generation procedure 300 includes, inter alia, an image acquisition procedure 310, an image enhancement procedure 315, a segmentation procedure 320, a measurement procedure 330, a thickness determination procedure 340, a flow turbulence determination procedure 350, and an RW determination procedure 370.

[0163] It is contemplated that portions of the RW map generation procedure 300 (i.e., some of the procedures) may be executed in a distributed manner (by two or more electronic devices). As a non-limiting example, a first computing device may execute the image acquisition procedure 310, the image enhancement procedure 315, the segmentation procedure 320, at least a portion of the measurement procedure 330, the thickness determination procedure 340, and the flow turbulence determination procedure 350, and a second computing device may execute another portion of the measurement procedure 330 and the RW determination procedure 370.

[0164] It will be appreciated that the image enhancement procedure 315 may be optional in some embodiments of the present technology.

[0165] Image acquisition procedure

[0166] The image acquisition procedure 310 is configured to, among other things, (i) receive an ultrasonic image from the medical imaging device 210 and (ii) receive a Doppler ultrasonic image from the medical imaging device 210.

[0167] The image acquisition procedure 310 is executed to obtain a plurality of ultrasonic images during a time period (e.g., a cardiac cycle). It will be appreciated that the ultrasonic images are collected as a sequence of frames during the time period and show movement over time (e.g., a loop). At a minimum, the ultrasonic images include one image at peak systole and one image at peak diastole.

[0168] The image acquisition procedure 310 may refer to ultrasonic images generated in the medical imaging device 210 via the user 202 operating the probe 214 of the medical imaging device 210 and / or ultrasonic images received by the server 230 or another processing device after being generated in the medical imaging device 210.

[0169] In one or more embodiments, the image acquisition procedure 310 may be performed online or in real time during at least a portion of the RW map generation procedure 300. In such embodiments, a physician, technician, or user 202 operating the medical imaging device 210 may collect images in real time, and the images may be transmitted to and received by the server 230 (or another processing device) according to requirements during the RW map generation procedure 300.

[0170] The image acquisition procedure 310 is configured to receive axial ultrasound images of the blood vessels of the patient 204 collected using the probe 214 operated by the user 202. The image acquisition procedure 310 is configured to collect Doppler ultrasound images of the patient's blood vessels collected from the same location as the non-Doppler ultrasound images. Similar to the ultrasound images, the Doppler images are collected as a sequence of frames during a time period (e.g., the cardiac cycle) to view the highest and lowest blood flow velocities throughout the cycle.

[0171] In one or more alternative embodiments, the image acquisition procedure 310 receives one static ultrasound image and M-mode (motion mode) information instead of a set of images.

[0172] It will be appreciated that the quality and quantity of the images depend on the medical imaging device 210, the user 202 operating the probe 214, and / or the anatomical structure of a given patient 204.

[0173] It will be appreciated that the blood vessels may include one or more of the aorta, iliac artery, and / or visceral artery of a given patient 204. The aorta may include one or more of the thoracic aorta, proximal aorta, intermediate aorta, and distal aorta.

[0174] In one or more other embodiments, the image acquisition procedure 310 may be performed offline. The image acquisition procedure 310 may be performed after all the necessary images of a given patient 204 have been acquired by the medical imaging device 210 and transmitted to the server 230 (or another processing device), after which no additional images of the given patient 204 are acquired.

[0175] In some embodiments of the present technology, the RW map generation procedure 300 includes an image enhancement procedure 315.

[0176] In one or more embodiments, the image enhancement procedure 315 is performed after the image acquisition procedure 310 or simultaneously with the image acquisition procedure 310 (i.e., in real time for each acquired image).

[0177] Image enhancement procedure

[0178] The image enhancement procedure 315 is configured to enhance the ultrasonic image in order to improve the quality, information content, and resolution of the ultrasonic image received from the medical imaging device 210 via the image acquisition procedure 310.

[0179] The image enhancement procedure 315 uses the resolution enhancer model 260 and the noise removal model 265 of the set of models 250.

[0180] The image enhancement procedure 315 uses the resolution enhancer model 260 for one or more ultrasonic images to obtain a high-resolution (HR) ultrasonic image, and uses the noise removal model 265 for the HR ultrasonic image to obtain a noise-removed HR ultrasonic image. In some embodiments of the present technology, it is contemplated that the image enhancement procedure 315 may use the noise removal model 265 to remove noise from the ultrasonic image before using the resolution enhancer model 260 to obtain the HR image.

[0181] In one or more embodiments, sufficient resolution and / or signal can be determined experimentally. It will be appreciated that the resolution in ultrasound can sometimes be difficult to define. As a non-limiting example, the axial resolution of a 5 MHz ultrasound scan is 0.3 mm, and for a 2 MHz transducer it is 0.8 mm. However, the lateral resolution depends on depth and on how well the transducer is focused, as it decreases as the structure gets deeper. The temporal resolution is generally 10 - 30 fps. In most cases, any resolution acceptable to a qualified clinician can be used, which is generally intended to be better than that obtained from CT. Signal quality is generally not limited by resolution. Signal quality is limited by noise, insufficient skin contact, insufficient adjustment of gain or focus, lack of coupling medium, presence of intestinal gas, improper handling of the transducer resulting in an image collected at an angle other than 90 degrees, and operator error resulting in an image where not the entire diameter is visible. In some embodiments, the ultrasound image can be evaluated to determine whether the signal quality is sufficient (e.g., above an experimentally determined threshold).

[0182] Furthermore, the image enhancement procedure 315 can also use pre - processing and post - processing techniques to improve the quality of the image.

[0183] The image enhancement procedure 315 can use, for example, filtering techniques. Non - limiting examples of filtering techniques include order - statistic filtering, Gaussian filtering, bilateral filtering, mean filtering, and Laplacian filtering, which can be used.

[0184] In some embodiments of the present technology, the image enhancement procedure 315 can be executed in real - time or near real - time such that the HR noise - removed image can be displayed to the user 202 operating the medical imaging device 210 (e.g., on the display of the medical imaging device 210 or on the display of a user device).

[0185] In one or more other embodiments, the image enhancement procedure 315 can be performed offline (i.e., after the necessary ultrasound images have been collected).

[0186] The image enhancement procedure 315 outputs a high-resolution (HR) ultrasound image of blood vessels and / or a noise-reduced ultrasound image.

[0187] In some embodiments of the present technology, the RW map generation procedure 300 can include a segmentation procedure 320 that is optional, for example, when segmentation is visually performed by the user 202 using an electronic caliper on the medical imaging device 210.

[0188] Segmentation procedure

[0189] The segmentation procedure 320 is configured to perform, among other things, (i) receiving an ultrasound image and (ii) segmenting the ultrasound image using a segmentation model 270 to obtain segmented tissue of blood vessels.

[0190] The segmentation procedure 320 is used to detect and demarcate different vascular tissues in the ultrasound image so that the composition, position, and movement of different vascular tissues in the ultrasound image can be evaluated.

[0191] In one or more embodiments, the segmentation procedure 320 receives the HR noise-reduced image from the image enhancement procedure 315.

[0192] The segmentation procedure 320 uses a segmentation model 270 to segment (i.e., classify each pixel) the aorta tissue in the ultrasound image collected by the medical imaging device 210. It should be understood that the segmentation procedure 320 can use one or more segmentation ML models 270 trained to segment tissue in an image of a blood vessel.

[0193] In one or more embodiments, the segmentation procedure 320 can be performed in real time, i.e., while the user 202 is operating the probe 214 of the medical imaging device 210 such that the segmented tissue is displayed on the display screen of the medical imaging device 210.

[0194] In one or more embodiments, the segmentation procedure 320 can segment the tissue based on specific images that depend on the collection angle and parameters of the probe 214. For example, the segmentation procedure 320 can perform segmentation based on an image scan along the entire aorta and also by measuring both static and dynamic images. As described above herein, the segmentation model 270 may be trained to perform segmentation using CT images and / or MRI images of the aorta in addition to ultrasound images of the aorta.

[0195] In one or more embodiments, the segmentation procedure 320 can use ultrasound Doppler images to perform segmentation.

[0196] Furthermore, in some embodiments, the user 202 may be able to confirm whether the segmentation of the tissue is accurate and / or correct the segmentation result. In such embodiments, the segmentation model 270 can be continuously trained to provide more accurate results based on the input of the user 202.

[0197] In one or more alternative embodiments, the segmentation procedure 320 can perform segmentation based on user input from the user 202 operating the medical imaging device 210. As a non-limiting example, the user 202 can give instructions via the input / output interface of the medical imaging device 210 (or via another computing device, such as a smartphone) to assist the segmentation model 270 in performing segmentation.

[0198] Although not limited, in some embodiments, such as an embodiment where the segmentation procedure 320 is executed in real time, the segmentation procedure 320 may color the boundaries (edges) of different segmented tissues that can be displayed on the display screen of the medical imaging device 210 or another electronic device.

[0199] FIG. 5 shows a non-limiting example of segmentation in an ultrasonic image 400, showing the boundary of the segmented outer wall 510, the boundary of the segmented inner wall 515, the boundary of the segmented lumen 520, and a thrombus 530 present in the lumen between the boundary of the segmented inner wall 515 and the boundary of the segmented lumen 520.

[0200] The segmentation procedure 320 outputs one or more segmented ultrasonic images. The segmentation procedure 320 outputs segmented tissues including an outer wall, an inner wall, a lumen, and a thrombus (if present).

[0201] In one or more embodiments, the segmentation procedure 320 outputs an indication (i.e., coordinates) of the definition of the segmented tissue in the ultrasonic image. As a non-limiting example, the segmentation procedure 320 may output a plurality of edge (boundary) coordinates for each segmented tissue.

[0202] Measurement procedure

[0203] The measurement procedure 330 is configured to perform, among other things, (i) receive a segmented ultrasonic image, (ii) measure a region of interest in a blood vessel based on the segmented ultrasonic images over time, and (iii) measure the movement in the region of interest of the blood vessel.

[0204] The measurement procedure 300 is configured to determine measurement values related to the segmented tissue in the blood vessel such that the movement of each of the segmented tissues in each ultrasonic image can be quantified. The movement of the segmented tissue in the blood vessel indicates deformation and strain and is used to evaluate local weakening in the blood vessel.

[0205] In one or more embodiments, the measurement procedure 330 receives the segmented image from the segmentation procedure 320. It will be appreciated that the segmented ultrasound image can be a segmented HR ultrasound image segmented after the image enhancement procedure 315.

[0206] In one or other embodiments, the measurement procedure 330 can be automatically performed with minimal user intervention. In such embodiments, the measurement procedure 330 segments tissue (e.g., inner wall, outer wall, and lumen) and uses one or more of a set of ML models 250, such as the segmentation model 270, to measure regions of interest (e.g., edges or boundaries) in the segmented tissue in each image taken over time. It will be appreciated that the measurement procedure 330 can be combined with the segmentation procedure 320 such that for a given segmented tissue in an image, a difference in the given segmented tissue over time (i.e., other images being compared taken over a time period) is automatically calculated.

[0207] In one or more other embodiments, the measurement procedure 330 can be semi-automatically performed when receiving an input by the user 202 operating the medical imaging device 210. As a non-limiting example, the user 202 can indicate a location (i.e., coordinates or a point) in the ultrasound image via the input / output interface to perform the measurement. In one or more embodiments, the measurement procedure 330 automatically measures the measurement values and reports them to the user 202 operating the medical imaging device 210 for verification.

[0208] In one or more embodiments, the measurement procedure 330 can be directly executed by the processing device of the medical imaging device 210, and its output is provided to the RW map generation procedure 300. In one or more other embodiments, the measurement procedure 330 can be executed by another processing device, such as a workstation computer (not shown), a server 230, a mobile device (not shown), etc.

[0209] In one or more embodiments, the measurement procedure 330 is configured to measure one or more of the supraceliac abdominal aorta diameter, the adrenal abdominal aorta diameter, the infrarenal abdominal aorta diameter, and the iliac diameter by receiving user input.

[0210] In one or more alternative embodiments, the measurement procedure 330 is configured to measure the inner diameter (i.e., from the inner anterior wall to the inner posterior wall (inner to inner or ITI) or from the intima to the intima), and / or the outer diameter (i.e., from the outer anterior wall to the outer posterior wall (OTO) or from the adventitia to the adventitia), and / or perform a measurement from leading edge to leading edge.

[0211] In one or more embodiments, the measurement procedure 330 determines the maximum transverse diameter of the aneurysm sac, the longitudinal length, the (saccular / fusiform / eccentric) shape measure, the upper extent measurement relative to the renal artery, and the lower extent measurement including the extension into the branches, side branches or visceral branches arising from the aneurysm.

[0212] The measurement procedure 330 is configured to perform measurements related to one or more of the outer wall, inner wall, lumen, and thrombus. It will be understood that the measurement procedure 330 may perform multiple measurements in each image for each of the outer wall, inner wall, lumen, and thrombus (if present). The multiple measurements of the outer wall, inner wall, lumen, and thrombus enable the evaluation of their characteristics and the determination of the above-described movement of tissue over time.

[0213] In one or more embodiments, the measurement procedure 330 outputs a set of diameters related to the movement of vascular tissue over time.

[0214] The measurement procedure 330 is configured to determine the minimum diameter, maximum diameter, and equivalent circle diameter and eccentricity for the segmented tissue by analyzing and comparing the measurements in each image (e.g., frame) taken over time.

[0215] FIG. 4 shows a non-limiting example of an axial ultrasonic image 400 of a patient's body having four radial points 402, 404, 406, 406 along the circumference of the patient's aorta. The axial ultrasonic image 400 includes a diameter measurement, i.e., a first diameter 410 of 7.12 cm (i.e., between point 402 and point 404) and a second diameter (i.e., between point 402 and point 404) of 7.82 cm, at the lower right.

[0216] In one or more embodiments, the measurement procedure 330 determines the radial deformation by analyzing a set of ultrasonic images taken over time based on a set of measurements. The measurement procedure 330 determines the variation of the boundaries (i.e., radial points) of the segmented tissue by comparing the ultrasonic images taken over time in order to obtain the radial deformation of the segmented tissue.

[0217] The radial deformation gives an indication of the strain in the blood vessel. The measurement procedure 330 outputs a radial deformation index or parameter that includes radial deformation values in the region of the blood vessel that at least partially indicates the strain.

[0218] In one or more embodiments, the measurement procedure 330 may use a machine learning (ML) model (not shown) or a heuristic to calculate strain values based on the radial deformation values.

[0219] In some embodiments, prior to measuring the deformation and thrombus load, the measurement procedure 330 sends the ultrasonic image to the segmentation procedure 320 to obtain a clear delineation of the aortic wall, lumen, and thrombus (if present). The measurement procedure 330 then measures and determines the deformation values of the wall and lumen.

[0220] The measurement procedure 330 outputs the radial deformation value.

[0221] The measurement procedure 330 includes a thickness determination procedure 340.

[0222] Thickness determination procedure

[0223] The thickness determination procedure 340 is configured to measure the thrombus burden or thickness based on the distance between the lumen and the inner wall of the blood vessel. The thrombus burden corresponds to the measured value of the intraluminal thrombus (ILT) thickness.

[0224] It will be appreciated that in some embodiments, there may be no thrombus in the aorta, and thus the measured thrombus burden may be equal to 0.

[0225] In one or more embodiments, the thickness determination procedure 340 receives the segmented ultrasound image from the segmentation procedure 320 and uses the segmented ultrasound image to determine the thrombus burden based on the distance between the lumen and the inner wall of the aorta.

[0226] In one or more embodiments, the thickness determination procedure 340 measures the thrombus burden based on the measurements performed during the measurement procedure 330.

[0227] It will be appreciated that the thickness determination procedure 340 may determine the distance based on the coordinates of the boundary between the segmented lumen and the segmented inner wall of the aorta. The distance can be measured at each radial point. It will be appreciated that the measured distances can be averaged to obtain the thrombus burden.

[0228] In one or more embodiments, if there is sufficient resolution to define the interface between the thrombus surface and the inner surface of the wall, the thickness determination procedure 340 determines the thrombus burden.

[0229] In one or more embodiments, the thrombus burden is measured from the inner surface of the artery to the edge of the thrombus. If the arterial wall is extremely thin or the wall cannot be segmented, the thrombus burden can be measured from the outer surface of the arterial wall to the edge of the thrombus.

[0230] The thickness determination procedure 340 outputs a thickness parameter including the thrombus thickness value.

[0231] The flow turbulence determination procedure 350 can be executed simultaneously with other procedures at any time prior to the RW determination procedure 360.

[0232] Flow turbulence determination procedure

[0233] The flow turbulence determination procedure 350 is configured to determine blood flow turbulence based on a 2D ultrasonic Doppler image.

[0234] The flow turbulence determination procedure 350 directly determines a flow turbulence value from the Doppler signal. The Doppler images taken over a time period (e.g., the cardiac cycle) include the highest and lowest speeds over the entire time period.

[0235] In one or more embodiments, the flow turbulence determination procedure 350 determines turbulent flow, low flow, and reverse flow based on the Doppler image.

[0236] As described above herein, in an alternative embodiment, the flow turbulence estimation model 275 may be pre-trained to estimate flow turbulence in a 2D ultrasonic Doppler image by correlating the measured local blood flow velocity with a CFD simulation analysis performed using a CT image or an MRI image.

[0237] The flow turbulence determination procedure 350 outputs blood flow parameters or blood flow measurements that include forward flow values and reverse flow values at different locations in the Doppler ultrasonic image during a time period (e.g., the cardiac cycle).

[0238] FIG. 6 shows a simulated Doppler image showing forward flow 610 and reverse flow 620. The Doppler image informs an image enhancement process regarding segmentation of the lumen and reveals the flow turbulence used to calculate the RW index.

[0239] RW determination procedure

[0240] The RW determination procedure 360 is configured to perform, among other things, (i) receiving one or more ultrasonic images, (ii) receiving thickness measurement values, flow turbulence measurement values, and deformation measurement values, (iii) generating a radial strain parameter including a radial strain value based on the thickness measurement values and the deformation measurement values, (iv) generating a local weakening (RW) parameter including a local weakening value based on the thickness measurement values, the flow turbulence measurement values, the deformation measurement values, and the radial strain parameter, (v) generating a radial strain map of the circumference of the blood vessel wall based on the radial strain parameter and the ultrasonic images, and (vi) generating an RW map of the circumference of the blood vessel wall based on the radial strain parameter, the RW parameter, and the ultrasonic images.

[0241] The RW determination procedure 360 receives the radial deformation measurement values from the measurement procedure 330, the blood flow turbulence measurement values from the flow turbulence determination procedure 350, and the thickness measurement values from the thickness determination procedure 340.

[0242] In one or more embodiments, the RW determination procedure 360 determines local material stiffness based on strain and blood pressure. In this context, it will be appreciated that the radial deformation measurement values and the blood flow velocity may be used as surrogates for strain and blood pressure.

[0243] In one or more other embodiments, the user 202 may evaluate the local material stiffness that may be used to improve the RW measurement values.

[0244] The RW determination procedure 360 determines a radial strain value based on the thickness measurement values and the deformation measurement values, and the radial strain parameter includes the radial strain value. The RW determination procedure 360 generates a radial strain parameter or index including the radial strain value at locations along the circumference of the blood vessel.

[0245] The RW determination procedure 360 determines a local weakening (RW) value based on the radial deformation value, the blood flow turbulence value, and the thrombus thickness value. The RW determination procedure 360 generates an RW parameter or index including the RW value at locations along the circumference on the blood vessel.

[0246] In one or more embodiments, the RW determination procedure 360 may indicate a segment of the circumference of a blood vessel or a continuous distribution around the circumference. It will be appreciated that the diameter is measured at 90 degrees (with some error due to the operator not being able to hold the transducer exactly at 90 degrees) with respect to the patient's torso.

[0247]

[0246] The RW determination procedure 360 determines a RW parameter or index that includes a RW value. The RW parameter represents the state of local weakening of a region or set of regions of the blood vessel wall and the probability of rupture. The RW parameter takes into account different factors for malignant remodeling and degeneration of the blood vessel wall, including but not limited to the wall, and indicates the localized state of weakening of the blood vessel and the resulting potential for dilation and rupture.

[0248] The RW determination procedure 360 displays the RW value on an ultrasonic image. It will be appreciated that the ultrasonic image may be an ultrasonic image with HR noise removed, and the RW determination procedure 360 may display the RW value around the circumference of the blood vessel.

[0249] In one or more embodiments, the RW determination procedure 360 displays a strain score and a RW score in segments around the circumference of the image. As a non-limiting example, there may be 12 patches, but it will be appreciated that the size and number of patches may vary. In some embodiments, a continuous display of strain and RW instead of using patches is also contemplated.

[0250]

[0247] In one or more embodiments, time resolution above a threshold may enable the evaluation of viscoelasticity in the wall and in a thrombus, if present. Viscoelastic properties, particularly energy loss, have been linked to wall degeneration in the thoracic aorta (see Chung et al. (2014). It has also been shown that high RAW corresponds to higher energy loss in the abdominal aorta. See Forneris et al. (2021)).

[0251] The temporal resolution depends on the maximum depth being measured. That is, it will be appreciated that ultrasound has a velocity in tissue, and the time defined by this velocity (which enables both the transmission and reception of echoes) limits the temporal resolution. The temporal resolution can be further limited by the quality of the electronics of the medical imaging device 210.

[0252] Referring to FIG. 7, shown is a radial strain map 720 on the wall of the aorta superimposed on an ultrasound image 400, with a radial value scale 710 on the right side, according to one or more non-limiting embodiments of the present technology.

[0253] Referring to FIG. 8, shown is a RAW map 720 of the circumference of the wall of the aorta superimposed on an ultrasound image 400, with an RW value scale. The RW map 720 is determined based on deformation values, flow turbulence values, and thrombus thickness values at every point around the circumference of the wall of the aorta.

[0254] Description of the method

[0255] FIG. 9 shows a flowchart of a method 900 for determining local weakening of a blood vessel based on an ultrasound image of a patient 204's body, according to one or more non-limiting embodiments of the present technology.

[0256] In one or more embodiments, the method 900 is executed by a processing device, such as a processor 110 and / or a GPU 111, operably connected to a non-transitory computer-readable storage medium, such as a solid-state drive 120 and / or a random-access memory 130 that stores computer-readable instructions. The processing device is configured or operable to execute the method 900 when executing the computer-readable instructions.

[0257] It will be appreciated that the processing device can be the processing device of the medical imaging device 210, a server 230, and / or another computerized device, such as a mobile device (smartphone or laptop), a desktop computer, a laptop computer, etc.

[0258] The processing device has access to a set of ML models 250.

[0259] Method 900 starts at processing step 902.

[0260] According to processing step 902, the processing device receives a set of images of the body of a given patient 204 collected using the medical imaging device 210, and the set of images has been collected over a time period (e.g., a cardiac cycle). The medical imaging device 210 is implemented as a 2D ultrasound imaging device.

[0261] The set of images includes ultrasound images and Doppler images taken over a time period. The set of images includes axial ultrasound images of a part of the body of patient 204. The ultrasound images of a part of the body of patient 204 include at least one blood vessel. The blood vessels can include, by way of non-limiting example, the aorta and / or the iliac artery. The aorta can include one or more of the thoracic aorta, the proximal aorta, the intermediate aorta, and the distal aorta.

[0262] In one or more embodiments, the processing device receives a plurality of ultrasound images. The processing device accesses the resolution enhancer model 260 and the noise removal model 265 among the set of models 250. The processing device uses the resolution enhancer model 260 for a plurality of ultrasound images to obtain high-resolution (HR) ultrasound images, and uses the noise removal model 265 for the HR ultrasound images to obtain a set of images, and the set of images includes a plurality of noise-removed HR ultrasound images of higher quality than the plurality of ultrasound images received from the medical imaging device 210. In one or more embodiments, the trained resolution improvement ML model includes an extremely deep super-resolution (VDSR) neural network, and the trained noise removal ML model includes a cycle-consistent generative adversarial network (CycleGAN).

[0263] In one or more other embodiments, the processing device receives a plurality of ultrasound images corresponding to a set of images from the medical imaging device 210.

[0264] According to processing step 904, the processing device receives a set of measurements representing the movement of the walls and lumen of the blood vessel, the set of measurements being determined based on the set of images.

[0265] In one or more embodiments, prior to performing processing step 904, the processing device receives an ultrasound image including coordinates on the tissue of the blood vessel, such as the outer wall, inner wall, lumen, and thrombus (if present). As a non-limiting example, the coordinates on the tissue of the blood vessel may include the diameters of the outer wall, inner wall, and lumen.

[0266] The processing device calculates the movement of the tissue by analyzing the coordinates in the tissue in the ultrasound images over time to obtain a set of measurements representing the movement of the walls and lumen of the blood vessel.

[0267] In one or more embodiments, prior to performing processing step 904, the processing device accesses one or more trained segmentation ML models 270 to segment (i.e., classify each pixel) the blood vessel tissue in the ultrasound image. The trained segmentation ML model 270 may output segmented tissue including the outer wall, inner wall, lumen, thrombus (if present), and background pixels. The segmented tissue may include the locations of all pixels on the edges and boundaries.

[0268] The processing device then determines the diameter of the segmented outer wall, the diameter of the segmented inner wall, and the diameter of the segmented lumen. In one or more alternative embodiments, the user 202 may perform the measurements and segmentation manually (i.e., using calipers) and enter the measurements in the medical imaging device 210 (or another electronic device) such that the measurements are received by the processing device.

[0269] The processing device is configured to perform one or more diameter measurements of the outer wall of the aorta, the inner wall of the aorta, and the lumen. In one or more embodiments, the processing device may perform at least a portion of the measurements based on an input from user 202. In one or more other embodiments, the processing device may perform the diameter measurements automatically.

[0270] It will be appreciated that the processing device may perform multiple measurements for each of the outer wall, inner wall, lumen, and thrombus (if present).

[0271] In one or more embodiments, the processing device is configured to determine, for each of the diameter measurements, the minimum diameter, the maximum diameter, the equivalent circular diameter, and the eccentricity. It will be appreciated that the minimum diameter, the maximum diameter, the equivalent circular diameter, and the eccentricity for each of the diameter measurements may be determined by the processor based on user input from user 202 performed in medical imaging device 210.

[0272] In one or more embodiments, the processing device measures the thrombus burden or thickness. The thrombus burden corresponds to the intraluminal thrombus (ILT) thickness. In one or more embodiments, the thrombus burden is measured from the inner surface of the artery to the edge of the thrombus. If the wall of the artery is extremely thin or the wall cannot be segmented, the thrombus burden is measured from the outer surface of the arterial wall to the edge of the thrombus.

[0273] It will be appreciated that if the thrombus burden is equal to 0, it indicates that no thrombus is present.

[0274] The processing device thus obtains a radial deformation value in the blood vessel that indicates the strain in the blood vessel. In one or more embodiments, the processing device generates a radial strain parameter or index based at least on the radial deformation value.

[0275] In one or more embodiments, the processing device may correct the radial deformation value or calculate the actual strain, for example, by using an ML model trained for that purpose.

[0276] According to processing step 906, the processing device determines a blood flow turbulence value in a blood vessel based on a Doppler image in a set of images.

[0277] In one or more embodiments, the processing device analyzes Doppler images over time, each including a blood flow velocity value, to determine the blood flow turbulence value.

[0278] In one or more alternative embodiments, the processing device accesses a trained flow turbulence estimation model 275. The flow turbulence estimation model 275 is pre-trained to estimate flow turbulence in 2D ultrasound Doppler images by correlating measured local blood flow velocities with fluidity analysis performed using CT or MRI images.

[0279] The processing device outputs blood flow parameters or blood flow values including forward flow values and reverse flow values at different locations in a Doppler ultrasound image during a cardiac cycle, as well as pressure values.

[0280] According to processing step 908, the processing device generates a regional aortic weakening (RW) parameter indicative of a state of weakening in the region of the aorta based on at least a set of measurement values representing movement in the blood vessel and the blood flow turbulence value.

[0281] In one or more embodiments, the processing device generates the RW parameter based on a radial deformation value indicative of strain, a thrombus thickness, and the blood flow turbulence value.

[0282] In one or more embodiments, the processing device defines a plurality of patches on a blood vessel shape including an outer wall and a lumen, perpendicular to the lumen centerline, and determines a patch-averaged distribution for each of the radial deformation value, the blood flow turbulence value, and the thickness value.

[0283] The processing device determines RW parameters or an exponent. The RW parameters represent the state of local weakening of a region or set of regions of the blood vessel wall and the probability of rupture. The RW parameters take into account different factors for malignant remodeling and degeneration of the blood vessel wall, including but not limited to the aortic wall, and indicate the localized state of weakening of the blood vessel, as well as the resulting potential for dilation and rupture.

[0284] In processing step 910, the processing device outputs the RW parameters. In one or more embodiments, the processing device outputs the radial strain parameter together with the RW parameters.

[0285] In one or more embodiments, the processing device uses a set of ultrasonic images to generate a strain map based on the strain parameters and a local weakening map based on the RW parameters.

[0286] In one or more embodiments, the processing device causes a strain score and an RW score to be displayed in a patch around the circumference of the ultrasonic image.

[0287] In one or more embodiments, the processing device determines the local material stiffness based on the strain and blood pressure.

[0288] Method 900 then ends.

[0289] In some cases, examples of modifications of the technology that may be useful may also be described. This is done merely to aid understanding and, in this case too, is neither for the purpose of limiting the scope of the technology nor for the purpose of describing the limits of the technology. These modifications are not an exhaustive list and those skilled in the art may make other modifications while still remaining within the scope of the technology. Further, where examples of modifications are not described, it should not be construed that the modification is not possible and / or that what is described is the only way to implement that element of the technology.

[0290] Modifications and improvements to the implementation examples described above of the present technology may be apparent to those skilled in the art. The above description is illustrative rather than limiting.

Claims

1. A method for determining local weakening parameters of the blood vessels of a given patient, the method being executed by at least one processing device, the method comprising: Receiving a set of images of a part of the body of the given patient collected over time using an ultrasonic imaging device, the set of images including Doppler images; Receiving a set of measurements representing the movement of the walls and lumen of the blood vessel, the set of measurements being determined based on the set of images; Determining a blood flow turbulence value in the blood vessel based on the Doppler images in the set of images; Generating a local weakening parameter indicating the state of weakening in the region of the blood vessel based on at least the set of measurements and the blood flow turbulence value; Outputting the local weakening parameter A method comprising:

2. Before generating the local weakening parameter indicating the state of weakening in the region of the blood vessel based on at least the set of measurements and the blood flow turbulence value, Determining a radial deformation in the blood vessel by analyzing the set of images over time based on the set of measurements, the radial deformation at least partially indicating strain in the blood vessel, further comprising determining the radial deformation; The method according to claim 1, wherein generating the local weakening parameter is based on the radial deformation in the blood vessel.

3. After receiving the set of images, Detecting respective wall positions and lumen positions in the blood vessel for the set of images over time; Calculating the movement of the walls and the lumen of the blood vessel based on the respective wall positions and lumen positions over time, thereby receiving the set of measurements; calculating the movement The method according to claim 1 or 2, further comprising:

4. Calculating the movement of the walls and the lumen of the blood vessel based on the respective wall positions and lumen positions over time includes determining a thrombus thickness based on the positions of the walls and the lumen, The method according to claim 3, wherein generating the local weakening parameter is based on the thrombus thickness.

5. The method according to claim 4, wherein a thrombus thickness of 0 indicates that no thrombus is present in the blood vessel.

6. The method according to any one of claims 1 to 5, wherein the blood vessel includes at least one of the aorta, the common iliac artery, and the visceral artery.

7. Generating a local weakening map based on the local weakening parameter and the set of images of the blood vessel; Outputting the local weakening map, wherein the local weakening map shows the circumference of the blood vessel including local weakening values. The method according to any one of claims 1 to 6, further comprising:

8. For the set of images over time, detecting each wall position and lumen position in the blood vessel, Segmenting the outer wall, inner wall, and lumen of the blood vessel using a segmentation model trained on the set of images. The method according to any one of claims 3 to 7, comprising:

9. Determining the minimum diameter, maximum diameter, equivalent circle diameter, and eccentricity of each of the outer wall diameter, inner wall diameter, and lumen diameter. The method according to claim 8, further comprising:

10. The method according to claim 8 or 9, wherein the trained segmentation model includes a fully convolutional neural network (FCN).

11. Before receiving the set of images of the blood vessel, Receiving a plurality of images of the given patient collected using the ultrasonic imaging device, wherein the plurality of images includes at least one Doppler image; Generating a plurality of high-resolution (HR) images of the blood vessel using a resolution improvement machine learning (ML) model trained on the plurality of images, wherein the resolution of the plurality of HR images is higher than the resolution of the plurality of images; Denosing the plurality of HR images using a trained ML denoising model to obtain the set of images. The method according to any one of claims 1 to 10, further comprising:

12. The trained resolution improvement ML model includes a very deep super-resolution (VDSR) neural network; The trained noise removal ML model includes a cycle-consistent generative adversarial network (CycleGAN). The method according to claim 11.

13. Generating a strain map based on the radial deformation value and the set of images of the blood vessel. Outputting the strain map, wherein the strain map shows the circumference of the blood vessel including strain values, and outputting the strain map The method according to any one of claims 2 to 12, further comprising.

14. The method according to any one of claims 1 to 13, further comprising determining the viscoelasticity of the blood vessel wall.

15. The method according to any one of claims 8 to 14, further comprising determining the viscoelasticity of the thrombus.

16. The method according to any one of claims 1 to 15, wherein the at least one processing device is operably connected to the ultrasonic imaging device.

17. Displaying the local weakening map on a display operably connected to the at least one processing device The method according to any one of claims 5 to 16, further comprising.

18. A system for determining local weakening (RW) of a blood vessel in a given patient, wherein The system is At least one processing device and A non-transitory storage medium operably connected to the at least one processing device, the non-transitory storage medium storing computer-readable instructions, a non-transitory storage medium Comprising When the processor executes the computer-readable instructions, Receiving a set of images of a part of the body of the given patient collected over time using an ultrasonic imaging device, wherein the set of images includes Doppler images, and receiving the set of images Receiving a set of measurement values representing the movement of the wall and lumen of the blood vessel, wherein the set of measurement values is determined based on the set of images, and receiving the set of measurement values Determining a blood flow turbulence value in the blood vessel based on the Doppler image in the set of images Generating a local weakening parameter indicating the state of weakening in the region of the blood vessel based on at least the set of measurement values and the blood flow turbulence value Outputting the local weakening parameter A system configured to perform.

19. Before the at least one processing device generates the local weakening parameter indicating the state of weakening in the region of the blood vessel based on at least the set of measurement values and the blood flow turbulence value, Determining radial deformation in the blood vessel by analyzing the set of images over time based on the set of measurement values, further configured to perform determining radial deformation, wherein the radial deformation at least partially indicates strain in the blood vessel, generating the local weakening parameter based on the radial deformation in the blood vessel, The system according to claim 18.

20. after the at least one processing device receives the set of images, for the set of images over time, detecting respective wall positions and lumen positions in the blood vessel, calculating the movement of the wall and the lumen of the blood vessel based on the respective wall positions and lumen positions over time, thereby receiving the set of measurement values, calculating the movement, The system according to claim 18 or 19, further configured to perform.

21. calculating the movement of the wall and the lumen of the blood vessel based on the respective wall positions and lumen positions over time includes determining a thrombus thickness based on the positions of the wall and the lumen, generating the local weakening parameter based on the thrombus thickness, the system according to claim 20.

22. The system according to claim 21, wherein a thrombus thickness of 0 indicates that no thrombus is present in the blood vessel.

23. The system according to any one of claims 18 to 22, wherein the blood vessel includes at least one of an aorta, a common iliac artery, and a visceral artery.

24. the at least one processing device, generating a local weakening map based on the local weakening parameter and the set of images of the blood vessel, outputting the local weakening map, wherein the local weakening map shows the circumference of the blood vessel including local weakening values, outputting the local weakening map, The system according to any one of claims 18 to 23, further configured to perform.

25. for the set of images over time, detecting the respective wall positions and lumen positions in the blood vessel, segmenting an outer wall, an inner wall, and the lumen of the blood vessel using a segmentation model trained for the set of images, The system according to any one of claims 20 to 24, including.

26. the at least one processing device being further configured to determine a minimum diameter, a maximum diameter, an equivalent circle diameter, and an eccentricity of each of an outer wall diameter, an inner wall diameter, and a lumen diameter The system according to claim 25, further configured to perform **Claim 27** The system according to claim 25 or 26, wherein the trained segmentation model includes a fully convolutional neural network (FCN). **Claim 28** before receiving the set of images of the blood vessel, the at least one processing device receiving a plurality of images of the given patient collected using the ultrasonic imaging device, the plurality of images including at least one Doppler image, and generating a plurality of high-resolution (HR) images of the blood vessel using a resolution improvement machine learning (ML) model trained on the plurality of images, the resolution of the plurality of HR images being higher than the resolution of the plurality of images, and denoising the plurality of HR images using a trained ML denoising model to obtain the set of images The system according to any one of claims 18 to 27, further configured to perform **Claim 29** wherein the trained resolution improvement ML model includes a very deep super-resolution (VDSR) neural network, and the system according to claim 28, wherein the trained noise removal ML model includes a cycle-consistent generative adversarial network (CycleGAN). **Claim 30** generating a strain map based on the radial deformation value and the set of images of the blood vessel, and outputting the strain map, the strain map showing a circumference of the blood vessel including strain values The system according to any one of claims 18 to 29, further comprising **Claim 31** The system according to any one of claims 18 to 30, wherein the at least one processing device is further configured to determine the viscoelasticity of the blood vessel wall. **Claim 32** The system according to any one of claims 25 to 31, wherein the at least one processing device is further configured to determine the viscoelasticity of the thrombus. **Claim 33** The system according to any one of claims 18 to 32, wherein the at least one processing device is operably connected to the ultrasonic imaging device.

34. The at least one processing device is further configured to display the local weakening map on a display operably connected to the at least one processing device. The system according to any one of claims 22 to 33.

Images