System and method for detecting image quality degradation of an ultrasound imaging system

Abstract

A method for detecting degradation of an ultrasound imaging system (140) includes receiving ultrasound images of a heart (S311); automatically determining a cardiac view of the heart by applying at least one ultrasound image to a trained machine learning algorithm (S312); automatically selecting an image quality (IQ) model from a plurality of IQ models based on the determined cardiac view of the heart (S313); determining an IQ metric associated with each ultrasound image by applying each ultrasound image to the selected IQ model (S314); aggregating the IQ metrics to provide an aggregated IQ score associated with each ultrasound image, thereby creating a time series of aggregated IQ scores (S315); estimating a level of degradation of the ultrasound imaging system based on the time series of aggregated IQ scores associated with the ultrasound images over time using a degradation detection machine learning algorithm (S316); and notifying a user of the estimated level of degradation (S317).

Description

Background Technology

[0001] The quality of images provided by an ultrasound imaging system (image quality (IQ)) decreases over time. This is especially true for the ultrasound transducer probes of an ultrasound imaging system. Acute equipment problems or malfunctions often cause a sudden drop in IQ that can be immediately identified by the user, resulting in the need for repairs and potentially extended downtime of the ultrasound imaging system.

[0002] However, potential detectable degradation in IQ often begins long before equipment failure and / or a sudden drop in IQ. It would be beneficial to detect image quality degradation in ultrasound imaging systems before acute equipment problems or failures occur and to promptly notify users of this degradation so that the equipment involved can be proactively replaced and / or repaired without interrupting current inspections or scheduling. This improves the overall quality of ultrasound images within the field of medical imaging technology. Summary of the Invention

[0003] In a representative embodiment, a method for detecting degradation of an ultrasound imaging system is provided. The method includes receiving multiple ultrasound images of a heart of an object acquired by the ultrasound imaging system; automatically determining a cardiac view of the heart based on the at least one ultrasound image by applying a trained first machine learning algorithm to the at least one ultrasound image, wherein the cardiac view of the heart is one of a plurality of predetermined cardiac views of the heart; automatically selecting an IQ model from a plurality of image quality (IQ) models based on the determined cardiac view of the heart, wherein the plurality of IQ models respectively correspond to the plurality of predetermined cardiac views of the heart, and wherein the selected IQ model includes a trained second machine learning algorithm; and automatically determining a cardiac view of the heart based on the at least one ultrasound image acquired by the ultrasound imaging system to detect degradation of an ultrasound imaging system. Each ultrasound image in the acoustic imaging is applied to a selected IQ model to determine multiple IQ metrics associated with each ultrasound image; the multiple IQ metrics are aggregated to provide an aggregated IQ score associated with each of the multiple ultrasound images; a time-series trend of image quality is predicted based on the time-series of the aggregated IQ scores associated with the multiple ultrasound images over time; the degradation level of the ultrasound imaging system is estimated by analyzing the time-series trend of the aggregated IQ scores associated with the multiple ultrasound images over time using a degradation detection machine learning algorithm; and the user is notified when the estimated degradation level is based on a predetermined degradation threshold.

[0004] In another representative embodiment, a system for detecting degradation of an ultrasound imaging system is provided. The system includes a display; a processor; and a non-transient memory storing instructions. When executed by the processor, the instructions cause the processor to: receive multiple ultrasound images of the heart of an object acquired by the ultrasound imaging system; automatically determine a cardiac view of the heart based on at least one of the multiple ultrasound images by applying it to a trained cardiac view machine learning algorithm, wherein the cardiac view of the heart is one of a plurality of predetermined cardiac views of the heart; and automatically select an IQ model from a plurality of IQ models based on the determined cardiac view of the heart, wherein the plurality of IQ models respectively correspond to the plurality of predetermined cardiac views of the heart, and wherein the selected IQ model… The selected IQ model includes a trained IQ machine learning algorithm; multiple IQ measures associated with each ultrasound image are determined by applying the selected IQ model to each of the multiple ultrasound images; the multiple IQ measures are aggregated to provide an aggregated IQ score associated with each of the multiple ultrasound images, thereby creating a time series of the aggregated IQ score; a degradation detection machine learning algorithm is used to estimate the degradation level of the ultrasound imaging system based on the time series of the aggregated IQ scores associated with the multiple ultrasound images over time; and a notification regarding the estimated degradation level is displayed on the display.

[0005] In another representative embodiment, a non-transient computer-readable medium stores instructions for detecting degradation of an ultrasound imaging system. When executed by the processor, the instructions cause the processor to: receive multiple ultrasound images of the heart of an object acquired by the ultrasound imaging system; automatically determine a cardiac view of the heart based on at least one ultrasound image by applying it to a trained cardiac view machine learning algorithm, wherein the cardiac view of the heart is one of a plurality of predetermined cardiac views of the heart; and automatically select an IQ model from a plurality of IQ models based on the determined cardiac view of the heart, wherein the plurality of IQ models respectively correspond to the plurality of predetermined cardiac views of the heart, and wherein the selected IQ model... The IQ model includes a trained IQ machine learning algorithm; multiple IQ measures associated with each ultrasound image are determined by applying each of the multiple ultrasound images to the selected IQ model; the multiple IQ measures are aggregated to provide an aggregated IQ score associated with each of the multiple ultrasound images, thereby creating a time series of the aggregated IQ score; a degradation detection machine learning algorithm is used to estimate the degradation level of the ultrasound imaging system based on the time series of the aggregated IQ scores associated with the multiple ultrasound images over time; and a notification regarding the estimated degradation level is displayed to the user on the display. Attached Figure Description

[0006] When combined with attachment Figure 1 When reading this, the exemplary embodiments are best understood in light of the following specific implementation. It should be emphasized that the various features are not necessarily drawn to scale. In fact, dimensions may be increased or decreased arbitrarily for clarity of discussion. Similar reference numerals refer to similar elements, provided they are applicable and practical.

[0007] Figure 1 This is a simplified block diagram of a system for detecting degradation of an ultrasound imaging system according to a representative embodiment.

[0008] Figure 2A This is a graph illustrating the predicted time-series trend of an ultrasound imaging system filtered by a patient's body mass index (BMI) according to a representative embodiment.

[0009] Figure 2B This is a graph illustrating the predicted time series trend of an ultrasound imaging system filtered by a transducer probe according to a representative embodiment.

[0010] Figure 3 This is a flowchart illustrating a method for detecting degradation of an ultrasound imaging system according to a representative embodiment. Detailed Implementation

[0011] In the following detailed description, representative embodiments with specific details disclosed are set forth for purposes of explanation and not limitation, in order to provide a thorough understanding of embodiments according to this teaching. Descriptions of known systems, devices, materials, methods of operation, and methods of manufacture may be omitted to avoid obscuring the description of representative embodiments. Nevertheless, systems, devices, materials, and methods within the knowledge of those skilled in the art are also within the scope of this teaching and may be used according to representative embodiments. It should be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting. The defined terminology is supplementary to its technical and scientific meaning as commonly understood and accepted in the art field of this teaching.

[0012] It will be understood that although the terms first, second, third, etc., may be used herein to describe various elements or components, these elements or components should not be limited by these terms. These terms are used only to distinguish one element or component from another. Therefore, without departing from the teachings of the inventive concept, the first element or component discussed below may be referred to as the second element or component.

[0013] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting. As used in the specification and claims, the singular terms “a,” “an,” and “the” are intended to include both the singular and plural forms unless the context clearly specifies otherwise. Furthermore, the terms “comprising,” “including,” and / or similar terms specify the presence of stated features, elements, and / or components, but do not exclude the presence or addition of one or more other features, elements, components, and / or groups thereof. As used herein, the term “and / or” includes any and all combinations of one or more of the associated listed items.

[0014] Unless otherwise stated, when an element or component is considered "connected to," "coupled to," or "proximity to" another element or component, it is understood that the element or component may be directly connected to or coupled to the other element or component, or that there may be intermediate elements or components. That is, these and similar terms cover situations where one or more intermediate elements or components may be used to connect two elements or components. However, when an element or component is considered "directly connected to" another element or component, this only covers situations where two elements or components are connected to each other without any intermediate or intermediary elements or components.

[0015] Therefore, this disclosure is intended to present one or more of the advantages specifically pointed out below through its various aspects, embodiments, and / or specific features or sub-components. Example embodiments with specific details disclosed are set forth for purposes of explanation and not limitation in order to provide a thorough understanding of embodiments based on this teaching. However, other embodiments consistent with this disclosure that depart from the specific details disclosed herein remain within the scope of the claims. Furthermore, descriptions of well-known apparatuses and methods may be omitted so as not to obscure the description of the example embodiments. Such methods and apparatuses are within the scope of this disclosure.

[0016] In general, the various embodiments described herein provide systems and methods for detecting degradation in ultrasound imaging systems. One embodiment provides a processing pipeline that derives image quality (IQ) scores from ultrasound images acquired by a monitored ultrasound imaging system. This pipeline includes a cardiac view model that identifies cardiac views of the acquired ultrasound images and selects an IQ model specific to that cardiac view. The selected IQ model determines an IQ metric for each ultrasound image, and the outputs of the IQ models are aggregated into an IQ metric for each ultrasound image. The aggregated IQ metrics are continuously monitored and predicted over time by a degradation detection model to enable the detection of early signs of IQ degradation.

[0017] Detected IQ degradation can be traced back to various characteristics of operating the ultrasound imaging system, such as age, frequency of use, and total duration of use. In various embodiments, IQ degradation can be traced back to characteristics unrelated to the age or use of the ultrasound imaging system. For example, IQ metrics can be grouped and filtered using different criteria, such as the transducer probe in use, system settings, and the body mass index (BMI) of the ultrasound physician or patient acquiring the images. One or more machine learning algorithms can support IQ analysis over time by providing estimates of whether IQ degradation is occurring for specific filter options.

[0018] Figure 1 This is a simplified block diagram of a system for detecting degradation of an ultrasound imaging system according to a representative embodiment.

[0019] refer to Figure 1System 100 includes a workstation 105 for implementing and / or managing the process described herein for detecting degradation of ultrasound images from ultrasound imaging system 140. Workstation 105 includes one or more processors indicated by processor 120, one or more memories indicated by memory 130, a user interface 122, and a display 124. Processor 120 communicates with ultrasound imaging system 140 via an imaging interface (not shown). Ultrasound imaging system 140 includes an ultrasound controller (base system) 143 and a transducer probe 145 operable by an operator to obtain ultrasound images of portions of a subject (patient) 150. Transducer probe 145 can be manually operated by an operator (e.g., an ultrasound physician, a doctor), automatically operated by a robot under the control of a robot controller (not shown), or a combination of both.

[0020] The transducer probe 145 may include a 2D matrix array of transducer elements capable of scanning in two or three dimensions for, for example, transmitting ultrasonic waves into the body of the object 150 and receiving echo signals in response. For example, the transducer elements may include capacitive micromechanical ultrasonic transducers (CMUTs) or piezoelectric transducers formed of materials such as lead zirconate titanate (PZT) or polyvinylidene fluoride (PVDF), but other types of transducer materials may be included without departing from the scope of this teaching. The transducer array may be coupled to a microwave beamformer in the ultrasonic transducer probe 145, which controls the transmission and reception of signals by the transducer elements.

[0021] Transducer probe 145 is connected to ultrasound controller 143 via probe cable 147. Ultrasound controller 143 is configured to control the ultrasound imaging process and includes known elements for performing ultrasound imaging, such as a transmit / receive (T / R) switch configured to switch between transmit and receive modes, and a main beamformer configured to provide final beamforming. One of the functions performed by ultrasound controller 143 is the direction in which the beam is steered and focused. For example, the beam may be steered directly forward from the transducer array of transducer probe 145 (orthogonal to the transducer array of transducer probe 145), or at different angles to obtain a wider field of view. Generally, the transmission of ultrasound signals and the reception and processing of echo signals in response are known, and therefore additional details in this regard are not included herein. In various embodiments, all or part of the functionality of ultrasound controller 143 may be implemented by processor 120.

[0022] Memory 130 stores instructions executable by processor 120. When executed, the instructions cause processor 120 to perform one or more procedures for evaluating muscle reduction in subject 150 using ultrasound images acquired by ultrasound imaging system 140. Ultrasound images may be provided from ultrasound imaging system 140 in real-time or near real-time during the scan procedure, or may be retrieved from storage after the scan procedure. For illustrative purposes, memory 130 is shown as including software modules, each of which includes instructions executable by processor 120 corresponding to an associated capability of system 100.

[0023] Processor 120 represents one or more processing devices and may be implemented using any combination of hardware, software, firmware, hardwired logic circuitry, or a combination of general-purpose computers, central processing units (CPUs), digital signal processors (DSPs), graphics processing units, computer processors, microprocessors, state machines, programmable logic devices, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), or similar devices. Any processor or processing unit herein may include multiple processors, parallel processors, or both. Multiple processors may be included in or coupled to a single device or multiple devices. As used herein, the term "processor" encompasses an electronic component capable of executing a program or machine-executable instructions. A processor may also refer to a collection of processors within a single computer system or distributed across multiple computer systems (such as in cloud-based applications or other multi-site applications). A program has software instructions that can be executed by one or more processors that may be within the same computing device or distributed across multiple computing devices.

[0024] Memory 130 may include main memory and / or static memory, wherein such memories can communicate with each other and with processor 120 via one or more buses. Memory 130 may be implemented, for example, by any number, type, and combination of random access memory (RAM) and read-only memory (ROM), and may store various types of information, such as software algorithms, artificial intelligence (AI) machine learning models, and computer programs, all of which can be executed by processor 120. Various types of ROM and RAM may include any number, type, and combination of computer-readable storage media, such as disk drives, flash memory, electrically programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disks, removable disks, magnetic tapes, optical disc read-only memory (CD-ROM), digital universal disk (DVD), floppy disks, Blu-ray discs, universal serial bus (USB) drives, or any other form of storage media. Memory 130 is a tangible storage medium for storing data and executable software instructions, and is non-transient during the time the software instructions are stored therein. As used herein, the term "non-transient" should not be interpreted as a permanent property of a state, but rather as a property of a state that will persist for a period of time. The term "non-transient" explicitly negates transient properties, such as carrier waves or signals, or other forms of properties that exist only temporarily at any time and place. Memory 130 may store software instructions and / or computer-readable code that enable the execution of various functions. Memory 130 may be secure and / or encrypted, or insecure and / or unencrypted.

[0025] System 100 may also include a database 112 for storing information that can be used by various software modules of memory 130. For example, database 112 may include image data from previously obtained ultrasound images of object 150 and / or other similar locations. For example, the stored image data may be used to train AI machine learning models, such as neural network models, as described below. Database 112 may be implemented by, for example, any number, type, and combination of RAM and ROM. Various types of ROM and RAM may include any number, type, and combination of computer-readable storage media, such as disk drives, flash memory, EPROM, EEPROM, registers, hard disks, removable disks, magnetic tape, CD-ROMs, DVDs, floppy disks, Blu-ray discs, USB drives, or any other form of storage media known in the art. Database 112 includes tangible storage media for storing data and executable software instructions, and is non-transient during the time the data and software instructions are stored therein. Database 112 may be secure and / or encrypted, or insecure and / or unencrypted. For illustrative purposes, database 112 is shown as a separate storage medium, but it should be understood that it may be combined with and / or included in memory 130 without departing from the scope of this teaching.

[0026] Processor 120 may include or have access to an artificial intelligence (AI) engine, which may be implemented as software that provides artificial intelligence (e.g., neural network models) and applies the machine learning described herein. For example, the AI ​​engine may reside in any of a variety of components other than or different from processor 120, such as memory 130, an external server, and / or the cloud. When the AI ​​engine is implemented in the cloud (e.g., at a data center), for example, the AI ​​engine may be connected to processor 120 via the Internet or other communication networks using one or more wired and / or wireless connections. In various embodiments, for example, all or part of the processes provided by the first, second, and third machine learning algorithms discussed below may be implemented by the AI ​​engine. The first, second, and third machine learning algorithms cannot actually be executed in the human brain.

[0027] User interface 122 is configured to provide a user with information and data output from processor 120, memory 130, and / or ultrasound imaging system 140 and / or to receive information and data input by the user. That is, user interface 122 enables a user to input data and control or manipulate aspects of the processes described herein, and also enables processor 120 to indicate the effects of user input, which may include control or manipulation of the transducer probe when the user / operator is present. All or part of user interface 122 may be implemented by a graphical user interface (GUI), such as by GUI 128 viewable on display 124, as discussed below. User interface 122 may include one or more interface devices, such as a mouse, keyboard, trackball, joystick, microphone, camera, touchpad, touchscreen, or voice or gesture recognition captured by a microphone or camera.

[0028] For example, display 124 may be a monitor, such as a computer monitor, television, liquid crystal display (LCD), organic light-emitting diode (OLED), flat panel display, solid-state display, or cathode ray tube (CRT) display or electronic whiteboard. Display 124 includes a screen 126 for viewing ultrasound images of object 150, various features described herein for conveying the degree of image degradation (if any) to the user, and a GUI 128 for enabling the user to interact with the displayed images and features. In an embodiment, ultrasound imaging system 140 may include a separate dedicated display for acquiring ultrasound images, wherein the dedicated display is also represented by display 124.

[0029] Reference memory 130 stores a set of data and instructions executable by processor 120 to detect image degradation, as described above. Ultrasound image module 131 is configured to receive and process ultrasound images of a region of interest in object 150 acquired by ultrasound imaging system 140 (including transducer probe 145). For illustrative and non-limiting purposes, it may be assumed that the region of interest is the heart of object 150, and that the ultrasound images are two-dimensional (2D) cardiac ultrasound images or 2D slices of three-dimensional (3D) cardiac ultrasound images. Of course, in practice, the same procedures described herein can be applied to other regions of interest and / or other types of ultrasound imaging, such as fetal imaging or general ultrasound imaging, without departing from the scope of this teaching. Ultrasound image module 131 may also store data associated with the ultrasound images, such as the time and date of image acquisition, the identifier of ultrasound imaging system 140, and the identifier of the operator (e.g., sonographer, physician) who operated ultrasound imaging system 140 when the ultrasound images were acquired.

[0030] Ultrasound images can be displayed on display 124. Ultrasound images can be received from ultrasound imaging system 140 in real time or near real time, for example, during a simultaneous imaging session of subject 150. In particular, the display of real-time images allows the operator to visualize the anatomy of subject 150 while manipulating transducer probe 145. Alternatively or additionally, ultrasound images can be previously acquired (historical) images obtained during one or more previous imaging sessions, which have been retrieved from storage devices (e.g., database 112), as described above.

[0031] The cardiac view module 132 is configured to automatically determine a cardiac view of the heart from at least one ultrasound image received by the ultrasound image module 131. The cardiac view module 132 may include a cardiac view (first) machine learning algorithm that takes at least one ultrasound image as input, compares the at least one ultrasound image with a plurality of predetermined cardiac views trained using it, and outputs the cardiac view most similar to (closest match to) the cardiac view provided by the input ultrasound image. Examples of different cardiac views that can be recognized by the cardiac view module 132 include a apical two-chamber view, a apical three-chamber view, a apical four-chamber view, a parasternal short-axis view, a parasternal long-axis view, and a subcostal view.

[0032] The cardiac view machine learning algorithm can be implemented as any suitable type of trainable machine learning algorithm or model, such as, for example, a convolutional neural network (CNN), an artificial neural network (ANN), a visual transformer, or a U-net model. The cardiac view machine learning algorithm can be trained in a supervised manner using historical ultrasound images of the heart from various cardiac views as input and manually annotated views as targets. This provides a training dataset that associates cardiac features from historical ultrasound images of the heart with various cardiac views. For example, the various cardiac views may correspond to different angles from which the images were acquired. For example, the training dataset can be stored in database 112. After training on historical ultrasound images and view targets, the cardiac view machine learning algorithm does not necessarily require additional historical data, as it is capable of processing new ultrasound images and predicting the corresponding cardiac views.

[0033] IQ model module 133 is configured to automatically select an IQ model from a plurality of previously provided IQ models based on a cardiac view of the heart, as determined by cardiac view module 132, and to use the selected IQ model to determine one or more IQ metrics associated with the ultrasound image. IQ metrics are objective measures of quality that correlate well with a human observer's subjective perception of image quality, while also rating unperceived errors, as will be apparent to those skilled in the art. Examples of IQ metrics include sharpness, signal-to-noise ratio (SNR), visibility of anatomical structures (e.g., heart wall, valves), correctness of the viewing plane, and other metrics mentioned below.

[0034] Multiple IQ models correspond to multiple predetermined cardiac views of the heart, respectively, in the training dataset of the cardiac view machine learning algorithms constituting cardiac view module 132, as discussed above. Each of the IQ models includes a trained IQ (second) machine learning algorithm that predicts one or more IQ metrics for an input ultrasound image at the corresponding cardiac view in order to determine image quality. Each of the IQ machine learning algorithms can be implemented as any suitable type of trainable machine learning algorithm or model, such as CNN, ANN, visual transformer, or U-net models. Each of the IQ machine learning algorithms can be trained in a supervised manner using historical ultrasound images of the heart from the corresponding cardiac views as input and manually annotated views showing the IQ metrics as targets. This provides a training dataset that correlates image quality metrics in ultrasound images (such as sharpness, SNR, visibility of anatomical structures, visual plane correctness, and contrast, noise, wall visibility, and / or valve visibility) with IQ metrics. For example, IQ metrics for the tip 2D view may include the visibility of the walls in the left ventricle (LV), the visibility of the right wall of the LV, and the visibility of the LV tip, but other types of IQ metrics may be included without departing from the scope of this teaching. A similar approach could be used for valve visibility, where an annotator marks whether the valve is correctly visible in the image. For example, the training dataset could be stored in database 112.

[0035] The IQ time series module 134 is configured to provide time series of image quality. In an embodiment, the IQ time series module 134 receives IQ metrics output by a selected IQ model from the cardiac view module 132 and aggregates the IQ metrics into a single value for each time point in the time series. For example, IQ metrics for the same image can be averaged simultaneously to provide an aggregated value. The aggregation of IQ metrics may also include a weighted average, where more weight is given to certain IQ metrics that are considered more important. The IQ time series module 134 may also score the aggregated IQ metrics associated with the ultrasound image to provide an aggregated IQ score, where higher values ​​of the aggregated IQ metrics receive a higher score.

[0036] The degradation detection module 135 is configured to estimate the degradation level of the ultrasound imaging system 140 based on a time series of image quality from the IQ time series module 134. According to various embodiments, the degradation level can be estimated using a degradation detection (third) machine learning algorithm. In an illustrative first embodiment, the degradation detection machine learning algorithm estimates the current degradation level of the ultrasound imaging system directly from a time series of aggregated IQ scores. In an illustrative second embodiment, the degradation detection machine learning algorithm estimates (predicts or forecasts) the future degradation level of the ultrasound imaging system 140 based on a time series trend of image quality, which is estimated, for example, from a time series of aggregated IQ scores using extrapolation. The degradation detection machine learning algorithm can be implemented as any suitable type of trainable machine learning algorithm or model, such as, for example, transformer-based networks, recurrent neural networks (RNNs), long short-term memory (LSTM) networks, or gated recurrent units (GRUs) discussed below. Architecturally, the degradation detection machine learning algorithm can be a recurrent model comprising one or more recurrent layers followed by a fully connected output layer.

[0037] Regarding the first embodiment, a degradation detection machine learning algorithm can be trained in a supervised manner using historical ultrasound imaging system data and associated historical fault data from multiple ultrasound imaging systems, where the historical fault data provides a degradation level based on a corresponding time series of image quality. For example, the multiple ultrasound imaging systems may or may not include ultrasound imaging system 140. The degradation level indicated by the time series may include a fault in the ultrasound imaging system, which would be 100% degradation. The degradation level may also include specific indicators of no fault occurrence, such as 25% or 50% degradation. In other words, a fault in the ultrasound imaging system is considered complete degradation, and other degradation levels of no fault occurrence are measured or otherwise quantified relative to this complete degradation.

[0038] The target used to train the degradation detection machine learning algorithm can be labeled based on historical data indicating whether a hardware failure of the corresponding ultrasound imaging system occurred after the time series was fed into the algorithm. In other words, the degradation detection machine learning algorithm is trained to predict hardware failures from time series inputs. Put another way, the goal of the degradation detection machine learning algorithm is to identify hardware failures when the time series indicates a hardware failure; this requires manual annotation and historical failure data. The percentage level of degradation can also be annotated based on this target. When the training data does not include hardware failures, the training data can be annotated to identify unacceptable image quality levels as targets. For example, a degradation threshold can be assigned to image quality corresponding to either a hardware failure or an unacceptable image quality level. The degradation detection machine learning algorithm can then estimate the degradation level of the ultrasound imaging system 140 by analyzing the time series of image quality from the IQ time series module 134 in the context of historical ultrasound imaging system failures. When a hardware failure is predicted, an alert can be issued to the user, for example, via display 124 or a networked display.

[0039] Time series can be tracked to provide internal quality checks for users (e.g., operators, managers, and / or service engineers) via a dashboard, for example, displayed on display 124. The dashboard can display additional information related to the time series of image quality, providing a more complete picture. For example, the time series can be filtered according to filtering criteria, which can be extracted from metadata, for example. Filtering criteria may include scan date, the type of transducer probe of the ultrasound imaging system, the type of ultrasound base system of the ultrasound imaging system, the operator's identity (e.g., sonographer), the physical characteristics of the object (e.g., BMI), and / or system settings (e.g., gain). For example, the sonographer's experience and skills are important parameters that directly affect image quality, and the object's BMI affects the difficulty of the scan. Scan date is important for time analysis. The data source for the dashboard may be a relational database or table containing filtering criteria, for example, stored in database 112. Alternatively, the degradation detection module 135 may provide a machine learning algorithm already trained for each filtering criterion, such as an algorithm for each sonographer, an algorithm for the object's BMI, etc.

[0040] In this embodiment, available filtering criteria can be provided to the user via a GUI 128 on screen 126. The GUI 128 can display optional elements or fields corresponding to different filters, each with different filtering criteria. For example, the GUI 128 can provide a list of filters or a drop-down list of filters with associated optional checkboxes, allowing the user to easily select each filter to be applied to the time series of image quality, for example using a mouse or touchscreen. Alternatively or additionally, the GUI 128 can provide a text field where the user can type in the desired filter. In response, the processor 120 applies the selected filter to the time series of image quality. The GUI 128 can provide other parameters to be set by the user, such as the duration of the time series or a degradation threshold for which the GUI 128 generates notifications or alarms. For example, the degradation threshold can be set to correspond to an impending failure of the ultrasound imaging system or a predetermined time prior to a failure.

[0041] Filtered time series and / or IQ metrics can be used to create meaningful charts to be displayed on monitor 124. For example, charts can be created showing the aggregated IQ metric of a specific ultrasound physician over time, or showing the aggregated IQ metric of a patient with a BMI greater than 30 over time. Displaying such charts allows operators, managers, or service engineers to analyze quality issues in more detail. Examples of time series filtered based on patient BMI and transducer probe are discussed below. Figure 2A and Figure 2B As shown in the image.

[0042] Regarding the second embodiment, the degradation detection machine learning algorithm can be a time-series prediction algorithm with self-supervised training and no manual annotation. Training uses historical ultrasound imaging system data over time, including IQ metrics. In this case, the degradation detection machine learning algorithm predicts the time-series trend and associated degradation over a predetermined period of time, which can be a relatively short time interval. The degradation detection machine learning algorithm can predict the time-series trend of image quality by extrapolating the time series provided by the IQ time-series module 134 over the predetermined time interval.

[0043] The predicted time series trend can be compared with a predetermined degradation threshold, which can be manually or automatically set as part of the degradation detection machine learning algorithm to determine the degradation level of the ultrasound imaging system 140 at future times. The length of the future time that can be resolved by the prediction can be a function of the time scale. For example, if the time series from the IQ time series module 134 spans several days, the degradation detection machine learning algorithm can predict the degradation level approximately one day in advance. If the time series from the IQ time series module 134 spans several weeks, the degradation detection machine learning algorithm can predict the degradation level approximately one week in advance or approximately +10% of the current time boundary.

[0044] In this embodiment, when the estimated degradation level drops below or near a predetermined degradation threshold, the degradation detection machine learning algorithm can, for example, notify or alert the user via display 124 or a networked display. The degradation detection machine learning algorithm targets the future IQ value itself and does not require manual annotation and historical fault data, as described above.

[0045] Figure 2A It is a graph showing the time series of ultrasound imaging systems filtered by the patient's BMI, and Figure 2B This is a graph showing the time series of an ultrasound imaging system filtered by a transducer probe according to a representative embodiment.

[0046] refer to Figure 2A Trajectory 201 depicts a time series of ultrasound image quality over a predetermined time period. The time series is filtered based on patients with a BMI exceeding 30. In this case, the time series provided by the degradation detection module 135 shows that the ultrasound imaging system has not degraded within the predetermined time period. Therefore, the user does not recommend or take any action.

[0047] refer to Figure 2B Trace 202 depicts a time series of ultrasound image quality over the same predetermined time period. The time series is filtered according to a specific transducer probe (US probe ID 12). In this case, the time series provided by the degradation detection module 135 shows the possible degradation of the ultrasound imaging system near the end of the predetermined time period, where the estimated degradation is highlighted by dashed box 212. The left edge of box 212 may be located at the time when the image quality degrades below a predetermined threshold T indicated on the image quality axis. Therefore, prior to box 212, the degradation detection module 135 may recommend and / or allow the user to take action to remove the specific transducer probe.

[0048] Figure 3This is a flowchart of a method for detecting degradation of an ultrasound imaging system according to a representative embodiment. For example, the method may be implemented at least in part using instructions stored in memory 130 and executable by processor 120 in system 100.

[0049] refer to Figure 3 In box S311, multiple ultrasound images of the subject's heart are received. The ultrasound images are acquired by an ultrasound imaging system and can be received, for example, directly from the ultrasound system during a patient examination, or from a database of previously acquired ultrasound images.

[0050] In block S312, a cardiac view of the heart is automatically determined based on at least one of multiple ultrasound images acquired by an ultrasound imaging system. For example, the cardiac view can be determined by applying at least one ultrasound image to a trained cardiac view (first) machine learning algorithm. The cardiac view of the heart is one of multiple predetermined cardiac views. The cardiac view machine learning algorithm is configured to compare at least one ultrasound image with multiple predetermined cardiac views and identify the closest match as the determined cardiac view of the heart.

[0051] In box S313, an IQ model is automatically selected from multiple previously provided IQ models based on the cardiac view of the heart determined from box S312. The multiple IQ models correspond to multiple predetermined cardiac views of the heart, respectively. Each IQ model, including the selected IQ model, includes a trained IQ (second) machine learning algorithm.

[0052] In box S314, an IQ metric is determined for each of the multiple ultrasound images. The IQ metric can be determined by applying each of the multiple ultrasound images received in box S311 to a selected IQ model chosen in box S313. The IQ metric indicates the quality of the ultrasound image.

[0053] In box S315, the IQ metrics associated with each ultrasound image are aggregated to provide an aggregated IQ score associated with that ultrasound image. For example, a mean or average value can be determined to provide the aggregated IQ score. The aggregated IQ scores are collected over time to provide a time series of the aggregated IQ scores.

[0054] In box S316, the degradation level of the ultrasound imaging system is estimated by analyzing the time-series trend of the aggregated IQ score associated with the ultrasound images over time. A degradation detection machine learning algorithm is also used to estimate the degradation level.

[0055] In this embodiment, the estimated degradation level is the current degradation level of the ultrasound imaging system, which is estimated directly from the time series of aggregated IQ scores. In this case, a degradation detection machine learning algorithm is previously trained using historical ultrasound imaging system data (including IQ metrics) and associated historical failure data. The trained degradation detection machine learning algorithm is applied to the time series from block S315 to estimate the current degradation level of the ultrasound imaging system. The degradation level can be provided as a percentage of total failure (100%). For example, an estimated degradation level equal to 75 would indicate a 75% failure level.

[0056] In another embodiment, for example, the estimated degradation level is a predicted degradation level of the ultrasound imaging system based on the trend of a time series of aggregated IQ scores predicted to future times and compared with a predetermined threshold. In this case, a degradation detection machine learning algorithm is previously trained using historical ultrasound imaging system data (including IQ metrics) over a period of time. The trained degradation detection machine learning algorithm is applied to the time series of aggregated IQ scores from block S315 and predicts a corresponding time series trend of image quality to future times based on the aggregated IQ scores, for example, by extrapolating the time series trend. The degradation level of the ultrasound imaging system at future times is estimated by analyzing the predicted time series trend of image quality. The estimated degradation level may be based on a predetermined degradation threshold, wherein the predicted aggregated IQ scores at future times according to the time series trend are compared with the predetermined degradation threshold. Closeness to the predetermined degradation threshold indicates the estimated degradation level, and exceeding the threshold indicates a predicted failure of the ultrasound imaging system at future times.

[0057] In block S317, the estimated degradation level is notified to the user. In an embodiment, the user may be notified based on a predetermined degradation threshold. For example, the user may be notified when the estimated degradation level falls below the threshold, or when the estimated degradation level approaches the threshold, allowing the user to take action before the threshold is reached. The user may be notified visually via a display, but other types of notification, such as audio tone or flashing lights, may be provided without departing from the scope of this teaching.

[0058] According to various embodiments of this disclosure, the methods described herein can be implemented using a hardware computer system that executes a software program stored on a non-transient storage medium. Furthermore, in exemplary, non-limiting embodiments, implementations may include distributed processing, component / object distributed processing, and parallel processing. Virtual computer system processing can implement one or more of the methods or functions described herein, and the processors described herein can be used to support virtual processing environments.

[0059] Although the evaluation of the quality of an ultrasound imaging system has been described with reference to exemplary embodiments, it should be understood that the language used is descriptive and illustrative, not limiting. Changes may be made within the scope and spirit of the embodiments, as stated and modified in the claims. Furthermore, while the evaluation of the quality of an ultrasound imaging system has been described with reference to specific modules, materials, and embodiments, it is not intended to be limited to the disclosed details; rather, the evaluation of the quality of an ultrasound imaging system extends to functionally equivalent structures, methods, and uses within the scope of the claims.

[0060] The illustrations of the embodiments described herein are intended to provide a general understanding of the structure of the various embodiments. These illustrations are not intended to be a complete description of all elements and features of the disclosure described herein. Many other embodiments will be apparent to those skilled in the art upon viewing this disclosure. Other embodiments can be utilized and derived from this disclosure, allowing structural and logical substitutions and changes to be made without departing from the scope of this disclosure. Furthermore, the illustrations are merely representative and may not be drawn to scale. Some scales within the illustrations may be exaggerated, while others may be minimized. Therefore, this disclosure and the accompanying drawings should be considered illustrative rather than restrictive.

[0061] One or more embodiments of this disclosure may be referred to individually and / or collectively by the term "invention" herein, merely for convenience and not intended to voluntarily limit the scope of this application to any particular invention or inventive concept. Furthermore, although specific embodiments have been illustrated and described herein, it should be appreciated that any subsequent arrangements designed to achieve the same or similar purpose may replace the specific embodiments shown. This disclosure is intended to cover any and all subsequent modifications or variations of the various embodiments. After reviewing this specification, combinations of the above embodiments and other embodiments not specifically described herein will be apparent to those skilled in the art.

[0062] This abstract of disclosure is provided in accordance with 37 C. FR § 1.72(b) and is submitted with the understanding that it is not intended to interpret or limit the scope or meaning of the claims. Furthermore, in the foregoing detailed description, various features may be grouped together or described in a single embodiment for the purpose of simplifying this disclosure. This disclosure should not be construed as reflecting an intention that the claimed embodiments require more features than expressly recited in each claim. Rather, as reflected in the following claims, the inventive subject matter may relate to fewer features than all of any of the disclosed embodiments. Therefore, the following claims are incorporated into the detailed description, each claim serving as itself to define a separately claimed subject matter.

[0063] The prior description of the disclosed embodiments is provided to enable any person skilled in the art to practice the concepts described in this disclosure. Thus, the subject matter of the above disclosure should be considered illustrative rather than restrictive, and the claims are intended to cover all such modifications, enhancements, and other embodiments falling within the true spirit and scope of this disclosure. Therefore, to the maximum extent permitted by law, the scope of this disclosure shall be determined by the broadest permissible interpretation of the claims and their equivalents, and should not be construed as limited or restricted by the foregoing detailed description.

Claims

1. A method for detecting degradation of an ultrasound imaging system (140), the method comprising: Receive multiple ultrasound images of the heart of the object (150) acquired by the ultrasound imaging system (S311); The heart view of the heart is automatically determined based on the at least one ultrasound image from the plurality of ultrasound images by applying a trained heart view machine learning algorithm, wherein the heart view of the heart is one of a plurality of predetermined heart views of the heart (S312). Based on the determined cardiac view of the heart, an IQ model is automatically selected from a plurality of image quality (IQ) models, wherein the plurality of IQ models correspond to the plurality of predetermined cardiac views of the heart, and wherein the selected IQ model includes a trained IQ machine learning algorithm (S313). Multiple IQ measures associated with each ultrasound image are determined by applying each of the multiple ultrasound images to a selected IQ model (S314). The multiple IQ measures are aggregated to provide an aggregated IQ score associated with each of the multiple ultrasound images, thereby creating a time series of the aggregated IQ score (S315). Using a degradation detection machine learning algorithm, the degradation level of the ultrasound imaging system is estimated based on the time series of the aggregated IQ scores associated with the multiple ultrasound images over time (S316); and Inform the user of the estimated level of degradation (S317).

2. The method according to claim 1, wherein, Estimating the degradation level of the ultrasound imaging system includes: The degradation detection machine learning algorithm is trained in a supervised manner using historical ultrasound imaging system data from multiple ultrasound imaging systems and associated historical fault data, the historical ultrasound imaging system data including IQ metrics; The time series of aggregated IQ scores is received as input to a trained degradation detection machine learning algorithm; and The degradation level of the ultrasound imaging system is estimated based on the time series of the aggregated IQ score.

3. The method according to claim 1, wherein, Estimating the degradation level of the ultrasound imaging system includes: The degradation detection machine learning algorithm is trained in a self-supervised manner using historical ultrasound imaging system data over time from multiple ultrasound imaging systems, including IQ metrics, without manual annotation. The time series of the aggregated IQ scores is received as input to a trained degradation detection machine learning algorithm; Predicting time-series trends of image quality in the future based on aggregated IQ scores; The degradation level of the ultrasound imaging system at future times is estimated by analyzing the predicted time series trend of image quality; and The estimated level of degradation is notified to the user based on a predetermined degradation threshold.

4. The method according to claim 3, wherein, Predicting the time series trend involves extrapolating the aggregated IQ score in the time series to the future time.

5. The method according to claim 2, further comprising: The aggregated IQ score is filtered using a filtering criterion, wherein the time series of the aggregated IQ score is based on the filtered aggregated IQ score associated with the multiple ultrasound images.

6. The method according to claim 5, wherein, The filtering criteria include one of the following: the type of transducer probe of the ultrasound imaging system, the type of the ultrasound underlying system of the ultrasound imaging system, the identity of the ultrasound physician, or the physical characteristics of the object.

7. The method according to claim 5, further comprising: Display multiple filters on the screen using the graphical user interface; Receives a user's selection of filters from a plurality of filters to be applied to the aggregated IQ score of the time series, wherein the selected filters include the filtering criteria; In response to the selection, the filtering criteria are used to perform the filtering of the aggregated IQ score; and The filtered aggregated IQ score over time is displayed on the monitor.

8. The method according to claim 1, wherein, The plurality of IQ measures include at least one of the following: wall visibility or valve visibility.

9. A system for detecting degradation of an ultrasound imaging system (140), the system comprising: Display (124); Processor (120); as well as A non-transient memory (130) stores instructions that, when executed by the processor, cause the processor to: Receive multiple ultrasound images of the heart of the object acquired by the ultrasound imaging system; The cardiac view of the heart is automatically determined based on the at least one ultrasound image from the plurality of ultrasound images by applying a trained cardiac view machine learning algorithm, wherein the cardiac view of the heart is one of a plurality of predetermined cardiac views of the heart. An IQ model is automatically selected from multiple image quality (IQ) models based on the determined cardiac views of the heart, wherein the multiple IQ models correspond to the multiple predetermined cardiac views of the heart, and wherein the selected IQ model includes a trained IQ machine learning algorithm. Multiple IQ measures associated with each ultrasound image are determined by applying each of the multiple ultrasound images to a selected IQ model; The multiple IQ measures are aggregated to provide an aggregated IQ score associated with each of the multiple ultrasound images, thereby creating a time series of the aggregated IQ score; The degradation level of the ultrasound imaging system is estimated using a degradation detection machine learning algorithm based on the time series of the aggregated IQ scores associated with the multiple ultrasound images over time; and The notification regarding the estimated level of degradation is displayed on the display.

10. The system according to claim 9, wherein, The instructions cause the processor to estimate the degradation level of the ultrasound imaging system by performing the following operations: The degradation detection machine learning algorithm is trained in a supervised manner using historical ultrasound imaging system data from multiple ultrasound imaging systems and associated historical fault data, the historical ultrasound imaging system data including IQ metrics; The time series of aggregated IQ scores is received as input to a trained degradation detection machine learning algorithm; as well as The degradation level of the ultrasound imaging system is estimated based on the time series of the aggregated IQ score.

11. The system according to claim 9, wherein, The instructions cause the processor to estimate the degradation level of the ultrasound imaging system by performing the following operations: The degradation detection machine learning algorithm is trained in a self-supervised manner using historical ultrasound imaging system data over time from multiple ultrasound imaging systems, including IQ metrics, without manual annotation. The time series of the aggregated IQ scores is received as input to a trained degradation detection machine learning algorithm; Predicting time-series trends of image quality in the future based on aggregated IQ scores; The degradation level of the ultrasound imaging system in the future time is estimated by analyzing the predicted time series trend of image quality; as well as The estimated level of degradation is notified to the user based on a predetermined degradation threshold.

12. The system according to claim 11, wherein, Predicting the time series trend involves extrapolating the aggregated IQ score in the time series to the future time.

13. The system according to claim 9, wherein, The trained cardiac view machine learning algorithm includes a convolutional neural network (CNN), an artificial neural network (ANN), or a U-net model, which has been trained in a supervised manner using multiple historical ultrasound images, which have been manually annotated to identify the corresponding cardiac views of the heart as the first training dataset.

14. The system according to claim 9, wherein, The trained IQ machine learning algorithm includes a CNN, ANN, or U-net model, which has been trained in a supervised manner using multiple historical ultrasound images, which have been manually annotated to classify the quality of at least one parameter in each of the multiple historical ultrasound images as second training data.

15. The system according to claim 9, wherein, Trained degradation detection machine learning algorithms include transformer-based networks, recurrent neural networks (RNNs), long short-term memory (LSTM) networks, or gated recurrent units (GRUs).

16. The system of claim 10, further comprising: A graphical user interface (GUI) capable of being displayed on the monitor, wherein the GUI is configured to receive selection by a user from a plurality of filters for the time series to be applied to the aggregated IQ score, wherein the selected filter includes corresponding filtering criteria. In response to the selection of the filter, the instruction further causes the processor to: The aggregated IQ score is filtered using the corresponding filtering criteria, wherein the time series of the aggregated IQ score is based on the filtered aggregated IQ score associated with the multiple ultrasound images; and The filtered aggregated IQ score over time is displayed on the monitor.

17. The system according to claim 16, wherein, The filtering criteria include one of the following: the type of transducer probe of the ultrasound imaging system, the type of the ultrasound underlying system of the ultrasound imaging system, the identity of the ultrasound physician, or the physical characteristics of the object.

18. A non-transient computer-readable medium storing instructions for detecting degradation of an ultrasound imaging system, the instructions causing the processor, when executed by a processor, to: Receive multiple ultrasound images of the heart of the object acquired by the ultrasound imaging system; A cardiac view of the heart is automatically determined based on at least one of the multiple ultrasound images by applying it to a trained cardiac view machine learning algorithm, wherein... The heart view of the heart is one of a plurality of predetermined heart views of the heart; An IQ model is automatically selected from multiple image quality (IQ) models based on the determined cardiac views of the heart, wherein the multiple IQ models correspond to the multiple predetermined cardiac views of the heart, and wherein the selected IQ model includes a trained IQ machine learning algorithm. Multiple IQ measures associated with each ultrasound image are determined by applying each of the multiple ultrasound images to a selected IQ model; The multiple IQ measures are aggregated to provide an aggregated IQ score associated with each of the multiple ultrasound images, thereby creating a time series of the aggregated IQ score; The degradation level of the ultrasound imaging system is estimated using a degradation detection machine learning algorithm based on the time series of the aggregated IQ scores associated with the multiple ultrasound images over time; and This allows a notification about the estimated level of degradation to be displayed to the user on the monitor.

19. The non-transient computer-readable medium according to claim 18, wherein, The instructions cause the processor to estimate the degradation level of the ultrasound imaging system by performing the following operations: The degradation detection machine learning algorithm is trained in a supervised manner using historical ultrasound imaging system data from multiple ultrasound imaging systems and associated historical fault data, the historical ultrasound imaging system data including IQ metrics; The time series of aggregated IQ scores is received as input to a trained degradation detection machine learning algorithm; as well as The degradation level of the ultrasound imaging system is estimated based on the time series of the aggregated IQ score.

20. The non-transient computer-readable medium according to claim 18, wherein, The instructions cause the processor to estimate the degradation level of the ultrasound imaging system by performing the following operations: The degradation detection machine learning algorithm is trained in a self-supervised manner using historical ultrasound imaging system data over time from multiple ultrasound imaging systems, including IQ metrics, without manual annotation. The time series of the aggregated IQ scores is received as input to a trained degradation detection machine learning algorithm; Predicting time-series trends of image quality in the future based on aggregated IQ scores; The degradation level of the ultrasound imaging system in the future time is estimated by analyzing the predicted time series trend of image quality; as well as The estimated level of degradation is notified to the user based on a predetermined degradation threshold.

Images