Artificial intelligence-based diagnosis using an in-motion multi-pulse X-ray source tomosynthesis imaging system.

Abstract

Presented is an X-ray diagnostic method and system that uses tomosynthesis imaging technology with an in-motion multi-pulse actuated X-ray source. While acquiring X-ray instrument image data, an artificial intelligence (AI) analyzes the patient response and compares the current pathology with the patient history and other patient information that may be part of the patient. The AI ​​reports changes in lesion location, sets severity thresholds and warning status, and generates treatment information. The AI ​​also recommends X-ray region of interest (ROI) scans, full X-ray CT scans, or other medical professionals and specialists.

Description

[[Technical Field]]

[0001] The present invention claims priority from Provisional Application No. 63182426 filed on April 30, 2021, Provisional Application No. 63226508 filed on July 28, 2021, Provisional Application No. 63170288 filed on April 2, 2021, Provisional Application No. 63175952 filed on April 16, 2021, Provisional Application No. 63194071 filed on May 27, 2021, Provisional Application No. 63188919 filed on May 14, 2021, Provisional Application No. 63225194 filed on July 23, 2021, Provisional Application No. 63209498 filed on June 11, 2021, Provisional Application No. 63214913 filed on June 25, 2021, Provisional Application No. 63220924 filed on July 12, 2021, Provisional Application No. 63222847 filed on July 16, 2021, Provisional Application No. 63224521 filed on July 22, 2021, and U.S. Application No. 17149133 filed on January 24, 2021 (which application claims priority from Provisional Application No. 62967325 filed on January 29, 2020), the contents of which are incorporated herein by reference.

[0002] The present invention generally relates to artificial intelligence (AI)-based methods and systems for diagnosis of lung and breast diseases, and more specifically relates to a method and system for artificial intelligence (AI)-based diagnosis using an in-motion multi-pulse pulsed X-ray source tomosynthesis imaging system. [[Background Art]]

[0003] Tomosynthesis (also referred to as digital tomosynthesis (DTS)) is a method for performing high-resolution limited-angle tomography at a radiation dose level comparable to that of projection radiography. Tomosynthesis is being investigated for a variety of clinical applications, including angiography, dental imaging, orthopedic imaging, mammography imaging, and lung imaging. A major advantage is that the DTS X-ray dose level is much lower than the dose level of CT imaging. DTS is also much faster and much lower cost than that of CT.

[0004] The in-motion multi-pulse tomosynthesis imaging system is a new type of DTS. It operates much faster and covers a much wider angle than conventional single-source DTSs. The in-motion multi-pulse tomosynthesis imaging system can be used for rapid lung cancer screening and breast cancer screening. The entire screening process takes only a few seconds.

[0005] While new types of DTS (Diagnostic Screening Systems) are ultra-fast, they still rely primarily on human physicians for diagnostic purposes. Therefore, the overall overhead adds up, slowing down the diagnostic process. Thus, it is desirable to provide much improved systems and methods for diagnosing lung or breast conditions to enable rapid lung and breast cancer screening for everyone. Artificial intelligence-based diagnostics are becoming necessary. [Overview of the project]

[0006] The presented method is an X-ray diagnostic technique using an in-motion multi-pulse activated source tomosynthesis imaging system. While acquiring X-ray instrument image data, artificial intelligence (AI) analyzes patient responses and compares the current condition with patient history and other patient information in a local computer or neural network that may be part of the patient's system. The AI ​​reports changes in the location of lesions, sets severity thresholds and warning statuses, and generates treatment information. The AI ​​also recommends scanning of the region of interest (ROI), a full CT scan, or referral to other medical professionals and specialists.

[0007] The advantages of this system include one or more of the following: This system leverages the high speed of DTS in conjunction with computer learning for diagnostic purposes. AI is used to ensure that diagnostic tools are used to accelerate data acquisition and diagnostic decision-making.

[0008] The In-Motion Multi-Pulse Tomosynthesis Imaging System is a rapid X-ray diagnostic device. It can perform near real-time low-dose X-ray scanning. The In-Motion Multi-Pulse Tomosynthesis Imaging System is also capable of performing 4D and sequential scanning. AI can manage all processes associated with the ultrafast tomosynthesis imaging system with very little human intervention.

[0009] Before performing the scan, a normal AI model is created from the selected data collection, and thresholds are determined for identifying anomalous measurements. Subsequently, anomalous measurements are identified from the newly acquired data and the created normal model. These anomalous measurements are then compared to the new thresholds to detect device anomalies.

[0010] Today, computers and networks are high-speed, and local storage is large, fast, and low-cost. For standalone systems, networks are not conveniently accessible; however, lesion diagnostic systems can preferably be implemented with locally running software.

[0011] The lesion diagnosis system is also implemented with software that runs on a neural network for networked systems. The lesion diagnosis system builds performance models for each subsystem in its normal operating mode and for each of several different possible failure modes.

[0012] Next, the AI ​​preferably dynamically predicts the performance of each subsystem based on the model's response to dynamically changing operating conditions, compares the actual performance and results of each subsystem with its dynamically predicted performance in both normal mode and possible failure modes, and determines the operating conditions based on these comparisons.

[0013] Using AI, it's possible to perform sequential scanning during the data acquisition phase. After obtaining results from partial scanning, the AI ​​can immediately decide whether continuous scanning is desirable. The AI ​​will stop if sufficient information has been collected.

[0014] The AI ​​behaves like a doctor, and the AI ​​scan behaves like a doctor's visit for a patient. If no changes are found, it probably means everything is normal. The process then stops. If some changes are found, the AI ​​decides to investigate the extent and location of the changes. The AI ​​then also determines how serious the changes are. If the changes are not serious, the process ends. If the changes are serious, the AI ​​generates an alert for the doctor and sends a recommendation to perform a scan of the region of interest (ROI), a CT scan, or even other more comprehensive diagnostic tools. [Brief explanation of the drawing]

[0015] [Figure 1] An exemplary in-motion multi-pulse X-ray source tomosynthesis imaging system is illustrated. [Figure 2] This shows an exemplary diagnostic scan of a patient subject. [Figure 3] This shows a flowchart of an exemplary AI-based diagnostic process performed by a multi-X-ray source in-motion tomosynthesis imaging system. [Modes for carrying out the invention]

[0016] The following paragraphs describe the present invention in detail, with reference to the accompanying drawings. Throughout this description, preferred embodiments and examples shown should be considered examples, not limitations, of the present invention. As used herein, “the present invention” refers to any one of the embodiments of the present invention described herein and any equivalents. Furthermore, reference to various features of “the present invention” throughout this specification does not mean that all claimed embodiments or methods must include the features(s) referenced.

[0017] However, this invention may be embodied in many different forms and should not be construed as being limited to the embodiments described herein. These embodiments are provided so as to ensure that this disclosure is detailed and complete and fully conveys the scope of the invention to those skilled in the art. Furthermore, all descriptions herein enumerating embodiments of the invention, and their specific examples, are intended to encompass both their structural and functional equivalents. In addition, such equivalents are intended to include both currently known equivalents and equivalents to be developed in the future (i.e., any elements to be developed that perform the same function regardless of their structure).

[0018] Therefore, it will be understood by those skilled in the art that, for example, figures, schematic diagrams, illustrations, etc., represent conceptual diagrams or processes illustrating systems and methods embodying the present invention. The functions of the various elements shown in the figures may be provided through the use of dedicated hardware and hardware capable of running associated software. Similarly, any switches shown in the figures are purely conceptual. Their functions may be performed through the operation of programmed logic, through dedicated logic, through the interaction of programmed control and dedicated logic, or even manually, and this particular technique may be selectable by the entity implementing the present invention. Those skilled in the art will further understand that the exemplary hardware, software, processes, methods, and / or operating systems described herein are for illustrative purposes only and are therefore not intended to be limited to any particular manufacturer.

[0019] This invention relates to a medical diagnostic method using an in-motion multi-pulse-actuated source tomosynthesis imaging system. The primary objective of this invention is to provide a method for image-based lesion diagnosis. A second objective of this invention is to provide image-based lesion diagnosis using artificial intelligence (AI). While acquiring X-ray instrument image data, the AI ​​analyzes the patient's response and compares the current condition with the patient's medical history and other patient information that may be part of the patient's condition. The AI ​​reports changes in the location of the lesion, sets severity thresholds and warning statuses, and generates treatment information. The AI ​​also recommends scanning of the region of interest (ROI), a full CT scan, or other medical professionals and specialists. An ultrafast, in-motion multi-pulse-actuated X-ray source tomosynthesis imaging system is designed. Dynamic changes occur due to sudden equipment failures during system operation. However, dynamic changes can be properly diagnosed and analyzed by comparing the dynamic changes with normal system responses.

[0020] In one embodiment, Figure 1 shows a novel type of X-ray imaging system. This X-ray imaging system is an in-motion multi-pulse-actuated X-ray source tomosynthesis imaging system 4 for performing highly efficient and ultrafast 3D X-ray imaging. There are multiple pulse-actuated X-ray sources mounted on a structure that is moving to form an array of sources. The multiple X-ray sources move simultaneously relative to the subject on a predetermined arc-shaped track at a constant velocity as a group. Each individual X-ray source can also move rapidly around the stationary position of the X-ray source, which is a short distance away. When the X-ray sources have a velocity equal to the group velocity but in the opposite direction of movement, the X-ray sources and X-ray flat panel detector are actuated via an external exposure control unit so that they remain stationary for a moment. This results in a significant reduction in the source travel distance for each X-ray source. As a result, 3D scanning can cover a much wider sweep angle in a much shorter time, and image analysis can also be performed in real time. This type of X-ray machine utilizes far more X-ray sources than other types of X-ray imaging machines to achieve much higher scanning speeds.

[0021] Multiple X-ray sources are mounted on a rotating gantry. The X-ray sources emit pulsed X-ray beams toward a target that the patient can image. The in-motion imaging device acquires image data from the region of interest as the patient moves from left to right. The in-motion imaging device preferably includes a microprocessor that collects and processes X-ray images of the patient, compares these images to previously acquired reference images, and generates various outputs. Data acquisition can be performed using one of the many digital imaging systems currently available for collecting digital image data. Examples of such devices include X-ray panel detectors that are either flat or curved.

[0022] The X-ray flat panel detector 3 receives an X-ray beam. The flat panel detector consists of an array of pixels. Each pixel has an individual intensity value corresponding to the X-ray energy received at its location. At each moment, each pixel may or may not receive X-ray radiation, depending on the number of times it has been exposed to the X-ray source within a time interval. The flat panel detector is positioned close to the X-ray source and the object during scanning. The flat panel detector receives X-ray photons from the X-ray source. The intensity of each pixel value depends on the energy of the X-ray photon reaching the location of that pixel. As a result, a single pixel may have a variety of pixel values ​​based on a variety of X-ray photon energies. These pixel values ​​represent the detector's response to a particular X-ray imaging modality. The particular modality depends on the application and imaging parameters used during acquisition. The X-ray detector can be used as a standalone device or as part of a larger imaging system.

[0023] The support frame structure comprises a substantially planar base, a substantially vertical support member, and a rotatable frame portion rotatably attached to the support member. The frame structure is sized and configured to support an X-ray tube at a preselected angular orientation relative to a patient on a movable platform. This enables the frame portion to rotate freely about the pivot arm. This configuration allows the angular orientation of the X-ray tube to be varied as needed during a scanning procedure without requiring rotation of the support member or the base. It will be understood that other configurations for attaching the frame portion to the base are possible without departing from the scope of the present invention.

[0024] Figure 2 shows a diagnostic scan of a patient 2 or a subject. The novel X-ray imaging system performs X-ray imaging primarily for pulmonary imaging or breast mammography using a plurality of pulsed X-ray sources in motion. The system can easily image up to 120 degrees or more of the area to be scanned within a few seconds. The system can already cover a relatively large span of angles from the first set of scans. For example, for a system with 5 X-ray sources and a total of 25 scans, the X-ray data sets from the emission sequence are a first set of 1-6-11-16-21, a second set of 2-7-12-17-22, a third set of 3-8-13-18-23, a fourth set of 4-9-14-19-24, a fifth set of 5-10-15-20-25, and so on. For five X-ray sources, this type of machine can easily achieve a total of more than 120 scans. After the first data set is acquired, artificial intelligence can immediately perform diagnosis without waiting for the second set. In Figure 2, there are a plurality of X-ray sources 1, and the patient 2 is positioned in front of an X-ray flat panel detector 3.

[0025] The in-motion multi-pulse-actuated X-ray source tomosynthesis imaging system 4 uses a high-speed pulse-actuated X-ray source to perform sequential scanning, near real-time low-dose tomosynthesis imaging. Multiple sequential in-motion tomosynthesis images are combined to form a higher-resolution image or video, but the two sequential images do not overlap. Sequential scanning is typically achieved by simultaneously performing imaging at two or more separate locations, each using a different tomosynthesis protocol. Multiple X-ray sources are preferably distributed along a tomosryngraph around the object of examination to obtain high-resolution X-ray images in any desired cross-sectional plane. A dedicated radiation plate collects the incident X-rays and directs them to a corresponding flat-panel detector via a suitable collimator. Sequential scanning using multiple sources offers several advantages, including speed and spatial resolution. Multiple exposures from the same tomosynthesis dataset can be merged into a higher-resolution image without any loss of diagnostic information. Therefore, it is not necessary to rescan the patient with a higher-resolution scanning technique to obtain a better-quality image. This increases efficiency and reduces radiation exposure for patients and healthcare providers.

[0026] In this vertical configuration, patient 2 is placed on an X-ray table. In other horizontal configurations, patient 2 can stand between the X-ray source 1 and the X-ray flat panel detector 3. The system operator or user directs or controls the overall procedure using a user interface that includes a display for visualizing selected data. A workstation with processing power (not shown) can be used instead of the user interface to control the overall procedure. The user interface is connected to local storage main memory and a processor via a network interface. The ultrafast X-ray source 1 directs multiple pulsed X-ray beams at the patient.

[0027] The X-ray flat panel detector 3, the X-ray source 1, and the collimator may be mounted on a table or a movable mount. The structure and operating principle of each of these elements are known in the art. The beam emitted by the X-ray source passes through the patient, then the collimator, and is then reflected by the beam splitter back to the patient. The in-motion multi-pulse activated X-ray source tomosynthesis imaging system 4 is an ultra-high-speed X-ray diagnostic apparatus. To achieve this object, one aspect of the disclosed invention includes the design of the in-motion multi-pulse activated X-ray source tomosynthesis imaging system 4 having a high data acquisition rate. This aspect also provides detailed X-ray imaging scanning.

[0028] During 3D X-ray imaging, artificial intelligence (AI) enables computer systems to perform tasks that require human intelligence such as vision, image recognition, and decision making. This type of task requires cognitive functions associated with the human mind, namely learning and problem solving. Machine learning is considered a subset of AI. Machine learning may be implemented using a deep learning (DL) process. DL is a machine learning method that employs mathematical models called neural networks. Neural networks may include numerous steps that attempt to mimic the human brain. When an X-ray imaging operation is performed, DL attempts to extract complex hierarchical features and patterns present in large image datasets. These features may then be synthesized together using a neural network to represent a model of X-ray image data.

[0029] This specification describes AI-based systems and methods for diagnosing the lung or breast, such as assisting in the detection of lung, breast cancer, or nodule conditions after an AI model has already been established, through a large amount of X-ray imaging data from actual patients. The AI-based systems and methods utilize machine learning. For example, the systems and methods may automate diagnosis by utilizing AI models in machine learning, such as deep learning models. The systems and methods described herein may be implemented as standalone or integrated applications for processing X-ray tomography images using artificial intelligence models related to image data.

[0030] Typically, one of the methods involves taking a human lung or breast X-ray scan as input and obtaining a model that outputs a diagnosis of this lung or breast. The diagnosis includes predicting lung or breast disease or nodules, the pathology of the lesion, predicting the severity level, and detecting important features in the input X-ray scan images.

[0031] Figure 3 shows an exemplary flowchart of AI-based diagnosis in an X-ray tomosynthesis imaging system after machine learning. Generally, before a patient undergoes a comprehensive X-ray scan, the patient's previous records or other standard information for a typical healthy individual are already available on either a local computer or a health network. The first step is to scan the patient to generate data. Image acquisition may be sequential. The second step is for the system to instantly construct images based on the sequential scan. The third step is to use artificial intelligence (AI) to compare this with available knowledge about the patient, such as the patient's medical history. The fourth step is to let the AI ​​make a decision. If there are no changes, the AI ​​may indicate that the patient scan is normal, and the process will stop. If there are changes in the scan indicating one or more lesions, the AI ​​will decide to obtain the extent of the newly detected lesions and the location of each lesion. The AI ​​will then also determine how serious the changes are. If the changes are not serious, the process will end. If the change is serious, the AI ​​will generate an alert for the doctor and send a recommendation to perform a more extensive scan, including a region of interest (ROI) scan or a CT scan.

[0032] Patient scanning generates raw data. To process the data, the acquired data must be compared to normal data in a reference memory unit. To verify that the raw data meets medical diagnostic criteria, a statistical model is constructed for the location of each anomaly in 3D space based on the geometric configuration of the patient-related device or device components. Based on the statistical model, the AI ​​calculates a decision on whether to perform a full scan or a partial scan. If the AI ​​then decides to perform a full scan, the in-motion multi-pulse activated source tomosynthesis imaging system proceeds to acquire an image, including exposing multiple pulse-activated sources, detectors, and tube elements for imaging. This can be done simultaneously or sequentially, but is preferably done simultaneously.

[0033] Reconstruction is preferably performed according to an in-motion multi-pulse-actuated source tomosynthesis imaging system technique. As a result, multiple images of a first subject or first patient taken from multiple different angles are obtained as a reconstructed tomosynthesis image set. The first subject of the patient is positioned relative to substantially stationary radiation sources and detectors, as described above. Subsequently, a radiation source control system, also referred to as a tracking system, generates a pulse-actuated X-ray emission pattern to direct the X-ray emission from multiple pulse-actuated sources onto the patient image. In a preferred embodiment, the pulse-actuated X-ray emission pattern corresponds to a collimated polygonal ring beam.

[0034] AI detection determines whether there is a change in the system structure. If AI detection determines that there is no change in the system structure, the AI ​​transmits a command to the X-ray detector to terminate the sequential scan and acquire the final image. The sequential scanning and acquisition of the final image may be repeated as needed to acquire one or more images. The AI ​​diagnostic module uses a variety of parameters, including but not limited to, dose value, partial volume effect (PVE), digital signal (DS), contrast (C), signal-to-noise ratio (SNR), image noise, thickness, histogram variance, pause time, X-ray scattering, attenuation coefficient, resolution, etc. The AI ​​scanning technique performs the scanning and analysis locally and notifies other locations via a network, using processing power and large-capacity storage for this analysis.

[0035] A threshold is set for the decision block in comparison with the patient's medical history or current data. The AI ​​scans multiple ROIs from the acquired image. In the area identification step, the AI ​​locates potential lesions from one or more ROIs and records the location of the lesion for the process block. In the detection confidence step, the AI ​​determines whether the recorded location is within the threshold. If the recorded location is not within the threshold, the process terminates. If the recorded location is within the threshold, the process proceeds to the severity classification determination step. In the severity classification determination step, the AI ​​determines whether the recorded location has a high severity rating or a low severity rating. If the recorded location has a high severity rating, the process proceeds to the second area identification step. In the second area identification step, the AI ​​may scan one or more additional ROIs from the acquired image to confirm the presence of a lesion. If the recorded location has a low severity rating, the process proceeds to the recommendation step. In the recommendation step, the AI ​​recommends scanning the lesion. Lesion scanning can be performed by standard scanning of a full CT scan, scanning of a region of interest (ROI), or focused scanning of multiple channels.

[0036] AI reports back to human doctors, provides recommendations on possible next steps, and generally shares and manages information with virtual (networked) medical professionals on a continuous basis.

[0037] A system with artificial intelligence (AI) for X-ray source in-motion diagnostic procedures can operate using different types of scanning. For example, one system may apply sequential scanning, while the other uses sequential scanning. Sequential scanning uses each data acquired in each snapshot to create the next image in a stack of images. This procedure continues until enough data has been collected to cover the region of interest (ROI). If the lesion is static, diagnosis becomes much easier using sequential scanning, as little or no information about the lesion's medical history is required. If the lesion is moving, sequential scanning can follow the lesion, but it requires high-frequency data acquisition, resulting in a significantly increased dose. Sequential scanning can also be performed in three stages: First, raw data is collected to build an initial normal model. Second, pre-processed data is collected to refine the normal model. Third, post-processed data is collected to further refine the normal model as needed. While acquiring X-ray instrument image data, the artificial intelligence analyzes the patient's response and compares the current condition with patient history and other patient information that may be part of the patient's condition. The AI ​​reports changes in the location of lesions, sets severity thresholds and warning statuses, and generates treatment information. The AI ​​also recommends scanning the region of interest (ROI), performing a full CT scan, or referring other healthcare professionals and specialists.

[0038] This application utilizes an in-motion multi-pulse activated source tomosynthesis imaging system. This application also relates to a method for a faster and more accurate X-ray diagnostic system by applying AI for real-time decision-making and scanning processes. Conventional X-ray diagnostic systems function as follows: acquire raw data, build a model of the acquired raw data against a model, build models of the normal operating mode and different failure modes, acquire real-time data from the X-ray detector array, analyze the real-time data based on each model, and detect whether there is an anomaly. In contrast, the system of the present invention functions as follows: an AI supercomputer is used to build performance models of each subsystem in the normal operating mode and each of several different possible failure modes. The AI ​​then dynamically predicts the performance of each subsystem based on the response of each model to dynamically changing operating conditions, compares the actual performance and results of each subsystem with its dynamically predicted performance in the normal mode and each of the possible failure modes, and determines the operating conditions based on these comparisons. Furthermore, the in-motion multi-pulse activated source tomosynthesis imaging system can perform near real-time low-dose X-ray scanning. The in-motion multi-pulse activated source tomosynthesis imaging system can also perform four-dimensional scanning or sequential scanning.

[0039] Image data obtained from each partial scan can be evaluated by artificial intelligence to detect potential lesions in various parts of the body, such as pulmonary nodules and breast cancer. Based on the detection, this disclosure generates alert information to help physicians decide whether they need to perform an ROI scan or CT scan, or use some other comprehensive diagnostic tools.

[0040] Various modifications and changes to the present invention will be apparent to those skilled in the art without departing from the spirit and scope of the invention as defined by the appended claims. It should be noted that the steps listed in the claims of any of the following methods do not necessarily have to be performed in the order they are listed. Those skilled in the art will recognize variations in order when performing the steps. In addition, the absence of mention or consideration of features, steps, or components provides a basis for a claim in which non-existent features or components are excluded by proviso or similar claim language.

[0041] While various embodiments of the present invention have been described above, it should be understood that they are presented only as examples and not as limitations. Various figures may depict exemplary architectures or other configurations for the present invention, and this is done to aid in understanding the features and functionalities that may be included in the present invention. The present invention is not limited to the exemplary architectures or configurations illustrated, and desired features may be implemented using a variety of alternative architectures and configurations. In fact, how alternative functional, logical, or physical divisions and configurations may be implemented to implement desired features of the present invention will be apparent to those skilled in the art. Furthermore, many different configuration module names other than those expressed herein may be applied to various divisions. In addition, with respect to flowcharts, operation descriptions, and method claims, the order in which the steps are presented herein does not obligate various embodiments to be implemented to perform the enumerated functionalities in the same order, unless the context indicates otherwise.

[0042] The terms and phrases used in this document, as well as their variations, should be interpreted as open-ended, rather than restrictive, unless otherwise explicitly stated. For example, the term “including” should be interpreted as “including, but not limited to,” the term “example” is used to provide illustrative examples of the items described, rather than an exhaustive or restrictive list, the terms “a” or “an” should be interpreted as “at least one,” “one or more,” and adjectives and similar terms such as “conventional,” “traditional,” “usual,” “standard,” and “known” should not be interpreted as limiting the items described in the items available in a given period or at a given time, but rather as encompassing conventional, traditional, usual, or standard techniques that are available or known now or at any future time. Therefore, where this document refers to a technique that is obvious or known to those skilled in the art, such a technique encompasses a technique that is obvious or known now or at any future time.

[0043] In addition, the various embodiments detailed herein are described with respect to exemplary block diagrams, flowcharts, and other explanatory diagrams. As will become apparent to those skilled in the art after reading this document, the exemplary embodiments and their various alternative forms may be implemented without being limited to the exemplary examples. For example, the block diagrams and their accompanying descriptions should not be construed as obligating a particular architecture or configuration.

[0044] The prior description of the disclosed embodiments is provided to enable those skilled in the art to create or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but rather to be granted the broadest scope consistent with the principles and novel features disclosed herein.

Claims

1. A diagnostic method using artificial intelligence with a tomosynthesis imaging system using a mobile multi-pulse X-ray source, The method involves acquiring a sequential scan of a patient using a tomosynthesis imaging system with a moving multi-pulse X-ray source, wherein the tomosynthesis imaging system is mounted on a moving structure and forms an array of X-ray sources that move simultaneously around an object as a group at a constant velocity along a predetermined arc trajectory, each X-ray source moves around a position that maintains a predetermined distance from the group, and when the velocity of the X-ray source is equal to the velocity of the group but in the opposite direction of movement, the X-ray source and the X-ray flat panel detector are activated via an external exposure control unit and temporarily maintain a stationary state, thereby acquiring the image. Based on the aforementioned sequential scanning, the image configuration is performed, The severity of the lesion is determined by using artificial intelligence (AI) to compare the differences between the previous scan and the sequential scan, A method comprising determining the location of the lesion and lesion proliferation using the AI.

2. The method according to claim 1, comprising using artificial intelligence (AI) to make a decision to stop if there is no change, otherwise to determine the severity of the lesion abnormality, and if so, recommend additional scanning by scanning a predetermined region of interest (ROI) or by CT scanning.

3. The method according to claim 1, comprising generating a warning and sending one or more recommendations to examine a region of interest (ROI) or a CT scan.

4. The method according to claim 1, comprising using AI to build a model of the performance of each subsystem in a normal operating mode and in each of a number of different possible failure modes.

5. Dynamically predict the performance of each subsystem based on the model's response to dynamically changing operating conditions, The actual performance and results of each subsystem are compared with the dynamically predicted performance of each subsystem in each of the normal operating mode and possible failure modes. The method according to claim 4, comprising determining the operating conditions based on these comparisons.

6. The method according to claim 1, comprising performing near real-time low-dose X-ray scanning.

7. The method according to claim 1, comprising performing a four-dimensional scan or sequential scan using a tomosynthesis imaging system with a moving multi-pulse X-ray source.

8. The method according to claim 1, which compares the current condition with the patient's medical history and patient information.

9. The method according to claim 1, comprising reporting changes in the location of a lesion, setting a severity threshold and a warning status, and generating treatment information.

10. The method according to claim 1, comprising scanning a region of interest (ROI) or a full CT scan, or recommending one or more medical professionals and specialists, based on the lesion and the location of lesion growth.

11. An X-ray tomosynthesis diagnostic system that uses artificial intelligence, A tomosynthesis imaging system using a moving multi-pulse X-ray source, wherein an array of X-ray sources is mounted on a moving structure and moves simultaneously as a group toward an object at a constant velocity along a predetermined circular trajectory, each X-ray source moves around a position that maintains a predetermined distance from the group, and when the velocity of the X-ray source is equal to the velocity of the group but in the opposite direction of movement, the X-ray source and the X-ray flat panel detector are activated via an external exposure control unit and temporarily maintain a stationary state, A processor coupled to a tomosynthesis imaging system using the aforementioned mobile multi-pulse X-ray source, wherein the processor comprises: Using the aforementioned tomosynthesis imaging system with a mobile multi-pulse X-ray source, sequential scanning of the patient is acquired, Based on the aforementioned sequential scanning, the image configuration is performed, The severity of the lesion is determined by using artificial intelligence (AI) to compare previous scans with the sequential scans, A system comprising a processor that executes code for determining the location of the lesion and lesion proliferation using the AI.

12. The system according to claim 11, further comprising artificial intelligence (AI) for making a decision to stop if there is no change, otherwise determining the severity of the lesion abnormality, and if so, recommending additional scanning by scanning a predetermined X-ray region of interest (ROI) or CT scan.

13. The system according to claim 11, comprising a code for generating a warning and transmitting one or more recommendations for inspecting X-ray region of interest (ROI) scanning or CT scanning.

14. The system according to claim 11, comprising code for using AI to build a model of the performance of each subsystem in a normal operating mode and in each of several different possible failure modes.

15. This involves dynamically predicting the performance of each subsystem based on the model's response to dynamically changing operating conditions, The actual performance and results of each subsystem are compared with the dynamically predicted performance of each subsystem in both the normal operating mode and the possible failure modes. The system according to claim 11, comprising a code for determining the aforementioned operating conditions based on these comparisons.

16. The system according to claim 11, comprising code for performing near real-time low-dose X-ray imaging scans.

17. The system according to claim 11, further comprising code for performing a four-dimensional scan or sequential scan using a tomosynthesis imaging system with a moving multi-pulse X-ray source.

18. The system according to claim 11, comprising a code for comparing the current medical condition with the patient's medical history and patient information.

19. The system according to claim 11, comprising a code for reporting changes in the location of a lesion, setting a severity threshold and a warning status, and generating treatment information.

20. The system according to claim 11, comprising a scan of a region of interest (ROI) or a full CT scan, or a code for recommending one or more medical professionals and specialists, based on the lesion and the location of lesion growth.

Images