Line-by-line scanning of a multi-motion pulse X-ray source tomosynthesis imaging system

Abstract

The present invention discloses a system and method for imaging acquisition from distributed wide-angle sparse partial scans. During real-time image reconstruction, artificial intelligence (AI) determines, based on an initial scan, whether there is sufficient information to perform a diagnosis. If there is sufficient information from a small partial scan, data acquisition stops; if more information is needed, the system progressively performs another round of wide-angle sparse scans at new positions until the result is satisfactory. The system reduces the X-ray dose applied to the patient and performs faster X-ray scans at multiple moving pulsed source tomosynthesis imaging systems. The method and system also significantly reduce the amount of time required to display high-quality three-dimensional tomosynthesis images.

Description

[0001] This invention claims priority to the following applications: Provisional Application Serial No. 63182426, filed Apr. 30, 2021; Provisional Application Serial No. 63226508, filed Jul. 28, 2021; Provisional Application Serial No. 63170288, filed Apr. 2, 2021; Provisional Application Serial No. 63175952, filed Apr. 16, 2021; Provisional Application Serial No. 63194071, filed May 27, 2021; Provisional Application Serial No. 63188919, filed May 14, 2021; Provisional Application Serial No. 63225194, filed Jul. 23, 2021; Provisional Application Serial No. 63209498, filed Jun. 11, 2021; Provisional Application Serial No. 63214913, filed Jun. 25, 2021; Provisional Application Serial No. 63220924, filed Jul. 12, 2021; Provisional Application Serial No. 63222847, filed Jul. 16, 2021; Provisional Application Serial No. 63224521, filed Jul. 22, 2021; and U.S. Application Serial 17149133, filed Jan. 24, 2021, which in turn claims priority to Provisional Serial 62967325, filed Jan. 29, 2020, the contents of which are incorporated by reference. Technical Field

[0002] The present invention generally relates to X-ray diagnostic imaging, and more particularly, to methods and apparatuses for sampling an imaging volume using a series of partial scans and using AI to determine whether information from image reconstruction is sufficient for multiple motion-pulsed X-ray source tomosynthesis imaging. Background Art

[0003] Tomosynthesis (also known as Digital Tomosynthesis (DTS)) is a method for performing high-resolution limited-angle tomography at a radiation dose level comparable to projection radiography. It has been studied for a variety of clinical applications, including angiography, dental imaging, orthopedic imaging, mammographic imaging, musculoskeletal imaging, and lung imaging. The DTS dose level is much lower than that of CT, DTS is also much faster than CT, and the cost of DTS itself is much lower. To further reduce the X-ray dose applied to the patient and even perform X-ray scans faster, the present invention relates to a method and system for image acquisition from a distributed wide-angle sparse partial scan. There are some prior arts for performing line-by-line image reconstruction using different resolutions. However, there are several disadvantages associated with the prior arts. The first disadvantage is that there is only a resolution change at the same position; the second disadvantage is that there is no perspective position information. The second disadvantage; the third disadvantage is that artificial intelligence (AI) is not involved. There are also other prior arts regarding line-by-line scans for CT. The first disadvantage is that the line-by-line scan is an incremental position and the angular coverage is quite small because it is only for CT. It includes an imaging technique by which an object is incrementally translated through a plurality of discrete scan positions to obtain CT data from an area of the object. In this regard, the object is not translated to the next scan position until valid or acceptable data is obtained for the current scan position. The second disadvantage is that it is slower in the case of one X-ray source. Most CT devices only have one X-ray source. It has to rotate a relatively large angle to get a larger coverage. The sampling imaging volume starts from a small angle and then increases incrementally. So, it will be slow. The third disadvantage in the prior art is that AI regarding incremental line-by-line scans is generally not involved in the prior art. Summary of the Invention

[0004] In one aspect, a method for line-by-line scanning using a multi-motion pulsed X-ray source tomosynthesis imaging system includes: placing an object at a predetermined position; controlling the multi-motion pulsed X-ray source tomosynthesis imaging system; using the tomosynthesis imaging system to obtain a first data set from different X-ray sources at different angles and performing image reconstruction; applying artificial intelligence with machine learning to perform diagnosis, and repeating the scan one or more times to achieve a predetermined image reconstruction quality; and constructing a 3D tomosynthesis volume therefrom.

[0005] On the other hand, in order to further reduce the X-ray dose applied to the patient and perform faster X-ray scans at a multi-motion pulsed source tomosynthesis imaging system, the present invention relates to a method for imaging acquisition from a distributed wide-angle sparse partial scan. During real-time image reconstruction, artificial intelligence (AI) determines based on the initial scan whether there is sufficient information to perform a diagnosis. The AI compares the diagnostic results with the patient's medical history in order to make an informed decision. If there is sufficient information from a small partial scan, data acquisition stops; if more information is needed, the system progressively performs another round of wide-angle sparse scans at new positions until the results are satisfactory.

[0006] Compared with other line-by-line scans in CT, this new invention has many advantages. The first advantage is its novel hardware device with multiple pulsed sources in motion. It runs in parallel and is much faster. The multi-motion pulsed source tomosynthesis imaging system has multiple X-ray sources. The second advantage is its sparsely distributed positions and much wider angles. The starting positions of the multiple sources have spanned a wide angle. Therefore, compared with the angles in CT, the first round of scans has covered a much wider angle. The third advantage is related to AI. AI is becoming increasingly popular now. Due to fast data acquisition and obtaining a first dataset covering a wider angle, AI can be used for real-time reconstruction to decide and make the next move. The AI compares the diagnostic results with the patient's medical history to make an informed decision. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 The flowchart shows a line-by-line scan using a multi-motion pulsed X-ray source tomosynthesis imaging system.

[0008] Figure 2 An exemplary multi-motion pulsed X-ray source tomosynthesis imaging system is shown.

[0009] Figure 3 The line-by-line scan positions of each X-ray source span a wide angle at the start. DETAILED DESCRIPTION

[0010] The present invention will be described in detail by way of example with reference to the accompanying drawings. Throughout the description, the preferred embodiments and examples shown should be considered exemplary rather than limiting 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. Additionally, references throughout the document to various features of "the present invention" do not mean that all claimed embodiments or methods must include the recited features.

[0011] However, the present invention may be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. These embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those of ordinary skill in the art. In addition, all statements herein reciting embodiments of the invention and specific examples thereof are intended to cover both structural and functional equivalents thereof. Moreover, it is intended that such equivalents include both currently known equivalents and equivalents developed in the future (i.e., any elements developed that perform the same function regardless of structure).

[0012] Thus, for example, those of ordinary skill in the art will recognize that diagrams, schematics, illustrations, etc. represent conceptual views or processes showing systems and methods embodying the invention. The functionality of the various elements shown in the figures can be provided by using dedicated hardware as well as hardware capable of executing associated software. Similarly, any switches shown in the figures are merely conceptual. Their functionality can be performed by the operation of program logic, by dedicated logic, by the interaction of program control and dedicated logic, or even manually, with the particular technique being selectable by the entity implementing the invention. Those of ordinary skill in the art also understand that the exemplary hardware, software, processes, methods, and / or operating systems described herein are for illustrative purposes and are thus not intended to be limited to any particular named manufacturer.

[0013] Figure 2 A plurality of moving pulse X-ray source tomosynthesis imaging systems 1 are shown. It is a novel X-ray imaging system that uses a plurality of moving pulse X-ray sources to perform efficient and ultrafast 3D radiography. There are a plurality of pulsed X-ray sources 2 mounted on a moving structure to form a source array. The plurality of X-ray sources 2 move simultaneously relative to the object at a constant speed as a group along a predefined arcuate track. Each individual X-ray source 2 can also move rapidly a small distance around its static position. When the X-ray source 2 has a speed equal to the group speed but in the opposite direction of movement, the X-ray source 2 and the X-ray flat panel detector 3 are activated by an external exposure control unit to be temporarily stationary. This results in a greatly reduced source travel distance for each X-ray source 2. 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 many more X-ray sources than other types of X-ray imaging machines in order to achieve a much higher scanning speed.

[0014] The primary motor is preferably an electric motor. An electrical signal generated by a motor controller is coupled to the primary motor. The motor controller is coupled to a programmable logic controller (PLC) that provides an output signal for controlling the rotational speed of the primary motor. A controller operably coupled to the motor can be configured to maintain a substantially constant spacing between the scanning arm and the patient table. The motor may also include a linear actuator to provide lateral movement of the arm and / or rotation of the scanning arm about its longitudinal axis.

[0015] The secondary motor of the scanning gantry includes a rotating platform mounted on motor drive bearings, each of which is movable along the longitudinal axis of the motor. The scanning platform is fixed to the rotating platform by a rotating shaft and is driven by the motor. A motion control subsystem controls the position and speed of the scanning platform together with the motor, which results in rotational movement of the gantry about its longitudinal axis. A control signal generation circuit is provided for generating a signal for controlling the rotational rate of the motor to vary the moving speed of the scanning platform together with the motor. The control signal generation circuit is connected to a control input of the motion control subsystem such that the control signal generation circuit can generate a control signal for causing the motor to produce a controlled movement of the scanning platform along the motor. The control signal generation circuit can also generate a control signal for causing rotation of the motor. The control signal generation circuit also receives X-ray event data from an X-ray event detector. The control signal generation circuit provides a drive signal to a drive element for controlling the energization of the X-ray tube and the modulator in response to the received X-ray event data and for controlling the speed of the motor and thus the moving rate of the scanning platform together with the motor.

[0016] A support frame structure is supported in the object receiving region of the tomosynthesis imaging system scanner, as is a plurality of pulsed X-ray sources that generate X-rays for transmission through the object receiving region. A fast gantry sweeps around the object receiving region such that the X-rays generated by the plurality of pulsed X-ray sources can pass through the object receiving region and impinge on a scintillator at different angles. As described above, the plurality of pulsed X-ray sources provide certain advantages compared to a single pulsed X-ray source.

[0017] The plurality of X-ray sources can be attached to the gantry via any suitable structure. The X-ray sources are configured to emit X-rays towards the patient in multiple directions simultaneously during movement to generate an image of the patient. Although depicted as linear sources in the drawings, any number of non-linear sources, such as elliptical or arcuate sources, can be used. The X-ray sources are also configured to move independently in multiple directions to achieve proper coverage for imaging. Additionally, the individual coverage of each source can be enhanced by applying some predetermined radiation filtering (e.g., using collimators and / or filters).

[0018] The X-ray flat panel detector 3 can be set to acquire a first set of images with a predetermined first exposure value. Suppose the first set of images is insufficient for diagnosis. In this case, the wide-angle multi-motion pulse source tomosynthesis imaging system 1 can gradually acquire at least one additional set of images with a second exposure value lower than the first exposure value until sufficient information is obtained to diagnose the object. The exposure value can be determined based on the energy spectrum or duration.

[0019] Figure 3 It shows that for the multi-motion pulse X-ray source tomosynthesis imaging system 1, the line-by-line scanning positions of each X-ray source 2 span a wide angle at the starting position. It has covered a wide angular span since the first scan set. For example, for a system with five sources A, B, C, D, E and a total of 25 exposures, each X-ray source makes five exposures or scans. The X-ray data sets of the activation sequence are: the first set 1-6-11-16-21; the second set 2-7-12-17-22; the third set 3-8-13-18-23; the fourth set 4-9-14-19-24; the fifth set 5-10-15-20-25, etc. For five sources, this type of machine can easily perform more than 120 or more scans in total. If there are a total of 120 scans, each X-ray source will move in small incremental steps 24 times and make 24 scans. Depending on the design of the system, the initial angular coverage can easily exceed 60 degrees. Five sources, six sources or even more X-ray sources can be utilized. In one embodiment, if five sources are used, there is a 12.5-degree angle between adjacent sources, and the total covered degrees are 62.5 degrees. Other configurations are also possible.

[0020] For the multi-motion pulse X-ray source tomosynthesis imaging system 1, the X-ray source 2 moves relative to the patient and scans the patient. The angular coverage of each partial scan is very wide and wide enough to provide diagnostic information without causing too much attenuation in the patient. Based on the initial scan, the AI will compare the results with a database of previous images to determine whether there is sufficient information to perform a diagnosis based on the initial scan. If there is sufficient information from a small partial scan, the data acquisition stops; if more information is needed, the system progressively performs another round of wide-angle sparse scans at a new position until the results are satisfactory. During real-time image reconstruction, artificial intelligence (AI) determines sufficient information to perform a diagnosis based on the initial scan. The AI will compare the diagnostic results with the patient's medical history to make an informed decision. If there is sufficient information from a small partial scan, the data acquisition stops; if more information is needed, the system progressively performs another round of wide-angle sparse scans at a new position until the results are satisfactory.

[0021] Figure 1Shows a flowchart of line-by-line scanning at a multi-motion pulse X-ray source tomosynthesis imaging system. The method includes: selecting an imaging task and loading a predetermined protocol into the imaging system; turning on the multi-motion pulse X-ray source tomosynthesis imaging system 1; placing an object at a predetermined position; obtaining a first data set from different angles and performing image reconstruction by acquiring one or more projection images from one or more selected sources of the imaging system; accumulating the one or more projection images and reconstructing a 3D tomosynthesis volume therefrom; applying machine learning to perform a diagnosis, repeating the scan one or more times to achieve a predetermined image reconstruction quality, and acquiring additional projection images, and applying machine learning to the reconstructed 3D tomosynthesis volume until a predetermined image reconstruction quality threshold is met; and applying machine learning to search for lung nodules in the 3D tomosynthesis volume.

[0022] Reconstruction can be performed even with one data set from different angles. After obtaining the first X-ray imaging scan set, the AI will determine how many more data sets are needed. It is entirely possible that the first few data sets only take less than a second to complete the diagnosis. The AI will compare the diagnosis results with the patient's medical history in order to make an informed decision. Most of the time spent is that the patient needs time to position themselves to be ready to obtain data, or the operator needs time to input patient information. The data acquisition itself can be carried out very quickly.

[0023] In one embodiment, the system performs the following operations. The system first collects patient information and selects a body anatomy, for example, the chest. The patient then enters the scan area. For example, in one embodiment where the patient leans on a table, the system can position the patient at 6 feet above the ground at the center of the table, where a large number of pulsed sources are above and below the patient at its corners to cover the chest area. There are moving worktables; one moving worktable is in front of the patient, and one moving worktable is behind the patient. Each has a plurality of pulsed sources. Use full-energy multiple pulsed sources. Between each round, the patient is rotated by a certain angle in his / her horizontal plane. One pulsed source rotates by a certain angle and moves both in-plane and out-of-plane simultaneously. Another pulsed source starts from the opposite direction at each time step and rotates at different angles in the same or opposite direction. Use at least parallel partial scans for each round. Sparsely distributed positions and much wider angles. The starting positions of multiple sources have spanned a wide angle. Therefore, compared with the angles of CT, the first round of scanning has covered a much wider angle. The multiple pulsed sources determine where to go next and at what angle to rotate.

[0024] In another embodiment for analyzing lung diseases, the operator selects a task (e.g., the lung module) via a button, and the 3D model of the lung moves on the screen accordingly. Then, the operator can place the pointer on top of the lung area on the 3D model and make a click to select the lung area for reconstruction. The same process can be performed for the heart, breast, etc. The checkbox "Match previous image" is used to compare with the patient's medical history to determine whether sufficient information has been obtained. Several challenges include: sparse distributed partial scans with wide angles, parallel data acquisition from multiple moving pulse sources, real-time image reconstruction, multi-layer contrast, radiography image resolution, high resolution with limited angles, diagnostic capabilities, etc. For such an application of distributed tomography, a specific implementation uses distributed wide-angle partial scans to further reduce the X-ray dose applied to the patient and perform faster X-ray scans at the multi-moving pulse X-ray source tomography imaging system 1. The system will have sparse distributed partial scans with wide angles. More particularly, several embodiments are disclosed in the present invention.

[0025] The system accumulates all projection images in parallel during each partial scan and reconstructs the 3D tomography image volume. With each projection image, the corresponding reconstructed 3D volume data from the previous scan is also accumulated and compared with the current 3D volume data of this partial scan to see if there is sufficient information for diagnosis. In this case, the system stops; otherwise, the system gradually performs another round of sparse partial scans at a new position until the result is satisfactory. The system works with multiple X-ray sources and runs in parallel. The number of sources depends on how wide an angle is needed. The system positions are sparsely distributed with wider angle starting points. The starting points have spanned a wide angle. Thus, compared with the angle of CT, the first round of scans has covered a much wider angle. It is covered in only one step in the case of continuous rotation. It is a device with a multi-moving pulse source tomography imaging system. There are some existing technologies for performing line-by-line partial scans and line-by-line image reconstruction.

[0026] The system has many advantages. One advantage over CT scans alone is that the line-by-line scan provides wide-angle coverage. Another advantage is that this system is faster than a CT scanner with one X-ray source, which requires large rotations / angular movements to obtain a larger coverage. Conventional CT sampling of the imaging volume starts from small angles and then incrementally increases, so it is slower. In contrast, a system with multiple scan sources can perform scans quickly.

[0027] An AI (Artificial Intelligence or Machine Learning) system analyzes images, determines a patient's health condition, and outputs a diagnostic report and the images. The AI uses multiple motion-pulsed source tomosynthesis imaging systems to analyze previous data, where each pulsed source emits X-rays at different angles. The hardware system starts with one pulsed source for a wide-angle distributed sparse partial scan, where during imaging the pulsed source points in multiple directions to cover a wide region of interest for imaging purposes. For the initial scan, the AI analyzes the images from the pulsed source positioned at a small angle. There is sufficient information from the small partial scan to perform a diagnosis. If the initial scan is insufficient, additional rounds of wide-angle sparse partial scans are performed step by step until the result is satisfactory. Since the AI decides the next scan direction based on previous data, the result enables high-resolution tomosynthesis scans to be performed even with faster scan times.

[0028] The AI analysis of the scan quality reduces radiation exposure: if there is sufficient information from the small partial scan, data acquisition stops; if more information is needed, the system performs another round of wide-angle sparse scans step by step at new positions until the result is satisfactory. The AI also processes the overall set of partial scans to generate a final report for the user and to determine the amount of inter-local information compared to the information between angles in the prior art for the radiologist to make a decision. In one embodiment, the system stores many images. In another embodiment, only different images are stored, thus compressing the image folder size. Comparing images from different positions for different angles can provide additional diagnostic information that is not obvious from single-slice images from prior art limited-angle tomosynthesis. In another embodiment, the device includes an overall scan transposition system for identifying images corresponding to the same position but different angles. This allows for a quick search to find images corresponding to a specific imaging region or region of interest. In one embodiment, after determining whether there is sufficient information to continue with the diagnostic process, the controller continues to use image fusion techniques to reconstruct a series of partial scans. In another embodiment, the system may include a patient data storage component for storing patient data (e.g., X-ray imaging data) from many different angles.

[0029] One embodiment warns the operator to make a decision to perform another round of scanning or stop data acquisition. Optionally, the AI system may have a library of patient images and their associated parameters, which can analyze the current partial scan for a previous partial scan for diagnosis and determine whether more information from additional partial scans is needed. If there is sufficient information from a small number of partial scans, data acquisition stops; if more information is needed, the system progressively performs another round of wide-angle sparse scanning at a new location until the result is satisfactory. The present invention makes it possible to save time and reduce the X-ray dose, which is made possible by the novel hardware device of multiple pulsed sources in motion. The starting positions of the multiple sources have spanned a wide angle. Thus, the first round of scanning has covered a much wider angle compared to the angle of CT. Image reconstruction during real-time image reconstruction can utilize such parallelism between X-ray sources. And, most importantly, it utilizes multiple iterations of "sparse" distributed imaging with different angles and parallelism to form the whole.

[0030] The diagnostic results include the probability of finding something of interest (such as cancer cells). It also includes the probability of finding something not of interest (such as healthy tissue). For each of the multiple acquired partial scans and the associated raw data, one or more of these probabilities can be considered. It is desirable to quickly acquire multiple slices without compromising image quality. Thus, the system applies multiple X-ray sources and many partial scan points to quickly acquire all slices with a minimum radiation dose. Each source starts with a wide angle and covers a small field of view (FOV), and progressively increases its FOV and progressively decreases its angle during successive acquisitions until the end of the volume is reached. After acquisition, the system does not need to repeat the beginning of the volume and start from there as in the case of single-source tomosynthesis. Alternative methods can be used, where the system first uses AI technology to reconstruct the initial partial scan (the first scan covering a large angular range). In this embodiment, if the AI finds anything suspicious, it will warn the operator and may instruct him / her to acquire more scans in a specific area before he / she stops. In this way, the system only stops when all diagnostic criteria are met, rather than stopping after a certain number of acquisitions as in the conventional case.

[0031] The multi-motion pulse source tomosynthesis imaging system 1 of the present system includes: at least two X-ray sources with a rotatable gantry, a plurality of rotating devices with input-output control mechanisms, and an imaging detector with an array detector on the input-output control mechanism. The input-output control mechanism further includes a plurality of moving devices. The gantry has at least two rotatable gantries that support the corresponding X-ray sources. The rotatable gantry can be made of any material that can maintain its structural integrity during operation. A clean environment is free of contaminants that may cause stickiness. A clean environment includes a production area where components are manufactured, a medical facility, a sterile environment, etc. A dirty environment includes an automotive production area, a power plant, a chemical processing plant, etc. The plurality of moving devices cooperate with each other to move the corresponding rotatable gantries.

[0032] Distributed wide-angle sparse partial scanning includes: obtaining a first wide-angle partial scan data set for an object target area through an X-ray source; and using an X-ray detector and using the first wide-angle partial scan data set to obtain a second wide-angle partial scan data set from the target area. The total radiation dose used to obtain the second wide-angle partial scan data set is less than a multiple of the radiation dose used to obtain the first wide-angle partial scan data set. In one embodiment, obtaining the first wide-angle partial scan data set for the object target area includes: obtaining a plurality of partially overlapping wide-angle partial scan data in a sparse distributed pattern from the target area. In another embodiment, the target area includes a patient's heart. In yet another embodiment, obtaining the second wide-angle partial scan data set from the target area includes: obtaining a third wide-angle partial scan data set from the target area. In another embodiment, the third wide-angle partial scan data set has a smaller lateral dimension than the lateral dimension of the second wide-angle partial scan data set. In another embodiment, obtaining the second wide-angle partial scan data set from the target area includes: obtaining a plurality of substantially non-overlapping wide-angle partial scan data in a sparse distributed pattern from the target area.

[0033] The patient will then be translated to another angle to acquire a second data set. As previously mentioned, the multi-motion pulse X-ray source tomosynthesis imaging system 1 has multiple X-ray sources. The spacing between each X-ray source is wider, which means there are three options for the position where the tomosynthesis scan begins. These positions are referred to as P1, P2, and P3. The same logic as in the previous example will apply here. P1 corresponds to the first position. This corresponds to the start of scan number one. P2 corresponds to the second position. This corresponds to the start of scan number two. P3 corresponds to the third position. This corresponds to the start of scan number three. After acquiring the second data set, the system compares the results with the diagnostic requirements for each slice position on each source. The system uses AI to decide whether data acquisition can be stopped without causing too much loss of diagnostic information. If so, the system stops data acquisition. If not, the system acquires another round of data. By doing so, the present invention significantly reduces the X-ray dose applied to the patient and performs faster X-ray scans. Since there are multiple X-ray sources, if the first two or three partial scans are not sufficient, the system can use another position to gradually acquire another partial scan until the results are satisfactory.

[0034] The system applies AI to perform real-time reconstruction and make the next move. AI compares the diagnostic results with the patient's medical history to make an informed decision, bringing many advantages. The first advantage is its novel hardware device with multiple motion pulses sources. It runs in parallel and is much faster. The multi-motion pulse X-ray source tomosynthesis imaging system has multiple X-ray sources. The second advantage is its sparsely distributed positions and wider angles. The starting positions of multiple sources have spanned a wide angle. Therefore, compared with the angle of CT, the first round of scans has covered a much wider angle. The third advantage involves AI. AI is becoming increasingly popular now. Due to the rapid data acquisition and obtaining the first data set covering a wider angle, AI can be used for real-time reconstruction to make the next move. AI will compare the diagnostic results with the patient's medical history to make an informed decision. If there is sufficient information from a partial scan, data acquisition stops; if more information is needed, the system progressively performs another round of wide-angle sparse scans at a new position until the results are satisfactory.

[0035] In other embodiments, the system may apply AI to segment organs and their vascular systems for early cancer detection. The system may utilize voxel data, one-dimensional intensity distributions, multi-dimensional intensity distributions, and multi-dimensional intensity gradients to operate. The system may automatically generate pathological mapping diagrams, intensity value mapping diagrams, or angle value mapping diagrams. The system can not only accept various inputs but also make intelligent decisions based on the patient's clinical information to automatically select the next partial scan direction. The system thus applies AI for image reconstruction of wide-angle distributed partial scans and uses the instant design of a multi-motion pulse source tomosynthesis imaging system to complete the partial scans. The system may perform distributed sparse partial scan acquisition to utilize the wide angle and the AI involved in real-time reconstruction to determine the next step.

[0036] Without departing from the spirit and scope of the invention as defined by the appended claims, various modifications and changes to the present invention will be apparent to those skilled in the art. It should be noted that the steps recited in any of the following method claims do not necessarily need to be performed in the order in which they are recited. Those of ordinary skill in the art will recognize variations in the performance of the steps based on the order in which the steps are recited. Additionally, the lack of mention or discussion of a feature, step, or component provides a basis for a claim that excludes the non-existent feature or component by means of a proviso or similar claim language.

[0037] Although the present invention has been described above in accordance with various exemplary embodiments and implementations, it should be understood that the various features, aspects, and functions described in one or more of the individual embodiments are not limited to their applicability to the particular embodiments in which they are described, but rather can be applied singly or in various combinations to one or more of the other embodiments of the present invention, whether or not such embodiments are described and whether or not such features are presented as part of the described embodiments. Therefore, the breadth and scope of the present valve should not be limited by any of the above exemplary embodiments.

[0038] Unless otherwise expressly stated, the terms and phrases used in this document and their variants shall be interpreted as open-ended rather than restrictive. As examples of the foregoing: the term "comprising" shall be construed to mean "including but not limited to" or words to that effect; the term "example" is used to provide exemplary instances of items in a discourse, rather than an exhaustive or restrictive list thereof; the term "a" or "an" shall be construed to mean "at least one", "one or more" or words to that effect; and adjectives such as "conventional", "traditional", "normal", "standard", "known" and terms of similar import shall not be construed to limit the items described to a given time period or to items available as of a given time, but rather shall be construed to cover conventional, traditional, normal or standard techniques that are available or known now or at any time in the future. Thus, where this document refers to techniques that are obvious or known to one of ordinary skill in the art, such techniques cover those that are obvious or known to a person skilled in the art now or at any time in the future.

[0039] In some cases, the presence of broad words and phrases such as "one or more", "at least", "but not limited to" or other phrases of that kind should not be taken to mean that the narrower cases are intended or required where such broad phrases might not be present. The use of the term "module" does not imply that all of the components or functionality described as part of the module are configured in a common enclosure. In fact, any or all of the various components of a module, whether control logic or otherwise, may be combined in a single enclosure or maintained separately and may further be distributed across multiple locations.

[0040] The foregoing description of the disclosed embodiments enables any person skilled in the art to make 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 present invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method of performing line-by-line scanning using a multi-motion pulse X-ray source tomosynthesis imaging system, the method comprises: Placing an object at a predetermined position; Controlling a multi-motion pulse X-ray source tomosynthesis imaging system, the tomosynthesis imaging system comprising a primary motor stage capable of moving on an arc-shaped guide rail, the primary motor stage having a plurality of X-ray sources, each X-ray source being mounted on the primary motor stage and aligned with an X-ray flat panel detector; Using the tomosynthesis imaging system, moving the plurality of X-ray sources relative to the object simultaneously on the arc-shaped guide rail, thereby obtaining a first data set from different X-ray sources at different angles and performing image reconstruction; Applying artificial intelligence with machine learning to perform diagnosis, and repeating the scan one or more times to achieve a predetermined image reconstruction quality; and Constructing a 3D tomosynthesis volume therefrom.

2. The method according to claim 1, which comprises: Selecting an imaging task and loading a predetermined protocol into the imaging system.

3. The method according to claim 1, which comprises: Obtaining one or more projection images from one or more selected sources.

4. The method according to claim 1, which comprises: Accumulating one or more projection images and reconstructing the 3D tomosynthesis volume therefrom.

5. The method according to claim 1, which comprises: Applying machine learning to the reconstructed 3D tomosynthesis volume to determine the image reconstruction quality.

6. The method according to claim 1, which comprises: Applying machine learning to search for lung nodules in the 3D tomosynthesis volume.

7. The method according to claim 1, which comprises: Applying artificial intelligence with machine learning to search for lung nodules in the 3D tomosynthesis volume; and Identifying nodule characteristics, including the size, shape, number and location of the nodules, and generating a report.

8. The method according to claim 5, which comprises: Obtaining additional projection images and applying machine learning to the reconstructed 3D tomosynthesis volume until a predetermined image reconstruction quality threshold is met.

9. The method according to claim 6, which comprises: Identifying nodule characteristics, including the size, shape, number and location of the nodules, and generating a report.

10. A method of performing line-by-line scanning using a multi-motion pulse X-ray source tomosynthesis imaging system, the method comprises: Selecting an imaging task and loading a predetermined protocol into the imaging system; Turning on a plurality of pulse sources using a multi-motion pulse X-ray source tomosynthesis imaging system, the tomosynthesis imaging system comprising a primary motor stage with a plurality of X-ray sources capable of moving on an arc-shaped guide rail, the X-ray sources being aligned with an X-ray flat panel detector; Placing an object at a predetermined position; Moving the plurality of X-ray sources relative to the object simultaneously on the arc-shaped guide rail, thereby obtaining a first data set from different X-ray sources at different angles, and performing image reconstruction by obtaining one or more sets of projection images from one or more selected sources of the imaging system; Accumulating the one or more sets of projection images and reconstructing a 3D tomosynthesis volume therefrom; Apply artificial intelligence with machine learning to perform a diagnosis, repeat the scan one or more times to achieve a predetermined image reconstruction quality, and obtain an additional set of projection images, and apply the artificial intelligence with machine learning to the reconstructed 3D tomosynthesis volume until a predetermined image reconstruction quality threshold is met; and Apply artificial intelligence with machine learning to search for lung nodules in the 3D tomosynthesis volume.

11. An X-ray imaging system, which comprises: A plurality of moving pulse X-ray source tomosynthesis imaging systems, the tomosynthesis imaging system including a primary motor workbench that can move freely on an arc-shaped guide rail with a predetermined shape, a primary motor that engages with the primary motor workbench and controls the speed of the primary motor workbench, a plurality of X-ray sources mounted on the primary motor workbench, a support frame structure that provides a housing for the primary motor workbench, and an X-ray flat panel detector for receiving X-rays and transmitting X-ray imaging data, for receiving an object at a predetermined position; A processor, the processor being coupled to the imaging system to run computer code to: Control the plurality of moving pulse X-ray source tomosynthesis imaging systems; Cause the plurality of X-ray sources to move simultaneously relative to the object on the arc-shaped guide rail, so as to obtain a first data set from different X-ray sources at different angles and perform image reconstruction; Apply artificial intelligence with machine learning to perform a diagnosis, and repeat the scan one or more times to achieve a predetermined image reconstruction quality; and Construct a 3D tomosynthesis volume therefrom.

12. The system according to claim 11, wherein the processor selects an imaging task and loads a predetermined protocol into the imaging system.

13. The system according to claim 11, wherein the processor controls the imaging system to obtain one or more projection images from one or more selected sources.

14. The system according to claim 11, wherein the processor accumulates the one or more projection images to reconstruct the 3D tomosynthesis volume.

15. The system according to claim 11, wherein the processor runs machine learning code on the reconstructed 3D tomosynthesis volume to determine the image reconstruction quality.

16. The system according to claim 11, wherein the processor applies machine learning to search for lung nodules in the 3D tomosynthesis volume.

17. The system according to claim 11, which includes computer-readable code for: Selecting an imaging task and loading a predetermined protocol into the imaging system; Turning on a plurality of pulsed sources using a motion tomosynthesis imaging system; Placing an object at a predetermined position; Obtaining a first data set from different X-ray sources at different angles, and performing image reconstruction by obtaining one or more projection images from one or more selected tubes of the imaging system; Accumulating the one or more projection images and reconstructing the 3D tomosynthesis volume therefrom; Applying artificial intelligence with machine learning to perform a diagnosis, repeating the scan one or more times to achieve a predetermined image reconstruction quality, and obtaining additional projection images; and Apply machine learning to the reconstructed 3D tomosynthesis volume until a predetermined image reconstruction quality threshold is met.

18. The system according to claim 11, comprising code for: obtaining additional projection images and applying machine learning to the reconstructed 3D tomosynthesis volume until a predetermined image reconstruction quality threshold is met.

19. The system according to claim 11, wherein the processor identifies nodule characteristics, including the size, shape, quantity, and location of the nodule.

20. The system according to claim 11, comprising computer-readable code for: applying artificial intelligence with machine learning to search for lung nodules in the 3D tomosynthesis volume; and identifying nodule characteristics, including the size, shape, quantity, and location of the nodule, and generating a report.

Images