Scatter correction for computed tomography imaging.

Abstract

To provide a system and method for a scatter correction of an X-ray image.SOLUTION: A scatter image of an object can be corrected using partial-scatter free images acquired using an aperture plate. The plate is positioned between the object and a radiation detector, and includes apertures in a grid. The original x-ray passes through the aperture, and a scattered x-ray can be blocked by the aperture plate. The aperture plate can be moved to a different position, thereby allowing a partial scatter-free image to be acquired at each position of the aperture plate. A full scatter-free image can be generated by combining the partial scatter-free image. The scatter and scatter-free images can be further used to enable training a scatter correction model.SELECTED DRAWING: Figure 7

Description

[Technical field]

[0001] (CROSS REFERENCE TO RELATED APPLICATIONS) This application claims the benefit of U.S. Provisional Patent Application No. 63 / 090,144, entitled “Scatter Correction For Computed Tomography Imaging,” filed October 9, 2020, which is incorporated by reference in its entirety. [Background technology]

[0002] Inspection of objects is commonly performed in the manufacturing and repair industries. Industrial inspection processes can use various types of inspection systems, such as computed tomography (CT), coordinate measuring machines (CMM), laser-based profilometry, optical gauges, infrared, etc. For example, these inspection systems can be used to measure dimensions or identify defects in manufactured parts (e.g., turbine blades).

[0003] Each of these inspection systems has its advantages and disadvantages. Modalities such as CMM and laser-based profilometry can be used to measure external surfaces with high accuracy, but internal features cannot be measured unless the object is cut open. To date, CT is the most versatile of the measurement / inspection systems for non-destructively revealing both the internal and external structure of industrial parts. Due to its ability to provide internal as well as external measurements, CT-based techniques can facilitate processes such as reverse engineering, rapid prototyping, casting simulation and validation, tire development, first article inspection, ceramic porosity inspection, process validation, part qualification, and defect detection, among others. Summary of the Invention

[0004] However, CT-based techniques may also have certain limitations that may inhibit their widespread use. For example, volumetric computerized tomography (VCT) imaging for industrial applications (e.g., imaging of metal parts) may provide unsatisfactory images with image artifacts due to radiation-material interaction-based artifacts, scanner-based artifacts, reconstruction technique-based artifacts, etc. Radiation-material interaction-based artifacts may further include beam hardening artifacts and artifacts due to X-ray scattered radiation. Scattered radiation is a strong function of imaging parameters such as the object under imaging, the beam spectrum used, the geometric distance, and the surrounding medium. In general, scattered radiation in a projection image may be undesirable because it reduces the contrast of the projection image, causing degradation or blurring of sharp features of the object in the generated volume image, which may reduce the accuracy of metrology applications and the detectability of smaller features.

[0005] Therefore, various techniques for estimating scatter have been developed to reduce or eliminate scatter from CT images. In general, accurate estimation of scatter signal content in projection imaging can be difficult due to various dependencies on imaging parameters. Physics-based models are often used to predict scatter content in X-ray images, but they are time-consuming and only predict scatter arising from the object under scan, provided that the material properties are known.

[0006] Different techniques exist for scatter measurement and scatter correction in acquired projection images. For example, one popular scatter measurement technique uses a beam stopper located between the radiation source and the object being scanned in a VCT system to measure the scatter at the corresponding location. However, most currently known techniques mainly deal with object scatter and involve time-consuming computer simulations.

[0007] As manufacturing tolerances become tighter, there is a corresponding increase in demands on metrology techniques to maintain the tolerances. The need for quality and performance testing has become an integral part of the production or manufacturing process. Therefore, more effective methods are needed to remove scatter radiation related artifacts to improve CT inspection accuracy and efficiency.

[0008] The embodiments of the present disclosure provide an improved system and method for scatter correction. The scatter image of an object can be corrected using additional images acquired with a scatter rejection aperture plate. The scatter rejection aperture plate can include multiple apertures positioned on a grid. In use, the aperture plate is positioned between the object being imaged and the detector. When collimated x-rays are directed from an x-ray source to the object, the original x-rays can pass through the apertures while a significant portion of the scattered x-rays can be blocked by the aperture plate. The aperture plate can also be configured to move between different positions. Thus, a partial scatter-free image can be acquired at each position of the aperture plate, which is referred to herein as a partial scatter-free image. By combining the partial scatter-free images, a scatter-free image (ground truth) can be obtained. The number of positions and the acquired partial scatter-free images can vary depending on the geometry of the aperture plate. In general, the number of positions can be sufficient such that the apertures cover the entire area of ​​the scatter image to be scatter-corrected.

[0009] The embodiments discussed above rely on the use of an aperture plate with individual holes, also referred to as beam holes, to acquire the partial scatter-free images. In alternative embodiments, a slit collimator or beam stop may be used instead of an aperture plate to acquire the partial scatter-free images. Similar to the aperture plate approach, the slit collimator or beam stop may be positioned and moved between the radiation source and the detector. Each partial scatter-free image may be acquired at a different position of the slit collimator or beam stop, and the partial scatter-free images may be combined to form the scatter-free image.

[0010] The scatter-free image acquired through the use of either an aperture plate, a slit collimator, or a beam stop can further be used to facilitate scatter correction. In one aspect, the scatter-free image can be used as a reference to train a neural network to determine an algorithm for scatter correction. In another aspect, the scatter-free image can be used to determine the point spread function (PSF) of a scatter deconvolution algorithm.

[0011] In further embodiments, the partially unscattered or unscattered images can be used to train an already trained neural network or to tune the parameters of a known PSF for convolution-based scatter correction.

[0012] In one embodiment, a method for scatter correction of an image of an object is provided. The method may include obtaining data representative of at least one scatter image of the object based on detection of radiation transmitted through an imaging volume of the object by a radiation detector of an imaging system. The method may further include disposing an aperture plate at a first position between the object and the radiation detector. The aperture plate may include a plurality of apertures configured to suppress scattered radiation from detection by the radiation detector. The method may also include obtaining data representative of at least one first partially scatter-free image based on detection of radiation transmitted through the imaging volume of the object when the aperture plate is in the first position by the radiation detector. The method may further include moving the aperture plate to one or more second positions different from the first position. The apertures at the first position and the one or more second positions may cover an area of ​​the object captured in the at least one scatter image. The method may also include obtaining data representative of at least one second partially scatter-free image based on detection of radiation transmitted through the imaging volume of the object by the radiation detector when the aperture plate is in the one or more second positions. The method may further include receiving, by an analyzer including one or more processors, at least one first and second partial free-scatter image data. The method may further include generating, by the analyzer, at least one free-scatter image based on a combination of at least a portion of the at least one first partial free-scatter image data and at least a portion of the at least one second partial free-scatter image data. The method may also include updating, by the analyzer, a scatter correction model using at least a portion of the at least one scatter image and the at least one free-scatter image. The method may further include outputting, by the analyzer, the updated scatter correction model.

[0013] In another embodiment, the scatter correction model may be a neural network model including a scatter correction algorithm. The method may further include operations performed by the analyzer including updating the scatter correction model by operations including applying a scatter correction algorithm to at least a portion of the at least one scatter image to generate a scatter corrected image and determining a deviation between the generated scatter corrected image and the at least one scatter free image. The operations may further include updating the scatter correction algorithm to reduce the deviation when the deviation is greater than a predetermined deviation. The operations may further include outputting an updated scatter correction model including an updated scatter correction algorithm when the deviation is less than or equal to the predetermined deviation.

[0014] In another embodiment, the scatter correction model may be a deconvolution including a scatter edge spread function (ESF). The scatter edge spread function may further include a point spread function (PSF) configured to correct for scatter in the scatter image. The method may further include updating the scatter correction model. The scatter correction model may be updated by operations including applying the PSF to at least a portion of the at least one scatter image to generate a scatter corrected image and determining a deviation between the scatter corrected image and the at least one scatter free image when the deviation exceeds a predetermined deviation, updating parameters of the PSF to reduce the deviation and repeating the applying and determining operations. The operations may further include outputting an updated scatter correction model including updated PSF parameters when the deviation is less than the predetermined deviation.

[0015] In another embodiment, the detected radiation may be collimated.

[0016] In another embodiment, the at least one scatter-free image is not generated by interpolation of the at least one first partial scatter-free image data or the at least one second partial scatter-free image data.

[0017] In another embodiment, moving the aperture plate may include at least one of unidirectional translation, bidirectional translation, or rotation.

[0018] In another embodiment, the at least one scattered image may be a scattered image set comprising a plurality of images.

[0019] In another embodiment, the openings may be arranged in a two-dimensional grid.

[0020] In another embodiment, the openings in the first location and the at least one second location may overlap one another.

[0021] In another embodiment, the openings in the first location and the at least one second location may be spaced a predetermined distance from each other.

[0022] In one embodiment, an imaging system is provided. The imaging system may include a radiation source, a movable aperture plate, and an analyzer. The radiation source may be configured to emit radiation directed toward the object. The radiation detector may be configured to detect the emitted radiation transmitted through an imaging volume of the object. The movable aperture plate may be positioned between the object and the radiation detector and may further include a plurality of apertures configured to suppress scattered radiation from detection by the radiation detector. The analyzer may include one or more processors. The analyzer may be further configured to receive data representing at least one scattered image of the object based on the detection of the radiation transmitted through the imaging volume of the object. The analyzer may also be configured to receive data representing at least one first partially unscattered image of the object based on the detection of the radiation transmitted through the imaging volume of the object when the aperture plate is in a first position. The analyzer may be further configured to receive data representing at least one second partially unscattered image based on the detection of the radiation transmitted through the imaging volume of the object when the aperture plate is in one or more second positions different from the first position. The apertures at the first location and the one or more second locations may cover an area of ​​the object captured in the at least one scatter image. The analyzer may also be configured to generate at least one scatter-free image based on a combination of at least a portion of the at least one first partial scatter-free image data and at least a portion of the at least one second partial scatter-free image data. The analyzer may be further configured to train a scatter correction model using at least a portion of the at least one scatter image and the at least one scatter-free image. The analyzer may be further configured to output the trained scatter correction model.

[0023] In another embodiment, the scatter correction model may be a neural network model including a scatter correction algorithm. The analyzer may be further configured to train the scatter correction algorithm by applying the scatter correction algorithm to at least a portion of the at least one scatter image to generate a scatter corrected image and determining a deviation between the generated scatter corrected image and the at least one scatter free image. When the deviation is greater than a predetermined deviation, the analyzer may further update the scatter correction algorithm to reduce the deviation. When the deviation is less than or equal to the predetermined deviation, the analyzer may further output the updated scatter correction algorithm.

[0024] In another embodiment, the scatter correction model may be a deconvolution including a scatter edge spread function (ESF). The scatter edge spread function may further include a point spread function (PSF) configured to correct for scatter in the scatter image. The analyzer may be further configured to train the PSF by applying the PSF to at least a portion of at least one scatter image to generate a scatter corrected image and determining a deviation between the scatter corrected image and at least one scatter free image. When the deviation exceeds a predetermined deviation, the analyzer may further update parameters of the PSF to reduce the deviation and repeat the applying and determining operations. When the deviation is less than the predetermined deviation, the analyzer may output the PSF parameters.

[0025] In another embodiment, the system may further include a collimator configured to collimate the emitted radiation.

[0026] In another embodiment, the analyzer is not configured to generate the at least one non-scattering image by interpolation of the at least one first partial non-scattering image data or the at least one second partial non-scattering image data.

[0027] In another embodiment, the first position and the at least one second position may differ by at least one of a unidirectional translation, a bidirectional translation, or a rotation.

[0028] In another embodiment, the at least one scattered image may be a scattered image set comprising a plurality of images.

[0029] In another embodiment, the openings may be arranged in a two-dimensional grid.

[0030] In another embodiment, the apertures in the first location of the aperture plate and at least one of the second locations of the aperture plate may overlap one another.

[0031] In another embodiment, the apertures in the first location of the aperture plate and the at least one second location of the aperture plate may be spaced a predetermined distance apart from one another.

[0032] In one embodiment, a method for scatter correction of an image of an object is provided. The method can include obtaining data representing a plurality of scatter images of the object based on detection of radiation transmitted through an imaging volume of the object by a radiation detector of an imaging system. The method can also include disposing an aperture plate between the object and the radiation detector. The aperture plate can include a plurality of apertures configured to suppress scattered radiation from detection by the radiation detector. The method can further include obtaining data representing a partial scatter-free image corresponding to each scatter image by the radiation detector. Each partial scatter-free image can be based on detection of radiation transmitted through an imaging volume of the object when the aperture plate is present. The scatter image and its corresponding partial scatter-free image can be obtained under substantially the same conditions except for the presence of the aperture plate. The method can further include receiving, by an analyzer including one or more processors, the plurality of scatter image data and the corresponding partial scatter-free image data. The method can further include receiving, by the analyzer, a trained scatter correction model. The method includes updating, by the analyzer, a trained scatter correction model based on the received plurality of scatter image data and corresponding partially scatter-free image data to obtain an updated trained scatter correction model, correcting at least one of the plurality of scatter images based on the updated trained scatter correction model, and outputting at least one corrected scatter image.

[0033] In one embodiment, the trained scatter correction model may be a trained neural network model.

[0034] In another embodiment, updating the trained scatter correction model may include, for each pair of corresponding scatter image and partial no-scatter image, performing an interpolation between an output of the trained neural network model for the scatter image and an output of the trained neural network model for the partial no-scatter image, and generating an updated trained scatter correction model based on the interpolation.

[0035] In another embodiment, the trained scatter correction model may include a previously determined deconvolved point spread function (PSF) estimate.

[0036] In another embodiment, updating the trained scatter correction model may include locally parameterizing the deconvolution PSF estimate using measurement points of each pair of corresponding scatter and partial scatter-free images, and updating the deconvolution PSF estimate based on the local parameterization.

[0037] In another embodiment, an imaging system is provided. The imaging system may include a radiation source, a radiation detector, and an analyzer. The radiation source may be configured to emit radiation directed towards the object. The radiation detector may be configured to detect the emitted radiation transmitted through an imaging volume of the object. The analyzer may include one or more processors. The analyzer may be further configured to receive data representing a plurality of scatter images of the object based on detection of radiation by the radiation source transmitted through the imaging volume of the object. The analyzer may also be configured to receive data representing a partial unscattered image corresponding to each scatter image. Each partial unscattered image may be based on detection of radiation transmitted through the imaging volume of the object when the aperture plate is present. The scatter image and its corresponding partial unscattered image may be acquired under substantially the same conditions, except for the presence of the aperture plate. The analyzer may be further configured to receive a trained scatter correction model and update the trained scatter correction model based on the received plurality of scatter image data and the corresponding partial unscattered image data to obtain an updated trained scatter correction model. The analyzer may be further configured to correct at least one of the plurality of scatter images based on the updated trained scatter correction model and output at least one corrected scatter image.

[0038] In another embodiment, the trained scatter correction model may be a trained neural network model.

[0039] In another embodiment, updating the trained scatter correction model may include, for each pair of corresponding scatter image and partial no-scatter image, performing, by the analyzer, an interpolation between an output of the trained neural network model for the scatter image and an output of the trained neural network model for the partial no-scatter image, and generating an updated trained scatter correction model based on the interpolation.

[0040] In another embodiment, the trained scatter correction model may include a previously determined deconvolved point spread function (PSF) estimate.

[0041] In another embodiment, updating the trained scatter correction model may include locally parameterizing the deconvolution PSF estimate using measurement points of each pair of corresponding scatter and partial scatter-free images, and updating the deconvolution PSF estimate based on the local parameterization.

[0042] In another embodiment, a method for scatter correction of an image of an object is provided. The method can include obtaining data representing a plurality of scatter images of the object based on detection of radiation transmitted through an imaging volume of the object by a radiation detector of an imaging system. The method can further include disposing an aperture plate between the object and the radiation detector. The aperture plate can include a plurality of apertures configured to suppress scattered radiation from detection by the radiation detector. The method can further include obtaining data representing a single partially scatter-free image by the radiation detector. The single partially scatter-free image can be based on detection of radiation transmitted through an imaging volume of the object when the aperture plate is present. The method can further include receiving, by an analyzer including one or more processors, the plurality of scatter image data and the single partially scatter-free image data. The method can also include receiving, by the analyzer, a trained scatter correction model. The method can further include updating, by the analyzer, the trained scatter correction model based on the received plurality of scatter image data and the single partially scatter-free image data to obtain an updated trained scatter correction model. The method may further include correcting, by the analyzer, at least one of the plurality of scatter images based on the updated trained scatter correction model, and outputting, by the analyzer, the at least one corrected scatter image.

[0043] In another embodiment, the trained scatter correction model may be a trained neural network model.

[0044] In another embodiment, updating the trained scatter correction model may include, for each image pair including a scatter image of the plurality of scatter images and a single partial non-scatter image, performing an interpolation between an output of the trained neural network model for the scatter image and an output of the trained neural network model for the single partial non-scatter image, and generating an updated trained scatter correction model based on the interpolation.

[0045] In another embodiment, the trained scatter correction model may include a previously determined deconvolved point spread function (PSF) estimate.

[0046] In another embodiment, updating the trained scatter correction model may include locally parameterizing the deconvolution PSF estimate using measurement points of each image pair including a scatter image of the plurality of scatter images and a single partially scatter-free image, and updating the deconvolution PSF estimate based on the local parameterization.

[0047] In another embodiment, an imaging system is provided and can include a radiation source, a radiation detector, and an analyzer. The radiation source can be configured to emit radiation directed towards the object. The radiation detector can be configured to detect the emitted radiation transmitted through an imaging volume of the object. The analyzer can include one or more processors. The analyzer can also be configured to receive data representing a plurality of scatter images of the object based on detection of radiation by the radiation source transmitted through the imaging volume of the object. The analyzer can be further configured to receive data representing a single partially scatter-free image. The single partially scatter-free image can be based on detection of radiation transmitted through the imaging volume of the object when an aperture plate is present. The analyzer can be further configured to receive a trained scatter correction model, update the trained scatter correction model based on the received plurality of scatter image data and the single partially scatter-free image data to obtain an updated trained scatter correction model, correct at least one of the plurality of scatter images based on the updated trained scatter correction model, and output at least one corrected scatter image.

[0048] In another embodiment, the trained scatter correction model may be a trained neural network model.

[0049] In another embodiment, updating the trained scatter correction model may include, for each image pair including a scatter image of the plurality of scatter images and a single partial non-scatter image, performing an interpolation between an output of the trained neural network model for the scatter image and an output of the trained neural network model for the single partial non-scatter image, and generating an updated trained scatter correction model based on the interpolation.

[0050] In another embodiment, the trained scatter correction model may include a previously determined deconvolved point spread function (PSF) estimate.

[0051] In another embodiment, updating the trained scatter correction model may include locally parameterizing the deconvolution PSF estimate using measurement points of each image pair including a scatter image of the plurality of scatter images and a single partially scatter-free image, and updating the deconvolution PSF estimate based on the local parameterization. [Brief description of the drawings]

[0052] These and other features will be more readily understood from the following detailed description taken in conjunction with the accompanying drawings.

[0053] [Figure 1A] FIG. 1 is a schematic diagram illustrating a top view of an exemplary embodiment of a CT imaging system.

[0054] [Figure 1B] FIG. 1B is a schematic diagram showing a side view of the CT imaging system of FIG. 1A.

[0055] [Figure 2A] 4 is a schematic diagram showing the positions of the apertures in the aperture plate when the aperture plate is in a first position. FIG.

[0056] [Figure 2B]2 is a schematic diagram showing the positions of apertures in an aperture plate when the aperture plate is in a first position and in a second position that is translated perpendicularly relative to the first position; FIG.

[0057] [Figure 2C] 2 is a schematic diagram showing the positions of the apertures in the aperture plate when the aperture plate is in a first position and in a third position that is horizontally translated from the first position; FIG.

[0058] [Figure 2D] 3A to 3C are schematic diagrams showing the positions of apertures in an aperture plate when the aperture plate is in a first position, a second position, and a third position.

[0059] [Figure 2E] 1 is a schematic diagram illustrating the positions of apertures in an aperture plate when the aperture plate is positioned in multiple positions covering substantially all of a desired field of view.

[0060] [Figure 3A] 13 is a scatter-corrected CT image using a single position of the aperture plate.

[0061] [Figure 3B] Scatter-corrected CT images using four aperture plate positions.

[0062] [Figure 4] FIG. 1 is a schematic block diagram illustrating a computational technique that uses scatter and non-scatter images as input for the development of a scatter correction algorithm or deconvolution.

[0063] [Diagram 5] FIG. 1 is a flow diagram illustrating an example embodiment of a method for scatter correction of CT images.

[0064] [Figure 6] FIG. 13 is a schematic diagram showing multiple measurement points for adjustment given an aperture plate.

[0065] [Figure 7] FIG. 13 is a flow diagram illustrating an exemplary embodiment of a method for scatter correction in which a previously trained neural network or a previously determined deconvolution PSF is adjusted based on corresponding image pairs of a scatter image set and a partially scatter-free image set.

[0066] [Figure 8] FIG. 11 is a flow diagram illustrating an exemplary embodiment of another method for scatter correction in which a previously trained neural network or a previously determined deconvolution PSF is adjusted based on corresponding image pairs of a scatter image set and a single partially scatter-free image.

[0067] [Figure 9] FIG. 11 is a flow diagram illustrating an exemplary embodiment of a further method for scatter correction in which a previously trained neural network or a previously determined deconvolution PSF is adjusted based on an undefined number of scatter image sets for different object types, where a single partially scatter-free image or a partially scatter-free image set is used.

[0068] [Figure 10A] FIG. 1 is a schematic diagram illustrating a top view of an exemplary embodiment of a CT imaging system.

[0069] [Figure 10B] FIG. 10B is a schematic diagram showing a side view of the CT imaging system of FIG. 10A.

[0070] It should be noted that the drawings are not necessarily to scale: the drawings are intended to depict only typical aspects of the subject matter disclosed herein, and therefore should not be considered as limiting the scope of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0071] When performing an X-ray inspection, X-rays may be passed through a target object and detected by a detector to generate an image. Some X-rays may be scattered from their initial trajectory, resulting in artifacts that reduce the contrast of the image. Techniques for scatter correction have been developed, but each has problems. In one example, simulations have been developed to model the scatter. However, such simulations may be based on the limitations of non-experimental data (pure simulations) and / or may be heavily dependent on interpolation (limited experimental data supplemented by interpolation). Furthermore, not all aspects of the scatter may be simulated or interpolated with sufficient accuracy. The missing accuracy may result in the introduction of artifacts in the scatter corrected image. Thus, the scatter estimates obtained from these simulations may deviate significantly from the actual scatter. Thus, improved systems and methods for scatter correction for X-ray inspections (e.g., computed tomography) are provided. The scatter-free image and the scatter image may be used as inputs for computational techniques such as deep learning (e.g., neural networks) or deconvolution models. The scatter-free image may be generated from a plurality of partially scatter-free images acquired using an aperture plate that includes apertures. Each partial scatter-free image is acquired with the aperture plate at a different position to block a portion of the scattered x-rays. Enough partial scatter-free images can be acquired such that a combination of the aperture positions of the aperture plate at different positions covers the entire scatter image. The combination of the partial scatter-free images gives the scatter-free image. The scatter-free images thus produced are measured within a small collimated area without interpolation, increasing the accuracy of the scatter estimate.

[0072]

[0004] Embodiments of the present disclosure are generally directed to scatter correction for computed tomography (CT) imaging to achieve improved image quality. Such imaging techniques may be useful in a variety of imaging contexts, such as medical imaging, industrial metrology and inspection, security screening, and baggage or package inspection. Furthermore, such imaging techniques may be employed in a variety of imaging systems, such as CT systems, tomosynthesis systems, and x-ray imaging systems. While this discussion provides examples of industrial inspection contexts for CT systems that result in improved measurement and inspection accuracy, one of ordinary skill in the art will readily appreciate that application of these techniques in other contexts and other systems is within the scope of the present technology.

[0073] 1A-1B show an imaging system 200 configured to generate high resolution images. The imaging system 200 may be a volumetric computed tomography (VCT) system designed to both acquire image data and process the image data for display and analysis. As shown, the imaging system 200 may include a radiation source 202, such as an X-ray source 202. A collimator 205 may be positioned adjacent to the radiation source 202 to collimate radiation 204 emitted by the radiation source 202 and to adjust the size and shape of the emitted radiation 204 emitted by the radiation source 202.

[0074] The stream of radiation 204 may be projected towards a detector array 206 disposed on an opposite side of the radiation source 202 to an object 208 being imaged. The object 208 may be any object suitable for X-ray inspection (e.g., a turbine blade). The stream of radiation 204 may enter an imaging volume where the object 208 being imaged is imaged. A portion of the radiation 204 passes through or around the object 208 and impinges on the detector array 206. The detector array 206 may generally be formed as a two-dimensional array of detector elements. Data collected by the detector array 206 may be output to an analyzer 207 that includes one or more processors.

[0075] The object 208, the radiation source 202, and the detector array 206 may be displaced relative to one another, and projection data may be acquired at various views relative to the object 208, if desired. As an example, the object 208 may be positioned on a table, such as a turntable, such that the object 208 may be rotated about an axis of rotation 210. In certain embodiments, the data collected from the detector array 206 may be pre-processed (e.g., by the analyzer 207) to condition data representing line integrals of the attenuation coefficients of the object 208 being scanned. The processed data or projections may then be reconstructed (e.g., by the analyzer 207 or another computing device) to formulate a volumetric image of the scanned area, as discussed in more detail in U.S. Pat. No. 9,804,106, which is incorporated herein by reference in its entirety.

[0076] The imaging system 200 may employ various scatter mitigation and / or correction techniques to improve image quality and resolution. For example, a scatter rejection aperture plate 212 may be employed to reject scattered radiation originating from the object 208 as well as scattered radiation originating from the background. To further improve resolution and image quality, the aperture plate 212 may be movable between multiple positions, which will be discussed in more detail below. By moving the aperture plate 212 between multiple positions, smaller structures on the object 208 may be recognized and artifacts may be better avoided.

[0077] As discussed in more detail in U.S. Patent No. 9,804,106, the aperture plate 212 may include a plurality of sub-centimeter sized apertures 48, such as circular apertures drilled in the plate. The apertures 48 may be positioned on a two-dimensional grid. Embodiments of the apertures 48 may have any geometric shape, such as circular, rectangular, or hexagonal shapes, among others. In certain embodiments, the circular apertures 48 may be about 1-2 millimeters in diameter spaced apart from each other (center-to-center) by about 5 millimeters.

[0078] As discussed above, the aperture plate 212 may be movable between multiple positions to increase the resolution and quality of the images generated. As shown in FIG. 2A, the aperture plate 212 may be initially positioned at a first position 211. After the first grid image is collected, the aperture plate 212 may be repositioned at a second position 213, 214 and a second grid image may be collected. In one embodiment, the aperture plate 212 may be moved in one direction. For example, the aperture plate 212 may be moved vertically from the first position 211 to the second position 213 as shown in FIG. 2B and a second image may be collected, or the aperture plate 212 may be moved horizontally from the first position 211 to the third position 214 as shown in FIG. 2C and a second image may be collected. In another embodiment, the aperture plate 212 may be moved bidirectionally. For example, the aperture plate 212 may be moved both vertically and horizontally as shown in FIG. 2D. In this embodiment, the aperture plate 212 may be placed at a first position 211, a first image may be collected, moved to a second position 213, a second image may be collected, moved to a third position 214, a third image may be collected, and moved to a fourth position 215, a fourth image may be collected. In another embodiment, the aperture plate 212 may be rotated relative to the object 208. Images may be collected at each of the aperture plate 212 positions 211, 213, 214, 215. The process of moving the aperture plate 212 and acquiring images may be repeated until the aperture positions cover the entire desired area, as shown in FIG. 2E.

[0079] In another embodiment, the image resolution can also be increased by moving the object 208 in front of the grid of aperture plate 212. Similar to repositioning the aperture plate 212 discussed above, in this embodiment the sample can be moved, for example, unidirectionally, bidirectionally, or rotated. By repositioning the apertures 48 of the aperture plate 212 relative to the object 208, the image resolution can be increased. The aperture plate 212 and / or the object 208 can be repositioned manually or automatically.

[0080] The resolution of the image can be determined by the number of positions at which the aperture plate 212 and / or the object 208 can be positioned. As the number of positions increases, the image resolution also increases. For example, using bidirectional movement of the aperture plate 212, positioning the aperture plate 212 in four positions increases the image resolution by a factor of two. In another example, positioning the aperture plate 212 in sixteen positions increases the image resolution by a factor of four. The improvement in image resolution by repositioning the aperture plate 212 is shown in Figures 3A-3B.

[0081] Figure 3A shows an image 216 of the object 208 that was generated with the aperture plate 212 positioned in a single position. Figure 3B shows an image 219 of the object 208 with the aperture plate 212 positioned in four positions during data collection. As indicated by the first location 217 and second location 218 shown in these figures, the additional positions of the aperture plate 212 result in the image of Figure 3B in which more detail is visible at each of the first location 217 and second location 218 compared to Figure 3A.

[0082] The scatter-free image can be used in combination with a computational method for scatter correction. This combination can produce better results and less scanning effort for scatter correction in X-ray tomography. Examples of computational methods can include deep learning (e.g., neural networks) or deconvolution methods.

[0083] As shown in FIG. 4, a set of scatter images 402 and a set of no-scatter images can be input to a model 406 (e.g., a neural network or a scatter edge spread function (ESF)). In the case of a neural network, one or more algorithms can be received or generated for scatter correction and applied to the scatter image set 402. In the case of deconvolution, a point spread function (PSF) for deconvolution can be received or generated and applied to the scatter image set 402 using the scatter ESF. In either case, the deviation between the scatter corrected image set and the no-scatter image set 404 is determined and used as feedback 410 for modification of the algorithm. This process repeats itself, updating the algorithm or parameters at each iteration until the deviation between the scatter corrected image set and the no-scatter image set 404 is less than a predetermined deviation. The determined algorithm or PSF is then output.

[0084] 5 is a flow diagram illustrating one exemplary embodiment of a method 500 for scatter correction of an image of an object, such as object 208, by employing advanced scatter measurement and correction techniques on imaging system 200. As shown, method 500 includes operations 502-512. It should be understood that this method is merely exemplary and selected operations can be modified, added, deleted, and / or rearranged as desired.

[0085] In operation 502, at least one scattering image may be acquired of the object 208. In certain embodiments, multiple scattering images (e.g., a scattering image set) may be acquired.

[0086] In operation 504, the aperture plate 212 is positioned at a first position between the object 208 and the detector 206. As discussed above, the aperture plate 212 may include multiple apertures 211 and may be configured to suppress scattered radiation (e.g., x-rays scattered from the object 208 and / or background) from detection by the detector array 206. When the aperture plate 212 is in the first position, at least one first partially unscattered image (e.g., a single partially unscattered image or a partially unscattered image set including multiple partially unscattered images) may be acquired.

[0087] In operation 506, the aperture plate 212 may be moved to one or more other positions (e.g., one or more second positions) different from the first position. As an example, the movement of the aperture plate 212 may include only translation, only rotation, or a combination of translation and rotation. The translation may include movement in at least one direction (e.g., horizontally, vertically, or a combination thereof). The apertures may overlap or be spaced apart by a predetermined distance between the first and second positions. A second partially unscattered image or partially unscattered image set may be acquired when the aperture plate 212 is positioned at a selected one of the second positions (e.g., at least a portion and up to all of the second positions). Operation 506 may be repeated to generate as many partially unscattered images / image sets as necessary such that the apertures cover substantially the entire scatter image.

[0088] In operation 510, the acquired partial non-scattered images may be received by an analyzer and combined to generate a non-scattered image / image set. In certain embodiments, at least a portion of the partial non-scattered images are used.

[0089] In operation 512, the scattered image / image set and the non-scattered image / image set are used to train a deep algorithm or estimate the PSF of a deconvolution algorithm.

[0090] It will be appreciated that changes in the X-ray inspection system may change the images acquired by the X-ray inspection system. Such changes may include, but are not limited to, the X-ray detector, the X-ray source, the target being inspected, and the environment. Changes in the X-ray detector may include one or more of replacing one X-ray detector with another and changes in detection capabilities due to changes over time. Changes in the X-ray source may include any change in the X-ray spectrum emitted by the X-ray source. Such changes may occur due to replacement of components of the X-ray source (e.g., filters, tubes, etc.) or replacement of the entire X-ray source with another X-ray source. Changes in the environment may include any change in the X-ray scattering behavior due to the environment surrounding the X-ray source.

[0091] As a result of these changes, a neural network algorithm trained prior to such changes may introduce errors into the output scatter corrected image, as well as deconvolution PSF estimates determined prior to such changes may introduce errors into the output scatter corrected image.

[0092] Thus, to address the effects of changes in the x-ray inspection system, a previously determined neural network algorithm can be retrained or the deconvolution PSF estimate can be updated. That is, the neural network training or deconvolution PSF estimate does not need to be completely regenerated. Beneficially retraining a previously determined neural network algorithm or updating a previously determined PSF estimate may require significantly less training data and time than generating one from scratch.

[0093] As will be discussed in more detail below, retraining the neural network algorithm or updating the deconvolution PSF estimate can be performed in a variety of ways. A single image or the entire image set can be employed. The adjustment can be valid for one data set and one sample type, or it can be valid for many data sets also having different sample types. The choice of training data depends on the neural network or the PSF evaluated, whether they were created for general correction or for a more specific case.

[0094] FIG. 6 is a schematic diagram of the aperture plate 212 showing a number of different possible measurement points that can be used to acquire a partially scatter-free image for retraining a neural network or adjusting the PSF estimate of a deconvolution algorithm.

[0095] 7 is a flow diagram of an example embodiment of a method 700 for retraining a neural network or adjusting the PSF estimate of a deconvolution algorithm. As shown, method 700 includes operations 702-712. It should be understood that this method is merely exemplary and selected operations can be modified, added, deleted, and / or rearranged as desired.

[0096] In operation 702, multiple scattering images (eg, a set of scattering images) of the object 208 may be acquired.

[0097] In operation 704, an aperture plate 212 may be positioned between the object 208 and the detector 206, and a partial non-scattered image corresponding to each of the acquired scatter images may be acquired by the radiation detector 206. That is, the partial non-scattered images corresponding to the scatter images may be acquired under the same or nearly the same conditions (e.g., within instrument tolerances) as the scatter images, except that the difference is the presence of the aperture plate 212 during acquisition of the partial non-scattered images.

[0098] The acquired scattered image data and partial un-scattered image data may be further received by the analyzer 207. In certain embodiments, the analyzer 207 may receive the acquired scattered image data and partial un-scattered image data directly from the radiation detector 206. In other embodiments, the analyzer 207 may receive the acquired scattered image data and partial un-scattered image data from another source (e.g., a memory device).

[0099] In operation 706, the analyzer 207 may use the set of scatter images and the set of partially scatter-free images to retrain (tune or update) a previously trained scatter correction model (e.g., a previously trained scatter correction algorithm of a neural network, or a previously determined deconvolution PSF estimate). In the case of a scatter correction model using a neural network, the retraining may include performing an interpolation between the output of a neural network previously trained for a given scatter image and the corresponding acquired partially scatter-free image. Thus, there may be an interpolation determined by the analyzer 207 for each corresponding scatter image and partially scatter-free image pair. Each interpolation may be utilized by the neural network to generate an adjusted scatter correction algorithm. In a scatter correction model employing a deconvolution approach, the deconvolution function may be parameterized locally for each measurement point (e.g., for each corresponding scatter image and partially scatter-free image pair), rather than globally.

[0100] In operation 710, the analyzer 207 may use the updated scatter correction model (e.g., a neural network algorithm or a deconvolution PSF estimate) to correct at least one scatter image of the plurality of scatter images.

[0101] At operation 712, the analyzer 207 may output at least one of the corrected scatter images. For example, the at least one corrected scatter image may be output to a memory device for storage and / or a display device for viewing.

[0102] 8 is a flow diagram of an example embodiment of another method 800 that may be performed by analyzer 207 to retrain (tune or update) the PSF estimate of a neural network or deconvolution algorithm. As shown, method 800 includes operations 802-812. It should be understood that this method is merely exemplary and that selected operations may be modified, added, deleted, and / or rearranged as desired.

[0103] In operation 802 , multiple scatter images (eg, a set of scatter images) of an object 208 may be acquired by a radiation detector 206 .

[0104] In operation 804, an aperture plate 212 may be positioned between the object 208 and the detector 206, and a single partially unscattered image may be acquired by the radiation detector 206.

[0105] The acquired scattered image data and partial un-scattered image data may be further received by the analyzer 207. In certain embodiments, the analyzer 207 may receive the acquired scattered image data and partial un-scattered image data directly from the radiation detector 206. In other embodiments, the analyzer 207 may receive the acquired scattered image data and partial un-scattered image data from another source (e.g., a memory device).

[0106] In operation 806, the multiple scatter image data and the single partial no-scatter image data may be used to retrain (e.g., tune or update) a previously trained scatter correction model (e.g., a previously trained scatter correction algorithm of a neural network, or a previously determined deconvolution PSF estimate). In the case of a scatter correction model using a neural network, the retraining may include performing an interpolation between the output of a neural network previously trained for a given scatter image and the single acquired partial no-scatter image. Thus, there may be an interpolation determined by the analyzer 207 for each pair of scatter image and single partial no-scatter image. Each interpolation may be utilized by the neural network to generate an adjusted scatter correction algorithm. In a scatter correction model employing a deconvolution approach, the deconvolution function may be parameterized locally for each measurement point (e.g., for each pair of scatter image and single partial no-scatter image), rather than globally.

[0107] In operation 810, the analyzer 207 may use the updated scatter correction model (e.g., a neural network algorithm or a deconvolution PSF estimate) to correct at least one scatter image of the multiple scatter images.

[0108] At operation 812, the analyzer 207 may output at least one of the corrected scatter images. For example, the at least one corrected scatter image may be output to a memory device for storage and / or a display device for viewing.

[0109] In situations where a single partially scatter-free image is acquired for alignment, the X-ray inspection system can optionally omit the ability to move the aperture plate 212, as shown in Figures 10A-10B.

[0110] The choice of whether to employ method 700 or method 800 to adjust the neural network or deconvolution ESF may depend on the use case and the required accuracy of the scatter correction. As an example, in situations where each inspection object 208 is the same, a single partially scatter-free image approach of method 800 may be sufficient. This measurement may be performed periodically to adjust the neural network or deconvolution ESF to account for changes in components over time.

[0111] 9 is a flow diagram of an example embodiment of a further method 900 for retraining (tuning) the PSF estimate of a neural network or deconvolution algorithm. As shown, method 900 includes operations 902-912. It should be understood that this method is merely exemplary and selected operations can be modified, added, removed, and / or rearranged as desired.

[0112] In operation 902, a scattering image or set of scattering images of an object 208 is obtained.

[0113] In operation 904, an aperture plate 212 is positioned between the object 208 and the detector 206 and a single partially scatter-free image or a set of partially scatter-free images is acquired. As described above, when a set of partially scatter-free images is acquired, each image in the partially scatter-free image set can correspond to a respective image in the scatter image set.

[0114] In operation 906, the set of scatter images and the single partial no-scatter image or no-scatter image set may be used to retrain (tune) a previously trained neural network or a previously determined deconvolution PSF estimate. The tuning performed using the partial no-scatter image set may be performed as discussed above in operation 706. The tuning performed using the single partial no-scatter image may be performed as discussed above in operation 806.

[0115] In operation 910, the scatter image set may be corrected using the tuned neural network algorithm or the deconvolution PSF estimate.

[0116] In operation 912, a scatter image or set of scatter images may be acquired of the same type of object 208 or of a different type of object 208. Scatter correction may be performed using a tuned neural network algorithm or a deconvolution PSF estimate. Operation 912 may be repeated as long as the correction quality is sufficient for the desired result without retuning.

[0117] Exemplary technical effects of the methods, systems, and devices described herein include, by way of non-limiting example, improved scatter correction of X-ray images. A scatter-free image can be generated from a plurality of partial scatter-free images acquired using aperture plates positioned at different positions. The scatter-free image can be used in combination with computational techniques for scatter correction (e.g., deep learning, deconvolution, etc.). This combination can create better results and less scanning effort for scatter correction in X-ray tomography.

[0118] Certain exemplary embodiments have been described to provide a general understanding of the principles of the structure, function, manufacture, and use of the systems, devices, and methods disclosed herein. One or more examples of these embodiments are shown in the accompanying drawings. Those skilled in the art will understand that the systems, devices, and methods explicitly described herein and illustrated in the accompanying drawings are non-limiting exemplary embodiments, and that the scope of the present invention is defined only by the claims. Features shown or described in connection with one exemplary embodiment may be combined with features of other embodiments. Such modifications and variations are intended to be included within the scope of the present invention. Furthermore, in this disclosure, similarly named components of the embodiments generally have similar features, and therefore each feature of each similarly named component in a specific embodiment has not necessarily been described in full detail.

[0119] The subject matter described herein can be implemented in analog electronic circuitry, digital electronic circuitry, and / or computer software, firmware, or hardware, including the structural means disclosed herein and their structural equivalents, or in combinations thereof. The subject matter described herein can be implemented as one or more computer program products, such as one or more computer programs tangibly embodied in an information carrier (e.g., in a machine-readable storage device) or embodied in a propagated signal for execution by or to control the operation of a data processing apparatus (e.g., a programmable processor, a computer, or multiple computers). Computer programs (also known as programs, software, software applications, or code) can be written in any form of programming language, including compiled or interpreted languages, and can be deployed in any form, such as as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. A computer program does not necessarily correspond to a file. A program can be stored in a portion of a file that holds other programs or data, in a single file dedicated to the program, or in multiple coordinated files (e.g., files that store one or more modules, subprograms, or portions of code). A computer program can be deployed to be executed on one computer or on multiple computers at one site, or distributed across multiple sites and interconnected by a communications network.

[0120] The processes and logic flows described herein, including the method steps of the subject matter described herein, may be performed by one or more programmable processors executing one or more computer programs to perform functions of the subject matter described herein by operating on input data and generating output. The processes and logic flows may also be performed by special purpose logic circuitry, such as a field programmable gate array (FPGA) or an application-specific integrated circuit (ASIC), and apparatus of the subject matter described herein may be implemented as such special purpose logic circuitry.

[0121] Processors suitable for executing computer programs include, for example, both general purpose and special purpose microprocessors, as well as any one or more processors of any kind of digital computer. Typically, a processor receives instructions and data from a read-only memory or a random access memory, or both. The essential elements of a computer are a processor for executing instructions and one or more memory devices for storing instructions and data. Typically, a computer also includes one or more mass storage devices for storing data, such as magnetic disks, magneto-optical disks, or optical disks, or is operatively coupled to receive data therefrom, transfer data thereto, or both. Information carriers suitable for embodying computer program instructions and data include all forms of non-volatile memory, including, by way of example, semiconductor memory devices (e.g., EPROM, EEPROM, and flash memory devices), magnetic disks (e.g., internal hard disks or removable disks), magneto-optical disks, and optical disks (e.g., CD and DVD disks). The processor and memory may be supplemented by, or incorporated in, special purpose logic circuitry.

[0122] To provide for interaction with a user, the subject matter described herein may be implemented on a computer having a display device, such as a CRT (cathode ray tube) or LCD (liquid crystal display) monitor, for displaying information to the user, as well as a keyboard and pointing device (e.g., a mouse or trackball) that the user may use to provide input to the computer. Other types of devices may also be used to provide interaction with a user. For example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback), and input from the user may be received in any form, including acoustic, speech, or tactile input.

[0123] The techniques described herein may be implemented using one or more modules. As used herein, the term "module" refers to computing software, firmware, hardware, and / or various combinations thereof. However, at a minimum, a module should not be interpreted as software that is not implemented on hardware, firmware, or on a non-transitory processor-readable and recordable storage medium (i.e., a module is not software itself). In fact, a "module" should always be interpreted to include at least some physical non-transitory hardware, such as a processor or part of a computer. Two different modules may share the same physical hardware (e.g., two different modules may use the same processor and network interface). The modules described herein can be combined, integrated, separated, and / or replicated to support various applications. Also, functions described herein as being performed in a particular module may be performed in one or more other modules and / or by one or more other devices instead of or in addition to the functions performed in the particular module. Furthermore, modules may be implemented across multiple devices and / or other components, local or remote from each other. Additionally, modules may be moved from one device and added to another device and / or may be incorporated into both devices.

[0124] The subject matter described herein may be implemented in a computing system that includes back-end components (e.g., data servers), middleware components (e.g., application servers), or front-end components (e.g., client computers having a graphical user interface or web browser through which a user can interact with an implementation of the subject matter described herein), or any combination of such back-end, middleware, and front-end components. The components of the system may be interconnected by any form or medium of digital data communication, e.g., a communication network. Examples of communication networks include a local area network ("local area network, LAN") and a wide area network ("wide area network, WAN"), e.g., the Internet.

[0125] As used herein throughout the specification and claims, approximation language may be applied to modify any quantitative expression that may be permissibly changed without resulting in a change in the basic function to which it relates. Thus, a value modified by one or more terms such as "about," "approximately," and "substantially" is not limited to the exact value specified. In at least some instances, the approximation language may correspond to the precision of an instrument for measuring the value. Combinations and / or interchanges of range limitations may be made herein throughout the specification and claims, and such ranges are identified and include all subranges contained therein unless the context or language dictates otherwise.

[0126] Those skilled in the art will appreciate further features and advantages of the present invention based on the above-described embodiments. Accordingly, this application is not to be limited by what has been particularly shown and described, except as indicated by the appended claims. All publications and references cited herein are expressly incorporated by reference in their entirety.

[0127] Preferred embodiments of the present invention will be described below in detail. EMBODIMENT 1 1. A method for scatter correction of an image of an object, comprising: obtaining data representative of a plurality of scattering images of the object based on detection of radiation transmitted through an imaging volume of the object by a radiation detector of the imaging system; disposing an aperture plate between the object and the radiation detector, the aperture plate including a plurality of apertures configured to suppress scattered radiation from detection by the radiation detector; acquiring, by the radiation detector, data representative of a partial unscattered image corresponding to each scatter image, each partial unscattered image being based on detection of radiation transmitted through the imaging volume of the object when the aperture plate is present, the scatter image and its corresponding partial unscattered image being acquired under substantially the same conditions except for the presence of the aperture plate; receiving, by an analyzer including one or more processors, the plurality of scattered image data and corresponding partially unscattered image data; receiving, by the analyzer, a trained scatter correction model; updating, by the analyzer, the trained scatter correction model based on the received plurality of scatter image data and corresponding partial scatter-free image data to obtain an updated trained scatter correction model; correcting, by the analyzer, at least one of the plurality of scatter images based on the updated trained scatter correction model; and outputting, by the analyzer, at least one corrected scatter image. EMBODIMENT 2 2. The method of embodiment 1, wherein the trained scatter correction model is a trained neural network model. EMBODIMENT 3 updating the trained scatter correction model for each pair of corresponding scatter and partial no-scatter images, performing an interpolation between an output of the trained neural network model for the scatter image and an output of the trained neural network model for the partial no-scatter image; and generating the updated trained scatter correction model based on the interpolation. EMBODIMENT 4 2. The method of embodiment 1, wherein the trained scatter correction model comprises a previously determined deconvoluted point spread function (PSF) estimate. EMBODIMENT 5 updating the trained scatter correction model; locally parameterizing the deconvolution PSF estimate using measurement points of each pair of corresponding scatter and partially scatter-free images; updating the deconvolution PSF estimate based on the local parameterization. EMBODIMENT 6 1. An imaging system, comprising: a radiation source configured to emit directed radiation towards the object; a radiation detector configured to detect the emitted radiation transmitted through an imaging volume of the object; and and an analyzer including one or more processors, the analyzer comprising: receiving data representing a plurality of scattering images of the object based on detection of radiation by the radiation source transmitted through an imaging volume of the object; receiving data representing a partial non-scatter image corresponding to each scatter image, each partial non-scatter image being based on detection of radiation transmitted through the imaging volume of the object when the aperture plate is present, the scatter image and its corresponding partial non-scatter image being acquired under substantially the same conditions except for the presence of the aperture plate; receiving a trained scatter correction model; updating the trained scatter correction model based on the received plurality of scatter image data and corresponding partial scatter-free image data to obtain an updated trained scatter correction model; correcting at least one of the plurality of scatter images based on the updated trained scatter correction model; and and outputting at least one corrected scatter image. EMBODIMENT 7 7. The imaging system of embodiment 6, wherein the trained scatter correction model is a trained neural network model. EMBODIMENT 8 updating the trained scatter correction model for each pair of corresponding scatter and partial no-scatter images, performing an interpolation between an output of the trained neural network model for the scatter image and an output of the trained neural network model for the partial no-scatter image; and generating the updated trained scatter correction model based on the interpolation. EMBODIMENT 9 7. The imaging system of embodiment 6, wherein the trained scatter correction model includes a previously determined deconvoluted point spread function (PSF) estimate. EMBODIMENT 10 updating the trained scatter correction model; locally parameterizing the deconvolution PSF estimate using measurement points of each pair of corresponding scatter and partially scatter-free images; 10. The imaging system of embodiment 9, comprising: EMBODIMENT 11 1. A method for scatter correction of an image of an object, comprising: obtaining data representative of a plurality of scattering images of the object based on detection of radiation transmitted through an imaging volume of the object by a radiation detector of the imaging system; disposing an aperture plate between the object and the radiation detector, the aperture plate including a plurality of apertures configured to suppress scattered radiation from detection by the radiation detector; acquiring, by the radiation detector, data representative of a single partially scatter-free image, the single partially scatter-free image being based on detection of radiation transmitted through the imaging volume of the object when the aperture plate is present; receiving, by an analyzer including one or more processors, the plurality of scattered image data and the single partially unscattered image data; receiving, by the analyzer, a trained scatter correction model; updating, by the analyzer, the trained scatter correction model based on the received plurality of scatter image data and the single partial no-scatter image data to obtain an updated trained scatter correction model; correcting, by the analyzer, at least one of the plurality of scatter images based on the updated trained scatter correction model; and outputting, by the analyzer, at least one corrected scatter image. EMBODIMENT 12 12. The method of embodiment 11, wherein the trained scatter correction model is a trained neural network model. EMBODIMENT 13 updating the trained scatter correction model for each image pair including a scatter image of the plurality of scatter images and the single partial non-scatter image, performing an interpolation between an output of the trained neural network model for the scatter image and an output of the trained neural network model for the single partial non-scatter image; and generating the updated trained scatter correction model based on the interpolation. EMBODIMENT 14 12. The method of embodiment 11, wherein the trained scatter correction model comprises a previously determined deconvoluted point spread function (PSF) estimate. EMBODIMENT 15 updating the trained scatter correction model; locally parameterizing the deconvolution PSF estimate using measurement points of each image pair including a scatter image of the plurality of scatter images and the single partially scatter-free image; and updating the deconvolution PSF estimate based on the local parameterization. EMBODIMENT 16 1. An imaging system, comprising: a radiation source configured to emit directed radiation towards the object; a radiation detector configured to detect the emitted radiation transmitted through an imaging volume of the object; and and an analyzer including one or more processors, the analyzer comprising: receiving data representing a plurality of scattering images of the object based on detection of radiation by the radiation source transmitted through an imaging volume of the object; receiving data representative of a single partially scatter-free image, the single partially scatter-free image being based on detection of radiation transmitted through the imaging volume of the object when the aperture plate is present; receiving a trained scatter correction model; updating the trained scatter correction model based on the received plurality of scatter image data and the single partial non-scatter image data to obtain an updated trained scatter correction model; correcting at least one of the plurality of scatter images based on the updated trained scatter correction model; and and outputting at least one corrected scatter image. EMBODIMENT 17 17. The imaging system of embodiment 16, wherein the trained scatter correction model is a trained neural network model. EMBODIMENT 18 updating the trained scatter correction model for each image pair including a scatter image of the plurality of scatter images and the single partial non-scatter image, performing an interpolation between an output of the trained neural network model for the scatter image and an output of the trained neural network model for the single partial non-scatter image; and generating the updated trained scatter correction model based on the interpolation. EMBODIMENT 19 17. The imaging system of embodiment 16, wherein the trained scatter correction model includes a previously determined deconvoluted point spread function (PSF) estimate. EMBODIMENT 20 updating the trained scatter correction model; locally parameterizing the deconvolution PSF estimate using measurement points of each image pair including a scatter image of the plurality of scatter images and the single partially scatter-free image; and updating the deconvolution PSF estimate based on the local parameterization.

Claims

1. 1. A method for scatter correction of an image of an object, comprising: acquiring, by a radiation detector of the imaging system, a plurality of scattered image data, each representing a scattered image of the object based on detection of radiation transmitted through an imaging volume of the object; disposing an aperture plate between the object and the radiation detector, the aperture plate including a plurality of apertures arranged in a two-dimensional grid, positioned such that when radiation is directed from a radiation source towards the object, unscattered radiation passes through the plurality of apertures and a portion of scattered radiation is blocked by the aperture plate; acquiring, by the radiation detector, partial free-scatter image data representative of a partial free-scatter image corresponding to each of the scatter images, each of the partial free-scatter images being based on detection of radiation transmitted through the imaging volume of the object when the aperture plate is present, the scatter image and its corresponding partial free-scatter image being acquired under substantially the same imaging conditions except for the presence of the aperture plate; receiving, by an analyzer including one or more processors, the plurality of scattered image data and corresponding partially unscattered image data; receiving, by the analyzer, a trained scatter correction model, the trained scatter correction model being a trained neural network model; updating, by the analyzer, the trained scatter correction model based on the received plurality of scatter image data and corresponding partial scatter-free image data to obtain an updated trained scatter correction model; correcting, by the analyzer, at least one of the plurality of scatter image data based on the updated trained scatter correction model; outputting, by the analyzer, at least one corrected scatter image data; the trained scatter correction model is trained using at least a portion of at least one scatter image based on detection of radiation transmitted through an imaged volume of an object and at least one non-scatter image based on detection of radiation transmitted through an imaged volume of the object; the at least one scatter-free image is generated by combining image portions corresponding to the aperture regions of a plurality of partially scatter-free images captured with different positions of the aperture plate, the at least one scatter image and the plurality of partially scatter-free images being captured under substantially the same imaging conditions except for the presence of the aperture plate; updating the trained scatter correction model for each pair of corresponding scattered image data and partial non-scattered image data, performing an interpolation between an output of the trained neural network model for the scattered image data and an output of the trained neural network model for the partial non-scattered image data; generating the updated trained scatter correction model based on the interpolation.

2. A method for scatter correction of an image of an object, comprising: acquiring, by a radiation detector of the imaging system, a plurality of scattered image data, each representing a scattered image of the object based on detection of radiation transmitted through an imaging volume of the object; disposing an aperture plate between the object and the radiation detector, the aperture plate including a plurality of apertures arranged in a two-dimensional grid, positioned such that when radiation is directed from a radiation source towards the object, unscattered radiation passes through the plurality of apertures and a portion of scattered radiation is blocked by the aperture plate; acquiring, by the radiation detector, partial free-scatter image data representative of a partial free-scatter image corresponding to each of the scatter images, each of the partial free-scatter images being based on detection of radiation transmitted through the imaging volume of the object when the aperture plate is present, the scatter image and its corresponding partial free-scatter image being acquired under substantially the same imaging conditions except for the presence of the aperture plate; receiving, by an analyzer including one or more processors, the plurality of scattered image data and corresponding partially unscattered image data; receiving, by the analyzer, a trained scatter correction model; updating, by the analyzer, the trained scatter correction model based on the received plurality of scatter image data and corresponding partial scatter-free image data to obtain an updated trained scatter correction model; correcting, by the analyzer, at least one of the plurality of scatter image data based on the updated trained scatter correction model; outputting, by the analyzer, at least one corrected scatter image data; the trained scatter correction model is trained using at least a portion of at least one scatter image based on detection of radiation transmitted through an imaged volume of an object and at least one non-scatter image based on detection of radiation transmitted through an imaged volume of the object; the at least one scatter-free image is generated by combining image portions corresponding to the aperture regions of a plurality of partially scatter-free images captured with different positions of the aperture plate, the at least one scatter image and the plurality of partially scatter-free images being captured under substantially the same imaging conditions except for the presence of the aperture plate; the trained scatter correction model includes a previously determined deconvolved point spread function estimate; updating the trained scatter correction model locally parameterizing the deconvoluted point spread function estimate using measurement points of each pair of corresponding scatter image data and partially scatter-free image data; and updating the deconvolved point spread function estimate based on the local parameterization.

3. 3. The method of claim 2, wherein the trained scatter correction model includes a deconvolution that includes a scattering edge spread function, the scattering edge spread function further including a point spread function configured to correct for scattering in a scatter image.

4. 1. An imaging system, comprising: a radiation source configured to emit directed radiation towards the object; a radiation detector configured to detect the emitted radiation transmitted through an imaging volume of the object; and an aperture plate including a plurality of apertures arranged in a two-dimensional grid, the aperture plate being positioned between the object and the radiation detector such that when radiation is directed from the radiation source towards the object, unscattered radiation passes through the plurality of apertures and a portion of scattered radiation is blocked by the aperture plate; an analyzer including one or more processors; wherein the analyzer comprises: receiving a plurality of scattered image data, each representing a scattered image of the object based on detection of radiation by the radiation source transmitted through the imaging volume of the object; receiving partial free-scatter image data representing a partial free-scatter image corresponding to each said scatter image, each said partial free-scatter image being based on detection of radiation transmitted through the imaging volume of the object when the aperture plate is present, and wherein the scatter image and its corresponding partial free-scatter image are captured under substantially the same imaging conditions except for the presence of the aperture plate; receiving a trained scatter correction model, the trained scatter correction model being a trained neural network model; updating the trained scatter correction model based on the received plurality of scatter image data and corresponding partial scatter-free image data to obtain an updated trained scatter correction model; correcting at least one of the plurality of scatter image data based on the updated trained scatter correction model; and outputting the at least one corrected scattering image data. the trained scatter correction model is trained using at least a portion of at least one scatter image based on detection of radiation transmitted through an imaged volume of an object and at least one non-scatter image based on detection of radiation transmitted through an imaged volume of the object; the at least one scatter-free image is generated by combining image portions of a plurality of partial scatter-free image data captured with different positions of the aperture plate, the at least one scatter image and the plurality of partial scatter-free images being captured under substantially the same imaging conditions except for the presence of the aperture plate; updating the trained scatter correction model for each pair of corresponding scattered image data and partial non-scattered image data, performing an interpolation between an output of the trained neural network model for the scattered image data and an output of the trained neural network model for the partial non-scattered image data; generating the updated trained scatter correction model based on the interpolation.

5. An imaging system, comprising: a radiation source configured to emit directed radiation towards the object; a radiation detector configured to detect the emitted radiation transmitted through an imaging volume of the object; and an aperture plate including a plurality of apertures arranged in a two-dimensional grid, the aperture plate being positioned between the object and the radiation detector such that when radiation is directed from the radiation source towards the object, unscattered radiation passes through the plurality of apertures and a portion of scattered radiation is blocked by the aperture plate; an analyzer including one or more processors; wherein the analyzer comprises: receiving a plurality of scattered image data, each representing a scattered image of the object based on detection of radiation by the radiation source transmitted through the imaging volume of the object; receiving partial free-scatter image data representing a partial free-scatter image corresponding to each said scatter image, each said partial free-scatter image being based on detection of radiation transmitted through the imaging volume of the object when the aperture plate is present, and wherein the scatter image and its corresponding partial free-scatter image are captured under substantially the same imaging conditions except for the presence of the aperture plate; receiving a trained scatter correction model; updating the trained scatter correction model based on the received plurality of scatter image data and corresponding partial scatter-free image data to obtain an updated trained scatter correction model; correcting at least one of the plurality of scatter image data based on the updated trained scatter correction model; and outputting the at least one corrected scattering image data. the trained scatter correction model is trained using at least a portion of at least one scatter image based on detection of radiation transmitted through an imaged volume of an object and at least one non-scatter image based on detection of radiation transmitted through an imaged volume of the object; the at least one scatter-free image is generated by combining image portions of a plurality of partial scatter-free image data captured with different positions of the aperture plate, the at least one scatter image and the plurality of partial scatter-free images being captured under substantially the same imaging conditions except for the presence of the aperture plate; the trained scatter correction model includes a previously determined deconvolved point spread function estimate; updating the trained scatter correction model locally parameterizing the deconvoluted point spread function estimate using measurement points of each pair of corresponding scatter image data and partially scatter-free image data; updating the deconvolved point spread function estimate based on the local parameterization.

6. 6. The imaging system of claim 5, wherein the trained scatter correction model includes a deconvolution that includes a scattering edge spread function, the scattering edge spread function further including a point spread function configured to correct for scattering in a scatter image.

7. 1. A method for scatter correction of an image of an object, comprising: acquiring, by a radiation detector of the imaging system, a plurality of scattered image data, each representing a scattered image of the object based on detection of radiation transmitted through an imaging volume of the object; disposing an aperture plate between the object and the radiation detector, the aperture plate including a plurality of apertures arranged in a two-dimensional grid, positioned such that when radiation is directed from a radiation source towards the object, unscattered radiation passes through the plurality of apertures and a portion of scattered radiation is blocked by the aperture plate; acquiring, by the radiation detector, a single partially scatter-free image data representative of a single partially scatter-free image, the single partially scatter-free image being based on detection of radiation transmitted through the imaging volume of the object when the aperture plate is present; receiving, by an analyzer including one or more processors, the plurality of scattered image data and the single partially unscattered image data; receiving, by the analyzer, a trained scatter correction model, the trained scatter correction model being a trained neural network model; updating, by the analyzer, the trained scatter correction model based on the received plurality of scatter image data and the single partial no-scatter image data to obtain an updated trained scatter correction model; correcting, by the analyzer, at least one of the plurality of scatter image data based on the updated trained scatter correction model; outputting, by the analyzer, at least one corrected scatter image data; the trained scatter correction model is trained using at least a portion of at least one scatter image based on detection of radiation transmitted through an imaged volume of an object and at least one non-scatter image based on detection of radiation transmitted through an imaged volume of the object; the at least one scatter-free image is generated by combining image portions of a plurality of partial scatter-free image data captured with different positions of the aperture plate, the at least one scatter image and the plurality of partial scatter-free images being captured under substantially the same imaging conditions except for the presence of the aperture plate; updating the trained scatter correction model for each image pair including scattered image data of the plurality of scattered image data and the single partial non-scattered image data, performing an interpolation between an output of the trained neural network model for the scattered image data and an output of the trained neural network model for the single partial non-scattered image data; generating the updated trained scatter correction model based on the interpolation.

8. A method for scatter correction of an image of an object, comprising: acquiring, by a radiation detector of the imaging system, a plurality of scattered image data, each representing a scattered image of the object based on detection of radiation transmitted through an imaging volume of the object; disposing an aperture plate between the object and the radiation detector, the aperture plate including a plurality of apertures arranged in a two-dimensional grid, positioned such that when radiation is directed from a radiation source towards the object, unscattered radiation passes through the plurality of apertures and a portion of scattered radiation is blocked by the aperture plate; acquiring, by the radiation detector, a single partially scatter-free image data representative of a single partially scatter-free image, the single partially scatter-free image being based on detection of radiation transmitted through the imaging volume of the object when the aperture plate is present; receiving, by an analyzer including one or more processors, the plurality of scattered image data and the single partially unscattered image data; receiving, by the analyzer, a trained scatter correction model; updating, by the analyzer, the trained scatter correction model based on the received plurality of scatter image data and the single partial no-scatter image data to obtain an updated trained scatter correction model; correcting, by the analyzer, at least one of the plurality of scatter image data based on the updated trained scatter correction model; outputting, by the analyzer, at least one corrected scatter image data; the trained scatter correction model is trained using at least a portion of at least one scatter image based on detection of radiation transmitted through an imaged volume of an object and at least one non-scatter image based on detection of radiation transmitted through an imaged volume of the object; the at least one scatter-free image is generated by combining image portions of a plurality of partial scatter-free image data captured with different positions of the aperture plate, the at least one scatter image and the plurality of partial scatter-free images being captured under substantially the same imaging conditions except for the presence of the aperture plate; the trained scatter correction model includes a previously determined deconvolved point spread function estimate; updating the trained scatter correction model locally parameterizing the deconvolved point spread function estimate using measurement points of each image pair including scatter image data of the plurality of scatter image data and the single partial non-scatter image data; and updating the deconvolved point spread function estimate based on the local parameterization.

9. 9. The method of claim 8, wherein the trained scatter correction model includes a deconvolution that includes a scattering edge spread function, the scattering edge spread function further including a point spread function configured to correct for scattering in a scatter image.

10. 1. An imaging system, comprising: a radiation source configured to emit directed radiation towards the object; a radiation detector configured to detect the emitted radiation transmitted through an imaging volume of the object; and an aperture plate including a plurality of apertures arranged in a two-dimensional grid, the aperture plate being positioned between the object and the radiation detector such that when radiation is directed from the radiation source towards the object, unscattered radiation passes through the plurality of apertures and a portion of scattered radiation is blocked by the aperture plate; an analyzer including one or more processors; wherein the analyzer comprises: receiving a plurality of scattered image data, each representing a scattered image of the object based on detection of radiation by the radiation source transmitted through the imaging volume of the object; receiving a single partially scatter-free image data representative of a single partially scatter-free image, the single partially scatter-free image being based on detection of radiation transmitted through the imaging volume of the object when the aperture plate is present; receiving a trained scatter correction model, the trained scatter correction model being a trained neural network model; updating the trained scatter correction model based on the received plurality of scatter image data and the single partial scatter-free image data to obtain an updated trained scatter correction model; correcting at least one of the plurality of scatter image data based on the updated trained scatter correction model; and outputting the at least one corrected scattering image data. the trained scatter correction model is trained using at least a portion of at least one scatter image based on detection of radiation transmitted through an imaged volume of an object and at least one non-scatter image based on detection of radiation transmitted through an imaged volume of the object; the at least one scatter-free image is generated by combining image portions of a plurality of partial scatter-free image data captured with different positions of the aperture plate, the at least one scatter image and the plurality of partial scatter-free images being captured under substantially the same imaging conditions except for the presence of the aperture plate; updating the trained scatter correction model for each image pair including scattered image data of the plurality of scattered image data and the single partial non-scattered image data, performing an interpolation between an output of the trained neural network model for the scattered image data and an output of the trained neural network model for the single partial non-scattered image data; generating the updated trained scatter correction model based on the interpolation.

11. An imaging system, comprising: a radiation source configured to emit directed radiation towards the object; a radiation detector configured to detect the emitted radiation transmitted through an imaging volume of the object; and an aperture plate including a plurality of apertures arranged in a two-dimensional grid, the aperture plate being positioned between the object and the radiation detector such that when radiation is directed from the radiation source towards the object, unscattered radiation passes through the plurality of apertures and a portion of scattered radiation is blocked by the aperture plate; an analyzer including one or more processors; wherein the analyzer comprises: receiving a plurality of scattered image data, each representing a scattered image of the object based on detection of radiation by the radiation source transmitted through the imaging volume of the object; receiving a single partially scatter-free image data representative of a single partially scatter-free image, the single partially scatter-free image being based on detection of radiation transmitted through the imaging volume of the object when the aperture plate is present; receiving a trained scatter correction model; updating the trained scatter correction model based on the received plurality of scatter image data and the single partial scatter-free image data to obtain an updated trained scatter correction model; correcting at least one of the plurality of scatter image data based on the updated trained scatter correction model; and outputting the at least one corrected scattering image data. the trained scatter correction model is trained using at least a portion of at least one scatter image based on detection of radiation transmitted through an imaged volume of an object and at least one non-scatter image based on detection of radiation transmitted through an imaged volume of the object; the at least one scatter-free image is generated by combining image portions of a plurality of partial scatter-free image data captured with different positions of the aperture plate, the at least one scatter image and the plurality of partial scatter-free images being captured under substantially the same imaging conditions except for the presence of the aperture plate; the trained scatter correction model includes a previously determined deconvolved point spread function estimate; updating the trained scatter correction model locally parameterizing the deconvolved point spread function estimate using measurement points of each image pair including scatter image data of the plurality of scatter image data and the single partial non-scatter image data; updating the deconvolved point spread function estimate based on the local parameterization.

12. the trained scatter correction model includes a deconvolution that includes a scatter edge spread function, the scatter edge spread function further including a point spread function configured to correct for scatter in a scatter image. The imaging system according to claim 11.

13. 1. A method for scatter correction of an image of an object, comprising: obtaining data representative of at least one scattering image of the object based on detection by a radiation detector of the imaging system of radiation transmitted through an imaging volume of the object; disposing an aperture plate at a first location between the object and the radiation detector, the aperture plate including a plurality of apertures arranged in a two-dimensional grid, positioned such that when radiation is directed from a radiation source towards the object, unscattered radiation passes through the plurality of apertures and a portion of scattered radiation is blocked by the aperture plate; obtaining, based on detection by the radiation detector of radiation transmitted through the imaging volume of the object when the aperture plate is in the first position, first partial unscattered image data representative of at least one first partial unscattered image; moving the aperture plate to one or more second positions different from the first position, wherein, with respect to the object and aperture plate positioned for imaging, a combination of an area of ​​the aperture when the aperture plate is disposed at the first position and an area of ​​the aperture when the aperture plate is disposed at the one or more second positions is configured to encompass an area through which radiation passes through the imaging volume of the object; obtaining, by the radiation detector, second partially unscattered image data representative of at least one second partially unscattered image based on detection of radiation transmitted through the imaging volume of the object when the aperture plate is in the one or more second positions; receiving the first and second partially unscattered image data by an analyzer including one or more processors; generating at least one non-scattering image by combining, by the analyzer, image portions corresponding to the areas of the plurality of apertures obtained from the first partial non-scattering image data and image portions corresponding to the areas of the plurality of apertures obtained from the second partial non-scattering image data; training, by the analyzer, a scatter correction model using at least a portion of the at least one scatter image and the at least one non-scatter image; outputting, by the analyzer, the trained scatter correction model; the scatter correction model is a neural network model including a scatter correction algorithm, and the method further comprises: applying the scatter correction algorithm to at least a portion of the at least one scatter image to generate a scatter corrected image; determining a difference between the generated scatter corrected image and the at least one scatter free image by a predetermined method; updating the scatter correction algorithm to reduce the difference when the difference is greater than a predetermined difference; when the difference is less than or equal to the predetermined difference, outputting the updated scatter correction model including the updated scatter correction algorithm.

14. 1. An imaging system, comprising: a radiation source configured to emit directed radiation towards the object; a radiation detector configured to detect the emitted radiation transmitted through an imaging volume of the object; and a movable aperture plate positioned between the object and the radiation detector, the movable aperture plate including a plurality of apertures arranged in a two-dimensional grid, the aperture plate being positioned such that when radiation is directed from the radiation source towards the object, unscattered radiation passes through the plurality of apertures and a portion of scattered radiation is blocked by the aperture plate; and an analyzer including one or more processors, the analyzer comprising: receiving data representative of at least one scattering image of the object based on detection of radiation transmitted through the imaging volume of the object; receiving first partially unscattered image data representative of at least one first partially unscattered image of the object based on detection of radiation transmitted through the imaging volume of the object when the aperture plate is in a first position; receiving second partially unscattered image data representative of at least one second partially unscattered image based on detection of radiation transmitted through the imaging volume of the object when the aperture plate is in one or more second positions different from the first position; receiving, configured with respect to the object positioned for imaging and the aperture plate such that a combination of an area of ​​the aperture when the aperture plate is disposed in the first position and an area of ​​the aperture when the aperture plate is disposed in the one or more second positions encompasses an area through which radiation passes through the imaging volume of the object; generating at least one non-scattering image by combining image portions corresponding to the areas of the plurality of apertures obtained from the first partial non-scattering image data and image portions corresponding to the areas of the plurality of apertures obtained from the second partial non-scattering image data; training a scatter correction model using at least a portion of the at least one scatter image and the at least one no-scatter image; outputting, by the analyzer, the trained scatter correction model; the scatter correction model is a neural network model including a scatter correction algorithm, and the analyzer applying the scatter correction algorithm to at least a portion of the at least one scatter image to generate a scatter corrected image; determining a difference between the generated scatter corrected image and the at least one scatter free image by a predetermined method; updating the scatter correction algorithm to reduce the difference when the difference is greater than a predetermined difference; and outputting the updated scatter correction algorithm when the difference is less than or equal to the predetermined difference.

Images