Image rendering method for tomographic image data
Patent Information
- Application Number
- JP2023526955
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2020-11-05
- Filing Date
- 2021-10-28
- Publication Date
- 2024-11-01
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Existing CT colonography methods face challenges in accurately generating endoluminal renderings due to the complex anatomy of the colon, lack of visual correspondence between different rendering types, and obscuration of polyps by fecal residue, requiring manual interpretation and additional display for comprehensive information.
A hybrid image generation method that embeds volume rendering information into 2D slice-by-slice tomographic reconstructions, providing a composite image that spatially correlates volume-rendered and standard tomographic views, enhancing depth and texture information while ensuring polyps are not obscured by contrast agent accumulation.
The method automatically establishes spatial correspondence between rendering types, reduces display clutter, and enhances visibility of anatomical structures by making polyps distinguishable from fecal debris, without increasing screen space requirements.
Smart Images

Figure 00000000_0000_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a method used for generating an image based on tomographic imaging data.
Background Art
[0002] Tomography is a method of imaging by cross-section or sectioning using any kind of penetrating wave. As an example, there is X-ray computed tomography (CT), which is most commonly used for medical purposes. However, this approach is more widely used in a wide range of scientific fields.
[0003] The generation of tomographic images is usually based on a mathematical procedure known as tomographic reconstruction. There are various reconstruction algorithms. Most algorithms are classified into one of two categories: filtered back projection (FBP) and iterative reconstruction (IR). Tomographic reconstruction is applied sequentially to each slice and is based on projection data obtained only for a single slice. This is called slice-by-slice image reconstruction, or more generally, simply tomographic image reconstruction.
[0004] In the most widespread use case of tomographic image reconstruction, only a 2D representation or rendering of a specific slice of the scanned body is provided. Therefore, the application of clinical experience and expertise is required to interpret the image. There is information (such as depth and texture information) that cannot be provided by slice-by-slice tomographic reconstruction.
[0005] To correct this drawback, it is known to additionally apply 3D volume rendering techniques to acquired tomography imaging data. 3D volume rendering techniques are applied to this data using reconstructed pixel values within the reconstructed tomography image data. Volume rendering is an image rendering technique used to display a 2D projection of a 3D image dataset. A 3D image dataset is formed by multiple slice images acquired from a tomography scan, which can be understood as a normal volumetric grid of pixels. Image data from multiple slices is used to generate the 3D representation. One exemplary technique is volume ray casting. In volume ray casting, sampling rays (sampling lines) are conceptually projected through the 3D image dataset toward (and possibly through) the image plane, and the value of each pixel in the image plane is determined based on the values of the set of pixels that intersect with the corresponding ray projected onto it.
[0006] X-ray spectral CT is an imaging modality that extends the functionality of conventional X-ray CT systems by acquiring projection data at multiple X-ray energies. This can be achieved by incorporating detectors that can distinguish between different X-ray energies (such as energy-distinguishing photon count detectors or energy-integrating detectors) or by sequentially changing the X-ray energy spectrum and acquiring corresponding detector data. Spectral X-ray data allows for the identification and quantification of substances contained in scanned objects. This is because different substances may have different X-ray absorption spectra. For example, it is possible to use X-ray energy spectra that are known to be absorbed most or least by the substance of interest.
[0007] One specific field within tomographic imaging is X-ray CT colonography.
[0008] CT colonography is an established test for the early detection and prevention of colorectal cancer. An oral contrast agent is administered to tag fecal residue in the colon.
[0009] CT colonography scans are interpreted using standard transsectional slice views (i.e., tomographic slice reconstructions) and virtual intraluminal rendering (fly-through views). Virtual intraluminal rendering provides a three-dimensional impression intended to resemble optical colonoscopy. Therefore, fluoroscopic projection geometry is used in virtual intraluminal rendering (as opposed to parallel rays in orthographic projection).
[0010] Spectral CT can be used in the field of CT colonography to generate multiple spectral channels (at least two) corresponding to low and high energy levels of projected X-ray energy. These spectral channels can be converted into various representations, including iodine concentration maps. This allows for the visualization of contrast agent distribution within the imaged colon. [Overview of the project] [Problems that the invention aims to solve]
[0011] Existing approaches to CT colonography have several problems.
[0012] First, to generate intraluminal renderings, a precise path along the midline of the colon must be selected. However, because the colon is divided into several components due to strictures and is closely intertwined with the small intestine and stomach, automated path tracing is often erroneous.
[0013] Furthermore, there is no visual correspondence between intraluminal renderings (generated using fluoroscopic geometry) and standard slice-by-slice tomographic reconstructions (using orthographic geometry). Clinicians must apply their experience and knowledge to mentally map the two renderings to each other and interpret the information provided by each. Consequently, information on how one spatially maps to the other is missing and must be supplemented by the user.
[0014] A further problem is that in standard volume rendering, the density of fecal residue is as high as that of the colon wall, making the residue appear opaque. Therefore, polyps may be overlooked if they are submerged and hidden within the fecal residue.
[0015] An improved approach to rendering tomographic image data is useful. [Means for solving the problem]
[0016] The present invention is defined by the claims.
[0017] According to an embodiment of one aspect of the present invention, a computer-aided method is provided for generating images based on a tomographic imaging dataset of a 3D object. This method is A step of acquiring reconstructed tomographic image data for each of multiple slices of a 3D object, wherein the image data for each slice includes the pixel values of the slice, and the reconstructed tomographic image data for the multiple slices forms at least one 3D image dataset. For each of at least a subset of the multiple slices, A selection step comprising selecting a subset of pixels in a slice to be rendered using pixel values obtained by at least one volume rendering of a 3D image dataset, wherein the selection is at least partially based on the pixel values of each pixel in the reconstructed tomographic image data, A step of performing volume rendering of a selected subset of pixels of a slice, based on the application of a volume rendering procedure to a 3D imaging dataset, wherein the plane defined by the slice in the 3D image dataset forms the imaging plane of the volume rendering, The process includes the steps of constructing a composite image rendering of slices based on correcting the pixel values of a selected subset of pixels from the reconstructed tomographic image data of the slices using the pixel values from volume rendering, The above method further includes the step of generating data output representing one or more constructed composite images.
[0018] Embodiments of the present invention are based on embedding volume rendering information into the reconstruction of each 2D slice of tomographic slices. This results in a hybrid image for each of at least several slices. The hybrid image contains visual information from both the volume rendering representation of the anatomical object or region of interest and the standard tomographic reconstruction of the tomographic slice data. This provides the viewer of the image with the advantage that relevant visual information from both types of image renderings is presented within a single image frame, and the spatial correspondence between the volume rendering view and the tomographic reconstruction view is essentially communicated. This spatial correspondence is achieved essentially because the 3D volume rendering is generated using the same 3D image data that constitutes each slice reconstruction and is based on a rendering viewing plane that corresponds to (is identical to) the plane of the tomographic slice. Visual information from both is then provided in a single clear image frame by swapping or substituting (or directly swapping for pixel values modified based on the volume rendering pixels) pixels of the tomographic reconstruction image of the slice with pixels of the 3D volume rendering on the same slice plane.
[0019] Therefore, information about how one rendered view is spatially mapped to another view is automatically provided by composite rendering, and the user does not need to resolve it mentally.
[0020] A further advantage is that this rendering approach saves screen space on viewing devices because it does not require displaying both tomography slice reconstructions and 3D representations simultaneously. Since both can be embedded together in a single viewing plane, display clutter is avoided, eliminating the need for additional displays to view all the information.
[0021] The selection of a subset of pixels can be performed by applying a selection procedure or evaluation, using predefined selection criteria for each pixel. For example, the selection criteria for pixels to be volume-rendered in the final image are based on identifying pixels that spatially correspond to a target anatomical structure (such as the colon wall of a subject). This may also be based on the expected pixel values of the region in the reconstructed image representing the target object (e.g., based on grayscale values, density thresholds). Thus, in the final image, the object of interest is rendered in 3D volume rendering, providing useful depth and texture information, while the rest of the image is rendered using pixel values from flat-field tomography slice reconstructions. For example, in the field of colonography, the target 3D object is the colon or a region thereof.
[0022] To avoid misunderstanding, tomography image reconstruction data refers to image data that has been individually reconstructed for each slice based on the reconstruction applied to the scan data corresponding only to that slice.
[0023] The volume rendering of a given slice is based on pixel values from multiple slices within a 3D image dataset. In particular, in at least one set of embodiments, it may also be based on projecting sampling rays through the 3D image dataset, where each pixel value in a given slice is determined based on the pixel values that intersect or encounter along the path of the corresponding ray passing through each pixel in the slice plane.
[0024] The reconstructed tomographic image data is data obtained as a result of applying image reconstruction to raw tomographic image scan data (projection data). This data includes at least one pixel value for each pixel of each of a plurality of slices of the scanned object.
[0025] In some embodiments, the reconstructed tomographic image data is X-ray CT image data. Here, in some embodiments, the pixel values of the reconstructed slices correspond to Hounsfield density values.
[0026] In some cases, the reconstructed tomographic image data includes a plurality of reconstructed images for each slice. For example, if the tomographic image data is spectral CT data, there are a plurality of data sets for each slice, each corresponding to a different spectral channel of data collection. Here, each pixel within a slice has a plurality of pixel values.
[0027] In some embodiments, volume rendering is generated only for a selected subset of the pixels of each plane.
[0028] Normally, since volume rendering is generated for each pixel, it is straightforward to generate pixel values only for a selected subset of the pixels of the volume-rendered image. For example, in a ray-tracing-based volume rendering technique, the corresponding sampling rays are projected through the 3D volume data from a starting point and / or through each pixel of the imaging plane. Thus, only the rays corresponding to the selected subset of pixels are projected.
[0029] The advantage of this approach is that processing resources and processing time are saved compared to generating a full volume rendering of the entire slice plane within the 3D image data set.
[0030] However, generating a volume rendering of the entire slice plane is also an option.
[0031] According to one or more embodiments, selecting a subset of pixels in a slice to be rendered using pixel values obtained by volume rendering is performed on a pixel-by-pixel basis. For example, this step includes iterating or indexing through the set of pixels that make up each slice and, for each pixel, deciding whether to render using volume rendering or tomographic reconstruction of the slice.
[0032] According to one or more embodiments, the volume rendering procedure is a volume raycasting method in which sampling rays (sampling lines) are cast (projected) orthogonally across a 3D dataset to the plane of each slice.
[0033] In other words, volume rendering provides an orthographic view rather than a perspective view. This ensures that the image data from the volume rendering is directionally aligned with the plane of each slice. That is, the viewpoint from which the volume rendering is generated corresponds to the same viewpoint as the tomography slice reconstruction; that is, it is viewed orthogonally toward the plane of the slice.
[0034] Raycasting is one-view raycasting; that is, the ray is cast from only one direction. Alternatively, the ray may be cast from multiple directions.
[0035] As mentioned above, the selection of a subset of pixels is based on identifying pixels that spatially correspond to the target anatomical structure.
[0036] This selection is based on one or more selection criteria applied to each pixel. The selection criteria are set to identify pixels corresponding to areas occupied by predefined anatomical structures.
[0037] This is based, for example, on the use of a simple threshold for pixel gray values. For instance, it relies on prior knowledge that tissues belonging to a particular anatomical object will have higher or lower gray values than surrounding tissues. A further option is to apply one or more image analysis algorithms to identify the boundaries of a target object within a given slice. These include segmentation algorithms and shape matching algorithms, among others.
[0038] One or more selection criteria used to select pixels may include predefined thresholds for the pixel value of each pixel in tomographic slice reconstruction.
[0039] The threshold may be the minimum or maximum threshold for pixel values so that the corresponding pixels are selected for rendering using the volume rendering process.
[0040] According to one or more embodiments, tomography imaging data is X-ray computed tomography imaging data.
[0041] Acquiring reconstructed image data involves receiving input tomographic projection data and performing tomographic image reconstruction on each slice in sequence.
[0042] In this embodiment, the method includes the step of generating tomographic reconstructions of each slice.
[0043] According to one or more embodiments, the tomographic image data is spectral CT image data, and the reconstructed tomographic image data includes multiple different image reconstructions for each slice, forming multiple different 3D image datasets, each of which is based on CT projection data corresponding to a different spectral channel.
[0044] Spectral CT is an extension of standard CT imaging in which X-ray energies are projected through an object being imaged at multiple X-ray energies. An X-ray detector acquires projection data measurements for each energy simultaneously or sequentially. Projection data from different X-ray energies are called projection data for different spectral channels.
[0045] According to one or more embodiments, the imaged volume is an anatomical region of the subject containing the administered contrast agent. The 3D imaging dataset for one of the spectral channels may include pixel values indicating the concentration of the contrast agent at the pixel location.
[0046] Volume rendering for one of the spectral channels can vary the transparency of the rendered pixels based on the pixel values of the tomographic image data for the spectral channel representing the contrast agent. Here, the transparency of a pixel is determined based on the concentration of the contrast agent in the corresponding spatial region within the 3D image dataset. For example, if the contrast agent concentration exceeds a certain threshold, the pixel may be made transparent or treated as transparent. This ensures that certain structures, such as polyps, are not overlooked even if their visibility is obscured by the accumulation of contrast agent (e.g., fecal residue accumulated in the colon).
[0047] In some embodiments, volume rendering is a raycasting method. The transparency of a given pixel when performing volume rendering is determined based on the pixel values accumulated along sampling rays that intersect with pixels on a slice plane in a 3D image dataset for the spectral channels representing the contrast agent described above. In some embodiments, this means that the transfer function used to determine the pixel values of the volume-rendered image is configured to treat pixels exceeding a certain value encountered along the ray path (i.e., indicating a high concentration of the contrast agent) as transparent or to give them a low or zero weighting. This results in rendered pixels in the volume-rendered image having pixel values that depend only on the density of the tissue, and not on the concentration of the contrast agent.
[0048] According to one or more embodiments, the method further includes the step of receiving a user input signal. The method further includes the step of selectively generating a first display output representing tomographic reconstruction of one or more slices, or a second display output representing composite reconstruction of one or more slices. The selection of the first or second display output depends on the user input.
[0049] Here, the data output is the display output. The display output can switch between displaying the tomographic reconstruction of a given slice and the composite rendering of the slice. Optionally, displaying volume rendering is also a further option that can be selected based on user input.
[0050] The above method includes the step of selectively switching between a first and a second display output, which is triggered by a user input signal.
[0051] Here, a function is implemented that allows the user to easily switch between different rendered images of a slice, for example, by pressing a single key or button (such as the spacebar) on a user input device. Pressing this single user control unit sends the associated user input signal.
[0052] Embodiments according to further aspects of the present invention provide a computer program product including computer program code. The computer program code is executable on a processor. The code causes the processor to perform a method according to any embodiment or model described above or below, or according to any claim of this application.
[0053] Embodiments according to further aspects of the present invention provide a processing apparatus for use in generating images based on a tomographic imaging dataset of a 3D object. This processing apparatus, The method involves obtaining reconstructed tomographic image data for each of multiple slices of a 3D object, wherein the reconstructed tomographic image data for each slice includes the pixel values of the slice, and the reconstructed tomographic image data for multiple slices forms at least one 3D image dataset. For each of at least a subset of the multiple slices, The selection involves selecting a subset of pixels in a slice to be rendered using pixel values obtained by at least one volume rendering of a 3D image dataset, wherein the selection is at least partially based on the pixel values of each pixel in the reconstructed tomographic image data. Based on the application of a volume rendering procedure to a 3D imaging dataset, the procedure involves performing volume rendering on at least a selected subset of pixels of a slice, such that the plane defined by the slice in the 3D image dataset forms the imaging plane of the volume rendering. This involves constructing a composite image rendering of the slices based on modifying the pixel values of a selected subset of the slice pixels from the reconstructed tomographic image data using the pixel values from the volume rendering.
[0054] Embodiments according to further aspects of the present invention provide a system comprising a processing apparatus according to any embodiment or model described above or below, or according to any claim of this application, and a display unit operably coupled to the processing apparatus for displaying one or more composite images.
[0055] These and other aspects of the present invention will become apparent from the embodiments described below and will be explained with reference to those embodiments. [Brief explanation of the drawing]
[0056] To gain a deeper understanding of the present invention and to more clearly illustrate how it is carried out, refer to the attached schematic drawings as just one example.
[0057] [Figure 1] Figure 1 outlines the steps of an exemplary method according to one or more embodiments. [Figure 2] Figure 2 schematically illustrates an exemplary workflow in one or more embodiments. [Figure 3] Figure 3 shows exemplary image rendering results of the application of the method according to one or more embodiments. [Figure 4] Figure 4 shows exemplary image rendering results of the application of the method according to one or more embodiments. [Figure 5] Figure 5 shows exemplary image rendering results of the application of the method according to one or more embodiments. [Figure 6] Figure 6 is a block diagram showing components of an exemplary system according to one aspect of the present invention. [Figure 7] Figure 7 shows the components of an exemplary CT imaging system. [Modes for carrying out the invention]
[0058] The present invention will be described with reference to the figures.
[0059] The detailed descriptions and specific examples illustrate exemplary embodiments of the apparatus, system, and method, but should be understood to be for illustrative purposes only and not intended to limit the scope of the invention. These and other features, aspects, and advantages of the apparatus, system, and method of the present invention will be better understood from the following description, the appended claims, and the appended drawings. It should be understood that the figures are schematic diagrams and are not drawn to scale. It should also be understood that the same reference numerals are used throughout the figures to indicate the same or similar parts.
[0060] The present invention provides a method for generating image representations of body slices based on tomographic imaging data of the body. This method involves processing reconstructed tomographic image slices to selectively embed image information from at least one 3D volume rendering performed on a 3D tomographic image dataset into each slice. This is done by a selection process, which determines, based on a predefined set of selection criteria, whether for each pixel in each reconstructed tomographic slice, the pixel value should be replaced with a new modified pixel value determined based on at least one volume rendering. This may involve simply swapping the pixel value with the value of the corresponding pixel value in the volume rendering of the corresponding slice plane, or it may involve a more complex process, such as blending the two values or adjusting the transparency of the pixel value based on at least one volume rendering.
[0061] Figure 1 shows a block diagram illustrating the basic steps of an exemplary computer implementation method 10 in one or more embodiments. This method is for generating images based on a tomographic imaging dataset of a 3D object.
[0062] Method 10 includes obtaining reconstructed tomographic image data for each of several slices of a 3D object (12). The image data for each slice includes the pixel values of the slice, and the image data for several slices forms at least one 3D imaging dataset.
[0063] Method 10 further applies to each of at least a subset of the multiple slices: (14) Selecting a subset of pixels of a slice to render using pixel values obtained by at least one volume rendering representation of the slice plane in a 3D image dataset, wherein the selection is at least partially based on the pixel values of each pixel in the tomographic reconstruction. Based on the application of volume rendering procedures to a 3D imaging dataset, at least volume rendering of a selected subset of pixels on a slice plane (16), Based on modifying the pixel values of a selected subset of pixels in the tomographic reconstruction of the slice using the pixel values from the volume rendering, construct a composite image rendering of the slice (18), Includes.
[0064] Method 10 further includes generating a data output (20) representing one or more constructed composite images. In some embodiments, this may be a display output or a data representation of the constructed images for recording to a local or remote data store. The generated data output may be combined with the local or remote data store for later use. Furthermore, or alternatively, the data output may be transmitted to a display device for display to a user.
[0065] Acquiring tomographic image data involves receiving input tomographic projection data and performing tomographic image reconstruction on each slice sequentially using a filter-corrected back projection algorithm, etc. Another option of the above method involves receiving reconstructed tomographic image data, which has already been reconstructed externally.
[0066] Two or more 3D image datasets are acquired. For example, tomography image data is spectral CT image data, which includes multiple reconstructions of each slice based on different spectral channels of the acquired projection data.
[0067] The process of selecting a subset of pixels to modify pixel values based on 3D rendering is based on the application of one or more selection criteria. At least one of the selection criteria is related to the pixel value of each pixel in the tomographic reconstruction of the relevant slice. For example, a threshold for pixel values in the tomographic reconstruction is used as at least one of the criteria. The selection criteria may be predefined and stored, for example, in a local or remote data store. In some embodiments, the selection criteria are user-configurable. Several different image rendering modes and settings are pre-stored, each associated with a different set of selection criteria for producing rendered images with specific properties and characteristics. These different modes are user-selectable.
[0068] As an example, in a set of advantageous embodiments, the pixel selection criteria are based on identifying pixels that spatially correspond to a target anatomical structure. This may also be based on the expected pixel values of the region of the reconstructed image representing the target object (e.g., based on thresholds of gray value, density, or other pixel values). In this way, in the final image, the object of interest is rendered in 3D volume rendering, thus providing useful depth and texture information, while the rest of the image is rendered using pixel values from flat-field tomography slice reconstructions. As an example, the target anatomical structure is the wall of the subject's colon.
[0069] A wide range of different volume rendering methods can be used according to embodiments of the present invention. Volume rendering is applied to at least one 3D image dataset formed by reconstructed tomographic image data.
[0070] In a preferred embodiment, a direct volume rendering method is used.
[0071] One preferred embodiment is volumetric raycasting or volumetric ray tracing, which is an example of direct volumetric rendering. In volumetric raycasting, a corresponding sampled ray is generated for each desired pixel of the rendered image. An observer viewpoint is defined. Each ray starts from the center of the viewpoint projection and passes through each image pixel on the defined image plane. The rays are sampled at regular intervals along the ray path, and a transfer function is applied to the set of sampled pixel values to derive the pixel value for each pixel on the imaging plane to which the ray corresponds. As an example, the pixel value may be a gray value, an RGB value, an opacity value, or a value representing any other visual or graphic characteristic or property.
[0072] In the context of this invention, a plane defined by a slice in a 3D image dataset is used as a conceptual image plane in the volume rendering to generate a volume rendering of a particular slice. The transfer function of the volume rendering can be configured as needed to generate a desired visual representation of a 3D object imaged with the imaging data.
[0073] For example, the transfer function applies different weights to pixels encountered along the path of each ray, depending on the pixel values in the 3D image dataset. The weighting may vary non-linearly with respect to the pixel values. For instance, there may be a threshold, above or below which very low or zero weightings are applied. This allows, for example, selective rendering of only specific tissue types in the final volumetric rendering image.
[0074] As an example, volume rendering is configured to represent a view of a specific anatomical object within a 3D image dataset. This object is characterized by pixel values that are, on average, higher or lower than the pixel values of the surrounding area. For example, the material of the object is denser than the surrounding area. This is the case, for example, for a solid object imaged in air. It is also the case for anatomical objects such as organs, or other anatomical structures within the tissues surrounding the body. A transfer function defines a minimum or maximum threshold. This ensures that only pixels below the minimum threshold or above the maximum threshold are considered for the rendered pixel values. Those that do not meet the thresholds are ignored. This results in a volume rendering in the final rendered image that shows only the target anatomical object or other objects in the image data. In one advantageous embodiment, this can be used to render, for example, the wall of an imaged colon.
[0075] In a preferred embodiment using a volumetric ray tracing approach, sampling rays are cast orthogonally through the 3D dataset to the plane of each slice. This means that volumetric rendering provides an orthographic view rather than a perspective view, which facilitates the interpretation of the combined composite image. In particular, per-slice rendering and volumetric rendering blend seamlessly with each other, as they depict objects of the same size and position.
[0076] In some embodiments, tomographic image data includes multiple image reconstructions for each slice, each reconstruction being based on a different subset of tomographic projection data. In one set of embodiments, each image dataset is data acquired using different spectral channels (e.g., different spectral CT channels) of the tomographic scanner. In some embodiments, a volume rendering of a given slice considers two or more of these multiple reconstructions.
[0077] As one specific example, one type of image reconstruction involves reconstructing from CT projection data acquired using spectral channels sensitive to the concentration of contrast agent administered to the imaged subject (such as the iodine content of the contrast agent). Another type of image reconstruction is sensitive to the general biological tissues that make up the subject's body, and therefore involves reconstructing from different spectral channels that represent the density of tissues within the body.
[0078] For each pixel sampled along each cast ray, volumetric raycasting is performed, where the pixel values from both image reconstructions are considered in the transfer function. For example, pixel values from contrast-sensitive channels are used to define the transparency applied to the pixel along the sampled ray, while pixel values from other spectral channels are used to define the gray value of the RGB values for each pixel. For example, pixels with pixel values indicating a high concentration of contrast agent are made transparent or semi-transparent. For example, a transparency threshold is set.
[0079] For example, at each sampling point of a direct volume rendering sampling ray, a slight increase in iodine concentration is accumulated. The amount of iodine concentration represents tagging by the contrast agent. This can be used to control the transparency of the ray, making the tagged areas effectively semi-transparent in volume rendering. Optionally, tagged areas can also be color-coded ("stained"), giving the observer a visual cue about where the tagged material is located.
[0080] This is advantageous in preventing areas where contrast agent accumulates from obscuring anatomically or physiologically relevant parts of the imaged object in the final rendered image. For example, in the context of colonography, this can be used to prevent accumulated fecal residue (containing contrast agent) in the colon from obscuring anatomical structures within the colon, such as polyps.
[0081] The above embodiments assume the use of a direct volume rendering method, but this is not mandatory. In further embodiments, other classes of volume rendering techniques can be used. For example, surface rendering of a triangular mesh to map the surface of an anatomical object of interest can be used. In this embodiment, as a preliminary step, the surface mesh must be determined in advance, for example, using a segmentation technique.
[0082] In some embodiments, the volume rendering results of each slice can be reused to generate the volume renderings of subsequent slices. This reduces the processing resources required to generate each subsequent 3D rendering slice.
[0083] Figure 2 shows a schematic overview of the workflow of an exemplary method, which is briefly explained below.
[0084] During operation, a set of reconstructed tomographic image data is acquired (Figure 2(a)). This includes at least one 3D image dataset 24 containing the corresponding tomographic reconstructions 22 of each of the multiple slices of the scanned body. As mentioned above, in some cases, multiple 3D image datasets are acquired that correspond to different spectral channels of the same volumetric region, for example, in a spectral CT imaging system.
[0085] The method further includes selecting a subset of pixels 25 in each slice to modify the pixel values based on the pixel values of a 3D volume rendering of the slice plane. This is schematically shown in Figure 2(b). For example, the selection procedure is performed pixel by pixel, based on iteration or indexing, or by stepping through each pixel one by one and deciding whether to re-render the pixel (i.e., modify the pixel value) based on the volume rendering. The decision for each pixel is based on a predefined selection criterion that is at least partially related to the pixel value of each pixel in the reconstructed tomographic slice image 22. For example, this is a simple threshold. Once a pixel is determined, this decision is recorded in local memory or a cache, and the process proceeds to the next pixel.
[0086] Once a subset of pixels to be modified using volume rendering is selected (Figure 2(b)), the method further includes generating a volume rendering from a 3D dataset of at least the selected subpixels of the slice plane (Figure 2(c)). For a given slice, this means generating a volume rendering in which the plane defined by the slice in the 3D image dataset is used as the conceptual image plane in the volume rendering. For example, in a raycasting method, each ray is cast from a conceptual viewpoint, through each relevant pixel of the image plane, and through the 3D image dataset. This is explained in detail above. It is sufficient to generate a volume rendering of only a subset of pixels. For example, in a volume ray tracing method, this means that it is sufficient to cast rays that pass through a selected subset of pixels in the slice plane. In other embodiments, a volume rendering of the entire slice plane is generated. If volume rendering is performed on the entire slice plane, the volume rendering of each slice can be performed before selecting a subset of pixels. Alternatively, the volume rendering may be generated after selecting a subset of pixels.
[0087] When volume rendering is performed, a composite image of each slice is generated based on modifying the pixel values of a selected subset of pixels in the tomography slice reconstruction based on the pixel values obtained for the corresponding pixels in the volume rendering (Figure 2(d)). In one example, the pixel values from the tomography reconstruction image are simply replaced with the pixel values of the corresponding pixels in the volume rendering image. In a further example, the pixel values of selected pixels are blended with the pixel values from the volume rendering. In yet another example, the appearance of pixels in the slice reconstruction is modified in a manner defined according to a modification function. The modification function depends on the pixel values of the corresponding pixels in the volume rendering. For example, the transparency, opacity, color, saturation, intensity, or other graphic properties of a pixel are adjusted based on the pixel values in the volume rendering.
[0088] The steps in Figures 2(b) to 2(d) are repeated for at least each of the subsets of slices in the 3D image dataset to generate a composite image for each slice.
[0089] In some embodiments, two or more composite images are generated, such as when there are two or more 3D image datasets.
[0090] Figure 3 shows exemplary results of applying Method 10 according to an embodiment to a 3D image dataset of a region including the colon of a subject. In this embodiment, the images are obtained from CT scan data of the above region. The leftmost image 22 shows tomographic reconstruction of a specific slice of the above region. This represents a “standard” slice-by-slice view provided by a standard CT scan of the colon. It can be seen that structural information regarding the morphology and texture of the colon is not visible. Only the cross-sectional shape and contour of the colon are visible. The middle column (images 28a and 28b) shows examples of composite images generated according to embodiments of the present invention. It can be seen that the region of the colon wall is rendered based on volume rendering, and as a result, the morphology of the colon wall is more clearly visible as the volume rendering provides depth information. The volume rendering in this embodiment is generated using a ray tracing method. The images in the top column correspond to ray tracing where rays are cast in the positive z direction (+1) through the 3D image dataset, and the images in the bottom column correspond to ray tracing where rays are cast in the negative z direction (-1). The z direction means the direction perpendicular to the plane of the slice. Images 28a and 28b correspond to composite images formed from volume rendering performed on non-spectral CT data. Images 30a and 30b correspond to composite images formed from volume rendering performed on spectral CT data that includes at least one spectral channel representing an iodine concentration map (i.e., sensitive to iodine contrast agent). In images 30a and 30b, areas with high iodine (contrast agent) concentrations are rendered semi-transparently (optionally color-coded). As a result, areas where the contrast agent has accumulated (indicated by label 32) are semi-transparent in images 30a and 30b, but large areas are obscured in images 28a and 28b.
[0091] Figure 4 shows further exemplary results of applying Method 10 according to the embodiment to a 3D image dataset of a region including the colon of a subject. The setting details of this embodiment are the same as in the embodiment of Figure 3, in that Image 22 represents a tomographic reconstruction of a slice, Images 28a and 28b represent composite images formed using non-spectral CT data, and Images 30a and 30b represent composite images formed using spectral CT data with an iodine-containing contrast agent rendered semi-transparently. In this embodiment, the imaged colon has a polyp 34 near the center. The three-dimensional cues of volume rendering improve the visual distinction of the polyp from the haustrale folds of the colon. In a standard slice-by-slice image 22, the polyp is barely discernible.
[0092] Figure 5 shows further exemplary results of applying Method 10 according to the embodiment to a 3D image dataset of a region including the colon of a subject. The details of the setting in this embodiment are the same as in the embodiments of Figures 3 and 4, in that Image 22 represents a tomographic reconstruction of a slice, Image 28 represents a composite image formed using non-spectral CT data, and Image 30 represents a composite image formed using spectral CT data in which an iodine-containing contrast agent is rendered semi-transparently. In this embodiment, there is also a polyp 34 in the imaged colon.
[0093] The polyp is not visible in standard tomographic slice reconstruction 22. The polyp is visible in the composite images 28 and 30, respectively, because volume rendering provides additional depth information.
[0094] Further aspects of the present invention provide an apparatus for carrying out a method according to any of the above-described embodiments or any of the claims of this application.
[0095] Further aspects of the present invention also provide a system. Figure 6 shows one exemplary system 40 according to one or more embodiments in block diagram form.
[0096] The system 40 includes the above-described processing unit 42 and a display unit 44 which is operably coupled to the processing unit. In this embodiment, the generated data output 20 is a display output, and the processing unit 42 transmits the display output to the display unit 44 to display the generated composite image on the display screen of the display unit.
[0097] It is beneficial for the user to be able to switch between displaying a composite image constructed from a given slice and the reconstruction of the original slice-by-slice tomography (and optionally, a full-volume rendering of the generated slices). To facilitate this, a user interface 46 is provided. In the system shown in Figure 6, this is operably coupled to a processing unit 42, which receives one or more user input signals from the user interface.
[0098] For example, the method performed by the processing unit 42 further includes selectively generating a first display output representing tomographic reconstruction of one or more slices, or a second display output representing composite reconstruction of one or more slices, the selection of the first or second display output depending on user input.
[0099] For an optimal viewing experience, it is preferable to be able to instantly switch between various rendering modes, for example, between a standard tomography reconstruction view and one or more composite images. To facilitate this, the above method involves selectively switching the display output between a first and a second display output, and this switching is triggered by a user input signal. For example, the generation of a user input signal can be triggered by pressing a single key, button, or other user control (such as the spacebar) on the UI, thereby allowing the user to quickly and easily switch between different rendering views.
[0100] Other optional interactive settings configurable from the UI include, for example, changing the translucency spectral threshold to adjust to lower density levels for tagging, or changing one or more 3D image datasets used to generate the composite image. For example, you can switch between a 3D dataset representing conventional Hånsfield values and virtual mono-energy Hånsfield unit values.
[0101] The embodiments of the present invention described above offer numerous advantages compared to the use of standard tomographic slice reconstruction and standard volume rendering, such as intraluminal rendering. In particular, by embedding volume rendering information into the tomographic reconstruction image, the image information provided by both types of images can be presented in a single image frame. This means that the spatial correspondence between the two is directly communicated to the observer. It also means that there is no need to change the workflow compared to standard tomographic imaging techniques. Another advantage is that there is no increased screen space requirement compared to standard slice-by-slice ("transsectional") reading. This is particularly beneficial for handheld devices such as tablets and mobile phones. When the volume rendering is an orthographic volume rendering, all size dimensions are directly mapped to each other, enabling an intuitive visual correspondence with standard slice-by-slice viewing. In particular, in the context of colonography applications, the in-slice embedding of orthographic volume renderings allows for an intuitive visual distinction between normal colonic bulging and possible polyps in the colon. As mentioned above, when using spectral data, volume rendering allows for a "visualization" of contrast-tagged stool residue even at low concentration levels by utilizing spectral information.
[0102] In a favorable set of embodiments, the acquired tomographic image data is X-ray CT image data. To better understand the nature and origin of the tomographic image data, an exemplary X-ray CT imaging system is briefly described. In one aspect of the present invention, the X-ray CT imaging system is provided in combination with the processing apparatus 42 described above and performs method 10 to generate one or more composite images. However, this is not essential, and the following description is simply used to help understand the embodiments of the present invention described earlier.
[0103] Figure 7 shows an imaging system 100, such as a computed tomography (CT) scanner.
[0104] The imaging system 100 generally includes a fixed gantry 102 and a rotating gantry 104. The rotating gantry 104 is rotatably supported by the fixed gantry 102 and rotates around the inspection area around a longitudinal axis or the z-axis.
[0105] The patient support 120, such as a couch, supports an object or a subject, such as a human patient, within the examination area. The support 120 moves the object or subject for loading, scanning, and / or unloading.
[0106] A radiation source 108, such as an X-ray tube, is rotatably supported by a rotating gantry 104. The radiation source 108 rotates with the rotating gantry 104, emitting radiation that traverses the inspection area 106.
[0107] A radiation-sensitive detector array 110 defines the range of the angular arc on the opposite side of the inspection area 106 from the radiation source 108. The detector array 110 includes one or more rows of detectors extending along the z-axis, which detect radiation traversing the inspection area 106 and generate projection data indicating it.
[0108] A general-purpose computing system or computer functions as an operator console 112 and includes input devices 114 such as a mouse and keyboard, and output devices 116 such as a display monitor and film. The console 112 allows the operator to control the operation of the system 100.
[0109] The reconstruction device 118 processes the projection data and reconstructs the volumetric image data. This data is displayed via one or more display monitors of the output device 116.
[0110] The reconstruction device 118 may employ filtered back projection (FBP) reconstruction, low-noise reconstruction algorithms (such as iterative reconstruction) (for the image domain and / or projection domain), and / or other algorithms. It is understood that the reconstruction device 118 may be implemented via a microprocessor that executes computer-readable instructions encoded or embedded in a computer-readable storage medium such as physical memory or other non-temporary media. Furthermore, the microprocessor may also execute computer-readable instructions carried by carrier waves, signals, and other temporary (or non-temporary) media.
[0111] The embodiments of the present invention described above employ a processing apparatus. The processing apparatus generally comprises a single processor or multiple processors. The processing apparatus may be installed within a single containing device, structure, or unit, or it may be distributed among multiple different devices, structures, or units. Therefore, reference to a processing apparatus adapted or configured to perform a particular step or task corresponds to that step or task being performed by any one or more of the multiple processing components, either individually or in combination. Those skilled in the art will understand how such a distributed processing apparatus can be implemented.
[0112] One or more processors in a processing unit can be implemented in various ways using software and hardware to perform a variety of required functions. Typically, a processor uses one or more microprocessors programmed using software (e.g., microcode) to perform the required functions. A processor can be implemented as a combination of dedicated hardware for some functions and one or more programmed microprocessors and associated circuits for other functions.
[0113] Examples of circuits that may be used in various embodiments of this disclosure include, but are not limited to, conventional microprocessors, application-specific integrated circuits (ASICs), and field-programmable gate arrays (FPGAs).
[0114] In various implementations, a processor may be associated with one or more storage media, such as volatile and non-volatile computer memory, including RAM, PROM, EPROM, and EEPROM®. The storage media may be encoded with one or more programs that, when executed on one or more processors and / or controllers, perform the necessary functions. The various storage media may be fixed within a processor or controller, or they may be transportable so that one or more programs stored therein can be loaded into the processor.
[0115] Modifications of the disclosed embodiments can be understood and implemented by those skilled in the art in carrying out the claimed invention, based on a review of the drawings, disclosures, and appended claims. In the claims, the word “including” does not exclude other elements or steps, and singular elements do not exclude plural elements.
[0116] A single processor or other unit may perform the functions of some of the items described in the claims.
[0117] The mere fact that certain means are described in mutually different dependent claims does not mean that combinations of these means cannot be used advantageously.
[0118] Computer programs may be stored and distributed on any suitable medium, such as optical storage media or solid-state media, supplied together with or as part of other hardware, but they may also be distributed in other forms, such as via the Internet or other wired or wireless communication systems.
[0119] Please note that when the term "adapted to..." is used in the claims or description, it is intended to be equivalent to the term "configured to...".
[0120] Any reference numeral in the claims should not be construed as limiting the scope.
Claims
1. 1. A computer-implemented method for generating an image based on a tomographic imaging dataset of a 3D object, the computer-implemented method comprising: acquiring reconstructed tomographic image data for each of a plurality of slices of the 3D object, the reconstructed tomographic image data for each slice including pixel values for the slice, the reconstructed tomographic image data for the plurality of slices forming at least one 3D image data set; For each of at least a subset of the plurality of slices, selecting a subset of pixels of the slice for rendering using pixel values acquired by at least one volume rendering of the 3D image dataset, the selection being based at least in part on the pixel values of each pixel of the reconstructed tomographic image data; performing a volume rendering of at least the subset of selected pixels of the slice based on application of a volume rendering procedure to the 3D image dataset, wherein a plane defined by the slice in the 3D image dataset forms an imaging plane of the volume rendering; constructing a composite image rendering of the slice based on modifying the pixel values of the subset of selected pixels of the slice from the reconstructed tomographic image data using the pixel values from the volume rendering; Including, The computer-implemented method further includes generating a data output representative of one or more constructed composite images.
2. The computer-implemented method of claim 1 , wherein the volume rendering is generated for only the selected subset of pixels in each slice.
3. 3. The computer-implemented method of claim 1 or 2, wherein the volume rendering procedure is a volume ray casting method, in which sampling rays are cast through the 3D image dataset orthogonal to the plane of each slice.
4. The computer-implemented method of claim 1 , wherein the selection of the subset of pixels is based on identifying pixels that spatially correspond to a target anatomical structure.
5. The computer-implemented method of claim 1 , wherein the selection of the subset of pixels is based on a predefined threshold of the pixel value of each pixel of the tomographic slice reconstruction.
6. The computer-implemented method of claim 5, wherein the tomographic imaging data is X-ray computed tomography (CT) imaging data.
7. 7. The computer-implemented method of claim 1, wherein obtaining the reconstructed tomographic image data comprises receiving input tomographic projection data and performing tomographic image reconstruction on each slice in turn.
8. 8. The computer-implemented method of claim 1, wherein the reconstructed tomographic image data is spectral CT image data, the reconstructed tomographic image data including a plurality of different image reconstructions for each slice, forming a plurality of different 3D image data sets, each of the plurality of different image reconstructions being based on CT projection data corresponding to a different spectral channel.
9. 9. The computer-implemented method of claim 8, wherein the imaged volume represents an anatomical region of a subject, the anatomical region including an administered contrast agent, and the 3D image data set for one of the spectral channels includes pixel values indicative of a concentration of the contrast agent at a pixel location.
10. The computer-implemented method further includes receiving a user input signal; The computer-implemented method further comprises: a first display output representing a tomographic reconstruction of one or more slices; or a second display output representing the composite image rendering of one or more slices; The computer-implemented method of claim 1 , wherein the selection of the first or second display output depends on a user input.
11. The computer-implemented method of claim 10 , further comprising selectively switching a display output between the first and second display outputs, the switching being triggered by the user input signal.
12. A computer program comprising computer program code, the computer program code being executable on a processor, the computer program code causing the processor to perform a computer implemented method according to any one of claims 1 to 11.
13. 1. A processing device for use in generating an image based on a tomographic imaging dataset of a 3D object, the processing device comprising: acquiring reconstructed tomographic image data for each of a plurality of slices of the 3D object, the reconstructed tomographic image data for each slice including pixel values for the slice, the reconstructed tomographic image data for the plurality of slices forming at least one 3D image data set; For each of at least a subset of the plurality of slices, selecting a subset of pixels of the slice for rendering using pixel values acquired by at least one volume rendering of the 3D image dataset, the selection being based at least in part on the pixel values of each pixel of the reconstructed tomographic image data; performing a volume rendering of at least the selected subset of pixels of the slice based on application of a volume rendering procedure to the 3D image dataset, wherein a plane defined by the slice in the 3D image dataset forms an imaging plane of the volume rendering; constructing a composite image rendering of the slice based on modifying the pixel values of a selected subset of pixels of the slice from the reconstructed tomographic image data using the pixel values from the volume rendering; Run The processor is further configured to generate a data output representative of one or more constructed composite images.
14. A processing device according to claim 13; a display unit operatively coupled to the processing unit for displaying one or more of the composite images; Including, the system.
