Suppression of landmark elements in medical images
Patent Information
- Application Number
- JP2023529061
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2021-03-01
- Filing Date
- 2021-11-15
- Publication Date
- 2024-11-21
- Estimated Expiration
- 2041-11-15
AI Technical Summary
The presence of landmark elements in medical images, such as contrast agent markers, often obscures structural elements and causes bright artifacts, hindering clinical evaluation.
A computer-implemented method that generates a suppressed image with reduced apparent brightness of landmark elements by combining it with the original medical image, allowing partial suppression without complete removal, facilitating advanced image processing and maintaining valuable information.
Enhances visibility of non-landmark elements while preserving landmark element information, reducing artifacts, and enabling more accurate clinical analysis.
Smart Images

Figure 00000000_0000_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of medical imaging, and in particular to the suppression of labeling elements in medical images.
Background Art
[0002] The use of medical images for evaluating the condition of a patient / individual is increasing. In medical treatment, it is reasonably common to use a contrast agent to facilitate the identification of non-structural objects (such as feces, some fluid / blood, or a bolus) in the medical image of the patient / individual. The part of the image representing the non-structural object resulting from the use of the contrast agent is generally referred to as the "labeling element" or "contrast agent labeling element" of the image, and usually has a higher brightness than the part of the image representing other elements of the body (such as bones, surrounding tissues, etc.).
[0003] If there is one or more labeling elements in the image, it may interfere with the clinician viewing the medical image, such as (partially) hiding the structural elements near the labeling element or generating bright artifacts in the medical image. There are known suppression concepts that virtually remove the labeling elements from the medical image to make it easier for the clinician to evaluate the image.
[0004] As an example, during colonoscopy in a computed tomography (CT) imaging process, the fecal residues in the colon are labeled by oral administration of a contrast agent. Using this contrast agent labeling, the visibility of the colon wall and possible polyps can be enhanced, and the labeled fecal residues can be digitally suppressed for observation so that they are not hidden by the fecal residues.
Summary of the Invention
Problems to be Solved by the Invention
[0005] There is still a desire to improve the suppression of labeling elements in medical images.
Means for Solving the Problems
[0006] The present invention is defined by the claims.
[0007] According to one embodiment of the present invention, a computer-aided method is provided for performing a dilution (suppression) of a labeling element in a medical image having one or more labeling elements.
[0008] The computer implementation method includes the steps of: acquiring a medical image having one or more labeling elements, where each labeling element is part of a medical image representing a material containing a contrast agent; processing the medical image to generate a suppressed image containing only one or more labeling elements of the medical image; and combining the medical image and the suppressed image to generate a processed medical image, wherein in the processed medical image, the apparent brightness of one or more labeling elements is reduced, so that the labeling elements are diluted compared to the medical image, but not completely suppressed.
[0009] This disclosure proposes reducing the brightness of one or more identifying elements in a medical image without completely removing them. This is achieved by generating a suppression image (containing the identifying elements) and appropriately combining the suppression image with the medical image. Using a suppression image containing one or more identifying elements makes it possible to manipulate, modify, and / or weight the suppression image without affecting the content of the original medical image. By combining the above two, it is possible to reduce the apparent brightness, i.e., appearance, of the identifying elements while preserving the features of the original medical image.
[0010] This disclosure provides, in effect, processed medical images in which labeling elements are “faded” rather than deleted, erased, or completely suppressed. This avoids the removal of potentially valuable information from labeling elements (e.g., for contextual understanding) while mitigating the effects of contrast agent use. Specifically, the present invention proposes a novel image processing technique for reducing the apparent brightness of labeling elements.
[0011] The proposed approach facilitates the identification of cysts, lesions, proliferations, or defects (such as polyps or fecal residues) within medical structures (in medical images) without them being "obscured" by brighter marking elements. Furthermore, the proposed approach still conveys information about the original location of the marking elements (since the marking elements are not completely removed), which facilitates a better contextual understanding of suppression artifacts and / or streak artifacts. Additionally, by not completely removing the marking elements, the location of air (bubbles) in medical images can be identified, as these appear as areas with even lower brightness than the marking elements in the processed medical image.
[0012] By using suppressed images that contain only one or more marker elements, modifications can be made to the marker elements without affecting the brightness of other elements depicted in the (original) medical image. This facilitates more specific and directional brightness modifications.
[0013] Using suppressed images (which are separate from medical images) increases the flexibility of image processing for medical images. Specifically, using suppressed images allows for the adoption of more complex and advanced image processing techniques.
[0014] For example, one advantage of using suppressed images is the increased ease and speed of interactively changing the apparent brightness or visibility of remaining ambiguous areas. This is because only the suppressed images need to be processed, without requiring, for example, reprocessing of the entire image volume.
[0015] Another advantage is the ability to modify medical images "on the fly." For example, the same suppressed image can be combined with conventional CT Hounsfield images, virtual single-energy images from spectral CT, or other derived images (all of which depict the same region of interest, and whose labeling elements have already been identified in the suppressed image).
[0016] Another advantage is that the suppressed images themselves can be further processed, for example, in terms of the size of the structures, before being applied to or combined with medical images. For example, thin structures can be made less thin than thick structures, or small structures can be made less thin than thick structures.
[0017] In the context of this disclosure, “apparent luminance” refers to the brightness or luminance of an image element as seen by the user. Therefore, pixels in an image with higher apparent luminance appear whiter than pixels in an image with lower apparent luminance. The concept of luminance is well-established in the field of image processing.
[0018] As described below, this disclosure envisions various mechanisms for reducing the apparent brightness of a labeling element by combining medical images and suppression images.
[0019] The step of combining a medical image and a suppressed image may be configured such that the apparent brightness of one or more indicator elements in the processed medical image is lower than the apparent brightness of one or more other elements in the processed medical image.
[0020] Therefore, the brightness of labeled elements in the processed medical image is made lower than the brightness of unlabeled elements in the medical image (i.e., elements that do not represent materials containing contrast agents). This further improves the visibility of unlabeled elements, for example, for evaluation or analysis, without completely discarding potentially valuable information from the labeled elements.
[0021] The step of combining a medical image and a suppressed image may include modifying the suppressed image to reduce the apparent brightness of one or more indicator elements in the suppressed image, and combining the medical image and the modified suppressed image to generate a processed medical image.
[0022] The above computer implementation method may be adapted so that the step of combining the medical image and the suppression image includes a step of subtracting the modified suppression image from the medical image.
[0023] Optionally, the step of combining the medical image and the suppression image includes the step of multiplying the medical image and the modified suppression image together.
[0024] The step of modifying the suppression image may include the step of reducing the value of one or more pixel parameters of all pixels representing the labeling elements in the suppression image by a predetermined amount or a predetermined ratio.
[0025] Images (including both 2D and 3D images) are formed from pixels. In the case of 3D images, these are also labeled as voxels. Each pixel specifies one or more pixel values representing the values of one or more different pixel parameters for that pixel. The number of pixel parameters for one pixel varies depending on the format of the image. For example, an RGB image includes at least three pixel parameters for one pixel, and a CT image has only a single pixel parameter (representing HU value) for one pixel.
[0026] It is proposed that the value of the pixel parameter be reduced by a predetermined amount. Thus, different pixel parameters are reduced by different predetermined amounts, but the same reduction is applied to the same pixel parameter for all pixels.
[0027] This provides a flexible mechanism for reducing the apparent brightness of the labeled element suppression image or for modifying the appearance of the labeled elements in the suppression image.
[0028] The step of modifying the suppression image may further include the step of performing a smoothing operation on the suppression image. Smoothing the suppression image improves the visual appearance of one or more labeled elements, for example, by blending the edges of one or more labeled elements of the processed medical image with other elements.
[0029] The step of combining the medical image and the suppression image may include the step of weighting the medical image and the suppression image during the combination. Thereby, another mechanism for combining the medical image and the suppression image is provided, which can reduce or modify the appearance of the labeled elements in the medical image.
[0030] The medical image is a medical image of the colon, and each labeled element represents a labeled stool in the colon. It is recognized that the proposed approach is particularly advantageous for reducing the appearance of stools in the colon. This is because it is recognized that labeled stools are particularly problematic by hiding potential sites of defects, growths, or diseases in the colon.
[0031] The medical image can be a computed tomography (CT) image. It is recognized that the labeled elements in the CT image are particularly problematic during image analysis because of the large difference in luminance between the labeled elements and the non-labeled elements. A large difference in luminance makes the evaluation or analysis of the labeled image extremely difficult. However, it will be understood that other medical image modalities such as, for example, magnetic resonance (MR) images (e.g., the labeled elements represent elements whose magnetic response is impaired during MR imaging using an MRI contrast agent), X-ray images (e.g., the elements are labeled using a radiopaque contrast agent), or ultrasound images (e.g., using an ultrasound contrast agent) can be processed using the proposed approach.
[0032] The above computer-implemented method may further include the step of displaying the processed medical image on a display. Alternatively, the processed medical image can be processed using, for example, a machine learning method to identify potential defects represented by the medical image and areas of (medical) concern. Of course, the processed medical image may be subjected to further processing such as noise reduction. Thus, the above method may include the step of outputting the processed medical image to a further processing device.
[0033] The above method may further include a step of further processing the processed medical image using a machine learning method to identify one or more characteristics of the medical image and / or one or more anatomical structures represented by the medical image.
[0034] A computer implementation method for generating one or more virtual medical images is also proposed. This computer implementation method includes the steps of: acquiring a first base medical image, the first base medical image having one or more labeling elements; acquiring a second base medical image generated using a different imaging modality than the first base medical image, the second base medical image having one or more labeling elements, and the first and second base medical images being usable together to generate one or more virtual medical images; performing labeling element suppression on the first base medical image by performing any of the aforementioned methods to generate a processed first base medical image; performing labeling element suppression on the second base medical image by performing any tactical method to generate a processed second base medical image; and processing the processed first and second base medical images to generate one or more virtual medical images.
[0035] The first and second base medical images image the same region of interest of the patient. For example, they have the same virtual camera position and / or orientation, but are acquired using different modalities (e.g., using different energy levels to generate the images).
[0036] A computer program product including computer program code means is also proposed. When the computer program code means is executed on a computing device having a processing system, it causes the processing system to perform all the steps of any of the aforementioned methods.
[0037] A processing system is also proposed for performing label element suppression on medical images having one or more label elements. This processing system performs the following: acquiring a medical image having one or more label elements, where each label element is part of a medical image representing a material containing a contrast agent; processing the medical image to generate a suppressed image containing at least one or more label elements of the medical image; and combining the medical image and the suppressed image to generate a processed medical image, wherein in the processed medical image, the apparent brightness of one or more label elements in the medical image is reduced, so that the label elements are diluted but not completely suppressed.
[0038] Medical images can be obtained, for example, from a medical image database or from a medical image generator (such as a CT scanner or MRI scanner). Processed medical images can be output to a display (to show the processed medical images) or to a further processor, for example, to perform further image processing on the processed medical images or to analyze the content of the processed medical images.
[0039] The processing system can be adapted to perform any of the aforementioned methods, and those skilled in the art can adapt the processing system accordingly.
[0040] 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]
[0041] To gain a deeper understanding of the present invention and to more clearly illustrate how it is implemented, please refer to the attached drawings as just one example.
[0042] [Figure 1] Figure 1 shows a method according to one embodiment. [Figure 2] Figure 2 shows a method according to one embodiment. [Figure 3] Figure 3 shows a medical image obtained through a method according to one embodiment. [Figure 4] Figure 4 shows another medical image obtained through the method according to one embodiment. [Figure 5] Figure 5 shows yet another medical image obtained through a method according to one embodiment, compared to an existing method. [Figure 6] Figure 6 shows the processing system. [Modes for carrying out the invention]
[0043] The present invention will be described with reference to the figures.
[0044] 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.
[0045] The present invention provides a mechanism for reducing the appearance of label elements in medical images. This is achieved by processing the medical image to generate a separate suppression image containing only the label elements. The medical image and the suppression image are then combined to reduce the appearance of label elements in the medical image. This can be achieved by modifying the suppression image before combination or by weighting the medical image and the suppression image during combination.
[0046] This disclosure is based on the recognition that it is possible to extract a label element in a medical image from other elements to form a separate suppression image, and that it is advantageous to use a second separate image to process the label element separately from other elements. In particular, it is recognized that using a separate suppression image makes it easier to use more complex and advanced image processing techniques and allows for greater flexibility (and less computational complexity) in adjusting the apparent brightness of label elements in medical images.
[0047] Embodiments can be used to reduce the appearance of labeled stool in, for example, CT images of the colon, without completely removing or deleting stool (in order to make it easier to understand and reduce the possibility of potentially valuable clinical information being removed).
[0048] The proposed approach to reduce the brightness and appearance of labeling elements (without completely removing them) can generate enhanced medical images that provide clinicians with more clinically relevant data. Specifically, it generates medical images that closely resemble labeling-free medical images (which are clinically more useful) while reducing the likelihood of subtraction artifacts going unnoticed.
[0049] Figure 1 shows method 100 according to one embodiment of the present invention.
[0050] This method includes step 110 of acquiring a medical image 150.
[0051] A medical image can be any suitable medical image that contains one or more labeling elements. For example, a CT image in which areas representing stool are labeled (e.g., areas with high brightness) or an MRI image in which areas representing blood are labeled (sometimes called a "bright blood" image).
[0052] In particular, medical image 150 represents the patient's region of interest. Medical image 150 is a 2D image, a 3D image, and / or a frame from a 2D / 3D video or cineloop.
[0053] Those skilled in the art will readily understand how to obtain medical images with one or more identifying elements. Typically, contrast agents or media are used to highlight fluids, foreign bodies, or digestive materials (such as boluses, chimes, chyle, stool, or fecal matter) moving within the body in a medical image.
[0054] Therefore, the labeling element represents a substance that appears more easily identifiable (in a medical image) with greater brightness under the influence of a contrast agent / contrast medium than without the influence of a contrast agent / contrast medium.
[0055] Method 100 processes the medical image 150 in step 120 to generate a suppression image. The suppression image is an image that contains only the labeling elements of the medical image (for example, at the same resolution as the medical image). Thus, the suppression image is essentially a version of the medical image in which only the labeling elements are duplicated.
[0056] Step 120 is performed using several approaches. Generally, step 120 involves identifying any labeling elements in a medical image and using this information to generate a “suppression image.” This image contains the labeling elements, but everything else is blank or zeroed out.
[0057] For example, step 120 includes initializing a suppression image having the same resolution as the medical image, identifying the pixels in the medical image corresponding to the labeling elements, and replicating the pixel values of these pixels to the corresponding pixels in the suppression image. The initialized suppression image may have a predetermined value (e.g., 0) for all the pixel values of the pixels.
[0058] Another approach involves generating a suppression image by duplicating the medical image (generating an initial suppression image) and setting the pixel values of all pixels that do not correspond to the labeling elements to 0.
[0059] In one example, since the labeling element appears to have high brightness, the suppression image is generated by identifying all areas of the medical image that have brightness greater than some predetermined threshold. Therefore, the suppression image may be generated using a simple thresholding process.
[0060] In some embodiments, the marker elements are identified using image segmentation techniques. Any suitable image segmentation technique can be used, such as edge detection approaches (since the marker elements should have clear, bright edges), model-based segmentation techniques (such as those disclosed in International Patent Publication WO2018 / 224363A1), thresholding-based segmentation techniques (such as those disclosed in International Patent Publication WO2007 / 046019A1), or deep learning approaches, i.e., machine learning methods such as those utilizing one or more convolutional neural networks (CNNs).
[0061] Naturally, a combination of these techniques can also be used.
[0062] Next, method 100 performs process 130 to generate a processed medical image 190 by combining the medical image 150 and the suppressed image. Process 130 is configured such that the marked elements in the processed medical image have the apparent brightness of one or more marked elements, which is reduced but not zero compared to the medical image.
[0063] This process 130 can be achieved by implementing one of several approaches.
[0064] In one example, medical images and suppression images can be directly combined. For example, they can be averaged, multiplied, or the suppression image can be subtracted from the medical image (for example, in step 135). In this process, the medical image and / or suppression image are appropriately weighted to reduce the apparent brightness of the marked elements.
[0065] For example, consider a scenario where a subtractive approach is used in the step of combining a medical image and a suppression image. In this case, the pixel value of each pixel in the combined medical image is the (weighted) pixel value of the corresponding pixel in the suppression image, which is obtained by subtracting the (weighted) pixel value of the corresponding pixel in the medical image. In this scenario, if all pixel values in the medical image are weighted by X and all pixel values in the suppression image are weighted by Y, the resulting pixel value of the marking element in the medical image will be (YX) times the pixel value of the medical image. This effectively reduces the apparent brightness of the marking element compared to the medical image.
[0066] In a preferred embodiment, process 130 includes step 131 of reducing the apparent brightness of the suppressed image before combining the medical image and the suppressed image (by subtraction or multiplication). This allows for better control of the brightness of the indicator elements in the processed image.
[0067] In particular, the apparent brightness of the suppressed image is reduced, but not to zero, when subtracted or multiplied from the medical image. That is, the indicator elements are still visible in the processed image, but their brightness is reduced.
[0068] The process of reducing apparent brightness differs depending on the type of suppressed image. For example, if the suppressed image is a brightness map (such as a CT Hounsfield image), reducing brightness simply means decreasing the single pixel value of each pixel. As another example, if the suppressed image is an RGB image, reducing brightness means decreasing the R, G, and B values of each pixel, and / or (if present) adjusting the alpha value of each pixel, i.e., the α value (which defines the transparency of the pixel).
[0069] The amount of brightness reduction may be, for example, in response to user input or may be predetermined. Preferably, the amount of brightness reduction is in response to user input in the user interface, allowing the user to control the apparent brightness of the indicator element. Using suppressed images is particularly advantageous when the amount of brightness reduction is determined in response to user input. This is because the processed medical image can be easily reproduced without completely reprocessing the original medical image.
[0070] In some embodiments, the value of one or more pixel parameters of all pixels representing the labeling element in the suppressed image is reduced by a predetermined amount or by a predetermined percentage.
[0071] The amount of brightness reduction does not need to be uniform across all pixels of the suppressed image. In some embodiments, different marker elements are reduced in brightness by different amounts. For example, thin marker elements may be reduced in brightness less than thick marker elements (to reduce the likelihood of clinicians missing thin marker elements), and small marker elements may be reduced in brightness less than large marker elements.
[0072] Therefore, the amount of brightness reduction for pixels in a suppressed image varies depending on the size and shape of the marker element to which the pixel belongs.
[0073] The combining step 130 includes combining the suppression image and the medical image by subtracting the suppression image from the medical image or by multiplying the medical image and the suppression image in step 135.
[0074] Preferably, process 130 includes step 132, which performs a smoothing operation on the suppression image (before combining it with the medical image). Imaging smoothing operations are well known to those skilled in the art and employ the application of one or more smoothing filters to the suppression image.
[0075] The smoothing operation improves the visual appearance of the processed medical image by reducing the appearance of any sharp or abrupt changes in brightness between labeled and unlabeled elements in the processed image. This reduces the appearance of artifacts in the processed medical image (e.g., in these abrupt gradients or changes) and provides a representation of the region of interest represented by the processed medical image that is closer to a truly labeled-element-free image.
[0076] The advantages of using a separate suppression image are further demonstrated in the smoothing operation. This is because the smoothing operation is performed only on the labeled elements, resulting in improved visual appearance and reduced artifacts without compromising the details of other parts of the medical image (which would occur if the smoothing operation were performed on the entire medical image).
[0077] In the combining step 130, other processing steps (such as denoising or colorization) may be performed. This may be done in addition to or instead of step 132, which performs smoothing.
[0078] Therefore, using suppressed images, additional processing (such as smoothing or denoising) can be performed on the labeled elements to generate processed suppressed images 170 without affecting other components of the medical image. This provides a new mechanism for image processing of labeled elements without affecting other components of the medical image.
[0079] Using separate suppression images also has the advantage of allowing the same suppression image to be applied to different versions of a medical image (for example, medical images of the same region of interest but acquired using different imaging modalities). This reduces the workload involved in generating processed medical images by reusing or repurposing the generated suppression images for other medical images of the same region of interest.
[0080] Therefore, in some embodiments, once a suppression image is generated (and optionally processed) for a particular medical image, the same suppression image can be combined with other medical images of the same region of interest to reduce the brightness of marking elements in the other medical images.
[0081] Similarly, the same suppression image can be combined with different derived images from the original medical image (such as further processed medical images like virtual non-calcium medical images or virtual non-contrast images).
[0082] Method 100 further includes step 183 of acquiring additional medical images 150B (images derived from medical image 150 or images taken at the same location as medical image 150). In process 185, the additional medical images 150B and the (processed) suppression image 170 are combined. This is similar to the combination process 130, for example, to generate a processed additional medical image.
[0083] Therefore, the suppression image 170 can be reused to combine with other medical images, such as images obtained from the original medical image.
[0084] This disclosure proposes an approach for generating processed medical images. This approach generates suppressed images containing only the labeling elements of the medical image from the original medical image. This facilitates the reuse of the suppressed images and dedicated image processing of the labeling elements (by image processing of the suppressed images). Furthermore, this approach facilitates simple adjustments of the brightness of the labeling elements in the processed medical image (for example, by adjusting the brightness of the suppressed image or the weighting of the suppressed image during combination) without completely reprocessing the original medical image.
[0085] In processed medical images, the brightness / appearance of labeling elements is reduced but not eliminated. This allows clinicians to more easily identify inconsistencies and features of anatomical structures represented in processed medical images without being "masked" by labeling elements, while still maintaining the ability to distinguish between labeling elements and bubbles and identify the causes of artifacts resulting from labeling elements in processed medical images.
[0086] Processed medical images can be displayed in a user interface or further processed (e.g., using machine learning methods) to identify one or more characteristics of the medical image or the region of interest represented by the medical image. For example, machine learning methods can be used to further process a processed medical image of the colon to identify the presence and characteristics of one or more polyps present in the colon.
[0087] Therefore, Method 100 includes the step of displaying the processed medical image, for example, in a user interface. Similarly, Method 100 includes the step of further processing the medical image to perform characterization, analysis, and / or classification of a region of interest represented by the medical image and / or processed medical image.
[0088] In some embodiments, this method includes the step of performing further image processing, such as denoising, on the processed medical image. The further processed medical image is then displayed in a user interface or the like.
[0089] Naturally, this method may be configured to operate in a mode where the label elements are completely suppressed (i.e., the values of the label elements in the processed medical image become 0). This mode can be switched, for example, in response to user input.
[0090] Figure 2 illustrates the application of a method for generating virtual medical images. In particular, Figure 2 shows a method 200 for generating one or more virtual medical images.
[0091] Method 200 includes step 210 of obtaining a first base medical image 291 having one or more marker elements.
[0092] Method 200 further includes step 220 of obtaining a second base medical image 292 generated using an imaging modality different from that of the first base medical image. The second base medical image has one or more marker elements. The first and second base medical images can be used together to generate one or more virtual medical images.
[0093] In other words, the first and second base medical images are designed to be combined to synthesize a new virtual medical image.
[0094] As just one example, the first base medical image 291 is a CT image acquired using a first energy level (of emitted X-rays), and the second base medical image 292 is a CT image acquired using a second different energy level (of emitted X-rays). These two base images are blended at different ratios to generate different virtual single-energy images. This synthesizes a medical image with a hypothetical energy level, i.e., a medical image that is functionally equivalent to a medical image obtained from an X-ray beam at that hypothetical energy level.
[0095] Both base medical images represent the same region of interest; that is, they contain the same elements, but acquired with different imaging modalities. For example, CT images are acquired using X-rays at multiple energy levels.
[0096] Next, the method 100 described herein is performed on each medical image to generate each processed base image. Thus, the method 200 includes performing label element suppression 100 on a first base medical image 291 to generate a processed first base medical image 293, and performing label element suppression 100 on a second base medical image to generate a processed second base medical image 294.
[0097] The processed first base medical image 293 and second base medical image 294 are processed together to generate or synthesize one or more virtual medical images 250. Mechanisms for synthesizing virtual medical images from base images are well known to those skilled in the art, and among several processes, a decomposition process is used.
[0098] For example, a method for generating a virtual single-energy image from two base (CT) medical images acquired using different energy levels is described by Matsumoto, Kazuhiro et al. in "Virtual monochromatic spectral imaging with fast kilovoltage switching: improved image quality as compared with that obtained with conventional 120-kVp CT" (Radiology 259.1(2011):257-262), and by D'Angelo, Tommaso et al. in "Dual energy computed tomography virtual monoenergetic imaging: technique and clinical applications" (The British journal of radiology 92.1098(2019):20180546).
[0099] Other approaches to synthesizing a virtual medical image from two base medical images will be obvious to those skilled in the art.
[0100] Therefore, it will be understood that by utilizing the proposed approach to reduce the appearance of labeling elements in medical images, it is possible to generate one or more virtual images in which the appearance of labeling elements is similarly reduced (but not zero).
[0101] Figure 3 shows the effect of applying the method according to one embodiment to an exemplary medical image 300.
[0102] Medical image 300 is a CT Hounsfield image of the colon and includes multiple labeling elements, as shown in the brightest region of medical image 300. If medical image 300 is subject to the approach described herein (for example, according to method 100 described with reference to Figure 1), a processed medical image 350 is generated.
[0103] The processed medical image 350 still contains the labeling elements (i.e., the labeling elements have not been removed or are not zero), but their apparent brightness has been reduced. The difference between the diluted labeling areas and the bubbles remains intuitively recognizable.
[0104] Figure 4 shows the effect of applying the method according to one embodiment to another exemplary medical image 400.
[0105] Medical image 400 is another example of a CT Hounsfield image of the colon, and includes multiple labeling elements, indicated in the brightest area of medical image 400. Medical image 400 shows strong streak artifacts (indicated by arrows) due to high absorption of the contrast agent material. Beam hardening artifacts lead to cavities that appear pseudo-in the colonic folds and surrounding soft tissue.
[0106] The processed medical image 450 is obtained by performing a method described herein, such as method 100, as described with reference to Figure 1, on the medical image 400.
[0107] Although streak artifacts are still present in the processed medical image 450, the original position of the marker element is still visible (because the luminance of the marker element is non-zero), and therefore the cause of the artifact is still indicated / visible, making it intuitive to the observer.
[0108] Therefore, the proposed approach provides images that hold useful information to help clinicians make correct clinical decisions by providing information about the causes of potential artifacts in the images that confuse and / or mislead (to make an inaccurate diagnosis).
[0109] Figure 5 shows a comparison between performing method 100 according to an embodiment of the present invention on a medical image 500 and performing method 590 according to a prior art technique.
[0110] When a medical image 500 is processed using method 590 according to prior art techniques, all identifying elements are completely removed from the medical image, and a first processed medical image 540 is generated. This process is called "suppression" of identifying elements.
[0111] When the medical image 500 is processed using the method 100 according to an embodiment of the present invention, as described with reference to Figure 1, for example, the brightness of the labeling elements is reduced, but not to zero. This process can be called "partial suppression" of the labeling elements. Thus, the labeling elements are not removed but are effectively "diluted". This generates a second processed medical image 550.
[0112] In the first processed medical image 540, removing the labeling elements leaves cleansing artifacts at the boundaries of the cleansed region. These cleansing artifacts can mislead clinicians and fail to provide a true representation of the imaged region of interest.
[0113] In the second processed medical image 550, the cause of any surface artifact resulting from partial suppression can be easily identified (e.g., by the presence of less brightly lit labeling elements). This avoids misleading clinicians with artifacts due to defects in the colon, and instead improves the performance of patient evaluation and / or diagnosis by providing clinicians with clinical information useful in identifying the presence and cause of artifacts.
[0114] Figure 6 shows a processing arrangement 600 according to one embodiment. The processing arrangement 600 includes a processing system 610 which is itself one embodiment of the present invention.
[0115] The processing system performs label element suppression on medical images that have one or more label elements.
[0116] Specifically, the processing system 610 acquires a medical image having one or more labeling elements. Each labeling element is a part of the medical image representing a material containing a contrast agent. The processing system 610 can also acquire the medical image from memory 620 (of the processing arrangement 600) and / or directly from a scanner (not shown).
[0117] The processing system processes a medical image to generate a suppressed image containing at least one or more marker elements of the medical image, and combines the medical image and the suppressed image to generate a processed medical image in which the apparent brightness of one or more marker elements in the medical image is reduced but not zero.
[0118] Therefore, the processing system 610 generates processed medical images by performing any method described herein.
[0119] The processing configuration 600 further includes a display 630. The processing system displays the processed medical images (and optionally medical images and / or suppression images) on the display 630.
[0120] The processing configuration 600 further includes an additional processor 640. The additional processor performs further processing (denoising, analysis, etc.) on the processed medical images. The information and / or images (e.g., the results of further processing) generated by the additional processor are displayed on the display 630, and the additional processor can be configured accordingly.
[0121] Those skilled in the art will be able to easily develop a processing system for carrying out the method described herein. Thus, each step in the flowchart represents a different action performed by the processing system, which can be performed by the corresponding module of the processing system.
[0122] As mentioned above, the system performs data processing using a processing system. The processing system can be implemented in various ways using software and hardware to perform a variety of necessary functions. Typically, a processing system uses one or more microprocessors programmed with software (e.g., microcode) to perform the required functions. A processing system can be implemented as a combination of dedicated hardware for some functions and one or more programmed microprocessors and associated circuits for other functions.
[0123] 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).
[0124] In various implementations, a processing system 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 perform the necessary functions when running on one or more processing systems and / or controllers. The various storage media may be fixed within a processing system or controller, or they may be transportable so that one or more programs stored therein can be loaded into the processing system.
[0125] Therefore, a computer program product including computer program code means is also proposed. When the computer program code means is executed on a computing device having a processing system, it causes the processing system to perform all steps of the method.
[0126] 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 “includes” does not exclude other elements or steps, and singular elements do not exclude plural elements. The mere fact that certain means are described in different dependent claims does not mean that combinations of these means cannot be used advantageously.
[0127] A single processor or other unit may perform the functions of some of the items described in the claims. Computer programs may be stored / distributed on any suitable medium, such as optical storage media or solid-state media, supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless communication systems.
[0128] 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...". Any reference numerals in the claims should not be construed as limiting the scope.
Claims
1. 1. A computer-implemented method for performing marking element attenuation on a medical image having one or more marking elements, the computer-implemented method comprising: acquiring the medical image having the one or more landmark elements, each landmark element being a portion of the medical image that represents a material that includes a contrast agent; processing the medical image to generate a suppression image that includes only the one or more landmark elements of the medical image; combining said medical image and said suppression image to generate a processed medical image, in which the apparent brightness of said one or more marking elements is reduced so that said marking elements are dimmed but not completely suppressed compared to said medical image, such that said marking elements are still visible in said processed image but less bright; 4. A computer-implemented method comprising:
2. 2. The computer-implemented method of claim 1, wherein the step of combining the medical image and the suppression image is performed such that an apparent brightness of the one or more sign elements in the processed medical image is less than an apparent brightness of one or more other elements in the processed medical image.
3. The step of combining the medical image and the suppression image includes: modifying the suppression image to reduce an apparent brightness of the one or more sign elements in the suppression image; combining the medical image with the modified suppression image to generate the processed medical image; 3. The computer-implemented method of claim 1, comprising:
4. The computer-implemented method of claim 3 , wherein combining the medical image and the suppression image comprises subtracting a modified suppression image from the medical image.
5. The computer-implemented method of claim 3 , wherein combining the medical image and the suppression image comprises multiplying the medical image and the modified suppression image together.
6. 4. The computer-implemented method of claim 3, wherein modifying the suppression image comprises reducing values of one or more pixel parameters of all pixels representing sign elements in the suppression image by a predetermined amount or by a predetermined percentage.
7. The computer-implemented method of claim 3 , wherein modifying the suppression image further comprises performing a smoothing operation on the suppression image.
8. The computer-implemented method of claim 1 , wherein combining the medical image and the suppression image comprises weighting the medical image and the suppression image during combination.
9. The computer-implemented method of claim 1 , wherein the medical image is a medical image of a colon, and each landmark element represents landmark stool within the colon.
10. The computer-implemented method of claim 1 , wherein the medical image is a computed tomography image.
11. The computer-implemented method of claim 1 , further comprising the step of displaying the processed medical image on a display.
12. 4. The computer-implemented method of claim 1, further comprising the step of further processing the processed medical image using machine learning methods to identify one or more characteristics of the medical image and / or one or more anatomical structures represented by the medical image.
13. 1. A computer-implemented method for generating one or more virtual medical images, comprising: acquiring a first base medical image having one or more landmark elements; acquiring a second base medical image generated using a different imaging modality than the first base medical image, the second base medical image having one or more landmark elements, the first base medical image and the second base medical image being usable together to generate one or more virtual medical images; performing landmark element suppression on the first base medical image by executing the computer-implemented method according to any one of claims 1 to 3 to generate a processed first base medical image; performing landmark element suppression on the second base medical image by executing the computer-implemented method according to any one of claims 1 to 3 to generate a processed second base medical image; processing the processed first base medical image and the processed second base medical image to generate one or more virtual medical images; 4. A computer-implemented method comprising:
14. A computer program comprising computer program code means which, when executed on a computing device having a processing system, causes the processing system to perform all the steps of the computer implemented method of any one of claims 1 to 3.
15. 1. A processing system for performing marking element attenuation on a medical image having one or more marking elements, the processing system comprising: acquiring the medical image having the one or more landmark elements, each landmark element being a portion of the medical image that represents a material that includes a contrast agent; processing the medical image to generate a suppression image including at least the one or more landmark elements of the medical image; combining the medical image and the suppression image to generate a processed medical image, in which an apparent brightness of the one or more marking elements in the medical image is reduced so that the marking elements are dimmed but not completely suppressed, such that the marking elements are still visible in the processed image but less bright; A processing system that executes the above.
