Inspiratory metrics for chest X-ray images

Abstract

In order to enhance the monitoring of inspiratory quality in enhanced X-ray images, a metric is proposed that can reproducibly provide an index of visible ribs to be used in the assessment of inspiratory status. In one example, the diaphragm detected in a chest X-ray image can be projected onto an atlas containing labels for all intercostal spaces (i.e., the spaces between the centerlines of the ribs). A spatial representation of the clavicle and ribs is provided in the atlas, and a cumulative histogram is constructed for all points (i.e., pixels) of the diaphragm. For each point, the rib label counter of the rib in the rib label map is incremented at that point and all ribs above it. The rib label counter is normalized by dividing by the number of points, and the median (or different quantiles) of this distribution can be taken to serve as the inspiratory index. Thus, an objective metric of inspiratory status is achieved.

Description

Technical Field

[0001] The present invention relates to X-ray image inhalation quality monitoring, and in particular to a method for X-ray image inhalation quality monitoring, an apparatus for X-ray image inhalation quality monitoring, an X-ray imaging system, a computer program element and a computer readable medium. Background Art

[0002] Chest X-ray examination is an important procedure for detecting and monitoring lung abnormalities and diseases in the early stages of a patient's lungs. Medical personnel (e.g., doctors or radiologists) detect abnormalities directly from the X-ray images. It is the task of the medical imaging professional to evaluate the image quality directly after acquisition. Before the patient leaves the examination room, the medical imaging professional checks whether the image is of sufficient quality. If the chest X-ray image is not of sufficient quality, a new image should be taken before the patient leaves the examination room. The image quality depends on the skill of the system operator. US2018 / 0325481 A1 discusses the evaluation of image quality based on the rib intersection rule. However, several posterior ribs may intersect with the diaphragm, making this task inherently ambiguous.

[0003] US 2013 / 156267 A1 discloses a diagnosis assistance system. Summary of the Invention

[0004] There is a need to provide techniques for enhanced X-ray image inhalation quality monitoring.

[0005] It should be noted that the aspects of the invention described below also apply to the method, the apparatus, the X-ray imaging system, the computer program element and the computer-readable medium.

[0006] A first aspect of the present invention provides a computer-implemented method for monitoring the quality of inhalation in X-ray images. The method comprises the following steps:

[0007] a) receiving a chest X-ray image of a patient;

[0008] b) detecting ribs in the chest X-ray image;

[0009] c) identifying a plurality of intercostal spaces, each intercostal space representing a space between detected adjacent ribs;

[0010] d) detecting a diaphragm in the chest X-ray image and generating a diaphragm line, the diaphragm line comprising a plurality of points representing a path of the diaphragm in the chest X-ray image; and

[0011] e) applying a score assignment rule to assign a point score to each point of the septal line based on the corresponding intercostal space in which the point is located in the image, wherein the score assignment rule defines a correspondence between point scores to be assigned according to the intercostal space in which the point is located; and

[0012] f) determining an inspiration index based on the point score of the point of the diaphragm line.

[0013] In other words, a chest X-ray image is evaluated to determine the location of points on the diaphragm relative to the intercostal spaces. Each point on the diaphragm is associated with a corresponding point score based on the intercostal space in which the point is located. For example, each intercostal space can be labeled (e.g., referred to) from "1" to "10." The point score of a point can be associated with the label of the corresponding intercostal space in which the point is located. For example, if the point is located in an intercostal space labeled "10," a point score of "10" is assigned to the point. Thus, if a rib intersects the diaphragm, the diaphragm can have points with different point scores. Because higher-quality images and lower-quality images have different intersection conditions, the diaphragm in a good-quality image has a different distribution of point scores along the diaphragm line than the diaphragm in a poor-quality image. Therefore, an inspiration index can be determined based on the distribution of point scores along the diaphragm. In some examples, the inspiration index can be reduced to a simple overall value, such as the mean or median of the point scores of all points on the diaphragm. In this way, a unique metric is provided that provides an index of the visible ribs above the diaphragm. Finally, image quality can be assessed based on the calculated inhalation index. In one example, the operator can use the inhalation index to measure image quality. In another example, the inhalation index can be further classified into three groups: "poor, moderate, and good" based on predefined X-ray image inhalation quality monitoring rules as described below.

[0014] The proposed method thus enables more reliable and easier assessment and determination of chest X-ray image quality. Calculation of the inspiratory index for each chest X-ray image can be performed without human oversight or interference, thus avoiding dependence on variations or biases in the skill level of different system operators. Automated quantification of a patient's inspiratory status can also be integrated into automated radiology interpretation and reporting processes. Consequently, the success rate of medical imaging can be improved and the assessment process can be streamlined.

[0015] As used herein, the term "rib" or "rib" may refer to the X-ray projection of the rib onto an X-ray detector, thereby forming an X-ray image of the rib. Thus, the term "rib centerline" defines a line in an X-ray image that corresponds to the shape of a rib in a patient. Depending on the imaging angle, image type, and the degree to which the patient inhales or exhales, the rib centerline may be more or less curved and may intersect the diaphragm or other rib centerlines.

[0016] As used herein, the term "intercostal space" may refer to the space between the centerlines of adjacent ribs.

[0017] As used herein, the term "diaphragm" may refer to an X-ray projection of the diaphragm onto an X-ray detector, thereby forming an X-ray image of the diaphragm (e.g., from the posterior or anterior body direction). Thus, the term "diaphragm line" defines a path that traces the shape of the diaphragm in an X-ray image. The diaphragm typically appears as a cup-shaped boundary at the base of a lung lobe, but this boundary can move depending on the degree of chest expansion.

[0018] The term "score assignment rule" as used in this document defines the correspondence between point scores and intercostal spaces. For example, a lookup table can be used to define the association between point scores and intercostal spaces. Once the intercostal space in which a point is located is determined, each intercostal space can be checked equally in turn, and the associated point score (if any) can be used as the search result (i.e., the point score assigned to the point). In another example, each intercostal space can be labeled, for example, from "1" to "10". The point score of a point can be associated with the label of the corresponding intercostal space in which the point is located. For example, if a point is located in an intercostal space with the label "10", a point score of "10" is assigned to the point.

[0019] According to an embodiment of the present invention, the method further comprises comparing the determined inhalation index with a predetermined X-ray image inhalation quality monitoring rule specifying an image inhalation state in terms of the inhalation index to generate an X-ray image inhalation quality metric.

[0020] Predetermined X-ray image inhalation quality monitoring rules can enable determination of whether a good image or a poor image exists. The predetermined X-ray image inhalation quality monitoring rules can classify the values of the inhalation index into two or more categories. Each category corresponds to a specific range of values for the inhalation index, for example, a specific range of values for the inhalation index selected from the group of "poor, medium, good." For example, the category "good" corresponds to an inhalation index ranging from 9 to 10, the category "medium" corresponds to an inhalation index ranging from 8 to 9, and the category "poor" corresponds to the remaining inhalation indices. The X-ray image inhalation quality metric can be output on a display (e.g., a computer display screen or a mobile device display). In another example, the X-ray image inhalation quality metric can be displayed as an illumination warning light or a warning sound. In yet another example, the X-ray image inhalation quality metric can be output as a bitmap "watermark" inserted into a portion of the X-ray image provided by the X-ray device, or the X-ray image inhalation quality metric can simply be stored in the image metadata (DICOM tag).

[0021] This may facilitate the system operator's determination of image quality compared to a continuous value of the inspiration index.

[0022] According to an embodiment of the present invention, step b) further comprises:

[0023] b1) constructing a statistical atlas by co-registration of training images comprising a plurality of chest X-ray images of one or more patients; and

[0024] b2) registering the statistical atlas with the chest X-ray image of the patient to construct a patient-specific rib model of the patient.

[0025] In other words, one or more rib atlases can be used to automatically detect ribs in chest X-ray images. The purpose of the statistical atlas is to construct a rib model for a patient's X-ray using other patient X-rays with their corresponding rib models. Using a patient-specific atlas of the ribs, or equivalently, a patient-specific atlas of the intercostal spaces, can robustly account for the inherent ambiguity caused by the intersection of several posterior ribs with the diaphragm.

[0026] A statistical atlas can be constructed by co-registering the training images. In this atlas, a rib index is assigned to each location by construction. This atlas is then registered with the chest X-ray image, thereby creating a rib index for each pixel in the chest X-ray image. In other words, the algorithm calculates corresponding pixels between the statistical atlas and the chest X-ray image. The rib model in the statistical atlas is then aligned with the chest X-ray image by applying a transformation mapping.

[0027] According to an embodiment of the present invention, in step c), the plurality of intercostal spaces are identified in the patient-specific rib model.

[0028] In other words, the intercostal spaces can be identified in atlas space.

[0029] According to an embodiment of the present invention, step d) further comprises:

[0030] d1) Mapping the detected diaphragm into the patient-specific rib model based on the registration to generate the diaphragm line.

[0031] For example, the diaphragm can be detected by gradient detection, texture detection of the intrathoracic region, water filling algorithms, or by deploying knowledge learned from training images about expected shapes to be found in chest X-ray images (e.g., via convolutional neural networks).

[0032] In step d1), the detected diaphragm may be mapped into the patient-specific rib model using the mapping parameters of the above-mentioned transformation mapping.

[0033] According to an embodiment of the present invention, in step f), the inspiration index is determined based on the point score of the right diaphragm line representing the path of the diaphragm in the right lung field part and / or the left diaphragm line representing the path of the diaphragm in the left lung field part.

[0034] In other words, the inspiration index can be collected separately for the right lung or the left lung.

[0035] According to an embodiment of the present invention, the method further includes:

[0036] labeling the plurality of intercostal spaces with intercostal labels; and / or

[0037] Use rib labels to mark the detected ribs;

[0038] The score allocation rule is defined as follows:

[0039] (i) the correspondence between point scores and intercostal labels; and / or

[0040] (ii) Correspondence between point scores and rib labels.

[0041] For example, if a point falls within the tenth intercostal space, a point score of "10" may be assigned to that point. For example, if a point falls within the intercostal space between the ninth and tenth ribs, a point score of "9.5" (i.e., the average of the ninth and tenth rib labels) may be assigned to that point.

[0042] According to an embodiment of the present invention, the score assignment rule further defines a correspondence between a point score and a spatial relationship of a point in an intercostal space relative to two detected ribs defining the intercostal space. Alternatively, the score assignment rule provides an interpolated rib label and / or an intercostal label at each position in the statistical atlas, such that a point score for a given image pixel of the chest X-ray image can be obtained by registering the statistical atlas with the chest X-ray image.

[0043] In other words, the point scores of the points can be spatially interpolated at their relative positions within the intercostal spaces. For example, if the point is located in the middle of the ninth intercostal space, its point score can be "9". If the point is located at the upper rib of the ninth intercostal space, its point score can be "8.5". If the point is located at the lower rib of the ninth intercostal space, its point score can be "9.5". This can further improve the accuracy of estimating the position of each point of the diaphragm relative to the corresponding intercostal space. Alternatively, the statistical atlas can provide interpolated rib labels and / or interpolated intercostal labels at each position in the atlas space. The interpolated rib labels and / or interpolated intercostal labels in the atlas space can then be mapped to corresponding image pixels in the chest X-ray image by applying a transformation mapping (i.e., using registration). In this way, the point score for a given image pixel in the chest X-ray image can be directly read out using a transformation.

[0044] According to an embodiment of the invention, the inhalation index comprises at least one of a measure of central tendency of the point fractions of the diaphragm line.

[0045] Examples of measures of central tendency may include the mean, median, and mode.Thus, the complexity of monitoring the distribution of point scores along the diaphragm line may be reduced to a simple overall value.

[0046] A second aspect of the present invention provides a device for monitoring the inhalation quality of X-ray images. The device comprises:

[0047] input unit; and

[0048] processing unit;

[0049] Wherein, the input unit is configured to receive a chest X-ray image of a patient;

[0050] Wherein, the processing unit is configured to:

[0051] detecting ribs in the chest X-ray image;

[0052] identifying a plurality of intercostal spaces, each intercostal space representing a space between detected adjacent ribs;

[0053] detecting a diaphragm in the chest X-ray image and generating a diaphragm line comprising a plurality of points representing a path of the diaphragm in the chest X-ray image; and

[0054] applying a score assignment rule to assign a point score to each point of the septal line based on the corresponding intercostal space in which the point is located in the image, wherein the score assignment rule defines a correspondence between point scores to be assigned according to the intercostal space in which the point is located; and

[0055] An inspiration index is determined based on the point score of the point of the diaphragm line.

[0056] Thus, an apparatus is provided having advantages similar to those of the first aspect discussed above.

[0057] According to an embodiment of the present invention, the processing unit is further configured to compare the determined inhalation index with predetermined X-ray image inhalation quality monitoring rules specifying image inhalation status in terms of the inhalation index to generate an X-ray image inhalation quality metric.

[0058] This can facilitate the system operator's determination of image quality.

[0059] According to an embodiment of the present invention, the output unit is configured to output at least one of the determined inhalation index and the X-ray image inhalation quality metric.

[0060] In an example, the output unit may be configured to output at least one of the determined inspiratory index and the X-ray image inspiratory quality metric on a display. Thus, the inspiratory index and / or the X-ray image inspiratory quality metric may be displayed to a medical professional. In another example, the determined inspiratory index and / or the X-ray image inspiratory quality metric may be output as an irradiation warning light or a warning sound. In yet another example, the determined inspiratory index and / or the X-ray image inspiratory quality metric may be output as a bitmap "watermark" inserted into a portion of an X-ray image provided by the X-ray device, or the X-ray image inspiratory quality metric may simply be stored in the image's metadata (DICOM tag).

[0061] A third aspect of the present invention provides an X-ray imaging system. The X-ray imaging system comprises:

[0062] an X-ray image acquisition device having an X-ray source and an X-ray detector; and

[0063] An apparatus as described above and below;

[0064] The X-ray image acquisition device is configured to: acquire chest X-ray image data of a patient, and provide the X-ray image data to the apparatus for monitoring inhalation quality of X-ray images.

[0065] Thus, an X-ray imaging system is provided that has the capability to provide a user with an assessment of lung image quality.

[0066] According to a fourth aspect of the present invention, a computer program element for controlling an apparatus as described above and below is provided. The computer program element, when run by a processing unit, is adapted to perform the steps of the method as described above and below.

[0067] According to a fifth aspect of the invention, there is provided a computer readable medium having stored thereon a computer program element as described above and below.

[0068] These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter.

[0069] definition

[0070] As used herein, a "user" is a medical person who is at least partially involved in the imaging process in an administrative or organizational manner.

[0071] A "patient" is a human being to be imaged, or in a veterinary context, an animal (particularly a mammal) to be imaged. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] These and other aspects of the invention will be apparent from and further elucidated with reference to the embodiments described by way of example in the following description and with reference to the accompanying drawings, in which:

[0073] Figure 1A An example of a chest X-ray image is shown.

[0074] Figure 1B Another example of a chest X-ray image is shown.

[0075] Figure 2 A flow chart illustrating a method for monitoring inhalation quality of X-ray images according to some embodiments of the present disclosure is shown.

[0076] Figure 3A An example of a patient-specific rib model is shown.

[0077] Figure 3B An example of a patient-specific rib model with identified intercostal spaces is shown.

[0078] Figure 3C An example of a patient-specific rib model with an identified diaphragm is shown.

[0079] Figure 3D A simplified version of the right half of the thorax is shown, the level of inspiration of which results in a poor quality image.

[0080] Figure 4 A schematic block diagram illustrating an exemplary implementation of an apparatus for monitoring inhalation quality of X-ray images according to some embodiments of the present disclosure is shown.

[0081] Figure 5 Depicted are exemplary implementations of X-ray imaging systems according to some embodiments of the present disclosure.

[0082] It should be noted that these drawings are purely diagrammatic and are not drawn to scale. In the drawings, elements corresponding to elements already described may have the same reference numerals. Examples, embodiments, or optional features, whether or not indicated as non-limiting, should not be construed as limiting the claimed invention. DETAILED DESCRIPTION

[0083] Chest X-rays are typically acquired while the patient is fully inhaling and holding their breath. In images that do not conform to this rule, the lung parenchyma may appear different, with tissue clumping and becoming opaque, which can easily be interpreted as abnormal. European guidelines for X-ray acquisition state that a posteroanterior (PA) chest X-ray should be performed "during full inspiration (assessed by the position of the ribs above the diaphragm (either 6 anterior or 10 posterior)) and with apnea."

[0084] When assessing image quality with respect to the patient's inspiratory status, radiologists often rely on intuition and only count ribs in doubt. To automate this task, algorithms can be used to assess image quality criteria, such as inspiratory status. Automated quality checks can be more reproducible, robust, and accurate. When implementing these guidelines with an algorithm, counting ribs above the diaphragm proved challenging. Often, ribs close to the diaphragm intersect the diaphragm rather than being clearly located above or below it.

[0085] For example, Figure 1A and Figure 1B Two examples of chest X-ray images are shown. In particular, Figure 1A Shown is a situation where the determination is still quite easy. In this case, the rib centerline 10 does not intersect the diaphragm 12. Only the rib centerline 14 may be considered to intersect. Figure 1B Shown is a situation where the two rib counts on both sides may not agree when considering the ribs above the diaphragm according to the criteria. In this case, the two rib centerlines 14 on both sides intersect the diaphragm 12, and only the rib centerline completely above the diaphragm is the rib centerline 10.

[0086] Figure 2 A flow chart representing a method 100 for enhanced X-ray image inspiratory quality monitoring is shown, according to some embodiments of the present disclosure.

[0087] In step 110 (i.e., step a), a chest X-ray image of the patient is received. In one example, the chest X-ray image may be received from a digital recording tablet (i.e., a digital flat panel X-ray detector). In another example, the X-ray image may be received from a digital scanning device used to scan film exposures of a region of interest.

[0088] In step 120 (i.e., step b), ribs are detected in the chest X-ray image. There are many methods for segmenting ribs in chest radiographs. In an example, an edge-based method can be used to detect ribs. The method first uses an edge detection algorithm to extract rib pixels and then groups candidate pixels / lines into complete rib boundaries by applying, for example, curve fitting techniques or geometric models such as parabolas. In another example, a classification scheme is used to classify rib / non-rib pixels. For example, iterative background pixel classification can be used to segment posterior ribs in chest radiographs. Starting from the initial rib segmentation obtained based on pixel classification, the method updates the pixel classification by reclassifying each pixel based on the original features and additionally based on the type label information of the pixels in the neighborhood of the pixel to be classified. In another example, a segmentation method based on prior information can be used to detect ribs. One method of incorporating prior knowledge is to use a prototype atlas that includes a set of model chest X-ray images and their corresponding rib boundaries.

[0089] Combine Figure 3A To describe an exemplary embodiment of an atlas. Figure 3A A patient-specific rib model 20 of a patient is shown. To construct the patient-specific rib model, a statistical atlas is constructed by co-registering training images comprising a plurality of chest X-ray images of one or more patients. The training images may have manually drawn rib boundaries. A rib index may be assigned to each pixel in the statistical atlas by construction. The statistical atlas is then registered with the patient's chest X-ray image to construct the patient-specific rib model of the patient. In other words, the statistical atlas is registered to the patient's chest X-ray image, resulting in a transformation for each pixel, which allows the corresponding atlas rib mask to be transformed and processed into a segmentation result for the patient's ribs. In this way, the rib index for each pixel is transformed from the statistical atlas to the chest X-ray image, thereby forming a patient-specific rib model in the form of a density of rib centerlines 30 in atlas space.

[0090] Optionally, as Figure 3AAs shown, a rib labeling algorithm can be used to number the ribs. For example, the rib centerlines 30 can be labeled according to rib position. Conventionally, the rib closest to the patient's head is designated "1." "True ribs" are numbered downward as ribs 1 to 7. "False ribs" are numbered forward as 8 to 12. Ribs 11 and 12 are sometimes referred to as "floating ribs."

[0091] In step 130 (ie, step c), a plurality of intercostal spaces are identified. Each intercostal space represents a space between adjacent ribs detected.

[0092] In an example, if an edge-based method or classification scheme is used for rib detection in step b), intercostal spaces can be identified in the chest X-ray image.

[0093] In another example, if the atlas is used for rib detection, the intercostal spaces may be identified in a patient-specific rib model of the patient.

[0094] Combine Figure 3B An exemplary embodiment of an atlas for identifying intercostal spaces is described below. In this example, Figure 3A A plurality of intercostal spaces 40 are identified in the patient-specific rib model 20. Each intercostal space 40 corresponds to a respective space between adjacent rib centerlines 30.

[0095] Optionally, as Figure 3B As shown, the intercostal spaces can be numbered using an intercostal space labeling algorithm. For example, the intercostal space below the rib centerline X is designated as "10." The intercostal space between the rib centerlines IX and X is designated as "9."

[0096] In step 140 (i.e., step d)), a diaphragm line is generated that represents the path of the diaphragm in the chest X-ray image. Typically, in a chest X-ray image, the region of interest is the human chest. The diaphragm follows a curved path through the base of the thoracic region. The diaphragm divides different tissue textures and has a different shape, with a right-hand "dome" that is slightly higher than the dome on the left side of the diaphragm. Therefore, the line that follows the path of the diaphragm in the chest X-ray image is amenable to detection results obtained through image processing techniques. Although the human diaphragm is a single connected boundary, different parts of the diaphragm can also be detected, for example, a left part and a right part.

[0097] The diaphragm can be detected by gradient detection, texture detection of the intrathoracic region, water filling algorithms, or by deploying knowledge learned from training images about expected shapes to be found in the image. The shape of the diaphragm follows a curve and is therefore amenable to modeling using spline methods.

[0098] Combine Figure 3C To describe an exemplary example of a diaphragm. Similar to Figure 3A and Figure 3B , providing the diaphragm 50 in the atlas space. This can be done by first detecting the diaphragm in the chest X-ray image (e.g. by gradient detection or water filling algorithm) and then Figure 3A The registration in maps the detected septa into the atlas space with the same mapping. Figure 3C In the example shown, the left diaphragm line 50a and the right diaphragm line 50b are both located in the tenth intercostal space. Figure 3D In the poor quality inspiratory chest X-ray shown, the left diaphragm line 50a may be located in two or more intercostal spaces.

[0099] In step 150 (ie, step e)), a score assignment rule is applied to assign a point score to each point of the diaphragm line based on the corresponding intercostal space that the point is in. The score assignment rule defines the correspondence between the point scores and the intercostal spaces.

[0100] For example, the score assignment rule may define a correspondence between point scores and intercostal labels. In the example, the point score of a point of the diaphragm line is equal to the label of the intercostal space in which the point is located. For example, if a point of the diaphragm line is located in the tenth intercostal space, a point score of "10" is assigned to the point. Figure 3C In the exemplary example, since the left diaphragm line 50a is located in the tenth intercostal space, the point on the left diaphragm line 50a is assigned a point score of "10." This also applies to the right diaphragm line 50b. In another example, the score assignment rule may define a correspondence between point scores and rib labels, such that, for example, the point score of a point on the diaphragm line is equal to the average of the rib labels of the adjacent ribs that define the intercostal space in which the point is located.

[0101] Optionally, the score assignment rule may further define a correspondence between a point score and a spatial relationship of a point in the intercostal space relative to two detected ribs defining the intercostal space. In other words, the point score of a point may be further adjusted based on its relative position in the intercostal space. Figure 3D illustrates an exemplary embodiment of the correspondence between point scores and positions within the corresponding intercostal spaces, Figure 3D The diagram shows a simplified version of the right half of the thorax, where the inspiratory level results in a poor quality image. Point 60 of the diaphragm line 50b is located between the eighth and ninth rib centerlines. Point 60 has a distance "a" from the eighth rib centerline and a distance "b" from the ninth rib centerline. In this example, the point score for point 60 can be calculated using the following formula:

[0102]

[0103] where x is the intercostal label of the intercostal space where the point lies.

[0104] Alternatively, the score assignment rule provides an interpolated rib label and / or an interpolated intercostal label at each location in the statistical atlas, so that a point score for a given image pixel of the chest X-ray image can be obtained by registering the statistical atlas with the chest X-ray image. For example, the atlas provides an interpolated rib index at each location, and the interpolated rib index for a given image pixel can be directly read out using a transformation.

[0105] In step 160 (i.e., step f), an inspiration index is determined based on the score of points of the diaphragm line. Optionally, the inspiration index may be determined based on the score of points of the right diaphragm line 50b representing the path of the diaphragm in the right lung field portion and / or the left diaphragm line 50a representing the path of the diaphragm in the left lung field portion.

[0106] Examples of an aspiration index may include, but are not limited to, the mean, median, and mode of the point fractions (i.e., pixels) of the diaphragm line. In an example, a cumulative histogram may be constructed for all points of the left and / or right diaphragm line. For each point, the rib label counter for the rib in the rib label map is incremented at that point and all ribs above it. After processing all points, all rib label counters may be normalized by dividing by the number of points so that all rib label counters have values between 0 and 1. Finally, the median (or mode) of this distribution may be taken to serve as the aspiration index.

[0107] Optionally, an inhalation index can be output. In an example, the inhalation index can be output on a display (e.g., a computer display screen). This can allow an operator to gauge image quality. In another example, the inhalation index can be output for further analysis, e.g., to generate an X-ray image inhalation metric as explained below.

[0108] Optionally, a step 170 (i.e., step g)) may be provided for comparing the determined inhalation index with predetermined X-ray image inhalation quality monitoring rules that specify the image inhalation status in terms of the inhalation index to generate an X-ray image inhalation quality metric.

[0109] X-ray aspiration quality monitoring rules can be provided as data records associated with specific imaging standards (e.g., the "European Guidelines for Diagnostic Image Quality") to enable compliance with regional or national imaging standards. Alternatively or additionally, X-ray image aspiration quality monitoring rules can be pre-programmed by medical imaging professionals to predefined standards or specific protocols required by healthcare facilities.

[0110] The X-ray image inhalation quality metric can be a discrete category, for example, selected from the group consisting of "poor, fair, and good." Each category of the X-ray image inhalation quality metric can correspond to a specific range of inhalation indices. For example, the category "good" corresponds to inhalation indices in the range of 9.2 and above, the category "fair" corresponds to inhalation indices in the range of 8.1 to 9.2, and the category "poor" corresponds to inhalation indices in the remaining ranges. The X-ray image inhalation quality metric can also be a continuous value, for example, a percentage. This allows experienced operators to gauge image quality.

[0111] Optionally, the X-ray image inspiratory quality metric can be output. For example, the X-ray image inspiratory quality metric can be output on a display (e.g., a computer display screen or a display of a mobile device). In another example, the X-ray image inspiratory quality metric can be output as an illumination warning light or a warning sound. In another example, the X-ray image inspiratory quality metric can be output as a bitmap "watermark" inserted into a portion of the X-ray image provided by the X-ray device.

[0112] Figure 4 A schematic block diagram of an exemplary embodiment of an apparatus 200 for monitoring inhalation quality of X-ray images according to some embodiments of the present disclosure is shown.

[0113] The apparatus 200 includes an input unit 210 and a processing unit 220. The apparatus 200 may be part of or include an application specific integrated circuit (ASIC), an electronic circuit, a processor and / or memory running one or more software or firmware programs, combinational logic circuits, and / or other suitable components that provide the described functionality.

[0114] The input unit 210 is configured to receive a chest X-ray image of a patient, for example, from an analog or digital X-ray detector.

[0115] The processing unit 220 is configured to detect ribs in the chest X-ray image (e.g., using an edge detection algorithm, a classification scheme, or a segmentation method based on prior information (e.g., an atlas-based segmentation method). For example, the processing unit 220 can be configured to co-register training images including a plurality of chest X-ray images of one or more patients to construct a statistical atlas, and to register the statistical atlas with the chest X-ray image of the patient to construct a patient-specific rib model of the patient.

[0116] The processing unit 220 is further configured to identify a plurality of intercostal spaces. Each intercostal space represents a space between adjacent ribs detected. In an example, the processing unit 220 can be configured to identify a plurality of intercostal spaces in the patient-specific rib model.

[0117] The processing unit 220 is further configured to generate a diaphragm line representing the path of the diaphragm in the chest X-ray image. The diaphragm can be detected by gradient detection, texture detection of the intrathoracic region, a water-filling algorithm, or by deploying knowledge learned from training images of expected shapes to be found in the image. Optionally, the processing unit 220 can be configured to detect the diaphragm in the chest X-ray image (e.g., using gradient detection or a water-filling algorithm), and map the detected diaphragm onto the patient-specific rib model based on registration to generate the diaphragm line.

[0118] The processing unit 220 is further configured to apply a score assignment rule to assign a point score to each point of the diaphragm line based on the corresponding intercostal space in which the point is located in the image. The score assignment rule defines a correspondence between the point scores and the intercostal spaces. For example, the processing unit 220 can be configured to use intercostal labels to mark multiple intercostal spaces and / or use rib labels to mark detected ribs. The score assignment rule can define (i) a correspondence between the point scores and the intercostal labels and / or (ii) a correspondence between the point scores and the rib labels. Optionally, the score assignment rule can also define a correspondence between the point scores and the spatial relationship of the points in the intercostal space relative to the two detected ribs that define the intercostal space.

[0119] The processing unit 220 is further configured to determine an inspiration index based on the point scores of the diaphragm line. The inspiration index includes at least one of the measures of central tendency of the point scores of the diaphragm line, such as the mode, median or mean.

[0120] Optionally, the processing unit 220 may further be configured to compare the determined inhalation index with a predetermined X-ray image inhalation quality monitoring rule that specifies the image inhalation status in terms of the inhalation index to generate an X-ray image inhalation quality metric, which may be a discrete category, for example, a discrete category selected from the group of “very poor, moderate, good”.

[0121] Optionally, the apparatus 200 may further include an output unit 220. The output unit 220 may be configured to output at least one of the determined inspiratory index and the X-ray image inspiratory quality metric. For example, the determined inspiratory index and / or the X-ray image inspiratory quality metric may be output on a display.

[0122] Figure 5 An exemplary implementation of an X-ray imaging system 300 according to some embodiments of the present disclosure is illustrated.

[0123] The X-ray imaging system 300 includes an X-ray image acquisition device 310 having an X-ray source 312 and an X-ray detector 314. The X-ray source 312 (e.g., a rotating anode generator) can be positioned so as to emit radiation in the direction of the X-ray detector 314 (e.g., a pixelated flat-panel X-ray detector), where the radiation is received. Optionally, the X-ray detector 314 can be located in a stand-alone detector 320a or in a bed 320b. Thus, the patient can be imaged while standing or sitting.

[0124] The X-ray detector 310 may be connected to the device 200 via a physical cable or a wireless connection. Figure 5 In the exemplary example in , the device 200 is a personal computer (PC). In another example, the device 200 can be a server with more powerful computing power to provide services to multiple users, thereby performing image quality assessment in many examination rooms. In other words, the image quality assessment algorithm can be provided on a server (e.g., a cloud server). In this case, a chest X-ray image taken by the X-ray imaging device 310 can be sent to the device 200 configured as a remote server, and the remote server identifies the inhalation index and optionally also identifies the X-ray image inhalation quality metric. The corresponding results can be sent to a PC or mobile device for display to the user. In another example, the image quality assessment algorithm can be used for a mobile device (e.g., a mobile computing device).

[0125] Optionally, the apparatus 200 may be configured to output at least one of the determined inhalation index and the X-ray image inhalation quality metric to a user via an output 330. In the exemplary embodiment, the output 330 is a display. In other examples, the output may be a light or a simple audio warning.

[0126] Optionally, the apparatus 200 may be provided with a user interface 340 (e.g., a keyboard 340a and a mouse 340b). Other examples of user interfaces may include, but are not limited to, a touch screen and a joystick device. The user interface 340 may allow the user to define score assignment rules and / or X-ray image inspiratory quality monitoring rules. The user interface 340 may be arranged remotely from the X-ray imaging device 310, for example, outside the imaging room. Alternatively, the user interface may be arranged inside the room.

[0127] In use, a region of interest of a patient is positioned in front of the X-ray source 312. The X-ray source 312 is activated (e.g., via the user interface 340), and the patient is exposed. As described above, the apparatus 200 assesses image quality. The apparatus 200 generates an inspiratory index and, optionally, an X-ray image inspiratory quality metric, which is then output to the user on the display 330. In this manner, a more reliable and easier assessment of the image quality of the chest X-ray images is achieved. The calculation of the inspiratory index for each chest X-ray image can be performed without human oversight or interference, thereby avoiding dependence on variations or deviations in the skill level of different system operators.

[0128] In a further exemplary embodiment of the present invention, a computer program or a computer program element is provided, characterized in that it is adapted to execute the method steps of the method according to one of the preceding embodiments on a suitable system.

[0129] Therefore, a computer program element can be stored in a computer unit, which can also be part of an embodiment of the present invention. The computer unit can be adapted to execute or cause the execution of the steps of the above-described method. In addition, the computer unit can be adapted to operate components of the above-described apparatus. The computer unit can be adapted to operate automatically and / or execute user commands. The computer program can be loaded into the working memory of a data processor. Thus, the data processor can be equipped to execute the method of the present invention.

[0130] This exemplary embodiment of the invention covers both a computer program that right from the beginning uses the invention and a computer program that is converted into a program that uses the invention by means of an update of an existing program.

[0131] Furthermore, the computer program element may be able to provide all necessary steps to fulfill the flow of an exemplary embodiment of the method as described above.

[0132] According to a further exemplary embodiment of the present invention, a computer-readable medium, for example a CD-ROM, is proposed, wherein the computer-readable medium has a computer program element stored on the computer-readable medium, the computer program element being described in the preceding sections.

[0133] The computer program may be stored and / or distributed on suitable media, 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 telecommunication systems.

[0134] However, the computer program may also be present on a network, like the World Wide Web, and can be downloaded from such a network into the working memory of a data processor. According to a further exemplary embodiment of the present invention, a medium is provided for making a computer program element available for downloading, said computer program element being arranged to perform a method according to one of the previously described embodiments of the present invention.

[0135] It should be noted that embodiments of the present invention are described with reference to different subject matters. In particular, some embodiments are described with reference to method claims, while other embodiments are described with reference to apparatus claims. However, unless otherwise indicated, a person skilled in the art will infer from the above and following descriptions that, in addition to any combination of features belonging to one type of subject matter, any combination of features relating to different subject matters is also considered to be disclosed in this application. However, all features can be combined to provide synergistic effects that are more than the simple sum of the features.

[0136] Although the present invention has been illustrated and described in detail in the drawings and the foregoing description, such illustration and description should be considered illustrative or exemplary rather than restrictive. The present invention is not limited to the disclosed embodiments. Other variations of the disclosed embodiments will be understood and effected by those skilled in the art in practicing the claimed invention by studying the drawings, the disclosure, and the claims.

[0137] In the claims, the word "comprising" does not exclude other elements or steps, and the word "a" or "an" does not exclude a plurality. A single processor or other unit may perform the functions of several items recited in a claim. The fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. Any reference signs in the claims should not be construed as limiting the scope.

Claims

1. A computer-implemented method (100) for monitoring the quality of inhalation in X-ray images, comprising the following steps: a) receiving (110) a chest X-ray image of a patient; b) detecting (120) ribs in the chest X-ray image; c) identifying (130) a plurality of intercostal spaces, each intercostal space representing a space between detected adjacent ribs; d) detecting a diaphragm in the chest X-ray image and generating (140) a diaphragm line, the diaphragm line comprising a plurality of points representing a path of the diaphragm in the chest X-ray image; as well as e) applying (150) a score assignment rule to assign a point score to each point of the diaphragm line based on the corresponding intercostal space in which the point is located in the image, wherein the score assignment rule defines a correspondence between a point score to be assigned based on the intercostal space in which the point is located and the intercostal space in which the point is located; and f) determining (160) an inspiration index based on the point score of the point of the diaphragm line.

2. The computer-implemented method of claim 1 , further comprising: g) comparing the determined inhalation index with predetermined X-ray image inhalation quality monitoring rules specifying an image inhalation state with respect to said inhalation index (170) to generate an X-ray image inhalation quality metric.

3. The computer-implemented method according to claim 1 or 2, in, Step b) further comprises: b1) constructing a statistical atlas by co-registration of training images comprising a plurality of chest X-ray images of one or more patients; and b2) registering the statistical atlas with the chest X-ray image of the patient to construct a patient-specific rib model of the patient.

4. The computer-implemented method of claim 3, in, In step c), the plurality of intercostal spaces are identified in the patient-specific rib model.

5. The computer-implemented method of claim 3, wherein: Step d) further comprises: d1) Mapping the detected diaphragm into the patient-specific rib model based on the registration to generate the diaphragm line.

6. The computer-implemented method of claim 4, in, Step d) further comprises: d1) Mapping the detected diaphragm into the patient-specific rib model based on the registration to generate the diaphragm line.

7. A computer-implemented method according to any one of claims 1, 2, 4, 5 and 6, in, In step f), the inspiration index is determined based on the point fraction of a right diaphragm line representing the path of the diaphragm in the right lung field portion and / or a left diaphragm line representing the path of the diaphragm in the left lung field portion.

8. The computer-implemented method of any one of claims 1, 2, 4, 5, and 6, further comprising: labeling the plurality of intercostal spaces using intercostal labels; and / or Use rib labels to mark the detected ribs; The score allocation rule is defined as follows: (i) Correspondence between point scores and intercostal labels; and / or (ii) Correspondence between point scores and rib labels.

9. A computer-implemented method according to any one of claims 4, 5 and 6, in, The score assignment rule further defines a correspondence between: i) a point score, and ii) a spatial relationship of a point in an intercostal space relative to two detected ribs defining the intercostal space; or The score assignment rule provides an interpolated rib label and / or an interpolated intercostal label at each position in the statistical atlas, so that a point score for a given image pixel of the chest X-ray image can be obtained by aligning the statistical atlas with the chest X-ray image.

10. The computer-implemented method of any one of claims 1, 2, 4, 5, and 6, in, The aspiration index includes at least one of a measure of central tendency of the point fraction of the diaphragm line.

11. A device (200) for monitoring the inhalation quality of an X-ray image, comprising: Input unit (210); as well as a processing unit (220); Wherein, the input unit is configured to receive a chest X-ray image of a patient; The processing unit is configured to: detecting ribs in the chest X-ray image; identifying a plurality of intercostal spaces, each intercostal space representing a space between detected adjacent ribs; detecting a diaphragm in the chest X-ray image and generating a diaphragm line comprising a plurality of points representing a path of the diaphragm in the chest X-ray image; and applying a score assignment rule to assign a point score to each point of the diaphragm line based on the corresponding intercostal space in which the point is located in the image, wherein the score assignment rule defines a correspondence between the point score to be assigned based on the intercostal space in which the point is located and the intercostal space in which the point is located; and An inspiration index is determined based on the point score of the point of the diaphragm line.

12. The device according to claim 11, in, The processing unit is further configured to compare the determined inhalation index with predetermined X-ray image inhalation quality monitoring rules specifying an image inhalation state in terms of the inhalation index to generate an X-ray image inhalation quality metric.

13. The apparatus according to claim 11 or 12, further comprising: An output unit (230) is configured to output at least one of the determined inhalation index and the X-ray image inhalation quality metric.

14. An X-ray imaging system (300), comprising: an X-ray image acquisition device (310) having an X-ray source (312) and an X-ray detector (314); as well as The device according to any one of claims 11 to 13; The X-ray image acquisition device is configured to: acquire chest X-ray image data of a patient, and provide the X-ray image data to the apparatus for monitoring inhalation quality of X-ray images.

15. A computer program product for controlling an apparatus according to one of claims 11 to 13, said computer program product being adapted to perform the steps of the computer-implemented method according to any of claims 1 to 10 when run by a processing unit.

16. A computer readable medium having stored thereon the computer program product of claim 15.

Images