Control unit of a medical X-ray machine

Abstract

Method for controlling a medical X-ray device (37), comprising: a) Taking (ACQ-RB) at least one X-ray image (RB) of an examination area of ​​an examination object (31) using the medical X-ray device (37), wherein a medical object is positioned in the examination area, b) Generating (DET-OB) an object image (OB) by averaging at least a subset (RB.sub) of the previously recorded X-ray images (RB.1, RB.2, ... RB.n), c) Determining (DET-BP) a determinability parameter (BP) to assess the determinability of the medical object based on the object image (OB), whereby the procedure is repeated from step a) until (E) a termination condition (AB) occurs based on the last determined determinability parameter (BP).

Description

[0001] The invention relates to a method for controlling a medical X-ray device, a computer-implemented method for providing a trained function, a computer-implemented method for providing a further trained function, a medical X-ray device, a training unit, a computer program product, and a computer-readable storage medium.

[0002] For imaging monitoring of a treatment, particularly a cardiovascular treatment, on an examination subject, a reliable and qualitatively sufficient image of the examination area of ​​the examination subject is necessary. To image the examination area, particularly a medical object arranged therein, such as a guidewire and / or a catheter and / or an endoscope and / or a laparoscope and / or an implant, particularly a stent, as quickly and as promptly as possible, X-ray-based imaging is frequently used. To improve image quality, particularly the signal-to-noise ratio (SNR) and / or the contrast-to-noise ratio (CNR), a particularly high X-ray dose is often selected with a short application time.However, the technical limitations of the X-ray source, especially the X-ray tube, in terms of the maximum possible output are often a disadvantage.

[0003] Alternatively or additionally, the image quality can be improved by averaging a series of X-ray images. For this purpose, a defined number of X-ray images can be taken of the examination area in chronological sequence. The individual X-ray images in the series often have an image quality that is insufficient to identify the medical object. Furthermore, the individual X-ray images in the series can each depict at least partially different examination areas, particularly due to a movement of at least part of the examination object. By retrospectively applying motion correction and / or registration, as well as averaging the series of X-ray images, improved image quality can often be achieved in the resulting image.The disadvantage, however, is that the defined number of X-ray images to be taken must be specified, and the image quality of the resulting image is only apparent after the entire X-ray dose has been applied. However, different examination subjects often require an adjustment of the number of X-ray images to be taken in order to achieve sufficient image quality in the resulting image.

[0004] Document DE 10 2011 090 047 A1 discloses a control method for determining a quality indicator of medical imaging result data from a contrast-enhanced tomography scan of an examination structure using a tomography system. Furthermore, document DE 10 2010 028 446 A1 discloses a method for reconstructing a current three-dimensional image data set of an object during a monitoring process. Furthermore, document DE 10 2017 217 543 A1 discloses a method for creating a classification unit for the automatic classification of materials.

[0005] The invention is therefore based on the object of minimizing the X-ray dose when imaging a medical object.

[0006] The object is achieved according to the invention by the respective subject matter of the independent claims. Advantageous embodiments with expedient further developments are the subject matter of the dependent claims.

[0007] The inventive solution to the problem is described below both with regard to methods for controlling a medical X-ray device and with regard to methods and devices for providing a trained function and methods and devices for providing a further trained function. Features, advantages, and alternative embodiments of data structures and / or functions in methods for controlling a medical X-ray device can be transferred to analog data structures and / or functions in methods and devices for providing a trained function, and methods and devices for providing a further trained function can be transferred. Analog data structures can be identified in particular by the use of the prefix "training."Furthermore, the trained functions used in methods for controlling a medical X-ray device can have been adapted and / or provided in particular by methods and devices for providing a trained function and / or by methods for providing a further trained function.

[0008] In a first aspect, the invention relates to a method for controlling a medical X-ray device. In a first step a), at least one X-ray image of an examination region of an examination object is recorded using the medical X-ray device, wherein a medical object is arranged in the examination region. Furthermore, in a second step b), an object image is generated based on the at least one X-ray image. Thereafter, in a third step c), a determinability parameter for evaluating a determinability of the medical object is determined based on the object image. Furthermore, the proposed method is repeatedly executed starting from step a) until a termination condition occurs based on the last determined determinability parameter.

[0009] The at least one X-ray image can comprise a plurality of image points, in particular pixels and / or voxels. Advantageously, the at least one X-ray image can at least partially depict the examination area of ​​the subject under examination.

[0010] The examination object can be, for example, a human and / or animal patient and / or an x-ray phantom. Furthermore, the at least one x-ray image can comprise a two-dimensional projection image and / or a three-dimensional image of the examination region. Furthermore, a plurality of x-ray images can be acquired using the medical x-ray device, wherein the plurality of x-ray images can at least partially have different acquisition parameters, for example an angulation and / or a field-of-view (FOV). Furthermore, the at least one x-ray image can comprise metadata, wherein the metadata can comprise information relating to acquisition parameters and / or operating parameters and / or physiological parameters of the examination object and / or patient information.

[0011] Advantageously, the medical object, for example a guide wire and / or a catheter and / or an endoscope and / or a laparoscope and / or an implant, in particular a stent, which is arranged in the examination area of ​​the examination object, can be imaged in the at least one X-ray image.

[0012] To acquire the at least one x-ray image of the examination region of the examination subject to be imaged in step a), the medical x-ray device can comprise an x-ray source and a detector unit. Furthermore, the x-ray source can emit an x-ray beam, in particular a cone beam and / or fan beam and / or parallel beam, whereby the examination region is illuminated by the x-ray beam. Upon impact of the x-ray beam, following interaction with the examination region of the examination subject to be imaged, on a surface of the detector unit, the detector unit can provide a signal. The at least one x-ray image can be reconstructed and / or provided based on the signal.

[0013] In step b), the object image can be generated based on the at least one x-ray image. The object image can advantageously comprise a segmented, in particular isolated, two-dimensional and / or three-dimensional image of the medical object and / or a marker structure, for example a balloon marker, arranged in particular on the medical object. Advantageously, generating the object image based on the at least one x-ray image can comprise filtering and / or segmenting and / or identification of the medical object and / or the marker structure. For example, the medical object and / or the marker structure can be segmented and / or identified based on a geometric shape, in particular a contour, and / or based on a comparison of image intensity values, in particular with a predetermined threshold value, in the at least one x-ray image.

[0014] Advantageously, a determinability parameter can be determined in step c) based on the object image. The determinability parameter can advantageously include information regarding the confidence and / or image quality of the object image. The determinability parameter can include at least one value, in particular a numerical value, that evaluates the determinability of the medical object from the at least one X-ray image, in particular during the generation of the object image in step b). In particular, the determinability parameter can include information regarding the confidence during the filtering and / or segmentation and / or identification of the medical object and / or the marker structure in the at least one X-ray image.Furthermore, the determinability parameter can comprise information regarding the image quality of the object image and / or the at least one x-ray image, for example, a contrast-to-noise ratio and / or a signal-to-noise ratio and / or image sharpness. Furthermore, the determinability parameter can comprise information regarding the characteristics of the medical object in the at least one x-ray image and / or the object image. The information regarding the characteristics of the medical object in the at least one x-ray image and / or the object image can comprise, for example, an intensity ratio and / or a contrast value between pixels corresponding to the image of the medical object and other pixels.

[0015] Furthermore, the determinability parameter can comprise a determinability matrix, in particular a two-dimensional and / or three-dimensional one, wherein the values ​​of the determinability matrix evaluate the determinability of the medical object in at least one subregion of the object image. In particular, the determinability matrix can be determined pixel by pixel.

[0016] The termination condition can advantageously comprise a comparison of the determinability parameter last determined in step c), in particular during the current repetition, with a predefined threshold value. If the determinability parameter comprises multiple values ​​and / or a determinability matrix, the termination condition can comprise a comparison of the multiple values ​​and / or the individual values ​​of the determinability matrix with the predefined threshold value. The proposed method for controlling the medical X-ray device can be repeatedly executed, starting from step a), based on the last determined determinability parameter until the termination condition occurs. Furthermore, the termination condition can occur, for example, when the determinability parameter exceeds and / or falls below and / or reaches the predefined threshold value.Furthermore, the termination condition may have a maximum number of repetitions, wherein the termination condition occurs when the maximum number of repetitions is reached and / or exceeded.

[0017] In particular, in the case of temporally pulsed X-ray emission by the X-ray source, a time-limited X-ray pulse and / or a defined number of X-ray pulses can be emitted by the X-ray source in step a). Upon the occurrence of the termination condition, radiation can advantageously be terminated, in particular, the temporally pulsed X-ray emission by the X-ray source can be stopped.

[0018] Furthermore, when steps a) to c) are carried out repeatedly, at least one determinability parameter for the at least one X-ray image, in particular the last one taken, can advantageously be determined based on the object image.

[0019] Furthermore, the generation of the object image in step b) can additionally be based on at least one x-ray image acquired in a previous repetition during a repeated execution of the proposed method. This advantageously allows an improvement in the determinability parameter to be achieved during the repeated execution of steps a) to c).

[0020] By evaluating the identifiability of the medical object after each recording of at least one X-ray image, the duration of the examination and / or the applied X-ray dose can advantageously be reduced to a minimum necessary for the identification of the medical object.

[0021] Furthermore, the object image is generated by a, in particular weighted and / or adaptive, averaging of at least a subset of the X-ray images taken so far.

[0022] When steps a) to c) are repeatedly performed, the object image in step b) can advantageously be generated based on at least a subset of the x-ray images acquired so far, in particular in each of the previous repetitions in step a). The at least one subset can, in particular, be a subset that includes at least one or all of the x-ray images acquired so far. Furthermore, the object image can be generated based on multiple subsets of the x-ray images acquired so far.

[0023] In this case, the generation of the object image in step b) can comprise a, in particular weighted and / or adaptive, averaging of at least a subset of the x-ray images acquired so far. With a weighted averaging of at least a subset of the x-ray images acquired so far, the x-ray images can advantageously be weighted depending on their respective image quality, in particular a signal-to-noise ratio and / or a contrast-to-noise ratio. Furthermore, an adaptive averaging can take place in at least one image region of the x-ray images that contains an image of the medical object. In this case, the averaging of at least a subset of the x-ray images acquired so far can take place in particular by means of the detector unit of the medical x-ray device, for example by binning at least some of the detector pixels.

[0024] If the object image is generated based on several subsets of the x-ray images taken so far, an intermediate image can be generated for each of the subsets, for example by means of a, in particular weighted and / or adaptive, averaging, wherein the object image can be generated by a, in particular weighted and / or adaptive, averaging of the intermediate images.

[0025] This can advantageously lead to an improvement in the image quality of the object image, in particular the determinability parameter.

[0026] In a further advantageous embodiment of the proposed method for controlling a medical X-ray device, the at least one subset of the X-ray images taken so far can be determined for generating the object image based on the determinability parameter.

[0027] Advantageously, upon repeated execution of steps a) to c), at least one determinability parameter can be determined for the at least one x-ray image, in particular the most recently acquired one, based on the object image. The at least one subset of the x-ray images acquired so far can be further determined for generating the object image in step b) based on the at least one determinability parameter. In particular, determining the at least one subset of the x-ray images acquired so far can comprise a comparison of the determinability parameters of the x-ray images acquired so far with a threshold value, in particular an average value of the determinability parameters.If the determinability parameter evaluates the determinability of the medical object based on the object image in such a way that the value of the determinability parameter increases monotonically with the determinability, advantageously only those X-ray images acquired so far whose determinability parameter has a value above the threshold value, in particular the mean value of the determinability parameters, can be included in the at least one subset for generating the object image. Furthermore, the X-ray images that have a determinability parameter with a value below the threshold value, in particular the mean value of the determinability parameters, can be excluded from the at least one subset for generating the object image. If there is a monotonically decreasing dependence between the value of the determinability parameter and the determinability of the medical object, the comparison can be adjusted analogously.

[0028] This advantageously allows an improvement, in particular an increase, of the determinability parameter of the current repetition. Furthermore, X-ray images that exhibit image artifacts and / or interference and / or motion artifacts can advantageously be excluded from the at least one subset for generating the object image.

[0029] The determination of the at least one subset for generating the object image may further include taking into account the possible permutations of the x-ray images taken so far with the respective determinability parameters.

[0030] In a further advantageous embodiment of the proposed method for controlling a medical X-ray device, the termination condition can comprise a comparison of the determinability parameter with a predetermined threshold value. The comparison of the determinability parameter, in particular of at least one value of the determinability parameter, with the predetermined threshold value can comprise a difference and / or a quotient and / or a scalar product. Advantageously, the threshold value can be specified such that the medical object in the object image can be fully determined, in particular captured and / or identified and / or localized, in particular by operating personnel, when the termination condition occurs.

[0031] This can advantageously reduce the duration of the examination and / or the applied X-ray dose.

[0032] In a further advantageous embodiment of the proposed method for controlling a medical X-ray device, generating the object image can comprise a motion correction and / or registration of the at least one X-ray image. The motion correction and / or registration can advantageously comprise a rigid and / or non-rigid transformation of the at least one X-ray image. When acquiring multiple X-ray images in step a) and / or when performing steps a) to c) repeatedly, the multiple X-ray images can preferably be registered with one another. The registration can be based, for example, on geometric and / or anatomical landmarks and / or a marker structure arranged on the medical object.This can improve the image quality of the object image, which is advantageously generated based on at least a subset of the X-ray images acquired so far. Furthermore, the determinability parameter can be advantageously improved.

[0033] Furthermore, the motion correction can be performed based on geometric and / or anatomical landmarks in the at least one X-ray image. Alternatively or additionally, the motion correction can be performed based on a physiological signal, such as a heart rate and / or respiratory rate, from the subject under examination. Furthermore, the physiological signal can be derived from the at least one X-ray image and / or metadata contained therein.

[0034] In a further advantageous embodiment of the proposed method for controlling a medical X-ray device, the determination of the determinability parameter in step c) can be carried out by applying a trained function to input data. The input data can be based on the object image. Furthermore, at least one parameter of the trained function can be based on a comparison of a training determinability parameter with a comparison determinability parameter.

[0035] The trained function can advantageously be trained using a machine learning method. In particular, the trained function can be a neural network, in particular a convolutional neural network (CNN) or a network comprising a convolutional layer.

[0036] The trained function maps input data to output data. In particular, the output data can continue to depend on one or more parameters of the trained function. The one or more parameters of the trained function can be determined and / or adapted through training. The determination and / or adaptation of the one or more parameters of the trained function can be based in particular on a pair of training input data and associated training output data, wherein the trained function is applied to the training input data to generate training mapping data. In particular, the determination and / or adaptation can be based on a comparison of the training mapping data and the training output data. In general, a trainable function, i.e., a function with one or more parameters that have not yet been adapted, is also referred to as a trained function.

[0037] Other terms for trained functions are trained mapping rule, mapping rule with trained parameters, function with trained parameters, algorithm based on artificial intelligence, and machine learning algorithm. An example of a trained function is an artificial neural network, where the edge weights of the artificial neural network correspond to the parameters of the trained function. Instead of the term "neural network," the term "neural network" can also be used. In particular, a trained function can also be a deep artificial neural network. Another example of a trained function is a "support vector machine." Furthermore, other machine learning algorithms can also be used as trained functions.

[0038] The trained function can be trained, in particular, by means of backpropagation. First, training image data can be determined by applying the trained function to training input data. Subsequently, a deviation between the training image data and the training output data can be determined by applying an error function to the training image data and the training output data. Furthermore, at least one parameter, in particular a weighting, of the trained function, in particular of the neural network, can be iteratively adjusted based on a gradient of the error function with respect to the at least one parameter of the trained function. This advantageously minimizes the deviation between the training image data and the training output data during training of the trained function.

[0039] Advantageously, the trained function, in particular the neural network, has an input layer and an output layer. The input layer can be configured to receive input data. Furthermore, the output layer can be configured to provide mapping data. The input layer and / or the output layer can each comprise multiple channels, in particular neurons.

[0040] Preferably, at least one parameter of the trained function can be based on a comparison of a training determinability parameter with a comparison determinability parameter. The training determinability parameter and / or the comparison determinability parameter can advantageously be determined as part of a proposed computer-implemented method for providing a trained function, which will be explained later in the description. In particular, the trained function can be provided by an embodiment of the proposed computer-implemented method for providing a trained function.

[0041] This can improve the assessment of the determinability of the medical object, in particular an identification and / or localization and / or detectability, in the object image.

[0042] In a further advantageous embodiment of the proposed method for controlling a medical X-ray device, the object image can be generated in step b) by applying a further trained function to further input data. The further input data can be based on the at least one X-ray image. Furthermore, at least one parameter of the further trained function can be based on a comparison of a further training object image with a comparison object image.

[0043] The advantages and properties of the additional trained function described above essentially correspond to the advantages and properties of the trained function. The features, advantages, or alternative embodiments mentioned with regard to the trained function can also be transferred to the additional trained function, and vice versa.

[0044] Preferably, at least one parameter of the further trained function can be based on a comparison of a further training object image with a comparison object image. The further training object image and / or the comparison object image can advantageously be determined as part of a proposed computer-implemented method for providing a further trained function, which will be explained later in the description. In particular, the further trained function can be provided by an embodiment of the proposed computer-implemented method for providing a further trained function.

[0045] This enables particularly reliable and robust generation of the object image.

[0046] In a further advantageous embodiment of the proposed method for controlling a medical X-ray device, the further trained function can also be configured to determine the determinability parameter. At least one further parameter of the further trained function can be based on a comparison of a training determinability parameter with a comparison determinability parameter.

[0047] Preferably, the at least one further parameter of the further trained function can be based on a comparison of a training determinability parameter with a comparison determinability parameter. The training determinability parameter and / or the comparison determinability parameter can advantageously be determined as part of a proposed computer-implemented method for providing a further trained function, which will be explained later in the description. In particular, the further trained function can be provided by an embodiment of the proposed computer-implemented method for providing a further trained function.

[0048] This can improve the assessment of the determinability of the medical object, in particular an identification and / or localization and / or detectability, in the object image.

[0049] In a second aspect, the invention relates to a computer-implemented method for providing a trained function. In a first step, at least one training x-ray image is received from an examination region of an examination object, wherein a medical object is arranged in the examination region. The medical object comprises a guide wire and / or a catheter and / or an endoscope and / or a laparoscope and / or an implant. In a second step, a training object image is generated based on the at least one training x-ray image. Furthermore, in a third step, a comparison determinability parameter for evaluating a determinability of the medical object based on the training object image is determined. In a fourth step, a training determinability parameter is determined by applying the trained function to input data.The input data is based on at least one training X-ray image. In a fifth step, at least one parameter of the trained function is adjusted based on a comparison of the training determinability parameter with the comparison determinability parameter. The trained function is then provided.

[0050] Receiving the at least one training x-ray image can, in particular, comprise capturing and / or reading from a computer-readable data storage device and / or receiving from a data storage unit, for example, a database. Furthermore, the at least one training x-ray image can be provided by a provision unit of a medical x-ray device.

[0051] The at least one training X-ray image can, in particular, have all the properties of the at least one X-ray image described in relation to the method for controlling a medical X-ray device, and vice versa. In particular, the at least one training X-ray image can be an X-ray image. Furthermore, the at least one training X-ray image can be simulated.

[0052] The at least one training x-ray image can comprise a plurality of image points, in particular pixels and / or voxels. Advantageously, the at least one training x-ray image can at least partially depict the examination region of the examination subject. The at least one training x-ray image can comprise a two-dimensional projection image and / or a three-dimensional image of the examination region. Furthermore, a plurality of training x-ray images can be received, wherein the plurality of training x-ray images can at least partially comprise different acquisition parameters, for example an angulation and / or an acquisition region. Furthermore, the at least one training x-ray image can comprise metadata, wherein the metadata can comprise information relating to acquisition parameters and / or operating parameters and / or physiological parameters of the examination subject and / or patient information.

[0053] Advantageously, the medical object, for example a guide wire and / or a catheter and / or an endoscope and / or a laparoscope and / or an implant, in particular a stent, which is arranged in the examination region of the examination object, can be depicted in the at least one training X-ray image.

[0054] The medical object described here, which is depicted in the at least one training X-ray image, can be the same as or different from the medical object depicted in the at least one X-ray image described in relation to the method for controlling a medical X-ray device, and vice versa. Similarly, the examination object and / or the examination region in which the medical object is arranged and which is depicted in the at least one training X-ray image can be the same as or different from the examination object and / or examination region described in relation to the method for controlling a medical X-ray device, and vice versa.

[0055] The training object image can be generated based on the at least one training X-ray image in a manner analogous to the generation of the object image in step b) of the proposed method for controlling a medical X-ray device. The training object image can, in particular, have all the properties of the object image described in relation to the method for controlling a medical X-ray device, and vice versa.

[0056] The training object image can advantageously comprise a segmented, in particular isolated, two-dimensional and / or three-dimensional image of the medical object and / or of a marker structure, in particular arranged on the medical object. Advantageously, generating the training object image based on the at least one training X-ray image can comprise filtering and / or segmenting and / or identifying the medical object and / or the marker structure. For example, the medical object and / or the marker structure can be segmented and / or identified based on a geometric shape, in particular a contour, and / or based on a comparison of image intensity values, in particular with a predetermined threshold value, in the at least one training X-ray image.

[0057] Furthermore, the comparative determinability parameter can be determined analogously to step c) of the proposed method for controlling a medical X-ray device based on the training object image. The comparative determinability parameter can, in particular, have all the properties of the determinability parameter described in relation to the method for controlling a medical X-ray device, and vice versa. Furthermore, the comparative determinability parameter can be determined by manual and / or semi-automatic annotation based on the training object image.

[0058] By applying the trained function to input data based on the at least one training X-ray image, the training determinability parameter can advantageously be determined. At least one parameter of the trained function can be adjusted based on a comparison of the training determinability parameter with the comparison determinability parameter. Furthermore, the comparison can include, for example, a difference and / or a scalar product and / or a quotient.

[0059] Providing the trained function may in particular comprise storing it on a computer-readable storage medium and / or transmitting it to a provisioning unit.

[0060] Advantageously, the method proposed here can provide a trained function which can be used in one embodiment of the method for controlling a medical X-ray device.

[0061] In a third aspect, the invention relates to a computer-implemented method for providing a further trained function. In a first step, at least one training x-ray image is received from an examination region of an examination object, wherein a medical object is arranged in the examination region. The medical object comprises a guide wire and / or a catheter and / or an endoscope and / or a laparoscope and / or an implant. In a second step, a comparison object image is generated based on the at least one training x-ray image. Furthermore, in a third step, a further training object image is generated by applying the further trained function to further input data. The further input data is based on the at least one training x-ray image.In a fourth step, at least one parameter of the further trained function is adjusted based on a comparison of the further training object image with the comparison object image. The further trained function is then provided.

[0062] Receiving the at least one training x-ray image can, in particular, comprise capturing and / or reading from a computer-readable data storage device and / or receiving from a data storage unit, for example, a database. Furthermore, the at least one training x-ray image can be provided by a provision unit of a medical x-ray device.

[0063] The medical object described here, which is depicted in the at least one training X-ray image, can be the same as or different from the medical object depicted in the at least one X-ray image described in relation to the method for controlling a medical X-ray device, and vice versa. Similarly, the examination object and / or the examination region in which the medical object is arranged and which is depicted in the at least one training X-ray image can be the same as or different from the examination object and / or examination region described in relation to the method for controlling a medical X-ray device, and vice versa.

[0064] The at least one training X-ray image can, in particular, have all the properties of the at least one X-ray image described in relation to the method for controlling a medical X-ray device, and vice versa. In particular, the at least one training X-ray image can be an X-ray image. Furthermore, the at least one training X-ray image can be simulated.

[0065] The comparison object image can be generated analogously to the generation of the object image in step b) of the proposed method for controlling a medical X-ray device based on the at least one training X-ray image. The comparison object image can, in particular, have all the properties of the object image described in relation to the method for controlling a medical X-ray device, and vice versa.

[0066] By applying the further trained function to the further input data based on the at least one training X-ray image, the further training object image can advantageously be generated. At least one parameter of the further trained function can be adjusted based on a comparison of the further training object image with the comparison object image. Furthermore, the comparison can include, for example, a difference and / or a scalar product and / or a quotient.

[0067] Providing the further trained function may in particular comprise storing it on a computer-readable storage medium and / or transmitting it to a provisioning unit.

[0068] Advantageously, the method proposed here can provide a further trained function which can be used in one embodiment of the method for controlling a medical X-ray device.

[0069] Furthermore, a comparative determinability parameter is determined for evaluating the determinability of the medical object based on the further training object image. Furthermore, a training determinability parameter is determined by applying the further trained function to the further input data. Subsequently, at least one further parameter of the further trained function is adjusted based on a comparison of the training determinability parameter with the comparative determinability parameter.

[0070] The comparative determinability parameter can be determined analogously to step c) of the proposed method for controlling a medical X-ray device based on the training object image. The comparative determinability parameter can, in particular, have all properties of the determinability parameter that were described in relation to the method for controlling a medical X-ray device, and vice versa. Furthermore, the comparative determinability parameter can be determined by manual and / or semi-automatic annotation based on the further training object image.

[0071] By applying the further trained function to the further input data based on the at least one training X-ray image, the further training object image can advantageously be determined. In this case, the at least one further parameter of the further trained function can be adjusted based on a comparison of the training determinability parameter with the comparison determinability parameter. Furthermore, the comparison can include, for example, a difference and / or a scalar product and / or a quotient.

[0072] This can advantageously improve the accuracy of the further trained function when generating the further training object image with regard to the determinability of the medical object depicted therein.

[0073] In a fourth aspect, the invention relates to a medical X-ray device comprising a provision unit. The medical X-ray device, in particular the provision unit, is designed to carry out a proposed method for controlling a medical X-ray device. In particular, the medical X-ray device can be designed as a medical C-arm X-ray device and / or as a computed tomography system. The medical X-ray device can further be designed to record and / or receive and / or provide the at least one X-ray image and / or the object image and / or the determinability parameter.

[0074] The advantages of the proposed medical X-ray device essentially correspond to the advantages of the proposed method for controlling a medical X-ray device. Features, advantages, or alternative embodiments mentioned herein can also be applied to the other claimed subject matter, and vice versa.

[0075] In a fifth aspect, the invention relates to a training unit configured to execute the previously described computer-implemented method according to the invention for providing a trained function and / or for providing a further trained function and its respective aspects. The training unit advantageously comprises a training interface and a training computing unit. The training unit is configured to execute these methods and their aspects by executing the corresponding method steps in the training interface and the training computing unit.

[0076] In an advantageous embodiment of the proposed training unit, the training interface can be configured to receive at least one training x-ray image of an examination region of an examination subject. Furthermore, the training computing unit can be configured to generate a training subject image based on the at least one training x-ray image. Furthermore, the training computing unit can be configured to determine a comparative determinability parameter for evaluating the determinability of the medical subject based on the training subject image. Furthermore, the training computing unit can be configured to determine a training determinability parameter by applying a trained function to input data based on the at least one training x-ray image.Furthermore, the training computing unit can be configured to adapt at least one parameter of the trained function based on a comparison of the training determinability parameter with the comparison determinability parameter. Furthermore, the training interface can be configured to provide the trained function.

[0077] In a further advantageous embodiment of the proposed training unit, the training interface can be configured to receive at least one training x-ray image of an examination region of an examination object. Furthermore, the training computing unit can be configured to generate a comparison object image based on the at least one training x-ray image. Furthermore, the training computing unit can be configured to generate a further training object image by applying a further trained function to further input data based on the at least one training x-ray image. Furthermore, the training computing unit can be configured to adapt at least one parameter of the further trained function based on a comparison of the further training object image with the comparison object image. Furthermore, the training interface can be configured to provide the further trained function.

[0078] The advantages of the proposed training unit essentially correspond to the advantages of the proposed computer-implemented method for providing a trained function and / or the proposed computer-implemented method for providing another trained function. Features, advantages, or alternative embodiments mentioned herein can also be applied to the other claimed subject matter, and vice versa.

[0079] In a sixth aspect, the invention relates to a computer program product comprising a computer program that can be loaded directly into a memory of a provision unit, with program sections for executing all steps of the method for controlling a medical X-ray device when the program sections are executed by the provision unit; and / or that can be loaded directly into a training memory of a training unit, with program sections for executing all steps of the proposed method for providing a trained function and / or for providing a further trained function and its respective aspects when the program sections are executed by the training unit.

[0080] In a seventh aspect, the invention relates to a computer-readable storage medium on which program sections readable and executable by a provision unit are stored in order to execute all steps of the method for controlling a medical X-ray device when the program sections are executed by the provision unit; and / or on which program sections readable and executable by a training unit are stored in order to execute all steps of the method for providing a trained function and / or for providing another trained function and its respective aspects when the program sections are executed by the training unit.

[0081] The invention relates in an eighth aspect to a computer program or computer-readable storage medium comprising a trained function provided by a proposed computer-implemented method or one of its aspects.

[0082] A largely software-based implementation has the advantage that even previously used provisioning units and / or training units can be easily upgraded to operate in the manner according to the invention via a software update. Such a computer program product may, in addition to the computer program, optionally include additional components such as documentation and / or additional components, as well as hardware components such as hardware keys (dongles, etc.) for using the software.

[0083] Embodiments of the invention are illustrated in the drawings and are described in more detail below. In different figures, the same reference numerals are used for the same features. Fig. 1 to 5 are schematic representations of various embodiments of a proposed method for controlling a medical X-ray device, Fig. 6 a schematic representation of a proposed computer-implemented method for providing a trained function, Fig. 7 and Fig. 8 schematic representations of various embodiments of a proposed computer-implemented method for providing a further trained function, Fig. 9 a schematic representation of a training session, Fig. 10 a schematic representation of a medical C-arm X-ray machine.

[0084] In Fig. 1 schematically shows an advantageous embodiment of the proposed method for controlling a medical X-ray device. In a first step a), at least one X-ray image RB of an examination region of an examination object can be recorded using the medical X-ray device ACQ-RB. In this case, a medical object, for example a guide wire and / or a catheter and / or an endoscope and / or a laparoscope and / or an implant, in particular a stent, can be arranged in the examination region. In a second step b), an object image OB can be generated based on the at least one X-ray image RB DET-OB. Thereafter, in a third step c), a determinability parameter BP for evaluating a determinability of the medical object can be determined based on the object image OB DET-BP.Furthermore, the proposed method for controlling the medical X-ray device can be executed repeatedly starting from step a) until a termination condition AB occurs, which is based on the last determined determinability parameter.

[0085] Furthermore, the generation of the object image DET-OB may comprise a movement correction and / or registration of the at least one X-ray image RB.

[0086] In each repetition of the proposed method, at least one X-ray image RB can be acquired in step a). After n repetitions, at least n X-ray images RB.1, RB.2, ... RB.n can advantageously be acquired.

[0087] The termination condition AB can advantageously comprise a comparison of the determinability parameter BP last determined in step c), in particular during the current repetition, with a predefined threshold value. The termination condition AB can occur, for example, when the determinability parameter BP exceeds and / or falls below and / or reaches the predefined threshold value Y. Furthermore, the termination condition AB can have a maximum number of repetitions, with the termination condition AB occurring when the maximum number of repetitions is reached and / or exceeded.

[0088] In particular, in the case of a pulsed emission of X-rays by the X-ray source for recording ACQ-RB of the at least one X-ray image RB, a time-limited X-ray pulse and / or a defined number of X-ray pulses can be emitted by the X-ray source in step a). Upon the occurrence Y of the termination condition AB, radiation termination, in particular a stopping E of the pulsed emission of X-rays by the X-ray source, can advantageously occur.

[0089] In the Fig. In the embodiment of the proposed method for controlling a medical X-ray device schematically illustrated in Figure 2, at least one subset RB.sub, comprising at least one X-ray image RB.x, RB.y of the previously acquired X-ray images RB.1, RB.2, ... RB.n, can be determined DET-RB.sub. The object image OB can be generated based on the at least one subset RB.sub of the previously acquired X-ray images RB.1, RB.2, ... RB.n, DET-OB.

[0090] Advantageously, the object image OB can be generated by a, particularly weighted and / or adaptive, averaging of the at least one subset RB.sub of the previously acquired X-ray images RB.1, RB.2, ... RB.n DET-OB. Furthermore, the determination DET-RB.sub of the at least one subset RB.sub can be based on the determinability parameter BP.

[0091] Advantageously, upon repeated execution of steps a) to c), at least one determinability parameter BP can be determined for the at least one x-ray image RB, in particular the most recently acquired one, based on the object image OB DET-BP. In this case, the at least one subset RB.sub of the x-ray images RB.1, RB.2, ... RB.n acquired so far can be further determined in step b) based on the at least one determinability parameter BP. This advantageously makes it possible to achieve an improvement, in particular an increase, of the determinability parameter BP of the current repetition. Furthermore, x-ray images RB that have image artifacts and / or interference and / or motion artifacts can advantageously be excluded from the at least one subset RB.sub for generating the object image DET-OB.

[0092] In the Fig. In the embodiment of the proposed method for controlling a medical X-ray device schematically illustrated in Figure 3, the determination of the determinability parameter DET-BP in step c) can be performed by applying a trained function TF-BP to input data. The input data can be based on the object image OB. Furthermore, at least one parameter of the trained function TF-BP can be based on a comparison of a training determinability parameter with a comparison determinability parameter.

[0093] In Fig. Figure 4 schematically illustrates a further advantageous embodiment of the proposed method for controlling a medical X-ray device. The object image DET-OB can be generated in step b) by applying a further trained function TF2 to further input data. The further input data can be based on the at least one X-ray image RB. Furthermore, at least one parameter of the trained function TF2 can be based on a comparison of a further training object image with a comparison object image.

[0094] In the Fig. In the embodiment of the proposed method for controlling a medical X-ray device schematically illustrated in Figure 5, the further trained function TF2 can also be configured to determine the determinability parameter DET-BP. At least one further parameter of the further trained function TF2 can be based on a comparison of a training determinability parameter with a comparison determinability parameter.

[0095] Fig. 6 shows a schematic representation of an embodiment of the computer-implemented method for providing a trained function TF-BP. In a first step, at least one training x-ray image TRB can be received REC-TRB, which images an examination region of an examination object. A medical object can be arranged in the examination region. In a second step, a training object image TOB can be generated based on the at least one training x-ray image TRB DET-OB. Furthermore, in a third step, a comparison determinability parameter VBP for evaluating a determinability of the medical object based on the training object image TOB can be determined DET-BP.In a fourth step, which can occur in particular before and / or during and / or after the third step, a training determinability parameter TBP can be determined by applying the trained function TF-BP to input data based on the at least one training x-ray image TRB. Subsequently, in a fifth step, at least one parameter of the trained function TF-BP can be adjusted based on a comparison of the training determinability parameter TBP with the comparison determinability parameter VPB. In a sixth step, the trained function TF-BP can be provided PROV-TF-BP.

[0096] In Fig. 7 schematically shows an embodiment of the computer-implemented method for providing a further trained function TF2. In a first step, at least one training x-ray image TRB can be received from an examination region of an examination object REC-TRB. A medical object can be arranged in the examination region. In a second step, a comparison object image VOB can be generated based on the at least one training x-ray image TRB DET-OB. In a third step, which can take place in particular before and / or during and / or after the second step, a further training object image TOB2 can be generated by applying the further trained function to further input data. In this case, the further input data can advantageously be based on the at least one training x-ray image TRB.Furthermore, in a fourth step, at least one parameter of the further trained function TF2 can be adjusted based on a comparison of the further training object image TOB2 with the comparison object image VOB ADJ-TF2. The further trained function TF2 can then be provided in a fifth step PROV-TF2.

[0097] In the Fig. In the embodiment of the proposed computer-implemented method for providing a further trained function TF2, which is schematically illustrated in FIG. 8, a comparison determinability parameter VBP for evaluating a determinability of the medical object can be determined based on the further training object image TOB2 DET-BP. Furthermore, a training determinability parameter TBP can be determined by applying the further trained function TF2 to the further input data. In this case, at least one further parameter of the further trained function TF2 can be adjusted based on a comparison of the training determinability parameter TBP with the comparison determinability parameter VBP ADJ-TF2.

[0098] Fig. 9 schematically shows a proposed training unit TRS comprising a training interface TIF, a training computing unit TCU, and a training storage unit TMU. The training unit TRS can be configured to execute a proposed computer-implemented method for providing a trained function PROV-TF and / or a proposed computer-implemented method for providing a further trained function PROV-TF2 and its respective aspects, by configuring the training interface TIF and the training computing unit TCU to execute the corresponding method steps.

[0099] In an advantageous embodiment of the proposed training unit TRS, the training interface TIF can be configured to receive REC-TRB of the at least one training x-ray image TRB of the examination region of the examination subject. Furthermore, the training computing unit TCU can be configured to generate DET-OB of the training subject image TOB based on the at least one training x-ray image TRB. Furthermore, the training computing unit TCU can be configured to determine DET-BP of the comparison determinability parameter VBP for evaluating the determinability of the medical subject based on the training subject image TOB. Furthermore, the training computing unit TCU can be configured to determine the training determinability parameter TBP by applying the trained function TF-BP to input data based on the at least one training x-ray image TRB.Furthermore, the training computing unit TCU can be configured to adapt ADJ-TF-BP of at least one parameter of the trained function TF-BP based on a comparison of the training determinability parameter TBP with the comparison determinability parameter VBP. Furthermore, the training interface TIF can be configured to provide PROV-TF-BP of the trained function TF-BP.

[0100] In a further advantageous embodiment of the proposed training unit TRS, the training interface TIF can be configured to receive REC-TRB of the at least one training x-ray image TRB of the examination region of the examination subject. Furthermore, the training computing unit TCU can be configured to generate DET-VOB of the comparison object image VOB based on the at least one training x-ray image TRB. Furthermore, the training computing unit TCU can be configured to generate the further training object image TOB2 by applying the further trained function TF2 to the further input data based on the at least one training x-ray image TRB. Furthermore, the training computing unit TCU can be configured to adapt ADJ-TF2 of the at least one parameter of the further trained function TF2 based on a comparison of the further training object image TOB2 with the comparison object image VOB.In addition, the training interface TIF can be designed to provide PROV-TF2 of the further trained function TF2.

[0101] A training interface (TIF) can be a hardware or software interface (e.g., PCI bus, USB, or Firewire). A training processing unit (TCU) can comprise hardware or software elements, such as a microprocessor or a so-called FPGA (field programmable gate array). A training memory unit (TMU) can be implemented as a temporary random access memory (RAM) or as a permanent mass storage device (hard drive, USB stick, SD card, solid state disk).

[0102] The training interface TIF can, in particular, comprise a plurality of subinterfaces that execute different steps of the respective methods. In other words, the training interface TIF can also be understood as a plurality of training interfaces TIF. The training computing unit TCU can, in particular, comprise a plurality of sub-computation units that execute different steps of the respective methods. In other words, the training computing unit TCU can also be understood as a plurality of training computing units TCU.

[0103] The training unit (TRS) can be, in particular, a computer, a microcontroller, or an integrated circuit. Alternatively, the training unit (TRS) can be a real or virtual network of computers (the technical term for a real network is "cluster," and the technical term for a virtual network is "cloud"). The training unit (TRS) can also be implemented as a virtual system running on a real computer or a real or virtual network of computers (virtualization).

[0104] In Fig.10 schematically illustrates a medical C-arm X-ray device 37 as an example of a proposed medical imaging device. The medical C-arm X-ray device 37 can advantageously include a proposed provision unit PRVS for controlling the medical X-ray device 37. The medical imaging device 37, in particular the proposed provision unit PRVS, is designed to execute a proposed method for controlling the medical X-ray device 37.

[0105] The medical C-arm X-ray device 37 further comprises a detector unit 34 and an X-ray source 33. To acquire the at least one X-ray image RB, the arm 38 of the C-arm X-ray device 37 can be mounted so as to be movable around one or more axes. Furthermore, the medical C-arm X-ray device 37 can comprise a movement device 39, which enables the C-arm X-ray device 37 to move in space.

[0106] To acquire the at least one x-ray image RB of the examination area of ​​the examination subject 31 arranged on a patient support device 32, the provision unit PRVS can send a signal 24 to the x-ray source 33. The x-ray source 33 can then emit an x-ray beam, in particular a cone beam and / or fan beam and / or parallel beam. When the x-ray beam strikes a surface of the detector unit 34 after interacting with the examination area of ​​the examination subject 31 to be imaged, the detector unit 34 can send a signal 21 to the provision unit PRVS. The provision unit PRVS can, for example, receive the at least one x-ray image RB based on the signal 21.

[0107] Furthermore, the medical C-arm X-ray device 37 can comprise an input unit 42, for example a keyboard, and / or a display unit 41, for example a monitor and / or display. The input unit 42 can preferably be integrated into the display unit 41, for example in the case of a capacitive input display. An input from an operating personnel at the input unit 42 can thereby enable, in particular supplementary, control of the medical C-arm X-ray device 37, in particular of the proposed method for controlling the medical X-ray device 37. For this purpose, the input unit 42 can, for example, send a signal 26 to the provision unit PRVS.

[0108] Furthermore, the display unit 41 can be configured to display information and / or graphic representations of information from the medical imaging device 37 and / or the provision unit PRVS and / or other components. For this purpose, the provision unit PRVS can, for example, send a signal 25 to the display unit 41. In particular, the display unit 41 can be configured to display a graphic representation of the at least one x-ray image RB and / or the at least one subset of the x-ray images RB.sub acquired so far and / or the object image OB and / or the determinability parameter BP.

[0109] The PRVS provisioning unit can be, in particular, a computer, a microcontroller, or an integrated circuit. Alternatively, the PRVS provisioning unit can be a real or virtual network of computers (the technical term for a real network is "cluster," and the technical term for a virtual network is "cloud"). The PRVS provisioning unit can also be implemented as a virtual system that runs on a real computer or a real or virtual network of computers (virtualization).

[0110] The schematic representations contained in the figures described do not represent any scale or proportions.

[0111] Finally, it should be noted once again that the methods described in detail above, as well as the devices illustrated, are merely exemplary embodiments that can be modified in a variety of ways by those skilled in the art without departing from the scope of the invention. Furthermore, the use of the indefinite articles "a" or "an" does not exclude the possibility that the respective features may be present in multiple instances. Likewise, the terms "unit" and "element" do not exclude the possibility that the respective components consist of several interacting subcomponents, which may also be spatially distributed.

Claims

[1] A method for controlling a medical X-ray device (37), comprising: a) taking (ACQ-RB) at least one X-ray image (RB) of an examination area of ​​an examination object (31) by means of the medical X-ray device (37), wherein a medical object is arranged in the examination area, b) generating (DET-OB) an object image (OB) by averaging at least a subset (RB.sub) of the X-ray images (RB.1, RB.2, ... RB.n) taken so far, c) Determining (DET-BP) a determinability parameter (BP) for evaluating a determinability of the medical object based on the object image (OB), wherein the method is repeatedly executed starting from step a) until the occurrence (E) of a termination condition (AB) based on the last determined determinability parameter (BP). [2] Method according to claim 1, characterized bythat at least one subset (RB.sub) of the x-ray images (RB.1, RB.2, ... RB.n) taken so far is determined for generating (DET-OB) the object image (OB) based on the determinability parameter (BP). [3] Method according to one of the preceding claims, characterized by that the termination condition (AB) comprises a comparison of the determinability parameter (BP) with a predetermined threshold value. [4] Method according to one of the preceding claims, characterized by that the generation (DET-OB) of the object image (OB) comprises a movement correction and / or registration of the at least one X-ray image (RB). [5] Method according to one of the preceding claims, characterized bythat the determination (DET-BP) of the determinability parameter (BP) in step c) is carried out by applying a trained function (TF-BP) to input data, wherein the input data is based on the object image (OB), wherein at least one parameter of the trained function (TF-BP) is based on a comparison of a training determinability parameter (TBP) with a comparison determinability parameter (VBP). [6] Method according to one of the preceding claims, characterized by that the generation (DET-OB) of the object image (OB) in step b) is carried out by applying a further trained function (TF2) to further input data, wherein the further input data is based on the at least one X-ray image (RB), wherein at least one parameter of the further trained function (TF2) is based on a comparison of a further training object image (TOB2) with a comparison object image (VOB). [7] Method according to claim 6, characterized bythat the further trained function (TF2) is also designed to determine (DET-BP) the determinability parameter (BP), wherein at least one further parameter of the further trained function (TF2) is based on a comparison of a training determinability parameter (TBP) with a comparison determinability parameter (VBP). [8] Computer-implemented method for providing (PROV-TF-BP) a trained function (TF-BP), comprising: - receiving (REC-TRB) at least one training X-ray image (TRB) from an examination area of ​​an examination object (31), wherein a medical object is arranged in the examination area, wherein the medical object comprises a guide wire and / or a catheter and / or an endoscope and / or a laparoscope and / or an implant, - generating (DET-OB) a training object image (TOB) based on the at least one training X-ray image (TRB), - Determining (DET-BP) a comparative determinability parameter (VOB) to evaluate the determinability of the medical object based on the training object image (TRB), - Determining a training determinability parameter (TBP) by applying the trained function (TF-BP) to input data, wherein the input data is based on the at least one training X-ray image (TRB), - Adjusting (ADJ-TF-BP) at least one parameter of the trained function (TF-BP) based on a comparison of the training determinability parameter (TBP) with the comparison determinability parameter (VBP), - Providing (PROV-TF-BP) the trained function (TF-BP). [9] Computer-implemented method for providing (PROV-TF2) a further trained function (TF2), comprising: - receiving (REC-TRB) at least one training X-ray image (TRB) of an examination area of ​​an examination object (31), wherein a medical object is arranged in the examination area, wherein the medical object comprises a guide wire and / or a catheter and / or an endoscope and / or a laparoscope and / or an implant, - generating (DET-OB) a comparison object image (VOB) based on the at least one training X-ray image (TRB), - Creating another training object image (TOB2) by applying the further trained function (TF2) to further input data, wherein the further input data are based on the at least one training X-ray image (TRB), - adjusting (ADJ-TF2) at least one parameter of the further trained function (TF2) based on a comparison of the further training object image (TOB2) with the comparison object image (VOB), - Determining (DET-BP) a comparative determinability parameter (VBP) to evaluate the determinability of the medical object based on the training object image (TRB), - Determining a training determinability parameter (TBP) by applying the further trained function (TF2) to the further input data, - adjusting (ADJ-TF2) at least one further parameter of the further trained function (TF2) based on a comparison of the training determinability parameter (TBP) with the comparison determinability parameter (VBP), - Providing (PROV-TF2) the further trained function (TF2). [10] Medical X-ray device (37) comprising a provision unit (PRVS) which is designed to carry out a method according to one of claims 1 to 7, wherein the medical X-ray device (37) is designed to record (ACQ-RB) and / or to receive and / or to provide the at least one X-ray image (RB) and / or the object image (OB) and / or the determinability parameter (BP). [11] Training unit (TRS) designed to carry out a computer-implemented method according to one of claims 8 to 9. [12] Computer program product with a computer program which is directly loadable into a memory of a provision unit (PRVS), with program sections to carry out all steps of the method according to one of claims 1 to 7 when the program sections are executed by the provision unit (PRVS); and / or which is directly loadable into a training memory (TMU) of a training unit (TRS), with program sections to carry out all steps of the computer-implemented method according to one of claims 8 to 9 when the program sections are executed by the training unit (TRS). [13] Computer-readable storage medium on which program sections readable and executable by a provision unit are stored in order to carry out all steps of the method according to one of claims 1 to 7 when the program sections are executed by the provision unit (PRVS); and / or on which program sections readable and executable by a training unit (TRS) are stored in order to carry out all steps of the computer-implemented method according to one of claims 8 to 9 when the program sections are executed by the training unit (TRS).

Images