DE102021201600A1 - Improvement of a set of imaging parameters in transmission X-ray imaging - Google Patents

Abstract

The invention relates to a computer-implemented method for creating a corrected volume data set (KVD) representing an examination object (10) based on initial projection data sets (IPD). In a first step, the initial projection data sets (IPD) are provided (S10; S11, S12), which are based on at least one initial set of imaging parameters (IAP). In a further step, manipulated projection data sets (MPD) are generated (S20) by applying a manipulation operation (MO) to at least part of the initial projection data sets (IPD). In a further step, a corrected set of imaging parameters (KAP) is generated (S30) based on the manipulated projection data sets (MAP). In a last step, the corrected volume data set (KVD) is generated (S40) based on at least part of the initial projection data sets (IVD) and the corrected set of imaging parameters (KAP).

Description

The invention is in the field of transmission imaging and relates in particular to a method for creating a corrected volume data set representing an examination object, a device for creating a corrected volume data set, an imaging system, computer program products and computer-readable storage media.

Transmission imaging is used to create a transmission image of an examination object. It can be used in medical imaging, but also for non-medical purposes such as industrial material testing. In transmission imaging, radiation is directed through an object to be examined and then impinges on a 2D detector. Computed tomography or classical radiography are two examples of transmission imaging. In both cases, an x-ray source is used as the radiation source, which emits x-rays of an adjustable x-ray energy spectrum in the direction of the object to be examined. The 2D detector aligned with the X-ray source records the extent to which the radiation was changed by the object to be examined, in particular weakened. The x-ray radiation leaves the x-ray source, for example, in the form of a cone beam or a fan beam corresponding to a very flattened cone. In the case of transmission imaging, 2D/two-dimensional (e.g. in radiography) or 3D/three-dimensional (e.g. in computed tomography) transmission images of the examination object can be created. A 2D transmission image is created when the examination object is irradiated from a specific viewing angle/a specific viewing direction/a specific perspective. A 2D transmission image, also known as a projection/projection image, corresponds to a direction-specific (X-ray) attenuation profile of the examination object. A 3D transmission image also includes depth information about the examination object, in particular a spatial density distribution of tissue or material. In order to generate a 3D transmission image, a number of projections of the examination object must be recorded from different perspectives and a 3D transmission image, also known as a 3D volume, must be reconstructed from them. The projection plurality is typically recorded while the examination object and the pair of X-ray source and 2D detector are moving relative to one another, as a result of which the viewing direction of each projection is different.

In computed tomography, for example, a set of 2D X-ray projections of an examination object is converted into a 3D volume using known reconstruction algorithms, with the 3D volume representing the spatial density distribution of the examination object. In order to carry out the object reconstruction, information about the imaging parameters used during image acquisition is required in addition to the projection images. Imaging parameters describe, for example, specific scattered beam properties or beam hardening properties of an examination object or the imaging geometry present during image data acquisition.

Exact information about one or more imaging parameters is often missing. However, the imaging parameters have a direct influence on an image quality of the 3D volume. If the accuracy of the imaging parameters can be increased, the image quality of the 3D volume usually also improves.

It is known to automatically determine mapping parameters more precisely by applying consistency conditions. A consistency condition describes, for example, data redundancies in projection data sets or their 3D Radon transform. A consistency condition can affect global data consistency or local data consistency. Assuming that an inconsistency in the data is due to a deviation between actual and assumed values of mapping parameters, a cost function can be formed using the consistency condition. With (iterative) variation of one or more mapping parameters, the cost function is then solved in the sense of an optimization method until a function minimum is reached or sufficiently approximated.

Consistency conditions can basically be derived mathematically exactly. They are regularly based on functions that use projection data sets and imaging parameters as input values and provide a quantitative statement about the quality/accuracy of the input values used. Consequently, consistency functions are cost functions that are typically used to at least qualitatively predict the image quality of a 3D volume without extremely computationally expensive CT image reconstruction.

Consistency conditions are often applied successfully even though the constraints of their mathematical derivation are violated. A typical and frequently occurring injury concerns, for example, the convex or complete shell of the examination object. Consistency conditions usually presuppose that the examination object is completely mapped in all projection data records. In practice, this is rarely the case because the X-ray detector is sufficiently large gates are too expensive or unavailable. However, experience shows that a number of consistency conditions can still be used to improve mapping parameters, provided that reasonable (value) ranges for the mapping parameters are defined in advance. When searching for the correct imaging geometry, the mechanical accuracy/tolerance of the imaging system can be used as a range of values, for example. The range limits may vary from application to application. By defining application-specific range limits, at least one local minimum of the cost function can be found or approximated.

An example of a consistency function that still delivers good results even if the secondary conditions of its mathematical derivation are violated is the consistency condition of epipolar geometry.

There are also use cases where a consistency function does not provide the desired result. For example, if an examination object is fundamentally unsuitable for the application of a consistency condition, for example due to its shape. In addition, various effects that influence data consistency can occur superimposed during image data acquisition. A minimum of the cost function can then still be found, but this no longer represents the optimum of the mapping parameters. This is also the problem of the present invention.

• - Setzen von strengen Randbedingungen, also Definieren von sinnvollen Wertebereichen für die zu optimierenden Abbildungsparameter, wie eingangs bereits erläutert. Dies erfordert menschliche Interaktion und kann im Einzelfall schwierig sein. Daneben kann es die Brauchbarkeit der Korrekturmethode weiter senken.
• - Wenn ein einzelner, dominanter, qualitätsmindernder Effekt identifiziert werden kann, der die Kostenfunktion überlagert, kann dieser Effekt zuerst reduziert werden. Hier kann die Kostenfunktion selbst oder auch ein anderes gängiges Korrekturverfahren angewandt werden.
• - Kann der dominante, qualitätsmindernde Effekt nicht identifiziert und losgelöst korrigiert werden, können verschiedene Effekte gemeinsam mittels der Kostenfunktion optimiert werden. Dies erhöht jedoch die zu optimierende Parameteranzahl und damit den Rechenaufwand. Sprich, der oder die einen qualitätsmindernden Effekt repräsentierenden Abbildungsparameter werden über das Lösen der Kostenfunktion jeweils in Abhängigkeit des oder der den weiteren Effekt repräsentierenden Abbildungsparameter optimiert.
• - Wird empirisch mit obigen Maßnahmen kein Erfolg festgestellt, kann noch eine andere Konsistenzbedingung/Kostenfunktionen herangezogen werden, mit welcher wiederum obige Maßnahmen auf Erfolg getestet werden können.
• - Schlägt auch das fehl, bleibt nur, die automatisierte Bestimmung der bildqualitätsrelevanten Abbildungsparameter einzustellen.

If it is determined that the optimum parameter cannot be achieved directly using a cost function, the following countermeasures can be taken so far:

• - Setting strict boundary conditions, i.e. defining meaningful value ranges for the imaging parameters to be optimized, as already explained at the beginning. This requires human interaction and can be difficult in individual cases. Besides, it may further lower the usefulness of the correction method.
• - If a single, dominant, quality-decreasing effect can be identified that is superimposed on the cost function, then this effect can be reduced first. The cost function itself or another common correction method can be used here.
• - If the dominant, quality-reducing effect cannot be identified and corrected separately, various effects can be optimized together using the cost function. However, this increases the number of parameters to be optimized and thus the computational effort. In other words, the mapping parameter(s) representing a quality-reducing effect are optimized by solving the cost function in each case as a function of the mapping parameter(s) representing the further effect.
• - If no success is determined empirically with the above measures, another consistency condition/cost function can be used, with which the above measures can in turn be tested for success.
• - If that also fails, the only option is to stop the automated determination of the imaging parameters relevant to image quality.

In contrast, the object of the present invention is to provide alternative means that allow a less complex, automatic optimization of at least one imaging parameter relevant to image quality. It is also an object of the present invention to provide means for an easier reduction of image errors in a 3D volume.

This object is achieved by a method for creating a corrected volume data set representing an examination object, corresponding device, imaging system, computer program and computer-readable data carrier according to the independent claims. Preferred and/or alternative, advantageous design variants are the subject matter of the dependent claims.

The solution to the problem according to the invention is described below in relation to the claimed method and also in relation to the claimed devices. Features, advantages or alternative embodiments mentioned here are also to be transferred to the other claimed subjects and vice versa. In other words, specific claims (which are directed, for example, to a method) can also be developed with features that are described or claimed in connection with one of the devices. The corresponding functional features of the method are formed by corresponding physical modules or units.

In a first aspect, the present invention relates to a method for generating a corrected set of imaging parameters from at least one initial set of imaging parameters. The process involves a number of steps.

The method is a computer-implemented method.

A first step is aimed at providing initial projection data sets. The initial projection data sets are based on the at least one initial set of imaging parameters. The initial projection data records represent an examination object. More precisely, the initial projection data records each represent an X-ray absorption profile of the examination object for a specific projection direction or a projection area. A projection data set essentially determines (ignoring multiple scattering or the like) for the multiplicity of individual rays of a conical or fan-shaped x-ray beam which proportion of the original x-ray intensity has passed through the examination object.

In embodiments, the initial projection data sets can all be based on an initial set of imaging parameters. In other versions, the initial projection data sets are based on a plurality, in particular two, different initial mapping parameter sets. In embodiments, the initial mapping parameter sets can be subdivided into at least two projection data set subsets corresponding to the respective initial mapping parameter sets. If subsets of the initial projection data sets are included, there are also different, respectively assigned, initial sets of imaging parameters. These differ from one another in the value of at least one imaging parameter.

An initial set of imaging parameters should be understood within the meaning of the invention as a parameter set comprising geometric parameters characterizing the imaging geometry, physical parameters characterizing, for example, the X-ray radiation used and/or parameters characterizing imaging properties of the examination object. An initial set of mapping parameters includes at least a large number of geometric parameters.

• - Rotationsachse (Achse relativ zu einem Referenzkoordinatensystem, um welche Untersuchungsobjekt oder Strahler-Detektor-Paar gedreht wird)
• - Linien-Trajektorie (Achse relativ zu einem Referenzkoordinatensystem, entlang welcher das Untersuchungsobjekt oder das Strahler-Detektor-Paar verschoben wird),
• - Winkel Röntgenstrahler-Rotationsachse und/oder Winkel Röntgendetektor-Rotationsachse (Winkel zwischen Zentralstahl des Röntgenstrahls und der Rotationsachse),
• - Abstand Untersuchungsobjekt-Röntgenstrahler,
• - Abstand Untersuchungsobjekt-Röntgendetektor,
• - Abstand Röntgenstrahler-Röntgendetektor,
• - Größe Detektorpixel,
• - Anzahl Detektorpixel,
• - Form Detektorpixel,
• - Winkelinkrement (Winkelbereich, der bei einer Rotationsbewegung um eine Rotationsachse abgeschritten wird),
• - Anzahl Projektionswinkel/Schrittweite = Anzahl Projektionsdatensätze,
• - Anzahl Aufnahmepositionen = Anzahl Projektionsdatensätze,
• - Kegelstrahlwinkel oder auch Fächerstrahlwinkel.

Geometric parameters can be parameters of the non-exhaustive group of parameters:

• - Axis of rotation (axis relative to a reference coordinate system around which the examination object or emitter-detector pair is rotated)
• - line trajectory (axis relative to a reference coordinate system, along which the examination object or the emitter-detector pair is moved),
• - angle of X-ray emitter axis of rotation and/or angle of X-ray detector axis of rotation (angle between the central beam of the X-ray beam and the axis of rotation),
• - distance between examination object and X-ray source,
• - distance between examination object and X-ray detector,
• - X-ray emitter-X-ray detector distance,
• - size of detector pixels,
• - number of detector pixels,
• - shape detector pixels,
• - Angular increment (angular range that is covered during a rotational movement around a rotational axis),
• - number of projection angles/increment = number of projection data records,
• - number of recording positions = number of projection data sets,
• - Cone beam angle or fan beam angle.

• - Röntgenröhrenstrom,
• - Röntgenröhrenspannung,
• - Anodenmaterial
• - Material oder Dicke eines Vorfilters für die Röntgenstrahlung
• - Detektormaterial
• - Detektionsmechanismus (energieintegrierend oder photonenzählend) des Röntgendetektors
• - Dynamikbereich des Röntgendetektors
• - Eigenschaften wir Material oder Geometrie eines Streustrahlungsrasters.

Physical parameters can be parameters from the non-exhaustive group of parameters:

• - X-ray tube current,
• - X-ray tube voltage,
• - anode material
• - Material or thickness of a pre-filter for X-rays
• - Detector material
• - Detection mechanism (energy integrating or photon counting) of the X-ray detector
• - Dynamic range of the X-ray detector
• - Properties such as material or geometry of a scattered radiation grid.

• - Parameter repräsentierend Streustrahleigenschaften,
• - Parameter repräsentierend Strahlaufhärtungsverhalten,
• - Parameter repräsentierend Metallstrukturen des Untersuchungsobjektes.

Parameters characterizing imaging properties of the examination object can be parameters from the non-exhaustive group of parameters:

• - parameters representing scattered radiation properties,
• - parameters representing jet hardening behavior,
• - Parameters representing metal structures of the object under investigation.

In embodiments, an initial set of imaging parameters can include real, actual, exact imaging parameters which are based, for example, on precise knowledge of the imaging system used (for example by means of calibration) and/or the examination object. An initial sentence fig However, the formation parameter also includes at least one theoretical, fictitious or virtual mapping parameter, the exact value of which is not exactly known and for which a theoretical, fictitious or virtual value is used instead. According to the invention, this value should be determined exactly. In embodiments, an initial set of imaging parameters can also only include theoretical, fictitious or virtual imaging parameters, for example when at least a subset of the initial projection data sets is only based on an already reconstructed 3D volume and no knowledge of the original projection data sets, the examination object and/or or the imaging system used.

The at least one theoretical, fictitious or virtual imaging parameter can be specified manually by a user based on empirical values and/or a data sheet of an imaging system and/or an inspection of the same, the 3D volume and/or the examination object. Alternatively or additionally, the specification can at least partly also be done automatically by a data processing unit, for example with automatic evaluation of image information of the 3D volume or based on data processing of a user input.

The examination subject is a patient, typically a human being. The patient can also be an animal. Alternatively, the examination object can be a plant or a non-living object, e.g. an industrially manufactured workpiece or other object, e.g. a piece of luggage. In this respect, the invention can be used in a wide variety of imaging fields of application.

A second step of the method according to the invention relates to generating manipulated projection data sets by applying a manipulation operation to at least some of the projection data sets. In embodiments, only one or more subsets of the initial projection data sets are processed with a manipulation operation, each corresponding to an initial set of imaging parameters. In other implementations, all projection data sets are processed with one manipulation operation. Different or the same manipulation operations can be used for each subset.

According to the invention, a manipulation operation corresponds to at least one arithmetic operation or function which is suitable for manipulating image information from projection data sets, preferably pixel by pixel. In embodiments, the manipulation operation is thus designed such that it changes the image information contained in at least one pixel of an initial projection data record. Only the entry of this pixel, the entries of several pixels, preferably neighboring pixels of the manipulated pixel and/or at least one entry of a corresponding pixel of a further initial projection data set can be included in the manipulation of a pixel of an initial projection data set. According to the invention, the manipulation operation is preferably determined empirically.

In this step, changed, modified projection data sets are now generated in the form of the manipulated projection data sets.

A further step relates to generating a corrected set of imaging parameters based on the projection data sets manipulated as above. In other words, using the manipulated projection data sets, an error correction method is now applied, which results in a corrected set of imaging parameters. Generating the corrected set of mapping parameters is also based on at least one initial set of mapping parameters. The error correction method can be applied to correct only one or more mapping parameters to generate the set of corrected mapping parameters. Consequently, in embodiments, this may include corrected and initial values for mapping parameters. In other implementations, it may only include corrected values for mapping parameters. In embodiments, the corrected set of imaging parameters approximates an actual, real but at least partially unknown set of imaging parameters. The error correction method used is preferably a correction method known per se, which is typically applied to real projection data sets having imaging errors before a 3D reconstruction.

By creating manipulated projection data sets by means of a manipulation operation, the present invention makes these automatic correction methods useful for the present application.

The invention uses the knowledge mentioned at the outset that there are specific functions that define the error correction method and that use projection data and imaging parameters relevant to image quality as input values. The knowledge lies in the fact that not only the imaging parameters, but also the projection data themselves can influence the function. The proposed method is based on an algorithmic approach. The corrected set of imaging parameters can be used for various applications and, if necessary, processed further.

For example, in the industrial environment, the aim of a measurement using transmission X-ray imaging can be the qualitative assessment of a workpiece in a non-destructive material or workpiece test. It can be observed here that this assessment is less about the conventional workflow “1. Measuring projection images, 2. creating a 3D volume, 3. application-specific evaluation of the 3D volume”, but increasingly with direct evaluation of the projection images and the corrected set of imaging parameters, in particular using a trained neural network or an artificial intelligence algorithm ( AI) takes place. Neural networks or AI algorithms are designed to carry out the evaluation directly and without a computationally intensive detour via the 3D volume.

Another application is in the medical environment. There, on the basis of the set of corrected imaging parameters, a certain undesired movement of the examination subject, here a patient, can be inferred without volume reconstruction. This is of particular interest, for example, if a target volume position (tumor) has to be readjusted as part of radiation therapy.

The corrected set of imaging parameters can also be used for a volume reconstruction. Accordingly, in a preferred embodiment of the invention, a further step of the method relates to generating a corrected volume data set based on at least a part of the initial projection data sets and the corrected set of imaging parameters. The corrected volume data set corresponds to a volume data set having reduced imaging errors compared to a volume data set reconstructed using the initial set of imaging parameters. So its picture quality is improved. The corrected volume data set is generated by a known transmission x-ray imaging reconstruction method, for example a filtered back projection or the like. During the reconstruction, the initial projection data sets are back-projected according to parameters of the corrected set of imaging parameters.

In this way, the present invention brings about an improvement in the image quality of a volume data set, the impairment of the image quality being based on deviations between an actual, real but at least partially unknown set of imaging parameters and a theoretical, virtual set of imaging parameters.

In embodiments of the invention, the provision of the initial projection datasets includes capturing projection datasets by means of an imaging system of the transmission X-ray according to the at least one initial set of imaging parameters. At least one subset, in embodiments also several, of the initial projection data sets is consequently based on real measurement data of an imaging system using an initial set of imaging parameters in each case. Subsets of the initial projection data sets each depict the same examination object. Subsets of the initial projection data sets are therefore imaging parameters in the respective initial sets, preferably in at least one of the geometric parameters and/or one of the physical parameters, for example the X-ray tube voltage.

The initial projection datasets can be made available, for example, directly after being created by the imaging system or via a step of intermediate storage from a memory, arranged centrally or decentrally, via appropriate communication paths and interfaces. The associated initial set of mapping parameters can also be provided in this way.

In further embodiments of the invention, the provision of the initial projection data sets includes generating projection data sets by forward projection of an initial volume data set corresponding to a virtual initial set of imaging parameters. At least a subset of the initial projection data sets is based on purely virtual measurement data. The associated initial set of mapping parameters is a virtual parameter set that exclusively includes virtual mapping parameters, as already mentioned at the beginning. This procedure enables the present idea to be used even if only a completely reconstructed 3D volume is available, but no real initial projection data.

The initial projection data sets can be created here, for example, directly by a data processing unit or retrieved from a memory. The initial volume data record can also be stored in a memory so that it can be called up by the data processing unit and, like the initial projection data records, can be made available via corresponding communication paths and interfaces. In embodiments, the associated initial set of imaging parameters can preferably be specified manually by a user based on empirical values or knowledge of the imaging system and/or the examination object, or manually or automatically based on image information of the initial volume data set.

In embodiments of the invention, the generation of the manipulated projection data sets includes apply the manipulation operation to all initial projection datasets. In embodiments of the invention, the entire image information of the initial projection data sets is therefore manipulated. In other versions, only parts of the initial projection data records, in particular individual subsets, can be manipulated instead. How exactly the procedure is carried out can depend in particular on the design of the manipulation operation and/or the initial projection data sets.

In further embodiments of the invention, the generation of the manipulated projection data sets includes forming a manipulated projection data set by applying the manipulation operation to one of the initial projection data sets in each case. In other words, each individual initial projection data record is processed individually with the manipulation operation, independently of other initial projection data records. In other words, the manipulation operation converts an initial projection data record into a manipulated projection data record, wherein at least one, preferably several, manipulated, ie modified, pixel entries are generated. Further pixel entries can be taken from the initial volume data set.

In other embodiments, the generation of the manipulated projection data sets includes forming a manipulated projection data set by the manipulation operation offsetting two or more initial projection data sets with one another. In embodiments, the manipulation operation is thus designed to offset several, preferably two, initial projection data sets, more precisely, the respective image information of corresponding pixels. A manipulated projection data set corresponds here to a new, further projection data set, which preferably does not include any pixel with original image information of one of the calculated initial projection data sets.

In embodiments of the invention, the design of the manipulation operation depends on the examination and/or the examination object. In other words, the manipulation operation includes at least one function dependent on the type of examination and/or the type of object under examination. According to the invention, the change in the initial projection images can therefore be adapted more generally to an image content (even if it cannot be traced back to the examination object per se) or a type of examination. One or more functions can be specified here, for example by manual input by a user, or they can be selected from a number of possible functions. Alternatively, a known automatic evaluation of image information in the initial projection datasets for recognizing, for example, an examination object can be used in a data processing unit and various functions for automatic selection can be stored in the sense of a look-up table for different examination objects. The evaluation of image information can include segmentation and/or classification.

In further embodiments of the invention, the manipulation operation comprises at least one non-linear function. Other functions are also possible. In particular, the manipulation function can include a plurality of arithmetic operations or functions which, according to the invention, are executed one after the other. In this respect, the manipulation operation can also correspond to a combination of several functions.

• - Pixelweise lokale oder globale Addition/Subtraktion von konstanten Grauwerten in einem initialen Projektionsdatensatz. Derart wird bspw. kontrastarmer Hintergrund (Muskelgewebe, Objektträger) im Vergleich zu einer kontraststarken Struktur (Knochen, Stent, elektronisches Bauteil) mehr oder weniger betont. Dies kann bspw. bei Dental-CT-Aufnahmen Anwendung finden, wenn es darum geht, Bewegungsartefakte durch eine Schluckbewegung zu unterdrücken, wenn eine Optimierung der Rekonstruktion der Zähne, anstatt einer Bewegungskompensation der Zunge erreicht werden soll.
• - Pixelweise lokale Grauwertoperation, bspw. Setzen eines oberen oder unteren Schwellwertes, um bspw. dominante Hochkontraststrukturen in den initialen Projektionsdatensätzen zu betonen oder zu unterdrücken. Bspw. kann derart ein Signal von Objekthalter, Liege, Kabel oder dergleichen unterdrückt werden, welches sich erfahrungsgemäß anders als das Untersuchungsobjekt insbesondere in Form eines Patienten bezüglich Bewegung oder physikalischer Artefakte wie bspw. einer Strahlaufhärtung verhält.
• - Nachbarpixel-Operation bspw. umfassend eine Gradientenbildung. Auch diese kann eine Unterdrückung eines Hintergrundes bzw. dazu gehöriger Objekte bewirken. Insbesondere eine Gradientenbildung kann in Ausführungen auch eine Betragsbildung des Gradienten umfassen.
• - globale oder lokale Histogrammequalisierung, um gewünschte Bildmerkmale zu betonen und/oder unerwünschte Bildmerkmale zu reduzieren.
• - Pixelweise Verrechnung von Projektionsdatensätzen. Bspw. können (in Bezug auf die Abbildungsgeometrie benachbarte) Projektionsdatensätze innerhalb einer Untermenge von initialen Projektionsdatensätze miteinander verrechnet werden. Alternativ können Projektionsdatensätze einer Untermenge mit Projektionsdatensätzen einer anderen Untermenge verrechnet werden. Die zu den Untermengen der initialen Projektionsdatensätze gehörigen initialen Sätze Abbildungsparamater können sich dabei bspw. im Wert wenigstens eines Abbildungsparameters unterscheiden, bspw. der bei der Bilddatenerfassung eingestellten Röntgenröhrenspannung. Oder sie können sich in wenigstens einem geometrischen Parameter unterscheiden. Ein initialer Satz Abbildungsparameter kann dabei insbesondere ein rein virtueller Parametersatz sein (entsprechend einem virtuellen initialen Projektionsdatensatz), dann unterscheiden sich die initialen Sätze der Abbildungsparameter in den Werten aller Abbildungsparameter. Eine einfache Art der Verrechnung ist bspw. eine Subtraktion von korrespondierenden Pixelwerten.

The at least one function can be a function from the non-exhaustive group of the following functions:

• - Pixel by pixel local or global addition/subtraction of constant gray values in an initial projection data set. In this way, for example, a low-contrast background (muscle tissue, slide) is more or less emphasized in comparison to a high-contrast structure (bones, stent, electronic component). This can be used, for example, in dental CT recordings when it is a matter of suppressing movement artifacts caused by a swallowing movement, when the aim is to optimize the reconstruction of the teeth instead of compensating for movement of the tongue.
• - Pixel-by-pixel local gray value operation, e.g. setting an upper or lower threshold value, in order to emphasize or suppress dominant high-contrast structures in the initial projection data sets, for example. For example, a signal from the object holder, bed, cable or the like can be suppressed in this way, which experience has shown to behave differently than the examination object, particularly in the form of a patient, with regard to movement or physical artefacts such as beam hardening.
• - Neighboring pixel operation, for example, including gradient formation. This can also suppress a background or associated objects. In particular, a gradient formation can also include an absolute value formation of the gradient in embodiments.
• - Global or local histogram equalization to emphasize desired image features and/or reduce unwanted image features.
• - Calculation of projection data sets per pixel. For example, projection datasets (adjacent in terms of imaging geometry) can be offset against one another within a subset of initial projection datasets. Alternatively, projection data sets from a subset can be offset against projection data sets from another subset. The initial sets of imaging parameters belonging to the subsets of the initial projection data sets can differ, for example, in the value of at least one imaging parameter, for example the x-ray tube voltage set during image data acquisition. Or they can differ in at least one geometric parameter. An initial set of mapping parameters can in particular be a purely virtual parameter set (corresponding to a virtual initial projection data set), then the initial sets of mapping parameters differ in the values of all mapping parameters. A simple type of calculation is, for example, a subtraction of corresponding pixel values.

In further, particularly preferred embodiments of the invention, generating the corrected set of imaging parameters includes applying at least one cost function based on a consistency condition to the manipulated projection data sets and simultaneously to an initial set of imaging parameters.

Basically, the application of a consistency condition exploits the existence of data redundancies in the initial projection data sets or their 3D Radon transform. In other words, global or local consistency conditions can be derived based on the initial projection data sets. The assumption is that inconsistencies in the data are due to a discrepancy between actual, real and the initial set of mapping parameters. These deviations result in an increase in the value of the cost function. The application of the cost function includes an optimization of the same by adapting at least one, preferably a large number of the initial mapping parameters. The optimization thus includes searching for or approaching a minimum of the cost function while varying the value of at least one mapping parameter. Finding a function minimum corresponds at the same time to finding or determining a value of a corrected imaging parameter. In this case, the cost function can preferably be applied iteratively in relation to the value of an imaging parameter and/or iteratively in relation to the plurality of imaging parameters for which the function minimum is sought. In particular, an imaging parameter can be optimized as a function of another imaging parameter to be optimized. Each of the iteration loops can be terminated after a predetermined number of iterations or after the value of the cost function falls below a predetermined value.

A subset, a subset of mapping parameters of the initial set of mapping parameters, is thus determined for the optimization of mapping parameters using the cost function. The subset includes at least one mapping parameter, preferably a plurality of or all mapping parameters of the initial parameter set. The optimization now includes the calculation of the cost function for a large number of modified values of the at least one imaging parameter, until the minimum of the cost function or a previously defined termination criterion is reached. The value of the at least one mapping parameter that leads to a minimum of the cost function and/or with which the termination criterion is reached is then used according to the invention as the value of the at least one mapping parameter of the corrected set of mapping parameters.

• Bspw. kann in die Auswahl des zu optimierenden Abbildungsparameters die der Bildgebung zugrunde liegende Aufgabenstellung oder Anamnese mit einfließen. Bspw. kann der zu optimierende Abbildungsparameter indirekt über ein Bildaufnahme-Protokoll festgelegt sein, dass ein Benutzer aus einer Liste möglicher Protokolle in Abhängigkeit des Untersuchungsobjektes auswählt (bspw. ein Protokoll für die Bildgebung eines Elektronikbauteils, eines Steckers, eines Thorax, eines Kopfes oder eines Herzens).

In embodiments of the invention, the at least one imaging parameter to be optimized can be selected manually by a user (application engineer, MTRA) or at least partially automatically as part of data processing. According to the invention, the selection can advantageously be based on the following considerations/assumptions:

• For example, the task or anamnesis on which the imaging is based can be included in the selection of the imaging parameter to be optimized. For example, the imaging parameter to be optimized can be defined indirectly via an image recording protocol that a user selects from a list of possible protocols depending on the examination object (e.g. a protocol for imaging an electronic component, a plug, a thorax, a head or a heart).

Alternatively, the imaging parameter to be optimized is selected based on an (erroneous) first reconstruction of initial volume data. This is now analyzed by a user visually or automatically by the data processing unit using appropriate image or pattern recognition methods. Such identified imaging errors can be traced back to underlying imaging errors in this way and the at least one imaging parameter to be optimized can thus be determined.

• - Indirekt: Bspw. kann in der Dental-Computertomographie basierend auf den initial rekonstruierten Bilddaten durch einen Benutzer die Korrektur „Metallartefaktreduktion“ ausgewählt werden, wobei dann wenigstens ein physikalischer Parameter bezüglich des Materials des Untersuchungsobjektes optimiert wird.
• - Indirekt: Bspw. kann in der Industrie basierend auf den initial rekonstruierten Bilddaten durch einen Benutzer die Korrektur „Multimaterial-Strahlaufhärtungsreduktion“ ausgewählt werden, wobei dann wieder wenigstens ein physikalischer Parameter bezüglich des Materials des Untersuchungsobjektes optimiert wird.
• - Indirekt: Benutzer erkennt in den initial rekonstruierten Bilddaten bewegungsbedingte Artefakte wie bspw. unscharfe Kanten und wählt die Korrektur „Bewegungskompensation“ (bei lebenden Objekten) oder „Jitter-Korrektur“ (bei industriellen Anwendungen) aus, wobei die Projektionsmatrix, also wenigstens ein geometrischer Abbildungsparameter für jeden einzelnen Projektionsdatensatz optimiert wird.
• - Indirekt: Benutzer erkennt in den initial rekonstruierten Bilddaten Doppelstrukturen und wählt die Korrektur „Auto A-lignment“, wobei die Projektionsmatrix (geometrische Abbildungsparameter) für jeden einzelnen Projektionsdatensatz optimiert wird.

This can be done either directly or indirectly, that is, a parameter to be optimized or a preferred correction method can now be determined directly.

• - Indirectly: For example, in dental computed tomography, a user can select the “metal artifact reduction” correction based on the initially reconstructed image data, with at least one physical parameter relating to the material of the examination object then being optimized.
• - Indirect: For example, in industry, based on the initially reconstructed image data, a user can select the “multi-material beam hardening reduction” correction, with at least one physical parameter relating to the material of the examination object then being optimized again.
• - Indirect: The user recognizes motion-related artifacts in the initially reconstructed image data, such as blurred edges, and selects the correction "motion compensation" (for living objects) or "jitter correction" (for industrial applications), whereby the projection matrix, i.e. at least one geometric Imaging parameters are optimized for each individual projection data set.
• - Indirect: The user recognizes double structures in the initially reconstructed image data and selects the "Auto A-lignment" correction, whereby the projection matrix (geometric imaging parameters) is optimized for each individual projection data set.

• - Klinische Computertomographie: Geometrische und physikalische Abbildungsparameter sind sehr genau und oft bekannt und unterliegen keinen relevanten Schwankungen. Auch die physikalischen Eigenschaften von menschlichen Untersuchungsobjekten sind sehr gut erforscht und genau bekannt. Dafür tritt unwillkürlich Patientenbewegung auf (Atmung, Herzschlag, Zappeln) .
• - C-Bogen: Hier unterliegen die geometrischen Abbildungsparameter des Gerätes aufgrund mechanischer Instabilität und bewegungsbedingt induzierten Eigenschwingungen starken Schwankungen.

The selection of the imaging parameter to be optimized can also take into account the imaging system used and its possibly characteristic imaging parameters, which are subject to uncertainties, while other characteristic imaging parameters are very reliable. Examples:

• - Clinical computed tomography: Geometric and physical imaging parameters are very accurate and often known and are not subject to any relevant fluctuations. The physical properties of human examination objects are also very well researched and well known. Instead, involuntary patient movement occurs (breathing, heartbeat, fidgeting).
• - C-arm: Here the geometric imaging parameters of the device are subject to strong fluctuations due to mechanical instability and natural vibrations induced by movement.

In industrial applications, the geometric imaging parameters are typically affected and must be optimized because high spatial resolutions are to be achieved. This requires a small X-ray focal spot, which results in low X-ray power and a long measurement time. The longer a measurement lasts, the greater the temperature fluctuation effects (e.g. X-ray tube and X-ray detector) and signs of wear (especially X-ray tube), which can cause a continuous change in various imaging parameters.

In particular, in the German patent DE 102013200329 B4 or in AICHERT et al., "Redundancies in X-ray images due to the epipolar geometry for transmission imaging", The third international conference on image formation in X-ray computed tomography, pp. 333-337 or in MAASS, N. ET . AL., "Geometrical Jitter Correction in Computed Tomography", In: CT-Meeting, pp. 338-342, 2014, the methods described therein and the cost functions based on the consistency condition corresponding to the epipolar geometry are applied. Alternative consistency conditions that can also be used according to the invention are, for example, the Helgason-Ludwig consistency condition, described in: S. Helgason, "Support of radon transforms," Advances in Mathematics, vol. 38, no. 1, pp. 91-100, 1980, or the Fourier consistency condition described in: PR Edholm, RM Lewitt, and B. Lindholm, "Novel properties of the fourier decomposition of the sinogram," Proceedings SPIE 0671, pp. 8-18, 1986.

Although the consistency condition, as described above, is applied to the manipulated projection datasets, in particularly preferred embodiments of the invention, the consistency condition is selected as a consistency condition based on the initial projection datasets. In other words, a consistency condition is selected that is based on the initial, non-manipulated projection data sets. By using the manipulated projection data sets for parameter optimization, a new, changed cost function with advantageous properties is created.

For example, through the targeted application of a manipulation operation to the initial projection data sets, a cost function can be derived that has a pronounced, clear minimum at a point where a cost function based on the same consistency condition but the initial projection data sets would only have an insignificant secondary minimum. By applying the manipulation operation, according to the invention, an effect in the cost function that covers the minimum corresponding to a parameter optimum can be eliminated without a further correction this undesired effect would be necessary before or as part of the parameter optimization described.

According to the invention, the manipulated projection data sets are therefore used as input data for a consistency condition that starts from the initial, that is to say unmanipulated, projection data sets. This creates a new cost function with advantageous properties.

It should be noted that the manipulation operation is preferably determined empirically and application-specifically and is typically not consistent with the theoretical derivation of the respective consistency condition on which the cost function is based.

In further embodiments of the invention, the generation of a corrected volume data set includes a back-projection of at least one subset of the plurality of projection data sets corresponding to the corrected set of imaging parameters. This step relates to reconstructing the corrected volume data set according to the corrected set of imaging parameters. In embodiments, only a subset of the initial projection datasets is used for the reconstruction. In other versions, more or all subsets of the projection data records are used individually for a reconstruction. The reconstruction process can include a filtered back projection (FBP-Filtered Back Projection), preferably an FDK (Feldkamp-Davis-Kress) algorithm, which is suitable in particular for the reconstruction of projection data sets acquired by means of cone beam geometry. The reconstruction process can also include an iterative reconstruction rule, in particular an iterative reconstruction rule that uses filtering.

In summary, the present invention is based on the finding that parameter optimization can be carried out and a minimal cost value can be achieved even without a mathematically and theoretically correct consistency function. This is possible in that the initial projection datasets are modified by means of an application-dependent manipulation operation prior to input into a consistency condition based on the initial projection datasets. The manipulation operation is selected in such a way that the cost function based on the manipulated projection data sets has at least one clear minimum corresponding to a parameter optimum compared to the cost function based on the initial projection data sets. Ideally, the manipulation operation should be selected in such a way that no secondary minima occur within a meaningfully selectable search area. In other words, it was recognized that the initial projection data records can be manipulated before being entered into a consistency function in such a way that the derived cost function reacts more strongly to a variation in the desired imaging parameters to be optimized and less significantly to the variation in other imaging parameters. The manipulation operation can be derived and formulated by a person skilled in the art or automatically, for example using the initial projection data or an initial, reconstructed volume data set. The invention thus enables an expansion of the fields of application of consistency conditions and associated cost functions. Furthermore, already existing applications for the automatic quality improvement of imaging systems that record a large number of X-ray images of the same scene can be designed more reliably.

In a further aspect, the invention relates to a device for generating a corrected set of imaging parameters from at least one initial set of imaging parameters. The device includes a data interface. This is designed to receive initial projection data sets representing an examination object. The initial projection data sets are based on the at least one initial set of imaging parameters. The device also includes a data processing unit. This is designed to generate manipulated projection data sets by applying a manipulation operation to at least part of the initial projection data sets and to generate a corrected set of imaging parameters based on the manipulated projection data sets.

In preferred embodiments, the data processing unit is also designed to generate a corrected volume data set based on at least part of the initial projection data sets and the corrected set of imaging parameters.

The data processing unit can be designed as a central or decentralized processing unit or computer. The computing unit can have one or more processors. The processors can be embodied as a central processing unit (a technical term for this is “central processing unit”, or CPU for short) and/or as a graphics processor (a technical term for this is “graphics processing unit”, or GPU for short). Alternatively, the data processing unit can be implemented as a local or cloud-based processing server.

The data interface can generally be designed for data exchange between the data processing unit and other components be. The data interface can be implemented in the form of one or more individual interfaces which, for example, can be a hardware and/or software interface, eg a PCI bus, a USB interface, a FireWire interface, a ZigBee or a Bluetooth -May have interface. The data interface can also have an interface of a communication network, wherein the communication network can have a local area network (LAN), for example an intranet or a wide area network (WAN). Correspondingly, the one or more data interfaces can have a LAN interface or a wireless LAN interface (WLAN or Wi-Fi).

In embodiments of the invention, the device can have a data memory which has a data connection to the data interface. One or more initial volume data records can be stored in the data memory so that they can be called up. The data memory can be embodied as local memory of the device, e.g. as main memory or hard disk. Or be designed as a decentralized or central memory, for example a PACS system of a hospital. Furthermore, the data store can be part of a server system.

In a further aspect, the invention provides an imaging system with a device according to the invention for generating a corrected set of imaging parameters from at least one initial set of imaging parameters. In some variants, the imaging system can be configured as a medical, veterinary-medical, or industrial computer tomography system, as a mammography system, or as an industrial transmission X-ray imaging system or the like, or as the imaging part thereof. The device according to the invention is advantageously integrated into the imaging system. Alternatively, the device can also be located remotely or remotely. The device can be designed, in particular to carry out the method according to the invention, for an imaging system or for a large number of systems.

The invention also relates to a computer program with program code in order to carry out the method according to the invention for creating a corrected volume data set representing an examination subject when the computer program is executed on a computer.

The invention also relates to a computer-readable data carrier with program code of a computer program to execute the inventive method for creating a corrected volume data set representing an examination object when the computer program is executed on a computer, in particular the data processing unit.

A computer program product may include software with source code that still needs to be compiled and linked or that just needs to be interpreted, or executable software code that just needs to be loaded into a computing device for execution. The method can be carried out quickly, identically repeatable and robustly by the computer program product. The computer program product is configured in such a way that it can carry out the method steps according to the invention using the data processing unit. In this case, the data processing unit must in each case have the prerequisites such as, for example, a corresponding working memory, a corresponding processor, a corresponding graphics card or a corresponding logic unit, so that the respective method steps can be carried out efficiently.

The computer program product is stored, for example, on a computer-readable storage medium or stored on a network or server, from where it can be loaded into the processor of the respective data processing unit. The server, memory or network can be connected directly to the data processing unit or can be formed as part of the same. Furthermore, control information of the computer program product can be stored on a computer-readable storage medium. The control information of the computer-readable storage medium can be designed in such a way that it carries out a method according to the invention when the data carrier is used in a data processing unit. Examples of computer-readable storage media are a DVD, a magnetic tape or a USB stick on which electronically readable control information, in particular software, is stored. If this control information is read from the data carrier and stored in a data processing unit, all of the inventive embodiments of the methods described above can be carried out. The invention can also proceed from said computer-readable medium and/or said computer-readable storage medium. The advantages of the proposed computer program product and the associated computer-readable media essentially correspond to the advantages of the proposed method.

• 1 eine schematische Darstellung eines erfindungsgemäßen Verfahrens in einer ersten Ausführungsform,
• 2 eine schematische Darstellung eines erfindungsgemäßen Verfahrens in einer weiteren Ausführungsform,
• 3 eine Vorrichtung zum Erstellen eines korrigierten Volumendatensatzes nach einer Ausführungsform der Erfindung,
• 4 eine schematische Darstellung einer Bildgebungsanlage in einer Ausführungsform der Erfindung,
• 5 die Erzeugung eines korrigierten Volumendatensatzes mit verbesserter Kantengenauigkeit und Bildschärfe aus initialen Projektionsdatensätzen gemäß der vorliegenden Erfindung, und
• 6 eine Gegenüberstellung der auf initialen und erfindungsgemäß manipulierten Projektionsdatensätzen basierenden Kostenfunktionen umfassend mehrere Funktionsminima.

The properties, features and advantages of this invention described above, and the manner in which they are achieved, will become clearer and more clearly understood in connection with the following description of the exemplary embodiments, which are given in connection with FIGS Drawings are explained in more detail. This description does not limit the invention to these exemplary embodiments. The same components are provided with identical reference symbols in different figures. The figures are generally not to scale. Show it:

• 1 a schematic representation of a method according to the invention in a first embodiment,
• 2 a schematic representation of a method according to the invention in a further embodiment,
• 3 a device for creating a corrected volume data set according to an embodiment of the invention,
• 4 a schematic representation of an imaging system in an embodiment of the invention,
• 5 the generation of a corrected volume data set with improved edge accuracy and image sharpness from initial projection data sets according to the present invention, and
• 6 a comparison of the cost functions based on initial and manipulated projection data sets according to the invention, comprising several function minima.

1 shows a schematic representation of a method according to the invention in an embodiment of the invention. The method is computer implemented. The method serves to generate a corrected set of imaging parameters KAP or to create a corrected volume data set KVD representing an examination subject 10. The corrected volume data set KVD is characterized by high image quality. The process involves a number of steps.

A first step S10 relates to providing initial projection data records IPD. These are based on at least one initial set of mapping parameters IAP.

Providing may include providing, sending, transmitting to a device 30 . The initial projection data records IPD can be stored in a local, central or decentralized storage unit, for example storage unit 60 or 36, and can be made available from there, for example via data interface 32. Alternatively, the initial projection datasets IPD are recorded by means of an imaging system 20 of the transmission X-ray according to an initial set of imaging parameters IAP and are then made available to the device 30 directly by the imaging system 20 .

The initial set of imaging parameters IAP is a parameter set comprising geometric and physical imaging parameters and parameters characterizing imaging properties of the examination object 10. An initial set of imaging parameters IPA includes at least one, typically several, fictitious or virtual imaging parameters whose exact values are not known and for which fictitious ones are used instead or virtual values are applied/estimated. According to the invention, these values should be determined exactly.

An initial mapping parameter can be specified by a specific value or specific number, but also by a function.

Step S10 also includes providing the initial set of mapping parameters IAP. In particular, the provision of the initial set of mapping parameters IAP is based on knowledge of exact, perfect mapping parameter values. This can be available or accessed in various ways.

If, for example, there is precise knowledge of the imaging system 20 used, e.g. via a calibration method and/or of the examination object 10, values for e.g. geometric parameters and/or imaging properties of the examination object 10 can be specified exactly.

Without this knowledge, fictitious, virtual estimated values are specified for imaging parameters. This specification is made manually by a user based on empirical values and/or a data sheet of the imaging system 20 and/or an inspection of the same and/or the examination object 10. Alternatively or additionally, the specification is made at least partly. also automatically by a data processing unit 34, for example with automatic evaluation of a user input or an electronic data sheet.

A further step S20 relates to generating manipulated projection data records MPD by applying a manipulation operation MO to at least a portion of the initial projection data records IPD.

The manipulation operation MO includes an arithmetic operation or function dependent on the type of examination and/or the type of the examination object 10 . The manipulation operation MO particularly preferably includes at least one non-linear function. The manipulation operation MO changes image information from initial projection data sets IPD, preferably pixel by pixel. In other words, the manipulation operation MO at least manipulates the pixel entry of a pixel of an initial projection data set IPD. In this case, pixel entries of other initial projection data sets or the same projection data set can be taken into account. The manipulation operation MO is determined empirically according to the invention and is based on empirical values of a user.

Generating the manipulated projection datasets MPD includes applying the manipulation operation MO to all or only one or more subsets of the initial projection datasets.

In this case, the manipulation operation MO can be designed in variant 1 such that a manipulated projection data set MPD is based solely on an offsetting of image information, be it just one or more preferably adjacent pixels, of just an initial projection data set IPD. Otherwise, the manipulation operation MO can be designed in variant 2 in such a way that a manipulated projection data set MPD is based on the calculation of image information from two or more initial projection data sets IPD. The manipulation operation MO can also be designed in such a way that it initially continues to use a provisional, manipulated projection data record generated according to variant 1 or 2 in order to generate a further manipulated projection data record MPD according to variant 2.

A further step S30 relates to generating a corrected set of imaging parameters KAP based on the manipulated projection data sets MPD. In order to derive the corrected set of imaging parameters KAP, the initial set of imaging parameters IAP is adjusted or changed using the manipulated projection data sets MPD. In particular, the values or the function of individual imaging parameters are adjusted or changed.

To this end, an optimization method can advantageously be used which utilizes data redundancy in the manipulated projection data records MPD and/or in their Radon transform. Correspondingly, step S30 can include the application of at least one cost function KF based on a consistency condition, in particular the epipolar geometry consistency condition.

This procedure is based on the consideration that deviations between redundantly present object information are based on a deviation of at least one imaging parameter from its exact value, for example due to a misalignment of the imaging system 20 . This deviation is compensated for with the procedure described.

The optimization includes varying the value of at least one, preferably several, of the mapping parameters of the initial set of mapping parameters IAP in order to seek a minimum of the cost function KF. The cost function KF is preferably optimized iteratively. The function minimum can first be sought individually for each desired mapping parameter, then for the next mapping parameter, etc. Optimization can be terminated after a predetermined number of iterations has been run or after the value falls below a previously defined function value. The values of the corrected mapping parameters KAP correspond to the values that led to the minimum of the cost function KF being reached.

According to the invention, the consistency condition is selected as a consistency condition based on the initial projection data sets IPD and is then nonetheless applied to the manipulated projection data sets MPD. The manipulated projection data sets MPD are therefore included in a consistency condition or the associated cost function, which starts from the initial, that is to say unmanipulated, projection data sets IPD. The use of the manipulated projection data sets MPD in the parameter optimization violates the mathematical-theoretical derivation of the consistency condition. Nevertheless, this results in a changed cost function KF with advantageous properties. This shows a pronounced, clear minimum at a point where a cost function KF' based on the same consistency condition but the initial projection data sets IPD would only have an insignificant secondary minimum. A function minimum corresponding to a parameter optimum in the cost function KF thus clearly emerges, since an effect covering this minimum was eliminated using the manipulated projection data sets MPD. A correction of this undesired effect before or as part of the described parameter optimization can be omitted.

A step S40, which is optional for the method according to the invention and is separated from the other steps by a dashed line, relates to the generation of the corrected volume data set KVD based on at least part of the initial projection data sets IVD and the corrected set of imaging parameters KAP. For this purpose, a known reconstruction algorithm is used that includes a back-projection of the initial projection datasets IPD according to the corrected imaging geometry KAG. This reconstruction algorithm can include a filtered back projection, such as the FDK algorithm. The reconstruction algorithm can also include an iterative reconstruction. According to the invention at least one subset, in particular a subset of the initial projection data sets IPD, is back-projected according to the corrected set of imaging parameters KAP.

As an alternative to this, the corrected set of imaging parameters generated here can also be supplied to one of the further application options mentioned at the outset.

2 shows a schematic representation of a method according to the invention in a further embodiment. Step S10, the provision of the initial projection data sets IPD, here includes generating projection data sets by forward projection of an initial volume data set IVD corresponding to a virtual initial set of imaging parameters IAP. The initial volume data set IVD is a completely reconstructed volume data set of an examination object 10. The raw data of the image data acquisition, ie the actual, real projection data sets or the imaging parameters used in imaging are at least partially unknown or not available. In this respect, the initial projection data sets IPD are virtual projection data sets in this embodiment. The initial volume data set IVD has a low imaging quality, with the imaging errors being due to deviations between real, unknown imaging parameters and theoretical, correct imaging parameters.

The initial set of imaging parameters IAP is also a completely or partially virtual parameter set or one based on estimated values, in particular when there is no knowledge of the examination object 10 and/or the imaging system 20 used. In this embodiment, individual imaging parameters can instead be specified based on the initial volume data set IVD either manually by a user or automatically by the data processing unit 34 , for example. For a manual determination, it can be provided that the initial volume data set IVD is displayed to the user via a user interface, for example comprising a display, for a visual assessment. Selection inputs relating to the parameterization can then be received via the user interface, which also includes a keyboard, for example. As an alternative to this, an automatic image evaluation method can be applied to the initial volume data set IVD. This can capture and classify image features. In particular, for example, object recognition including segmentation can be carried out on the initial volume data set IVD in order to recognize represented structures, tissue, organs or the like. On the basis of the recognized image features or recognized objects/organs, parameterization, ie the selection of a set of predefined imaging parameters, can then take place and/or be displayed to the user for selection and/or confirmation.

In step S20, manipulated projection data sets MPD are generated using a manipulation operation MO. In step S30, these flow into a cost function KF based on a data consistency condition. In this embodiment, the cost function KF varied on the basis of the manipulated projection data sets MPD is checked beforehand. If the manipulated projection data sets do not bring about the desired function minimum, step S20 is repeated at least once more via a repetition loop REP with a modified manipulation operation MO, until the cost function KF shows the desired profile. In other words, different manipulation operations MO for generating multiple sets of manipulated projection data records MPD are tested here.

Steps S30 and S40 are carried out only when the cost function shows the desired course. Steps S30 and S40 correspond to the embodiment according to FIG 1 .

3 12 schematically shows a device 30 for creating a corrected volume data set KVD, for example of the examination subject 10, based on initial projection data sets IPD, according to an embodiment of the present invention. The initial projection data sets IPD are based on at least one initial set of imaging parameters.

Device 30 here has a data processing unit 34 , a data memory 36 and the data interface 32 . Device 30 is designed as a data processing device that is detached from an imaging system 20 . The data interface 32 is designed to receive initial two-dimensional projection data sets IPD or also an initial three-dimensional volume data set IVD from the data memory 36 and to transmit them to the data processing unit 34 . In the present case, the data memory 36 is designed as a sub-component of the device 30; alternatively, it can be designed as a separate, for example central, memory unit. The device 30, in particular the data processing unit 34, is set up to provide a method for creating a corrected volume data set KVD representing the examination object 10 based on initial projection data sets IPD according to an embodiment of the present invention and in particular according to an embodiment 1 to execute.

The data processing unit 34 is also designed to interact with a computer-readable storage medium 38 in order to carry out the method according to the invention in one embodiment according to a computer program with program code. The computer program can be stored on the computer-readable storage medium 38 in a retrievable manner. The computer-readable storage medium 38 can be embodied, for example, as a CD, DVD, Blu-ray disk, memory stick or hard disk.

A computer program comprising program code can alternatively be stored in the data memory 36 of the device 30 . The computer program is designed to execute a method according to the invention with all method steps when the computer program is executed by device 30, in particular data processing unit 34.

4 12 shows a schematic representation of an imaging system 20 in an embodiment of the present invention.

The imaging system 20 is a transmission X-ray system. In the exemplary embodiment shown, the imaging system 20 includes a device 30 according to the invention for creating a corrected volume data set KVS of an examination object 10. The device 30 for creating a corrected volume data set KVD is related to 3 shown in an embodiment of the present invention and explained in more detail there. The imaging system 20 can be embodied as a computed tomography device or also as a mammography system. The focus here is on medical use. Alternatively, the imaging system 20 can also image non-living examination objects 10, for example an industrially manufactured workpiece, a piece of luggage or the like.

In the exemplary embodiment shown, the imaging system 20 has a radiation source 21 in the form of an X-ray tube for generating X-ray radiation and a 2D detector 23 . This is designed to detect X-ray radiation that has passed through the examination object 10 . The x-ray tube 21 is preferably designed to generate a cone beam and/or a fan beam. For this purpose, the x-ray tube 21 preferably comprises a collimation device in order to collimate the x-ray beam as desired. The 2D detector 23 is designed in particular as a flat detector, i.e. the surface of its X-ray-sensitive area is arranged in one plane. The 2D detector 23 can be designed as a scintillation detector or as a photon-counting x-ray detector. Radiation source 21 and 2D detector 23 are here along a holding device 26 - arranged in particular as a so-called C-arm. Alternatively, the holding device 26 can be designed as a closed gantry. The holding device 26 is designed to move the radiation source 21 and the 2D detector 23 during image data acquisition on a circular path about an axis of rotation R (here the z-axis of a reference coordinate system) around the examination object 10 and thereby a relative movement between the examination object 10 and the To effect radiator-detector pair. Alternative embodiments of the holding device 26 are also possible, for example an embodiment in which the holding device only holds the emitter-detector pair in one position or moves it along a line trajectory in which both are aligned with one another. In embodiments, including the one shown, the examination object 10 can be arranged on a carrier device 28—for example a patient table or an object carrier plate. The carrier device 28 is also part of the imaging system 20. The carrier device can also be designed to adjust the examination object 10 relative to the emitter-detector pair, for example the carrier device can be designed to be rotatable (e.g. about the y-axis of the reference coordinate system) or tiltable be. Furthermore, the carrier device 28 can be designed to be linearly adjustable, in particular horizontally. Combinations are also conceivable. In this way, a relative movement for acquiring a plurality of projection data sets IPD can be generated via the carrier device 28 even when the radiator-detector pair is fixed. Holding device 26 and/or carrier device 28 can be designed to adapt the distance of the emitter-detector pair or emitter 21 or 2D detector 23 to examination object 10 in each case, and thus to adapt a magnification factor of imaging system 20 . When carrying out a transmission X-ray imaging, the imaging system 20 can preferably be used to create a large number of projection data sets IPD in the form of 2D X-ray transmission images, each representing the X-ray attenuation profile of the examination object 10 in the transmission direction or in the transmission region, in that the X-ray radiation from the radiation source 21 after passing through the examination object 10 impinges on the 2D detector 23 and is detected there.

The device 30 for creating a corrected volume data set KVD of an examination subject 10 is designed as part of a control and monitoring device 40 of the imaging system 20 . The control and monitoring device 50 comprises a data processing unit 50 which is designed to transmit control signals for the imaging system, for example for adjusting the holding device 26 or for controlling the radiation source 21 or to create something like that. The data processing unit 50 can also be designed to acquire projection data sets IPD of a transmission imaging examination and to generate a three-dimensional volume data set representing the examination object 10 using a reconstruction rule. The control and monitoring device 40 can also include a local storage device 60 in which the completely reconstructed volume data set can be stored, for example by the device 30, so that it can be called up.

Device 30 may be in data communication with data processing unit 50 and/or storage device 60 via data interface 32 . Furthermore, the control and monitoring unit 60 can be in data communication via the data interface 32 with the imaging system 20 , in particular with the 2D detector 28 or the holding device 26 and/or the carrier device 28 .

5 shows the generation of a corrected volume data set KVD with improved image quality from initial projection data sets IPD and a corrected set of imaging parameters KAP according to the present invention.

gebildet. Andere Ausgestaltungen der Manipulationsoperation sind jedoch ebenfalls denkbar.Initial projection data IPD are converted into manipulated projection data records MPD by means of a manipulation operation MO. The manipulation operation MO used here as an example includes various functions that are combined. The manipulation operation MO first forms the difference between two initial projection data sets p ₁ and p ₂ pixel by pixel. Then a manipulated projection data set MPD as the amount of the gradient in the two dimensions u and v of the 2D difference projection image according to the rule $$\left|\frac{\mathrm{i.e}}{{\mathrm{i.e}}_{\mathrm{and}}{\mathrm{i.e}}_v}\left[p_2\left(\mathrm{and},v\right)-p_1\left(\mathrm{and},v\right)\right]\right|$$

educated. However, other configurations of the manipulation operation are also conceivable.

The two initial projection data sets p ₁ and p ₂ differ in their image information. That also means that the value of at least one imaging parameter of the projection data sets p ₁ and p ₂ differs. For example, a geometric mapping parameter and/or a physical mapping parameter can be different. The initial projection data set p ₁ here specifically corresponds to a projection data set recorded by means of the imaging system 20 according to a first set of initial imaging parameters. The initial projection data set p ₂ , on the other hand, is calculated from an initial volume data set IVD by forward projection according to a second, virtual set of imaging parameters. The initial volume data set IVD here corresponds to a three-dimensional image data set reconstructed from the projection data set p ₁ and the first set of initial imaging parameters, which has imaging errors.

6 shows a comparison of the cost functions KF, KF′ based on initial and manipulated projection data sets according to the invention, comprising a plurality of function minima M1, M2. The cost functions are each based on a consistency condition that takes data consistency into account. In 5 the consistency condition of the epipolar geometry was chosen. The desired function minimum M2 of the cost function KF' appears as a weak secondary minimum to the further minimum M1. The cost function KF based on the manipulated projection data sets MPD shows a different course. By changing the image information by using the manipulation operation MO, the cost function KF shows the desired minimum M2 at the desired position as the main minimum, while the further minimum M1 recedes into the background as a secondary minimum.

In 5 a set of corrected set of mapping parameters KAP is subsequently formed by varying the desired function minimum M2 of the cost function KF by varying at least one mapping parameter of the initial set of mapping parameters. The initial set of imaging parameters IAP to be optimized corresponds to an initial set of imaging parameters for a subset of the initial projection data sets IPD. In concrete terms, the first set of initial imaging parameters can be optimized in accordance with the projection data set p ₁ . Alternatively, the second, purely virtual set of imaging parameters is optimized in accordance with the projection data set p ₂ .

Finally, in this embodiment, the corrected set of imaging parameters KAP and the subset of the initial projection data sets IVD associated with the optimized set of imaging parameters are included in the reconstruction of a corrected volume data set having an improved image quality.

Alternatively, after creating a corrected set of imaging parameters KAP according to 6 the method according to the invention can also be ended.

In summary, the invention is based on the finding that it is not necessary to work with a mathematically and theoretically correct consistency function, but that it is sufficient that the consistency used function has a minimum in the right place. Secondary minima should be prevented within a meaningfully selectable search area, which can be achieved empirically by changing the manipulation operation MO or the manipulated projection data sets MPD. The invention thus makes it possible to change projection data sets before entering them into a consistency function in such a way that the consistency function reacts more strongly to the desired parameters to be optimized and less significantly to other undesirable effects. In this way, known consistency functions can also be used where they have previously only produced unusable results. Existing applications can be designed more reliably. As a result, the image quality of tomographic three-dimensional x-ray images can be reliably improved in an automated manner.

Where this has not yet been done explicitly, but makes sense and is within the meaning of the invention, individual exemplary embodiments, their individual aspects or features can be combined with one another or exchanged without departing from the scope of the present invention. Advantages of the invention described with reference to one exemplary embodiment also apply to other exemplary embodiments, where applicable, without explicit mention.

QUOTES INCLUDED IN DESCRIPTION

This list of documents cited by the applicant was generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Patent Literature Cited

• DE 102013200329 B4 [0053]

Claims (16)

A computer-implemented method for generating a corrected set of mapping parameters (KAP) from at least an initial set of mapping parameters (IAP), comprising the steps of - Providing (S10; S11, S12) initial projection data sets (IPD) representing an examination object (10), the initial projection data sets (IPD) being based on the at least one initial set of imaging parameters (IAP), - Generating (S20) manipulated projection data sets (MPD) by applying a manipulation operation (MO) to at least part of the initial projection data sets (IPD), and - Generating (S30) a corrected set of imaging parameters (KAP) based on the manipulated projection data sets (MAP). procedure after claim 1 , comprising the further step - generating (S40) a corrected volume data set (KVD) based on at least a part of the initial projection data sets (IVD) and the corrected set of imaging parameters (KAP). procedure after claim 1 or 2 , wherein the provision of the initial projection data sets comprises capturing projection data sets by means of an imaging system (20) of the transmission X-ray according to the at least one initial set of imaging parameters. procedure after claim 1 , 2 or 3 , wherein the provision of the initial projection data sets comprises generating projection data sets by forward projection of an initial volume data set (IVD) corresponding to a virtual initial set of imaging parameters (IAP). A method according to any one of the preceding claims, wherein generating the manipulated projection datasets comprises applying the manipulation operation to all initial projection datasets. Method according to one of the preceding claims, wherein the generation of the manipulated projection data sets comprises forming a manipulated projection data set by applying the manipulation operation to an initial projection data set in each case. Procedure according to one of Claims 1 until 5 , wherein the generation of the manipulated projection data sets includes forming a manipulated projection data set by the manipulation operation offsetting two or more initial projection data sets with one another. Method according to one of the preceding claims, wherein the manipulation operation comprises at least one function dependent on the type of examination and/or the type of examination object. A method according to any one of the preceding claims, wherein the manipulation operation comprises at least one non-linear function. Method according to one of the preceding claims, wherein the generation of the corrected set of imaging parameters comprises applying at least one cost function (KF) based on a consistency condition to the manipulated projection data sets. procedure after claim 10 , where the consistency constraint is chosen as a consistency constraint based on the initial projection data sets. Procedure according to one of claims 2 until 11 , wherein the generation of the corrected volume data set includes a back-projection of at least a subset of the initial projection data sets (IPD) corresponding to the corrected set of imaging parameters (KAP). Apparatus (30) for generating a corrected set of mapping parameters (KAP) from at least an initial set of mapping parameters (IAP), comprising - A data interface (32) designed to receive initial projection data sets (IPD) representing an examination object (10), the initial projection data sets (IPD) being based on the at least one initial set of imaging parameters (IAP); and - a data processing unit (34), designed, - to generate manipulated projection data sets (MPD) by applying a manipulation operation (MO) to at least part of the initial projection data sets (IPD), and - generate a corrected set of mapping parameters (KAP) based on the manipulated projection data sets (MPD). Imaging system (20) with a device (30) for generating a corrected set of imaging parameters (KAP) from at least one initial set of imaging parameters (IAP). Claim 13 . Computer program product, which includes a program and can be loaded directly into a memory (36, 60) of a programmable computing unit of a data processing unit (34), with program means to implement a method according to claims 1 until 11 run when the pro program is executed in the data processing unit. Computer-readable storage medium (38) on which readable and executable program sections are stored in order to carry out all the steps of the method according to one of Claims 1 until 11 to be executed when the program sections are executed by the data processing unit (34).

Images