DE102005037426A1 - Image processing device for use in catheter angiography, has allocation unit assigning two-dimensional data set to N-dimensional data set based on heart action and/or respiration signals representing respective heart and respiration actions - Google Patents

Abstract

The invention relates to an image processing device for assigning a 2-D data record to a portion of an ND data record. DOLLAR A The ND data record is a result of a detection of an object in at least three, in particular spatial dimensions, which at least partially represents the object. The 2-D data set is a result of recording the object in two dimensions, which at least partially represents the object. The 2-D data record can in particular represent a top view of the object, a view through the object or a section through the object. The invention is characterized in that the assignment unit is designed to transfer the 2-D data set received on the input side as a function of a heart action signal received on the input side and representing a heart action or a respiration signal representing a breathing activity or both, a portion of the multi-dimensional, in particular at least three-dimensional, ND- Assign the data record.

Description

The The invention relates to an image processing apparatus for assigning a 2D dataset for a portion of an ND dataset.

Of the ND record is a result of detecting an object in at least three in particular spatial and / or temporal dimensions, which the object at least partially represents. The 2D data set is a result of detecting the object in two dimensions which at least partially represent the object. The 2D data set can in particular be a top view of the object, a look through the object or a cut through the object represent.

The Image processing device also has an object memory for the ND record and an allocation unit. The allocation unit is with the Object memory connected and trained, one input side received 2D data set, in particular from a sequence time-dependent consecutive recorded 2D datasets, allocate a corresponding proportion of an ND data record and from the corresponding assigned portion a 2D ND record to generate and 2D ND record output on the output side. The 2D ND record can determine the proportion of Represent an ND record in a representation in two dimensions.

Out US 2005/0015006 A1 is a method and an apparatus for the Visualization of 2D / 3D-fused image data sets for catheter angiography known.

While one Intervention, for example, a catheter intervention in vessels the problem, preoperative registered 3D images of an object, such as a heart, with while the intervention recorded X-ray, for example a C-arm X-ray system, in accordance bring to.

at the known from the prior art devices for fusing of image records, as in the case of the device known from US 2005/0015006 A1, In particular, the problem arises that in-vivo detected objects due to a patient moving, or as a result of an elastic time-dependent Deforming organs or vessels with preoperatively captured 3D acquisition results not at all or only with difficulty in accordance can be brought to give a surgeon a comparative perspective.

The The object underlying the invention is therefore, a device specify the type mentioned, which is a simple and safe Image fusion of in-vivo acquired 2D acquisition results with preoperative captured 3D acquisition results.

These The object is achieved by the device of the type mentioned, in which the allocation unit is formed, the received on the input side 2D dataset depending on from an input side, representing a heart action Heart action signal or a respiratory activity representative Respiration signal or both, a portion of the multidimensional, In particular, assign at least three-dimensional ND record.

By the capture of 2D datasets dependent on from a body signal, in particular a heart action signal or a respiration signal Advantageously, a secure assignment of registered 2D data records can take place. Further advantageously, an elastic deformation of objects, for example, of a heart.

One ND data set can be obtained, for example, by magnetic resonance imaging (MRI), X-ray computed tomography (CT), positron emission tomography (PET), or single photon emission computerized tomography (SPECT) be.

One Associating a 2D dataset may be beneficial to a corresponding one Proportion of an ND data record.

For example may be a 2D dataset from a C-arm X-ray device, a computed tomograph or a comparable imaging device and a view through an object, a view of the object or represent a section through the object.

The Allocation unit can be from an already generated, the object representing ND record create a 2D dataset corresponding to a review, which by the C-arm X-ray device generated 2D data set corresponds. Contains an object, for example a heart, several vessels to be represented, which in a review are projected to a two-dimensional view, so can through the allocation unit from several, the vessels corresponding proportions be assigned to the ND record and a corresponding review be generated.

In another embodiment, the allocation unit is formed from the ND record assign a complementary to the corresponding 2D dataset share.

For example can advantageous results of imaging methods with mutually complementary contrasts associated image areas with each other for merging become. For example, if a tumor is dark in a 2D dataset displayed and displayed in an ND record bright, so can a merger to better recognize the areas of interest done with a complement of the ND record, which is the tumor does not match.

A View in three spatial Dimensions can be determined by the allocation unit, for example Volume rendering can be generated.

One Heart action signal, for example, by a from the state of Technology known multipoint derivation, in particular a three-point derivative be detected and thus the input side of the allocation unit for disposal be put.

To the image processing device can have an ECG sensor, which is configured to detect and input a heart action signal to generate the ECG signal corresponding to the heart action signal.

In a preferred embodiment the image processing device has a respiration sensor which is designed to detect a breathing activity and a the breathing activity representing Respiration signal to generate. This can advantageously a local Displacement and / or elastic deformation of an object, for example, of a heart due to a respiratory activity be recorded.

In a further preferred embodiment If the respiration sensor is designed to receive the respiration signal in dependence from one representing a heart action To generate heart action signal. As a result, a simultaneous detection is advantageous of heart action and respiration with a small number of electrodes allows. By a sensor for detecting an electrocardiogram, which corresponds to a heart action signal, can advantageously be a 3D detection dependent on another dimension, namely dependent on from the heart action signal.

For example The respiration sensor may have the respiration signal beneficial in Dependence on a heart action waveform, in particular depending from a QT interval, an R-R interval or a Q-R-S interval produce. For this purpose, the respiration sensor can be a waveform analyzer which is formed, periodically recurring signal sections to capture a heart action signal. For example, a waveform analyzer at least one sample holding member having one with the sample holding member connected memory unit for storing sampled signal amplitude values include.

In an advantageous embodiment the respiration sensor is formed, a thoracic impedance and a thoracic impedance signal representing thoracic impedance to create. For example, the respiration sensor may be at least comprise two electrodes, which are formed on a thorax to be created.

Of the Respiration sensor preferably has a to be connected to the electrodes Power source, wherein the power source is formed, in time successive alternating currents each with different frequencies, for example to produce in the range from 500 hertz to 100 thousand hertz.

Further Preferably, the respiration sensor may include a voltage detection unit and a current detection unit each having a Quotient member are connected. The quotient member is to make a quotient of detected voltage and detected current is formed. The Ratio quotient can generate a quotient signal, wherein the quotient signal corresponds to the quotient formed and the detected thoracic impedance represents.

alternative For this purpose, a respiration sensor can be configured, a change of a Thorax circumference to capture. For this purpose, the respiration sensor a belt to enclose a thorax, wherein the belt connected to a strain gauge is. The strain gauge is formed, its electrical properties, in particular its electrical resistance in dependence Wi to change a deformation of the strain gauge.

By the respiration sensor can advantageously be dependent on a respiration thorax movement determined and associated with the thorax movement elastic Deformation of objects to be detected, such as organs be achieved.

Prefers can be a dependent a sequence of 2D data sets acquired via a respiration signal Be formed time interval, which at least one inhalation phase, an exhalation phase or both phases.

Preferably, a detection result of 2D data sets or 3D data sets, respectively in Dependence of each body signal to be complemented by calculating intermediate states between two temporally successive detection results. Such supplementing can take place-for example, by means of an interpolation unit connected to the allocation unit. The interpolation unit is designed to incur such intermediate states by warping or interpolation or both. As a result, a lower sampling rate of a heart action signal or of a respiration signal, and consequently advantageously a data reduction of the ND data set and / or the sequence of 2D data records, is made possible.

In an advantageous embodiment the image processing device is formed, the object in three dimensions in at least one other dimension depending on from a body signal, in particular a motion signal, a respiration signal or a heart action signal assigned. For this purpose, the object memory as a multi-dimensional (ND) memory formed which, for example, preoperatively captured 3D objects in dependence from one of the at least one other dimension corresponding time-dependent signal can save.

The Motion signal, for example, by an accelerometer be generated, which is adapted for attachment to a patient is. Such an accelerometer can advantageously be a 3-axial Be accelerometer. The allocation unit can be input side connected to the accelerometer and be designed dependent on an input received motion signal an assignment perform. As a result, advantageously a patient movement can be detected and reassigning a received after patient movement 2D data record done.

Of the Object memory may be advantageous for storing an object representing in three dimensions Record be formed. Preferably, the object memory for storing the record representing the object in at least be formed another dimension. For example, the Object storage as 4D object storage, as 5D object storage, or as 6D object storage be educated. The further dimension can be a temporal dimension be. For example, a stored in the object memory 5D record an object, such as a body organ, in three spatial Dimensions, taking into account of respiration and heartbeat in two other temporal dimensions and thus in a total of five Dimensions included.

A sixth temporal dimension can be formed by a motion signal be.

In an advantageous embodiment contains the ND record data representing a plurality of voxel object points and the voxel object points together at least partially Represent object in at least three dimensions. A voxel object point corresponds to a location in an object. The voxel object point contains thus information about an object location.

Independent of the aforementioned embodiment variant, in which the ND record contains data representing a plurality of Voxel object points may correspond to an ND record coefficient an approximation function included. The by the with the coefficients certain approximation function at least partially represents the Object in at least three dimensions.

A Approximation function can be achieved, for example, by a spline function, formed a polynomial or another suitable function for approximation be.

Thereby can advantageously be a reduction of the multidimensional to be detected Data can be achieved.

For example A 4D record can be consecutive in time, respectively At various times recorded detection results of an object contained in three dimensions. The fourth dimension is thus through the time is formed.

each the time-dependent consecutive 3D registrations can thus be exactly one point in time within a heartbeat period. In this way can advantageous elastic heart deformations with an in-vivo 2D detection be compared.

Independent of the embodiment variant, at which the allocation unit is formed, the input side received 2D data set in dependence from a body signal, in particular, from a heart action signal or a respiration signal assign, the allocation unit may be formed on the input side received 2D data set in dependence from a similarity parameter assign, with the similarity parameter a similarity between the 2D dataset and the corresponding portion of the ND dataset represents.

In an advantageous embodiment is the similarity parameter a spatial and thus location-dependent formed similarity parameter. A location-dependent formed similarity parameter can be formed by a mutual information.

Mutual information can be through little at least two probability densities may be formed, preferably by a Kullback-Leibler distance between the at least two probability densities.

Prefers the allocation unit may have the similarity parameter generate with a Fourier transformer contained in this so that the similarity parameter at least partially a spectrum of the 2D data set and at least partially corresponds to a spectrum of the ND data set. The spectrum of the records represented Contains object location-dependent formed frequencies. At a location-dependent formed frequency an amplitude - not as usual - in dependence of time, but depending on a place, in this example from a place where represented by the ND record Object.

In another advantageous embodiment the allocation unit has a correlation unit which the similarity parameter can generate. The correlation unit may for example consist of two on the input side received 2D data records Cross correlation produce a correlation result which corresponds to the similarity parameter corresponds and output this output side.

By the correlation unit can advantageously have a fast allocation perform an in-vivo captured 2D dataset to a 2D ND dataset.

One dependent on at least one body signal detected ND-Datzensatz, in particular a 5D record, in particular be saved before a intervention as a 5D look-up table. A dependent on detected at least one body signal Sequence of 2D datasets can in particular be saved before an intervention as a 2D look-up table be. This can be done during intervention can be done quickly. Further advantageous can be an attribution result generated in particular before an intervention between a sequence of 2D datasets and a 5D dataset be.

Prefers is such an assignment result by a DRR (Digitally Reconstructed Radiograph) generated. The allocation unit can be designed for this purpose be such a DRR, in particular by using degrees of freedom, Advantageously exploiting 11 degrees of freedom of a 2D-3D assignment result to create. This can advantageously be a reassignment of a 2D dataset during an intervention is eliminated.

The Allocation unit, for example, as a microprocessor with a Control program be realized. The object store and the cache can each by a random access memory, such as a personal computer be realized.

The Allocation unit may preferably comprise two arithmetic processors, which can each carry out an assignment at the same time. This can, for example a first processor an assignment of a currently detected Execute a 2D dataset, while a second processor an assignment depending from a body signal or a calibration signal or both.

• – Zuordnen des 2D-Datensatzes in Abhängigkeit von einem eine Herzaktion repräsentierenden Herzaktionssignal und/oder einem eine Atmungstätigkeit repräsentierenden Respirationssignal zu wenigstens einem Anteil des ND-Datensatzes;
• – Erzeugen eines 2D-ND-Datensatzes als Zuordnungsergebnis, wobei der 2D-ND-Datensatz den wenigstens einen zugeordneten Anteil des ND-Datensatzes in einer Darstellung in zwei Dimensionen repräsentiert.

The invention also relates to a method for associating a 2D dataset with a 2D ND dataset from an ND dataset, wherein the ND dataset is a result of registration in at least three dimensions and at least partially represents an object in at least three dimensions. The 2D data set is a result of registration in two dimensions which at least partially represents the object, in particular a plan view of the object, a view through the object or a section through the object. The procedure is characterized by the following steps:

• Assigning the 2D data set as a function of a heart action signal representing a heart action and / or a respiration activity representing a respiration activity to at least a portion of the ND data set;
• Generating a 2D ND data set as an assignment result, the 2D ND data set representing the at least one assigned proportion of the ND data set in a representation in two dimensions.

• – Gemeinsamens Darstellen, insbesondere räumlich und/oder zeitlich gemeinsames Darstellen der durch den 2D-Datensatz und durch den 2D-ND-Datensatz jeweils repräsentierten Objekte auf einer Bildwiedergabeeinheit.

The aforementioned method may advantageously additionally comprise the method step:

• Representing in common, in particular spatially and / or temporally joint presentation of the objects respectively represented by the 2D data record and by the 2D ND data record on an image display unit.

The Invention will now be described below with reference to FIGS embodiments explained. Show it:

1 schematically an arrangement with an image processing device for assigning a 2D data set to a portion of a 5D data set;

2 schematically that too 1 described generating a 3D data set from each temporally consecutively detected 2D data sets in response to a heart action signal;

3 schematically that too 1 previously described generation of a 3D data set from a sequence of temporally successively acquired 2D data sets in dependence on a respiration signal;

4 schematically that too 1 described generating a 4D data set from a sequence of temporally successively detected 2D data sets in dependence on a heart action signal and a respiration signal;

5 schematically that too 1 described generating a 4D data set from a sequence of sequentially acquired 3D data sets in response to a respiration signal;

6 generating a 4D data set in response to a heart action signal;

7 generating a 5D data set in response to a heart action signal and a respiration signal;

8th schematically associating a 2D dataset from a 4D dataset with a portion of a 5D dataset; and

9 schematically a method for assigning a 2D data set from a sequence of 2D data sets to a portion of a 5D data set.

1 shows an arrangement 1 with an image processing device 3 , The image processing device 3 has an allocation unit 5 , an object store 7 , a cache 9 , a 2D memory 13 for storing 2D data sets, an image processing unit 11 , and a timer 15 on.

The order 1 also has a C-arm x-ray device 23 for acquiring image data by means of X-radiation and for generating a temporal sequence of 2D data sets.

The order 1 also has a detection device 21 for detecting an object in three spatial dimensions, which is formed in this embodiment as a magnetic resonance tomograph.

The order 1 also includes sensors for detecting body signals, namely an ECG sensor 16 for detecting a heart action signal (ECG = electrocardiogram), a respiration sensor 14 for detecting a respiratory activity and generating a respiration signal representing the respiratory activity and a motion sensor 18 for detecting a body movement, in particular the movement of an object.

The order 1 also has a picture display unit 17 with a touch-sensitive surface 19 on. The touch-sensitive surface 19 may depend on touching it - for example, by a user's hand 62 Generate a touch signal indicating the location of contact of the touch-sensitive surface 19 - Represents and output the touch signal on the output side.

The image processing device 3 has an entrance 33 for a 3D data record signal with which the allocation unit 5 over a connecting line 39 is connected on the input side. The detection device 21 for detecting an object in three dimensions, also referred to below as 3D detection devices, is on the output side with the input 33 connected. The 3D detection device 21 For example, it may be a magnetic resonance tomograph, a single photon emission computer tomograph, a computer tomograph, or a positron emission tomograph.

The allocation unit 5 is via a bidirectional data bus 33 with the object store 7 connected, and via a bidirectional data bus 37 with the cache 9 connected. The allocation unit 5 is on the input side via a connecting line 53 with the touch-sensitive surface 19 the image display unit 17 connected.

The allocation unit 5 is on the output side via a connecting line 49 with an image processing unit 11 connected, which output side via a connecting line 51 with the image display unit 17 connected is.

The image playback unit 17 can be formed for example by a TFT display (TFT = Thin Film Transistor).

The image processing device 3 has an entrance 27 for a 2D data set signal, which is via a connection line 47 with the 2D memory 13 connected is. The 2D memory 13 is on the output side via a connecting line 45 with the allocation unit 5 connected and can generate temporally successive 2D data sets from a received on the input side 2D data set signal and store these successive 2D data sets. The 2D memory 13 can store the stored, time-sequential 2D datasets over the connection line 45 out on the output side to the allocation unit 5 output.

The C-arm X-ray device 23 is configured to generate a 2D data set signal which at least partially is a top view of an object 60 , a view through the object or a section through the object 60 represents and output this output side. For this purpose, the C-arm X-ray device is the output side with the input 27 the image processing device 3 connected.

The C-arm X-ray device 23 can also generate a calibration signal which a detection setting, in particular a detection direction to an object 60 and / or a detection distance to the object 60 equivalent.

The C-arm X-ray device is the output side with the input 25 the image processing device 3 connected and can the calibration signal through the input 25 , which via a connecting line 43 with the allocation unit 5 connected to the allocation unit 5 send.

The image processing device 3 has an entrance 31 for a respiration signal, which via a connecting line 44 with the allocation unit 5 connected is. The entrance 31 is with the respiration sensor 14 connected.

The image processing device 3 also has an entrance 29 for an ECG signal, which via a connecting line 41 with the allocation unit 5 connected is. The ECG sensor 16 is the output side with the input 29 connected.

The image processing device 3 has an entrance 30 for, for example, a body movement representing motion signal, which via a connecting line 42 with the allocation unit 5 connected is. The motion sensor 18 is the output side with the input 30 connected.

The operation of the arrangement 1 will now be explained below:
The 3D detection device 21 For example, before an intervention, it can detect an object, in particular a patient or, for example, a patient's heart, in 3 dimensions and generate a 3D data record signal which represents the object at least partially in 3 dimensions.

The 3D data record signal may be from the 3D acquisition device 21 output on the output side and over the input 33 and the connection line 39 to the allocation unit 5 be sent.

The allocation unit 5 is formed, for example, depending on an input side via the connecting line 53 received touch signal - via the connection line 39 to receive the 3D data set signal and to generate a 3D data set from the 3D data set signal and this via the data bus 33 to the object memory 7 to send and in the object store 7 save.

The assignment device 5 is designed to handle the 3D data sets in response to a via the input 31 received respiration signal and / or in response to a via the input 29 to generate the received ECG signal and the 4D data records or 5D data records thus generated in the object memory 7 save.

The allocation unit 5 is via a connection line 40 with the timer 15 connected. The allocation unit 5 can thus generate the 4D data set or the 5D data record in a time-synchronized manner as a function of time-dependent signals, for example a respiration signal and / or an ECG signal.

Thus, before a planned intervention, a detection result in 3 spatial dimensions in 1 or 2 further temporal dimensions can be stored in the object memory 7 be stored.

Through the C-arm X-ray device 23 For example, the object may be detected - for example, prior to a planned intervention - and a 2D data set signal representative of, for example, a view, a view, or a section through the object may be generated.

Timing successive 2D data sets can be generated from the 2D data set signal and stored in the 2D memory 13 be kept in stock.

The allocation unit 5 can over the connecting line 45 the 2D datasets from the 2D memory 13 received and this over the bidirectional data bus 37 in the cache 9 save.

The allocation unit 5 can - for example, depending on one over the connecting line 53 received touch signal - the 2D data sets from the buffer 9 read out and over the connecting line 49 and the image processing unit 11 to the image display unit 17 send.

A user 62 then, for example, has the option of using the touch-sensitive surface 19 the image display unit 17 an area a 2D data set, which in this embodiment, a heart 60 represents, which with a corresponding proportion of the object memory 7 stored 5D record should be merged.

The allocation unit 5 is designed, depending on a via the connecting line 53 received touch signal to select a part of the 2D data set and this part of the 2D data set exactly a portion of the in the object memory 7 Assign stored 5D record and create a displayable in two dimensions 2D 5D record for merged representation with the 2D data set.

For this purpose, the allocation unit 5 a - not shown in this figure - correlation unit, which is formed, depending on a via the input 31 received respiration signal and / or one via the input 29 Assign received ECG signal to a corresponding portion of the 5D record. Allocation may thus occur simultaneously with in vivo generation of 2D data sets by the C-arm X-ray device 23 in response to a respiration signal detected in vivo and / or a cardiac action signal detected in vivo.

In addition, can the correlation unit be formed, the assignment by means of cross-correlation, Fourier transformation, Contour line comparison or a combination of these perform.

The allocation unit 5 is formed, the assignment result in the case of a 2D-4D fusion as a 2D-4D data set or in the case of a 2D-5D merger, the assignment result as a 2D-5D data set, respectively together with the corresponding 2D data set in the cache 9 save.

The allocation unit 5 may be formed, depending on a via the connecting line 53 received touch signal the previously generated, and in the buffer 9 stored 2D-5D data record and together with the corresponding, used for the assignment 2D data set via the connecting line 49 , the image processing unit 11 , and the connection line 51 for sharing with the image display unit 17 to send. In this way, the data sets to be displayed together, and thus the images represented by the data sets to be displayed together, are fused together.

Thereby a surgeon can while an intervention, such as surgery, and thus captured in-vivo 2D object data with preoperatively acquired 3D object data to compare.

In addition, the allocation unit 5 be formed, 2D datasets on the connection line 45 depending on one over the connecting line 43 received calibration signal in the buffer 9 save. In this way, advantageously acquired before an intervention 2D data sets each having different detection settings of the C-arm X-ray device 23 in the cache 9 be stored. The allocation unit 5 may be formed, depending on the over the connecting line 43 received calibration signal generated 2D data sets and in the object memory 7 stored 5D data record in the manner described above corresponding assignment results in the form of 2D-4D data sets (in the case of assignment depending on a time-dependent body signal) or 2D-5D data sets (in the case of an assignment depending on two time-dependent body signals) to produce , which respectively different detection settings of the C-arm X-ray device 23 correspond.

Regardless of an in vivo assignment described above in dependence on a respiration signal detected in vivo and / or in dependence on a heart action signal detected in vivo, the allocation unit 5 be formed on the input side received 2D data sets in response to a body signal received on the input side, in particular a heart action signal or a respiration signal to capture sequentially in time and the detection result as a 4D record in the buffer 9 save.

Alternatively, the allocation unit 5 be formed to record temporally successive 2D data sets in response to 2 body signals, in particular a respiration signal and a heart action signal, and the detection result as a 4D record in the cache 9 save.

The ones in the cache 9 Saved records may be merged with the one in the object store for later merging 7 stored 5D record can be used.

For such Fusion is an in vivo detection of a respiration signal and / or a heart action signal is no longer required because the respiration information or the her zaktions information already in the on the above Way generated 3D record or 4D record is included.

When the detection setting of the C-arm X-ray device 23 During an intervention, the C-arm X-ray may be used direction 23 a modified according to the change calibration signal at the input 25 provide the image processing device. The allocation unit 5 may be formed, a change of the calibration signal, which a change in the detection setting of the C-arm X-ray device 23 corresponds to detecting and reassigning on the input side received 2D data records with one in the object memory 7 stored 5D record in the prescribed manner.

The allocation unit 5 may be formed depending on a via the input 30 received motion signal to perform a previously described assignments. In this way, moving a patient can advantageously be detected. The motion sensor 18 can contain at least one accelerometer, advantageously the motion sensor includes a 3-axial acceleration sensor.

2 shows a 2D dataset 100 which in this embodiment represents an image of a cardiac vessel.

This in 2 illustrated diagram 150 shows a heart action signal curve 110 which represents a heart action signal detected as a function of time. The diagram 150 has a timeline 114 and a signal amplitude axis 112 on. Illustrated by way of example is the heart signal amplitude value 116 at time th1, the cardiac signal amplitude value 118 at time th2 and the Herzsignalamplitudewert 120 at the time thn.

A 2D dataset 100h1 corresponds to the cardiac signal amplitude value 116 and has been detected at time th1. The 2D dataset 100h2 corresponds to the cardiac signal amplitude value 118 and has been detected at time th2. The 2D dataset 100HN corresponds to the cardiac signal amplitude value 120 and has been recorded at time thn. The 2D datasets 100h1 . 100h2 and 100HN together form a sequence of 2D datasets, which are recorded in chronological succession and, for example, from the in 1 illustrated C-arm X-ray device 23 have been generated. The sequence of 2D datasets can be in the 2D memory 13 be saved.

The 2D dataset 100 is thus extended by a detection in response to a heart action signal by a further temporal dimension. The sequence of 2D datasets 100h1 . 100h2 and 100HN together form a 3D dataset.

3 shows the 2D dataset 100 , which has been detected in response to a respiration signal.

The diagram 152 shows a respiration signal curve 140 which represents a respiration signal detected as a function of time.

The diagram 152 has a timeline 124 and a signal amplitude axis 122 on. The respiration signal amplitude value is shown as an example 126 at time ta1, the respiration signal amplitude value 128 at time ta2 and the respiration signal amplitude value 130 at the time tam.

By capturing 2D datasets in response to a respiration signal, a sequence of 2D datasets is formed, representing the 2D datasets 100a1 . 100a2 and the 2D dataset 100am includes. The 2D dataset 100a1 corresponds to the respiration signal amplitude value 126 at time ta1. The 2D dataset 100a2 corresponds to the respiration signal amplitude value 128 at time ta2 and the 2D dataset 100am corresponds to the respiration signal amplitude value 130 at the time tam.

By capturing the 2D dataset 100 depending on a respiration signal, this has been extended by a further time dimension. The sequence of 2D data records formed as a function of a respiration signal thus forms a 3D data record. The 3D dataset can be found in the in 1 shown arrangement 1 in the cache 9 be kept in stock.

4 shows a detection of a 2D data set 100 in response to a respiration signal and in response to a heart action signal. The detection result thus formed is a 4D data set 140 , 4 also shows that in 2 described diagram 150 and that in 3 described diagram 152 ,

The 4D record 140 has 2D datasets, which are in a temporal dimension 142 in response to a heart action signal and in a further temporal dimension 144 have been detected in response to a respiration signal.

The temporal dimension 144 is by the times th1, th2 and thn according to the diagram 150 educated. The temporal dimension 142 is by the times ta1, ta2 and tam according to the diagram 152 formed as a function of a respiration signal.

The 4D record 140 can in the in 1 illustrated cache 9 be stored.

5 schematically shows the detection of a 2D data set 200 depending on a Re spiration signal. By capturing a 3D data set as a function of a respiration signal, a sequence of 3D data sets can be formed, which together form a 4D data record.

The 3D record 200 For example, by the in 1 illustrated 3D detection device 21 be captured.

The sequence of 3D data records formed by detection as a function of a respiration signal comprises the 3D data records in this exemplary embodiment 200a1 . 200a2 , and the 3D record 200am , The 3D record 200a1 corresponds to the respiration signal amplitude value 126 and is formed at time ta1. The 3D record 200a2 corresponds to the respiration signal amplitude value 128 and is formed at time ta2. The 3D record 200am is formed at the time tam and corresponds to the respiration signal amplitude value 130 ,

5 also shows the diagram 152 which is already in 3 is described.

6 shows a 3D dataset 200 which is detected in response to a heart action signal and thus generates a sequence of 3D data sets. The sequence of 3D datasets together form a 4D dataset.

This is already shown in 2 and in 4 described diagram 150 , The 4D data set comprises the 3D data record in this exemplary embodiment 200h1 . 200h2 and 200hn , The 3D record 200h1 corresponds to the cardiac signal amplitude value 116 and is formed at time th1. The 3D record 200a2 corresponds to the cardiac signal amplitude value 118 and is formed at time th2. The 3D record 200hn corresponds to the cardiac signal amplitude value 120 and is formed at the time thn.

7 shows the capture of a 3D data set 200 in response to a heart action signal and in response to a respiration signal. By detecting in response to 2 time-dependent signals, namely a heart action signal and a respiration signal, is a 5D record 136 educated. A temporal dimension 132 of the 5D record 136 has the times th1, th2 and thn. The time dimension 134 has the times ta1, ta2 and tam, and represents the time course of a respiration signal. 7 also shows that already in 2 described diagram 150 and that in 3 described diagram 152 ,

8th schematically shows an embodiment for assigning the in 4 represented 4D data set 140 to the in 7 illustrated 5D record 136 ,

The assignment can be made, for example, by the in 1 described allocation unit 5 respectively. The arrows 160 . 162 and 164 each show the assignment of a 2D dataset from the 4D dataset 140 to a 3D dataset from the 5D dataset 136 , The by the arrows 160 . 162 and 164 labeled 3D records from the 5D record 136 thus correspond to the corresponding assigned portion of the 5D record 136 , In this embodiment, the records are 140 and 136 each independently generated by a heart action signal.

The 4D record 140 and the 5D record 136 can advantageously be stored in each case as a look-up table.

In this embodiment, the data sets represent 136 and 140 each a matrix of image data sets. In case of heart rate signal and respiration signal dependent generated data sets 136 and 140 matrix indices of the data sets correspond to one another, for example, then the 2D data set a1-h1 corresponds to the matrix of the 4D data record 140 the 3D data set a1-h1 of the matrix of the 5D data record 136 ,

An assignment can be made by the in 1 shown allocation unit 5 in response to a heart action signal and a respiration signal, or alternatively - independently of a heart action signal and a respiration signal - based on a similarity parameter. A similarity parameter can be achieved by location-dependent contour line comparison, by cross-correlation or by a comparable similarity measure. For example, a similarity parameter may also be formed by interpolation or by warping.

9 schematically shows a method for assigning a 2D dataset to a portion of a 5D dataset.

In a process section 250 For example, before an intervention, a sequence of 2D data sets can be detected as a function of a heart action signal and as a function of a respiration signal. The acquisition of the 2D data sets can be done by, for example, an in 1 illustrated C-arm X-ray system 23 respectively.

As a next step, each 2D data record is assigned a corresponding 3D data record from a 5D data record. The next step is to save the assignment result as the current 2D-3D registration. The 2D 3D registry For example, in the in 1 illustrated cache 9 be kept in stock.

In a further step 252 can the 2D-3D-registration thus generated in another part of the process 254 be available.

• – Erfassen wenigstens eines 2D-Datensatzes in Abhängigkeit eines Herzaktionssignals und/oder eines Respirationssignals (beispielsweise durch die in 1 dargestellte C-Bogen-Röntgenanlage 23 in Verbindung mit der Zuordnungseinheit 5);
• – Zuordnen eines entsprechenden Anteils aus einem 5D-Datensatz (beispielsweise einem 5D-Datensatz 136, dargestellt in 7);
• – Gemeinsames Darstellen des 2D-Datensatzes mit dem entsprechenden 3D-Datensatz aus dem 5D-Datensatz, insbesondere unter Verwendung der im Verfahrensteil 250 erzeugten 2D-3D-Registrierung.

In the further part of the process 254 During an intervention, the following procedural steps can take place:

• - Detecting at least one 2D data set in response to a heart action signal and / or a respiration signal (for example by the in 1 illustrated C-arm X-ray system 23 in connection with the allocation unit 5 );
• - Assign a corresponding portion from a 5D record (for example, a 5D record 136 represented in 7 );
• - Joint representation of the 2D data set with the corresponding 3D data set from the 5D data set, in particular using the in the process part 250 generated 2D 3D registration.

In one process step 256 During an intervention, additional 2D datasets, such as those in the 1 illustrated C-arm X-ray system 23 be captured and again through the allocation unit 5 be associated with a corresponding proportion of a stored 5D data record.

The process part 260 can depend on an event 262 and / or depending on an event 264 respectively. The event 262 can be a patient movement within an intervention, the event 264 may be a change in the detection setting of the C-arm X-ray system 23 out 1 be, wherein the change in the setting by sending a changed calibration signal to the input 25 out 1 can be done.

• – Erfassen eines 2D-Datensatzes oder einer Folge von 2D-Datensätzen in Abhängigkeit von einem Herzaktionssignal, einem Respirationssignal, einem Bewegungssignal, oder einem Kalibriersignal;
• – Zuordnen eines entsprechenden 3D-Datensatzes aus einem abgespeicherten 5D-Datensatz;
• – Erzeugen eines Zuordnungsergebnisses aus dem 2D-Datensatz und dem entsprechenden 3D-Datensatz;
• – Abspeichern des Zuordnungsergebnisses als 2D-3D-Registrierung.

The process part 260 can thus comprise the following process steps:

• Detecting a 2D data set or a sequence of 2D data sets in response to a heart action signal, a respiration signal, a motion signal, or a calibration signal;
• Assigning a corresponding 3D data record from a stored 5D data record;
• Generating an assignment result from the 2D data set and the corresponding 3D data set;
• - Save the assignment result as 2D-3D-registration.

During an intervention, for example, in one step 272 that in the process section 260 illustrated method in response to a motion signal 262 and / or in response to a calibration signal 264 simultaneously with one in the process part 254 described during an intervention.

The process part 260 can be done for example by its own processor in the background. Alternatively, the process part 254 and the process part 260 take place on a common processor, wherein the method part 260 with a lower priority than the procedure part 254 ,

In one process step 258 This can be done in the process section 260 generated 2D-3D assignment result as current 2D-3D registration in the process part 254 be available.

Regardless of the method described above, a method part 250 detecting a sequence of 2D data sets independently of a heart action signal and independent of a respiration signal. Such a sequence of 2D data sets may be generated, for example, during a predetermined period of time including at least one complete respiratory period and / or one complete heartbeat period. As a result, although movements or elastic deformations of an object represented by the 2D data set are detected, there is no association of each individual detection result with a time of a cardiac activity or a time within a respiratory activity. An assignment of a 2D data record to a 3D data record from a 5D data record takes place in this exemplary embodiment on the basis of a similarity parameter.

An assignment in the process part 254 takes place in this case also independent of a heart action signal and independent of a respiration signal, an assignment is also made here by a similarity parameter.

In the process section 260 in this embodiment, the detection of a 2D data set in response to an event, in this embodiment of motion signal 262 and / or in response to a calibration signal 264 , and irrespective of a heart action signal and a re-breathing signal. The 2D data record is assigned to a portion of the 5D data record in the method section 260 this embodiment variant based on a similarity parameter.

Claims (15)

Image processing device ( 3 ) for assigning a 2D dataset to a portion of ei an ND record ( 140 ), where the ND record ( 136 ) a result of detecting an object ( 60 ) is in at least three dimensions and the object ( 60 ) is at least partially represented in at least three dimensions, and the 2D data set is a result of detecting the object ( 60 ) in two dimensions, which the object ( 60 ) at least partially, in particular a plan view of the object ( 60 ), a view through the object or a section through the object, wherein the image processing device ( 3 ) an object memory ( 7 ) for the ND record ( 136 ) and an allocation unit ( 5 ) and the allocation unit ( 5 ) with the object memory ( 7 ) is connected and formed, a received on the input side 2D data set ( 100 ), in particular from a sequence of time-dependent consecutively acquired 2D data sets ( 100a1 . 100a2 . 100an ), a portion of the ND dataset ( 136 ) and to generate a 2D ND record, which determines the assigned portion of the ND record ( 136 ) in particular represented in a representation in two dimensions and to output the 2D ND data record on the output side, characterized in that the allocation unit ( 5 ) is configured to assign the 2D data set received on the input side to the corresponding portion of the ND data set as a function of a heart action signal received on the input side, representing a heart action, and / or a respiration signal representing a respiration activity. Image processing device according to one of the preceding claims, characterized in that the image processing device is a respiration sensor ( 14 ), wherein the respiration sensor is configured to detect a thoracic impedance and to generate a thoracic impedance signal representing the thoracic impedance. Image processing device according to claim 2, characterized in that the respiration sensor ( 14 ) has a current source to be connected to the electrodes, wherein the current source is adapted to generate temporally successive alternating currents each having different frequencies, in particular in the range of 500 hertz up to 100 thousand hertz. Image processing device according to one of the preceding claims, characterized in that the image processing device comprises an ECG sensor ( 16 ) configured to detect a heart action signal and generate an ECG signal corresponding to the heart action signal. Image processing device according to claim 1 or 4, characterized in that the image processing device comprises a respiration sensor ( 14 ) configured to generate a respiration signal representing a respiratory activity in response to a heart action signal representative of a heart action. Image processing apparatus according to claim 5, characterized characterized in that the respiration sensor is formed, the Respiration signal depending from a heart action waveform, especially in dependence from a QT interval, an R-R interval or to generate a Q-R-S interval. Image processing apparatus according to one of the preceding Claims, characterized in that the image processing device is formed is the object in three dimensions in at least one other Dimension in dependence from a time-dependent Body signal in particular a movement signal, a respiration signal or Assign a heart action signal and the object memory as a multi-dimensional Memory is formed, which 3D objects in dependence from one of the at least one other dimension corresponding time-dependent Can store signal. Image processing apparatus according to one of the preceding Claims, characterized in that the ND record contains data which correspond to a plurality of voxel object points and the voxel object points together at least partially represent the object in at least three dimensions, where a voxel object point corresponds to a location in an object. Image processing apparatus according to one of the preceding Claims, characterized in that the ND record has coefficients of a Contains approximation function and by the approximation function determined with the coefficients at least partially represents the object in at least three dimensions is. An image processing apparatus for mapping a 2D dataset to a 2D ND dataset from an ND dataset, the ND dataset being a result of detecting an object in at least three dimensions and representing the object at least partially in at least three dimensions, and the 2D Dataset is a result of detecting the object in two dimensions which at least partially represents the object, in particular a top view of the object, a view through the object or a section through the object, the image processing device storing an object memory for the ND Data record and an allocation unit and the allocation unit is connected to the object memory and is formed, a 2D data set received on the input side, in particular from a sequence of time-dependent successively detected 2D data sets, allocate a portion of the ND data set and generate a 2D ND data set which displays the assigned portion of the ND data set, in particular in a representation in two dimensions represents output on the output side, characterized in that the allocation unit is adapted to assign the input side received 2D data set in dependence on a similarity parameter, wherein the similarity parameter represents a similarity between the 2D data set and the corresponding portion of the 2D ND data set. Image processing apparatus according to claim 10, characterized characterized in that the similarity parameter a location-dependent similarity parameter which is formed in particular by mutual information. Image processing apparatus according to claim 10, characterized characterized in that the similarity parameter at least partially from a spectrum of the 2D dataset and at least partially formed from a spectrum of the ND data set, wherein the spectrum is location-dependent formed frequencies of the object represented by the records contains. Image processing apparatus according to one of the preceding Claims, characterized in that the ND-data set by magnetic resonance tomography, X-ray computed tomography, Positron emission tomography or single photon emission computer tomography is generated. Method for allocating a 2D dataset to a 2D ND record from an ND record, the ND record being a result of a capture of an object in at least three dimensions and at least partially represents the object in at least three dimensions, and the 2D dataset is a result of detecting the object in two dimensions, which at least partially, in particular the object a view of the object, a view through the object or represents a section through the object, characterized by the steps - Assign of the 2D data set in dependence from one representing a heart action Heart action signal and / or a respiratory activity representative Respiration signal to at least a portion of the ND record. - Produce of a 2D ND record as a mapping result, the 2D ND record the at least one associated portion of the ND record in one Representation represented in two dimensions. Method according to claim 14, characterized, that the procedure additionally the process step comprises: - sharing, in particular spatially and / or timely representation of the through the 2D dataset and through each represented the 2D ND record Objects on a picture display unit.