POSITION DETERMINATION DEVICE FOR DETERMINING THE POSITION OF AN OBJECT WITHIN A TUBULAR STRUCTURE

Description

[0001] The invention relates to a positioning device, a method, and a computer program for determining the position of an elongated medical instrument, such as a catheter, bronchoscope, or guidewire, within a tubular structure, e.g., blood vessels or airways. The invention further relates to an imaging system comprising the positioning device, as well as an imaging method and an imaging computer program.

[0002] In medical applications, positioning devices are known that employ electromagnetic (EM) or optical shape sensing (OSS) tracking techniques to determine the position of an interventional instrument within a patient's body. Optical shape sensing is typically performed by evaluating multiple optical curvature sensors, which in turn are composed of strain sensors arranged concentrically around the fiber's cross-section. However, EM and OSS tracking techniques can be relatively inaccurate due to, for example, inhomogeneities in the EM field or errors in the interpolation and integration of curvature information.

[0003] US patent application US 2015 / 254526 A1 discloses a registration system for registering the coordinate system of a shape-sensing system with a coordinate system of pre-procedural or intra-procedural imaging data. An instrument, such as a catheter, comprises an optical fiber with shape-sensing sensors, wherein the three-dimensional shape of the instrument is determined based on signals from the shape-sensing sensors. The coordinate system of the shape-sensing system is registered with the coordinate system of the imaging data by identifying a stable curvature of the optical fiber and comparing it with a curvature from the pre-procedural or intra-procedural imaging data, aligning the compared curvatures relative to each other.

[0004] US patent application US 2016 / 0302869 A1 discloses a method in which the shape of an instrument that lies at least partially within an anatomical passage is determined. A set of deformation forces is determined for a plurality of sections of the instrument, and this set of deformation forces, together with the determined shape of the instrument and anatomical data describing the anatomical passage, is used to generate a composite model that indicates the position of the instrument relative to the anatomical passage.

[0005] US patent application US 2014 / 0336501 A1 discloses a diagnostic endoscopic imaging support device that compares shape data of tubular tissue and endoscopic path data and uses the result of this comparison together with three-dimensional image data and the endoscopic path data to represent a path of an endoscope in a three-dimensional stereoscopic image.

[0006] It is an object of the present invention to provide a positioning device, a method, and a computer program that enable improved determination of the position of an elongated medical instrument within a tubular structure. A further object of the present invention is to provide an imaging system comprising the positioning device, as well as an imaging method and an imaging computer program.

[0007] The problem is solved by a position determination device for determining a position of an elongated medical instrument within a tubular structure according to claim 1, wherein the position determination device comprises: a first provisioning unit to provide a first distribution of curvature values ​​at several first locations along a path within the tubular structure, a second provisioning unit to provide a second distribution of strain values ​​or curvature values ​​at several second locations along the instrument, a position determination unit to determine the position of the instrument relative to the path based on the first and second distributions.

[0008] The curvature values ​​of the first distribution at the multiple first locations along the tubular structure and the strain values ​​or curvature values ​​of the second distribution at the multiple second locations along the instrument are local values. By determining the position of the instrument within the tubular structure based on this local information, rather than global information, the accuracy of the position determination can be significantly improved, as this method is less sensitive to measurement errors. For example, the accuracy of the position determination can be less sensitive to, and in particular, completely insensitive to, global measurement errors such as drift, field inhomogeneities, etc. Furthermore, the position of the instrument can be determined without, for example, the need to reconstruct the shape of the instrument based on the curvature values ​​of the second distribution.

[0009] The first provisioning unit can be a storage unit in which the curvature values ​​of the first distribution are stored, and the first provisioning unit can be suitable for making the stored curvature values ​​of the first distribution available. However, the first provisioning unit can also be a receiving unit for receiving the curvature values ​​of the first distribution from, for example, a first destination unit and for making the received curvature values ​​of the first distribution available. Furthermore, the first provisioning unit can itself be the first destination unit. The second provisioning unit can also be a storage unit or a receiving unit. Furthermore, the second provisioning unit can be a second destination unit.

[0010] The curvature values ​​of the first distribution are preferably predefined. They may have been determined based on an image of the tubular structure, such as a computed tomography scan, a magnetic resonance imaging scan, or another image. The curvature values ​​of the first distribution can also be determined in another way. For example, in a prior procedure, one or more curvature sensors may be moved within the tubular structure so that they are located at known first positions within the tubular structure, while the curvatures of the tubular structure are measured using the curvature sensors. Another example for generating the curvature values ​​is the simulation of the position of the medical instrument within the tubular structure extracted from the image data.

[0011] The tubular structure can be, for example, a circulatory system, a respiratory system, or another branched tubular structure. The tubular structure can also be an engineering tubular structure, such as a pipeline system, a tunnel system, a cable duct system, etc. The medical instrument is, for example, a guidewire, a catheter, or a bronchoscope.

[0012] In a preferred embodiment, the curvature values ​​of the first and second distributions are scalar. Therefore, the position of the instrument within the tubular structure can be determined without using direction-dependent curvature information. In one embodiment, two local curvatures in two different directions, for example, in the transverse and vertical axes, can be determined at a given location, and these two local curvatures can be combined to obtain a scalar curvature value that is indicative of the local curvature at that location. For example, a scalar mean curvature value or a scalar total Gaussian curvature value can be determined. However, the curvature values ​​can also be vector-valued.For example, the two local curvatures, such as the curvature relative to the vertical axis and the curvature relative to the transverse axis, can be viewed as the two-component curvature values ​​at the respective location.

[0013] It is preferred that the first provisioning unit is adapted to provide a scalar distribution of curvature values ​​along a path within the tubular structure as the first distribution. The second provisioning unit is then preferably adapted as the second distribution to provide a distribution of scalar strain values ​​at the multiple second locations along the instrument, such that the positioning unit determines the position of the instrument relative to the path based on the scalar curvature values ​​of the first distribution and the scalar strain values ​​of the second distribution. Since only scalar strain values ​​are required in this situation, relatively little technical effort is necessary to determine the position of the instrument within the tubular structure.

[0014] It is further preferred that the second delivery unit is adapted to determine the second distribution based on optical signals from optical strain sensors arranged along the instrument. The optical strain sensors are preferably integrated into a single optical single-core fiber. Preferably, only one strain sensor is present at each point along the instrument. This enables the determination of the second distribution of scalar strain values ​​with a further reduced technical effort compared to curvature sensors, and it is also relatively easy for the user to handle.

[0015] If, in another embodiment, the second distribution is a distribution of curvature values, these can be determined, for example, using OSS curvature sensors or groups of EM or RFID sensors. The relative position with respect to other sensors within the sensor group is sufficient for this group of EM or RFID sensors to calculate the local curvatures.

[0016] The first provisioning unit is adapted to a) apply a dissimilarity measure to the first distribution and the second distribution for different possible candidate positions of the instrument relative to the path, wherein the dissimilarity measure is adapted to provide, for each candidate position of the instrument relative to the path, a dissimilarity value indicative of dissimilarity between the two distributions at the respective candidate position, and b) determine the position of the instrument relative to the path based on the dissimilarity values ​​determined for the different candidate positions. In one embodiment, the position determination unit is adapted to determine the candidate position of the instrument relative to the path for which the smallest dissimilarity value has been determined. This can already lead to a relatively accurate determination of the position of the instrument relative to the path.

[0017] Furthermore, the positioning device is adapted to provide multiple first distributions of curvature values ​​for several paths within the tubular structure that arise from branching of the structure. The positioning unit is adapted to determine the position of the instrument relative to one of these paths based on the first and second distributions. The positioning unit is adapted to a) determine, for each path, a distribution of dissimilarity values ​​for different candidate positions of the instrument relative to the respective path, and b) determine the position of the instrument relative to one of the paths based on the dissimilarity values ​​determined for the different paths and different candidate positions.This can lead to an accurate determination of the position of the instrument relative to one of the pathways if the tubular structure has multiple possible pathways in which the instrument could be located.

[0018] The second provisioning unit is adapted to provide multiple second distributions along the instrument for different time points, wherein the multiple second distributions correspond to multiple positions of the instrument relative to a path of the tubular structure. The position determination unit is adapted to a) determine, for each path and for each time point, a distribution of dissimilarity values ​​for different candidate positions of the instrument relative to the respective path, and b) determine the positions of the instrument relative to one of the paths based on the dissimilarity values ​​determined for the different paths, the different time points, and the different candidate positions. The dissimilarity values ​​determined for the different time points and the different candidate positions can be interpreted as a map, with different locations,which are determined by the respective times and candidate positions, in which a dissimilarity value is entered in the map, whereby the position determination unit is adapted, a) provide a path measure which, for a path through the respective map that ends at a candidate position for a latest time and begins at a candidate position for an earlier time, yields a path value which depends on the dissimilarity values ​​along the respective path, b) for each path, determine a path through the respective map using the path measure for which a minimum path value is determined, so that an optimal path is determined for different paths and thus for different maps,and c) to determine the positions of the instrument relative to one of the paths based on the optimal paths determined for the different paths, taking into account the path values ​​determined for these optimal paths. The position determination unit is preferably adapted to provide a path measure that has a component which increases with the increasing sum of the dissimilarity values ​​along the respective path. In one embodiment, the position determination unit is adapted to determine the positions of the instrument relative to one of the paths as defined by the path with the lowest path value of all optimal paths. This can lead to a further improved determination of the position of the instrument within the tubular structure.

[0019] It is also preferred that the positioning unit is adapted to: a) determine optimal paths for different times, such that for different times and different paths an optimal path with a corresponding path value is determined, wherein, in order to determine an optimal path with a corresponding path value for a path and a specific time, the dissimilarity values ​​determined for the specific time and for earlier times and for the candidate positions, which can be understood as a map for the respective path and the specific time, are used; b) at each time, select from the determined optimal paths the optimal path for which a minimum path value has been determined, wherein, before selecting the optimal path with the minimum path value, for an optimal path for which no minimum path value was found at earlier times compared to the path values,which have been determined for the other optimal paths, the path value is increased, and c) to determine the positions of the instrument relative to one of the paths, the positions defined by the selected optimal path. This can lead to a further improved accuracy in determining the position of the instrument within the tubular structure.

[0020] The position determination unit can be adapted to apply the dissimilarity measure by a) determining a spatial first gradient distribution of the first distribution, b) determining a spatial second gradient distribution of the second distribution, c) comparing, for each location along the instrument, the respective gradient of the second distribution with the respective curvature gradient of the first distribution, where the respective candidate position defines which respective gradient of the second distribution is compared with which respective curvature gradient of the first distribution at the respective location, where, for the respective location and for the respective candidate position, a sub-dissimilarity measure is used which depends on i) the directions of the gradients relative to each other and / or ii) the magnitude of the gradient of the first distribution and the magnitude of the gradient of the second distribution.wherein a sub-dissimilarity value is determined for each location by applying the sub-dissimilarity measure, and d) the sub-dissimilarity values ​​determined for a candidate location are summed to determine a respective dissimilarity value for that candidate location. In one embodiment, the value of the sub-dissimilarity measure decreases with increasing similarity of the directions of the two gradients relative to each other. This decrease can be monotonic and / or continuous.

[0021] Furthermore, in one embodiment, the sub-dissimilarity measure provides a first value if the magnitude of the gradient of the first distribution is less than a first magnitude threshold and the magnitude of the gradient of the second distribution is less than a second magnitude threshold, and a second value if the magnitude of the gradient of the first distribution is not less than the first magnitude threshold and / or the magnitude of the gradient of the second distribution is not less than the second magnitude threshold, wherein the first value is less than the second value.The sub-dissimilarity measure can also have two components: a first component that yields a value decreasing with increasing similarity of the gradients, and a second component that yields a value that is smaller when the magnitude of the gradient of the first distribution is less than the first absolute value threshold and the magnitude of the gradient of the second distribution is less than the second absolute value threshold, and a larger value otherwise. The first and second components can be combined linearly, for example, which includes simple addition. These two components can also be combined non-linearly. In one embodiment, the two components are combined multiplicatively. The result of this combination is a scalar value for the sub-dissimilarity measure.

[0022] In another embodiment, the value of the sub-dissimilarity measure decreases as the magnitudes of the gradients of the first and second distributions decrease. That is, this value decreases when the magnitudes of both gradients decrease. This decrease can be monotonic and / or continuous. This value can also be combined linearly or non-linearly with the value of the aforementioned first component. For example, this value can be multiplied by the value of the aforementioned first component, or these values ​​can be added together.

[0023] The positioning unit can also be adapted to apply the dissimilarity measure by a) determining local maxima of the first distribution, b) determining local maxima of the second distribution, c) assigning each local maximum of the first distribution to a local maximum of the second distribution such that the sum of all spatial distances between assigned local maxima is minimal, with each local maximum of the second distribution being assigned to only one local maximum of the first distribution, and d) adding the spatial distances between the assigned local maxima of the first distribution and the assigned local maxima of the second distribution. Furthermore, the positioning unit can be adapted to apply the dissimilarity measure by a) comparing, for each location along the instrument, the respective value of the second distribution with the respective value of the first distribution, where the respective candidate position is defined.which respective value of the second distribution is compared with which respective value of the first distribution at the respective location, using for the respective location and for the respective candidate position a sub-dissimilarity measure that is zero if i) the respective value of the first distribution is above a given first threshold or ii) the respective value of the first distribution is below the first threshold and the respective value of the second distribution is below a given second threshold, and which otherwise has a positive value, and b) the sub-dissimilarity values ​​determined for a candidate position are summed to determine a respective dissimilarity value for that candidate position. It has been found thatthat the determination of the instrument's position within the tubular structure can be further improved by using these dissimilarity measures. However, it is of course also possible to use other dissimilarity measures.

[0024] The position determination unit can also be adapted to apply the dissimilarity measure by applying a cross-correlation, in particular a normalized cross-correlation, to the first distribution and the second distribution for the respective candidate positions. In particular, the position determination unit is adapted to the mean. E D the strain values ​​or curvature values ​​of the second distribution, the mean value E T the curvature values ​​of the first distribution, the standard deviation σ D the strain values ​​or curvature values ​​of the second distribution, the standard deviation σ T to calculate the curvature values ​​of the first distribution and finally the normalized cross-correlation according to the following formula: 1 n ∑ i 1 σ D σ T D i − E D T i − E T , where n is the number of curvature values ​​of the first distribution and the number of strain values ​​or curvature values ​​of the second distribution, where the index i the individual curvature values ​​of the first distribution and the individual strain values ​​or curvature values ​​of the second distribution are indexed, whereby D ( i ) the respective strain value or curvature value of the second distribution and T ( i ) denotes the respective curvature value of the first distribution. The strain value or curvature value D ( i ) the second distribution and the curvature value T ( i ) of the first distribution are each the iThe values ​​that overlap at the respective candidate position are called the -th values. Since the first and second distributions are shifted differently relative to each other for different candidate positions, different first digits of the first distribution and different second digits of the second distribution overlap for different candidate positions. This also generally results in different strain or curvature values ​​of the second distribution and curvature values ​​of the first distribution overlapping, leading to different normalized cross-correlations for different candidate positions. It has been found that using this dissimilarity measure based on a normalized cross-correlation can further improve the determination of the instrument's position within the tubular structure.

[0025] The above-mentioned task is also solved by an imaging system, the imaging system having: a position determination device for determining a position of an elongated medical instrument within a tubular structure according to claim 1, an imaging unit for providing an image of the tubular structure, a visualization generation unit for generating a visualization of the tubular structure based on the provided image and the determined position.

[0026] Furthermore, the above-mentioned problem is solved by a position determination method for determining a position of an elongated medical instrument within a tubular structure according to claim 11, wherein the method comprises: Providing a first distribution of curvature values ​​at multiple first locations along a path within the tubular structure by a first provisioning unit, wherein multiple first distributions of curvature values ​​are provided for multiple paths within the tubular structure; providing a second distribution of strain values ​​or curvature values ​​at multiple second locations along the instrument by a second provisioning unit, wherein multiple second distributions are provided along the instrument for different time points, wherein the multiple second distributions correspond to multiple positions of the instrument relative to a path of the tubular structure; determining, for each path and for each time point, a distribution of dissimilarity values ​​for different candidate positions of the instrument relative to the respective path by a position determination unit.wherein a dissimilarity measure is applied to the corresponding first distribution and the corresponding second distribution for different possible candidate positions of the instrument relative to the path, wherein the dissimilarity measure is adapted to provide, for each candidate position of the instrument relative to the respective path, a dissimilarity value indicative of a dissimilarity between the two distributions at the respective candidate position, wherein the dissimilarity values ​​determined for the different time points and the different candidate positions can be viewed as a map, wherein, for different locations determined by the respective time points and candidate positions, a dissimilarity value is entered in the map, providing a path measure that, for a path through the respective map,which ends at a candidate position for a latest time and begins at a candidate position for an earlier time, provides a path value that depends on the dissimilarity values ​​along the respective path, by the positioning unit, determine a path for each path through the respective map using the path measure for which a minimum path value is determined, so that an optimal path is determined for different paths and thus for different maps, by the positioning unit, and determine the positions of the instrument relative to one of the paths based on the optimal paths determined for the different paths, taking into account the path values ​​determined for these optimal paths by the positioning unit.

[0027] Furthermore, the above-mentioned task is solved by an imaging procedure, the imaging procedure comprising: Providing an image of a tubular structure by an imaging unit, determining a position of an elongated medical instrument within the tubular structure according to the position determination method of claim 11 by a position determination device of claim 1, and generating a visualization of the tubular structure based on the provided image and the determined position by a visualization generation unit.

[0028] Furthermore, the above-mentioned problem is solved by a computer program for determining the position of an elongated medical instrument, wherein the computer program comprises program code means adapted to cause the position determination device according to claim 1 to perform the position determination method according to claim 11 when executed on the position determination device.

[0029] The program code means can be configured to cause an imaging system according to claim 10, comprising the position determination device, to perform the following steps when the computer program is executed on the imaging system: a) providing an image of the tubular structure and b) generating a visualization of the tubular structure based on the provided image and the determined position. In other words, the above-mentioned problem can further be solved by an imaging computer program comprising program code means for causing an imaging system according to claim 10 to execute the imaging method according to claim 12 when the computer program is executed on the imaging system.

[0030] It should be understood that the positioning device according to claim 1, the imaging system according to claim 10, the positioning method according to claim 11, the imaging method according to claim 12, the computer program according to claim 13 and the imaging computer program according to claim 14 have similar embodiments.

[0031] The following describes embodiments with reference to the following figures, wherein Fig. 1 shows a schematic and exemplary embodiment of an imaging system, Fig. 2 shows a flowchart that exemplifies an embodiment of an imaging method, Fig. 3 shows an exemplary image of a tubular structure, Fig. 4 shows an exemplary vascular graph representing the tubular structure and from which the following is shown: Fig. 3 The image shown has been extracted, Fig. 5 shows, by way of example, different first distributions of curvature values ​​for different paths within the tubular structure, Fig. 6 shows, by way of example, a distribution of dissimilarity values ​​for different candidate positions and a path, Fig. 7 shows, by way of example, dissimilarity values ​​for different candidate positions and times for a path within the tubular structure, thereby forming a map, Fig. 8 shows, by way of example, an optimal path through the in Fig. 7 The map shown represents this, and Fig. 9 shows various cost curves as examples.

[0032] Fig. 1 Figure 1 schematically and exemplarily shows an embodiment of an imaging system 1, comprising a position determination device 5 for determining the position of an instrument 4 within a tubular structure, an imaging unit 10 for providing an image of the tubular structure, and a visualization generation unit 11 for generating a visualization of the tubular structure based on the provided image and the determined position. In this embodiment, the tubular structure is a blood vessel structure of a patient 3 lying on a patient bed 2.The positioning device 5 comprises a first provisioning unit 6 for providing curvature values ​​indicative of curvatures of the tubular structure at several first locations along the tubular structure, wherein the first provisioning unit 6 in this embodiment is suitable for determining the curvature values ​​based on the image provided by the imaging unit 10. In particular, the imaging unit 10 is suitable for providing a three-dimensional image of the tubular structure, such as a computed tomography image or a magnetic resonance image, which may have been acquired before the insertion of the instrument 4 into the patient 3; i.e., the provided image may be a pre-interventional three-dimensional image.The first provisioning unit 6 can be suitable for segmenting the tubular structure within the provided image and determining the curvature values ​​based on the segmentation of the tubular structure in the image. These curvature values ​​are local curvature values ​​and form a first distribution.

[0033] In this embodiment, the instrument 4 is a catheter having optical strain sensors at several second locations along the catheter 4. The position determination device 5 includes a second delivery unit 7 for providing strain values ​​based on optical signals received from the strain sensors. The strain values ​​are scalar and indicative of the strains at the respective second locations and are therefore local strain values.

[0034] The positioning device 5 further comprises a positioning unit 8 for determining the position of the instrument 4 within the tubular structure based on the first curvature values ​​and the second strain values. In this embodiment, these first and second values ​​are scalar. In particular, at a specific spatial location, two local curvatures of the tubular structure in two different directions can be determined, and these two local curvatures can be combined to form a scalar curvature value that is indicative of the local curvature at the respective spatial location. For example, a scalar mean curvature value can be determined according to the following equation: K m = 1 2 ⋅ k 1 + k 2 , where K m the scalar mean curvature value is and k 1 k 2. The local curvatures in the two different directions. Additionally, a (Gaussian) curvature value can be used. K t calculated according to the following equation: K t = k 1 k 2 .

[0035] The variables k 1 and k 2 can be considered as the principal curvatures of, for example, the tubular structure, i.e., the eigenvalues ​​of the shape operator of the tubular structure at the respective location.

[0036] Position determination unit 8 is adapted to apply a dissimilarity measure to the first distribution and the second distribution for various possible candidate positions of catheter 4 relative to the path. The dissimilarity measure is adapted to provide, for each candidate position of catheter 4 relative to the path, a dissimilarity value indicative of dissimilarity between the two distributions at the respective candidate positions. Position determination unit 8 is also adapted to determine the position of catheter 4 relative to the path based on the dissimilarity values ​​determined for the different candidate positions. For example, position determination unit 8 can be adapted to determine the candidate position of catheter 4 relative to the path for which the smallest dissimilarity value has been determined.

[0037] The positioning device is also adapted to determine the position of the catheter 4 within the tubular structure even when the tubular structure has multiple paths. In particular, the first provisioning unit 6 is adapted to provide multiple first distributions of curvature values ​​for multiple paths within the tubular structure, and the positioning unit 8 is adapted to determine the position of the catheter 4 relative to one of these paths based on the first distributions and on the second distribution. Preferably, the positioning unit 8 is adapted to determine a distribution of dissimilarity values ​​for different candidate positions of the catheter 4 relative to each path and to determine the position of the catheter 4 relative to one of the paths based on the dissimilarity values ​​determined for the different paths and different candidate positions.

[0038] Furthermore, the second provisioning unit 7 is adapted to provide multiple second distributions along the catheter 4 at different time points, where the multiple second distributions correspond to multiple positions of the catheter 4 relative to a path of the tubular structure. The position determination unit 8 is adapted to determine, for each path and at each time point, a distribution of dissimilarity values ​​for different candidate positions of the medical instrument relative to the respective path, and to determine the positions of the catheter 4 relative to one of the paths based on the dissimilarity values ​​determined for the different paths, the different time points, and the different candidate positions.In particular, the dissimilarity values ​​determined for the different time points and candidate positions can be interpreted as a map, with a dissimilarity value plotted on the map for each of the different locations defined by the respective time points and candidate positions. Position determination unit 8 is then adapted to provide a path measure that yields a path value for a route through the respective map, ending at a candidate position for a latest time point and beginning at a candidate position for an earlier time point. This path value depends on the dissimilarity values ​​along the respective route. Position determination unit 8 is further adapted to determine, for each path, a route through the respective map using the path measure, for which a minimum path value is determined, thus ensuring that an optimal route is determined for each of the different paths.Furthermore, the position determination unit 8 is adapted to determine the positions of the catheter 4 relative to one of the paths based on the optimal paths determined for the different paths, taking into account the path values ​​determined for these optimal paths.

[0039] The positioning unit 8 is specifically adapted to provide a path measure that increases with the increasing sum of the dissimilarity values ​​along the respective path. Furthermore, the positioning unit 8 is preferably adapted to determine, as the position of the catheter 4 relative to one of the paths, the positions defined by the path with the lowest path value of all optimal paths.

[0040] The positioning unit 8 is also preferably adapted to determine optimal paths for different times, so that for different times and for different paths an optimal path with a corresponding path value is determined, wherein, in order to determine an optimal path with a corresponding path value for a path and a specific time, the dissimilarity values ​​determined for the specific time and for earlier times and for the candidate positions, which can be understood as a map for the respective path and the specific time, are used.Furthermore, the positioning unit 8 is preferably adapted to select, at any given time, from the determined optimal paths, the optimal path for which a minimum path value has been determined, wherein, before selecting the optimal path with the minimum path value, the path value is increased for any optimal path for which no minimum path value has been determined at earlier times compared to the path values ​​determined for the other optimal paths. In addition, the positioning unit 8 is preferably adapted to determine the positions of the catheter 4 relative to one of the paths, the positions defined by the selected optimal path.

[0041] To determine a dissimilarity value, the positioning unit 8 can be adapted to calculate gradients of the measured strain sensor profile and gradients of the curvature profile of the tubular structure. The positioning unit 8 can therefore be adapted to apply the dissimilarity measure by determining a spatial first gradient distribution of the first distribution and a spatial second gradient distribution of the second distribution.The position determination unit 8 is then further adapted to compare, for each location along the catheter 4, the respective gradient of the second distribution with the respective curvature gradient of the first distribution, where the respective candidate position defines which gradient of the second distribution is compared with which curvature gradient of the first distribution at that location, and where a sub-dissimilarity measure is used for each location and for each candidate position, which decreases with increasing similarity of the gradient directions relative to each other. This decrease with increasing similarity can, for example, be monotonic.Alternatively or additionally, the sub-dissimilarity measure can yield a smaller value if the magnitude of the gradient of the first distribution is less than a predetermined first magnitude threshold and the magnitude of the gradient of the second distribution is less than a predetermined second magnitude threshold, and otherwise yield a larger value. The sub-dissimilarity measure can therefore have at least one or at least two components, where a first component yields a value that decreases with increasing similarity of the gradient directions, and / or where a second component yields a value that is smaller if the magnitude of the gradient of the first distribution is less than a predetermined first magnitude threshold and the magnitude of the gradient of the second distribution is less than a predetermined second magnitude threshold, and otherwise yields a larger value.This means that for all corresponding gradient values, a penalty term is calculated which is high if a) both gradients point in dissimilar directions and / or b) neither of the two gradients is small, whereby a comparison with a corresponding absolute value threshold is used to determine whether the respective gradient value is small.

[0042] By applying the sub-dissimilarity measure, a sub-dissimilarity value is determined for each location. The position determination unit 8 is also adapted to sum the sub-dissimilarity values ​​determined for a candidate position in order to determine a respective dissimilarity value for that candidate position. This means that the sum of all penalty terms is a measure of the agreement between the strain sensor profile and the curvature profile.

[0043] The positioning unit 8 can also be adapted to apply a different dissimilarity measure by determining local maxima of the first distribution and local maxima of the second distribution. That is, the positioning unit 8 can be adapted to find all local maxima in the strain gauge profile and the curvature profile. The positioning unit 8 is then further adapted to assign each local maximum of the first distribution to a local maximum of the second distribution such that the sum of all spatial distances between assigned local maxima is minimal, with each local maximum of the second distribution being assigned to only one local maximum of the first distribution. For this purpose, the positioning unit 8 can be adapted, for example, to apply a brute-force approach to the assignment.The positioning unit 8 is then further adapted to add the spatial distances between the associated local maxima of the first distribution and the associated local maxima of the second distribution in order to determine a dissimilarity value.

[0044] The positioning unit 8 can also be adapted to apply the dissimilarity measure by comparing, for each location along the catheter 4, the respective value of the second distribution with the respective value of the first distribution, wherein the respective candidate position defines which respective value of the second distribution is compared with which respective value of the first distribution at the respective location, wherein a sub-dissimilarity measure is used for the respective location and for the respective candidate position, which is zero if a) the respective value of the first distribution is above a predefined first threshold or b) the respective value of the first distribution is below the first threshold and the respective value of the second distribution is below a predefined second threshold, and which otherwise yields a positive value.This is based on the observation that zeros in the curvature of the first distribution result in corresponding zeros in the strain measurement. This dissimilarity measure can also be viewed as an injective zero comparison, where a binary convolution is calculated between small curvatures and small strain values. The sub-dissimilarity values ​​determined for a candidate position are then summed to determine a dissimilarity value for that specific candidate position.

[0045] The positioning unit 8 can also be adapted to apply the dissimilarity measure by applying a cross-correlation, in particular a normalized cross-correlation, to the first distribution and the second distribution for the respective candidate position. Specifically, the positioning unit 8 can be adapted to determine the mean. E D the strain values ​​of the second distribution, the mean E T the curvature values ​​of the first distribution, the standard deviation σ D of the aforementioned strain values, the standard deviation σ T to calculate the aforementioned curvature values ​​and finally the normalized cross-correlation according to the following formula: 1 n ∑ i 1 σ D σ T D i − E D T i − E T , where n is the number of curvature values ​​and thus, in this example, also the number of strain values, where the index i the individual curvature values ​​and the individual strain values ​​are indexed, whereby D ( i ) the respective elongation value and T ( i ) denotes the respective curvature value. The elongation value D ( i ) and the curvature value T ( i ) are each the iThe values ​​that overlap at the respective candidate positions are called the -th values. Since the first and second distributions are shifted differently relative to each other for different candidate positions, different first digits of the first distribution and different second digits of the second distribution overlap for different candidate positions. This also typically results in different stretch and curvature values ​​overlapping, leading to different normalized cross-correlations for different candidate positions. In this example, the dissimilarity measure is the normalized cross-correlation or depends on it.

[0046] The visualization generation unit 11 is suitable for generating a visualization of the tubular structure based on the provided image and the specific position of the instrument 4 within the tubular structure, such that the specific position of the instrument 4 is visualized in the provided image or in a further image derived from the image provided by the imaging unit, showing the desired structures within the patient. The imaging system 1 further comprises an output unit 13, such as a display for showing the visualization to a user, and an input unit 12, such as a keyboard, computer mouse, touchpad, etc.While instrument 4 is located within the tubular structure of patient 3, the user can see precisely where instrument 4 is positioned within the tubular structure, thus assisting the user in its use within the patient. For example, instrument 4 may possess detection and / or treatment capabilities, and the displayed visualization can support the user in performing detection and / or treatment procedures by showing the exact position, for example, of the tip of instrument 4 relative to the tubular structure on the output unit 13.

[0047] The following is an exemplary embodiment of an imaging method with reference to the one described in Fig. 2 The flowchart shown is described.

[0048] In step 101, an image of the tubular structure is provided by the imaging unit 10, and in step 102, the curvature values ​​indicative of tubular structure curvatures at several first points along the tubular structure are provided by the first provisioning unit 6. The first provisioning unit 6 thus provides the first distribution. These curvature values ​​can be provided before, during, or after the insertion of the instrument 4 into the tubular structure of the patient 3. After the instrument 4 has been inserted into the tubular structure of the patient 3, in step 103, strain values ​​indicative of instrument 4 strains at several second points along the instrument 4 are provided by the second provisioning unit 6. The second provisioning unit 7 thus provides the second distribution.In step 104, the position of instrument 4 within the tubular structure is determined by the position determination unit 8 based on the first and second distributions. In step 105, a visualization of the tubular structure is generated by the visualization generation unit 11 based on the provided image and the determined position. The generated visualization is displayed on the output unit 13 in step 106. Steps 102 to 104 can be considered steps of a position determination procedure for determining the position of an instrument within a tubular structure.

[0049] The positioning device described above enables endotubular navigation based on strain profiles, which are determined, for example, by means of fiber optic strain sensors that can be integrated into a medical instrument such as a catheter, a guide wire, a bronchoscope or another medical instrument.

[0050] The described positioning device and the positioning method are preferably adapted to enable real-time correlation of strain measurement with a given tubular structure. The tubular structure is preferably extracted from pre-interventional image data, and based on this extracted structure, a scalar curvature profile of the medical instrument is determined for each possible path through the tubular structure. This can be achieved, for example, by extracting the respective vessel midline or by biophysical simulation of the expected instrument position within the vessel. These curvature profiles, which can be multiple for several paths, are then provided as initial distributions for the intra-interventional mapping step.

[0051] During the intervention, the strain profile of the medical instrument, i.e., the second distribution, is determined and correlated with the pre-defined curvature profiles, i.e., the first distributions. Distance measures, i.e., dissimilarity measures, between the strain profile and the respective curvature profile are preferably chosen to account for the non-linear relationship between curvature and strain.

[0052] If a strain profile along the medical instrument is used as the second distribution, it is not necessary to determine a curvature profile based on at least two strain profiles in order to determine, for example, a three-dimensional shape of the medical instrument based on the curvature profile, which could then be mapped onto the tubular structure. Therefore, a single optical fiber or a single optical fiber core with strain sensors is sufficient to determine the position of the medical instrument. These are relatively inexpensive to manufacture and relatively easy to integrate into a medical instrument.

[0053] Fig. 3 shows an example of a three-dimensional image of a tubular structure and Fig. 4 illustrates a vessel graph 31, which is based on the one in Fig. 3 The image shown was generated and represents the tubular structure. Vessel graph 31, and thus the tubular structure, is branched and non-cyclic.

[0054] Fig. 5 Figure 32...35 illustrates four curvature profiles for four paths within the tubular structure, where the curvature K is plotted along the vertical axis and the respective location x along each path is plotted along the horizontal axis. This means that a local curvature is calculated for each location along the corresponding path based on the tubular structure. The various paths, which can also be referred to as vascular tree paths (GBPs), result from the branching within the tubular structure. The local curvature is preferably determined along all GBPs. Fig. 5 This shows several initial distributions.

[0055] Strain gauges positioned along the medical instrument are used to determine localized strain values ​​that can correlate with the curvature of the tubular structure. It should be noted that, depending on the orientation of the optical fiber relative to the path, individual local strain gauges may not be affected by the path's curvature. This is particularly true when the curvature is orthogonal to the strain gauge's axis of rotation within the medical instrument. Nevertheless, precise determination of the medical instrument's position within the tubular structure is still possible using the positioning device and method.

[0056] Preferably, the position determination unit 8 first determines the most plausible position of the medical instrument for each possible GBP. For this purpose, for each possible position of the medical instrument, i.e., for each candidate position, the curvature values ​​along the respective first distribution of the respective GBP and the measured strain values ​​in the medical instrument are compared using a distance measure, i.e., a dissimilarity measure. Various distance measures are conceivable, whereby for each possible distance measure or dissimilarity measure, characteristics are extracted from the curvature and strain profiles, i.e., from the first and second distributions, and based on these characteristics, a distance, i.e., a dissimilarity value, between the first and second distributions is determined for a specific candidate position. Thus, for each GBP through the tubular structure, a plausibility value is obtained.a dissimilarity value for each possible position, i.e., each candidate position, of the medical instrument along the respective GBP. Fig. 6 This illustrates, by way of example, the plausibility or dissimilarity value as different gray values ​​for different candidate positions p.

[0057] In one embodiment, the most plausible candidate position of the medical instrument at a given GBP can be considered the candidate position for which the lowest dissimilarity value has been determined, i.e., for which the lowest distance has been determined. However, such a position determination can be erroneous due to factors such as noise, so it is preferable to integrate knowledge about past time points into the determination of the most plausible candidate positions. Therefore, for each GBP and each time point, a distribution of dissimilarity values ​​for different candidate positions of the medical instrument relative to the respective GBP is preferably determined, resulting in a map that could also be called an energy map (EAM) and which is displayed in Fig. 7 is illustrated.

[0058] In Fig. 7 The different dissimilarity values ​​of EKP 38 are again represented by different shades of gray, with EKP 38 being spanned by different candidate positions p and different times t. That is, EKP 38 comprises corresponding dissimilarity values ​​for different candidate positions p and different times t, which are shown in Fig. 7 are shown in grayscale. The most plausible position of the medical instrument in a GBP is then preferably obtained by an optimal path through the EKP 38. Optimal here preferably means that the accumulated dissimilarity value along the path through the EKP 38 is minimal, optionally taking into account a regularity condition. The optional regularity condition could, for example, stipulate that the path through the EKP 38 should be as short as possible. It could also stipulate that the path through the EKP 38 should be as straight as possible. Technically, every path through the EKP 38 is associated with certain costs, which result from the accumulated dissimilarity values ​​and the optional regularity. The path with the lowest costs is the optimal path, which defines the position of the medical instrument for a GBP and for different times. Such an optimal path 39 is shown in Fig. 8 Shown as an example.

[0059] The position of the medical instrument within the entire tubular structure is determined from the most plausible positions per GBP. For this purpose, the optimal EKP cost values, i.e., the path values ​​of the optimal paths, of the individual GBPs are compared. In one embodiment, the GBP and its corresponding most plausible position with the smallest path value are selected. However, this approach is also subject to noise, which can lead to erroneous error determination, especially since the finally determined position of the medical instrument can jump between GBPs. To prevent this, a temporal regularization is preferably used, which penalizes jumps between GBPs. For this purpose, the temporal evolution of the optimal path value per GBP is stored, for example, as optimal cost curves (OCCs). Such OCCs 40... 42 are in Fig. 9 illustrated. Fig. 9 illustrates OKKs 40 ... 42, which have different optimal path values ​​W for different times t.

[0060] The temporal regularization is based on imposing a penalty term on all OKKs that were not minimal at a previous point in time. Fig. 9 This penalty term is symbolized, for example, by the double arrow 45. Curve 44 is based on OKK 41 after the application of penalty term 45, and curve 43 is based on OKK 42 again after the application of the penalty term. The penalty term is preferably simply an additive value that is predefined and can be determined in advance, for example, during system calibration. The threshold values ​​and absolute value thresholds mentioned above can also be determined in advance during calibration.The positioning unit is therefore preferably adapted, as described above, to select at any given time from the determined optimal paths the one for which a minimum path value has been determined, wherein, before selecting the optimal path with the minimum path value, the path value is increased for any optimal path for which no minimum path value was determined at earlier times compared to the path values ​​determined for the other optimal paths. The determined current position of the catheter within the tubular structure is given in . Fig. 8 marked with an X.

[0061] Although different dissimilarity measures have been shown separately in the embodiments above, combinations of these dissimilarity measures can also be used. For example, two or more dissimilarity measures can be combined linearly, and the resulting linear combination can then be used to determine the position.

[0062] Although scalar strain values ​​are measured along the medical instrument in the embodiments described above to determine the second distribution, other embodiments allow for the determination of multiple strain values ​​at the respective locations along the medical instrument. This means that the strain can be determined in at least two directions at each location along the medical instrument. The position can then be determined based on these multi-component strain values ​​and the first distribution of curvature values. Furthermore, in another embodiment, the second distribution can also be a distribution of curvature values, so that the position of the medical instrument within the tubular structure is determined based on both the first and second distributions of curvature values.The second distribution of curvature values ​​can also be determined based on the aforementioned two-component or multi-component strain values. If the position of the medical instrument within the tubular structure is determined based on the first and second distributions of curvature values, the curvature values ​​of both distributions are preferably scalar.

[0063] Other modifications of the disclosed embodiments can be understood and carried out by persons skilled in the art after studying the drawing, the description and the attached claims.

[0064] In the claims, the word "comprehensive" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality.

[0065] A single device or apparatus can fulfill the functions of several objects listed in the claims. The mere fact that certain measures are listed in different dependent claims does not preclude the possibility of using a combination of these measures to one's advantage.

[0066] Procedures such as determining the position of the instrument within the tubular structure, determining dissimilarity measures, determining displacement values, etc., which are performed by one or more devices or apparatuses, can be performed by any other number of devices or apparatuses. These procedures and / or the control of the imaging system according to the imaging procedure and / or the control of the positioning device according to the positioning procedure can be implemented as program code means of a computer program and / or as dedicated hardware.

[0067] A computer program may be stored / delivered on a suitable medium, such as an optical storage medium or a solid-state medium, delivered together with or as part of other hardware, but may also be delivered in other forms, such as via the Internet or other wired or wireless telecommunications systems.

[0068] All reference numerals in the claims should not be interpreted as limiting the scope of application.

[0069] The invention is defined in the attached claims.

Claims

1. A position determining device for determining the position of an elongate medical instrument within a tubular structure, said position determining device (5) comprising: - a first providing unit (6) for providing a first distribution of curvature values at a plurality of first points along a path within the tubular structure, wherein the first providing unit (6) is adapted to provide a plurality of first distributions of curvature values for a plurality of paths within the tubular structure, - a second providing unit (7) for providing a second distribution of strain values or curvature values at a plurality of second points along the instrument (4), wherein the second providing unit (7) is adapted to provide a plurality of second distributions along the instrument (4) for different points in time, wherein the plurality of second distributions correspond to a plurality of positions of the instrument (4) relative to a path within the tubular structure, - a position determining unit (8), adapted to - to determine for each path and for each point in time a distribution of dissimilarity scores for different candidate positions of the instrument (4) relative to the respective path, wherein a dissimilarity measure to the corresponding first distribution and the corresponding second distribution for different possible candidate positions of the instrument (4) relative to the path is applied, wherein the dissimilarity measure is adapted to return a dissimilarity score indicative of a dissimilarity of the two distributions in the respective candidate position for each candidate position of the instrument (4) relative to the path, wherein dissimilarity scores determined for the different points in time and the different candidate positions can be interpreted as a map, wherein a respective dissimilarity score is entered in the map for different locations specified by the respective points in time and candidate positions, - to provide a route measure which returns a route score for a route through the respective map that ends at a candidate position for a latest point in time and begins at a candidate position for an earlier point in time, said route score depending on the dissimilarity scores along the respective route, - to determine for each path a route through the respective map by means of the route measure, for which a minimum route score is determined such that a respective optimal route is calculated for different paths and thus for different maps, and - to determine the positions of the instrument (4) relative to one of the paths on the basis of optimal routes calculated for the different paths by taking into account the route scores determined for said optimal routes.

2. The position determining device according to claim 1, characterised in that the first providing unit (6) is adapted to provide a scalar distribution of curvature values along a path within the tubular structure as the first distribution.

3. The position determining device according to any one of the preceding claims, characterised in that the second providing unit (7) is adapted to determine the second distribution on the basis of optical signals from optical strain sensors arranged along the instrument (4).

4. The position determining device according to any one of the preceding claims, characterised in that the position determining unit (8) is adapted to provide a route measure which includes a component that becomes greater as the sum of the dissimilarity scores increases along the respective route.

5. The position determining device according to any one of the preceding claims, characterised in that the position determining unit (8) is adapted - to calculate optimal routes for different points in time such that a respective optimal route with a corresponding route score is determined for different points in time and different paths, wherein, in order to determine an optimal route with a corresponding route score for a path and a particular point in time, use is made of the dissimilarity scores calculated for the particular point in time and for earlier points in time and for the candidate positions, which can be interpreted as a map for the respective route and the particular point in time, - to select at each point in time, from the calculated optimal routes, the optimal route for which a minimum route score has been determined, wherein the route score is increased, before the optimal route with the minimum route score is selected, for an optimal route for which no minimum route score was determined at earlier points in time compared to the route scores determined for the other optimal routes, - to specify the positions defined by the selected optimal route as the positions of the instrument (4) relative to one of the paths.

6. The position determining device according to any one of the preceding claims, characterised in that the position determining unit (8) is adapted to apply the dissimilarity measure by - determining a spatial first gradient distribution of the first distribution, - determining a spatial second gradient distribution of the second distribution, - comparing, for each location along the instrument (4), the respective gradient of the second distribution with the respective curvature gradients of the first distribution, wherein the respective candidate position defines which respective gradient of the second distribution is compared with which respective curvature gradient of the first distribution at the respective location, wherein, for the respective location and for the respective candidate position, a sub-dissimilarity measure is applied which is dependent a) on the directions of the gradients relative to each other and / or b) on the amount of the gradient of the first distribution and the amount of the gradient of the second distribution, wherein a sub-dissimilarity score is determined for each location by applying the sub-dissimilarity measure, - summing the sub-dissimilarity scores determined for a candidate position, in order to determine a respective dissimilarity score for the respective candidate position.

7. The position determining device according to any one of the preceding claims, characterised in that the position determining unit (8) is adapted to apply the dissimilarity measure by - determining local maxima of the first distribution, - determining local maxima of the second distribution, - assigning a local maximum of the second distribution to each local maximum of the first distribution, so that the sum of all the spatial distances between assigned local maxima is minimal, wherein each local maximum of the second distribution is assigned to only one local maximum of the first distribution, - adding the spatial distances between the assigned local maxima of the first distribution and the assigned local maxima of the second distribution.

8. The position determining device according to any one of the preceding claims, characterised in that the position determining unit (8) is adapted to apply the dissimilarity measure by - comparing the respective value of the second distribution with the respective value of the first distribution for each location along the instrument (4), wherein the respective candidate position defines which respective value of the second distribution is compared with which respective value of the first distribution at the respective location, wherein, for the respective location and for the respective candidate position, a sub-dissimilarity measure is used which is zero if a) the respective value of the first distribution is above a predefined first threshold or b) the respective value of the first distribution is below the first threshold and the respective value of the second distribution is below a predefined second threshold, and which otherwise has a positive value, - summing the sub-dissimilarity scores determined for a candidate position, in order to determine a respective dissimilarity score for the respective candidate position.

9. The position determining device according to any one of the preceding claims, characterised in that the position determining unit (8) is adapted to apply the dissimilarity measure by applying a cross-correlation, in particular a normalised cross-correlation, to the first distribution and the second distribution for the respective candidate position.

10. An imaging system, comprising: - a position determining device (5) for determining the position of an elongate medical instrument (4) within a tubular structure in accordance with claim 1, - an imaging unit (10) for providing an image of the tubular structure, - a visualisation generating unit (11) for generating a visualisation of the tubular structure on the basis of the image provided and the position determined.

11. A position determining method for determining the position of an elongate medical instrument (4) within a tubular structure, said method comprising the steps of: - providing a first distribution of curvature values at a plurality of first points along a path within the tubular structure by a first providing unit (6), wherein a plurality of first distributions of curvature values for a plurality of paths within the tubular structure is provided, - providing a second distribution of strain values or curvature values at a plurality of second points along the instrument (4) by a second providing unit (7), wherein a plurality of second distributions along the instrument (4) for different points in time is provided, wherein the plurality of second distributions correspond to a plurality of positions of the instrument (4) relative to a path within the tubular structure, - determining for each path and for each point in time a distribution of dissimilarity scores for different candidate positions of the instrument (4) relative to the respective path by a position determining unit (8), wherein a dissimilarity measure to the corresponding first distribution and the corresponding second distribution for different possible candidate positions of the instrument (4) relative to the path is applied, wherein the dissimilarity measure is adapted to return a dissimilarity score indicative of a dissimilarity of the two distributions in the respective candidate position for each candidate position of the instrument (4) relative to the path, wherein dissimilarity scores determined for the different points in time and the different candidate positions can be interpreted as a map, wherein a respective dissimilarity score is entered in the map for different locations specified by the respective points in time and candidate positions, - providing a route measure which returns a route score for a route through the respective map that ends at a candidate position for a latest point in time and begins at a candidate position for an earlier point in time, said route score depending on the dissimilarity scores along the respective route by the position determining unit (8), - determining for each path a route through the respective map by means of the route measure, for which a minimum route score is determined such that a respective optimal route is calculated for different paths and thus for different maps by the position determining unit (8), and - determining the positions of the instrument (4) relative to one of the paths on the basis of optimal routes calculated for the different paths by taking into account the route scores determined for said optimal routes by the position determining unit (8).

12. An imaging method, comprising the steps of: - providing an image of a tubular structure by an imaging unit (10), - determining the position of an elongate medical instrument (4) within the tubular structure according to the position determining method according to claim 11 by a position determining device according to claim 1, and - a visualisation generating unit generating a visualisation of the tubular structure on the basis of the image provided and the position determined.

13. A computer program for determining the position of an elongate medical instrument, wherein the computer program includes program code means adapted to cause the position determining device according to claim 1 to carry out the position determining method for determining the position of an elongate medical instrument (4) within a tubular structure according to claim 11 when it is executed on the position determining device (1).

14. The computer program according to claim 13, characterized in that the program code means are further adapted to cause the imaging system according to claim 10 comprising the position determining device to carry out, when the computer program is executed on the imaging system, following steps: - providing an image of the tubular structure - generating a visualisation of the tubular structure on the basis of the image provided and the position determined.