Blood vessel registration using functional information

Abstract

A method and apparatus for analyzing diagnostic image data is provided, wherein correspondence detection between a first diagnostic image and a second diagnostic image of a vessel of interest in a patient's vasculature is performed based on at least one functional parameter by matching one or more values ​​of the functional parameter at specific locations along the vessel of interest shown in the first diagnostic image and the second diagnostic image with each other, thereby determining a correlation between locations based on the functional parameter values ​​rather than only on vessel geometry.

Description

[Technical field]

[0001] The present invention relates to a method for analyzing diagnostic image data, in some embodiments multi-modality diagnostic image data, a corresponding apparatus, and respective computer programs. In particular, the present invention relates to an improved method and apparatus that allows for more robust and accurate correspondence detection to vessels of interest in a patient's vasculature, as represented in multiple diagnostic images, by using functional parameter values. [Background technology]

[0002] Today, coronary artery disease is a widespread condition in many societies. In order to establish the appropriate treatment options for each individual patient, it is of great importance to obtain as much information as possible about the coronary vasculature. This makes an accurate assessment of the geometrical characteristics of the vasculature and the hemodynamic characteristics through the vasculature with sufficient reliability particularly relevant.

[0003] To that end, a large number of different medical measurement modalities are known to obtain the respective medical data related to these characteristics, among which are medical imaging modalities such as computed tomography (CT) angiography, invasive angiography, intravascular ultrasound (IVUS) imaging, magnetic resonance (MR) imaging, intracardiac echo (ICE), transesophageal echocardiography (TEE), fluid dynamics measurement modalities such as invasive functional flow reserve (FFR) or instantaneous wave free ratio (IFR) measurements, ultrasonic blood flow measurements such as Doppler blood flow measurements.

[0004] Each medical imaging modality makes it possible to collect specific diagnostic image data that reveals specific information about the vasculature. This specific information may vary from one imaging modality to another, and even from one diagnostic image acquired using the same imaging modality, since acquisition settings such as projection direction, cardiac motion, etc., may alter the visualization of the vasculature in different diagnostic images.

[0005] Therefore, in order to perform an accurate and reliable assessment of the vasculature and therefore any potential (coronary) artery disease, it is beneficial not only to use a single diagnostic image, but to use a combination of multiple diagnostic images to derive a more complete picture of the vasculature.

[0006] Therefore, methods have been developed in which different diagnostic images are correlated with each other to derive more information about the coronary arteries. Applications that enable such correlation are, for example, co-registration of vascular roadmaps, image fusion for different imaging modalities, or motion compensated vascular modeling.

[0007] Thus, in order to properly combine the different diagnostic images, a so-called correspondence detection has to be performed, during which the same vessel of interest or a segment thereof is identified in all diagnostic images and thus co-registered to all diagnostic images, i.e. it is determined which vessel (segment) in a first diagnostic image corresponds to which vessel (segment) in a second diagnostic image.

[0008] However, accurate correspondence detection is often hindered by factors such as changed projection direction, cardiac motion, and usually uncalibrated C-arm geometry. As an example, in two or more two-dimensional angiograms acquired from different projection directions, the same vessel (segment) may appear to have different characteristics such that the correspondence between the vessel (segment) represented in the first diagnostic image and the same vessel segment represented in the second diagnostic image can be established only with low accuracy. This problem becomes even more severe when the diagnostic images are acquired using different imaging modalities. As an example, the coregistration of CT angiograms with invasive angiography data is mentioned. Such registration of CT angiograms with angiography data can be very challenging, since cardiac motion and / or unknown relative geometry can cause ambiguities.

[0009] Therefore, various feature detection modes have been developed that are believed to allow for accurate determination of these correspondences. These feature detection modes are typically based on a combination of many features such as branch locations, characteristic curves, geometric constraints, etc. Using these features often yields the most robust correspondences.

[0010] Despite this, an approach that allows establishing a particularly accurate and reliable correspondence between the same vessel (segment) represented in different diagnostic images has not yet been developed. Summary of the Invention [Problem to be solved by the invention]

[0011] It is therefore an object of the present invention to provide a method and apparatus that allows correspondence detection for one or more vessel segments of a vasculature in a reliable, accurate and efficient manner.

[0012] More particularly, it is an object of the present invention to provide a method and apparatus for analyzing diagnostic image data that allows combining multiple diagnostic images by performing reliable associations between the diagnostic images used individually.

[0013] A further object is to provide a method and apparatus that allows to perform multi-modality image visualization and / or multi-modality modeling of a patient's vascular system, i.e., one or more vessels or vessel segments therein, in a reliable and accurate manner. [Means for solving the problem]

[0014] This object is achieved by a method for analyzing diagnostic image data, comprising the steps of acquiring a first diagnostic image and a second diagnostic image of a blood vessel of interest in a patient's vascular system, defining at least one determined position for the blood vessel of interest in the first diagnostic image and at least one corresponding determined position for the blood vessel of interest in the second diagnostic image, determining a first value for at least one functional parameter at the at least one determined position for the first diagnostic image and a second value for the same at least one functional parameter at the corresponding determined position for the second diagnostic image, and associating one or more blood vessel positions along the blood vessel of interest represented in the first diagnostic image with one or more blood vessel positions along the blood vessel of interest represented in the second diagnostic image based on a comparison of the first value of the at least one functional parameter and the second value of the at least one functional parameter.

[0015] That is, the two diagnostic images are co-registered with one another based on functional parameters determined for at least one particular determined location in both diagnostic images.

[0016] In that context, the term diagnostic image may specifically refer to an image representing the vasculature of a patient. Here, the term vasculature may refer to a vascular tree or a single vessel. The term "vasculature" may specifically refer to a vessel segment of a vessel of interest. In some embodiments, the diagnostic image may represent a vasculature including one or more vessels of interest of the coronary vasculature.

[0017] The first and second diagnostic images may in particular be acquired by a diagnostic imaging modality, which may in particular be gated. A gated diagnostic imaging modality may typically use a gated reconstruction, where acquisition of diagnostic images is performed in parallel with acquisition of data providing information across the cardiac cycle, such as electrocardiogram (ECG) or photoplethysmography (PPG) data, whereby this data can be used to gate image acquisition and reconstruction with respectively selected phase points of the cardiac cycle.

[0018] The term "determined location" may specifically refer to a particular location along the vessel of interest as represented in the first and second diagnostic images, at which a value of a functional parameter may be determined, for example, by actual invasive measurements such as intravascular pullback measurements and / or non-invasive measurements such as Doppler flow measurements, and / or by simulating the fluid dynamics through the vessel of interest using a respective fluid dynamics model. Multiple determined locations may be used, at each of which a corresponding value of one or more functional parameters may be determined.

[0019] The term functional parameters may in particular refer to parameters indicative of the dynamic behavior of the vessel of interest, in particular the fluid dynamics through the vessel of interest, i.e. functional parameters may refer to parameters such as pressure, flow, vascular resistance, etc. that vary along the vessel of interest.

[0020] The term vessel position may specifically refer to a position along the vessel of interest. A plurality of vessel positions may be determined along the length of the vessel of interest. The vessel positions in the first diagnostic image and the second diagnostic image are correlated with each other by comparing first and second values ​​of at least one functional parameter obtained for the determined positions in the first diagnostic image and the determined positions in the second diagnostic image.

[0021] That is, if a plurality of first values ​​are obtained for a particular functional parameter in a first diagnostic image and a plurality of second values ​​are obtained for the same functional parameter in a second diagnostic image, these values ​​may be compared with each other, where it is assumed that the determined positions whose values ​​correspond to each other are related to each other, i.e. correspond to the same vessel position represented in the two images.

[0022] Thus, according to the method, at least one determination position is determined in both diagnostic images. After this, a value of a functional parameter is determined for the at least one determination position. The thus determined values, respectively determined for the first diagnostic image and the second diagnostic image, are then compared with each other in order to establish an association. As an example, if a value for one particular determination position in the first diagnostic image corresponds to a value for one particular determination position in the second diagnostic image, it is assumed that these determination positions show the same value for the particular functional parameter and therefore correspond to the same vessel position in the vessel of interest. This makes it possible to perform a correspondence detection between the first diagnostic image and the second diagnostic image even if the diagnostic images were acquired with different acquisition settings and / or different imaging modalities.

[0023] In some embodiments, acquiring the first diagnostic image and the second diagnostic image includes acquiring the first diagnostic image by a first diagnostic imaging modality using a first acquisition setting, and acquiring the second diagnostic image by the first diagnostic imaging modality using a second acquisition setting that is different from the first acquisition setting.

[0024] In some embodiments, the first diagnostic image and the second diagnostic image may be acquired with different acquisition settings using one particular diagnostic or medical imaging modality such as CT angiography, invasive angiography, IVUS imaging, MR imaging, ICE, TEE, etc. In one particular embodiment, the first diagnostic image and the second diagnostic image may be acquired using two-dimensional X-ray angiography, each having two different acquisition settings.

[0025] In that context, the term acquisition settings may refer to settings of acquisition parameters, such as radiation time, projection direction, acquisition timing with respect to the cardiac cycle, among others, which may be adjusted for each patient or for each image. Thus, a change in the acquisition settings corresponds to a change in one or more acquisition parameters for performing the respective acquisition. Such a change typically results in a change in the representation of the vessel of interest in the diagnostic image.

[0026] Therefore, acquiring two or more diagnostic images and subsequently using different acquisition settings to correlate these images makes it possible to gather more information about the vessel of interest or a segment thereof.

[0027] In some embodiments, acquiring the first diagnostic image and the second diagnostic image comprises acquiring the first diagnostic image with a first diagnostic imaging modality and acquiring the second diagnostic image with a second diagnostic imaging modality different from the first diagnostic imaging modality, hi some embodiments, the first diagnostic imaging modality corresponds to a non-invasive diagnostic imaging modality and the second diagnostic imaging modality corresponds to an invasive diagnostic imaging modality.

[0028] In some embodiments, the first and second diagnostic images may also be acquired using different imaging modalities. As an example, the first diagnostic image may be acquired using a non-invasive imaging modality such as CT angiography, X-ray angiography, MR imaging, US imaging, etc., and the second diagnostic image may be acquired using an invasive imaging modality such as invasive angiography, IVUS imaging, ICE or TEE imaging. In other embodiments, the first and second diagnostic images may both be acquired using an invasive imaging modality, or the first and second diagnostic images may both be acquired using a non-invasive imaging modality. It should further be understood that the respective acquisition settings may depend, among other things, on the particular imaging modality used.

[0029] The use of different imaging modalities allows to gather more information than would be available from the use of a single imaging modality, since each imaging modality allows to visualize different aspects of the vessel of interest. By using functional parameter values ​​for the association between the first and second diagnostic images, the corresponding detection of the vessel of interest in the first and second diagnostic images, respectively, is still possible, regardless of the different imaging methods used to acquire them. This is the case when the functional parameters are related to the dynamics of the vessel of interest and not (mainly) to the geometric properties, which may be represented differently when using different imaging modalities. In particular, the combination of non-invasive and invasive techniques allows to acquire a wide variety of information about the vessel of interest. This improves the accuracy of the vessel evaluation and thus the diagnosis of any possible diseases.

[0030] In some embodiments, determining the first and / or second value of the at least one functional parameter comprises obtaining intravascular measurement data for the vessel of interest and deriving the first and second values ​​of the at least one functional parameter at the at least one determined location based on the intravascular measurement data. In some embodiments, determining the first and / or second value of the at least one functional parameter comprises generating a fluid dynamics model for the vessel of interest and deriving the first and second values ​​of the at least one functional parameter at the at least one determined location based on the fluid dynamics model.

[0031] In some embodiments, the functional parameter value may in particular be derived from an invasive measurement of said functional parameter. For that purpose, a catheter or the like fitted with a respective measuring device may be introduced into the patient's vascular system, in particular into the blood vessel of interest. Inside the blood vessel of interest, the measuring device may be used to obtain intravascular measurement data, such as values ​​for one or more of pressure, flow, vascular resistance, etc. at dedicated positions along the length of the blood vessel of interest.

[0032] In some embodiments, a first value of the at least one functional parameter may be determined using first intravascular measurement data acquired by a first measurement modality, and a second value of the at least one functional parameter may be determined using second intravascular measurement data acquired by a second measurement modality.

[0033] Furthermore, in some particular embodiments, the measurement data collected for the vessel of interest in the first diagnostic image and the vessel of interest in the second diagnostic image may correspond to pullback data, such as pressure pullback data. The pullback data may then comprise a plurality of first functional parameter values, one for each measurement location in the vessel of interest, i.e., each location at which a pullback measurement was performed, as shown in the first diagnostic image. The pullback data may further comprise a plurality of second functional parameter values, one for each measurement location in the vessel of interest, as shown in the second diagnostic image.

[0034] In some embodiments, functional parameter values ​​may be derived from each fluid flow simulation using a fluid dynamics model representing blood flow through a vessel of interest and / or through multiple vessels in a vasculature, where the term fluid dynamics model may specifically refer to a model of blood flow through a vessel of interest.

[0035] The fluid dynamics model is generated by simulating the interaction of blood with the vessel walls. Since blood is a fluid and the vessel walls may be considered as the respective surfaces with which the fluid interacts, the interaction of blood with the vessel walls may be most accurately defined by respective boundary conditions that take into account the properties of the vessel walls and the blood interacting therewith. These properties may include vessel wall composition, vessel wall elasticity and vascular impedance, branches within the vessel, outflow through these branches, blood viscosity, vessel outlet resistance at specific locations along the length of the vessel, etc. The fluid dynamics model may be integrated with a two-dimensional or three-dimensional geometric model, thereby representing the fluid dynamics of blood at each location within one or more blood vessels of interest depicted in the geometric model.

[0036] In some embodiments, one, e.g., a first, functional parameter value may correspond to a measured value, whereas the other, e.g., a second, functional parameter value may correspond to a simulated value simulated using a fluid dynamics model. As an example, the first functional parameter value may be obtained using a pressure, in particular a pullback measurement, such as an FFR or iFR pullback measurement, or an ultrasonic flow measurement, such as a Doppler flow measurement, and the second functional parameter may be obtained by means of simulating the corresponding functional parameter using a fluid dynamics model.

[0037] In that context, in some exemplary embodiments, the first diagnostic image may correspond to a CT angiography image and the second diagnostic image may correspond to an invasive angiography image, whereby a first functional parameter value for the first diagnostic CT angiography image corresponds to a simulated iFR value derived from the fluid dynamics model and a second functional parameter value for the second diagnostic invasive angiography image corresponds to a measured FFR or iFR pullback value.

[0038] In some exemplary embodiments, the first diagnostic image may correspond to an IVUS image and the second diagnostic image may correspond to a CT or invasive angiography image, and correspondence detection is performed based on first and second functional parameter values, such as pressure or flow distribution, simulated using a fluid dynamics model for both the first diagnostic image and the second diagnostic image.

[0039] In some exemplary embodiments, two CT or invasive angiography images may be co-registered, thereby performing correspondence detection based on the simulated functional parameter values. In some exemplary embodiments, a first diagnostic image may correspond to a neuro-MR image and a second diagnostic image may correspond to an ultrasound (US) image, whereby a first functional parameter value for the first diagnostic image may be obtained using MR-based flow measurements and a second functional parameter value for the second diagnostic image may be obtained using Doppler flow measurements.

[0040] In other exemplary embodiments, the first diagnostic image may correspond to a CT image provided with a simulated flow distribution as a first functional parameter value and registered to an intravascular US image provided with Doppler flow measurements as a second functional parameter value.

[0041] In some exemplary embodiments, the first diagnostic image may correspond to a CT image and the second diagnostic image may be an ICE image or a TEE image, and correspondence detection may be performed by matching a first functional parameter value from the Doppler flow measurement to a second functional parameter value derived from the simulated flow distribution. Other combinations are possible as well.

[0042] In some embodiments, the method further includes acquiring one or more acquisition parameters used to acquire the first diagnostic image and / or the second diagnostic image, respectively, and associating one or more vessel positions along the vessel of interest as represented in the first and second diagnostic images, respectively, based on the one or more acquisition parameters.

[0043] In some embodiments, acquisition parameters used to acquire the first diagnostic image and the second diagnostic image may be used as further information for performing correspondence detection.

[0044] Acquisition parameters that may be used as such further information may in particular relate to the projection geometry, cardiac phase, and specific image features of the first and second diagnostic images, respectively, such that one or more of these acquisition parameters may be combined to remove ambiguities and / or inaccuracies that may result from uncertainties that may exist when performing correspondence detection.

[0045] As an example, by knowing the projection geometry, the orientation of the vessel of interest in the diagnostic image may be determined. Thus, since the projection geometry is known, the inherent error of incorrectly identifying the vessel of interest in the first diagnostic image and / or the second diagnostic image when performing a correspondence detection between the two can be reduced. Thus, for a particular projection geometry (or other factors that affect the vessel of interest and its representation in the diagnostic image in a similar manner), it is possible to indicate how the vessel of interest should be represented in the diagnostic image, thereby allowing the correspondence detection to be performed accurately.

[0046] As discussed above, acquisition parameters may be used to remove uncertainty regarding vessel identification. As a specific example, an FFR or iFR pullback measurement may result in the acquisition of multiple pressure values ​​that can be matched to either a simulated pressure value for the left anterior descending artery (LAD) or a simulated pressure value for the left circumflex artery (LCx). Knowledge of the projection geometry, and in particular the projection angle of one or more invasive angiographic images recorded along with the FFR or iFR pullback data, may resolve whether the data should be matched to the LAD or LCx.

[0047] In some embodiments, the acquisition parameters may be used as further information to provide an initial starting point for correspondence detection. As a specific example, knowledge of the projection geometry of one or more invasive angiography images and of the reconstruction geometry of one or more computed tomography (CT) angiography images may be useful in performing an initial coregistration of image data from the invasive angiography and the CT angiography, respectively. This initial coregistration may then be used as a starting point for further correspondence detection, which in this case takes into account (coronary) motion and potential miscalibration, for example by matching pressure values ​​from FFR or iFR pullback measurements to the respective simulated pressure values.

[0048] In some embodiments, further additional information, e.g., geometric information about the vessel of interest itself, may also help to provide a more accurate correspondence detection. In particular, the further geometric information may allow a vessel location to be uniquely identified in cases where multiple vessel locations can be matched to one particular functional parameter value. As an example of this, a pressure value obtained by invasive measurement for one particular vessel location, e.g., as visible in an invasive angiography image, may be matched as a first value to two or more simulated pressure values ​​at two or more vessel locations in the respective CT angiography images, second values. That is, the vessel location in the CT angiography image that corresponds to the vessel location in the invasive angiography image is ambiguous. In that case, considering that the vessel size is derivable from both the invasive angiography image and the CT angiography image at a particular vessel location allows to determine the vessel location where the vessel size seen in the invasive angiography image best matches the vessel size in the CT angiography image. This best match then allows for unique identification of a particular vessel location in the CT angiogram image that should be matched to one vessel location in the invasive angiogram image.

[0049] In some embodiments, associating one or more vascular positions along the vessel of interest as represented in a first diagnostic image with one or more vascular positions along the vessel of interest as represented in a second diagnostic image includes determining a point-to-point correspondence for the one or more vascular positions along the vessel of interest as represented in the first and second diagnostic images, respectively.

[0050] In some embodiments, the correspondence detection is performed by creating a point-to-point correspondence between the vasculature as shown in the first and second diagnostic images by matching first and second values ​​of corresponding functional information with each other. The point-to-point correspondence may then be used for multi-modality vascular modeling or multi-modality image visualization. In some embodiments, the point-to-point correspondence may in particular be achieved by determining the outflow, in particular the virtual outflow, and successively matching the peaks of said (virtual) outflow derived for the first diagnostic image and the second diagnostic image, respectively, with each other. In some embodiments, the point-to-point correspondence determination is further performed based on additional information regarding acquisition parameters, such as the projection geometry, as described herein above.

[0051] In some embodiments, the first diagnostic image and / or the second diagnostic image may have been acquired using one or more of computed tomography (CT) angiography and / or invasive angiography and / or intravascular ultrasound (IVUS) imaging and / or magnetic resonance (MR) imaging and / or intracardiac echocardiography (ICE) and / or transesophageal echocardiography (TEE).

[0052] A variety of different (medical) imaging modalities can be envisaged for development purposes. Furthermore, various functional parameter measurements may be obtained by IVUS, TEE, ICE etc.

[0053] In some embodiments, the method further comprises training a classifier using a training data set indicating an association between the vessel of interest in the first diagnostic image and the second diagnostic image to obtain a classification result, and comparing a first value of at least one functional parameter and a second value of at least one functional parameter based on the classification result.

[0054] In some embodiments, the matching of the derived functional parameter values ​​for the first and second diagnostic images, respectively, may be performed using a machine learning algorithm, which may be implemented as a specifically trained classifier or in terms of a neural network. The machine learning algorithm, and in particular the classifier, may be trained on a training data set, which in particular indicates the association of the vessels of interest and the functional parameter values. In some embodiments, the classifier uses the training results to perform the matching of the functional parameter values ​​between the first and second diagnostic images, respectively, thereby generating a point-to-point correspondence.

[0055] In some embodiments, the method further includes a step of segmenting the vessel of interest as represented in the first diagnostic image and the second diagnostic image, respectively, and generating a physiological model of the vessel of interest based on the segmentation.

[0056] In some embodiments, the first and second diagnostic images may be used to generate a physiological model of the vessel of interest. That is, once correspondence detection has been performed, the information collected by using different acquisition settings and / or different imaging modalities may be used to generate a more complete physiological model of the vessel of interest. For this purpose, the vessel of interest as represented in the first and second diagnostic images may be segmented into a number of segments. Based on this segmentation, a geometric model of the vessel of interest may be generated, i.e. a physiological model of the vessel of interest having a geometric representation of the vessel of interest and / or the entire vasculature. This geometric model may typically be a two-dimensional or three-dimensional model, depending on the information that may be derived from the first and second diagnostic images. That is, if the first and / or second diagnostic images allow to derive information in three dimensions, the geometric model may be a three-dimensional model, and if the first and / or second diagnostic images allow to derive information in two dimensions, the geometric model may be a two-dimensional model.

[0057] In other embodiments where only two-dimensional information is available, the geometric model may be a quasi-three-dimensional model with the third dimension interpolated from the other two dimensions. According to a further aspect, an apparatus for analyzing diagnostic image data is provided, the apparatus comprising: an input unit configured to acquire a first diagnostic image and a second diagnostic image of a vessel of interest in a patient's vasculature, a definition unit configured to define at least one determined position for the vessel of interest in the first diagnostic image and at least one corresponding determined position for the vessel of interest in the second diagnostic image, a determination unit configured to determine a first value for at least one functional parameter in the at least one determined position for the first diagnostic image and a second value for the same at least one functional parameter in the corresponding determined position in the second diagnostic image, and an association unit configured to associate one or more vessel positions along the vessel of interest represented in the first diagnostic image with one or more vessel positions along the vessel of interest represented in the second diagnostic image based on a comparison of the first value of the at least one functional parameter and the second value of the at least one functional parameter. In some embodiments, the apparatus further comprises a modelling unit configured to generate a physiological model of the vessel of interest based on the first diagnostic image and the second diagnostic image.

[0058] In a further aspect, a computer program for controlling an apparatus according to the invention is provided, which computer program, when executed by a processing unit, is arranged to perform the method steps according to the invention. In yet another aspect, a computer readable medium having stored thereon a computer program as cited above is provided.

[0059] It is to be understood that the method of claim 1, the apparatus of claim 12, the computer program of claim 14 and the computer readable medium of claim 15 have similar and / or identical preferred embodiments, in particular as defined in the dependent claims.

[0060] It is to be understood that a preferred embodiment of the invention can be any combination of the dependent claims or the above embodiments with the respective independent claim.

[0061] These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter. [Brief description of the drawings]

[0062] [Figure 1] 1 illustrates a schematic diagram of an apparatus for analyzing diagnostic image data according to one embodiment; [Diagram 2] 2A and 2B illustrate schematic diagrams of a correlation between a first diagnostic image and a second diagnostic image according to one embodiment; [Diagram 3] 2 shows a variant of the device according to FIG. [Figure 4] 1 shows a flowchart of a method for analyzing diagnostic image data according to one embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0063] The figures in the drawings are schematic: in different drawings, similar or identical elements are provided with the same reference numbers.

[0064] 1 shows a schematic diagram of an exemplary embodiment of an apparatus 1 for analysing diagnostic image data. The apparatus 1 comprises an input unit 100, a definition unit 200, a determination unit 300, a correlation unit 400, a modelling unit 500 and a display unit 600.

[0065] The input unit 100 is configured to receive a first diagnostic image 10 and a second diagnostic image 20. In an exemplary embodiment, the first diagnostic image 10 and the second diagnostic image 20 correspond to two-dimensional images acquired using the same imaging modality, namely X-ray angiography, using different acquisition settings. In particular, the first diagnostic image 10 has been acquired in a different projection direction than the second diagnostic image 20. It should be understood that, in the particular embodiment according to FIG. 1, the device 1 is configured to receive two diagnostic images acquired by the same imaging modality, but the device 1 is also configured to process the first diagnostic image 10 acquired using the first imaging modality and the second diagnostic image 20 acquired using the second imaging modality.

[0066] In the exemplary embodiment of Fig. 1, the input unit 100 is further configured to receive data 30 indicative of respective acquisition settings used to acquire the first diagnostic image 10 and the second diagnostic image 20. In the exemplary embodiment according to Fig. 1, it is the input unit 100 that is further configured to retrieve acquisition parameters from the data 30. In other embodiments, the definition unit 200, the determination unit 300 and / or the correlation unit 400 may alternatively or additionally be configured to retrieve said acquisition parameters.

[0067] The input unit 100 is further configured to provide the first diagnostic image 10, the second diagnostic image 20 and the derived acquisition parameters to the definition unit 200. The definition unit 200 is configured to receive the first diagnostic image 10 and the second diagnostic image 20 and to define at least one determination position for each vessel of interest as shown in the first diagnostic image 10 and the second diagnostic image 20. In a particular embodiment according to FIG. 1, the definition unit 200 defines three determination positions along a longitudinal axis of the vessel of interest in the first diagnostic image 10 and three corresponding determination positions along a longitudinal axis of the vessel of interest in the second diagnostic image 20. It should be understood that in other embodiments the definition unit 200 may define more or less determination positions for the vessel of interest. The definition unit 200 then provides data indicative of the defined determination positions together with the first diagnostic image 10 and the second diagnostic image 20 to the determination unit 300.

[0068] The determination unit 300 is configured to receive the first diagnostic image 10 and the second diagnostic image 20 as well as data indicative of the defined determination position from the definition unit 200. Furthermore, the determination unit 300 is configured to receive intravascular measurement data 40 which, in the particular embodiment according to Fig. 1, corresponds to pressure pullback data acquired by slowly pulling back the pressure wire through the blood vessel of interest.

[0069] The determining unit 300 is further configured to determine a respective value for the at least one functional parameter at each determined location based on the intravascular measurement data 40. In the example embodiment according to Fig. 1 the determining unit 300 is configured to obtain a first pressure value for the first determined location, a first pressure value for the second determined location and a first pressure value for the third determined location from the first diagnostic image 10. Further the determining unit 300 is configured to obtain a second pressure value for the first corresponding determined location, a second pressure value for the second corresponding determined location and a second pressure value for the third determined location from the second diagnostic image 20.

[0070] The determination unit 300 then provides to the correlation unit 400 data indicating the first and second pressure values ​​respectively obtained from the first diagnostic image 10 and the second diagnostic image 20 at the three respective determination positions, together with the first diagnostic image 10, the second diagnostic image 20 and, optionally, acquisition parameters.

[0071] The correlation unit 400 is configured to receive data indicative of the first and second pressure values, the first diagnostic image 10 and the second diagnostic image 20, and optionally acquisition parameters. The correlation unit 400 then compares the first and second pressure values ​​with each other in order to determine a correlation between a first plurality of vessel positions along a longitudinal axis of the vessel of interest represented in the first diagnostic image 10 and a second plurality of vessel positions along a longitudinal axis of the same vessel of interest represented in the second diagnostic image 20. In a particular implementation according to FIG. 1, the correlation unit 400 evaluates in particular a pressure distribution derived from the first pressure value and the second pressure value and matches the pressure distribution determined for the first pressure value to the pressure distribution determined for the second pressure value. By doing so, the correlation unit 400 may establish a point-to-point correspondence between the vessel of interest shown in the first diagnostic image 10 and the vessel of interest shown in the second diagnostic image 20. That is, for each vessel position of a vessel of interest in the first diagnostic image 10, a corresponding vessel position of the vessel of interest in the second diagnostic image 20 is determined. In the particular embodiment according to Fig. 1, the derivation of the point-to-point correspondence is further supported by information regarding the acquisition parameters.

[0072] Upon registering the vessel of interest in the first diagnostic image 10 with the vessel of interest in the second diagnostic image 20, the registered images are provided to a modeling unit 500. The modeling unit 500 segments the vessel of interest in the first diagnostic image 10 and in the second diagnostic image 20 and generates a physiological model of the vessel of interest. In a particular embodiment according to Fig. 1, the physiological model comprises in particular a geometric model that allows to derive the geometric configuration of the vessel of interest. In some embodiments, the physiological model may also comprise a fluid dynamics model.

[0073] The modelling unit 500 is further configured to, upon generating the model, provide the generated model, optionally together with the first diagnostic image 10 and the second diagnostic image 20, to a display unit 600 having a screen 601 and a user interface 602.

[0074] The display unit 600 is configured to generate a graphical representation of the physiological model and present this graphical representation to a user, optionally together with one or both of the first diagnostic image 10 and the second diagnostic image 20. In some embodiments, the user may interact with the model via a user interface 602.

[0075] Fig. 2 illustrates an approach for relating vessels of interest depicted in a first diagnostic image 10 and a second diagnostic image 20. In a particular embodiment according to Fig. 2, the first diagnostic image 10 and the second diagnostic image 20 correspond to coronary angiograms acquired using different projection orientations.

[0076] In an exemplary embodiment according to Fig. 2, a fluid dynamics model is generated from the first diagnostic image 10 and the second diagnostic image 20. The fluid dynamics model is then used to simulate functional parameters for at least the determined positions 11, 12, 13, 21, 22, and 23 along the vessel of interest. In a particular embodiment according to Fig. 2, the fluid dynamics model is used to derive a virtual outflow through the vessel of interest at the determined positions 11, 12, and 13 for the diagnostic image 10 and at the determined positions 21, 22, and 23 for the diagnostic image 20. The correlation unit 400 then performs matching of peaks of the virtual outflow through the vessel of interest, thereby providing a point-to-point reference between at least two images.

[0077] Figure 3 shows diagrammatically an apparatus 1' which is a variant of the apparatus 1 according to figure 1. Similar components of the apparatus 1' compared to the apparatus 1 are referred to by the same reference numerals, i.e. the apparatus 1' also comprises an input unit 100, a definition unit 200, a decision unit 300, a modelling unit 500 and a display unit 600. The apparatus 1' further comprises a correlation unit 400' which comprises a classifier 401.

[0078] To avoid repetition, in the following only the differences between the device 1 and the device 1′ are described in detail, i.e. the input unit 100, the definition unit 200 and the decision unit 300, as well as the modeling unit 500 and the display unit 600 are configured as described in connection with FIG.

[0079] In the particular embodiment according to Fig. 3, the input unit 100 receives a first diagnostic image 10' acquired by a first imaging modality, which in the particular case of Fig. 2 corresponds to X-ray angiography. Furthermore, the input unit 100 receives a second diagnostic image 20' acquired using a second imaging modality, which in this particular case corresponds to magnetic resonance imaging. It should be understood that although the device 1' is configured to receive (and process) the first diagnostic image 10' and the second diagnostic image 20' acquired with two different imaging modalities, the device 1' can similarly receive and process the first diagnostic image 10 and the second diagnostic image 20 acquired with one particular imaging modality.

[0080] The first diagnostic image 10' and the second diagnostic image 20' are processed as described in connection with Fig. 1. The difference between the device 1 according to Fig. 1 and the device 1' according to Fig. 3 resides in the correlation unit 400'. In contrast to the embodiment according to Fig. 1, the correlation unit 400' according to Fig. 2 comprises a classifier 401. The classifier 401 is trained by a training data set 50, which is typically provided to the classifier 401 before the association and comprises a correspondence between at least one functional parameter along the vessel of interest represented in the first diagnostic image 10 and the second diagnostic image 20.

[0081] In a particular embodiment according to Fig. 3, the classification device receives the first diagnostic image 10 and the second diagnostic image 20 and classifies the two diagnostic images to establish a point-to-point correspondence. Once the first diagnostic image 10 is registered to the second diagnostic image 20, the registered diagnostic images are provided to the modeling unit 500 and proceed as described in relation to Fig. 1.

[0082] 4 shows diagrammatically a method 1000 for analysing diagnostic image data executed by the apparatus 1. In step S101, the input unit 100 receives the first diagnostic image 10 and the second diagnostic image 20. In step S102, the input unit 100 further receives data 30 indicative of acquisition settings used to acquire the first diagnostic image 10 and the second diagnostic image 20. In step S103, the input unit 100 derives relevant acquisition parameters from the data 30 and provides the first diagnostic image 10 and the second diagnostic image 20 together with the acquisition parameters derived from the data 30 to the definition unit 200.

[0083] In step S201, the definition unit 200 receives the first diagnostic image 10 and the second diagnostic image 20 together with the acquisition parameters. In step S202, the definition unit 200 defines at least one determination position for the respective vessel of interest shown in the first diagnostic image 10 and the second diagnostic image 20. In the particular embodiment according to Fig. 4, the definition unit 200 again defines three determination positions along the vessel of interest in the first diagnostic image 10 and three corresponding determination positions along the vessel of interest in the second diagnostic image 20. The definition unit 200 then provides the first diagnostic image 10 and the second diagnostic image 20 and data indicative of the defined determination positions to the determination unit 300.

[0084] In step S301, the determination unit 300 receives from the determination unit 300 the first diagnostic image 10, the second diagnostic image 20 and data indicative of the determination positions and further receives intravascular measurement data 40 indicative of at least one functional parameter. In step S302, the determination unit 300 derives respective values ​​for the at least one functional parameter at each determination position based on the intravascular measurement data 40. In the exemplary embodiment according to Fig. 4, the determination unit 300 in particular derives a first pressure value for the first determination position, a first pressure value for the second determination position and a first pressure value for the third determination position from the first diagnostic image 10 and derives a second pressure value for the first corresponding determination position, a second pressure value for the second corresponding determination position and a second pressure value for the third corresponding determination position from the second diagnostic image 20.

[0085] In step S303, the determination unit 300 provides the first diagnostic image 10, the second diagnostic image 20 and, optionally, the acquisition parameters indicated by the data 30 to the correlation unit 400, together with data indicating first and second pressure values ​​obtained from the first diagnostic image 10 and the second diagnostic image 20 at three corresponding determination positions.

[0086] In step S401, the correlation unit 400 receives the first diagnostic image 10 and the second diagnostic image 20 together with data indicative of the first and second pressure values ​​and optionally acquisition parameters. In step S402, the correlation unit 400 compares the first and second pressure values ​​with each other and in step S403 determines a correlation between a first plurality of vessel positions along the vessel of interest represented in the first diagnostic image 10 and a second plurality of vessel positions of the same vessel of interest represented in the second diagnostic image 20, thus performing a vessel registration for the vessel of interest. The correlation unit 400 then provides the registered plurality of images to the modeling unit 500.

[0087] In step S501, the modeling unit 500 receives the registered images and in step S502 segments the vessel of interest in the first diagnostic image 10 and the second diagnostic image 20. In step S503, the modeling unit 500 generates a physiological model of the vessel of interest and provides the generated model, optionally together with the first diagnostic image 10 and the second diagnostic image 20, to the display unit 600.

[0088] Next, in step S601, the display unit 600 generates a graphical representation of the physiological model, optionally including the first diagnostic image 10 and / or the second diagnostic image 20. Next, in step S602, the display unit 600 displays the graphical representation to a user, on which the user may interact with the displayed information.

[0089] In the above embodiment, the first diagnostic image and the second diagnostic image are acquired using X-ray angiography with different acquisition settings, however, it should be understood that the first diagnostic image and the second diagnostic image may be acquired using different imaging modalities such as invasive angiography, neuro-MR imaging, US imaging, CT imaging, ICE imaging, TEE imaging, etc.

[0090] In the above embodiment the functional parameter value corresponds to a pressure value, however in other embodiments further functional (mechanical) parameter values ​​such as flow rate, vascular resistance etc. may be used.

[0091] Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims.

[0092] In the claims, the word "comprise" does not exclude other elements or steps and the indefinite article "a" or "an" does not exclude a plurality.

[0093] A single unit or device may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

[0094] The procedures such as defining at least one determination position, determining the first and second values ​​of at least one functional parameter, correlating one or more vascular positions, generating a fluid dynamics model and deriving the first and second values ​​of the at least one functional parameter based on the fluid dynamics model, generating a physiological model, etc., performed by one or more units or devices, can be performed by any other number of units or devices. These procedures, in particular the control of the devices for analyzing the diagnostic image data according to the control methods performed by the respective device controllers, can be implemented as program code means of a computer program and / or as dedicated hardware.

[0095] The computer program may be stored / distributed on a suitable medium, such as an optical storage medium or a solid-state medium, supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunications systems.

[0096] Any reference signs in the claims should not be construed as limiting the scope.

[0097] The present invention relates to a method of analyzing diagnostic image data comprising the steps of acquiring a first diagnostic image and a second diagnostic image of a blood vessel of interest in a patient's vascular system, defining at least one determined position for the blood vessel of interest in the first diagnostic image and at least one corresponding determined position for the blood vessel of interest in the second diagnostic image, determining a first value for at least one functional parameter at the at least one determined position for the first diagnostic image and a second value for the same at least one functional parameter at the corresponding determined position for the second diagnostic image, and associating one or more blood vessel positions along the blood vessel of interest represented in the first diagnostic image with one or more blood vessel positions along the blood vessel of interest represented in the second diagnostic image based on a comparison of the first value of the at least one functional parameter and the second value of the at least one functional parameter.

Claims

1. A method for operating an apparatus for analyzing diagnostic image data having an input unit, a definition unit, a decision unit and a correlation unit, comprising: said input unit acquiring a first diagnostic image and a second diagnostic image of a vessel of interest in a patient's vascular system; - the definition unit defining at least one determined position for the vessel of interest in the first diagnostic image and at least one determined position for the vessel of interest in the second diagnostic image; said determining unit determining a first value of at least one functional parameter at said at least one determination location for said first diagnostic image and a second value of the same at least one functional parameter at said at least one determination location for said second diagnostic image; the correlation unit associating one or more vessel positions along the vessel of interest represented in the first diagnostic image with one or more vessel positions along the vessel of interest represented in the second diagnostic image based on a comparison of the first value of the at least one functional parameter and the second value of the at least one functional parameter; The method comprising:

2. The acquisition of the first diagnostic image and the second diagnostic image includes: acquiring the first diagnostic image with a first diagnostic imaging modality using a first acquisition setting; acquiring the second diagnostic image with the first diagnostic imaging modality using a second acquisition setting different from the first acquisition setting; 2. The method of claim 1, comprising:

3. The acquisition of the first diagnostic image and the second diagnostic image includes: acquiring a first diagnostic image with a first diagnostic imaging modality; acquiring a second diagnostic image with a second diagnostic imaging modality different from the first diagnostic imaging modality; 2. The method of claim 1, comprising:

4. The method of claim 3 , wherein the first diagnostic imaging modality corresponds to a non-invasive diagnostic imaging modality and the second diagnostic imaging modality corresponds to an invasive diagnostic imaging modality.

5. Determining the first value and / or the second value of the at least one functional parameter comprises: acquiring intravascular measurement data of the vessel of interest; deriving the first and second values ​​of the at least one functional parameter at the at least one determined location based on the intravascular measurement data; 2. The method of claim 1, comprising:

6. Determining the first value and / or the second value of the at least one functional parameter comprises: generating a fluid dynamics model for the vessel of interest; deriving the first and second values ​​of the at least one functional parameter at the at least one determined location based on the fluid dynamics model; 2. The method of claim 1, comprising:

7. The method comprises: said input unit acquiring one or more acquisition parameters used for acquiring said first diagnostic image and / or said second diagnostic image, respectively; having the associating step is based on the one or more acquisition parameters in addition to the comparison. The method of claim 1 further comprising:

8. 2. The method of claim 1, wherein associating the one or more vascular positions along the blood vessel of interest represented in the first diagnostic image with one or more vascular positions along the blood vessel of interest represented in the second diagnostic image comprises determining a point-to-point correspondence for the one or more vascular positions along the blood vessel of interest represented in the first and second diagnostic images, respectively.

9. 2. The method of claim 1, wherein the first diagnostic image and / or the second diagnostic image are acquired using one or more of computed tomography angiography and / or invasive angiography and / or intravascular ultrasound imaging and / or magnetic resonance diagnostic imaging and / or intracardiac echocardiography and / or transesophageal echocardiography.

10. The correlation unit trains a machine learning algorithm using a training data set having a correspondence between at least one functional parameter along a blood vessel of interest represented in the first diagnostic image and the second diagnostic image; the correlation unit using the trained machine learning algorithm to determine, from the first value of the at least one functional parameter and the second value of the at least one functional parameter, a point-to-point correspondence to the one or more vessel locations along the vessel of interest represented in the first diagnostic image and the second diagnostic image; The method of claim 1 further comprising:

11. The apparatus according to claim 1, further comprising: a modeling unit; said modeling unit segmenting said vessel of interest represented in said first diagnostic image and said second diagnostic image, respectively; said modeling unit generating a physiological model of said vessel of interest based on said segmentation; The method of claim 1 further comprising:

12. an input unit configured to acquire a first diagnostic image and a second diagnostic image of a vessel of interest in the patient's vascular system; a definition unit configured to define at least one determined position for the vessel of interest in the first diagnostic image and at least one determined position for the vessel of interest in the second diagnostic image; a determination unit configured to determine a first value of at least one functional parameter at the at least one determination location for the first diagnostic image and a second value of the same at least one functional parameter at the at least one determination location for the second diagnostic image; a correlation unit configured to associate one or more vessel positions along the vessel of interest represented in the first diagnostic image with one or more vessel positions along the vessel of interest represented in the second diagnostic image based on a comparison of the first value of the at least one functional parameter and the second value of the at least one functional parameter; 16. An apparatus for analyzing diagnostic image data, comprising:

13. a modeling unit configured to generate a physiological model of the blood vessel of interest based on the first diagnostic image and the second diagnostic image; The apparatus of claim 12 further comprising:

14. A computer program arranged to, when executed by a processing unit, cause an apparatus according to any one of claims 12 or 13 to carry out the method according to any one of claims 1 to 11.

15. A computer readable medium storing a computer program according to claim 14.

Images