Calculation of fractional flow reserve

Abstract

To provide a method for vascular assessment.SOLUTION: The method, in some embodiments, comprises receiving a plurality of 2-D angiographic images of a portion of a vasculature of a subject, and processing the images to produce a stenotic model over the vasculature, the stenotic model having measurements of the vasculature at one or more locations along vessels of the vasculature. The method, in some embodiments, further comprises obtaining a flow characteristic of the stenotic model, and calculating an index indicative of vascular function, based, at least in part, on the flow characteristic in the stenotic model.SELECTED DRAWING: Figure 9

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of priority to U.S. Provisional Patent Application No. 61 / 752,526, filed January 15, 2013, U.S. Provisional Patent Application No. 14 / 040,688, filed September 29, 2013, and International Patent Application No. PCT / IL2013 / 050869, filed October 24, 2013, the contents of which are incorporated herein by reference in their entireties.

[0002] This application is one of three concurrently filed applications, attorney docket numbers 58285, 58286, and 58287. [Background technology]

[0003] In some of its embodiments, the present invention relates to vascular modeling, and more particularly, but not exclusively, to the use of vascular models to generate indices related to vascular function and diagnosis in real time, for example, during a catheterized imaging procedure.

[0004] Arterial stenosis is one of the most serious forms of arterial disease. In clinical practice, the severity of stenosis is estimated either by using simple geometric parameters, such as determining the percentage of diameter of the stenosis, or by measuring hemodynamic-based parameters, such as the pressure-based fractional flow reserve (FFR). FFR is an invasive measurement of the functional severity of coronary artery stenosis. The FFR measurement technique involves the insertion of a 0.014” guidewire equipped with a miniature pressure transducer that is placed across the arterial stenosis. It represents the ratio of maximum blood flow in the area of ​​stenosis to the maximum blood flow in the same area without stenosis. From previous studies, FFR<0.75 has been found to be an accurate predictor of ischemia, and delayed percutaneous coronary intervention for lesions with FFR≥0.75 appears to be safe.

[0005] An FFR cutoff value of 0.8 is typically used in clinical practice to guide revascularization, which is supported by long-term outcome data. Typically, FFR values ​​in the range of 0.75 to 0.8 are considered a "gray zone" where clinical significance is uncertain.

[0006] Modeling of blood flow and assessment of blood flow are described, for example, in U.S. Published Patent Application No. 2012 / 0059246 to Taylor, "Method And System For Patient-Specific Modeling Of Blood Flow," which describes an embodiment including a system for determining cardiovascular information for a patient. The system may include at least one computer system configured to receive patient-specific data related to a geometry of at least a portion of the patient's anatomy. The portion of the anatomy may include at least a portion of the patient's aorta and at least a portion of a plurality of coronary arteries exiting the portion of the aorta. The at least one computer system may also be configured to create a three-dimensional model representing the portion of the anatomy based on the patient-specific data, create a physics-based model related to blood flow characteristics within the portion of the anatomy, and determine fractional flow reserve within the portion of the anatomy based on the three-dimensional model and the physics-based model.

[0007] Additional background art includes the following:

[0008] Taylor, U.S. Published Patent Application No. 2012 / 053918; U.S. Published Patent Application No. 2012 / 0072190 to Sharma et al. Taylor, U.S. Published Patent Application No. 2012 / 0053921; U.S. Published Patent Application No. 2010 / 0220917 to Steinberg et al. U.S. Published Patent Application No. 2010 / 0160764 to Steinberg et al. U.S. Published Patent Application No. 2012 / 0072190 to Sharma et al. U.S. Published Patent Application No. 2012 / 0230565 to Steinberg et al. U.S. Published Patent Application No. 2012 / 0150048 to Kang et al. U.S. Published Patent Application No. 2013 / 0226003 to Edic et al. U.S. Published Patent Application No. 2013 / 0060133 to Kassab et al. U.S. Published Patent Application No. 2013 / 0324842 to Mittal et al. U.S. Published Patent Application No. 2012 / 0177275 to Suri and Jasjit; U.S. Patent No. 6,236,878 to Taylor et al. Taylor's U.S. Patent No. 8,311,750; U.S. Patent No. 7,657,299 to Hizenga et al. Bullitt et al., U.S. Pat. No. 8,090,164; U.S. Patent No. 8,554,490 to Tang et al. U.S. Patent No. 7,738,626 to Weese et al. Hart et al., U.S. Pat. No. 8,548,778; Jerry T. Wong and Sabee Molloi, "Determination of fractional flow reserve (FFR) based on scaling laws: a simulation study", Phys. Med. Biol. 53 (2008) 3995-4011, Weickert, "A Scheme for Coherence-Enhancing Diffusion Filtering with Optimized Rotation Invariance," Journal of Visual Communication and Image Representation, Vol. 13, No. 1-2, March 2002, pp. 103-118 (2002); Editorial in the book "Anisotropic Diffusion in Image Processing" by J. Weickert, BG Teubner (Stuttgart), 1998; Paper by AF Frangi, WJ Niessen, KL Vincken, MA Viergever, titled "Multiscale vessel enhancement filtering", Medical Image Computing and Computer-Assisted Intervention-MICCA'98, Jerry T Wong and Sabee Molloi, entitled "Determination of fractional flow reserve (FFR) based on scaling laws: a simulation study", Phys. Med. Biol. 53 (2008) 3995-4011, S. Molloi, J.T. Wong, D.A. Chalyan, and H. Le, "Quantification of Fractional Flow Reserve Using Angiographic Image Data", O. Doessel and WC Schlegel (Eds.): WC 2009, IFMBE Proceedings 25 / II, pp. 901-904, 2009. Jerry T. Wong, Huy Le, William M. Suh, David A. Chalyan, Toufan Mehraien, Morton J. Kern, Ghassan S. Kassab, and Sabee Molloi, entitled "Quantification of fractional flow reserve based on angiographic image data," Int J Cardiovasc Imaging (2012) 28:13-22. The paper by Shigeho Takarada, Zhang Zhang, and Sabee Molloi, titled "An angiographic technique for coronary fractional flow reserve measurement: in vivo validation," was published online on August 31, 2012 in Int J Cardiovasc Imaging. A.M. Seifalian, D.J. Hawkes, A.C. Colchester, and K.E. Hobbs, entitled "A new algorithm for deriving pulsatile blood flow waveforms tested using stimulated dynamic angiographic data," Neuroradiology, Vol. 31, pp. 263-269, 1989. A.M. Seifalian, D.J. Hawkes, C.R. Hardingham, A.C. Colchester, and J.F. Reidy, entitled "Validation of a quantitative radiographic technique to estimate pulsatile blood flow waveforms using digital subtraction angiographic data," J. Biomed. Eng., Vol. 13, No. 3, pp. 225-233, May 1991; DJ Hawkes, AM Seifalian, AC Colchester, N. Iqbal, CR Hardingham, CF Bladin, and KE Hobbs, entitled "Validation of volume blood flow measurements using three dimensional distance-concentration functions derived from digital X-ray angiograms", Invest. Radiol, Vol. 29, No. 4, pp. 434-442, April 1994. A.M. Seifalian, D.J. Hawkes, C. Bladin, A.C.F. Colchester, and K.E.F. Hobbs, entitled "Blood flow measurements using 3D distance-concentration functions derived from digital X-ray angiograms," Cardiovascular Imaging, J.H.C. Reiber and E.E. van der Wall, Eds. Norwell, MA, The Netherlands: Kluwer Academic, 1996, pp. 425-442. KR Hoffmann, K. Doi, and LE Fencil, entitled "Determination of instantaneous and average blood flow rates from digital angiograms of vessel phantoms using distance-density curves", Invest. Radiol, Vol. 26, No. 3, p. 207212, March 1991; SD Shpilfoygel, R. Jahan, RA Close, GR Duckwiler, and DJ Valentino, entitled "Comparison of methods for instantaneous angiographic blood flow measurement," Med. Phys., Vol. 26, No. 6, pp. 862-871, June 1999. DW Holdsworth, M. Drangova, and A. Fenster, entitled "Quantitative angiographic blood flow measurement using pulsed intra-arterial injection," Med. Phys., Vol. 26, No. 10, pp. 2168-2175, October 1999. Paper by Joan C. Tuinenburg, Gerhard Koning, Andrei Rares, Johannes P. Janssen, Alexandra J. Lansky, and Johan HC Reiber, entitled "Dedicated bifurcation analysis: basic principles", Int J Cardiovasc Imaging (2011) 27:167-174, Papers by Salvatore Davide Tomasello, Luca Costanzo, and Alfredo Ruggero Galassi, entitled ``Quantitative Coronary Angiography in the Interventional Cardiology,'' Advances in the Diagnosis of Coronary Atherosclerosis, Papers by Johannes P. Janssen, Andrei Rares, Joan C. Tuinenburg, Gerhard Koning, Alexandra J. Lansky, and Johan HC Reiber, titled ``New approaches for the assessment of vessel sizes in quantitative (cardio-)vascular X-ray analysis”, Int J Cardiovasc Imaging (2010) 26:259–271, Kirkeeide R L., ed., Reiber JHC and Serruys PW, entitled "Coronary obstructions, morphology and physiologic significance of quantitative coronary arteriography", The Netherlands: Kluwer, 1991, pp. 229-244. Kevin Sprague, Maria Drangova, Glen Lehmann, Piotr Slomka, David Levin, Benjamin Chow, and Robert deKemp, entitled "Coronary x-ray angiographic reconstruction and image orientation", Med Phys, March 2006, 33(3):707-718. Paper by Adamantios Andriotis, Ali Zifan, Manolis Gavaises, Panos Liatsis, Ioannis Pantos, Andreas Theodorakakos, Efstathios P. Efstathopoulos, and Demosthenes Katritsis, entitled "A New Method of Three-dimensional Coronary Artery Reconstruction From X-Ray Angiography: Validation Against a Virtual Phantom and Multislice Computed Tomography", Catheter Cardiovasc Interv, January 1, 2008, 71(1):28-43. An article by Kenji Fusejima, MD, entitled "Noninvasive Measurement of Coronary Artery Blood Flow Using Combined Two-Dimensional and Doppler Echocardiography," JACC, Vol. 10, No. 5, November 1987, pp. 1024-31. Carlo Caiati, Cristiana Montaldo, Norma Zedda, Alessandro Bina, and Sabino Iliceto, entitled "New Noninvasive Method for Coronary Flow Reserve Assessment: Contrast-Enhanced Transthoracic Second Harmonic Echo Doppler", Circulation, the American Heart Association, 1999, 99:771-778. Harald Lethena, Hans P Triesa, Stefan Kerstinga, and Heinz Lambertza, entitled "Validation of noninvasive assessment of coronary flow velocity reserve in the right coronary artery-A comparison of transthoracic echocardiographic results with intracoronary Doppler flow wire measurements", European Heart Journal (2003) 24, pp. 1567-1575. Paper by Paolo Vocia, Francesco Pizzutoa, and Francesco Romeob, entitled "Coronary flow: a new asset for the echo lab?", European Heart Journal (2004) 25, pp. 1867-1879, Abstract of the paper by Siogkas et al., entitled “Quantification of the effect of Percutaneous Coronary Angioplasty on a stenosed Right Coronary Artery,” Information Technology and Applications in Biomedicine (ITAB), 2010 10th IEEE International Conference on, Review article by Patrick Meimoun and Christophe Tribouilloy, entitled "Non-invasive assessment of coronary flow and coronary flow reserve by transthoracic Doppler echocardiography: a magic tool for the real world", European Journal of Echocardiography (2008) 9, 449-457. Paper by Carlo Caiati, Norma Zedda, Mauro Cadeddu, Lijun Chen, Cristiana Montaldo, Sabino Iliceto, Mario Erminio Lepera, and Stefano Favale, titled "Detection, location, and severity assessment of left anterior descending coronary artery stenoses by means of contrast-enhanced transthoracic harmonic echo Doppler", European Heart Journal (2009) 30, pp. 1797-1806, Bullitt et al., "Determining malignancy of brain tumors by analysis of vessel shape", Medical Image Computing and Computer-Assisted Intervention-MICCAI 2004.

[0009] The disclosures of all documents cited above and throughout the specification, as well as all documents referenced therein, are hereby incorporated by reference. Summary of the Invention

[0010] According to aspects of some embodiments of the invention, there is provided a method for performing vascular assessment comprising receiving a first vascular model of a cardiovascular system, determining at least one characteristic based on the first vascular model representative of flow through a stenosed segment of the vascular system, generating a second vascular model comprising elements corresponding to the first vascular model and at least one modification including a difference in the at least one characteristic of the flow, and calculating a flow index comparing the first model and the second model.

[0011] According to some embodiments of the invention, the difference in at least one characteristic of the flow comprises a difference between at least one characteristic of the flow through the constricted segment and a characteristic of the flow in a corresponding segment of the second model.

[0012] According to some embodiments of the present invention, a vascular model is calculated based on a plurality of 2-D angiographic images.

[0013] According to some embodiments of the invention, the angiographic images have sufficient resolution to allow determination of vessel width within 10% for vessel segments following at least the third bifurcation from the main human coronary arteries.

[0014] According to some embodiments of the invention, the flow index comprises a prediction of the increase in flow achievable by an intervention to relieve the stenosis from the stenotic segment.

[0015] According to some embodiments of the invention, the comparative flow index is calculated based on a ratio of corresponding flow characteristics of the first and second vascular models.

[0016] According to some embodiments of the present invention, the comparative flow index is calculated based on the ratio of corresponding flow characteristics of stenosed and non-stenosed segments.

[0017] According to some embodiments of the invention, the method comprises reporting the comparative flow index as a single number per stenosis.

[0018] According to some embodiments of the present invention, the at least one characteristic of the flow comprises a flow rate.

[0019] According to some embodiments of the invention, the comparative flow index comprises an index representative of a flow reserve index comprising a ratio of maximum flow through a stenosed blood vessel to maximum flow through a stenosed blood vessel with the stenosis removed.

[0020] According to some embodiments of the invention, the comparative flow index is used in determining a recommendation for revascularization.

[0021] According to some embodiments of the invention, the comparative flow index comprises a value indicative of the capacity to restore flow by removing a stenosis.

[0022] According to some embodiments of the present invention, the first and second vascular models comprise connected branches of vessel segment data, each branch being associated with a corresponding vascular resistance to flow.

[0023] According to some embodiments of the present invention, the vascular model does not include a radially detailed 3-D description of the vessel wall.

[0024] According to some embodiments of the present invention, the second vascular model is a normal model, comprising a vessel having a relatively large diameter that replaces the stenotic vessel in the first vascular model.

[0025] According to some embodiments of the invention, the second vascular model is a normal model comprising a normalized vessel obtained by normalizing the stenosed vessel based on properties of adjacent non-stenosed vessels.

[0026] According to some embodiments of the invention, the at least one characteristic of the flow is calculated based on characteristics of a plurality of vessel segments that are flow-connected with the stenotic segment.

[0027] According to some embodiments of the present invention, the flow characteristic comprises a resistance to fluid flow.

[0028] According to some embodiments of the invention, the method comprises identifying in a first vascular model a stenotic vessel and a crown of a branch vessel downstream of the stenotic vessel, and calculating a resistance to fluid flow in the crown, wherein a flow index is calculated based on a volume of the crown and based on a contribution of the stenotic vessel to the resistance to fluid flow.

[0029] According to some embodiments of the invention, the first vascular model comprises a representation of a vascular position in three-dimensional space.

[0030] According to some embodiments of the invention, each vascular model corresponds to a portion of the vascular system that lies between two successive branches of the vascular system.

[0031] According to some embodiments of the present invention, each vascular model corresponds to a portion of the vascular system that includes a branch of the vascular system.

[0032] According to some embodiments of the invention, each vascular model corresponds to a portion of the vascular system that extends at least one branch of the vascular system beyond a stenotic segment.

[0033] According to some embodiments of the invention, each vascular model corresponds to a portion of the vascular system that extends at least three branches of the vascular system beyond the stenotic segment.

[0034] According to some embodiments of the present invention, the vascular model comprises paths along vascular segments, each of which is mapped along its extent to locations in the multiple 2-D images.

[0035] According to some embodiments of the invention, the method comprises acquiring an image of the cardiovascular system and constructing a first vascular model thereof.

[0036] According to some embodiments of the invention, each vessel model corresponds to a portion of the vasculature that extends as far distally as the image resolution allows determination of vessel width to within 10% of the correct value.

[0037] According to some embodiments of the invention, the vascular model is of an artificially expanded vasculature at the time of acquisition of the images used to generate the model.

[0038] In accordance with an aspect of some embodiments of the present invention, a computer software product is provided comprising a computer readable medium having program instructions stored thereon that, when read by a computer, cause the computer to receive a plurality of 2-D images of a subject's vasculature and perform a method for vascular assessment.

[0039] According to an aspect of some embodiments of the present invention, there is provided a method for performing vascular assessment comprising: A system is provided that includes a computer configured to receive a plurality of 2-D images, convert the plurality of 2-D images into a first vascular model of the vasculature, determine at least one characteristic based on the first vascular model representative of flow through a stenosed segment of the vasculature, generate a second vascular model having elements corresponding to the first vascular model and at least one modification including changing at least one characteristic of the flow through the stenosed segment to a characteristic of the flow as if through a corresponding segment in which the effect of the stenosis is reduced, and calculate a flow index that compares the first model and the second model.

[0040] According to some embodiments of the invention, the computer is configured to calculate the flow index within 5 minutes of receiving the first vascular model.

[0041] According to some embodiments of the invention, the computer is configured to calculate the flow index within 5 minutes of acquisition of the 2-D image.

[0042] According to some embodiments of the present invention, the computer is located remotely from the imaging device.

[0043] According to an aspect of some embodiments of the invention, there is provided a method for vascular assessment comprising receiving a vascular model of a cardiovascular system, determining at least a first flow characteristic based on the vascular model representative of flow through a stenosed segment of the vascular system and a coronary vessel to the stenosed segment, determining at least a second flow characteristic based on the vascular model representative of flow through the coronary vessel without flow restriction by the stenosed segment, and calculating a flow index comparing the first flow characteristic to the second flow characteristic.

[0044] According to an aspect of some embodiments of the present invention, there is provided a method for constructing a vascular tree model comprising receiving a plurality of 2-D angiographic images of a vessel segment included in a portion of a subject's vasculature; automatically extracting from each of the plurality of 2-D angiographic images a corresponding image feature set comprising 2-D feature locations of the vessel segment; automatically adjusting the 2-D feature locations to reduce relative position error in a common 3-D coordinate system into which each of the feature sets can be backprojected; automatically associating the 2-D feature locations across the image feature sets such that image features projected from a common vessel segment region are associated; and automatically determining a representation of the image features based on inspection of 3-D projections determined from the associated 2-D feature locations and selecting an optimal available 3-D projection therefrom.

[0045] According to some embodiments of the present invention, the extracted image feature set comprises a centerline dataset including 2-D centerline positions ordered along a vessel segment.

[0046] According to some embodiments of the invention, the determined representation is a 3-D spatial representation of the extent of the vessel segment.

[0047] According to some embodiments of the invention, the determined representation is a graphical representation of the extent of the vessel segment.

[0048] According to some embodiments of the present invention, the information necessary to automatically correlate 2-D image locations is provided in its entirety prior to review of the images by a human operator.

[0049] According to some embodiments of the present invention, the adjusting, correlating and determining are performed with elements of the centerline data set.

[0050] According to some embodiments of the invention, the adjusting comprises registering the 2-D images in 3-D space with parameters that cause the 2-D centerline positions to have a closer correspondence between their 3-D backprojections.

[0051] According to some embodiments of the present invention, the extracted set of image features comprises a landmark data set including at least one of the group consisting of the origin of the tree model, the location of locally reduced radii within a stenosed vessel segment, and bifurcations spanning a vessel segment.

[0052] According to some embodiments of the present invention, the extracted image feature set comprises a landmark data set that includes pixel intensity configurations below a predefined threshold of self-similarity over the transformation.

[0053] According to some embodiments of the present invention, the adjusting is performed on elements of the landmark dataset and the relating and determining is performed between elements of the centerline dataset.

[0054] According to some embodiments of the invention, the adjusting comprises registering the 2-D image in 3-D space with parameters that cause features of the landmark dataset to have a closer correspondence between their 3-D backprojections.

[0055] According to some embodiments of the invention, registration of the 2-D image comprises registration of the positions of elements of the centerline data set.

[0056] According to some embodiments of the invention, the method comprises estimating a radial vessel width metric based on values ​​of at least one of the plurality of 2-D angiographic images along a straight line perpendicular to the ordered 2-D centerline positions.

[0057] According to some embodiments of the invention, estimating a metric of radial vessel width comprises finding connected paths running along either side of the 2-D centerline location, the connected paths comprising pixels that are images of the boundary region of the vessel wall.

[0058] According to some embodiments of the present invention, the boundary region of the vessel wall is determined by analysis of the intensity gradient along the vertical line.

[0059] According to some embodiments of the present invention, a metric of radial vessel width is calculated as a function of centerline position.

[0060] According to some embodiments of the invention, the determining comprises adjusting the 2-D feature positions based on a projection of the 3-D representation into at least one 2-D plane of the plurality of 2-D angiographic images.

[0061] According to some embodiments of the invention, the adjusting comprises calculating 3-D representations of feature locations from the 2-D feature locations of a first subset of the plurality of 2-D angiographic images, adjusting the 2-D feature locations in a second subset of the plurality of 2-D angiographic images to better match the features of the 3-D representation as if the first 3-D representation were projected into an adjusted image plane of the second subset, and repeating the calculating and adjusting with changes to the first and second subsets until a stopping condition is met.

[0062] According to some embodiments of the present invention, a stopping condition is that no position aligns with the 2-D feature position above a distance threshold.

[0063] According to some embodiments of the invention, the method comprises defining a surface corresponding to the shape of the subject's heart and using the surface as a constraint for feature location association.

[0064] According to some embodiments of the invention, the images are acquired after injection of a contrast agent into the vasculature, and the method further comprises determining temporal characteristics of the movement of the contrast agent through the vasculature, and constraining the feature locations based on the temporal characteristics.

[0065] According to some embodiments of the invention, the portion of the vascular system comprises a coronary artery.

[0066] According to some embodiments of the present invention, capturing the multiple 2D angiographic images is performed by multiple imaging devices to capture the multiple 2D angiographic images.

[0067] According to some embodiments of the present invention, capturing a plurality of 2D angiographic images comprises synchronizing a plurality of imaging devices to capture the plurality of images at substantially the same phase within the cardiac cycle.

[0068] According to an aspect of some embodiments of the present invention, a computer software product is provided comprising a computer readable medium having program instructions stored thereon that, when read by a computer, cause the computer to receive a plurality of 2D angiographic images of a portion of a vascular system and perform a method for constructing a vascular tree model.

[0069] According to an aspect of some embodiments of the present invention, a system is provided for performing vascular assessment, the system comprising: a computer logically connected to an angiographic imaging device for capturing a plurality of 2-D images of a portion of a subject's vasculature, the computer configured to receive the plurality of 2-D angiographic images from the plurality of angiographic imaging devices, extract from each of the plurality of 2-D angiographic images an image feature data set comprising 2-D feature locations of vascular segments, adjust the 2-D feature locations to minimize a relative position error in a common 3-D coordinate system for the feature locations, find a correspondence of the 2-D feature locations between the image feature data sets such that 2-D feature locations projected from a common vascular segment region onto different images are associated, and determine a 3-D representation of the 2-D feature locations based on inspection of the 3-D projections determined from the associated 2-D feature locations.

[0070] According to some embodiments of the invention, the set of image features that the system is configured to extract comprises a centerline dataset that includes 2-D centerline locations ordered along a vessel segment.

[0071] According to some embodiments of the invention, the system is configured to use the positions of the elements of the centerline data set as the 2-D feature positions.

[0072] According to some embodiments of the invention, the system is configured to adjust the 2-D feature positions based on registration of the 2-D images in 3-D space with parameters that cause the 2-D centerline positions to have a closer correspondence between their 3-D backprojections.

[0073] According to some embodiments of the invention, the measurement of radial vessel width comprises the distance between connected paths running along either side of the 2-D centerline location, the connected paths comprising pixels that are an image of the boundary region of the vessel wall.

[0074] According to some embodiments of the present invention, the image transformation based adjustment may be performed iteratively for at least a second selection of images for the first and second sets of images.

[0075] According to some embodiments of the invention, the portion of the vascular system comprises the coronary artery tree from a main coronary artery to at least a third branching point.

[0076] According to one aspect of some embodiments of the present invention, there is provided a method comprising constructing a vascular tree model, receiving 2-D images of the vascular tree, each of the images being associated with a corresponding image plane location, automatically identifying vascular features in the 2-D images, identifying homologous vascular features between the images by geometrically projecting rays from the vascular features in the image plane locations through a common image target space, and associating features having intersecting rays as being homologous.

[0077] According to some embodiments of the present invention, the intersection of the rays comprises falling within a predetermined distance from each other.

[0078] According to some embodiments of the present invention, the image plane position is iteratively updated to reduce errors in ray intersections, and the identification of homologous vessel features is then repeated.

[0079] According to an aspect of some embodiments of the present invention, there is provided a method comprising constructing a vascular tree model, iteratively backprojecting rays from features in a plurality of 2-D images onto a common 3-D plane, determining an error in the intersection of rays from features that are common among the plurality of 2-D images, and repeating the aligning, backprojecting, determining, and aligning the 2-D images at least a first additional time.

[0080] According to one aspect of some embodiments of the invention, a model of a portion of a vasculature is provided, where elements of the model are associated with a number of geometry descriptions selected from the group consisting of a coordinate space of a number of 2-D angiographic images, a coordinate space of a common 3-D space, and a vascular graph space having 1-D extents branching off from connecting nodes.

[0081] According to an aspect of some embodiments of the present invention, there is provided a method for performing a vascular assessment, comprising receiving a plurality of 2-D angiographic images of a portion of a subject's vasculature, generating a first 3-D vascular tree model for the portion of the vasculature comprising a stenosed cardiac artery by automated processing of the images within 20 minutes of receipt, and automatically determining an index based on the vascular tree model that quantifies the capacity to restore flow by opening the stenosis.

[0082] According to some embodiments of the invention, the indication of the volume to restore flow by opening the stenosis comprises a calculation based on a change in vessel width.

[0083] According to some embodiments of the present invention, the automated process is performed in less than 10 quadrillion calculations.

[0084] According to some embodiments of the invention, the automated processing comprises the creation of a model that does not include a radially detailed 3-D representation of the vessel wall.

[0085] According to some embodiments of the present invention, the automatically determining and the automatic processing comprises forming a model that does not include dynamic flow modeling.

[0086] According to some embodiments of the invention, the automatically determining comprises linear modeling of flow characteristics of the blood vessel.

[0087] According to some embodiments of the present invention, the vascular tree model represents vascular width as a function of vascular extent.

[0088] According to some embodiments of the present invention, the vascular extent comprises a distance along a vascular segment that is located at a node location on the vascular tree model.

[0089] According to some embodiments of the present invention, the first 3-D vascular tree model comprises at least three branch nodes between the vascular segments.

[0090] According to some embodiments of the present invention, the first 3-D vascular tree model comprises vascular centerlines and vascular widths along them.

[0091] According to some embodiments of the present invention, a first 3-D vascular tree is generated in less than 5 minutes.

[0092] According to some embodiments of the invention, the method comprises calculating a characteristic FFR for at least one vascular segment of a vascular tree.

[0093] According to some embodiments of the invention, calculating the FFR characteristic comprises generating a second vascular tree model based on the first model, the difference being that vascular widths are represented as being larger in the second model, and comparing the first vascular tree model with the second vascular tree model.

[0094] According to some embodiments of the invention, the comparing comprises obtaining a ratio of flow rates modeled in the first vascular tree model and the second vascular tree model for the at least one vascular segment.

[0095] According to some embodiments of the invention, the FFR characteristics are calculated within one minute of generating the first 3-D vascular tree model. According to some embodiments of the invention, the FFR characteristics are calculated within 10 seconds of generating the first 3-D vascular tree model and the second 3-D vascular tree model. According to some embodiments of the invention, the FFR characteristic is a predictor of the manometry determined FFR index with at least 95% sensitivity.

[0096] According to some embodiments of the invention, the method comprises generating a projection of a portion of a first 3-D vascular tree into a 2-D coordinate reference frame shared by at least one of the plurality of 2-D angiographic images.

[0097] According to some embodiments of the present invention, at least one image is transformed from its original coordinate reference frame to a coordinate reference frame that is defined relative to the 3-D coordinate reference frame of the 3-D vascular tree.

[0098] According to some embodiments of the invention, a subject undergoes intravascular catheterization during imaging to generate a plurality of received 2-D angiographic images, and the subject remains catheterized during receipt of the images and during generation of a first 3-D vascular tree model.

[0099] According to some embodiments of the invention, the method includes imaging the subject to generate a second plurality of 2-D angiographic images after the first generation of the first vascular tree model, the second receiving of images comprising the second plurality of images, and the second generation of the first 3-D vascular tree model, wherein the subject remains with the catheter inserted in the blood vessel.

[0100] According to some embodiments of the invention, generation occurs interactively with the subject's ongoing catheter insertion procedure.

[0101] According to some embodiments of the present invention, the calculation of the FFR characteristics is performed interactively with the subject's ongoing catheter insertion procedure.

[0102] According to an aspect of some embodiments of the present invention, a system is provided that includes a computer for performing vascular assessment, logically connected to an angiography imaging device for capturing multiple 2-D images of a portion of a subject's vasculature, and configured to calculate a vascular tree model therefrom within 5 minutes, wherein an index of vascular function indicative of capacity to restore flow by opening a stenosis can be determined based on the vascular tree model within another minute.

[0103] According to some embodiments of the invention, determining based on the vascular tree model comprises generating a second vascular tree model derived from the vascular tree model by widening the modeled vessel width in the region of the stenosis.

[0104] In some embodiments of the invention, one or more models of the patient's vascular system are generated.

[0105] In some embodiments, the first model is generated from actual data collected from an image of the patient's vasculature. Optionally, the actual data includes a portion of the vasculature including at least one blood vessel with a stenosis. In these embodiments, the first model describes a portion of the vasculature including at least one blood vessel with a stenosis. This model is interchangeably referred to as a stenosis model. Optionally, the actual data includes a portion of the vasculature including at least one blood vessel with a stenosis and a crown. In these embodiments, the stenosis model further includes information related to the shape and / or volume of the crown and information related to the blood flow and / or resistance to blood flow within the crown.

[0106] In some embodiments, the first model is used to calculate an index indicative of vascular function. Preferably, the index also indicates the potential effect of revascularization. For example, the index can be calculated based on the volume of the crowns in the model and the contribution of stenosed blood vessels to the resistance to blood flow in the crowns.

[0107] In some embodiments of the invention, a second model is generated from actual data, and one or more stenoses present in the patient's vasculature are modified as if they were being revascularized.

[0108] In some embodiments, the first model and the second model are compared and an index indicative of the potential effect of revascularization is generated based on comparing the physical characteristics in the first model with the physical characteristics in the second model.

[0109] In some embodiments, the index is fractional flow reserve (FFR), as known in the art.

[0110] In some embodiments, the index is some other measure that potentially correlates to the effectiveness of performing revascularization of one or more blood vessels, optionally at the location of the stenosis.

[0111] According to an aspect of some embodiments of the present invention there is provided a method for vascular assessment comprising receiving a plurality of 2D angiographic images of a portion of a subject's vascular system and using a computer to process the images and generate a first vascular tree for the portion of the vascular system within less than 60 minutes.

[0112] According to some embodiments of the invention, the vasculature has at least one catheter therein other than an angiographic catheter, and where images are processed and a tree is generated while the catheter is within the vasculature.

[0113] According to some embodiments of the invention, the method comprises using a vascular model to calculate an index indicative of vascular function.

[0114] According to some embodiments of the invention, the index is indicative of the need for revascularization.

[0115] According to some embodiments of the present invention, the calculation is performed in less than 60 minutes.

[0116] According to an aspect of some embodiments of the present invention there is provided a method of analysing angiographic images comprising receiving a plurality of 2D angiographic images of a portion of a subject's vasculature and using a computer to process the images to generate a tree model of the vasculature.

[0117] According to an aspect of some embodiments of the present invention, there is provided a method of treating a vasculature comprising capturing a plurality of 2D angiographic images of the vasculature of a subject that is immobilized on a treatment surface, processing the images to generate a vascular tree for the vasculature while the subject is immobilized, identifying constricted vessels in the tree, and expanding a stent at a portion of the vasculature corresponding to the constricted vessels in the tree.

[0118] According to some embodiments of the present invention, the plurality of 2D angiographic images comprises at least three 2D angiographic images, and wherein the tree model is a 3D tree model.

[0119] According to some embodiments of the invention, the method comprises identifying a stenosed vessel and a crown of the stenosed vessel in a first vascular tree and calculating a resistance to fluid flow in the crown, wherein an index is calculated based on a volume of the crown and based on a contribution of the stenosed vessel to the resistance to fluid flow.

[0120] According to some embodiments of the invention, the vascular tree comprises data relating to the location, orientation and diameter of blood vessels at multiple points within a portion of the vascular system.

[0121] According to some embodiments of the invention, the method comprises processing the image to generate a second three-dimensional vascular tree for the vasculature, the second vascular tree corresponding to the first vascular tree in which stenotic vessels are replaced by dilated vessels, and wherein calculation of the index is based on the first tree and the second tree.

[0122] According to some embodiments of the invention, the method comprises processing the image to generate a second three-dimensional vascular tree for the vascular system, the second vascular tree corresponding to a portion of the vascular system that is geometrically similar to the first vascular tree without including the stenosis, and wherein calculation of the index is based on the first tree and the second tree.

[0123] According to some embodiments of the invention, the method comprises obtaining a fractional flow reserve (FFR) based on the index.

[0124] According to some embodiments of the invention, the method comprises determining, based on the index, a ratio of maximum blood flow in the region of the stenosis to maximum blood flow in the same region without the stenosis.

[0125] According to some embodiments of the invention, the method comprises minimally invasively treating a stenosed blood vessel.

[0126] According to some embodiments of the invention, treatment is performed within less than one hour of calculating the index.

[0127] According to some embodiments of the invention, the method comprises storing the tree in a computer readable medium.

[0128] According to some embodiments of the invention, the method comprises transmitting the tree to a remote computer.

[0129] According to some embodiments of the invention, the method comprises capturing a 2D angiographic image.

[0130] According to some embodiments of the present invention, capturing the plurality of 2D angiographic images is performed by a plurality of imaging devices to capture the plurality of 2D angiographic images.

[0131] According to some embodiments of the present invention, capturing a plurality of 2D angiographic images comprises synchronizing a plurality of imaging devices to capture the plurality of images at substantially the same phase within a cardiac cycle.

[0132] According to some embodiments of the present invention, the synchronizing is according to the subject's ECG signal.

[0133] According to some embodiments of the invention, the method comprises detecting corresponding image features in each of the N angiographic images, where N is an integer greater than 1, calculating image correction parameters based on the corresponding image features, and performing registration of the N-1 angiographic images to geometrically correspond to angiographic images other than the N-1 angiographic images based on the correction parameters.

[0134] According to some embodiments of the invention, the method comprises defining a surface corresponding to the shape of the subject's heart and using the surface as a constraint for detection of corresponding image features.

[0135] According to some embodiments of the invention, the method comprises compensating for respiration and patient movement.

[0136] According to an aspect of some embodiments of the present invention there is provided a computer software product comprising a computer readable medium having stored thereon program instructions which, when read by a computer, cause the computer to receive a plurality of 2D angiographic images of a subject's vasculature and perform the method as set forth above and, where appropriate, as further detailed below.

[0137] According to an aspect of some embodiments of the present invention there is provided a system for vascular assessment comprising a plurality of imaging devices configured to capture a plurality of 2D angiographic images of a subject's vasculature, and a computer configured to receive the plurality of 2D images and to perform the method as set forth above and as further detailed below where appropriate.

[0138] According to an aspect of some embodiments of the present invention, a vascular assessment system is provided that is functionally connected to a plurality of angiographic imaging devices for capturing a plurality of 2D images of a portion of a subject's vascular system, and is configured to receive data from the plurality of angiographic imaging devices and process the images to generate a tree model of the vascular system, wherein the tree model comprises geometric measurements of the vascular system at one or more positions along the vessels of at least one branch of the vascular system.

[0139] According to some embodiments of the invention, the system comprises a synchronization unit configured to provide a synchronization signal to the multiple angiographic imaging devices for synchronizing the capture of multiple 2D images of the vasculature.

[0140] According to some embodiments of the invention, a computer is configured to receive a subject ECG signal and select 2D images that correspond to substantially the same phase in a cardiac cycle based on the ECG signal.

[0141] According to some embodiments of the present invention, the system comprises an image registration unit configured to detect corresponding image features in each of the N angiographic images, where N is an integer greater than 1, calculate image correction parameters based on the corresponding image features, and perform registration of the N-1 angiographic images to geometrically correspond to angiographic images other than the N-1 angiographic images based on the correction parameters.

[0142] According to some embodiments of the invention, the computer is configured to define a surface corresponding to the shape of the subject's heart and to use the surface as a constraint for detection of corresponding image features.

[0143] According to some embodiments of the invention the computer is configured to compensate for respiration and patient movement.

[0144] According to some embodiments of the present invention, compensating comprises iteratively repeating the detection of corresponding image features each time for different subsets of the angiographic images, and updating image correction parameters in response to the repeated detection of the corresponding image features.

[0145] According to some embodiments of the invention, N is greater than 2. According to some embodiments of the invention, N is greater than 3.

[0146] According to some embodiments of the present invention, the corresponding image features comprise at least one of the group consisting of an origin of the tree model, a location of a minimum radius within a stenotic vessel, and a bifurcation of the vessel.

[0147] According to some embodiments of the invention, the tree model comprises data relating to the placement, orientation and diameter of blood vessels at multiple points within a portion of the vascular system.

[0148] According to some embodiments of the invention, the tree model comprises measurements of the vasculature at one or more locations along the vessels of at least one branch of the vasculature.

[0149] According to some embodiments of the invention, the geometric measurements of the vasculature are measurements at one or more locations along a centerline of at least one branch of the vasculature.

[0150] According to some embodiments of the invention, the tree model comprises data relating to blood flow characteristics at one or more of the plurality of points.

[0151] According to some embodiments of the invention, the portion of the vascular system comprises a cardiac artery.

[0152] According to an aspect of some embodiments of the present invention, there is provided a method for performing vascular assessment, comprising receiving a plurality of 2D angiographic images of a portion of a subject's vasculature and processing the images to generate a stenosis model for the vasculature, the stenosis model having measurements of the vasculature at one or more locations along blood vessels of the vasculature, obtaining flow characteristics of the stenosis model, and calculating an index indicative of vascular function based at least in part on the flow characteristics in the stenosis model.

[0153] According to some embodiments of the present invention, the flow characteristics of the stenosis model comprise a resistance to fluid flow.

[0154] According to some embodiments of the invention, the method comprises identifying a stenotic vessel and a crown of the stenotic vessel in a first stenosis model and calculating a resistance to fluid flow in the crown, wherein an index is calculated based on the volume of the crown and based on the contribution of the stenotic vessel to the resistance to fluid flow.

[0155] According to some embodiments of the present invention, the flow characteristics of the stenosis model comprise a fluid flow.

[0156] According to some embodiments of the invention, the stenosis model is a three-dimensional vascular tree.

[0157] According to some embodiments of the invention, the vascular tree comprises data relating to the location, orientation and diameter of blood vessels at multiple points within a portion of the vascular system.

[0158] According to some embodiments of the invention, this process comprises expanding the stenosis model at one bifurcation, calculating new flow characteristics in the expanded stenosis model, updating the index in response to the new flow characteristics and according to a predetermined criterion, and iteratively repeating the expanding, calculating and updating.

[0159] According to some embodiments of the invention, the method includes processing the image to generate a second model of the vascular system and obtaining flow characteristics of the second model, wherein calculation of the index is based on the flow characteristics in the stenosis model and the flow characteristics in the second model.

[0160] According to some embodiments of the present invention, in this method the second model is a normal model with a dilated blood vessel replacing the stenotic blood vessel in the stenosis model.

[0161] According to some embodiments of the invention, the stenosis model is a three-dimensional vascular tree and the second model is a second three-dimensional vascular tree.

[0162] According to some embodiments of the invention, each of these models corresponds to a portion of the vascular system that lies between two successive branches of the vascular system and includes a stenosis.

[0163] According to some embodiments of the invention, each of these models corresponds to a portion of the vascular system that includes a branch of the vascular system.

[0164] According to some embodiments of the invention, each of these models corresponds to a portion of the vasculature that includes a stenosis and extends at least one branch of the vasculature beyond the stenosis.

[0165] According to some embodiments of the invention, each of these models corresponds to a portion of the vasculature that includes a stenosis and extends at least three branches of the vasculature beyond the stenosis.

[0166] According to some embodiments of the present invention, each of the methods, models corresponds to a portion of the vasculature that includes the stenosis and extends as distally as image resolution allows.

[0167] According to some embodiments of the invention, the stenosis model corresponds to a portion of the vasculature that includes a stenosis, and the second model corresponds to a portion of the vasculature that does not include the stenosis and is geometrically similar to the stenosis model.

[0168] According to some embodiments of the invention, this process comprises extending the models with one branch each, calculating new flow characteristics in each extended model, updating the index in response to the new flow characteristics and according to a predetermined criterion, and iteratively repeating the extending, calculating and updating.

[0169] According to some embodiments of the invention, the index is calculated based on a ratio of flow characteristics in a stenosis model to flow characteristics in a second model.

[0170] According to some embodiments of the invention, the index is indicative of the need for revascularization.

[0171] According to an aspect of some embodiments of the present invention, there is provided a method for performing a vascular assessment, comprising generating a stenosis model of a subject's vasculature, the stenosis model including measurements of the subject's vasculature at one or more locations along a vascular centerline of the subject's vasculature, obtaining flow characteristics of the stenosis model, generating a second model of a similar extent of the subject's vasculature as the stenosis model, obtaining flow characteristics of the second model, and calculating an index indicative of the need for revascularization based on the flow characteristics in the stenosis model and the flow characteristics in the second model.

[0172] According to some embodiments of the invention, the second model is a normal model that includes a dilated blood vessel that replaces the stenotic blood vessel in the stenosis model.

[0173] According to some embodiments of the invention, the vascular system includes cardiac arteries of the subject.

[0174] According to some embodiments of the present invention, generating a stenosis model of a subject's vasculature includes using a plurality of angiography imaging devices to capture a plurality of 2D images of the subject's vasculature, and generating the stenosis model based on the plurality of 2D images.

[0175] According to some embodiments of the present invention, the flow characteristics include fluid flow.

[0176] According to some embodiments of the invention, obtaining the flow characteristics of the stenosis model includes measuring fluid flow within the subject's vasculature at one or more locations within the extent of the subject's vasculature included in the stenosis model, and obtaining the flow characteristics of the second model includes calculating fluid flow within the subject's vasculature at one or more locations within the extent of the subject's vasculature included in the second model based, at least in part, on correcting the fluid flow of the stenosis model to account for dilated blood vessels.

[0177] According to some embodiments of the present invention, the flow characteristic comprises resistance to fluid flow.

[0178] According to some embodiments of the invention, obtaining the flow characteristics of the stenosis model includes calculating resistance to flow based at least in part on a cross-sectional area of ​​the subject's vasculature at one or more locations within the extent of the subject's vasculature included in the stenosis model, and obtaining the flow characteristics of the second model includes calculating resistance to flow based at least in part on an expanded cross-sectional area of ​​the subject's vasculature at one or more locations within the extent of the subject's vasculature included in the second model.

[0179] According to some embodiments of the invention, the extent of each one of the stenosis model and the second model includes a segment of the vasculature between two successive branches of the vasculature that includes a stenosis.

[0180] According to some embodiments of the invention, the extent of each one of the stenosis model and the second model includes a segment of the vasculature that includes a branch of the vasculature.

[0181] According to some embodiments of the invention, each one of the stenosis model and the second model includes a widening of the vasculature that includes a stenosis and extends at least one branch of the vasculature beyond the stenosis.

[0182] According to some embodiments of the invention, each one of the stenosis model and the second model includes a stretch of vasculature, each including a stenosis and a dilated stenosis, expanding at least three branches of the vasculature beyond the stenosis.

[0183] According to some embodiments of the invention, each one of the stenosis model and the second model includes a stretch of vasculature that includes the stenosis and extends as distally as the resolution of the imaging modality allows.

[0184] According to some embodiments of the present invention, each one of the stenosis model and the second model includes a stenosis and a width of the vasculature that extends distally beyond the stenosis to at least one branch of the vasculature, and further includes storing flow characteristics of the stenosis model as previous flow characteristics of the stenosis model and storing flow characteristics of the second model as previous flow characteristics of the second model, extending the width of the stenosis model and the second model by another branch, calculating new flow characteristics in the stenosis model and calculating new flow characteristics in the second model, and determining whether to calculate an index indicative of the need for revascularization if a difference between the new flow characteristics of the stenosis model and the previous characteristics of the stenosis model is less than a first specified difference and a difference between the new flow characteristics of the second model and the previous characteristics of the second model is less than a second specified difference, calculating an index indicative of the need for revascularization, and otherwise repeating the storing, expanding, calculating, and determining.

[0185] According to some embodiments of the invention, the stenosis model includes a stretch of vasculature that includes the stenosis, and the second model includes a stretch of vasculature that is geometrically similar to the first model but does not include the stenosis.

[0186] According to some embodiments of the invention, the index is calculated as a ratio between the flow characteristics in a stenosis model and the flow characteristics in a second model.

[0187] According to some embodiments of the present invention, the calculated index is used to determine the fractional flow reserve (FFR).

[0188] According to some embodiments of the invention, the calculated index is used to determine the ratio of maximum blood flow in a region of a stenosis to the maximum blood flow in the same region without a stenosis.

[0189] According to some embodiments of the present invention, generating a stenosis model, obtaining flow characteristics of the stenosis model, generating a second model, obtaining flow characteristics of the second model, and calculating the index are all performed during diagnostic catheterization before the catheter used for the diagnostic catheterization is withdrawn from the subject's body.

[0190] According to an aspect of some embodiments of the present invention, there is provided a method for performing vascular assessment, comprising: capturing a plurality of 2D angiographic images of a subject's vasculature; generating a tree model of the subject's vasculature, the tree model including geometric measurements of the subject's vasculature at one or more locations along a vascular centerline of at least one branch of the subject's vasculature, using at least a portion of the plurality of captured 2D angiographic images; and generating a model of flow characteristics of the first tree model.

[0191] According to some embodiments of the invention, the vascular system includes cardiac arteries of the subject.

[0192] According to some embodiments of the present invention, capturing the plurality of 2D angiographic images includes capturing the plurality of 2D angiographic images using a plurality of imaging devices.

[0193] According to some embodiments of the present invention, capturing a plurality of 2D angiographic images includes synchronizing a plurality of imaging devices to capture the multiple images simultaneously.

[0194] According to some embodiments of the present invention, the synchronizing uses the subject's ECG signal.

[0195] According to some embodiments of the present invention, the synchronizing includes detecting corresponding image features in at least a first 2D angiographic image and a second 2D angiographic image of the plurality of 2D angiographic images, calculating image correction parameters based on the corresponding image features, and registering the at least second 2D angiographic image to geometrically correspond to the first 2D angiographic image, wherein the corresponding image features include at least one of the group consisting of an origin of a tree model, a location of a minimum radius in a stenotic vessel, and a bifurcation of the vessel.

[0196] According to one aspect of some embodiments of the present invention, a system is provided for performing vascular assessment, the system including a computer operatively connected to a plurality of angiographic imaging devices for capturing a plurality of 2D images of a subject's vasculature, receiving data from the plurality of angiographic imaging devices, and generating a tree model of the subject's vasculature using at least some of the plurality of captured 2D images, wherein the tree model includes geometric measurements of the subject's vasculature at one or more locations along a vascular centerline of at least one branch of the subject's vasculature, and configured to generate a model of flow characteristics of the tree model.

[0197] According to some embodiments of the invention, the vascular system includes cardiac arteries of the subject.

[0198] According to some embodiments of the present invention, the apparatus further comprises a synchronization unit configured to provide a synchronization signal to the plurality of angiographic imaging devices for synchronizing the capture of the plurality of 2D images of the vasculature of the subject.

[0199] According to some embodiments of the present invention, the method further comprises a synchronization unit configured to receive an ECG signal of the subject and to select a 2D image from data from the multiple angiographic imaging devices at the same cardiac phase within the 2D image.

[0200] According to some embodiments of the present invention, the system further comprises an image registration unit configured to detect corresponding image features in at least a first 2D image and a second 2D image from data from a plurality of angiographic imaging devices, calculate image correction parameters based on the corresponding image features, and register the at least second 2D image to geometrically correspond to the first 2D image, wherein the corresponding image features include at least one of the group consisting of an origin of the tree model, a location of a minimum radius in a stenotic blood vessel, and a bifurcation of the blood vessel.

[0201] According to an aspect of some embodiments of the present invention, a method for performing a vascular assessment includes generating a stenosis model of a subject's vasculature, the stenosis model including geometric measurements of the subject's vasculature at one or more locations along a vascular centerline of the subject's vasculature, the stenosis model including a stenosis and an extension of the vasculature beyond the stenosis, obtaining flow characteristics of the stenosis model, generating a second model of a similar extension of the subject's vasculature as the stenosis model, obtaining flow characteristics of the second model, and calculating an index indicative of a need for revascularization based on the flow characteristics in the stenosis model and the flow characteristics in the second model, and calculating an index indicative of a need for revascularization based on the flow characteristics in the stenosis model and the flow characteristics in the second model, The method further includes storing the flow characteristics of the stenosis model as dynamic characteristics and storing the flow characteristics of the second model as previous flow characteristics of the second model, further expanding the extent of the stenosis model and the second model by one branch, calculating new flow characteristics in the stenosis model, calculating new flow characteristics in the second model, and determining whether to calculate an index indicating the need for revascularization if a difference between the new flow characteristics of the stenosis model and the previous characteristics of the stenosis model is smaller than a first specified difference and a difference between the new flow characteristics of the second model and the previous characteristics of the second model is smaller than a second specified difference, calculating an index indicating the need for revascularization, and otherwise repeating the storing, expanding, calculating, and determining.

[0202] Unless otherwise specified, all technical and / or scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention pertains. Although many methods and materials similar or equivalent to those described herein can be used in carrying out or testing embodiments of the present invention, exemplary methods and / or materials are described below. In case of discrepancy, the present patent specification, including definitions, will control. In addition, the materials, methods, and examples are illustrative only and are not necessarily intended to be limiting.

[0203] As will be appreciated by those skilled in the art, aspects of the present invention may be embodied as a system, method, or computer program product. Accordingly, aspects of the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment (including firmware, resident software, microcode, etc.), or an embodiment combining software and hardware aspects, all of which may be generally referred to herein as a "circuit," "module," or "system." Furthermore, aspects of the present invention may take the form of a computer program product embodied in one or more computer readable medium(s) in which computer readable program code is embodied. Implementation of the method and / or system of embodiments of the present invention may involve performing or completing selected tasks manually, automatically, or a combination thereof.

[0204] For example, hardware for performing selected tasks according to embodiments of the invention may be implemented as a chip or circuit. As software, selected tasks according to embodiments of the invention may be implemented as a number of software instructions executed by a computer using a suitable operating system. In an exemplary embodiment of the invention, one or more tasks according to exemplary embodiments of the method and / or system as described herein are performed by a data processor, such as a computing platform for executing a number of instructions. Optionally, the data processor comprises volatile memory for storing instructions and / or data, and / or non-volatile storage, e.g., a magnetic hard disk and / or removable media, for storing instructions and / or data. Optionally, a network connection is provided as well. Optionally, a display and / or user input devices, such as a keyboard or mouse, are also provided.

[0205] A combination of one or more computer readable medium(s) may be utilized. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a suitable combination of the foregoing. More specific examples (non-exhaustive list) of computer readable storage media include an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or a suitable combination of the foregoing. In the context of this specification, a computer readable storage medium may be a tangible medium that contains or can store a program used by or in connection with an instruction execution system, apparatus, or device.

[0206] A computer-readable signal medium may include a data signal propagated with computer-readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated signal may take a variety of forms, including, but not limited to, electromagnetic, optical, or any suitable combination thereof. A computer-readable signal medium may be a computer-readable medium that is not a computer-readable storage medium and that can convey, propagate, or carry a program for use in or in connection with an instruction execution system, apparatus, or device.

[0207] The program code embodied on the computer readable medium may be transmitted using any suitable medium, including but not limited to wireless, wired, fiber optic cable, RF, etc., or any suitable combination of the foregoing.

[0208] Computer program code for carrying out operations for aspects of the present invention may be written in a combination of one or more programming languages, including object-oriented programming languages ​​such as Java, Smalltalk, C++, or the like, and traditional procedural programming languages ​​such as the "C" programming language or similar programming languages. The program code may run entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer, partially on a remote computer, or entirely on a remote computer or server. In the latter scenario, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or wide area network (WAN), or the connection may be made to an external computer (e.g., through the Internet using an Internet Service Provider).

[0209] Aspects of the present invention are described below with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be sent to a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, whereby the instructions, executed via a processor of the computer or other programmable data processing apparatus, create means for performing the functions / activities specified in one or more blocks of the flowchart illustrations and / or block diagrams.

[0210] These computer program instructions, which may direct a computer, other programmable data processing apparatus, or other device to function in a particular manner, may also be stored in a computer readable medium, such that an article of manufacture containing instructions that perform the functions / activities specified in one or more blocks of the flowcharts and / or block diagrams, with the instructions stored in the computer readable medium, may be produced.

[0211] These computer program instructions may then be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device, and the instructions executing on the computer or other programmable data processing apparatus may constitute a process that performs the functions / activities specified in one or more blocks of the flowcharts and / or block diagrams.

[0212] Some embodiments of the present invention are herein described, by way of example only, with reference to the accompanying drawings and images. With particular reference now to the drawings and images in detail, it should be noted that the particulars shown are exemplary and are intended to be illustrative of embodiments of the present invention. In this regard, the description accompanying the drawings and images will make clear to those skilled in the art how embodiments of the present invention may be practiced. [Brief description of the drawings]

[0213] [Figure 1] 1A-1C show an original image and a Franzi-filtered image processed according to some exemplary embodiments of the present invention; [Diagram 2] FIG. 2 illustrates a light-colored centerline superimposed on the original image of FIG. 1, according to some exemplary embodiments of the present invention. [Figure 3A] 1 is an image of a coronary vascular tree model generated according to some exemplary embodiments of the present invention; [Figure 3B]3B is an image of the coronary vascular tree model of FIG. 3A with branch tags added in accordance with certain exemplary embodiments of the present invention; [Figure 3C] 1 is a simplified diagram of a tree model of a coronary vascular tree generated according to some exemplary embodiments of the present invention; [Figure 4] A series of nine images showing vessel segment radii generated by example embodiments of the present invention along the branches of the coronary vascular tree model shown in Figure 3C as a function of distance along each branch, in accordance with some exemplary embodiments of the present invention. [Diagram 5] 1 illustrates a coronary artery tree model, a combination matrix indicating tree branch tags, and a combination matrix indicating tree branch resistances, all generated by some exemplary embodiments of the present invention. [Figure 6] FIG. 1 illustrates a tree model of the vascular system with tags numbering the outlets of the tree model, where the tags correspond to streamlines, generated by an example embodiment of the present invention, in accordance with some exemplary embodiments of the present invention. [Figure 7] FIG. 2 is a simplified diagram of a vascular tree model generated by an example embodiment of the present invention, including branch resistance Ri of each branch and calculated flow rate Qi of each streamline outlet, in accordance with some exemplary embodiments of the present invention. [Figure 8] 1 is a simplified flow diagram illustrating FFR index generation, according to some exemplary embodiments of the present invention. [Figure 9] 4 is a simplified flow diagram illustrating another method of FFR index generation, according to some exemplary embodiments of the present invention. [Figure 10] 4 is a simplified flow diagram illustrating yet another method of FFR index generation, according to some exemplary embodiments of the present invention. [Figure 11] 1 is a simplified drawing of a vasculature comprising stenosed and non-stenosed blood vessels as pertaining to some exemplary embodiments of the present invention; [Figure 12A] 1 is a simplified diagram of a hardware implementation of a system for vascular assessment constructed in accordance with some exemplary embodiments of the present invention; [Figure 12B]2 is a simplified diagram of another hardware implementation of a system for vascular assessment, constructed in accordance with some exemplary embodiments of the present invention; [Figure 13] 1 is a flow chart illustrating an example overview of stages in constructing a vascular model, according to some example embodiments of the present invention. [Figure 14] 1 is a flow chart outlining an exemplary detailed overview of steps in constructing a vascular model, according to some exemplary embodiments of the present invention. [Figure 15] 1 is a schematic diagram of an exemplary arrangement of imaging coordinates for an imaging system, according to certain exemplary embodiments of the present invention; [Figure 16] 1 is a simplified flow diagram of processing operations involving anisotropic diffusion, according to some exemplary embodiments of the present invention. [Figure 17A] 4 is a simplified flow diagram of processing operations including motion compensation, according to some exemplary embodiments of the present invention. [Figure 17B] 4 is a simplified flow diagram of processing operations including alternative or additional methods of motion compensation, according to some exemplary embodiments of the present invention. [Figure 18A] 4A-4C illustrate aspects of the computation of a "heart shell" constraint for ignoring bad ray intersections from computed correspondences between images, according to some exemplary embodiments of the present invention. [Figure 18B] 4A-4C illustrate aspects of the computation of a "heart shell" constraint for ignoring bad ray intersections from computed correspondences between images, according to some exemplary embodiments of the present invention. [Figure 18C] 1 is a simplified flow diagram of processing operations involving constraining pixel correspondences within a volume near a heart surface, according to some exemplary embodiments of the present invention. [Figure 19A] 1A-1C illustrate identification of homology between branch vessels, according to some exemplary embodiments of the present invention. [Figure 19B] 1A-1C illustrate identification of homology between branch vessels, according to some exemplary embodiments of the present invention. [Figure 19C] 1A-1C illustrate identification of homology between branch vessels, according to some exemplary embodiments of the present invention. [Figure 19D] 1A-1C illustrate identification of homology between branch vessels, according to some exemplary embodiments of the present invention. [Figure 19E] 1 is a simplified flow diagram of processing operations involving identifying homologous regions along vascular branches, according to some exemplary embodiments of the present invention. [Figure 20A] 1 is a simplified flow diagram of processing operations involving selecting projection pairs along a vascular centerline, according to some exemplary embodiments of the present invention. [Figure 20B] 3A-3C are schematic representations of epipolar determination of 3-D target positions from 2-D image positions and their geometrical relationships in space, according to some exemplary embodiments of the present invention; [Figure 21] 4 is a simplified flow diagram of processing operations involving generating an edge graph and finding connected paths along the edge graph, in accordance with some exemplary embodiments of the present invention. [Figure 22] 1 is a simplified schematic of an automated VSST scoring system, according to some exemplary embodiments of the present invention. [Figure 23] 1A-1C illustrate example branched structures having recombining branches, according to some example embodiments of the present invention. [Figure 24] 1 is a Bland-Altman plot of the difference between FFR index and image-based FFR index as a function of its mean, according to certain exemplary embodiments of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0214] In some of its embodiments, the present invention relates to vascular modeling, and more particularly, but not exclusively, to the use of vascular models to generate indices related to vascular function and diagnosis in real time, e.g., during catheter insertion imaging procedures.

[0215] A broad aspect of some embodiments of the present invention relates to calculating fractional flow reserve (FFR) based on imaging of a portion of the vasculature.

[0216] An aspect of some embodiments of the invention relates to the calculation of a model of blood flow in a subject. In some embodiments, the portion of the vasculature imaged is a coronary vessel. In some embodiments, the vasculature is an artery. In some embodiments, the blood flow is modeled based on vessel diameters in a 3-D reconstruction of the vascular tree. Optionally, vascular resistance is determined based on vessel diameters. Optionally, vascular resistance is calculated for the stenosed vessel and for the coronary portion of the vessel (the vessel downstream of the stenosed vessel). In some embodiments, the FFR is calculated from a general 3-D reconstruction of the vascular tree, e.g., a vascular tree reconstruction from a CT scan. In some embodiments, the reconstruction is performed ab initio, e.g., from 2-D angiographic image data. Optionally, a given vascular tree meets the specific requirements of the FFR calculation, e.g., by reduction to a graphical representation of vessel width as a function of vessel widening.

[0217] One aspect of some embodiments of the present invention relates to the calculation of FFR based on the difference in flow between a vascular model of a vasculature potentially having a stenosis and a different vascular model derived from and / or homologous to the stenotic vasculature model. In some embodiments, the modification to a non-stenotic version of the vascular model includes a determination of wall opening (dilation) in the stenotic region based on reference width measurements obtained in one or more other portions of the vasculature. In some embodiments, the reference width measurements are obtained from vessel portions on either side of the stenosis. In some embodiments, the reference width measurements are obtained from a vessel that is naturally non-stenotic with a branch order similar to the stenotic segment.

[0218] In some embodiments of the invention, the FFR index comprises a ratio of flow rates between a model with a potentially stenosed vascular segment and a model in which said segment has been replaced with a lower flow resistance segment and / or the resistance to flow caused by said segment has been removed. A potential advantage of determining this ratio is that the index comprises a representation of the effect of a potential therapeutic treatment on the vasculature, e.g., opening of a vascular region by percutaneous coronary intervention (PCI), such as stent implantation. Another potential advantage of this ratio is that it measures a parameter (fractional flow reserve) that is well accepted as providing an indication of the need for revascularization, but which is commonly determined in the art by invasive pressure measurements that require direct access to both sides of the stenotic lesion.

[0219] A broad aspect of some embodiments of the present invention relates to the generation of a vascular tree model.

[0220] An aspect of some embodiments of the invention relates to building a tree model of a portion of a mammalian vasculature based on automatic matching of homologous features among a plurality of vascular images. In some embodiments of the invention, the tree model comprises vascular segment centerlines. Optionally, the homologous matching is between vascular segment centerlines and / or portions thereof. In some embodiments, the modeled spatial relationship between vascular segment centerlines comprises association of segment ends at branch nodes.

[0221] A potential advantage of using vessel segment centerlines, and / or other features easily identifiable from vessel segment data in the images, is that they provide rich "meta-features" that can be subjected to ray intersection testing by back-projecting rays from several pairs (or more) of separate images into 3-D space based on the imaging configuration. In some embodiments, exact ray intersection is not required, and intersection within a volume is sufficient to establish homology. Note that while isolated features are potentially useful for such intersection-based homology identification, expanded features of paths along the vessel segments (for example) enable the use in some embodiments of correlation and / or constraint techniques for further refinement of initial potential tentative homology identification. Thus, ray intersections, in some embodiments, replace manual identification of homologous features in different vessel projections.

[0222] In some embodiments of the invention, the modeled vascular system comprises the vascular system of the heart (cardiovascular system), in particular the vascular system of the coronary arteries and their branches. In some embodiments, the tree model comprises 3-D positional information for the cardiovascular system.

[0223] An aspect of some embodiments of the invention relates to positions (and in particular 3-D spatial positions) in a model of the cardiovascular system that are derived based on and / or through coordinates defined by features of the vasculature itself. In some embodiments, the same vascular feature (e.g., a vascular centerline) both defines the 3-D space of the model and additionally comprises the backbone of the model itself. Optionally, the centerline is represented according to its 3-D position in space and at the same time as a position in the graph space defined by the vascular tree comprising connecting segments of the centerlines of the nodes. In some embodiments, a vascular segment comprises data associated with a vascular path (e.g., a vascular centerline) connecting two branch nodes.

[0224] In some embodiments of the invention, a "consensus" 3-D space defined by matching between vascular feature positions is the result of tree model construction.

[0225] A potential advantage of this vascular-centric modeling approach concerns the fact that the heart (to which the vasculature is mechanically coupled) is constantly moving. In some embodiments, the vasculature model is constructed from a series of 2-D images taken sequentially. During cardiovascular imaging and / or between imaging positions, regions of the vasculature potentially change their actual and / or calibrated positions in 3-D space (absolute and / or relative). This is due, for example, to heart beating, breathing, voluntary movements, and / or misalignments in determining the image projection plane. In some embodiments, the imaging protocol is modified to compensate to some extent for these movements, for example, by synchronizing the imaging instant to a particular phase of the cardiac cycle (e.g., end of diastole). However, errors potentially remain even after such time due, for example, to natural variations in the cardiac cycle, effects of different physiological cycles (cardiac and respiratory) being out of phase, and time limitations for available imaging. Thus, potentially, there is no "natural" 3-D space common to raw 2-D image data. Targeting a consensus space potentially allows one to reframe the modeling problem with respect to the consistency of modeling outcomes.

[0226] Although the 3-D location changes, other features of the vascular location, such as connectivity along the vasculature and / or ordering of regions, are invariant with respect to motion artifacts. It is therefore potentially advantageous to use features of the vasculature itself to determine a frame of reference on which the 3-D reconstruction can be established. In some embodiments, the features on which the 3-D reconstruction is based comprise 2-D centerlines of vessel segments present in multiple images whose similarity is established by an automatic, optionally iterative, method. A potential advantage of using centerlines as a basis for 3-D modeling is that the centerlines that anchor the construction of the model tree can also be used independently as a 1-D coordinate system. Thus, using centerlines as a basis for reconstruction can ensure consistency and / or continuity of tree model features associated with centerline locations.

[0227] In some embodiments, other features related to the vasculature are used as landmarks, e.g., points of minimum vessel width, vessel branching points, and / or vessel origins. Optionally, vascular features such as centerlines are transformed along with transformations to landmark features (without themselves being the target of cross image matching) before being incorporated into the vascular tree model.

[0228] One aspect of some embodiments of the present invention relates to using iterative projection and backprojection between 2-D and 3-D coordinate systems to arrive at a consensus coordinate system that relates the 2-D image plane to the 3-D system of target coordinates.

[0229] In some embodiments of the invention, the assignment of consensus 3-D positions to landmark vascular features (e.g., vascular centerlines) projected from a single target region onto multiple 2-D images during imaging comprises reprojection and / or reregistration of the 2-D images themselves to better match the "consensus" 3-D space. Optionally, the reprojection assigns the 2-D images to a different image plane than the one originally recorded relative to the image plane. Optionally, the reregistration comprises non-linear distortion of the images, e.g., to compensate for deformation of the heart during imaging. Optionally, the reprojection and / or reregistration is performed iteratively, e.g., by defining different sets of images as "target" and "matching" in different feature registration iterations. In some embodiments of the invention, different numbers of images are used to define homologous features and for subsequent analysis of additional image features (e.g., vessel width) to which the homologous features relate.

[0230] One aspect of some embodiments of the present invention relates to reducing the computational complexity of decision trees, thereby allowing faster processing for arriving at clinical conclusions.

[0231] In some embodiments, portions of the image data (e.g., "non-feature" pixel values) are optionally maintained in the 2-D representation without requiring a full 3-D reconstruction. In some embodiments, calculation of the 3-D positions of non-landmark features, such as, for example, vessel wall positions, is thereby avoided, simplified, and / or postponed. In particular, in some embodiments, vessel edges are recognized from direct processing of the 2-D image data (e.g., comprising inspection of image gradients perpendicular to the vessel centerline). Optionally, the determined edges are projected into 3-D space (e.g., represented as one or more radii extending perpendicularly from the 3-D centerline position) without requiring projection of the original image pixel data into a 3-D voxel representation.

[0232] Additionally or alternatively, vascular wall positions are determined and / or processed (e.g., to determine vascular resistance) in one or more "1-D" spaces defined by a frame of reference comprising positions along the centerline. Optionally, this processing is independent of, for example, a projection of the wall positions into a 3-D space. In some embodiments, further computational complexity is reduced by reducing the model to a 1-D function of the centerline positions, for example, to determine vascular flow characteristics.

[0233] An aspect of some embodiments of the present invention relates to relationships between vascular model components with different dimensionality. In some embodiments, 1-D, 2-D, and / or 3-D positions and / or logical connectivity and / or properties with non-positional or partially non-positional properties are related to each other by direct functions and / or indirectly through intermediate frames of reference.

[0234] In some embodiments, for example, the vascular model comprises one or more of the following features:

[0235] a 2-D image having a position in 3-D space defined by a relationship between homologous features in the 3-D image; a vessel extent comprising one or more 1-D axes for a function of one or more vessel characteristics, e.g., diameter, radius, flow rate, flow resistance, and / or curvature; A vessel extent comprising one or more 1-D axes as a function of position in 3-D space; connectivity between vessel extents described as nodes (e.g., nodes connecting the ends of vessel segments) with respect to position along the vessel extent; 2-D images in which the 1-D axis of vascular extent is mapped; A 2-D frame with the extent of the vessel along one axis, and image data orthogonal to the extent of the vessel along a second axis.

[0236] A broad aspect of some embodiments of the present invention relates to the real-time determination and / or use of a vascular tree model to provide clinical diagnostic information while a catheter insertion procedure is in progress on a subject.

[0237] An aspect of some embodiments of the present invention relates to utilizing real-time automatic vascular status determination to interactively manipulate a clinical diagnostic procedure as it progresses. Real-time determination, in some embodiments, comprises a determination within the time frame of a catheterization procedure, e.g., 30 minutes, 1 hour, or less, more, or intermediate times. More specifically, real-time determination comprises a determination that is just in time to affect the decision and / or outcome of the catheterization procedure, starting from the image on which the vascular status determination is based. For example, it is a potential advantage to select a particular portion of the vascular tree for an initial calculation, where the calculation is likely to be completed in a short enough time to affect the decision to perform a particular PCI procedure, such as implanting a stent. For example, a 5 minute delay for the calculation of an FFR with two main vessel branches can be reduced to a 2.5 minute delay by selecting the first branch to be the initial stage of the calculation, when the first branch appears to be particularly important based on the results of a cursory review of the image data. Additionally or alternatively, if the calculation is fast, the FFR result can be updated one or more times during the course of the catheterization procedure. For example, a first stent implantation potentially alters perfusion conditions at other sites sufficiently to cause autoregulatory changes in vessel width that may then alter the expected effect of subsequent stent implantations. Also, for example, the imaged effect of actual stent implantation on vessel width may be compared to the predicted effect to verify that a desired effect on flow capacity has been achieved. Some embodiments of the invention are configured to allow interface control of how vascular models and / or vascular properties are calculated, control model updates based on newly available image data, and / or select comparisons between actual and / or predicted vascular condition models.

[0238] An aspect of some embodiments of the invention relates to the construction of a vascular tree model suitable for targeted prediction of the outcome of a potential clinical intervention. Optionally, the clinical intervention is a PCI procedure, such as the implantation of a stent. In some embodiments of the invention, the goal comprises focusing the vascular tree model construction stage to arrive at pre / post results for vascular parameters that are directly available for clinical modification. In some embodiments, the vascular parameter is vessel width (e.g., modifiable by stent implantation). A potential advantage of focusing modeling on determining the difference between the pre-treatment and post-treatment states of the vasculature is that the effects of model modifications due to approximations of other vascular details are counterbalanced (and / or reduced in magnitude). In particular, these are potentially reduced in importance with respect to operational concerns such as "Do the changes brought about by the intervention usefully improve the clinical situation with respect to the known effects of the variables that the intervention is targeted at?". As one potential consequence, calculations that would otherwise be performed to fully model the functional and / or anatomical properties of the vasculature may be omitted. Potentially, this increases the speed at which flow indices can be generated.

[0239] One aspect of some embodiments of the invention relates to the formation of a model representation of the vasculature target definition of a framework for structuring one or more selected, clinically relevant parameters (such as vessel width, flow resistance, and flow itself). In some embodiments, the structuring comprises a vessel extent approach to modeling, where the position along the vessel segment defines a frame of reference. Optionally, the vessel extent frame of reference comprises a division between the branches linking the nodes of the vascular tree. Potentially, this provides a reduction in dimensions that reduces computation time.

[0240] In some embodiments, a model of the 3-D position in the vascular model is formed from potentially incomplete or accurate initial position information. This is achieved, for example, by annealing to a self-consistent framework by an iterative process of adjusting the position information to increase the consistency between the acquired data. Adjustment comprises operations such as, for example, transforming the image plane, distorting the image itself to increase similarity, and / or ignoring outliers that prevent consensus determination. Potentially, alternative approaches that seek to ensure the fidelity of the framework to a particular real-world configuration (e.g., one or more "real" 3-D configurations of a portion of the vasculature in space) perform a large amount of computation with respect to the benefits obtained for the estimation of the target parameters. In contrast, a framework that emphasizes internal consistency in the service that supports the computation related to the target parameters can potentially reduce the computational load using a consensus-like approach. In particular, this approach is potentially suitable for combination with the computation of changes in the vasculature, as described above.

[0241] Before describing at least one embodiment of the invention in detail, it will be understood that the invention is not necessarily limited in its application to the details of construction and the arrangement of components and / or methods set forth in the following description and / or illustrated in the drawings. The invention is capable of other embodiments or of being practiced or carried out in various ways.

[0242] It should be noted that in the exemplary embodiments described below, the coronary vasculature, and more specifically the coronary artery system, is used. The examples are not meant to limit embodiments of the invention to the coronary arteries, and embodiments of the invention potentially apply to other vascular systems, such as, for example, the venous system and lymphatic system.

[0243] In some embodiments, a first model of blood flow in a subject is constructed based on imaging the subject's vasculature. Typically, the first model is constructed from the vasculature including a problem segment of the vasculature, such as a stenosis in at least a portion of the blood vessel. In some embodiments of the invention, the first model corresponds to a portion of the vasculature including at least one blood vessel with a stenosis. In these embodiments, the first model describes a portion of the vasculature including at least one blood vessel with a stenosis and a crown. In these embodiments, the first model optionally further includes information related to the shape and / or volume of the crown and information related to the blood flow and / or resistance to blood flow in the stenotic blood vessel and / or the crown.

[0244] Typically, but not necessarily, a second model is constructed, optionally describing an at least partially healthy vasculature that corresponds to the first model. In some embodiments, the second model is constructed by modifying the stenosis of the first model to a more open state, as would be the case if a stent were to open the stenosis, and in some embodiments, the second model is constructed by selecting a segment of the subject's vasculature that includes a healthy vessel similar to the problematic vessel of the first model and using it to replace the stenotic vessel.

[0245] Vascular model construction is described below.

[0246] In some embodiments, an index indicative of the need for revascularization is calculated. This can be performed based on the first model or based on the results of a comparison of the first model and the second model of blood flow. The index is optionally used in a manner similar to the pressure measurement derived FFR index to evaluate whether the stenosed vessel affects the flow in the vasculature such that the prognosis for improvement of the subject's condition after dilatation of the stenosed vessel is higher than the probability of developing complications resulting from the dilatation itself.

[0247] The terms "FFR" and "FFR index" in all grammatical forms are used throughout this specification and are intended to refer to the above-mentioned indexes, and not just the FFR index referred to in the Background section as an invasive measurement involving the insertion of a guidewire equipped with a miniature pressure transducer over the stenosis. In some cases - particularly when differences between specific types of FFR and FFR-like indexes are being explained - subscripts are used to distinguish between them, e.g., FFR for FFR derived from pressure measurements. pressure , and / or FFR if FFR is expressed in terms of flow determination flow is used.

[0248] Acquiring data to build a vascular model In some embodiments, the data for modeling the vasculature comprises medical imaging data.

[0249] In some embodiments of the invention, the data is from minimally invasive angiographic images, e.g., x-ray images. In some embodiments, the angiographic images are two-dimensional (2-D). In some embodiments, 2-D angiographic images taken from different viewing angles are combined to generate a model that includes three-dimensional (3-D) data, e.g., from 2, 3, 4, or more viewing angles.

[0250] In some embodiments, the data is from a computed tomography (CT) scan. It is noted that with today's technology, angiographic images provide finer resolution than CT scans. A model of the vasculature constructed based on angiographic images, whether a one-dimensional (1-D) tree model or a full 3-D model, is potentially more accurate than a model based on a CT scan, and potentially provides a more accurate vascular assessment.

[0251] Speed ​​of Results In some embodiments of the invention related to real-time use, the aim is the fast computation of a vascular model and its anatomical and / or functional parameters, thereby providing feedback for real-time diagnostic decision making.

[0252] In some embodiments of the invention, the feedback regarding making a decision to intervene (e.g., in a particular area or all) can be divided into three broad categories: "Suggest to Intervene," "Suggest not to Intervene," and "No Suggestion." Optionally, the feedback is presented in such a format. Optionally, the classification itself is performed by the physician based on the provided index, which may be a graph, a categorical description, and / or another output, having multiple output states, a continuous range of output states, or any number of states in between. Additionally, in some embodiments of the invention, the diagnostic feedback may be, for example, a FFR. pressure , and / or can be readily correlated to clinical outcomes by generating indices that are readily related to (and potentially interchangeable with) indices already established in the field, such as scoring methods such as the SYNTAX score, or other methods of vascular assessment.

[0253] In some embodiments of the invention, vascular tree construction is optimized to generate vascular segment paths, e.g., vascular segment centerlines. From this stage (or from the outcome of another process that generates a vascular tree where the location of vascular extents is easily determined), calculations for the determination of one or more diagnostically significant indices are potentially very fast, as long as appropriate index targets are sought.

[0254] With regard to the selection of an appropriate index target, another objective in some embodiments of the present invention related to real-time use is the use of flow parameters that can be calculated very quickly given a vascular tree while still producing a diagnostic index that is sufficiently accurate to be usable as a clinical decision-making tool. One aid in obtaining such an index is, in some embodiments, the availability of a deep vascular tree (three, four or more branches), whereby the resistance to flow throughout a large expanse of the vascular network can be calculated in terms of its effect on flow through a particular segment in both stenosed (narrowed) and non-stenosed (widened) conditions. In some embodiments, a well-defined vascular tree, constructed as described herein for X-ray angiography images, but also as potentially available from other imaging methods such as rotational angiography and / or CT angiography, is used as input to the image-based FFR calculation. "Well-defined" comprises, for example, having a branching depth of three, four or more vascular branches. Additionally or alternatively, "well-defined" may comprise imaging resolution sufficient to model vascular width with an accuracy within, for example, 5%, 10%, 15%, or another greater, lesser, or intermediate value of the true vascular width.

[0255] In some embodiments of the invention, the selection of image analysis methods is guided by focusing on generating treatment suggestions, which makes it easier to provide rapid diagnostic feedback. In particular, in some specific embodiments, the goal is to perform an analysis of whether a particular revascularization intervention restores clinically meaningful blood flow. In creating a vascular model, it is a potential advantage to focus on modeling measurable parameters that target changes in clinical interventions, since it is through these changes that the effect of the proposed treatment (if any) will be felt. Moreover, such a focus allows for simplifying and / or ignoring unaltered and / or equivalent parameters, where appropriate, at least to the extent that they do not affect the desirability of the treatment outcome. Thus, for example, potentially no modeling of dynamic flow is required to arrive at a diagnostic index of vascular function.

[0256] In some embodiments of the invention, analysis sufficient to generate useful suggestions for PCI and / or CABG (coronary artery bypass grafting) comprises analysis of one or more features that are easily determined as a local function of 1-D parameters such as vessel segment location. For example, vascular resistance is affected by many variables that are potentially treatable by careful consideration of the fluid dynamics of the system, but has a strong dependency on the variable of vessel diameter. Vessel diameter is then the target of treatment options such as stent implantation. Furthermore, vessel diameter itself (and / or related metrics such as vessel radius, cross-sectional area, and / or cross-sectional profile) can be rapidly calculated from image data along a path that comprises a description of the vessel segment location.

[0257] Furthermore, a potential advantage of a vascular model that is optimized for centerline determination is that, for example, calculations related to details of less clinical significance, e.g., vascular wall shape, can be avoided and / or postponed. In some embodiments, the vascular centerlines constitute the central framework of the final model. It is a potential advantage to use the same vascular centerlines (and / or their approximations) as landmark features in one phase of the process of building the 3-D coordinate system in which the 2-D images in which the 3-D vascular tree is modeled are registered. Potentially, this eliminates the need for determining a second set of features. Potentially, using the same features set for both registration and model basis avoids some calculations, and asymmetries between registration features and model features are thereby reduced, thus eliminating mismatches due to image artifacts.

[0258] In some embodiments of the invention, using relatively modest computational resources (e.g., a PC with a commercially available multi-core CPU and four mid-range GPU cards - equivalent to about 8-12 teraflops of raw computational power), the complete processing period from receipt of images to availability of diagnostically useful metrics such as FFR comprises about 2-5 minutes. On a 5 minute time scale, using this type of equipment, in some embodiments, centerline segmentation of about 200 input images comprises about 0.5 minutes of processing time, about 4 minutes for conversion to a 3-D model, and about 10-30 seconds for the remaining tasks such as computing the FFR, depending on the extent of the tree being computed. It should be noted that further reductions in processing time are expected as long as the general processing power cost per teraflop continues to decrease. Furthermore, the division of processing tasks to multiple processors and / or multiple cores can of course be achieved by partitioning along the vessel boundaries, e.g., dividing the work into several processing resources based on spatial location. In some embodiments of the invention, the calculations to reconstruct the vessel tree and compute the flow index comprise less than about 100 quadrillion operations. In some embodiments, the computation comprises less than about 5 quadrillion, 2 quadrillion, 1 quadrillion, 500 quadrillion, or an intermediate, greater, or lesser number of operations.

[0259] Another objective in some embodiments of the invention is the integration of automatic vascular parameter determination from images into the clinical workflow. In some embodiments, this integration is interactive in that it comprises an interaction between the results and / or control of the automatic imaging process and other aspects of the catheter insertion procedure as the procedure is in progress. For example, in some embodiments of the invention, a medical professional may determine from a cursory manual inspection that one of two branches of a vascular tree is a likely first candidate for a vascular intervention such as PCI. In some embodiments of the invention, the first candidate branch may be selected such that, for example, the process of determining an FFR index for that branch is completed sooner than the calculation for the second branch. Potentially, this allows decision making to be done earlier and / or with less interruption in the procedure being performed on the patient.

[0260] In some embodiments of the invention, the tree processing is fast enough that two, three, or more imaging modalities may be performed and analyzed within the course of a single session with a patient. A single session may comprise a period during which, for example, a portion of a catheter and / or guidewire remains within a portion of the vascular tree for an intervention to, for example, open a stenosis therein, which may be, for example, 30 minutes to an hour, or shorter, longer, or intermediate times. FFR pressure For example, FFR is typically determined in conjunction with an injection of adenosine to maximally open the patient's vasculature. However, the safe frequency of adenosine injections is limited, and thus a method for determining an index equivalent to FFR without such injections offers potential advantages. A second imaging session is potentially valuable, for example, to verify the results of stent implantation, which is typically performed at the level of positioning verification for current stent implantation. Potentially, vascular autoregulation after stent implantation may result in changes in vessel width, whereby a second imaging session can help determine whether further stent implantation becomes and / or remains adviseable.

[0261] In some embodiments, the results of the intensive phase of calculations may be used as the basis for recalculation based on further acquired images and / or for recalculation of the indices, for example, an already calculated vascular tree may be used as the basis for registration of one or more images of the post-implant vasculature without the need to reacquire a complete image set.

[0262] In some embodiments of the invention, a user interface to a computer is provided, e.g., a graphical user interface in which one or more interactive user commands are supported. Optionally, for example, one or more user commands are available to focus image processing targets on one or more selected branches of the subject's vasculature. Optionally, one or more commands are available to modify an aspect of a vascular model (e.g., to model a non-stenosed state of a stenosed vessel). Optionally, one or more commands are available to select and / or compare vascular models from multiple image sets (e.g., image sets taken at distinct times during a procedure and / or image sets with views of the heart at different cardiac cycle phases).

[0263] Characteristics of Some Exemplary Vascular Models In some embodiments of the invention, the vascular system model comprises a tree model, optionally a 3-D tree model. However, the spatial dimensions of the model are optionally adjusted at different anatomical levels and / or processing stages to suit the requirements of the application. For example, 2-D images are optionally combined to extract 3-D vascular tree information that allows identification and construction of 1-D vascular segment models. The models of the 1-D segments are then logically linked, in some embodiments, according to their connectivity, with or without preserving details of other spatial relationships. In some embodiments, the spatial information is compressed or encoded, for example, by approximating the cross-sectional area by parameters of a circle (diameter), an ellipse (major / minor axis), or other representation. In some embodiments, the area along the vascular tree comprises non-spatial information, for example, flow resistance, calculated flow rate, elasticity, and / or sampled and / or expanded vascular segment area, and / or another dynamic or static property associated with the nodes of the vascular tree.

[0264] In some embodiments, the tree model comprises a tree data structure having nodes linked by curved segments. The nodes are associated with vascular bifurcations (e.g., bifurcations or trifurcations or multifurcations) and the curved segments are associated with vascular segments. The curved segments of the tree are also referred to below as branches and the entire part of the tree distal to the branches is referred to as the crown. Thus, in some embodiments of the invention, the tree model comprises a description of the vasculature that assigns the nodes of the tree to vascular bifurcations and the branches of the tree to vascular segments of the vasculature.

[0265] In some embodiments, sample points along the branches are associated with vascular diameter information. In such embodiments, the tree may be thought of as being represented as a series of disks or poker chips (e.g., circular or oval disks) that are linked together to form a 3-D structure that contains information related to the local size, shape, branching, and other structural features at any point in the vascular tree.

[0266] In some embodiments, trifurcations and / or multifurcations are systematically converted into bifurcations combinations. Where appropriate, for example, a trifurcations is converted into two bifurcations. The term "bifurcations" in all grammatical forms is used throughout this specification and is intended to mean bifurcated, trifurcated, or multifurcated.

[0267] In some embodiments, the tree model includes characteristic data associated with sample points along each branch in the model and / or aggregated for the entire branch and / or for an extension thereof. The characteristic data may include, for example, the placement, orientation, cross section, radius, and / or diameter of the vessel. In some embodiments, the tree model comprises flow characteristics at one or more of these points.

[0268] In some embodiments, the tree model comprises geometric data measured along vascular centerlines of the vasculature.

[0269] In some embodiments, the vasculature model comprises a 3-D model, for example, a 3-D model constructed from a series of 2-D angiographic images taken from different angles, such as may be obtained from a CT scan.

[0270] In some embodiments, the vasculature model comprises 1-D modeling of vessel segments along the centerline of a set of vessels in the vasculature.

[0271] In some embodiments, the tree model of the vascular system comprises data relating to a segment represented in 1-D describing the division of the segment into two or more segments.

[0272] In some embodiments, the model includes three-dimensional data associated with a 1-D collection of points, e.g., a collection of data along a segment of a vessel including data regarding the cross-sectional area at each point, data regarding the 3-D orientation of the segment, and / or data regarding the angle of bifurcations.

[0273] In some embodiments, a model of the vascular system is used to calculate a physical model of fluid flow, including physical properties such as pressure, flow rate, flow resistance, shear stress, and / or flow velocity.

[0274] It should be noted that performing calculations on a 1-D collection of points, such as calculating resistance to fluid flow, is potentially much more efficient than performing such calculations using a full 3-D model that includes all voxels of the vasculature.

[0275] Calculation of the blood vessel model Reference is now made to FIG. 13, which is a flow chart outlining an exemplary overview of stages in constructing a vascular model, according to certain exemplary embodiments of the present invention.

[0276] FIG. 13 is used as an overview of an exemplary vascular tree reconstruction method, which is first introduced as an overview and then described in more detail below.

[0277] At block 10, in some embodiments, images are acquired, for example, about 200 images, split across, for example, four imaging devices. In some embodiments, the acquired images are obtained by X-ray angiography. Potential advantages of using X-ray angiography include that, due to the current state of the art, devices for stereo X-ray angiography are commonly available in catheterization labs where diagnostic and interventional procedures are performed. X-ray angiography images also potentially have a relatively high resolution compared to alternative imaging methods such as CT.

[0278] In block 20, in some embodiments, vascular centerlines are extracted. Vascular centerlines have several properties that make them useful references for other stages of vascular tree reconstruction. Properties utilized in some embodiments of the invention include, as appropriate:

[0279] The centerline is a feature that can be determined from the 2-D image, which can then be used to relate separate images to each other in 3-D.

[0280] Vascular centerlines, by definition, are distributed throughout the imaged region of interest when the goal is to reconstruct a 3-D vascular model, and therefore they serve as attractive candidates for fiducial points within the reconstructed imaged region.

[0281] Vascular centerlines can be determined automatically, without prior human selection, based on image characteristics that are easily segmented, for example, as described below.

[0282] Vascular centerlines are extended features that preserve sufficient similarity between images, even images taken from different perspectives, such that their homology is easily identifiable, for example, in the 2-D image itself and / or by backprojection along rays into 3-D space, and ray intersections (and / or intersections between extended volumes based on backprojected rays) identify homologous targets found in different image projections.

[0283] The spatial ordering of samples along the centerline is preserved between images, even though the centerline itself may be distorted by viewing angle and / or motion artifacts, which facilitates comparisons used, for example, for 3-D reconstruction.

[0284] The centerline provides a convenient frame of reference for organizing and / or analyzing features related to position along the vessel. For example, by using distance along the centerline as a reference, morphological features such as diameter and / or functional features such as flow resistance can be expressed as functions in a simplified 1-D space.

[0285] Intersections of centerlines provide a convenient means for describing vascular branching points and / or dividing the vascular tree into different segments which can optionally be treated separately from one another and / or further simplified for the purposes of, for example, functional analysis of flow characteristics.

[0286] Additionally or alternatively, in some embodiments, another type of image feature is identified. Optionally, the image feature is, for example, a vascular bifurcation, a location of a minimum radius (a radius that is locally reduced compared to the surrounding vascular region) in a stenosed vessel. Optionally, the image feature is any configuration of image pixels that has a pattern of intensity that generally lacks identity in any translation direction (below a certain threshold of identity, for example always above a threshold of intensity difference, or always within a criterion for statistically insignificant identity), for example, a bend or bifurcation such as a corner.

[0287] At block 30, in some embodiments, correspondences are found between the extracted vascular centerlines in the individual 2-D images. These correspondences more generally indicate the relationship between the 2-D images. Additionally or alternatively, other features that are generally identifiable in the multiple 2-D images are the basis for finding the correspondences. Such correspondences are generally not uniquely revealed by a transformation determined a priori from calibration information related to the imaging system and / or the patient. A potential advantage of using centerlines to find the correspondences is that the feature of most interest in the vascular image (the vessel itself) is the basis for the determination.

[0288] In some embodiments of the invention, a surface corresponding to the shape of the subject's heart is defined, for example, by determining the projection of the heart surface onto different 2-D image planes using the pattern of cardiac vessels and calculating the shell volume therefrom. Optionally, this surface is used as a constraint for the detection of corresponding image features. In some embodiments, image data constraints and / or other sources of additional information are used in the process of reconstructing the vascular tree. For example, one or more knowledge-based (atlas-based) constraints may be applied, for example, by restricting the recognized vessel positions to those that are within a range of expected vessel positions and / or branch configurations. Also, for example, temporal information is available in some embodiments of the invention based on the filling times of positions along the vascular tree. The filling times are used in some embodiments, for example, to determine and constrain relative vessel positions (positions along the extent of the vascular tree). In some embodiments, the filling times are also used to establish homology between vessel features in different 2-D images (the same filling times should be seen at the vantage points of all images of a homologous arrangement). Additionally or alternatively, filling times are used in some embodiments of the invention to constrain vascular topology.

[0289] At block 40, in some embodiments, the vascular centerlines are mapped to a 3-D coordinate system. In some embodiments, the mapping comprises identifying pairs of homologous centerline locations in different 2-D images that best satisfy a set of optimization criteria, such as consistency with epipolar geometry constraints and / or consistency with vascular points whose 3-D locations have already been determined.

[0290] At block 50, in some embodiments, the vessel diameter is estimated. In some embodiments, the vessel diameter is calculated over sample points of the selected 2-D projection and extrapolated to the entire circumference of the vessel. In some embodiments, the diameter over multiple projection angles is determined. In some embodiments, the projection is selected from a single acquired image, optionally an image in which the vessel is seen at its longest point and / or without any intersections. Optionally, the projection is synthesized from two or more 2-D images.

[0291] Application of vascular tree The computational procedure of this embodiment potentially requires scaled-down computations with respect to conventional techniques employing computational fluid dynamics simulation and analysis. It is recognized that computational fluid dynamics requires substantial computational power and / or time. For example, if a fluid dynamics simulation is performed on a standard PC, it takes several days of CPU time. This time can be somewhat reduced by using a supercomputer that applies parallel processing, but such computational platforms are not generally available for such dedicated use in medical facilities. The computational procedure of this embodiment is not based on a fluid dynamics simulation and therefore can be implemented on a computing platform based on ordinary off-the-shelf components, for example configured as a standard PC, without the need for a supercomputer.

[0292] The present inventors have found that a tree model according to some embodiments of the present invention can be constructed in less than 60 minutes, or less than 50 minutes, or less than 40 minutes, or less than 30 minutes, or less than 20 minutes, or less than 5 minutes, or less than 2 minutes from the time the 2-D image is received by the computer. This time is potentially dependent on the computational resources available, but the inventors have found that run times of 2 to 5 minutes are easily achievable with ordinary commercially available computing hardware (e.g., available with total computing power in the range of about 8 to 12 teraflops).

[0293] This allows for an efficient combination of computation and therapy in some embodiments of the invention, where the tree model is generated conveniently while the subject is immobilized on a treatment surface (e.g., a bed) for purposes of catheter insertion. In some embodiments of the invention, the tree model is generated while the subject has a catheter in his or her vasculature. In some embodiments of the invention, the vasculature has at least one catheter other than an angiography catheter (e.g., a cardiac catheter or an intracranial catheter), where images are processed and the tree is generated while the catheter is in the vasculature.

[0294] Use of the calculated vascular tree is contemplated in the context of further processing and / or decision making in a clinical setting. A potential advantage of a method and / or system for rapid determination of the vascular tree lies in its usefulness in "real-time" applications that allow automated assisted diagnostic and / or therapeutic decisions to be made while the patient being imaged is immediately available for further procedures, perhaps even while still on the catheter insertion table.

[0295] Examples of such real-time applications include blood flow determination and / or vascular status scoring.

[0296] FFR In some embodiments of the present invention, the model calculated from the original imaging data is treated as a "stenosis model", so named because it potentially reflects the location of stenoses in the patient's vascular (cardiovascular) system. In some embodiments, this stenosis model is used to calculate an index indicative of vascular function. The index may also indicate the need for revascularization. Representative examples of indices suitable for embodiments of the present invention include, without limitation, FFR.

[0297] In some embodiments, the index is calculated based on the volume of the crown or other vascular parameters in the stenosis model and the contribution of the stenotic vessel to the resistance to blood flow in the crown. In some embodiments, the FFR index is calculated as the ratio of the flow resistance of a stenotic vessel in a vascular model including the stenotic vessel to the flow resistance of an expanded version of the same vessel in a similar vascular model where the stenotic vessel has been mathematically expanded.

[0298] In some embodiments, the index is calculated as the ratio of the flow resistance of a stenosed vessel in the vascular model to the flow resistance of an adjacent similar healthy vessel in the vascular model, in some embodiments, this ratio is multiplied by a constant that accounts for the different geometries of the stenosed vessel and adjacent vessels, as described in the section "Generating a Model of the Physical Properties of the Vascular System" below.

[0299] In some embodiments, a first tree model of the vasculature is generated based on actual patient measurements and optionally includes stenoses at one or more locations of the patient's blood vessels, a second tree model of the patient's vasculature is generated and optionally modified so that at least one of the stenosis locations is modeled as if after revascularization, and an index indicative of the need for revascularization is generated based on comparing physical characteristics in the first model with physical characteristics in the second model.

[0300] In some embodiments, actual pressure and / or flow measurements are used to calculate the physical properties of the model(s) and / or the indices mentioned above.

[0301] In some embodiments, actual pressure and / or flow measurements are not used to calculate the physical properties of the model(s) and / or the indices mentioned above.

[0302] It should be noted that the resolution of angiographic images is typically higher than that typically obtained by 3-D techniques such as CT. According to some embodiments, models constructed from higher resolution angiographic images are inherently of higher resolution, resulting in greater geometric accuracy and / or allowing the use of smaller geometric characteristics of the vessel than CT images and / or allowing calculations using more generation or branching downstream from the stenosis versus vascular branches distal to the stenosis compared to CT images.

[0303] Vascular Status Scoring In some embodiments of the invention, the automated determination of parameters based on the vascular images is used to calculate a vascular disease score, hi some embodiments, the imaged vessels are cardiac vessels.

[0304] In some embodiments of the present invention, the cardiac score is calculated according to the SYNTAX score calculation method. In some embodiments, the cardiac score is calculated by an alternative, derivative, and / or successor vascular status scoring tool (VSST) of the SYNTAX score. Alternative VSST approaches potentially include, for example, a "functional SYNTAX score" (integrating physiological measurements such as blood flow capacity, vascular elasticity, vascular autoregulatory capacity, and / or other measures of vascular function with a SYNTAX score-like tool) or a "clinical SYNTAX score" (integrating clinical variables such as patient history, and / or systemic and / or organ-specific laboratory results with a SYNTAX score-like tool). Examples also include the AHA classification of the coronary tree segments modified for the ARTS study, the Leaman score, the ACC / AHA lesion classification system, the total occlusion classification system, and / or the Duke and ICPS classification system for bifurcation lesions.

[0305] In some embodiments, two-dimensional images from an angiography procedure are converted to three-dimensional images, and intravascular lesions are identified and input as VSST parameters to arrive at a fast objective SYNTAX score during the procedure. In some embodiments, the VSST parameters are determined directly from the two-dimensional images. Thus, for example, a 2-D image with a determined spatial relationship to a 3-D vascular model (and optionally with vascular segments identified therein) is analyzed for vascular geometry characteristics that are then linked to positions in the vascular model (and optionally with vascular segments identified therein).

[0306] In some embodiments of the invention, automatically determined values ​​are provided as parameters to the VSST, such as the SYNTAX score, in real time during the catheter insertion procedure or after imaging.

[0307] Potentially, a reduction in the time to calculate the VSST has benefits by allowing patients to remain catheterized for a shorter period of time while waiting for a possible PCI (percutaneous coronary intervention) treatment, and / or by reducing the need for recatheterization of patients who have been temporarily released from the operating room pending a treatment decision. Potentially, a reduction in the time and / or labor of scoring could lead to increased use of the VSST, such as the SYNTAX score, as a tool for clinical decision making.

[0308] Generating a geometric model of the vascular system Image acquisition Reference is now made to Figure 14, which is a flow diagram outlining an exemplary overview of the details of the stages in constructing a vascular model, according to some exemplary embodiments of the present invention, and which are also described in additional figures referenced below in the course of executing the blocks of Figure 14 in sequence.

[0309] At block 10, multiple 2-D data images are acquired. In some embodiments of the invention, data images are acquired simultaneously from multiple vantage points, e.g., 2, 3, 4, or more vantage points (cameras) for imaging. In some embodiments, images are acquired at a frame rate of, e.g., 15 Hz, 30 Hz, or another lower, higher, or intermediate frame rate. In some embodiments, the number of frames acquired per vantage point for imaging is about 50 frames (a total of 200 frames for four vantage points for imaging). In some embodiments, the number of frames per vantage point for imaging is, e.g., 10, 20, 40, 50, 60, 100, or another larger, smaller, or intermediate number. In some embodiments of the invention, the number of cardiac cycles included in an imaging period is about 3-4 cardiac cycles. In some embodiments, the number of cardiac cycles is, for example, 3-4, 3-5, 4-6, 5-10, or another range of cardiac cycles having the same, smaller, larger, or intermediate range boundaries.

[0310] Reference is now made to FIG. 1, which illustrates an original image 110 and a Franzi filtered image 120 that are processed according to some exemplary embodiments of the present invention.

[0311] Original image 110 shows a typical angiographic 2-D image.

[0312] It should be noted that when using two or more 2-D projections of a subject's vessels, e.g., cardiac vessels, it is a potential advantage that the two or more 2-D projections are performed simultaneously, or at least at the same phase within the cardiac cycle, so that the 2-D projections are of the same vessel geometry.

[0313] Deviations between the 2-D projections can result from cardiac, and / or respiratory, and / or patient motion between the 2-D projection frames.

[0314] In some embodiments, the ECG output is used to select the same cardiac phase within the 2-D projection frame to reduce deviations that may result from lack of cardiac phase synchrony.

[0315] In some embodiments, to reduce deviations that may result from lack of cardiac phase synchronization, an ECG output or another cardiac / pulse synchronization means, such as a visible light pulse monitor, is used to select the same cardiac phase in the 2-D projection frame. Optionally, the cardiac synchronization output is recorded with a time scale, and a corresponding time scale is used to record when the image of the vasculature is captured. In some embodiments of the invention, the time of acquisition relative to the cycle of the physiological dynamics is used to determine candidates for co-registration. For example, image registration is optionally performed using image datasets comprising images taken at nearby phases of the cardiac cycle. In some embodiments, registration is performed multiple times across different sets of adjacent datasets, so that the registered landmarks are iteratively transformed in position to a common 3-D coordinate system.

[0316] In some embodiments, the 2-D projection frame is selected to be at the end of the diastole of the cardiac cycle. In some embodiments, the temporal and / or phase order in which the 2-D projection frames are acquired is used to perform registration between images captured in adjacent motion cycle phases. In some embodiments, images registered from a first phase to a second phase are then re-registered to a third and / or further phases, such that images captured at widely separated cardiac cycle phases may be registered to each other.

[0317] In some embodiments, the heart is imaged under the influence of intravenous adenosine, which potentially exaggerates the differences between normal and abnormal segments. Optionally, imaging with and without adenosine potentially allows for the determination of the (vasodilatory) effect of adenosine itself, which in turn provides information regarding vascular compliance and / or autoregulatory status.

[0318] Extracting centerlines Reference is now made to FIG. 15, which illustrates a schematic diagram of an exemplary imaging coordinate arrangement 1500 for an imaging system, according to certain exemplary embodiments of the present invention.

[0319] A number of different spatial relationships of the imaging arrangements are used in determining the 3-D relationships of the image data within the 2-D image set.

[0320] In some embodiments, the image coordinate systems 1510, 1520 and associated image planes 1525, 1530 describe how images taken of the same subject in different positions relate to each other, and that information is used to reconstruct 3-D information about the subject. In some embodiments of the invention, these coordinates reflect the axis of rotation of the C-arm of the angiographic imaging device. In some embodiments, the coordinate plane 1515 of the subject (e.g., lying on a bed 1505) is also used as part of the 3-D reconstruction.

[0321] This system configuration information is typically documented in the DICOM (image) file and / or elsewhere, but is not guaranteed to reflect the actual positions and orientations of the system components with sufficient accuracy and / or precision for a useful reconstruction of the coronary artery tree. In particular, the bed axis potentially does not align with the room coordinate system, the axis of C-arm rotation potentially does not intersect with the isocenter and / or is non-orthogonal, and / or the detector axes potentially do not align in-plane.

[0322] In block 20 - returning to FIG. 14 - the centerline of the vascular tree is extracted from the acquired 2-D image. In some embodiments of the invention, image filtering by anisotropic diffusion 21 comprises part of the processing operations preceding centerline extraction. Anisotropic diffusion of 2d grayscale images reduces image noise while preserving region edges - smoothing along image edges and removing noise gaps. In some embodiments, the basis of the method used is similar to that introduced by Weickert in "A Scheme for Coherence-Enhancing Diffusion Filtering with Optimized Rotation Invariance" and / or "Anisotropic Diffusion in Image Processing" (Thesis 1996).

[0323] Reference is now made to FIG. 16, which is a simplified flow diagram of processing operations involving anisotropic diffusion, according to certain exemplary embodiments of the present invention.

[0324] The operations are described for a single image for some embodiments of the invention. In block 21A, the Hessian transform is calculated from all pixels of the Gaussian smoothed input image (the Hessian transform involves the second derivative of the image data and is a form of edge detection). In block 21B, the Hessian transformed image is smoothed, for example by a Gaussian filter. In block 21C, the eigenvectors and eigenvalues ​​of the smoothed Hessian transformed image are calculated. The resulting eigenvalues ​​are generally larger when the original image contains edges, and the eigenvectors corresponding to the larger eigenvalues ​​describe the direction in which the edge runs. Additionally or alternatively, other edge detection methods are used as known in the art.

[0325] In block 21D, in some embodiments of the invention, a diffusion image is calculated. A finite difference scheme is used to perform the diffusion, in which, in some embodiments, eigenvectors are used as the diffusion tensor directions.

[0326] In block 21E, in some embodiments, a determination is made whether a diffusion time limit (e.g., a particular number of iterations that results in a desired level of image filtering) has been reached. If not, the flow chart returns to block 21A and continues. If so, the flow chart ends and flow continues within a higher level flow chart, such as the flow chart of FIG. 14.

[0327] In block 22, a Frangi filter is applied based on the eigenvectors of the Hessian matrix, which in some embodiments of the present invention comprises calculating the probability that an image region is within a vessel. Frangi filtering is described, for example, by Frangi et al., "Multiscale vessel enhancement filtering," Medical Image Computing and Computer-Assisted Intervention-MICCA'98. By way of non-limiting example, Frangi filtered image 120 (FIG. 1) shows the original image 110 after image processing using the Frangi filter. In some embodiments, another filter, for example a threshold filter or a hysteresis threshold filter, is used, whereby pixels of the image are identified as belonging to an image region of a vessel.

[0328] At block 23, in some embodiments of the present invention, the image is processed to generate black and white shapes representing vascular placement within the angiographic projection image. In some embodiments, a hysteresis threshold filter is run on the Franzi filter output with a high and low threshold. First, the algorithm indicates pixels that are brighter (e.g., with respect to image 120) than the higher threshold, and these are labeled as vascular pixels. In a second step, the algorithm labels as vascular those pixels that have a brightness higher than the lower threshold and that are also connected across the image to pixels already labeled as vascular pixels.

[0329] A potential disadvantage of the Frangi filter is the presence of light bulb-like shapes at vascular junctions that interfere with accurate detection. In some embodiments, a region growing algorithm is used to extract these regions as an improvement over hysteresis thresholding alone. Thresholded black and white and grayscale images obtained by anisotropic diffusion provide the input to this algorithm.

[0330] A square dilation is performed on the black and white image and the result is subtracted from the original black and white image. This subtraction image comprises a one pixel wide frame along which the growth of the area of ​​vessel labeling pixels is then examined. The values ​​(brightness) of pixels in this frame are locally compared to the brightness of existing vessel pixels and to the surroundings. High relative results lead to dilation. Optionally, this process is repeated until no vessel pixels are found.

[0331] At block 24, in some embodiments, a thinning convolution is applied, which thins the black and white image segments down to straight lines representing vascular centerlines.

[0332] In some embodiments, blocks 21-24 are performed image by image (e.g., sequentially, interleaved, and / or in parallel). If there are more images to process, assuming sequential processing in block 25, the next image is selected in block 26 and processing continues again from block 21.

[0333] If not, centerline extraction is complete in some embodiments of the present invention. Reference is now made to Figure 2, which shows a 2-D tree 218 with light-colored vascular centerlines overlaid on the original image 110 of Figure 1, in accordance with an exemplary embodiment of the present invention.

[0334] Motion Compensation In some embodiments of the invention, processing for finding centerline correspondence follows at block 30 (FIG. 14). The goal of finding centerline correspondence is to find correspondence between different 2-D images (potentially from different angles, but imaging the same region of space) so that a 3-D reconstruction of the target vasculature can be made.

[0335] In block 31, operations for motion compensation and / or imaging position artifact compensation are performed.

[0336] With ideal calibration information (e.g., each image plane is perfectly identified with respect to a common coordinate axis) and no artifacts due to motion or other positioning errors, backprojecting a sufficient number of 2-D images into 3-D space potentially yields intersecting rays (e.g., rays S1-P1 and S2-P2 in FIG. 20B) that uniquely define the extent of the vascular centerline in 3-D. In fact, deviations between images come from, for example, breathing, voluntary patient motion, and imprecise and / or unclear phase locking of the imaging exposure to the cardiac cycle. Eliminating this problem without increasing computational complexity is a goal in some embodiments of the operation for motion compensation. Calibration errors potentially introduce other forms of image position artifacts.

[0337] In some embodiments of the invention, this procedure compensates for respiration and / or other patient movements. Optionally, this comprises iteratively repeating the detection of corresponding image features, each time for a different subset of angiographic images, and updating image correction parameters in response to the repeated detection.

[0338] Reference is now made to FIG. 17A, which is a simplified flow diagram of processing operations including motion compensation, according to some exemplary embodiments of the present invention.

[0339] In block 31A, in some embodiments of the invention, a subset (comprising a plurality) of images for which a 2-D centerline has been identified is selected for processing. The centerline is optionally expanded in block 31B, and a centerline backprojection to 3-D is performed in block 31C based on the currently best known projection parameters for each image (initially these are expected parameters based on, for example, the known configuration of the imaging device). The resulting projected volume is in some embodiments skeletonized to form a "consensus centerline" in block 31D. In block 31E, the consensus centerline is backprojected into the coordinate system of the 2-D images that comprise those that were not used in forming the consensus centerline. In block 31F, the projection parameters for the 3-D centerline are adjusted in each 2-D image to more accurately fit the centerline found in the image itself. This adjustment is used to adjust the projection parameters associated with each image. In 31G, in some embodiments, a decision is made to repeat this procedure for a different image subset in order to improve the overall quality of the projection fit. The decision to repeat may be based, for example, on a predetermined number of iterations, a metric measurement quality of the fit (such as the average distance between the closest points in the centerline projection), or another criterion. If so, the flow chart returns to block 31A and continues. If not, the flow chart ends and the flow continues within a higher level operational ordering, for example, the ordering of FIG. 14.

[0340] It has been found by the present inventors that such an iterative process can significantly reduce one or more of the effects of respiration, patient movement, and cardiac phase differences.

[0341] Reference is now made to FIG. 17B, which is a simplified flow diagram of processing operations involving alternative or additional methods of motion compensation, according to some exemplary embodiments of the present invention.

[0342] In some embodiments of the invention, features are identified in the reference image R at block 31H based on feature detection methods known in the art. Such image features are, for example, vascular bifurcations and origins of the coronary vascular tree, locations of minimum radii in stenotic vessels, and / or configurations of image pixels having intensity patterns that generally lack self-identity in translation in any direction, such as bends or bifurcations that resemble corners. Similar features (presumed to be homologous to features in the reference image) are identified in the remaining image F.

[0343] In some embodiments of the invention, in block 31I, the image in F is then registered to image R. For example, the best known projection parameters of image F are used to transform into the best known projection plane of image R, and then optimized to obtain an improved fit, for example using epipolar geometry to calculate shift, rotation, and / or scaling parameters. Optionally, the registration comprises application of a geometric distortion function. The distortion function is, for example, a linear, quadratic, or other order function of the two image plane coordinates. Additionally or alternatively, the distortion function comprises parameters describing the adjustment of nodal points defined in the image coordinate plane to cause the registration to occur. In some embodiments, the best known projection parameters are the same and the geometric distortion function is applied.

[0344] In some embodiments, the operation in block 31I comprises image correction parameter calculation based on the identified corresponding image features. The correction parameters typically describe, for example, a translation and / or rotation of a coordinate system of a particular image. Based on the calculated parameters, the angiographic images are registered to provide a mutual geometric correspondence between them. In some embodiments of the invention, several images are registered with respect to one of the images. For example, if corresponding image features are identified in N images (e.g., N=2, 3, 4 or more), one of the images may be selected as a reference, while registration is applied to the remaining N-1 angiographic images such that each of those remaining images geometrically corresponds to the single angiographic image selected as the reference. In some embodiments, for example, another registration scenario is performed.

[0345] At block 31J, in some embodiments, candidate feature locations of identified features that are expected to be contained within a shell-like volume near the heart surface (particularly vascular features) are filtered based on whether they actually fall within such a volume. Calculation of this shell-like volume is described, for example, with respect to Figures 18A-18C.

[0346] In block 31K, in some embodiments, a decision is made whether to repeat this procedure for a different image subset to improve the overall quality of the projection fit. For example, a decision to repeat is made as described for block 31G. If so, the flow chart returns to block 31H and continues. If not, the flow chart ends and the flow continues within a higher level operational ordering, for example, the ordering of FIG. 14.

[0347] Cardiac surface constraints Reference is now made to FIGS. 18A-18B, which illustrate aspects of the calculation of a “heart shell” constraint to ignore bad ray intersections from the calculated correspondences between images, according to some exemplary embodiments of the present invention.

[0348] Reference is also now made to FIG. 18C, which is a simplified flow diagram of processing operations including constraining pixel correspondence to within a volume near the heart surface, according to certain exemplary embodiments of the present invention.

[0349] In some imaging procedures, a "sufficiently large" number of projections is potentially unavailable, and therefore errors in the determined locations of ray intersections potentially prevent convergence to the correct output. In some embodiments of the invention, at block 32, operations are performed to reduce the effect of this (and / or other) sources of location error based on cardiac surface constraints.

[0350] In block 32A, according to some embodiments of the invention, an image is selected that comprises features expected to be within the projected outline of the heart. In some embodiments, the features are representations of coronary arteries 452 that course on the heart surface. In some embodiments, the previously determined vascular centerlines 451 comprise the identified features. In block 32B, in some embodiments, a convex hull 450 defined by the vascular centerlines 451 is determined. This convex hull represents the shape of the heart as projected into the plane of the selected 2-D image (if it is covered by the identified arterial centerlines) as seen by the naked eye. In block 32C, in some embodiments, a decision is made whether another image should be selected to determine the projection of the heart shell from a different angle. The number of images from which the convex hull of the heart shape is calculated is at least two, thereby allowing for 3-D localization of the heart shell, and optionally more images, optionally all available images, are used to determine the heart shell. If another image is added, the flow continues to block 32A, otherwise the flow continues to block 32D.

[0351] At block 32D, in some embodiments, the 3-D convex hull position (heart shell) is determined from the various available 2-D convex hull projections, for example, by using the best known projection parameters for each 2-D image plane and / or 3-D polyhedron intersection. Such surfaces may be defined using techniques known in the art, including, without limitation, polyhedra stitching, based on the description presented herein. At block 32E, in some embodiments, the heart shell is expanded to a volume where the amount of expansion has been determined, for example, to correspond to an error margin where the "true" vessel region is expected to fall.

[0352] At block 32F, in some embodiments, candidate 3-D locations of vascular centerline points that fall outside the heart shell are eliminated. The flow diagram of Figure 17C ends and flow continues within a higher level operational ordering, for example, with the ordering of Figure 14.

[0353] Homolog Identification Reference is now made to Figures 19A-19D, which illustrate the identification of homology between branch vessels, according to some exemplary embodiments of the present invention.

[0354] Reference is also now made to FIG. 19E, which is a simplified flow diagram of processing operations including identifying homologous regions along vascular bifurcations, according to some exemplary embodiments of the present invention.

[0355] In block 33A, in some embodiments, a base 2-D image is selected for homology determination. The initial base selection is arbitrary. In block 33B, in some embodiments, a vascular centerline in one of the remaining images is projected into the plane of the base image. For example, the exemplary vascular centerline 503 in FIG. 19A is from a base image having a base coordinate system 504. A vascular centerline 501 taken from another image having a different coordinate system 502 is shown as centerline 501B transformed (translated in one direction for clarity in FIG. 19A) into coordinate system 504. In FIG. 19B, the two centerlines are shown superimposed to show their general similarity and some artifactual differences as they diverge.

[0356] At block 33C, in some embodiments, the projected vascular centerline 501B is dynamically expanded (501C), noting that intersections with vascular centerlines of the base image occur first, and continuing, for example, until all homologies have been identified. Dynamic expansion comprises, for example, gradually expanding the centerline, for example, by applying a morphological operator to the pixel values ​​of the image. In some embodiments, another method, for example, a nearest neighbor algorithm, is used to determine the correspondence (in addition to or as an alternative). FIG. 19D shows an example of correspondence between points of the vascular centerline at either end of the minimum distance lines 515 and 510.

[0357] In block 33D, in some embodiments, a decision is made whether another image should be selected for extension to the current base image. If yes, the flow chart continues to block 33B. If no, a decision is made in block 33E whether another base image should be selected. If yes, the flow chart continues to block 33A. If no, the flow chart ends and flow continues within a higher level operational ordering, for example, the ordering of FIG. 14.

[0358] It will be appreciated that the operations in block 30 of finding centerline correspondence (e.g., in sub-blocks 31, 32, 33) are operations that perform a function of finding correspondence between different 2-D images - and more specifically, in some embodiments, between vascular centerlines in the 2-D images - that allow the images to be reconstructed into a 3-D model of the vasculature. It will be appreciated that this function may be performed by modifications of the described method and / or by other methods familiar to those skilled in the art, based on the teachings of the present description. For example, whenever image A is projected, mapped, or otherwise transformed into the coordinate space of image B, in some embodiments, it is possible to reverse those transformations (transform B instead) or transform both into a common coordinate space. Also, for example, features and / or locations near the features named in describing the operations (e.g., vascular boundaries with respect to vascular centerlines) may be used in some embodiments to perform part of the task of finding correspondence. Moreover, the operations used to refine the result as a whole (including intermediate results) are optional in some embodiments, and other operations that additionally or alternatively refine the result (including intermediate results) are potentially determinable by one of ordinary skill in the art working based on the description herein. These examples of modifications are not intended to be exhaustive, but rather to illustrate the breadth of the methods that comprise embodiments of the present invention.

[0359] Considering even more generally the results of the operations described in relation to blocks 20 and 30, progress towards at least two related but separate goals is calculated in some embodiments based on a shared intermediate result - the vascular centerline. The first goal is to find a spatial relationship that relates the acquired 2-D images to a common 3-D imaging region. In principle, many possible reference features can be selected from the 2-D images as the basis for this determination, but using the centerline of the vascular tree as a reference is a potential advantage. In particular, the determination of the vascular tree in this 3-D space is itself a secondary goal, and thus in some embodiments the features for mutual registration of the images are also the features used as the skeleton of the vascular model itself. This is a potential advantage for the speed of calculation by reducing the need for separate determination of features for image registration, and of the vascular features as such. This is a potential advantage for the accuracy, precision, and / or consistency of the resulting vascular model, since the registration between the vascular features is the basis for the transformation of the image data that is reconstructed by those same features.

[0360] 3-D Mapping In some embodiments of the invention, a 3-D mapping of the 2-D centerline is performed, at block 40. In some embodiments, the 3-D mapping begins with the identification of an optimal projection pair, at block 41. If several different images are acquired, there are potentially several different (albeit homologous) projections of each region of the vascular centerline into 3-D space, each based on a different pair of 2-D images.

[0361] Reference is now made to Figure 20A, which illustrates a simplified flow diagram of processing operations including selecting projection pairs along a vascular centerline, in accordance with certain exemplary embodiments of the present invention. Entering at block 41, an initial segment comprising the vascular centerline is selected along with an initial homologous group of centerline points (e.g., points from the end points) along it in different 2-D images.

[0362] In block 41A, in some embodiments of the present invention, a point P1 on the vascular centerline (corresponding to some homologous group of centerline points P) is selected from the first base image. In block 42B, in some embodiments, other points P2...P3 are selected to find the location of the 3-D spatial phase. N is selected from homology group P to pair with P1.

[0363] Reference is now made to FIG. 20B, which is a schematic representation of epipolar determination of 3-D target positions from 2-D image positions and their geometric relationships in space, according to some exemplary embodiments of the present invention.

[0364] A point P1 associated with the image plane 410 is matched with a point P2 and, using the principles of epipolar geometry, a location P 1,2 Briefly, the ray from source S1 through the target region to point P1 lies on a plane 417 that is also determined by exiting and intersecting with S2. The continuation of these intersection lines intersects with plane 412 along epipolar line 415.

[0365] At block 41C, in some embodiments, the points are evaluated for their relative suitability as the best available projection pair to extend the vascular centerline in 3-D space.

[0366] In some embodiments of the present invention, the criterion for the optimal selection of the projection points is the distance of the projected point from its associated epipolar line. Ideally, each point P2...P N is S2...S N419. However, due to imaging position artifacts, such as those described with respect to calibration and / or motion, some error may remain, and thus point Pi, already determined to be homogeneous with P1, falls outside its associated epipolar plane 418 and is therefore a distance 420 away from its associated epipolar line 419. Optionally, the projection point closest to its associated epipolar line for a given homogeneity group is scored as the most suitable projection point for extending the vascular centerline.

[0367] In some embodiments of the present invention, one or more criteria for optimal selection of a projection point relate to the continuity of expansion that the projected point brings from an already determined projected point. For example, a set of points along the vascular centerline 421A, 421B may be used to determine the current direction of expansion 423 and / or the expected distance interval to the next expansion point in 3-D space. In some embodiments, a projection point that more closely matches one or more of these or another geometric criteria is scored as a correspondingly more suitable selection.

[0368] In some embodiments, multiple criteria are weighted together and the selection of the optimal projection pair is made based on the weighted result.

[0369] At block 41D, it is determined whether a different base point within the homology group should be selected. If so, the next base point is selected and further projection and evaluation continues from block 41A. If not, the point having the best (best of the available choices) score is selected for inclusion in the 3-D vessel centerline. The flow diagram of FIG. 20A ends and the flow continues within a higher level operational ordering, for example, the ordering of FIG. 14.

[0370] At block 42, in some embodiments, the current vessel segment centerline is expanded according to the points specified by the identified optimal pair of projections.

[0371] Vascular centerline determination, in some embodiments, continues at 43 where it is determined whether another sample (homology group) should be evaluated for the current vascular centerline. If so, operation continues by selecting the next sample at block 44 and continues re-entering block 41. If not, a determination is made at block 45 whether the last vascular segment centerline has been determined. If not, the next segment is selected at 46 and processing continues with the first sample of that segment at block 41. If so, flow continues to vascular diameter estimation in block 50 in some embodiments.

[0372] Estimation of vessel diameter Reference is now made to FIG. 21, which is a simplified flow diagram of processing operations including generating an edge graph 51 and finding connection paths along an edge graph 52, in accordance with some exemplary embodiments of the present invention.

[0373] Upon entering block 51, in some embodiments, an edge graph is determined. In block 51A, in some embodiments, a 2-D centerline projection is selected that maps to a configuration relative to the configuration of intensity values ​​of the 2-D imaging data. Optionally, the selected projection is one in which the vessel is projected at a maximum length. Optionally, the selected projection is one in which the vessel does not cross another vessel. In some embodiments, the projection is selected according to a sub-region of the 2-D centerline of the vessel segment, for example, to have a maximum length and / or non-crossing property within the sub-region. In some embodiments of the invention, an image from an orthogonal projection (and / or a projection with another defined angular relationship) is selected.

[0374] At block 51B, in some embodiments, a starting vessel width (e.g., radius) is estimated. The starting width is determined, for example, by generating an orthogonal profile to the centerline and selecting the peak of a weighted sum of the first and second derivatives of the image intensity along this profile.

[0375] In block 51C, in some embodiments, orthogonal profiles are created for points along the centerline, e.g., sampled at intervals approximately equal to the vessel starting width. The exact choice of interval is not critical. Using the radius as the interval is generally adequate to obtain sufficient resolution for diameter estimation.

[0376] In block 51D, in some embodiments, the orthogonal profiles for the sampled points are assembled in a rectangular frame somewhat as if a 3-D centerline convolution were straightened out, which puts the orthogonal profiles into a parallel alignment through the centerline.

[0377] Entering block 52, in some embodiments, a connected path along the vessel edge is then found. In block 52A, in some embodiments, a first side (vessel edge) is selected for path tracing. In some embodiments, in block 52B, a path is found along the edge at a distance of approximately the initial radius by minimizing the energy corresponding to a weighted sum of the first and second horizontal derivatives, optionally with the aid of the Dijkstra family of algorithms. If the second side has not been calculated in block 52C, the flow diagram branches to select a second side in block 52D and repeat the operations of 52B.

[0378] In some embodiments, following block 53 of Figure 14, the centerline is reset to the middle of the two vessel walls just determined. In block 55, in some embodiments, it is determined whether this is the last centerline to process. If not, in some embodiments, the next segment is selected in block 56 and processing continues to block 51. If so, in block 54, in some embodiments, a decision is made whether the procedure should be repeated. If so, the first segment is selected for the second iteration and operation continues to block 51. If not, the flow diagram of Figure 14 ends.

[0379] Constructing a segment-node representation of the vascular tree Referring now to FIG. 3A, this is an image 305 of a coronary vascular tree model 310 generated in accordance with an exemplary embodiment of the present invention.

[0380] Reference is now also made to FIG. 3B, which is an image of the coronary vascular tree model 315 of FIG. 3A with branch tags 320 added in accordance with an illustrative embodiment of the present invention.

[0381] It should be noted that tags 320 are just one exemplary method for tracking branches in a tree model.

[0382] In some embodiments of the invention, after reconstruction of a vascular tree model, such as a coronary artery tree, from an angiographic image, the tree model is optionally divided into several branches, where a branch is defined as a section of the vessel between bifurcations (e.g., along a frame of reference established by the vessel centerline). The branches are, for example, numbered according to their occurrence within the tree. Branch points (nodes) are determinable in some embodiments of the invention from points of the skeletal centerline representation that connect in more than two directions.

[0383] In some embodiments of the invention, the branch topology comprises a division of the vascular tree model into distinct branches along the branch structure, hi some embodiments, the branch topology comprises recombination of branches, e.g., by vascular side branches and / or bifurcations.

[0384] Reference is now made to FIG. 3C, which is a simplified diagram of a tree model 330 of a coronary vascular tree generated in accordance with an exemplary embodiment of the present invention.

[0385] For some embodiments of applying vascular tree models to coronary artery diagnostics and / or functional modeling, it is useful to abstract some details of spatial location to simplify the calculation of vascular tree properties (and in the present case, for further illustration).

[0386] In some embodiments, the tree model is represented by a 1-D array, e.g., the 9-branch tree of FIG. 3C is represented by a 9-element array, i.e., a=[0 1 1 2 2 3 3 4 4], which enumerates the tree nodes in breadth-first order.

[0387] In some embodiments, in the reconstruction process, the spatial location and radius of the segments in each branch are sampled at small distances, for example, every 1 mm, or in the range of every 0.1 mm to every 5 mm.

[0388] In some embodiments, the tree branches corresponding to vessel segments between the modeled bifurcations correspond to vessel segments of 1 mm, 5 mm, 10 mm, 20 mm, 50 mm, or even longer lengths.

[0389] In some embodiments, sampling at small distances improves the accuracy of the geometric model of the vessel, thereby improving the accuracy of the flow characteristics calculated based on the geometric measurements.

[0390] In some embodiments, the tree model is a reduced tree restricted to a single segment of a vessel between two successive branches of the vasculature, hi some embodiments, the reduction is to the region of the bifurcation, optionally including a stenosis.

[0391] Measuring flow from time-intensity curves in angiographic sequences. In some embodiments, a physical model of fluid flow within the coronary vascular tree is calculated, including physical properties such as pressure and / or flow rate and / or flow resistance and / or shear stress and / or flow velocity.

[0392] In one exemplary embodiment, techniques based on the analysis of concentration-distance-time curves are used. These techniques work well in pulsatile flow conditions. An exemplary concentration-distance-time curve technique is the concentration-distance curve matching algorithm.

[0393] Using the above technique, the concentration of a contrast agent, such as iodine, at a particular distance along a vessel segment is determined by integrating pixel intensities in the angiogram(s) over the vessel lumen perpendicular to the centerline. An optimal shift in the distance axis between successive concentration-distance curves is found. Blood flow velocity is then calculated by dividing the shift by the time interval between the curves. Several variations of the above technique have been reported in the following publications, the contents of which are incorporated herein by reference: the four above-mentioned papers by Seifalian et al., the above-mentioned paper by Hoffmann et al. entitled "Determination of instantaneous and average blood flow rates from digital angiograms of vessel phantoms using distance-density curves," the above-mentioned paper by Shpilfoygel et al. entitled "Comparison of methods for instantaneous angiographic blood flow measurement," and the paper by Holdsworth et al. entitled "Quantitative angiographic blood flow measurement using pulsed intra-arterial injection."

[0394] Measuring flow using other modalities In some embodiments, flow is calculated from ultrasound measurements. Variations of the above mentioned ultrasound techniques are reported in the above mentioned publications, the contents of which are incorporated herein by reference.The aforementioned paper by Kenji Fusejima, titled "Noninvasive Measurement of Coronary Artery Blood Flow Using Combined Two-Dimensional and Doppler Echocardiography", the paper by Carlo Caiati et al., titled "New Noninvasive Method for Coronary Flow Reserve Assessment : Contrast-Enhanced Transthoracic Second Harmonic Echo Doppler", the paper by Harald Lethena et al., titled "Validation of noninvasive assessment of coronary flow velocity reserve in the right coronary artery-A comparison of transthoracic echocardiographic results with intracoronary Doppler flow wire measurements", the paper by Paolo Vocia et al., titled "Coronary flow: a new asset for the echo lab?", the review paper by Patrick Meimoun et al., titled "Non-invasive assessment of coronary flow and coronary flow reserve by transthoracic Doppler echocardiography: a magic tool for the real world", and the paper by Carlo Caiati et al., titled "Detection, location, and severity assessment of left anterior descending coronary artery stenoses by means of contrast-enhanced transthoracic harmonic echo Doppler.”

[0395] In some embodiments, other modalities are used to measure flow within the coronary vascular tree. Exemplary modalities include MRI flow measurements and SPECT (single photon emission computed tomography), or gamma camera, flow measurements.

[0396] It should be noted that in some embodiments, a blood flow model is constructed based on geometric measurements taken from an image of the vasculature, without the use of flow or pressure measurements.

[0397] It should be noted that in some embodiments, flow measurements are used to validate calculated flow characteristics based on models constructed based on geometric measurements.

[0398] It should be noted that in some embodiments, pressure measurements are used to validate calculated flow characteristics based on models constructed based on geometric measurements.

[0399] An exemplary embodiment for generating a model in which a stenosis is modeled as if it had been revascularized - stenosis dilation In some embodiments, the structure of the diseased vessel is estimated as if the vessel were revascularized relative to a healthy structure, which is referred to as a dilated structure, as if a stenosed vessel were revascularized back to its original normal diameter.

[0400] In some embodiments, techniques are used as described in the following publications, the contents of which are incorporated herein by reference: Tuinenburg et al., entitled "Dedicated bifurcation analysis: basic principles," Tomasello et al., entitled "Quantitative Coronary Angiography in the Interventional Cardiology," and Janssen et al., entitled "New approaches for the assessment of vessel sizes in quantitative (cardio-)vascular X-ray analysis."

[0401] The stenosis dilation procedure is suitably implemented separately for each one of the 2-D projections. In some cases, the stenosis may occur in the area near the bifurcation, and in some cases, the stenosis may occur along the vessel. The stenosis dilation procedure in the two cases will now be described separately.

[0402] If the stenosis is not located in a bifurcation region, it is sufficient to assess the flow rate in the diseased vessel. The coronary vessel segments proximal and distal to the stenosis are relatively disease-free and are referred to as reference segments. The algorithm calculates the coronary vessel edges by interpolating coronary vessel segments that are considered disease-free and located proximal and distal to the region of stenosis with the edges of the region of stenosis, as appropriate. The algorithm reconstructs the reference coronary vessel segments as if they were disease-free, as appropriate.

[0403] In some embodiments, the technique involves calculating the average diameter of the vessel lumen in fiducial segments located upstream and downstream of the lesion.

[0404] When the stenosis is located in the bifurcation region, two exemplary bifurcation models are defined: the T-bifurcation model and the Y-bifurcation model.

[0405] A T-shaped or Y-shaped bifurcation model is detected by analyzing the arterial contours of the three vessel segments connected to the bifurcation, as appropriate. The calculation of the flow model for the expanded healthy vessel diameter is based on calculating as if each of the three segments connected to the bifurcation had a healthy diameter. Such a calculation ensures that both the proximal and distal main (interpolated) reference diameters are based on the arterial diameter outside the bifurcation core.

[0406] The reference diameter function of the branch core is optionally based on the reconstruction of a smooth transition between the proximal and distal vessel diameters, so that the reference diameter of the entire main section can be displayed as one function, consisting of three different reference straight lines linked together.

[0407] Example of generating a model of the physical properties of the vascular system Several exemplary embodiments of methods for generating a model of the physical properties of the vascular system are now described.

[0408] The exemplary vasculature used in the remainder of the following description is the coronary vasculature.

[0409] In some embodiments of the invention, a one-dimensional model of the vascular tree is used to estimate the FFR index in stenosed branches of the vascular tree, thereby estimating the flow in the stenosed branches before and, optionally, after stent implantation.

[0410] In some embodiments of the invention, a one-dimensional model of the vascular tree is used to estimate the FFR index in a stenosed branch of the vascular tree, thereby estimating the flow in the stenosed branch before and, optionally, after stent expansion.

[0411] Based on a maximum peak flow rate of 500 mL / min and an artery diameter of 5 mm, the maximum Reynolds number for the flow is:

number

[0412] The above calculations assume laminar flow, which assumes that blood, for example, is a homogeneous Newtonian fluid. Another assumption that is often made is that flow in a branch vessel is 1-D and occurs across the entire cross-section of the vessel.

[0413] Based on these assumptions, the pressure drop within each segment of the vascular tree is approximated by the Poiseuille equation in a straight tube.

number

[0414] Here, R i is the viscous resistance to flow in the vessel segment. Small losses due to vessel bifurcations, constrictions, and curvatures are added in series as additional resistance according to the Darcy-Weisbach equation as follows:

number

number

[0415] Here, K i is the corresponding loss factor.

[0416] Reference is now made to Figure 4, which is a series of nine diagrams 901-909 illustrating vessel segment radii generated by an exemplary embodiment of the invention along the branches of the coronary vascular tree model 330 shown in Figure 3C as a function of distance along each branch. The relative distance along the vessel segments is plotted in the X direction (horizontal) and the relative radii are plotted in the Y direction (vertical).

[0417] The branch's resistance to flow is calculated as the sum of the individual segment resistances along the branch.

number

[0418] or

number

[0419] The resistor array corresponding to the example shown in FIG. R s = [808 1923 1646 1569 53394 10543 55341 91454 58225], where the resistance to flow is in mmHg*s / mL.

[0420] The above resistance is for a stenosed vessel, as indicated by the peak of 91454 [mmHg*s / mL] in the resistance sequence.

[0421] Resistance sequences for the non-stenotic tree model are calculated based on quantitative coronary angiography (QCA) techniques to eliminate stenoses greater than 50% in area, where appropriate.

[0422] In some embodiments, the non-stenotic tree model is calculated by appropriately replacing stenotic vessels with dilated vessels, i.e., the geometric measurements of the stenotic vessel sections are replaced with measurements appropriate for dilated vessels.

[0423] In some embodiments, the geometric data (diameter and / or cross-sectional area) used for the dilated vessel is the maximum value of the geometric data of the unstenotic vessel in a configuration just proximal to the stenotic configuration and just distal to the stenotic configuration.

[0424] In some embodiments, the geometric data (diameter and / or cross-sectional area) used for the expanded vessel is the minimum of the geometric data of an unstenotic vessel in a configuration just proximal to the stenotic configuration and just distal to the stenotic configuration.

[0425] In some embodiments, the geometric data (diameter and / or cross-sectional area) used for the dilated vessel is the average of the geometric data of an unstenotic vessel in a configuration just proximal to the stenotic configuration and just distal to the stenotic configuration.

[0426] In some embodiments, the geometric data (diameter and / or cross-sectional area) used for the dilated vessel is calculated as a linear function of the geometric data of the unstenotic vessel between a location proximal to the stenotic location and a location distal to the stenotic location, i.e., the dilation value is calculated taking into account the distance of the stenotic location from the proximal location and from the distal location.

[0427] The stented, also referred to as expanded, resistance arrangement for the example shown in FIG.

number

[0428] The peak resistance, which was 91454 [mmHg*s / mL], was replaced by 80454 [mmHg*s / mL] in the inflated or stented model.

[0429] Reference is now made to FIG. 5, which illustrates a coronary artery tree model 1010, a combination matrix indicating tree branch tags 1020, and a combination matrix indicating tree branch resistances 1030, all generated in accordance with some exemplary embodiments of the present invention.

[0430] The tree model is an exemplary tree model with nine branches tagged with branch numbers 0, 1a, 1b, 2a, 2b, 3a, 3b, 4a, and 4b.

[0431] The combination matrix 1020 includes nine rows 1021-1029 that contain data for nine streamlines, i.e., nine paths that the fluid takes as it flows through the tree model. Five rows 1025-1029 contain data for five complete streamlines, in dark text, for the five paths that go all the way to the exit of the tree model. Four rows 1021-1024 contain data for partial streamlines, in light text, for the four paths that are not fully developed within the tree model and do not go all the way to the exit of the tree model.

[0432] Combination matrix 1030 shows rows for the same tree model as shown in combination matrix 1020 , with the branch resistances placed in cells of the matrix that correspond to the branch tags in combination matrix 1020 .

[0433] After calculating the resistance of each branch, streamlines are defined from the tree root, branch 0, to each outlet. To trace the streamlines, the branches that make up each streamline are listed in a combination matrix, e.g., as shown in Fig. 5.

[0434] In some embodiments, the defined streamlines are further numbered, as shown in FIG.

[0435] Reference is now made to FIG. 6 which shows a tree model 1100 of the vascular system generated in accordance with an exemplary embodiment of the present invention, with tags 1101-1105 numbering the outlets of the tree model 1100, where the tags correspond to flow lines.

[0436] The pressure drop along a streamline j is calculated as the sum of the pressure drops in each of its component branches (i) according to the following formula:

number

[0437] However, each branch has a different flow rate Q i When the person has the following:

[0438] Based on the principle of conservation of mass at each branch, the flow rate at the mother branch is the sum of the flows at the daughter branches. For example,

number

[0439] Thus, for example, the pressure drop along a streamline terminating at branch 4a is:

number

[0440] Here, Q j is the flow rate along streamline j, and ER 4,j is the sum of the common resistances between streamline j and streamline 4. A global equation is formulated for the pressure drop along streamline j, as follows, where appropriate:

number

[0441] For a tree with k exit branches, ie for k complete streamlines, a set of k linear equations is used accordingly.

number

[0442] where the indices 1...k represent streamlines in the tree and Q1...Q k represents the flow rate at the corresponding outlet branch. The k×k matrix A consists of elements ER and is calculated from the combination matrix. For example, for the five streamline trees shown in Figure 6, the ER matrix is:

number

number

[0443] Reference is now made to Figure 23, which illustrates an exemplary branched structure with recombining branches, according to some exemplary embodiments of the present invention. In some embodiments of the present invention, side branches and / or bifurcations of blood vessels are provided and the branches are recombined in the tree.

[0444] In some embodiments of the invention, streamlines 2302, 2303 may be modeled as comprising a loop, for example a loop comprising branch segments 2a and 4c, splitting off from branch segment 2b and then recombining, or a loop comprising branch segments 5a and 5b. In some embodiments of the invention, the necessary terms corresponding to the branches are written to reflect that the vascular resistances operate in parallel. Thus, for example, the pressure drop along streamline 2302 may be written as:

number

[0445] In some embodiments, a fluid pressure measurement, for example a blood pressure measurement, is performed. Given a fluid pressure boundary condition (P in and P out_i ), based on the vector

number

[0446] is defined, and Q i is calculated as follows:

number

[0447] For example, for a constant pressure drop of 70 mmHg between the base point and all outlets, the following flow distribution between the outlets is calculated:

[0448] Q = [1.4356, 6.6946, 1.2754, 0.7999, 1.4282], where the flow rates are in mL / s. The results are the output of the above method and are shown in Figure 7.

[0449] Reference is now made to FIG. 7, which is a simplified diagram of a vascular tree model 1210 generated in accordance with an exemplary embodiment of the present invention, showing the branch resistance R i 1220[mmHg*s / mL] and the calculated flow rate Q at each flow line outlet i Includes 1230 [mL / s].

[0450] In some embodiments, two models of the tree are calculated - the first with stenosis, optionally measured for a particular patient, and the second without stenosis. The FFR is calculated for each branch using the following formula:

number

[0451] For example, for the tree described above, the calculated FFR for each one of the nine branches is as follows:

number

[0452] The FFR calculated above is the flow rate Q S , Q N (FFR flow =Q S / Q N ) and the pressure difference distal to the stenosis, P d and proximal P a FFR derived from pressure measurements (FFR pressure =P d / P a ) in its determination. Furthermore, rather than being a comparison of two variables at a fixed system state, this is a comparison of two distinct states of the system.

[0453] FFR pressure Regarding the large differences in pressure measurements across stenotic lesions (e.g., FFR pressure If an FFR < 0.75 is found, this means that removing the lesion will remove a substantial resistance to flow, which will in turn substantially increase blood flow. "Substantially" in this case means "sufficient to be of medical value." This chain of reasoning is flow It relies on simplifying assumptions regarding the pressures and resistances remaining within the vascular system which differ in detail from those set forth above with respect to .

[0454] However, the two indexes are closely related in what they describe. FFR is therefore defined as the ratio of the maximum blood flow in a stenosed artery to the maximum blood flow in the same artery if it were normal, which is commonly measured by the pressure difference at fixed system conditions. Thus, FFR flow and FFR pressure can be characterized as indices that arrive at different forms of the same desired information, namely, what fraction of flow can, at least in principle, be restored by intervention in a particular region of the cardiovascular system.

[0455] On-site FFR pressure Acceptance of FFR also relates to experience and correlation with clinical outcomes. Therefore, it is important to develop FFR to describe the vascular system in terms that are familiar to medical professionals. pressure There is potential benefit to providing an alternative to FFR. FFR provides an index that estimates the potential for restoration of blood flow after treatment. Therefore, flow -At least as long as this has a flow ratio--even if they arrive by different routes, the FFR pressure As such it relates strictly to parameters of direct medical interest.

[0456] In addition, FFR pressure About FFR flowThe goal of determining FFR, in some embodiments of the present invention, is to guide medical decision making by providing an index that can be rapidly calculated and easily interpreted. A medical professional seeking diagnostic assistance may wish to use FFR to determine whether an intervention will result in a medically meaningful change in perfusion. flow It is potentially sufficient to determine this by vascular indices such as FFR. The ratio index is an example of an index that compactly represents such changes. Also, by describing an index that represents the potential for change, flow is FFR pressure As such, it should be noted that this potentially reduces the effects of error and / or distortion in the absolute determination of vascular perfusion characteristics.

[0457] Quality of Results Reference is now made to FIG. 24, which shows a graph of the FFR index (FFR pressure ) and image-based FFR index (FFR flow ) as a function of its mean.

[0458] In the exemplary graph, the mean difference between FFR index and image-based FFR 2415 is -0.01 (from N=34 lesions, 30 patients) and two standard deviation lines 2420, 2425 are found at 0.05 and -0.08. Lines 2405, 2410 mark the typical cutoff between a preferably "non-treatment" FFR (FFR>0.80) and a preferably "treatment" FFR (FFR<0.75) and intermediate FFR values ​​(0.075≦FFR≦0.80). In a linear correlation, R 2 The value was 0.85. The specificity was found to be 100%.

[0459] These validation results indicate that image-based FFR is potentially a direct replacement for pressure-derived FFR.

[0460] In some embodiments of the invention, the relationship between the image-calculated FFR and the baseline FFR (e.g., FFR measured by pressure before and after actual stent implantation, or FFR measured by control flow before and after actual stent implantation) has approximately, e.g., 90%, 95%, 100% specificity, or another intermediate or lesser specificity. R, which describes the correlation with the baseline FFR, 2 The value may be approximately, for example, 0.75, 0.80, 0.85, 0.90, or another larger, smaller, or intermediate value.

[0461] Some example implementations of calculating the index In some embodiments of the present invention, image processing techniques and numerical calculations are used to generate physiological indices (e.g., FFR (pressure)) that are functionally equivalent to pressure-derived fractional flow reserve (FFR). flow ) in some embodiments, the functional equivalence is direct, and in some embodiments, the functional equivalence comprises application of additional calibration factors (e.g., offsets to vessel width, changes in blood viscosity, or simply representing equivalence factors and / or functions). Integration of the above techniques potentially allows for minimally invasive assessment of blood flow at the time of diagnostic catheterization, providing a suitable estimate of the functional significance of coronary lesions.

[0462] In some embodiments of the present invention, novel physiological indices potentially allow assessment of the need for percutaneous coronary intervention and aid in making real-time diagnostic and interventional (treatment) decisions. Minimally invasive methods potentially prevent unnecessary risks to the patient and / or reduce the time and / or cost of angiography, hospitalization, and / or follow-up treatment.

[0463] In some embodiments, the geometric characteristics of the patient's vascular system, including its vascular tree, and even single vessels, and associated hemodynamic information (e.g., FFR flow) is realised, a scientific model based on patient data is realised that distinguishes between

[0464] Additionally, this model potentially allows for the combination of 3D reconstruction of the vessels with computational flow analysis to be explored to determine the functional significance of coronary lesions.

[0465] Some embodiments of the present invention perform 1-D reconstruction of one or more coronary artery segments and / or branches and computational / numerical flow analysis during coronary angiography to assess arterial pressure, flow, and / or flow resistance in parallel.

[0466] Some embodiments of the present invention perform 3-D reconstruction of one or more coronary artery segments and / or branches and computational / numerical flow analysis during coronary angiography to assess arterial pressure, flow, and / or flow resistance in parallel.

[0467] In the embodiment of the present invention where the vascular function index is calculated based only on the stenosis model, the resistance contributed by the stenosis to the total resistance of the crown of the lesion R S The volume of the crown distal to the stenosis, V crown The FFR index (FFR resistance ) but R S and V crown Representative examples of such functions include, but are not limited to, the following:

number

[0468] Here, P a is the aortic pressure, P0 is the pre-capillary pressure, and k is a scaling law coefficient that can be fitted to the aortic pressure. resistance The calculation is as follows:

[0469] Reference is now made to Figure 8, which is a simplified flow diagram of an exemplary embodiment of the present invention. This embodiment is particularly useful when a vascular function index, such as FFR, is calculated based on two models of the vascular system.

[0470] 8 illustrates portions of a method according to an exemplary embodiment, including receiving 1810 at least two 2-D angiographic images of a portion of a patient's coronary arteries and reconstructing 1815 a 3-D tree model of the coronary artery system, including lesions, if any.

[0471] Flow analysis of blood flow and, optionally, arterial pressure along the segment of interest is based on the tree model and, optionally, other available hemodynamic measurements, such as aortic pressure and / or volume of injected contrast agent.

[0472] The exemplary embodiment just described potentially provides a minimally invasive physiological index indicating the functional significance of coronary artery lesions.

[0473] The exemplary method is optionally performed when a coronary angiogram is being performed, and the calculations are optionally performed while the coronary angiogram is being performed, thereby providing a minimally invasive physiological index in real time.

[0474] Reference is now made to FIG. 9, which is a simplified flow diagram of another exemplary embodiment of the present invention.

[0475] Figure 9 shows generating a tree model of the subject's vasculature, the stenosis model comprising geometric measurements of the subject's vasculature at one or more locations along a vascular centerline of at least one branch of the subject's vasculature (1910); To obtain the flow characteristics of the stenosis model (1915) and (1920) to generate a second model of a similar dilation of the patient's vascular system as the stenosis model; To obtain the flow characteristics of the normal model (1925) and A method for vascular assessment is shown, comprising calculating (1930) an index indicative of the need for revascularization based on flow characteristics in a stenosis model and flow characteristics in a normal model.

[0476] Reference is now made to FIG. 10, which is a simplified flow diagram of yet another exemplary embodiment of the present invention.

[0477] Figure 10 shows (2010) capturing multiple 2-D images of a subject’s vasculature; generating a tree model of the subject's vasculature using at least some of the plurality of captured 2-D images, wherein the tree model comprises geometric measurements of the subject's vasculature at one or more locations along a vascular centerline of at least one branch of the subject's vasculature (2015); and generating a model of flow characteristics of the first tree model (2020).

[0478] Spread of the coronary artery tree model In some embodiments, the extent of the first stenosis model is sufficient to include the stenosis, a section of the vessel proximal to the stenosis, and a section of the vessel distal to the stenosis.

[0479] In one such embodiment, the extent of the first model is a segment of the vessel between the bifurcations, in some cases including a stenosis within the segment. In some cases, particularly when the stenosis is at the bifurcation, the extent includes the bifurcation and sections of the vessel proximal and distal to the stenotic bifurcation.

[0480] In some embodiments, the extent to which the first model extends proximally to the stenosis within a single segment ranges from as little as 1 or 2 millimeters to as much as 20 to 50 millimeters, and / or to the end of the segment itself.

[0481] In some embodiments, the extent to which the first model extends distal to the stenosis within a single segment ranges from as little as 1 or 2 millimeters to as much as 20 to 50 millimeters, and / or to the end of the segment itself.

[0482] In some embodiments, the extent to which the first model extends distal to the stenosis is measured by the bifurcations of the blood vessel. In some embodiments, the first model extends distal to the stenosis by as few as one or two branches, and in some embodiments by as many as three, four, five branches, or even more. In some embodiments, the first model extends distal to the stenosis as far as the resolution of the imaging process allows for the distal portion of the vasculature to be discerned.

[0483] A second model of the same extent as the first model is optionally generated with the stenosis dilated as if it had been revascularized back to its normal diameter.

[0484] Generating a model of the physical properties of the vascular system In an exemplary implementation, the proximal arterial pressure P a Given [mmHg], the flow rate Q through the segment of interest s [mL / s] is derived from the concentration of iodine contrast agent based on analysis of the concentration-distance-time curve and a geometric description of the segment of interest, including diameter d(l) [cm] and / or volume V(l) [ml] as a function of segment length, as appropriate.

[0485] In some embodiments, particularly for large vessels such as the left anterior descending artery (LAD), blood flow may be measured using transthoracic echo Doppler or other modalities such as MRI or SPECT to obtain a flow model.

[0486] For a given segment, the total resistance of the segment (R t [mmHg*s / mL]) is calculated by dividing arterial pressure by flow rate, as appropriate.

number

[0487] Here, R t corresponds to the total resistance, and P a corresponds to the arterial pressure, and Q s corresponds to the flow rate through the vessel segment.

[0488] From the geometric description of the segment, the local resistance of the stenosis within the segment, R s [mmHg*s / mL] is estimated. R s The estimation of may be performed using one or more of an empirical look-up table, and / or using a function such as that described in the Kirkeeide reference mentioned above, and / or a cumulative sum of Poiseuille resistance methods.

number

[0489] where the integral is over the sample segment (dl), d is the artery diameter of each sample, as appropriate, and μ is 0.035 g cm -1 ·s -1 , appropriate blood viscosity.

[0490] The downstream resistance of a segment is given by segment R as follows: n Calculated in terms of [mmHg*s / mL].

number

[0491] Normal flow rate through a non-stenotic segment Q n [mL / s] is calculated for the example as follows:

number

[0492] Here, Q n is the input flow rate to the segment, and Pa is the pressure proximal to the segment, and R n is the resistance to flow due to the vessels distal to the segment.

[0493] Fractional Flow Reserve (FFR) contrast-flow Another form of ) is optionally derived as the ratio of the measured flow through the stenosed segment to the normal flow through the non-stenosed segment.

number

[0494] In some embodiments, the FFR index (e.g., FFR contrast-flow ) are calculated using the data described below, indicating the potential effectiveness of revascularization.

[0495] Proximal arterial pressure P a [mmHg] is measured, The total inlet flow Q through a blood vessel origin, such as the origin of a coronary artery total [ml / s] is optionally derived from the concentration of the contrast agent (such as iodine) based on analysis of the concentration distance time curve. In some embodiments, particularly for large vessels such as the left anterior descending (LAD) coronary artery, flow is optionally recorded using transthoracic echo Doppler and / or other modalities such as MRI and SPECT.

[0496] The particular anatomical structure of the subject, including one or more of the following:

[0497] A geometric description of the arterial diameter along a vascular tree segment, e.g., generation of up to 3-4 as a function of the segment length dl [cm], Geometric description of arterial length along vascular tree segments (L i [cm]), e.g., a maximum of 1-2 generation downstream of the segment of interest, and a cumulative crown length downstream of the segment of interest (L crown [cm]), L crown =ΣL i , Geometric description of arterial volume along vascular tree segments (V i [ml]), e.g., a maximum of 1-2 generation downstream of the segment of interest, and a cumulative coronary volume downstream of the segment of interest (V crown [ml]), V crown =ΣV i , myocardial mass (LV mass) distribution M [ml] for the arterial segment of interest (in some embodiments, LV mass is calculated as appropriate, e.g., using transthoracic echo Doppler); and Anatomical parameters as described above are compared to normal flow (without stenosis) Q through the segment. n A reference parameter K or function F to be correlated with [mL / s], e.g.

number

[0498] Using the above data, an index indicative of the potential benefit of revascularization, such as FFR index, is calculated by performing the following calculations for each vessel segment under consideration, as appropriate:

[0499] The normal flow rate Q in a segment can be calculated from the geometric parameters of the tree, such as length, volume, mass, and / or diameter. n is obtained, From the arterial pressure, the resistance at the distal part of the segment (R n , [mmHg*s / mL]), for example, R n =P a / Q n And it is calculated as follows: From the geometrical shape, the local resistance of the stenosis in the segment R s [mmHg*s / mL] is estimated, for example, using one or more of the following methods: Lookup tables, An empirical function such as that described in Kirkeeide's reference above, and / or the cumulative sum of Poiseuille resistances R s =(128μ) / π∫(dl) / (d4 ), where the integral is over the sample of segments (dl), d is the artery diameter of each sample, and μ is 0.035 g cm -1 ·s -1 and optionally blood viscosity. Total resistance for the segment R t [mmHg*s / mL] is R t =R n +T s It is calculated as Flow rate through the stenotic segment Q s [mL / s] is Q s =P a / R t It is calculated as Indices such as fractional flow reserve (FFR) for segments are calculated as FFR=Q s / Q n It is calculated as:

[0500] The above calculation is checked for correctness by checking if the cumulative flow in the tree is Q total =ΣQ i This can be done by checking whether the measured total flow rate matches the measured total flow rate as follows:

[0501] In some embodiments, the first model extends, which includes the stenosis, as far as the resolution of the imaging modality that generated the vascular model allows, and / or extends distal to the stenosis by several branches, e.g., 3 to 4 branches. In some embodiments, the number of branches is limited by the resolution at which the vessel width can be determined from the image. For example, a cutoff is set for the branching order, at which the vessel width is no longer determinable within an accuracy of 5%, 10%, 15%, 20%, or another greater, lesser, or intermediate accuracy. In some embodiments, sufficient accuracy is not available, e.g., due to insufficient imaging resolution in the original image. The availability of a large number of measurable branches is a potential advantage for a more complete reconstruction of detailed vascular resistance in the coronary vessels of the stenosis. It is noted that in the state of the art, CT scans generally have a lower resolution than X-ray angiographic imaging, reducing the availability of vessels at which vascular resistance can be determined.

[0502] In an exemplary implementation, the total inlet flow rate through the origin of the coronary arteries is derived from the concentration of the contrast agent and, optionally, the particular anatomy of the subject.

[0503] In some embodiments, the anatomical data includes, as appropriate, a geometric description of arterial diameter along a vascular segment distal to the stenosis by up to 3-4 bifurcations, a geometric description of arterial length along a vascular tree segment, a geometric description of arterial volume along the tree segment, and / or myocardial mass (LV mass) distribution for the arterial segment of interest.

[0504] In some embodiments, the anatomical data includes, as appropriate, a geometric description of the arterial diameter along a vascular segment distal to the stenosis as the imaging modality permits, a geometric description of the arterial length along a vascular tree segment, a geometric description of the arterial volume along the tree segment, and / or myocardial mass (LV mass) distribution for the arterial segment of interest.

[0505] In some embodiments, LV mass is calculated by using a transthoracic echo Doppler, if desired.

[0506] In some embodiments, a reference scaling parameter or function is used that correlates the anatomical parameter with the normal flow rate through a non-stenotic segment.

[0507] In some embodiments, a first model extent includes a stenosed blood vessel and a second model includes a similar extent of the vasculature, with healthy blood vessels similar to the stenosed blood vessels.

[0508] FFR index (FFR 2-segment ) is optionally calculated from the ratio of the measured flow in the stenosed vessel to the flow in an adjacent healthy vessel. In some embodiments, the index is adjusted by the ratio between the total length of the vessel in the crown of the stenosed vessel and the healthy vessel. The crown of a vessel is defined herein as the sub-tree that branches off from the vessel. The total length of the crown is optionally derived from a 3-D reconstruction of the coronary artery tree.

[0509] Referring now to FIG. 11, this is a simplified drawing 2100 of a vasculature system including a stenosed blood vessel 2105 and a non-stenosed blood vessel 2107 .

[0510] 11 shows two blood vessels that are candidates for benchmarking flow characteristic comparisons between a stenosed blood vessel 2105 and a non-stenosed blood vessel 2107. FIG. 11 also shows a stenosed blood vessel corona 2115 and a non-stenosed blood vessel corona 2117.

[0511] Note that the two vessels in FIG. 11 seem to be particularly good candidates for comparison because they both appear to have similar diameters and both appear to have similar crowns.

[0512] Due to scaling laws, there is a linear relationship between the normal flow rate Q in an artery and the total length of the vessel in its coronary region. This relationship holds for both adjacent healthy vessels and stenosed vessels. For a healthy vessel,

number

[0513] where

number

[0514] is the flow rate of a healthy blood vessel, k is the correlation coefficient, and L h is the total length of the coronary vasculature in a healthy vessel.

[0515] The same applies to stenosed blood vessels:

number

[0516] where

number

[0517] is the flow rate of the stenotic blood vessel, k is the correlation coefficient, and L s is the total length of the coronary vasculature of the stenosed vessel.

[0518] FFR 2-segment is defined in some embodiments as the ratio of the flow rate in a stenosed artery during hyperemia to the flow rate in the same artery when there is no stenosis (normal flow rate), as described in Equation 6.3 below. The above relationship results in an FFR, calculated as the ratio between the flow rates measured in both vessels divided by the ratio between their respective coronary lengths.

[0519] It should be noted that the scaling law also defines the relationship between normal flow rate and total coronary volume. In some embodiments, the index or FFR is calculated as the above-mentioned ratio between the measured flow rates in the diseased and healthy arteries divided by the ratio between the respective total coronary volumes of the stenosed and normal vessels, raised to the power of 3 / 4.

number

[0520] Where:

number

[0521] is the existing flow rate in the stenotic vessel as measured by the methods described herein;

number

[0522] is the existing flow rate in a healthy vessel, measured by the method described herein, and L s is the total coronary length of the stenotic vessel, and L h is the total coronary length of a healthy vessel.

[0523] Scaling laws also define the relationship between normal flow rate and total crown volume.

number

[0524] where

number

[0525] is the flow rate of a healthy blood vessel, kv is the correlation coefficient, and V h is the total volume of the coronary vasculature in healthy vessels.

[0526] The same applies to stenosed blood vessels:

number

[0527] where

number

[0528] is the flow rate of the stenotic blood vessel, and k v is the correlation coefficient, and V h is the total volume of the coronary vasculature in the stenosed vessel.

[0529] The FFR is calculated from the above-mentioned ratio of the measured flow rates in the diseased and healthy vessels, as appropriate, divided by the ratio of the total coronary volumes of each raised to the power of 3 / 4.

number

[0530] Here, V h and V s By way of non-limiting example, is measured by using a 3-D model of the vasculature.

[0531] Hardware implementation example Reference is now made to FIG. 12A, which is a simplified diagram of a hardware implementation of a system for vascular assessment constructed in accordance with an exemplary embodiment of the present invention.

[0532] The exemplary system of FIG. 12A includes:

[0533] Two or more imaging devices 2205 for capturing multiple 2-D images of the patient's vasculature, a computer 2210 operatively connected 2209 to the two or more imaging devices 2205 .

[0534] The computer 2210 is optionally configured to receive data from the multiple imaging devices 2205 and use at least some of the multiple captured 2-D images to generate a tree model of the patient's vasculature, where the tree model comprises geometric measurements of the patient's vasculature at one or more locations along a vascular centerline of at least one branch of the patient's vasculature, and to generate a model of the flow characteristics of the tree model.

[0535] In some embodiments, a synchronization unit (not shown) is used to provide a synchronization signal to the imaging device 2205 to synchronize the capture of 2-D images of the patient's vasculature.

[0536] Reference is now made to FIG. 12B, which is a simplified diagram of another hardware implementation of a system for vascular assessment constructed in accordance with an exemplary embodiment of the present invention.

[0537] The exemplary system of FIG. 12B includes:

[0538] An imaging device 2235 for capturing multiple 2-D images of the patient's vasculature, and a computer 2210 operatively connected (2239) to the imaging device 2235.

[0539] 12B, the imaging device 2235 is configured to acquire 2-D images from two or more orientations relative to the subject. The orientation and positioning of the imaging device 2235 relative to the subject and / or relative to a fixed frame of reference are optionally recorded to assist in generating a vascular tree model, whether one-dimensional (1-D) or three-dimensional (3-D), from the 2-D images taken from the imaging device 2235.

[0540] The computer 2210 is optionally configured to receive data from the multiple imaging devices 2235 and use at least some of the multiple captured 2-D images to generate a tree model of the patient's vasculature, where the tree model comprises geometric measurements of the patient's vasculature at one or more locations along a vascular centerline of at least one branch of the patient's vasculature, and to generate a model of the flow characteristics of the tree model.

[0541] In some embodiments, a synchronization unit (not shown) is used to provide a synchronization signal to the imaging device 2235 to synchronize the capture of 2-D images of the patient's vasculature, optionally at the same phase in the cardiac cycle.

[0542] In some embodiments, the computer 2210 accepts an ECG signal (not shown) of the subject and selects 2-D images from the imaging device 2235 according to the ECG signal, for example, selecting 2-D images at the same cardiac phase.

[0543] In some embodiments, the system of FIG. 12A or 12B includes an image registration unit that detects corresponding image features in the 2-D image, calculates image correction parameters based on the corresponding image features, and registers the 2-D image such that the image features geometrically correspond.

[0544] In some embodiments, the image features are suitably selected as base points of the tree model and / or locations of minimum radii within a stenotic vessel and / or bifurcations of the vessel.

[0545] Example of a system for scoring vascular status Reference is now made to FIG. 22, which is a simplified schematic diagram of an automated VSST scoring system 700, according to some exemplary embodiments of the present invention.

[0546] 22, the broad white paths (e.g., path 751) represent simplified paths of data processing through the system. The broad black paths (e.g., path 763) represent one or more data connections to the system user interface 720. The black path data content is labeled with overlaying trapezoidal blocks.

[0547] The vascular tree reconstructor 702, in some embodiments of the invention, receives image data 735 from one or more imaging systems or system connectivity network 730. The stenosis determiner 704, in some embodiments, determines the presence of a stenotic vascular lesion based on the reconstructed vascular tree. In some embodiments, the metrics module 706 determines additional metrics related to the disease state of the vascular tree based on the reconstructed vascular tree and / or the determined stenosis location and other measurements.

[0548] In some embodiments, the metrics extractor 701 comprises the functionality of the vascular tree reconstructor 702, the stenosis determiner 704, and / or the metrics module 706. In some embodiments, the metrics extractor 701 is operable to receive image data 735 and extract therefrom a number of vascular condition metrics suitable, for example, as input to the parameter combiner 708.

[0549] In some embodiments, the parameter combiner 708 converts the determined metrics into sub-score values ​​(e.g., true / false values) comprising parameters that "answer" the vascular status scoring question and / or are otherwise mapped to specific operations of the VSST scoring procedure.

[0550] In some embodiments, the sub-score extractor 703 comprises the functionality of the vascular tree reconstructor 702, the stenosis determiner 704, the metrics module 706, and / or the parameter combiner 708. In some embodiments, the sub-score extractor 703 comprises the functionality of the metrics extractor 701. In some embodiments, the sub-score extractor 703 receives image data 735 and extracts therefrom one or more VSST sub-scores that are suitable as input for the score calculator 713.

[0551] The parameter finalizer 710, in some embodiments, ensures that the parameter data provided is sufficiently complete and correct to proceed to final scoring. In some embodiments, corrections to the automatically determined parameters are determined in the finalizer 710, optionally with operator oversight through the system user interface 720. In some embodiments, gaps in the automatically supplied parameter data are filled in, for example, by user input from the system user interface 720, or by other parameter data 725 provided, for example, from another diagnostic system or from a network allowing access to clinical data.

[0552] The score compositor 712, in some embodiments, combines the finalized outputs based on the determined parameters for the scores into a weighted score output 715. The scores are available, for example, on a system user interface to the network resources 730.

[0553] In some embodiments of the invention, the score calculator 713 comprises the functionality of the parameter finalizer 710 and / or the score combiner 712. In some embodiments, the score calculator 713 is operable to receive combined parameters and / or sub-scores (e.g., from the parameter combiner 708 and / or the sub-score extractor 703) and convert them into a VSST score output 715.

[0554] In some embodiments of the invention, intermediate results of the processing (e.g., the reconstructed vascular tree, various metrics determined therefrom, and / or parameter determinations) are stored in persistent or temporary storage on a storage device (not shown) of the system 700 and / or on the network 730.

[0555] The scoring system 700 is described in the context of modules that, in some embodiments of the invention, are implemented as program functions of a digital computer. It will be appreciated that the underlying system architecture may be implemented in a variety of ways that comprise embodiments of the invention, for example, as a single-process or multi-process application and / or as client-server processes running on the same or different computer hardware systems. In some embodiments of the invention, the system is implemented in code executed by a general-purpose processor. In some embodiments, some or all of the functionality of one or more modules is realized by an FPGA or another dedicated hardware component, such as an ASIC.

[0556] To provide an example of a client-server configuration, the sub-score extractor 703 is implemented as a server process (or a group of server-implemented processes) on one or more machines remote from the client computers that implement modules such as the score calculator 713 and the user interface 720. It will be understood that other divisions of modules (or even divisions within modules) described herein are encompassed by some embodiments of the invention. A potential advantage of such a division is, for example, achieving economies of scale by performing computationally intensive parts of the scoring on fast dedicated hardware while allowing the hardware to be shared among multiple end users. Such a distributed architecture potentially also has advantages for the maintenance and / or distribution of new software versions.

[0557] Potential Advantages of Embodiments of the Invention Some exemplary embodiments of the present invention are minimally invasive, i.e., they allow for the forgoing of guidewire probing of the coronary arteries, thus minimizing risk to the patient compared to invasive FFR catheter procedures.

[0558] It is noted that an exemplary embodiment of the present invention provides a cost-effective means of allowing reliable indices to be measured at the time of catheterization, potentially eliminating the need for processing of angiographic data after the catheterization procedure, and / or additional equipment at the catheterization procedure such as guidewires, and / or materials involved in the catheterization procedure such as adenosine. It is noted that another potential savings includes savings in hospitalization costs following more appropriate treatment decisions.

[0559] An exemplary embodiment of the present invention allows for trying different post-dilatation vessel cross sections in different post-dilatation models of the vasculature, as appropriate, and selecting an appropriate stent for a subject based on the desired flow characteristics of the post-dilatation models.

[0560] An exemplary embodiment of the present invention optionally automatically identifies geometric characteristics of blood vessels, defines the vessel contour, and optionally provides relevant hemodynamic information associated with the vessel that is compatible with current invasive FFR methods.

[0561] An embodiment of the present invention generates an index that indicates the need for coronary revascularization when appropriate. Minimally invasive embodiments of the present invention potentially prevent unnecessary risks to the patient and potentially reduce the overall time and cost of angiography, hospitalization, and follow-up care.

[0562] A system constructed according to an exemplary embodiment of the present invention potentially allows for shortening diagnostic angiography procedures. Unnecessary coronary interventions during angiography and / or in the future are also potentially prevented. Additionally, a method according to an exemplary embodiment of the present invention optionally allows for evaluation of vascular problems in other arterial territories, such as the carotid arteries, renal arteries, and diseased vessels of the limbs.

[0563] It should be noted that the resolution of angiographic images is typically higher than that typically obtained by 3-D techniques such as CT. Models constructed from higher resolution angiographic images are inherently of higher resolution, resulting in greater geometric accuracy and / or allowing the use of smaller geometric characteristics of the vessel than CT images and / or allowing calculations using more generation or branching downstream from the stenosis versus vessel branches distal to the stenosis compared to CT images.

[0564] A short list of potential non-invasive FFR benefits, one or more of which are realized in some embodiments of the invention, are as follows:

[0565] A non-invasive method that does not put the patient at risk; A calculation method that does not require additional time or invasive equipment; prognostic benefits in "borderline" and multivessel disease; provide a reliable index for assessing the need for coronary revascularization, Methods for evaluating and / or optimizing revascularization procedures; strategies to save costs of catheterization, hospitalization, and follow-up care; Prevent unnecessary coronary interventions following angiography, "One-stop shop" comprehensive lesion evaluation.

[0566] Another advantage of some embodiments of the present invention is that the tree model can be generated within a short time period. This allows the calculation of indices, particularly, but not necessarily, indices indicative of vascular function (e.g., FFR) within a also short time period (e.g., less than 60 minutes, or less than 50 minutes, or less than 40 minutes, or less than 30 minutes, or less than 20 minutes from the time the 2-D image is received by the computer). Preferably, the indices are calculated while the subject is immobilized and fixed on the treatment surface for catheter insertion. Such fast calculation of the indices is advantageous because it allows the physician to make an assessment of the lesion being diagnosed while the catheter insertion procedure is being performed, thereby allowing a quick decision regarding the appropriate treatment for that lesion. The physician can determine the need for treatment while in the catheter insertion room and does not have to wait for offline analysis.

[0567] Additional benefits of rapid index calculation include reduced risk to the patient, the ability to calculate the index without the need for drugs or invasive devices, reduced performance time of coronary diagnostic procedures, the benefit of established prognosis in borderline lesions, reduced costs, a reduction in the number of unnecessary coronary interventions, and a reduction in the volume of subsequent procedures.

[0568] In this "game" of assessing the hemodynamic severity of each lesion, non-real-time solutions are often not an option considered. Physicians need to know whether to treat a lesion in the catheterization lab and cannot afford to wait for offline analysis. CT-based solutions are also part of a different "game," but it is important to note that the availability of cardiac CT scans is low compared to PCI procedures, and the solutions are much lower in both time and space compared to angiography.

[0569] Another point to emphasize is that online image-based FFR assessment, unlike invasive assessment, potentially allows for the assessment of borderline lesions, not necessarily limited to the percentage of hospitals that are assessed today, since the risk to the patient (e.g., due to passing a guidewire) and the cost are much lower.

[0570] It is expected that many related methods and systems for imaging the vasculature will be developed during the life of the patent which matures from this application, and the scope of the terms describing imaging is intended to include a priori all such new technologies.

[0571] As used herein, the term "about" refers to 10%.

[0572] The words "comprises," "comprises," "includes," "including," "having," and variations thereof mean "including but not limited to."

[0573] The term "consisting of" means "including and including."

[0574] The phrase "consisting essentially of" means that a composition, method, or structure includes additional materials, steps, and / or components, but only if the additional materials, steps, and / or components do not materially alter the basic and novel characteristics of the claimed composition, method, or structure.

[0575] As used herein in the English language, the singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise. For example, the phrase "a compound" or "at least one compound" can include a plurality of compounds, including mixtures thereof.

[0576] The words "example" or "exemplary" are used herein to mean serving as an example, instance, or illustration. An embodiment described as "example" or "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments and / or to exclude the incorporation of features from other embodiments.

[0577] The phrase "optionally" is used herein to mean "provided in some embodiments and not provided in other embodiments." Particular embodiments of the invention may include multiple "optional" features unless such features are inconsistent.

[0578] As used herein, the term "method" refers to methods, means, techniques, procedures, and procedures for accomplishing a given task, including, but not limited to, methods, means, techniques, procedures, and procedures known or readily developed from methods, means, techniques, procedures, and procedures known to practitioners of chemistry, pharmacology, biology, biochemistry, and medicine.

[0579] Throughout this application, various embodiments of the present invention may be presented in a range format. It will be understood that the description in range format is merely for the convenience of brevity and should not be interpreted as an inflexible limitation on the scope of the present invention. Thus, the description of a range should be considered as having all possible subranges specifically disclosed, as well as individual numerical values ​​within that range. For example, a description of a range such as 1 to 6 should be considered as having specifically disclosed subranges such as 1 to 3, 1 to 4, 1 to 5, 2 to 4, 2 to 6, 3 to 6, as well as individual numbers within that range, for example, 1, 2, 3, 4, 5, and 6. This applies regardless of the breadth of the range.

[0580] Whenever a range of numerical values ​​is indicated herein, it is intended to include the numbers (fractional or integer) recited within the indicated range. The phrases "ranging between" a first indicated number and a second indicated number and "ranging from" a first indicated number to a second indicated number are used interchangeably herein and are intended to include the first and second indicated numbers and all fractional and integer numbers therebetween.

[0581] It will be understood that certain features of the invention that are, for clarity, described in the context of separate embodiments, can also be implemented in combination in a single embodiment. Conversely, various features of the invention that are, for brevity, described in the context of a single embodiment, can also be provided separately or in any suitable subcombination or as appropriate in other described embodiments of the invention. Certain features that are described in the context of various embodiments should not be considered essential features of those embodiments, unless the embodiment is inoperable without those elements.

[0582] While the present invention has been described in conjunction with specific embodiments thereof, it is evident that many alternatives, modifications, and variations will be apparent to those skilled in the art. Accordingly, the present invention is intended to embrace all such alternatives, modifications, and variations that fall within the spirit and broad scope of the appended claims.

[0583] All publications, patents, and patent applications mentioned in this specification are incorporated herein to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated herein. In addition, citation or identification of any document in this application shall not be construed as an admission that such document is available as prior art to the present invention. To the extent section headings are used, they should not be construed as necessary limitations.

Claims

1. A computer-implemented method comprising: acquiring sensor information reflective of at least a portion of the heart; determining a representation of a coronary vascular tree based on the sensor information, wherein the representation is based on centerlines identified for the vessel segments forming the coronary vascular tree and geometric information associated with the vessel segments; adjusting geometric information for the at least one vessel segment such that flow characteristics of the at least one vessel segment are adjusted; and calculating a flow index that quantifies vascular function of the cardiovascular system based on the representation and the adjusted geometric information.

2. The method described in claim 1, wherein the sensor information includes an angiographic image.

3. The method described in claim 2, wherein the sensor information includes computed tomography information.

4. The method of claim 2, wherein the geometric information includes a diameter associated with the vascular segment, and wherein adjusting the geometric information for the at least one vascular segment includes increasing one or more diameters associated with the at least one vascular segment.

5. The method described in claim 2, wherein the flow characteristics for the at least one vascular segment are determined based on resistance to fluid flow associated with the at least one vascular segment, and wherein adjusting the geometric information for the at least one vascular segment includes reducing resistance to the fluid flow.

6. The method described in claim 2, wherein the at least one vascular segment corresponds to a stenosis identified based on the representation.

7. The method described in claim 2, wherein the flow index indicates a fractional flow reserve.

8. A system comprising one or more processors and one or more non-transitory computer storage media storing instructions, the instructions, when executed by the one or more processors, causing the one or more processors to: acquiring sensor information reflective of at least a portion of the heart; determining a representation of a coronary vascular tree based on the sensor information, wherein the representation is based on centerlines identified for the vessel segments forming the coronary vascular tree and geometric information associated with the vessel segments; adjusting geometric information for the at least one vessel segment such that flow characteristics of the at least one vessel segment are adjusted; and calculating a flow index that quantifies vascular function of the cardiovascular system based on the representation and the adjusted geometric information.

9. The system described in claim 8, wherein the sensor information includes angiographic images.

10. The system described in claim 9, wherein the sensor information includes computed tomography information.

11. The system of claim 9, wherein the geometric information includes a diameter associated with the vascular segment, and wherein adjusting the geometric information for the at least one vascular segment includes increasing one or more diameters associated with the at least one vascular segment.

12. The system described in claim 9, wherein the flow characteristics for the at least one vascular segment are determined based on resistance to fluid flow associated with the at least one vascular segment, and wherein adjusting the geometric information for the at least one vascular segment includes reducing resistance to the fluid flow.

13. The system described in claim 9, wherein the at least one vascular segment corresponds to a stenosis identified based on the representation.

14. The system described in claim 9, wherein the flow index indicates a fractional flow reserve.

15. A non-transitory computer storage medium storing instructions that, when executed by a system of one or more processors, cause the one or more processors to: acquiring sensor information reflective of at least a portion of the heart; determining a representation of a coronary vascular tree based on the sensor information, wherein the representation is based on centerlines identified for the vessel segments forming the coronary vascular tree and geometric information associated with the vessel segments; adjusting geometric information for the at least one vessel segment such that flow characteristics of the at least one vessel segment are adjusted; and calculating a flow index that quantifies vascular function of the cardiovascular system based on the representation and the adjusted geometric information.

16. The computer storage medium of claim 15, wherein the sensor information includes an angiographic image.

17. The computer storage medium of claim 16, wherein the sensor information includes computed tomography information.

18. The computer storage medium of claim 16, wherein the geometric information includes a diameter associated with the vascular segment, and wherein adjusting the geometric information for the at least one vascular segment includes increasing one or more diameters associated with the at least one vascular segment.

19. The computer storage medium of claim 16, wherein the flow characteristics for the at least one vascular segment are determined based on resistance to fluid flow associated with the at least one vascular segment, and wherein adjusting the geometric information for the at least one vascular segment includes reducing resistance to the fluid flow.

20. The computer storage medium of claim 16, wherein the at least one vascular segment corresponds to a stenosis identified based on the representation.

Images