WO2017037287A1 - Procédé et dispositif d'identification de la connectivité vasculaire d'une image - Google Patents

Procédé et dispositif d'identification de la connectivité vasculaire d'une image Download PDF

Info

Publication number
WO2017037287A1
WO2017037287A1 PCT/EP2016/070842 EP2016070842W WO2017037287A1 WO 2017037287 A1 WO2017037287 A1 WO 2017037287A1 EP 2016070842 W EP2016070842 W EP 2016070842W WO 2017037287 A1 WO2017037287 A1 WO 2017037287A1
Authority
WO
WIPO (PCT)
Prior art keywords
volume
volume element
blood flow
arterial
venous
Prior art date
Application number
PCT/EP2016/070842
Other languages
English (en)
Inventor
Frederik MAES
David ROBBEN
Paul Suetens
Original Assignee
Katholieke Universiteit Leuven
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from GBGB1515661.5A external-priority patent/GB201515661D0/en
Priority claimed from GBGB1515660.7A external-priority patent/GB201515660D0/en
Priority claimed from GBGB1517063.2A external-priority patent/GB201517063D0/en
Application filed by Katholieke Universiteit Leuven filed Critical Katholieke Universiteit Leuven
Publication of WO2017037287A1 publication Critical patent/WO2017037287A1/fr

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/20—Analysis of motion
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30101—Blood vessel; Artery; Vein; Vascular
    • G06T2207/30104—Vascular flow; Blood flow; Perfusion
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30241—Trajectory

Definitions

  • the present invention relates to a method and device for identifying vascular connectivity of an anatomical region of interest in an image. It is particularly, though not exclusively, applicable in identifying cerebral vascular connectivity from dynamic CT and MRI perfusion studies.
  • the vascular system plays a crucial function in our body, permitting blood to circulate from the heart, through arteries and arterioles to the capillary bed, and through venules and veins back to the heart.
  • perfusion takes places: the process of exchanging nutrients, oxygen, waste, etc. with the surrounding cells. Perfusion is essential for the function and survival of living tissue and hence there is vasculature virtually everywhere in the body.
  • CTA CT angiography
  • MRA MR angiography
  • CTP CT perfusion imaging
  • each image of the movie can be considered a traditional CT angiography (CTA), thus allowing assessment of the vasculature in both arterial and venous phase images. More important, however, is the voxelwise calculation of clinically relevant parameters such as time to peak, mean transit time, blood flow and blood volume. These voxelwise parameter maps are used for the diagnosis of stroke— a.o. for the assessment of the tissue at risk— and help making therapeutic decisions. Organs of interest may for example include liver, lungs and brain.
  • WO 2013/159147 discloses a method for perfusion territory determination. Therein it is provided that, while the assumption of monotonically rising arrival times is true in theory, the high blood velocity in big arteries (cf. 50cm/s), the relatively low sampling frequency (cf. 1 image/s) and the small voxel size (cf. less than 1 mm 3 ) make their estimation of the arrival time too noisy for previous methods to work: the many local minima in the arrival time will result in as many incorrect arterial sources.
  • region growing with a low threshold is applied, the selected voxels are added to their respective seed regions, the threshold is increased and the process is repeated.
  • cerebral CTP images where manual segmentations of the anterior, middle and posterior arterial trees are used as seeds and their method subsequently finds the associated perfusion territories.
  • two masks— one for each hemisphere— are used to avoid the region growing process to flow over in the other hemisphere. This is done to capture the anatomical knowledge that the arteries do not cross the arachnoid mater (one of the three meninges), which also runs down the medial longitudinal fissure.
  • the present invention relates to a method for identifying blood flow trajectories in an imaging volume, the method comprising - obtaining for each volume element of a plurality of volume elements into which the imaging volume is divided, arrival or departure time data representing respectively an estimated arrival or departure time at the volume element of an administered contrast agent whereby each volume element indicating the administered contrast agent is a perfused volume element;
  • defining for each perfused volume element a blood flow trajectory from one of said arterial volume elements respectively towards said venous volume elements wherein said defining blood flow trajectories comprises optimising for all volume element a cost function of their paths from a peripheral arterial volume element to the perfused volume element respectively the cost of their paths to a peripheral venous volume element.
  • Said defining blood flow trajectories may comprise optimizing for every volume element a cost function of its path from a peripheral arterial volume element to the perfused volume element respectively the cost of its path to a peripheral venous element.
  • the cost function of the path respectively to a perfused volume element respectively from a perfused volume element towards a venous element may be the maximal time respectively the negative of the minimal departure time of the contrast bolus in the perfused volume elements of the path.
  • Trajectories may be obtained by dynamic programming.
  • the periphera region of the imaging volume may be the periphery of the imaging volume having a width e.g. between 1mm and 50mm, e.g. between 1mm and 30mm, e.g. between 10mm and 30mm, e.g. 20mm.
  • the peripheral region of the imaging volume may be the periphery of the imaging volume having a width of at least 1 voxels, preferably a width between 1 to 5 voxels, more preferably 1 to 20 voxels or more preferably 1 to 50 voxels.
  • Said blood flow trajectory may be an arterial blood flow trajectory
  • said obtaining may comprise obtaining for each volume element of a plurality of volume elements into which the imaging volume is divided, arrival time data representing an estimated arrival time at the volume element of an administered contrast agent whereby each volume element indicating the administered contrast agent is a perfused volume element;
  • said identifying may comprise identifying arterial volume elements towards peripheral regions of the imaging volume and
  • said defining may comprise defining towards each perfused volume element an arterial blood flow trajectory from one of said arterial volume elements.
  • the estimated arrival time in a volume element may be obtained by applying time to half max (THM) to the time-intensity time series in that volume element.
  • a trajectory may be a sequence of neighboring volume elements through which arterial blood flows from a peripheral arterial volume element towards each perfused volume element.
  • Said blood flow trajectory may be a venous arterial blood flow trajectory and wherein
  • said obtaining may comprise obtaining for each volume element of a plurality of volume elements into which the imaging volume is divided, departure time data representing an estimated departure time at the volume element of an administered contrast whereby each volume element indicating the administered contrast agent is a perfused volume element;
  • said identifying may comprise identifying venous volume elements towards peripheral regions of the imaging volume and
  • said defining may comprise defining for each perfused volume element of a plurality of volume elements a venous flow trajectory to one of said venous volume elements.
  • the estimated departure time of a volume element may be obtained by
  • a trajectory may be a sequence of neighboring volume elements through which blood flows from a perfused volume element towards a peripheral venous volume element.
  • the present invention furthermore relates to a computer-readable medium containing computer program instructions which, when executed by at least one processor, cause the at least one processor to perform the method as described above.
  • the method as claimed does not comprise the step of administering a contrast agent. More particularly, the method as claimed relates to data processing of image data and the different acts performed act on image data, e.g. time dependent image data, and not on the human body.
  • the present invention also relates to a system for identifying blood flow trajectories in an imaging volume, the system comprising
  • - pre-processing means programmed for obtaining for each volume element of a plurality of volume elements into which the imaging volume is divided, arrival or departure time data representing respectively an estimated arrival or departure time at the volume element of an administered contrast agent whereby each volume element indicating the administered contrast agent is a perfused volume element;
  • - identifying means programmed for identifying arterial respectively venous volume elements towards peripheral regions of the imaging volume
  • defining blood flow trajectories comprises optimising for all volume element a cost function of their paths from a peripheral arterial volume element to the perfused volume element respectively a cost function of their paths to a peripheral venous element.
  • Said blood flow trajectory may be an arterial blood flow trajectory
  • said pre-processing means may be a pre-processing means programmed for obtaining, for each volume element of a plurality of volume elements into which the imaging volume is divided, arrival time data representing an estimated arrival time of a contrast agent at the volume element;
  • said identifying means may be an identifying means programmed for identifying arterial volume elements towards peripheral regions of the imaging volume and;
  • Said blood flow trajectory may be a venous blood flow trajectory
  • said pre-processing means may be a pre-processing means for obtaining, for each volume element of a plurality of volume elements into which the imaging volume is divided, departure time data representing an estimated departure time of a contrast agent at the volume element;
  • said identifying means may be a identifying means for identifying venous volume elements towards peripheral regions of the imaging volume and;
  • the system further may comprise selecting means, said selecting means adapted to select volume elements of an artery respectively vein of interest.
  • the system for identifying blood flow trajectories may comprise inferring means, said inferring means adapted to deduce the perfusion territory of the selected volume elements, whereby the perfusion territory comprises all volume elements whose perfusion path goes through one of the selected volume elements.
  • the inferring means may be adapted to deduce the drainage territory of the selected volume elements, whereby the drainage territory is determined by all volume elements whose venous blood flow trajectory contains one of the selected volume elements.
  • the present invention also relates to a volumetric image of blood flow trajectories obtained by a method as described above.
  • the volumetric image may visualize an arrival time as a 3D iso-surface at an arrival time threshold, the arrival time being interactively controlled by the user or the arrival time advancing, e.g. automatically, through a range of values.
  • the present invention also relates to a method for visualizing an arrival time image as a 3D iso-surace at an arrival time threshold whereby the arrival time is either interactively controlled by the user, or advances automatically through a range of values.
  • Embodiments of the present invention provide a method to deduct or infer the vascular connectivity throughout an organ - even if the vasculature itself is on a subvoxel scale.
  • the organs of interest may for example include liver, lungs or brains, but are not limited thereto.
  • Embodiments of the present invention provide methods for identifying arterial blood flow trajectories and connectivity in an imaging volume as well as identifying venous blood flow trajectories in an imaging volume.
  • the present invention provides methods for identifying arterial blood flow trajectories in an imaging volume, the method comprising:
  • each perfused volume element defining towards each perfused volume element an arterial blood flow trajectory from one of said arterial volume elements.
  • the imaging volume may be a CT-scan, however any other suitable imaging volumes known and found suitable by the skilled person may be applied.
  • said imaging volume is a three dimensional image whereby an arrival time or a time series per image voxel is provided. Whereby from the time series one can derive the arrival time.
  • the estimated arrival time in a volume element is obtained by applying time to half max (THM) to the time-intensity time series in that volume element.
  • THM as parameter is better conditioned than time to peak (TTP).
  • TTP time to peak
  • a trajectory is a sequence of neighboring volume elements through which arterial blood flows from a peripheral arterial volume elements towards each perfused volume element. If the arterial blood flow trajectory to voxel A comprises voxel B, then the first part of the sequence will be the arterial blood flow trajectory to voxel B. In preferred embodiments obtaining the trajectories comprises minimizing for every volume element the cost of its path from a peripheral arterial volume element to the perfused volume element.
  • the cost of its trajectory to a perfused volume element is the maximal time of the contrast bolus in the perfused volume elements of the path.
  • the trajectories are obtained by dynamic programming.
  • the peripheral region of the imaging volume is the periphery of the imaging volume having a width of at least 1 voxels, preferably a width between 1 to 5 voxels, more preferably 1 to 20 voxels or more preferably 1 to 50 voxels.
  • Embodiments of the present invention provides a computer-readable medium containing computer program instructions which, when executed by at least one processor, cause the at least one processor to perform the method according to embodiments of the present invention.
  • the present invention provides systems for identifying blood flow trajectories in an imaging volume, the system comprising;
  • - pre-processing means for obtaining, for each volume element of a plurality of volume elements into which the imaging volume is divided, arrival time data representing an estimated arrival time of a contrast agent at the volume element;
  • - identifying means for identifying arterial volume elements towards peripheral regions of the imaging volume and
  • a system for identifying blood flow trajectories in an imaging volume further comprises selecting means, said selecting means adapted to select volume elements of an artery of interest.
  • a system for identifying blood flow trajectories in an imaging volume further comprises inferring means, said inferring means adapted to deduce the perfusion territory of the selected volume elements.
  • all volume elements whose perfusion path goes through one of the selected volume elements if inferred.
  • the perfusion territory being all volume elements whose perfusion path goes through one of the selected volume elements.
  • the present invention provides methods for identifying venous blood flow trajectories in an imaging volume, the method comprising:
  • each volume element indicating the administered contrast agent is a perfused volume element
  • each perfused volume element of a plurality of volume elements a venous flow trajectory to one of said venous volume elements.
  • the estimated departure time of a volume element is obtained by
  • a trajectory is a sequence of neighboring volume elements through which blood flows from a perfused volume element towards a peripheral venous volume element. If the venous blood flow trajectory from voxel A comprises voxel B, then the last part of the trajectory will be the venous blood flow trajectory from voxel B.
  • obtaining them comprises minimizing for every volume element the cost of its path to a peripheral venous element.
  • the cost of the path from a perfused volume element towards a venous element is the negative of the minimal departure time of the contrast bolus in the perfused volume elements of the path.
  • the perfusion paths are obtained by dynamic programming.
  • the peripheral region of the imaging volume is the periphery of the imaging volume having a width of at least 1 voxel, preferably a width between 1 to 5 voxels, more preferably 1 to 20 voxels or more preferably 1 to 50 voxels.
  • Embodiments of the present invention provides a computer-readable medium containing computer program instructions which, when executed by at least one processor, cause the at least one processor to perform the method for identifying venous blood flow trajectories according to the present invention.
  • the present invention provides systems for identifying venous flow trajectories in an imaging volume, the system comprising;
  • - pre-processing means for obtaining, for each volume element of a plurality of volume elements into which the imaging volume is divided, departure time data representing an estimated departure time of a contrast agent at the volume element;
  • - identifying means for identifying venous volume elements towards peripheral regions of the imaging volume and
  • each perfused volume element a venous flow trajectory to said venous volume elements.
  • system for identifying blood flow trajectories further comprises selecting means, said selecting means adapted to select volume elements of a vein of interest.
  • the system for identifying blood flow trajectories further comprises inferring means, said inferring means adapted to deduce the drainage territory of the selected volume elements. That is, all volume elements whose venous blood flow trajectory contains one of the selected volume elements.
  • CT perfusion imaging CTP
  • multiple, consecutive 3D CT scans of an organ are made during the administration of contrast agent. This results in a 3D movie of the contrast agent entering and subsequently leaving the organ.
  • this modality is mainly used for voxelwise analysis of perfusion parameters such as blood flow, blood volume, transit time etc.
  • Embodiments of the present invention enable analyzation of these images in a more global fashion and introduce methods to infer the connectivity of the vascular structure underlying the perfusion— even if the vasculature itself is on a subvoxel scale.
  • Embodiments of the present invention advantageously enable several new applications for CTP.
  • Embodiments of the present invention enables one to investigate the inter-subject variability in vascular connectivity, how this changes during for example a stroke or whether there is correlation with incidence and outcome of vascular pathology.
  • Embodiments of the present invention in addition provide overlay of the perfusion territory over the original images or over other images in the same coordinate system.
  • the overlay can, amongst others, be the boundary of the territories or a color overlay.
  • Embodiments of the present invention further may comprise visualization means, quantification means adapted and reporting means.
  • the present invention provides visualization of the arrival time image as a 3D iso-surface at an arrival time threshold.
  • the arrival time is either interactively controlled by the user, or advances automatically through a range of values. This allows the user to advantageously see the contrast agent front evolve throughout the imaged volume.
  • Fig. 1 illustrates the relation between the contrast enhancement during the arterial phase and perfusion territory volume (PTV).
  • PTV perfusion territory volume
  • Fig. 2 illustrates the perfusion territories of the anterior cerebral arteries (ACAs), the left and right middle cerebral arteries (MCAs) and the left and right posterior cerebral arteries (PCAs).
  • ACAs anterior cerebral arteries
  • MCAs middle cerebral arteries
  • PCAs posterior cerebral arteries
  • FIG. 3 illustrates a cerebral CT perfusion, as obtained in a method according to an embodiment of the present invention, illustrating an axial slice of a cerebral CTP during the late arterial phase and a time series of the markers used.
  • FIG. 4 illustrates the relation between the contrast enhancement during the arterial phase (left column) and perfusion territory volume (PTV), as can be obtained using a method according to embodiments of the present invention (middle column) and an alternative method.
  • FIG. 5 illustrates the perfusion territories of the anterior cerebral arteries (ACAs), the left and right middle cerebral arteries (MCAs) and the left and right posterior cerebral arteries (PCAs) for a healthy patient determined using a method according to an embodiment of the present invention (upper row) and a known state of the art technique (lower row).
  • ACAs anterior cerebral arteries
  • MCAs middle cerebral arteries
  • PCAs posterior cerebral arteries
  • FIG. 6 illustrates the perfusion territories of the anterior cerebral arteries (ACAs), the left and right middle cerebral arteries (MCAs) and the left and right posterior cerebral arteries (PCAs) for a patient having a stenosis in a distal MCA segment determined using a method according to an embodiment of the present invention (upper row) and a known state of the art technique (lower row).
  • ACAs anterior cerebral arteries
  • MCAs middle cerebral arteries
  • PCAs left and right posterior cerebral arteries
  • FIG. 7 illustrates the perfusion territories of the anterior cerebral arteries (ACAs), the left and right middle cerebral arteries (MCAs) and the left and right posterior cerebral arteries (PCAs) for a patient with a tumor determined using a method according to an embodiment of the present invention (upper row) and a known state of the art technique (lower row).
  • ACAs anterior cerebral arteries
  • MCAs middle cerebral arteries
  • PCAs posterior cerebral arteries
  • FIG. 8 illustrates the perfusion territories of the anterior cerebral arteries (ACAs), the left and right middle cerebral arteries (MCAs) and the left and right posterior cerebral arteries (PCAs) for a patient with a stenosis in the first segment of the right MCA (Ml -segment) determined using a method according to an embodiment of the present invention (upper row) and a known state of the art technique (lower row).
  • ACAs anterior cerebral arteries
  • MCAs middle cerebral arteries
  • PCAs left and right posterior cerebral arteries
  • a device comprising means A and B should not be limited to devices consisting only of components A and B. It means that with respect to the present invention, the only relevant components of the device are A and B.
  • the terms first, second, third and the like in the description and in the claims are used for distinguishing between similar elements and not necessarily for describing a sequential or chronological order. It is to be understood that the terms so used are interchangeable under appropriate circumstances and that the embodiments of the invention described herein are capable of operation in other sequences than described or illustrated herein.
  • the terms top, bottom, over, under and the like in the description and the claims are used for descriptive purposes and not necessarily for describing relative positions.
  • a method for inferring the arterial connectivity from a CTP sequence is discussed.
  • embodiments of the present invention are not limited to methods where a cost is determined for each path individually, but alternatively also embodiments wherein an overall cost is taken into account can be used.
  • reference is made to optimizing a cost function reference may be made to minimizing a cost, minimizing energy, maximizing fitness, maximizing probability, etc. Reference therefore alternatively can be made to optimizing an objective function.
  • reference is made to arterial volume elements or venous volume elements reference is made to volume elements that contain all or a part of an artery respectively vein.
  • An acceptable edge set can only contain arborescences, i.e. there is only a single directed path between a source and a perfused voxel and 2) Instead of optimizing for
  • the approximation for taking into account a cost of each path independently can be avoided by formulating a global cost function, i.e. one that is based on the complete connectivity E.
  • cost(E) ⁇ ; ⁇ 5 ⁇ ⁇ 3 ⁇ 4)
  • the cost for an edge could be:
  • cost(ei j ) ⁇ max (0, arrival ⁇ pi)— arrival(p j ) ⁇
  • the set Eall contains all potential edges plus an edge between the virtual vertex v v and each arterial source:
  • each volume element or each perfused volume element typically may refer to each volume element of interest or each perfused volume element of interest.
  • the present invention relates to a method for identifying blood flow trajectories in an imaging volume.
  • the method according to embodiments of the present invention can be for identifying of arterial blood flow trajectories or to venous blood flow trajectories.
  • the method according to embodiments of the present invention comprises obtaining for each volume element of a plurality of volume elements into which the imaging volume is divided, arrival or departure time data representing respectively an estimated arrival or departure time of an administered contrast agent at the volume element whereby each volume element indicating the administered contrast agent is a perfused volume element.
  • the method is not limited thereto and could furthermore comprise also identifying non-arterial or non- venous volume elements with respect to the peripheral region and identifying blood flow trajectories (next to the identification of arterial or venous volume elements). In some embodiments this could for example be implemented by first identifying and defining blood flow trajectories to/from the periphery and thereafter limiting the periphery to arterial/venous voxels.
  • peripheral region typically reference is made to the peripheral region of the imaging volume.
  • the peripheral region referred to may be the peripheral region of the region of interest.
  • the imaging volume considered may correspond with the imaging volume corresponding with the region of interest rather than the physically imaged imaging volume.
  • Embodiments of the present invention models the arterial and venous connections separately. Although in the description most embodiments are directed to the arterial case, the same approach is applicable for the veins, and consequently the claims and embodiments of the present invention encompass both the arterial case and the veins.
  • E * argmax Ee ⁇ E] logP(E ⁇ I) Eq. (1)
  • ⁇ E ⁇ is the set of acceptable edge sets.
  • V s C V be the set of arterial voxels on the image border; these are the sources of blood in the imaged region.
  • An edge set E is acceptable if every perfused voxel is reachable from a source v s G V s and the only perfused voxels without an incoming edge are those in V s .
  • vascular anastomoses these are connections between two vascular trees, such as for example present in the Circle of Willis. If there is an anastomosis, only one connection will be modeled.
  • G contains now only tree-like structures.
  • path(vi) (v s , .. . , Vj, . . . , Vi).
  • path(v j ) (v s , .. . , Vj, . . . , Vi).
  • E ⁇ (path(vi) j , path(vi) j+ i)lvi C V, 1 ⁇ j ⁇ (Ipath(vi)l-l) ⁇ .
  • PT(vi) ⁇ V j I Vi G path(v j ) ⁇ .
  • COSt(p) ⁇ 2 ⁇ j ⁇ p Vj) (l - A(pj)), Eq. (5) with A a binary arterial segmentation and "dist" giving the Euclidean distance between a pair of voxels.
  • Advantageously embodiments of the present invention enable a very accurate estimation of the arrival time, as it uses the time to half max (THM), whose determination is better conditioned than the frequently used time to peak (TTP). This is contrary to prior art methods like for instance that of Christensen et al in "Inferring origin of vascular supply from tracer arrival timing patterns using bolus tracking MRI" in J. Magn. Reson. Imaging. 2008 June;27(6):1371-81, which describe that the first moment of the time series is the best metric. The latter metric however, due to recirculation of the contrast agent, will be heavily influenced by the tail of the time series and will overestimate the arrival time by a variable factor depending on the length of the tail.
  • the time series of the THM metric is much flatter at the maximum than at its half maximum, making determination of the exact time when the maximum is reached much more susceptible to noise.
  • Each dataset was acquired on a Toshiba Acquilion One, which comprised 19 time points and had a voxel size of 0.443x0.443x0.5mm 3 , resampled to 0.886x0.886x1.0 mm 3 to speed up calculation and reduce noise.
  • motion correction is performed: every image I t with t > 0 is rigidly aligned to the first image I t using the Elastix Toolbox as described by Klein et al in IEEE Transactions on Medical Imaging 29(1), 196-205 (January 2010).
  • a noise processing step is performed whereby noise is suppressed through both spatial and temporal filtering.
  • Fig. 1 illustrates results obtained by our method using either THM or TTP for arrival estimation, showing that the proposed metric used in embodiments of the present invention outperforms the latter.
  • Results for the healthy subject are shown in Fig. 2.
  • the boundaries produced by the proposed method are more crotchety, the results are more in concordance with the medical literature: the posterior inferior part of the brain is supplied by the PCAs, there is symmetry between the left and right PCA, and the superior posterior part of the brain is supplied by the ACA.
  • vascular pathology such as stenosis, prediction of the territories based on the vasculature becomes even more unreliable.
  • the lenticulostriate perforators which branch off from the Al and Ml segment, provide blood to the putamen.
  • the fraction of putamen volume perfused by the lenticulostriate perforators is in the four datasets: 0.70, 0.80, 0.57 and 0.93.
  • the method of Selle et al. finds 0.31, 0.49, 0.38 and 0.56, which is significantly lower.
  • Embodiments of the present invention provide a method to infer the arterial perfusion paths.
  • embodiments of the present invention use an improved formulation and careful estimation of the arrival times, which results in tracing the blood flow trajectories back to the arterial sources.
  • the main advantage of our method over the competitive region growing approach of WO2013/159147, is the ability to determine the territory of any single artery or even voxel, without prior segmentation of all 'competing' arteries or voxels.
  • the same formulation or principles could be applied to infer the venous trajectories.
  • the venous trees are rooted in the contrast sinks i.e. the venous voxels on the image border and there is a directed edge ⁇ 3 ⁇ 4 if voxel Vi receives venous blood from voxel Vj.
  • the proposed cost function bears close resemblance to doing a watershed segmentation on the arrival time image, with the extension that the flow path of the water is stored.
  • anatomical knowledge about the vascular connectivity can be included: e.g. there is no arterial connectivity between the two hemispheres. This can be enforced by excluding edges (vi, Vj) with Vi and Vj in different hemispheres from the acceptable edge sets. Dynamic programming can naturally handle such constraints.
  • the condition number of a numerical problem measures states how much the result f(x) can vary due to a variation of the input x. It is a measure of the sensitivity of the problem to errors (noise) on the input. It is defined as:
  • the condition numbers of the determination of time to half max (THM) and the determination of time to peak (TTP) are compared. Hereto the determination of THM and TTP was expressed as functions.
  • the input x represents the time series in a voxel.
  • the function y x (t) gives the value at time
  • the first term states how the TTP will change by changing the value of i-th point of the timeseries, given that one knows the maximum value of the timeseries.
  • the second term gives how the TTP changes due to a variation of the estimated maximum value caused by a change of the value of i-th point of the timeseries.
  • max(y x ) or 0.5 max(y x ) depends on the type of interpolation. It is however expected to be relatively small. E.g. for piecewise linear interpolation, it will be smaller than or equal to 1.
  • FIG. 3 is an illustration of a cerebral CT perfusion, whereby in image (a) an axial slice of a cerebral CTP during the late arterial phase is shown. The marker on the top left is located in the grey matter, the marker at the top right is located in an artery and the marker at the bottom right is located in a vein.
  • the graph (b) illustrates a time series in the markers.
  • Another consideration relates to the perfusion territory volume.
  • FIG. 4 Three dimensional renderings of the subtraction image during the arterial phase and of the logarithm of the PTV calculated using the proposed THM and using the TTP (Fig. 4) are shown.
  • a brain mask suppresses the blood vessels in the skin and the base of the skull.
  • the subtraction image is rendered at its orginal resolution of 0.443x0.443x0.5mm3 and the PTV images are supersampled
  • FIG. 4 the relation between the contrast enhancement during the arterial phase and the perfusion territory volume (PTV) is shown.
  • Three dimensional renderings of a subtraction image during the arterial phase (left column) and of the logarithm of the PTV calculated using the proposed THM (central column) and TTP (right column) are shown.
  • the first row is a patient with a tumor hence the large middle meningeal arteries.
  • the second row is a patient with a stenosis in the first segment of the right MCA (Ml -segment).
  • the third row is a healthy patient.
  • the fourth row is a patient with a stenosis in a distal MCA segment.
  • Fig. 5 shows the result for a healthy patient.
  • the perfusion territories of the anterior cerebral arteries (ACAs), the left and the right middle cerebral arteries (MCAs) and the left and right posterior cerebral arteries (PCAs) are shown.
  • Three axial slices are shown in a healthy patient, the top row being calculated by the proposed method according to an embodiment of the present invention and the bottom row by the method of Selle et al. "Analysis of vasculature for liver surgical planning", IEEE Transactions on Medical Imaging 21 (11), 1344-1357 (November 2002).
  • the posterior inferior part of the brain is supplied by the PCAs, there is symmetry between the left and right PCA, and the superior posterior part of the brain is supplied by the ACA, as described by Tatu et al in "Arterial territories of the human brain", Neurology 50(6) 1699-1708 (1998).
  • Fig. 6 shows the result for a patient with a stenosis in a distal MCA segment.
  • the perfusion territories of the anterior cerebral arteries (ACAs), the left and the right middle cerebral arteries (MCAs) and the left and right posterior cerebral arteries (PCAs) are shown.
  • Three axial slices are shown in a patient with a stenosis in a distal MCA segment, the top row being calculated by the proposed method according to an embodiment of the present invention and the bottom row by the method of Selle et al.
  • Selle et al "Analysis of vasculature for liver surgical planning", IEEE Transactions on Medical Imaging 21 (11), 1344-1357 (November 2002).
  • the results obtained by the proposed method are in concordance with medical literature and largely similar to those obtained with the method of Selle et al.
  • Fig. 7 shows the results for a patient with a tumor.
  • the perfusion territories of the anterior cerebral arteries (ACAs), the left and the right middle cerebral arteries (MCAs) and the left and right posterior cerebral arteries (PCAs) are shown.
  • Three axial slices are shown in a patient with a tumor, the top row being calculated by the proposed method according to an embodiment of the present invention and the bottom row by the method of Selle et al. "Analysis of vasculature for liver surgical planning", IEEE Transactions on Medical Imaging 21 (11), 1344-1357 (November 2002).
  • the perfusion territories of the MCAs are very small, while the perfusion territories of the middle meningeal arteries are very large (coloured in grey).
  • the top left image in Fig. 7 shows that this patient has exceptionally large middle meningeal arteries. It is however difficult to assess whether this result is correct.
  • Fig. 8 shows the result for a patient with a stenosis in the first segment of the right MCA (Ml -segment), more particularly the perfusion territories of the anterior cerebral arteries (ACAs), the left and the right middle cerebral arteries (MCAs) and the left and right posterior cerebral arteries (PCAs) are shown.
  • ACAs anterior cerebral arteries
  • MCAs middle cerebral arteries
  • PCAs left and right posterior cerebral arteries
  • Three axial slices are shown, the top row being calculated by the proposed method according to an embodiment of the present invention the bottom row by the method of Selle et al. "Analysis of vasculature for liver surgical planning", IEEE Transactions on Medical Imaging 21 (11), 1344-1357 (November 2002).
  • the part of the right hemisphere that in a healthy person would be perfused by the MCA belongs now to the ACA and PCA territory. This is in concordance with medical literature, which describes leptomeningeal anastomoses that cause such a pattern, as known from Liebeskind, "Collateral circulation” in Stroke 34(9) 2279-2284 (September 2003).
  • the present invention also relates to a controller for performing a method as described in embodiments of the first aspect of the present invention.
  • the controller may be implemented in software or hardware.
  • Features and advantages may correspond with a feature or step of a method according to an embodiment of the first aspect.
  • the present invention also relates to a computer program product for, when executed on a processor, identifying blood flow trajectories in an imaging volume.
  • a computer program product may be programmed for
  • each volume element indicating the administered contrast agent is a perfused volume element
  • the computer program product may be stored on a processor.
  • a processor may for example include at least one programmable computing component coupled to a memory subsystem that includes at least one form of memory, e.g., RAM, ROM, and so forth.
  • the computing component or computing components may be a general purpose, or a special purpose computing component, and may be for inclusion in a device, e.g., a chip that has other components that perform other functions.
  • a device e.g., a chip that has other components that perform other functions.
  • one or more aspects of the present invention can be implemented in digital electronic circuitry, or in computer hardware, firmware, software, or in combinations of them.
  • each of the method steps may be a computer implemented step.
  • a processor as such is prior art
  • a system that includes the instructions to implement aspects of the methods for identifying blood trajectories in an imaging volume is not prior art.
  • the present invention thus also includes a computer program product which provides the functionality of any of the methods according to the present invention when executed on a computing device.
  • the present invention relates to a data carrier for carrying a computer program product for identifying blood trajectories in an imaging volume.
  • a data carrier may comprise a computer program product tangibly embodied thereon and may carry machine-readable code for execution by a programmable processor.
  • the present invention thus relates to a carrier medium carrying a computer program product that, when executed on computing means, provides instructions for executing any of the methods as described above.
  • carrier medium refers to any medium that participates in providing instructions to a processor for execution.
  • Such a medium may take many forms, including but not limited to, non-volatile media, and transmission media.
  • Non-volatile media includes, for example, optical or magnetic disks, such as a storage device which is part of mass storage.
  • Computer readable media include, a CD-ROM, a DVD, a flexible disk or floppy disk, a tape, a memory chip or cartridge or any other medium from which a computer can read.
  • Various forms of computer readable media may be involved in carrying one or more sequences of one or more instructions to a processor for execution.
  • the computer program product can also be transmitted via a carrier wave in a network, such as a LAN, a WAN or the Internet.
  • Transmission media can take the form of acoustic or light waves, such as those generated during radio wave and infrared data communications. Transmission media include coaxial cables, copper wire and fibre optics, including the wires that comprise a bus within a computer.
  • the present invention also relates to a system for identifying blood trajectories.
  • the system comprises pre-processing means programmed for obtaining for each volume element of a plurality of volume elements into which the imaging volume is divided, arrival or departure time data representing respectively an estimated arrival or departure time of an administered contrast agent at the volume element whereby each volume element indicating the administered contrast agent is a perfused volume element.
  • It also comprises an identifying means programmed for identifying arterial respectively venous volume elements towards peripheral regions of the imaging volume and for defining for each perfused volume element a blood flow trajectory from one of said arterial volume elements respectively towards said venous volume elements.
  • Defining blood flow trajectories comprises minimizing for all volume element the cost of their paths from a peripheral arterial volume element to the perfused volume element respectively the cost of their paths to a peripheral venous element. Further optional features of embodiments of the present invention may correspond with components having the functionality of one or more method steps of a method according to an embodiment of the first aspect.
  • the present invention relates to a volumetric image of blood flow trajectories obtained by a method as described above.
  • the volumetric image of blood flow trajectories, which visualizes the blood flow trajectories may for example be a visualization of the colored perfusion territories that are inspected slice by slice.
  • Another volumetric image of blood flow trajectories may be a volume renderings of the perfusion territory volumes (where every voxel has a value representative for its perfusion territory volume.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Apparatus For Radiation Diagnosis (AREA)

Abstract

L'invention concerne un procédé destiné à identifier des trajectoires d'écoulement de sang dans un volume d'imagerie. Le procédé comprend les étapes consistant à obtenir, pour chaque élément de volume parmi une pluralité d'éléments de volume en lesquels le volume d'imagerie est divisé, des données de temps d'arrivée ou de départ représentant respectivement un temps d'arrivée ou de départ estimé au niveau de l'élément de volume d'un agent de contraste administré, moyennant quoi chaque élément de volume indiquant l'agent de contraste administré est un élément de volume perfusé, identifier des éléments de volume artériel, respectivement veineux, vers des régions périphériques du volume d'imagerie et définir, pour chaque élément de volume perfusé, une trajectoire d'écoulement de sang à partir de l'un desdits éléments de volume artériel respectivement vers lesdits éléments de volume veineux, ladite définition des trajectoires d'écoulement de sang comprenant l'optimisation, pour tous les éléments de volume, d'une fonction de coût de leurs trajets, d'un élément de volume artériel périphérique vers l'élément de volume perfusé, respectivement une fonction de coût de leurs trajectoires vers un élément veineux périphérique.
PCT/EP2016/070842 2015-09-04 2016-09-05 Procédé et dispositif d'identification de la connectivité vasculaire d'une image WO2017037287A1 (fr)

Applications Claiming Priority (6)

Application Number Priority Date Filing Date Title
GBGB1515661.5A GB201515661D0 (en) 2015-09-04 2015-09-04 Method and device for identifying vascular connectivity of an image
GBGB1515660.7A GB201515660D0 (en) 2015-09-04 2015-09-04 Method and device for identifying vascular connectivity of an image
GB1515661.5 2015-09-04
GB1515660.7 2015-09-04
GB1517063.2 2015-09-28
GBGB1517063.2A GB201517063D0 (en) 2015-09-28 2015-09-28 Method and device for identifying vascular connectivity of an image

Publications (1)

Publication Number Publication Date
WO2017037287A1 true WO2017037287A1 (fr) 2017-03-09

Family

ID=56920690

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/EP2016/070842 WO2017037287A1 (fr) 2015-09-04 2016-09-05 Procédé et dispositif d'identification de la connectivité vasculaire d'une image

Country Status (1)

Country Link
WO (1) WO2017037287A1 (fr)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113792740A (zh) * 2021-09-16 2021-12-14 平安科技（深圳）有限公司 眼底彩照的动静脉分割方法、系统、设备及介质
CN114066969A (zh) * 2021-04-23 2022-02-18 数坤(北京)网络科技股份有限公司 一种医学图像分析方法以及相关产品

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2013159147A1 (fr) * 2012-04-24 2013-10-31 The University Of Melbourne Identification de territoires vasculaires
WO2014162273A1 (fr) * 2013-04-03 2014-10-09 Koninklijke Philips N.V. Segmentation de vaisseau

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2013159147A1 (fr) * 2012-04-24 2013-10-31 The University Of Melbourne Identification de territoires vasculaires
WO2014162273A1 (fr) * 2013-04-03 2014-10-09 Koninklijke Philips N.V. Segmentation de vaisseau

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
ANDREAS FIESELMANN ET AL: "Deconvolution-Based CT and MR Brain Perfusion Measurement: Theoretical Model Revisited and Practical Implementation Details", INTERNATIONAL JOURNAL OF BIOMEDICAL IMAGING, vol. 26, no. 6, 31 January 2011 (2011-01-31), pages 1 - 20, XP055128513, ISSN: 1687-4188, DOI: 10.3174/ajnr.A1274 *
ROBBEN DAVID ET AL: "Perfusion Paths: Inference of Voxelwise Blood Flow Trajectories in CT Perfusion", NETWORK AND PARALLEL COMPUTING; [LECTURE NOTES IN COMPUTER SCIENCE; LECT.NOTES COMPUTER], SPRINGER INTERNATIONAL PUBLISHING, CHAM, vol. 9350 Chap.49, no. 558, 20 November 2015 (2015-11-20), pages 407 - 414, XP047326937, ISSN: 0302-9743, ISBN: 978-3-642-36762-5, [retrieved on 20151120] *
SØREN CHRISTENSEN ET AL: "Inferring origin of vascular supply from tracer arrival timing patterns using bolus tracking MRI", JOURNAL OF MAGNETIC RESONANCE IMAGING, vol. 27, no. 6, 26 June 2008 (2008-06-26), US, pages 1371 - 1381, XP055264633, ISSN: 1053-1807, DOI: 10.1002/jmri.21386 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114066969A (zh) * 2021-04-23 2022-02-18 数坤(北京)网络科技股份有限公司 一种医学图像分析方法以及相关产品
CN114066969B (zh) * 2021-04-23 2022-04-26 数坤(北京)网络科技股份有限公司 一种医学图像分析方法以及相关产品
CN113792740A (zh) * 2021-09-16 2021-12-14 平安科技（深圳）有限公司 眼底彩照的动静脉分割方法、系统、设备及介质
CN113792740B (zh) * 2021-09-16 2023-12-26 平安创科科技(北京)有限公司 眼底彩照的动静脉分割方法、系统、设备及介质

Similar Documents

Publication Publication Date Title
US11399779B2 (en) System-independent quantitative perfusion imaging
JP2022169579A (ja) リアルタイムで診断上有用な結果
WO2018120644A1 (fr) Procédé et système d'extraction de vaisseau sanguin
JP5081390B2 (ja) 腫瘍量を監視する方法及びシステム
JP6484760B2 (ja) 非侵襲的血流予備量比（ｆｆｒ）に対する側副血流モデル化
CN108122616B (zh) 个体特异性的心血管模型的生成方法及其应用
JP2019500960A (ja) 心臓モデルガイド付き冠動脈瘤セグメンテーション
Arias-Lorza et al. Carotid artery wall segmentation in multispectral MRI by coupled optimal surface graph cuts
JP2014501126A (ja) 組織病変部の輪郭描出方法
JP5801316B2 (ja) 副側血流評価
CN109431531A (zh) 基于灌注成像的血管分割方法及装置和计算机装置
JP2024515635A (ja) 超音波及びカメラ画像から３ｄ画像を再構成するためのシステム及び方法
US20120209108A1 (en) System and Method for Treatment Planning of Organ Disease at the Functional and Anatomical Levels
CN109431532A (zh) 基于灌注成像的动静脉分离方法及装置和计算机装置
US20220151500A1 (en) Noninvasive quantitative flow mapping using a virtual catheter volume
Lidayova et al. Skeleton-based fast, fully automated generation of vessel tree structure for clinical evaluation of blood vessel systems
Luo et al. IVUS validation of patient coronary artery lumen area obtained from CT images
WO2017037287A1 (fr) Procédé et dispositif d'identification de la connectivité vasculaire d'une image
US9058651B2 (en) System and methods for functional analysis of soft organ segments in SPECT-CT images
Tönnes et al. Deterministic Arterial Input Function selection in DCE-MRI for automation of quantitative perfusion calculation of colorectal cancer
Racimora et al. Segmentation of polycystic kidneys from MR images
Robben et al. Perfusion paths: Inference of voxelwise blood flow trajectories in CT perfusion
Lee et al. Tumor segmentation from small animal PET using region growing based on gradient magnitude
Cho et al. Aortic annulus detection based on deep learning for transcatheter aortic valve replacement using cardiac computed tomography
Zhu et al. Automatic centerline extraction of cerebrovascular in 4D CTA based on tubular features

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16763748

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 16763748

Country of ref document: EP

Kind code of ref document: A1