DE102010009701A1 - Method for automatically identifying narrow points of coronary blood vessel in heart of patient, involves identifying narrow points in hollow vessel based on automated evaluation of profile course curve of hollow vessel - Google Patents

Abstract

The method involves assigning a profile course curve to a hollow vessel based on a cross section area of an idle lumen of the hollow vessel and/or a diameter of the hollow vessel. Narrow points in the hollow vessel are identified based on an automated evaluation of the profile course curve of the hollow vessel. A location or a portion of the hollow vessel is identified if an actual value of the cross section area is more than a preset value. The identified narrow points are prioritized based on the cross section area, the diameter and length of the narrow points. Independent claims are also included for the following: (1) a computing unit for automatically identifying narrow points of a hollow vessel (2) a data carrier comprising a set of instructions for executing a method for automatically identifying narrow points of a hollow vessel.

Description

The invention relates to a method for the automated identification of at least one constriction of at least one hollow vessel, in particular a blood vessel. The invention also relates to a computing unit for carrying out the method and to a data carrier which has a program implementing the method.

The examination of patients' hollow vessels for anomalies using imaging devices is of great importance in everyday clinical practice. Thus CT or MR angiographies for examining the vascular system of a patient make up about 80% of the daily examinations with a computed tomography device or an MR device in a clinic. In almost every angiography, a plurality of images are generated, for example, in a CT angiography up to 3000 slice images are generated, which must be viewed to find anomalies, especially for finding narrowing of the vasculature or the blood vessels of the vascular system. In patients with atherosclerotic vascular disease, there are often many blood vessels with constrictions and sometimes severe occlusions, all of which must be found in the slice images and identified on the basis of the slice images.

The finding and findings of constrictions of hollow vessels is time consuming and still requires many manual steps, eg. For example, the sequential sifting of the generated slices, the generation of suitable multiplanar reformations, the manual or the cause of the automated creation of center lines in hollow vessels to induce the creation of suitable Curved Planar Reformations, etc. to find the narrowings, can be suitably reviewed and to be able to judge.

The invention is therefore based on the object of specifying a method, a computing unit and a data carrier of the type mentioned above such that at least the prerequisite for a faster diagnosis of a narrowing of a hollow vessel of a living being is created.

According to the invention, this object is achieved by a method for automatically identifying at least one constriction of at least one hollow vessel of at least one animal according to claim 1. The starting point of the method is a reconstructed image data set which has the at least one vessel of the at least one animal. As a rule, the at least one hollow vessel is segmented in the image data record. The at least one hollow vessel is associated with a profile profile curve which, based on the cross-sectional area and / or the diameter of the at least one hollow vessel, identifies the free lumen of the at least one hollow vessel. The determination of such a profile curve is incidentally in the DE 10 2009 023 504.3 whose disclosure is included in the present patent application.

The profile progression curve of the at least one hollow vessel is automatically evaluated in such a way that one or more constrictions of the at least one hollow vessel are identified. Due to the automated evaluation of several profile progression curves of several hollow vessels of one or more organisms, multiple constrictions of several hollow vessels can be identified automatically. The constrictions identified in this way can finally be offered or visualized to a medical expert for a faster diagnosis.

According to a variant of the invention, a location or a portion of the at least one hollow vessel is identified as a constriction if the actual profile value of the cross-sectional area at the location or in the section originating from the profile profile curve exceeds more than one of a comparative profile value of the cross-sectional area at the location or in the section a first value or a first percentage and / or if the actual profile value of the diameter at the location or in the section derived from the profile profile curve deviates from a comparison profile value of the diameter at the location or in the section by more than a second value or a second percentage. Preferably, the first percentage is between 45% and 95%, preferably 75%, and / or the second percentage between 25% and 75%, preferably 50%. In the case of such a deviation, a narrowing must be assumed, which must be assessed. The specified percentages are to be understood as examples and may optionally be chosen or specified differently. Incidentally, a total occlusion of a hollow vessel as a constriction is also understood for the present invention.

According to a further variant of the invention, each identified constriction of the at least one hollow vessel is quantified, wherein the cross-sectional area, the diameter and / or the length of the constriction are determined. The determination of the cross-sectional area, the diameter and the length of an identified constriction preferably takes place on the basis of the profile profile curve of the hollow vessel having the constriction.

According to one embodiment of the invention, multiple constrictions in at least one or more vasculature of one or more animals are identified and automated prioritization of the identified constrictions occurs. Of the Automated prioritization of the identified constrictions is based on the idea of sorting the constrictions according to their severity so that a clinician can immediately identify the most severe constrictions and make the findings of those constrictions first.

Variants of the invention provide for the prioritization of the identified constrictions on the basis of the cross-sectional areas, the diameter and / or the lengths of the identified constrictions. Preferably, for the prioritization of a plurality of identified constrictions, a value tuple is formed for each identified constriction from the cross-sectional area, diameter, and length of the identified constriction, and for prioritizing the identified constrictions of the cross-sectional area of an identified constriction, the highest constriction identified is the diameter of the identified constriction second highest and the length of the identified narrowing the third highest importance is attributed. Thus, for example, if two identified constrictions have the same cross-sectional area and diameter, that of the identified constrictions is classified as more severe, comprising a greater length. In this way, an order of the identified constrictions can be made according to their severity.

An embodiment of the invention provides for assigning to each prioritized identified constriction, ie for example to each value tuple, at least one representation of the respective identified constriction, in which the identified constriction is displayed in a selection of the identified constriction. As a rule, each identified narrowing is associated with a plurality of display forms, so that, when selecting or selecting a specific identified narrowing, the different representations showing the identified constriction, for example, are immediately visible. B. be displayed on a viewing device. Preferably, an identified constriction of a hollow vessel in a Curved Planar Reformation (CPR) and / or in two mutually offset by 90 ° Curved Planar Reformations (CPR) and / or as a "Ribbon View" of the identified constriction having hollow vessel is shown. The two Curved Planar Reformations offset by 90 ° are created in order to examine a narrowing from different directions. A "Ribbon View" representation is a kind of elongated Curved Planar Reformations of a hollow vessel, in which the gray or color values of the pixels resulting from the Curved Planar Reformations remain the same, but the representation of the vessel is modified in such a way that the curved course of the hollow vessel is rectilinear or rectilinear. If one considers the hollow vessel as a kind of curved thread, this thread or for the imaging of the hollow vessel is pulled in the length to obtain a rectilinear view of the hollow vessel. Incidentally, the "ribbon-view" representation is also referred to as a straightened or smoothed Curved Planar Reformation.

A further embodiment of the invention provides, in each case, to visualize the profile progression curve of a hollow vessel having at least one identified constriction, preferably together with the mentioned forms of representation of the at least one identified constriction, wherein a marker which identifies the identified constriction is entered in the profile progression curve.

According to a variant of the invention, based on the prioritization of the identified constrictions, a ranking list of the living beings or of the patients is created whose constrictions identified on the basis of their severity are to be given priority. Such a ranking can be kept in a memory of a diagnostic calculator and continuously updated in newly arrived patients.

The object on which the present invention is based is also achieved by a computing unit which is set up to carry out one of the methods described above. The arithmetic unit can be, for example, a workstation for the diagnosis. However, the arithmetic unit can also be part of an imaging device for generating an image data set of at least one hollow vessel of at least one living being, for example a computed tomography device, an MR device or an ultrasound device.

Furthermore, the object underlying the present invention is achieved by a data carrier which has a computer program which implements one of the methods described above, wherein the computer program is stored on the data carrier and can be loaded from a data carrier by the data carrier to one of the methods described above execute when the computer program is loaded in the arithmetic unit. The data carrier can be a portable data carrier, for example a CD, or else a computer from which the computer program can be loaded, for example, via the Internet.

An embodiment of the invention is illustrated in the accompanying schematic drawings. Show it:

1 an X-ray computer tomography device and

2 . 3 a Curved Planar Reformation of a coronary vessel with a profile curve.

In the figures, identical or functionally identical elements, components, fabrics, etc. are provided with the same reference numerals throughout. The illustrations in the figures are schematic and not necessarily to scale, scales may vary between the figures. On the in 1 illustrated X-ray computer tomography device 1 is in the following and without restriction of generality only to the extent that it is considered necessary for the understanding of the invention.

This in 1 X-ray computer tomography device shown 1 has a patient bed 2 for storage of a patient P to be examined. The X-ray computer tomography device 1 also includes a gantry 4 with one around a system axis 5 rotatably mounted tube-detector system. The tube-detector system has an X-ray tube opposite each other 6 and an X-ray detector unit 7 on. In operation goes from the X-ray tube 6 X-rays 8th in the direction of the X-ray detector unit 7 from, and is detected by means of this.

The patient bed 2 has a lying base 9 on, on which provided for the actual storage of the patient P patient support plate 10 is arranged. The patient support plate 10 is so relative to the bed base 9 adjustable that the patient support plate 10 with the patient P in the opening 3 the gantry 4 for taking 2D X-ray projections from the patient P, e.g. In a sequence scan or a spiral scan. The computational processing of the 2D X-ray projections, ie the reconstruction of slice images or a volume data set from a body region of the patient P based on the 2D X-ray projections is performed with a schematically illustrated image computer 11 of the X-ray computer tomography device 1 , The image calculator 11 of the X-ray computer tomography device 1 is in the case of the present embodiment of the invention with a computing unit 12 connected to the the with the image calculator 11 reconstructed layer or volume images are to be viewed and evaluated.

With the X-ray computer tomography device 1 In the case of the present exemplary embodiment of the invention, an examination of the heart H of the patient P, in particular of the coronary vessels, is to be carried out for constrictions or stenoses. This is the arithmetic unit 12 with a corresponding computer program 13 provided, in the present case by means of a portable data carrier 16 or storage medium, for example a CD, in the arithmetic unit 12 was loaded. The computer program 13 sets a procedure for the automated identification and prioritization of narrowing in hollow vessels, here the coronary vessels.

First, with the image calculator 11 of the X-ray computer tomography device 1 in a manner known per se, a set of in the direction of the system axis 5 successively reconstructed axial slice images of the patient's heart H based on recorded 2D x-ray projections having the patient's coronary artery p. This set of layered images is stored in memory 14 stored and thus stands the arithmetic unit 12 available for further processing.

For processing the slice images is the arithmetic unit 12 provided with software with which the coronary arteries of the heart H in the slice images are preferably fully automatically identified and segmented. Possibly. If a manual post-processing of the slices is also possible, not all coronary arteries should have been automatically identified and segmented. Finally, all automatically and possibly manually identified coronary vessels are automatically provided with a center line by the software. Furthermore, the identified coronary vessels are provided with profile progression curves by the software.

This process is exemplary in 2 on the basis of a section of a coronary vessel 20 of the heart H, which has abnormalities in the form of constrictions. For the coronary vessel 20 is, as already mentioned, first the center line 22 certainly. Subsequently, the tubular sheath 23 of the coronary vessel 20 for example, pixel by pixel or voxel by voxel along the course of the coronary artery 20 determines what z. B. can be done using a gray value analysis in the CT image data of the reconstructed slices. Based on the detected shell 23 becomes the profile curve 24 for the coronary vessel 20 developed or derived, which is right next to the image of the coronary artery 20 is shown.

The profile curve 24 in the case of the present exemplary embodiment of the invention, embodies or illustrates the profile of the value of the cross-sectional area of the free lumen of the coronary vessel 20 along the course of the coronary vessel 20 , Wherein the cross-sectional planes belonging to the cross-sectional areas in the case of the present embodiment of the invention are always at right angles to the center line 22 stand as exemplified in 2 registered cross-sectional planes 25 to 29 the case is.

For the cross-sectional planes 25 to 29 are in 2 to the left of the illustration of the Coronary artery 20 the to the cross-sectional planes 25 to 29 belonging to five cross-sectional representations 30 to 34 of the coronary vessel 20 shown.

It thus becomes clear that for each cross-sectional plane, based on the set of slice images, the value of the cross-sectional area of the free lumen of the coronary vessel 20 can be determined, as well as the different diameters of a cross-sectional area, which is not circular in the rule. In the case of the present embodiment of the invention, in addition to the maximum diameter D1, in particular the minimum diameter D2 of a cross-sectional area is determined, which could likewise form the basis for a profile profile curve. The course of the minimum diameter of the coronary vessel 20 along the course of the coronary vessel for an alternative profile curve is the course of the maximum diameter of the coronary vessel 20 along the course of the coronary vessel, since the minimum diameter with respect to a narrowing represents the more critical value.

According to the inventive method are by means of the computer program 13 all profile curve of the coronary arteries of the heart H evaluated for constrictions, the constrictions are identified based on the profile profile curves and then prioritized.

In the case of the present embodiment of the invention, a point or portion of a coronary vessel is identified as a constriction if the actual profile value of the cross-sectional area at the point under consideration or in the section under consideration originates from a comparative profile value of the cross-sectional area at the point under consideration or in the section considered considered by more than 75%.

In order to be able to accomplish this in an automated manner, according to a first embodiment of the invention, an ideal profile value is determined for each coronary vessel as the comparative profile value. For this purpose, on the basis of a regression analysis, in each case an approximation curve is fitted to a profile progression curve, which represents the ideal course of the cross-sectional area of the free lumen of the respective coronary vessel, as shown in FIG DE 10 2009 023 504.3 is described.

In the case of the present embodiment of the invention was an approximation curve in the form of a straight line 35 to the profile curve 24 attached, the ideal course of the cross-sectional area of the free lumen of the coronary artery 20 represents. It can be seen that the free lumen of the coronary artery 20 down to smaller, which is based on the natural rejuvenation of the coronary artery. Now, the course of the profile curve 24 be evaluated automatically without any interaction of a user or an observer on constrictions. Based on the straight line 35 can do this to anyone from the profile curve 24 For the identification of a constriction, the actual profile value for the ideal profile value of the cross-sectional area of a specific location or a specific section of the coronary vessel is set in relation to the actual profile value of a cross-sectional area.

In this way, in the case of the present embodiment of the invention, two constrictions of the coronary vessel 20 identified. The first narrowing will be on the spot 36 identifies the cross-sectional plane 27 assigned. The second narrowing will be on the spot 37 identifies the cross-sectional plane 29 assigned.

According to a second embodiment of the invention, constrictions of a coronary vessel can also be determined on the basis of its profile curve so that the minima of the profile curve of the coronary vessel are first determined. Subsequently, in the immediate vicinity of each minimum, preferably symmetrically about the minimum, a value is determined in front of and a value according to the minimum is calculated from the profile profile curve and from this the mean value is formed. The values are determined at a distance of 1 mm to 5 mm before or after the minimum, so that they are not on the one hand in the potential narrowing and on the other hand are not too far away from the minimum and thus are no longer representative of the vascular site. This average value represents for each respective minimum the comparison profile value to which the respective minimum is set as an actual profile value.

Alternatively, the values in the immediate vicinity of each minimum, that is, for example, a value before and a value after the minimum outside the potential narrowing can be determined such that the profile profile curve 24 with the straight line 35 and those values before and after the minimum are used to determine the mean outside the constriction. For example, the intersections of the profile curve could 24 with the straight line 35 be used in the vicinity of a minimum for determining the mean value or the comparative profile value.

If the actual profile value deviates more than 75% from the comparative profile value, there is a narrowing. Also in this way can the first narrowing 36 and the second constriction 37 be identified.

For each of the two constrictions 36 and 37 is in each case a value tuple from the associated cross-sectional area, the minimum diameter of the cross-sectional area and in the direction of the center line 22 measured length of the respective constriction formed. The value of the cross-sectional area can be in each case directly from the profile curve 24 remove. The minimum diameter for each constriction has also been determined as part of determining the cross-sectional area and is also present for both constrictions. The length of a narrowing can also be determined from the profile curve 24 determining, as it progresses along the course of the coronary vessel, how long the condition for the presence of a constriction is fulfilled, according to which the actual profile value for the cross-sectional area must deviate more than 75% from the ideal profile value or from the comparative profile value of the cross-sectional area.

Since the inflammatory region of the hollow vessel, which belongs to a stenosis or a constriction, is generally longer than the constriction, the length of the stenosis or constriction can also be determined in such a way that the profile profile curve 24 with the straight line 35 is compared, wherein the length is determined such that the profile profile curve 24 in the area of the constriction by a certain value, z. At least the standard deviation, from the straight line 35 must deviate. This results in two stenosis or narrowing narrowing values that determine their length.

In the case of the present embodiment of the invention, the constrictions 36 and 37 the same cross-sectional area of the free lumen and the same minimum diameter. The narrowing 37 however, has a greater length than the constriction 36 on, so that the value tuple W1 from the cross-sectional area, the minimum diameter and the length of the constriction 37 is prioritized higher than the corresponding value tuple W2 for the constriction 36 ,

If the prioritization has been carried out for all identified constrictions, in the case of the present exemplary embodiment of the invention, in each case two Curved Planar Reformations offset by 90 ° with respect to the center line of the respective coronary vessel are generated, which are linked to the respective profile curve of the respective coronary vessel. Furthermore, a Ribbon View is generated for each coronary vessel.

The prioritized value tuples are, as in 1 is schematically illustrated, according to the order of their priority on the screen 15 presented for a surveyor. The value tuple W1 has the highest priority, the value tuple W2 has the second highest priority, etc. The finder can use the input means, for. B. the keyboard of the arithmetic unit 12 Find each constriction associated with a value tuple starting with the highest priority. The finder is free in the choice of the constriction to be found, he can browse freely through the value tuple and freely choose the constriction that he currently wants to find.

For example, does he choose the constriction 36 for the diagnosis, which in the present case the value tuple W2 is assigned, this becomes the constriction 36 having coronary vessel 20 for example as in 3 shown on the screen 15 shown. In the case of the present embodiment of the invention, therefore, on the left is a Curved Planar Reformation of the coronary artery 20 and on the right the profile curve 24 of the coronary vessel 20 with a steno marker 38 to mark the narrowing 36 shown. Additionally, a second Curved Planar Reformation of the coronary 20 which are at 90 ° to the in 3 shown Curved Planare Reformation is offset. During the first Curved Planar Reformation of the coronary artery 20 For example, a kind of top view of the coronary vessel 20 shows, shows the second Curved Planar Reformation of the coronary artery 20 a view from the side of the coronary vessel 20 ,

Furthermore, in addition, a "Ribbon View" representation of the coronary artery 20 are displayed. In the "Ribbon View" illustration, the coronary artery 20 quasi be elongated or straight, so that its curved course becomes straight or straight.

The values of the cross-sectional area, the minimum diameter and the length of the constriction are also preferred 36 displayed what was in 3 however, this is not the case. The steno marker 38 can be an interactive steno marker. In this case, the stenotic marker may be displaced, for example, with a computer mouse along the profile curve, in each case adjusting the image representations as required and the displayed values of the cross-sectional area and the minimum diameter of the coronary vessel known from the profile curve. The value for the length of a constriction can only be displayed if there is one at the location of the stenosis marker.

Incidentally, the prioritization of the identified constrictions need not necessarily be limited to the identified constrictions of a patient's lumen. Rather, the prioritization may refer to any identified constrictions of hollow organs of various patients that are pending for diagnosis. Of Further, based on the prioritization of the identified constrictions, for example, a patient having the highest grade identified constrictions may be visually highlighted in a report list or ranked first in the order of the report. In this way, an identified narrowing physician can immediately identify the patient (s) to be prioritized.

The invention has been described above using the example of coronary vessels as hollow vessels. However, the application of the invention is not limited to coronary arteries. Rather, other blood-carrying vessels of a living or other hollow vessels of a living being can be examined with the method according to the invention for the presence of constrictions.

Incidentally, the image data set underlying the identification and prioritization of constrictions need not necessarily be generated with an X-ray computer tomograph, but may also be generated with an MR device or an ultrasound device.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DE 102009023504 [0005, 0031]

Claims (15)

Method for the automated identification of at least one constriction ( 36 . 37 ) at least one hollow vessel ( 20 ) at least one living being (P), based on a the at least one hollow vessel ( 20 ) of the at least one living being (P) having, reconstructed image data set, wherein the at least one hollow vessel ( 20 ) a profile profile curve ( 24 ), which, based on the cross-sectional area and / or the diameter of the at least one hollow vessel ( 20 ) the free lumen of the at least one hollow vessel ( 20 ), based on an automated evaluation of the profile curve ( 24 ) of the at least one hollow vessel ( 20 ) at least one constriction ( 36 . 37 ) of the at least one hollow vessel ( 20 ) is identified. The method of claim 1, wherein a location or a portion of the at least one hollow vessel ( 20 ) as a narrowing ( 36 . 37 ) is identified when the profile profile curve ( 24 ) is an actual profile value of the cross-sectional area at the location or in the section of a comparative profile value of the cross-sectional area at the location or in the section by more than a first value or a first percentage and / or if the profile profile curve ( 24 ), the actual profile value of the diameter at the location or in the section deviates from a comparative profile value of the diameter at the location or in the section by more than a second value or a second percentage. The method of claim 2, wherein the first percentage is between 45% and 95% and / or the second percentage is between 25% and 75%. Method according to one of claims 1 to 3, wherein each identified constriction ( 36 . 37 ) of the at least one hollow vessel ( 20 ), whereby the cross-sectional area, the diameter and / or the length of the constriction are determined. Method according to one of claims 1 to 4, wherein a plurality of constrictions ( 36 . 37 ) in at least one or more hollow vessels ( 20 ) of one or more living beings (P) and in which a prioritization of the identified constrictions ( 36 . 37 ) he follows. Method according to claim 5, wherein the prioritization is based on the cross-sectional area of an identified constriction ( 36 . 37 ). Method according to claim 5 or 6, wherein the prioritization is based on the diameter of an identified constriction ( 36 . 37 ). Method according to one of Claims 5 to 7, in which the prioritization is based on the length of an identified constriction ( 36 . 37 ). Method according to claim 8, wherein for the prioritization of a plurality of identified constrictions ( 36 . 37 ) for each identified constriction ( 36 . 37 ) a value tuple of the cross-sectional area, the diameter and the length of the respective identified constriction ( 36 . 37 ), whereby for the prioritization of the identified constrictions ( 36 . 37 ) of the cross-sectional area of an identified constriction ( 36 . 37 ) the highest, the diameter of the identified constriction ( 36 . 37 ) the second highest and the length of the identified narrowing ( 36 . 37 ) the third highest importance is attributed. Method according to one of claims 5 to 9, wherein each prioritized identified constriction ( 36 . 37 ) at least one representation of the respective identified narrowing ( 36 . 37 ), in which the identified constriction ( 36 . 37 ) in a selection of the identified narrowing ( 36 . 37 ) is pictured. Method according to claim 10, wherein an identified constriction ( 36 . 37 ) of a hollow vessel ( 20 ) in a Curved Planar Reformation (CPR) and / or in two Curved Planar Reformations (CPR) offset by 90 ° and / or as a "Ribbon View" of the identified narrowing ( 36 . 37 ) having hollow vessel ( 20 ) is pictured. Method according to one of claims 1 to 11, wherein in each case the profile curve ( 24 ) of at least one identified constriction ( 36 . 37 ) having hollow vessel ( 20 ) in which a marker ( 38 ), which identifies the identified constriction ( 36 . 37 ). Method according to one of Claims 5 to 12, in which, based on the prioritization of the identified constrictions, a ranking list of the animals whose identified constrictions are to be given priority is drawn up. Arithmetic unit ( 12 ), which is adapted to carry out a method according to one of claims 1 to 13. Disk ( 16 ), comprising a computer program implementing a method according to one of claims 1 to 13 ( 13 ), which on the disk ( 16 ) and stored by a computing unit ( 12 ) from the disk ( 16 ) in order to carry out a method according to one of claims 1 to 13, when the computer program ( 13 ) in the arithmetic unit ( 12 ) is loaded.

Images