DE102013202063A1 - Method for determining characteristic value of hemodynamic state in blood vessel, involves taking temporal intensity flows of gray values of fixed image point coordinates of angiographic image series and transforming image point coordinates - Google Patents

Abstract

The method (1) involves taking the temporal intensity flow of gray values of a fixed image point coordinate of an angiographic image series and the temporal intensity flow of gray values of another fixed image point coordinates of another angiographic image series while determining (S8) the characteristic value. The image point coordinates of gray values are transformed (S5) in the frequency area, and the temporal intensity flows of gray values are taken while determining the characteristic value. The ratio of amplitudes of given frequencies of the frequency spectrums of the intensity flows of gray values of the angiographic image series are taken during determination of the characteristic value. An independent claim is included for a device for determining a characteristic value of a hemodynamic state in a vessel.

Description

The present invention relates to a method for determining a characteristic of a hemodynamic condition in a vessel. Moreover, the present invention relates to a corresponding device for determining a characteristic of a hemodynamic state in a vessel.

Angiographic recordings are an essential tool in certain endovascular treatment procedures in order to be able to estimate the success of the intervention during the intervention. Based on subtraction angiographic image series, for example, during treatment of a cerebral aneurysm with one or more special endoluminal vascular prostheses, an assessment of the induced hemodynamic changes by the attending physician must be made. In this connection, important questions are, in particular, to what extent and at which point in time an occlusion of the aneurysm takes place through a thrombus and whether the thrombus that is forming will be stable. It also assesses whether the ongoing treatment is already promising and can be completed. Such an assessment is essential, above all, in the case of therapeutic measures which do not lead to an immediate separation of the aneurysm from the blood flow in the afferent vessels.

The estimation of an expected therapeutic result has previously been carried out, for example, by a treating physician by visually assessing one or more subtraction angiographic image series. Here, for example, an angiographic image taken of one or more endoluminal vascular prostheses prior to implantation is compared to a photograph taken after implantation. On the basis of such a comparative examination and applying certain evaluation criteria, the attending physician forms an opinion on hemodynamic changes. Among other things, the problem with this method is that the viewer's experience and perceptive gifts are very important components, so that each assessment is inevitably subjectively very subjective. Changes that are difficult or even imperceptible to the naked human eye, for example, can not or only insufficiently be included in a rating. Also, the applied evaluation criteria are often not clearly defined or accessible only through time-consuming training. This lack of objectivity makes a retrospective assessment of the successful therapy and the formulation of reliable clinical guidelines difficult.

The object of the present invention is therefore to specify a method which makes it possible to determine a characteristic value of a hemodynamic state in a vessel. It is a further object of the invention to describe a device with which a determination of a characteristic value of a hemodynamic state in a vessel can be carried out.

The invention solves this problem with a method for determining a characteristic value of a hemodynamic state in a vessel having the features of the first independent patent claim and a device for determining a characteristic value of a hemodynamic state in a vessel having the features of the second independent patent claim. Advantageous embodiments are described in subclaims.

A basic idea of the invention is a method for determining a characteristic value of a haemodynamic condition in a vessel, wherein in the determination of the characteristic value the time intensity curve of gray values of at least a first, fixed pixel coordinate of a first angiographic image series and transformed into the frequency domain is transformed into the frequency domain , temporal intensity course of gray values of at least one second fixed pixel coordinate, a second angiographic image series, and / or wherein in the determination of the characteristic value the temporal intensity curves of gray values of at least two fixed pixel coordinates of a first angiographic image series transformed into the frequency domain are received.

In the context of the invention, vessels can be understood to be blood vessels, such as arteries, veins or capillaries. Hemodynamics describes the blood flow in the blood vessels, which depends on various factors. Influence on the flow mechanics of the blood have, inter alia, parameters of the geometry of the vessel, the elasticity of the vessel walls, the pressures prevailing in the vessel, the cardiac output, in particular the flow velocity, the blood volume and the blood composition. The aim of the method according to the invention is to determine a characteristic value of a hemodynamic state in a vessel. That is, a characteristic value that gives information about the fluid mechanics in a vessel.

This characteristic value is determined with the aid of one or more angiographic image series. Obtaining an angiographic image series is common clinical practice. Frequently, the so-called digital subtraction angiography, DSA, is used, which makes blood flow on x-ray images visible. DSA is a process in which temporally successively a plurality of X-ray images of the object to be examined, for example a blood vessel, are recorded. An X-ray image is generally composed of a matrix of pixels. Pixels, together, to each of which a brightness value or gray value is assigned. The gray value of a pixel depends on the degree of absorption of X-ray radiation when passing through matter, eg of bone, tissue, blood, etc. After the first X-ray of the object to be examined, for example, a contrast agent is injected into the artery. The following pictures of the object to be examined differ only in the flowing contrast agent. A subtraction of the second, third, etc. image from the first image (image without contrast agent minus image with contrast agent) makes the blood vessels visible, eliminating the background. The subtracted individual images form an angiographic image series.

Based on the individual images of such an angiographic image series, the temporal intensity profiles of gray values of at least two fixed pixel coordinates are initially determined according to the invention. An intensity progression of gray values of a fixed pixel coordinate means the sequence of gray values of pixels over time, wherein the pixel coordinate, ie the location of the pixel whose gray value is read out, is the same in each individual image and the time information can be calculated over the acquisition time of the X-ray image is. The temporal intensity profiles of gray values of the at least two fixed, preferably different, pixel coordinates are then transformed into the frequency domain so that at least two frequency spectra are obtained. These at least two frequency spectra then enter into the determination of the characteristic value of a hemodynamic state in a vessel.

Alternatively or in addition to the temporal intensity curves of at least two fixed pixel coordinates of a first angiographic image series transformed into the frequency domain, the temporal intensity profile of gray values of at least one first, fixed pixel coordinate of a first angiographic image series transformed into the frequency domain can also be used to determine the characteristic value and the temporal intensity course of gray values, transformed into the frequency domain, of at least one second, fixed pixel coordinate, of a second angiographic image series. That is, this case is not based on grayscale gradients of at least two point coordinates of an angiographic image series, but on gray scale gradients of two angiographic image series, wherein at least a first gray level profile at a first pixel coordinates of the first angiographic image series and at least one gray value curve at a second pixel coordinates of the second angiographic image series determined become. The first and the second pixel coordinates may also be the same, so that the gray scale gradients differ essentially in that they were obtained from different angiographic image series, which were taken at different times, for example. Again, the temporal gray scale gradients are transformed into the frequency domain and the characteristic of a hemodynamic state obtained by the fact that the frequency spectra are related to each other. This can be related, for example, a mathematical link.

The method described so far assumes that the same pixel coordinates depict the same anatomical image details. If this condition is violated, for example as a result of a movement of a patient between exposures of angiographic image series, the images of an angiographic image series can advantageously be shifted in such a way that the image contents are again congruent. For such so-called registration methods, various proposals are known from the specialist literature, in particular from the field of medical image processing.

Preferably, a ratio of amplitudes of predeterminable frequencies of the frequency spectra, of the temporal intensity gradients of gray values of the first and / or the second angiographic image series transformed into the frequency range, is included in the determination of the characteristic value.

In this embodiment of the invention, amplitudes of the previously determined frequency spectra at frequencies that can be predetermined, for example, by a user are determined and related to one another. For example, the amplitudes A1 and A2 of the frequency spectrums of a gray value curve could each be determined at an image coordinate X1 and at an image coordinate X2 at the frequency 60 Hz. The characteristic value of a hemodynamic state in the vessel would then be, for example, K = A1 / A2 and could indicate, for example, turbulent flow conditions. Or, the amplitude A1 of the frequency spectrum of a first gray level curve of a first angiographic image series at an image coordinate X1 at the frequency 100 Hz and the amplitude A1 'of the frequency spectrum of a second gray level curve of a second angiographic image series are also determined at the image coordinates X1 and at the frequency 100 Hz , The characteristic value of a hemodynamic state in the vessel would then be, for example, K '= A1' / A1. The characteristic value K 'thus provides information about how, at the point X1, the amplitude at the frequency 100 Hz of the first angiographic series has changed to the second series of angiographic images. For example, the angiographic image series could have been made every one year, so that the characteristic K 'could be interpreted as an indication of an improvement or deterioration of the fluid mechanics of a vessel. Furthermore, combinations with a plurality of angiographic image series, a plurality of image coordinates and / or a plurality of considered frequencies are conceivable. The mathematical combination to be used and / or the selection of the image coordinates can be determined, for example, by test series.

A mathematical combination of the intensities of the gray values of pixels of a predefinable target region of the individual images of the angiographic image series or of the angiographic image series is particularly advantageous in the temporal intensity profiles of gray values of the first and / or the second angiographic image series.

This feature makes it possible to combine several pixels or their gray values per image of an angiographic image series into a gray value. This has the advantage that in practice difficult to avoid tolerances in imaging can be compensated. For example, the gray values could be averaged or a region around an image coordinate whose gray values are weighted less strongly by the image coordinates radially outwards, such as grayscale values of image points near the image coordinates, could be averaged.

In an advantageous development, the first angiographic image series visualizes a fluid flow in the vessel prior to insertion of a flow-altering agent and the second angiographic image series visualizes a fluid flow in the vessel after insertion of a flow-altering agent.

For a flow modifying agent or a flow altering implant, it may be e.g. to trade a commercially available flow diverter, for example in the form of a stent act. A flow diverter is generally understood to mean a medical implant containing a flow of a fluid, e.g. a bloodstream, targeted in a target vessel to influence, for example, the inflow into and outflow from an aneurysm should be prevented without covering the branches and thereby close it. The described methods according to the invention allow the determination of a characteristic value of a hemodynamic state in a vessel and can thus be used profitably to determine a change in a hemodynamic state or a change in the fluid mechanics in a vessel. For this purpose, in this embodiment of the invention, the first angiographic image series used is an angiographic image series before inserting a flow-altering agent, and as the second angiographic image series an angiographic image series after inserting a flow-altering agent. The particular characteristic may then be e.g. Provide information about how a flow characteristic has changed through the use of the flow-altering agent.

In a further advantageous embodiment, the at least one fixed pixel coordinate of at least one of the angiographic image series is arranged within an aneurysm.

An aneurysm is a pathological, localized, often baggy dilatation of a blood vessel. In particular, an aneurysm can occur in a blood vessel in the region of the brain or the heart; however, the appearance of an aneurysm is generally not limited to a particular body region. The clinical significance of an aneurysm, which is localized for example in the brain, is in particular due to the risk of rupture, so a crack or fracture, which may, for example, lead to bleeding and thrombosis. The dynamics of blood flow in an aneurysm is often considered in today's medicine as an important factor for the pathogenesis of the aneurysm, ie for its formation and development. Thus, information about the flow characteristics within an aneurysm is of great importance.

In an alternative embodiment of the invention, the frequency spectrum of a predefinable frequency range of the temporal intensity curve of gray values of at least one first fixed pixel coordinate of at least one of the angiographic image series is amplified and / or filtered and / or averaged.

It may be desirable to amplify certain frequency ranges to emphasize or first visualize image information. Of particular interest here may be, for example, portions of the frequency spectrum corresponding to the frequency of a periodic signal causing an oscillatory variation in one or more pixels in an angiographic image series. An example of such a signal is the heartbeat, which can cause dilution during systole and at a rate of contrast injection rate that is constant in time and compression of the used X-ray attenuating contrast agent during diastole. The proportion of the frequency of this signal in an averaged signal from one or more pixels inside the Aneurysms before a therapy can be related to the value after the therapy. This quotient can be understood as a measure describing the decoupling of the aneurysm from the main circulation with respect to the considered frequency or band of frequencies. Such a measure could serve as an indication for predicting the success of treatment after endovascular aneurysm treatment.

A further advantageous embodiment provides that frequency spectra of several predefinable frequency ranges of the temporal intensity curve of gray values of at least one fixed pixel coordinate of at least one of the angiographic image series are included in the determination of the characteristic value and / or frequency spectra of one or more predefinable frequency ranges of the in the frequency range transformed, temporal intensity course of gray values of at least one, fixed pixel coordinate and gray values of a predetermined range by at least one, fixed pixel coordinate at least one of the angiographic image series in the determination of the characteristic.

The methodology is independent of the number and numerical spacing of the frequencies considered in the frequency spectrum. To reduce the variance of the estimated magnitude of a frequency component in a signal, it may be advantageous to average a plurality of overlapping frequency divisions of the same signal or signals of spatially adjacent pixels in frequency space.

In addition, further improvements can be achieved if information about an injection course of a contrast agent, which injection course was present during the extraction of at least one of the angiographic image series, is included in the determination of the characteristic value.

It would be conceivable, besides a natural signal, such as the heartbeat, to consider the consideration of an artificial variation induced by the injection protocol of the contrast agent. A contrast agent can be injected in a controlled manner or the influence can be measured. In both cases, the information of the injected contrast agent is available over time. A periodic variation can be detected in an angiographic series as a periodic increase or decrease of the blackening in one or more pixels. The proportion of the frequency of this signal in an averaged signal from one or more pixels, for example, in the interior of an aneurysm before a therapeutic measure can now be set again with the corresponding value after the therapy measure. This quotient can be interpreted as a measure describing the decoupling of the aneurysm from the main circulation with respect to a particular frequency or band of frequencies. Such a measure could again serve as an indication for predicting the success of treatment after an endovascular aneurysm treatment.

The temporal intensity course of gray values of at least one first, fixed pixel coordinate of at least one of the angiographic image series is advantageously normalized, the normalization value being determined by a gray value of a pixel located on the vessel or by a gray value obtained by filtering pixels of a predefinable image area, which is on the vessel is won.

A particularly advantageous embodiment would determine a characteristic value of a hemodynamic state in a vessel from one or more pixels which are located on the vessel from which, for example, an aneurysm branches off in order to normalize with its reciprocal the quotient of the image information, for example, inside an aneurysm to account for changes in blood flow in the afferent vessel before, during, and / or after therapy.

It has proven to be advantageous if, in the determination of the characteristic value, the temporal intensity gradients of gray values of fixed pixel coordinates transformed into the frequency domain respectively proximal and distal of an aneurysm of a first angiographic image series, and / or the temporal intensity gradients of gray values transformed into the frequency domain of solid pixel coordinates respectively proximal and distal to an aneurysm of a second angiographic image series.

In this embodiment, a characteristic value can be formed in different angiographic image series as a quotient of gray values of one or more pixels from two regions which are respectively proximal and distal of the aneurysm, one characteristic value in front of each and one characteristic value after insertion of a flow-altering agent which can then be compared with each other. Thus, the influence that the aneurysm has on certain periodic signals at different locations in the delivery vessel before, during, and / or after treatment can be quantified. For this embodiment, it may be advantageous to perform a frequency separation, ie a transformation of the time signal into the frequency range, for the signals of the proximal and the distal pixels, and to subsequently correlate the results of this frequency analysis. This makes it possible, for example, the influence of Aneurysm to determine the pulsatility of the flow of blood with and without flow diverter.

It is conceivable that, in addition to the determination of the characteristic value, information about flux-mechanical states can be received by analysis in spatial space.

A simulation of blood flow by CFD method, CFD for engl. Computational Fluid Dynamics, provides a three-dimensional distribution of flow parameters, such as WSS. Wall shear stress, wall shear stress, along a surface of a vessel lumen. In the DE 10 2010 039 312 A1 For example, a method for simulating a blood flow in a vascular segment of a patient is obtained, an image acquisition of an examination region comprising the vessel segment is determined, a 3D vascular model is determined from the image acquisition, a number of blood flow parameters are read in and, taking into account these blood flow parameters, the blood flow in The 3D vascular model is simulated. Advantageously, methods that obtain information about flux-mechanical states by analysis in spatial space are supplemented by an induced amplification of variations with a specific repetition frequency. An example of a method of analyzing fluid mechanical states in spatial space is the angiographic image series estimation of a vector field to describe the projected contrast media motion in a vessel known as optical flow from the field of computer vision. Here, for example, by means of frequency filtering or frequency amplification, in particular movements with a repeating pattern of interest can be considered.

It is proposed that a phase shift between two temporal intensity progressions of gray values of two adjacent pixel coordinates of an angiographic image series transformed into the frequency range is reduced by a windowing of the temporal intensity profiles.

If a phase shift with respect to a certain frequency can be detected between adjacent pixels, this can be determined and reduced by rectangular windowing of the signals in order to achieve a better averaging in the time domain. Independently of this windowing, fenestration can be provided which, for example, favors the frequency or amplitude resolution of the frequency spectrum.

In an advantageous development, at least one method step is carried out automatically.

Automatically executed procedural steps require less intervention by a user and can thus speed up the procedure. For example, the propagation of an image area through individual images of an angiographic image series, possibly with additional registration steps, can be carried out with the aid of automatically executed image processing algorithms of medical image processing. In addition, automatically executed procedures are often less error prone.

In a further advantageous embodiment of the specific characteristic value is compared with given characteristics and it is issued a Maßungsmaß with the given characteristics.

The given characteristic values, which are stored for example in a database and can be obtained by extensive test series, can be linked to information in the manner of an expert system known per se, such as e.g. the information that a parameter of size x indicates a laminar flow. If, with one of the above-described inventive methods and under the same boundary conditions as when obtaining the characteristic values from the database, a characteristic value of a hemodynamic state in a vessel is determined which has a large degree of agreement with a stored characteristic value, the probability is high in both cases at least similar flux-mechanical states are present.

Another aspect of the invention is a device for determining a characteristic value of a hemodynamic state in a vessel, the device comprising means for carrying out one of the methods described above.

The device according to the invention advantageously comprises at least one computing and control means, eg a computer, and a display means, eg a computer monitor. The computing and control means may be adapted to receive at least one angiographic image series of an examination region of an examination subject, for example by a data interface with the aid of which it can receive an angiographic image series. The computing and control means advantageously has an input means for input of image coordinates. Furthermore, the calculation and control means are designed to determine a characteristic value of a hemodynamic state in a vessel, for example by having a suitable computer program comprising the method steps of one of the previously described methods according to the invention for determining a characteristic of a hemodynamic state in a vessel executes and executes this computer program. The specific characteristic value can be visualized on the display means, eg by displaying it on the computer monitor.

The embodiments described in more detail below represent preferred embodiments of the present invention.

Further advantageous developments will become apparent from the following figures, including description. Show it:

1 a flow diagram of an embodiment of a method according to the invention for determining a characteristic of a hemodynamic condition in a vessel;

2 symbolically a first angiographic image series with an aneurysm;

3 a temporal intensity profile of averaged gray values of pixels of a first predefinable target area in the first angiographic image series;

4 an amplitude characteristic of the temporal intensity profile of averaged gray values of pixels of the first predefinable target area in the first angiographic image series which is transformed into the frequency domain;

5 a temporal intensity profile of gray values of pixels of a second, fixed pixel coordinate in the first angiographic image series;

6 an amplitude profile of the temporal intensity curve of gray values of pixels of the second, fixed pixel coordinate in the first angiographic image series which is transformed into the frequency domain;

7 a temporal intensity profile of averaged gray values of pixels of a predefinable target area in the aneurysm in the first angiographic image series;

8th an amplitude characteristic of the temporal intensity curve of averaged gray values of pixels of the predefinable target area in the aneurysm in the first angiographic image series which is transformed into the frequency domain;

9 symbolically a second angiographic image series with an aneurysm and a stent;

10 a temporal intensity profile of averaged gray values of pixels of a first predefinable target area in the second angiographic image series;

11 an amplitude curve of the temporal intensity curve of averaged gray values of pixels of the first predefinable target area in the second angiographic image series which is transformed into the frequency domain;

12 a temporal intensity profile of gray values of pixels of a second, fixed pixel coordinate in the second angiographic image series;

13 an amplitude profile of the temporal intensity curve of gray values of pixels of the second, fixed pixel coordinate in the second angiographic image series which is transformed into the frequency range;

14 a temporal intensity profile of averaged gray values of pixels of a predefinable target area in the aneurysm in the second angiographic image series;

15 an amplitude profile of the temporal intensity curve of averaged gray values of pixels of the predefinable target area in the aneurysm in the second angiographic image series which is transformed into the frequency domain;

16 symbolically an embodiment of a device for determining a characteristic value of a hemodynamic state in a vessel;

17 by way of example a flow chart of a further embodiment of a method according to the invention for determining a characteristic value of a hemodynamic state of a vessel.

• S1) Entgegennahme wenigstens einer angiographischen Bildserie;
• S2) Entgegennahme wenigstens eines Zielbereiches, der in der wenigstens einen angiographischen Bildserie dargestellt ist;
• S3) automatisches oder manuelles Propagieren des wenigstens einen Zielbereiches durch die Bilder der wenigstens einen angiographischen Bildserie und vorgebbar Filterung, insbesondere Mittelung, der Grauwerte der Bildpunktkoordinaten des wenigstens einen Zielbereiches;
• S4) vorgebbar Fensterung der zeitlichen Intensitätsverläufe von gefilterten Grauwerten der Bildpunktkoordinaten des wenigstens einen Zielbereiches;
• S5) Transformieren der gegebenenfalls gefensterten, zeitlichen Intensitätsverläufe von gefilterten Grauwerten der Bildpunktkoordinaten des wenigstens einen Zielbereiches in den Frequenzbereich;
• S6) Entgegennahme wenigstens einer Frequenz oder wenigstens eines Frequenzbereiches;
• S7) vorgebbar Bestimmung von Amplituden des Frequenzspektrums an den entgegengenommenen Frequenzen oder Mittelung von Amplituden des Frequenzspektrums über vorgebbare Mittelungsfrequenzen der entgegengenommenen Frequenzen oder über einen vorgebbaren Frequenzbereich des wenigstens einen entgegengenommenen Frequenzbereiches;
• S8) Bestimmung eines Kennwertes aus den Amplituden oder aus den gemittelten Amplituden aus Verfahrensschritt S7, insbesondere durch Bilden von Quotienten mindestens zweier Amplituden;
• S9) Vergleich des Kennwertes mit gespeicherten Kennwerten und Ausgabe eines Übereinstimmungsmaßes.

Vorteilhaft können ein oder mehrere Verfahrensschritte automatisch ausgeführt werden. 1 shows an example of a flowchart of an embodiment of a method according to the invention 1 for determining a characteristic value of a hemodynamic condition in a vessel. The procedure 1 includes the method steps S1 to S9. It begins with method step S1 and ends, "End", after method step S9. The individual process steps are:

• S1) receiving at least one angiographic image series;
• S2) receiving at least one target area represented in the at least one angiographic image series;
• S3) automatic or manual propagation of the at least one target area through the images of the at least one angiographic image series and predefinable filtering, in particular averaging, of the gray values of the pixel coordinates of the at least one target area;
• S4) predeterminable windowing of the temporal intensity profiles of filtered gray values of the pixel coordinates of the at least one target region;
• S5) transforming the optionally windowed, temporal intensity profiles of filtered gray values of the pixel coordinates of the at least one target region into the frequency domain;
• S6) receiving at least one frequency or at least one frequency range;
• S7) specifiable determination of amplitudes of the frequency spectrum at the received frequencies or averaging of amplitudes of the frequency spectrum over predefinable averaging frequencies of the received frequencies or over a predefinable frequency range of the at least one received frequency range;
• S8) determining a characteristic value from the amplitudes or from the averaged amplitudes from method step S7, in particular by forming quotients of at least two amplitudes;
• S9) Comparison of the characteristic value with stored characteristic values and output of a conformity measure.

Advantageously, one or more method steps can be performed automatically.

In 2 is symbolically a first angiographic series of pictures 10 of a blood vessel 14 with an aneurysm 15 shown. The angiographic image series 10 consists in this embodiment simplified of three frames 11 . 12 and 13 a digital subtraction angiography associated with the times of their admission. In the first picture 11 the angiographic image series 10 is a first target area 16 , for example, by a user, given. The first target area 16 has a disk-shaped shape with a known radius and a center through a first, fixed pixel coordinate 8th , symbolically represented by a location vector arrow, is defined. By averaging the gray values of the pixels, the first target area 16 To get a gray value at the first time of the first, fixed pixel coordinate 8th the first angiographic image series 10 , By propagating the first target area 16 through all pictures, ie here through the further single pictures 11 and 12 , and in each case averaging of the gray values of the target area, become further gray values at other times of the first, fixed pixel coordinate 8th the first angiographic image series 10 certainly. In general, the first image series will comprise a plurality of individual images, so that the sequence of gray values results in a temporal intensity profile of gray values. An intensity profile is thus understood to mean the temporal course of gray values.

3 shows an example of a temporal intensity course 20 , I (t), of averaged gray values of pixels of the first predefinable target area 16 in the first angiographic series 10 over time t. A sinusoidal course of the intensity profile can be seen 20 which is essentially caused by the pumping action of the heart.

In 4 is the amplitude curve 21 of the time domain intensity curve transformed into the frequency domain 20 of averaged gray values of pixels of the first predefinable target area 16 in the first angiographic series 10 shown. That is, by mathematical methods known per se, such as the fast Fourier transform, FFT, is the temporal intensity curve 20 the averaged gray values of pixels of the first predefinable target area 16 in the first angiographic series 10 out 3 transformed into the frequency domain and represents the amplitude curve P (f) over the frequency f. One recognizes a large amplitude value at a frequency that corresponds to the heart rate.

Exemplary is in 2 a pixel 17 with gray value at a second, fixed pixel coordinate 9 indicated by a location vector arrow, the first angiographic image series 10 represented. Since this is a single pixel, which has only one gray value per frame, there results a temporal intensity profile of gray values of pixels of the second, fixed pixel coordinate 9 in the first angiographic series 10 through the sequence of gray values over time. This is in 5 as intensity course 22 , I (t), represented over time t. Evident is a sinusoidal course, the shares of other frequencies are superimposed. The transformation in the frequency domain and representation of the amplitude curve over the frequency is in 6 to see. One recognizes the amplitude course 23 , P (f), of the time domain intensity curve transformed into the frequency domain 22 gray values of pixels 17 the second, fixed pixel coordinate 9 in the first angiographic series 10 over the frequency f, a large contribution to the fundamental frequency, which corresponds to the heart rate, and further, smaller contributions at frequencies smaller and larger than the heart rate.

Next is in 2 a target area 18 in the aneurysm 15 the first angiographic image series 10 highlighted. Analogous to the procedure as for the first target area 16 can be a temporal intensity curve of averaged gray values of pixels of the given target area 18 in the aneurysm 15 in the first angiographic series 10 represented in 7 as a temporal intensity course 24 , I (t) over the time t, and an amplitude characteristic of the transformed in the frequency domain, time intensity curve of averaged Gray values of pixels of the predefinable target area 18 in the aneurysm 15 in the first angiographic series 10 represented in 8th as amplitude curve 25 , P (f), over the frequency f.

According to the invention, in the determination of a characteristic value of a hemodynamic state in a vessel, the temporal intensity profiles of gray values of at least two fixed pixel coordinates of a first angiographic image series transformed into the frequency domain can be included. Thus, from the frequency spectra, in the 4 . 6 and 8th , are the amplitude spectra 21 . 23 . 25 , ie the amounts of the frequency spectra, recorded, determine a characteristic value that describes a flow mechanical state in a vessel. For example, a ratio of amplitudes of predeterminable frequencies of the frequency spectra, the temporal intensity gradients of gray values of the first angiographic image series transformed into the frequency range, enters into the determination of the characteristic value. For example, it is conceivable to use the characteristic value as a quotient of the amplitude values of the gray values of the pixel transformed into the frequency domain 17 the second, fixed pixel coordinate 9 and of the temporal intensity profile of averaged gray values of pixels of the predetermined target range which is transformed into the frequency domain 18 in the aneurysm 15 in the first angiographic series 10 to be calculated at twice the heart rate. Alternatively, a characteristic value could be formed by forming the quotient of the amplitudes 21 . 23 and 25 be determined at the heart rate.

By comparing the determined characteristic value with predetermined characteristic values, e.g. stored in a database, and determination of that characteristic closest to the determined characteristic may be inferred to a hemodynamic condition in the vessel.

In 9 is symbolically a second angiographic series of pictures 10 ' of the blood vessel 14 with the aneurysm 15 shown. The angiographic image series 10 ' consists in this embodiment simplified of three frames 11 ' . 12 ' and 13 ' a digital subtraction angiography associated with the times of their admission. The presentation of the 9 is similar to the representation of 2 with the significant difference that the blood vessel 14 in addition a stent 19 ' having. The stent 19 ' should, for example, the blood flow in the aneurysm 15 limit, so a rupture of the aneurysm 15 to prevent. In the first picture 11 ' the angiographic image series 10 ' is again a first target area 16 , for example, by a user, given. The first target area 16 again has a disk-shaped shape of known radius and a center point through the first, fixed pixel coordinate 8th is defined. By averaging the gray values of the pixels that are the first target area 16 To get a gray value at the first time of the first, fixed pixel coordinate 8th the second angiographic image series 10 ' , By propagating the first target area 16 through all pictures, ie through the other single pictures 11 ' and 12 ' , and in each case averaging of the gray values of the target area, become further gray values at other times of the first, fixed pixel coordinate 8th the second angiographic image series 10 ' certainly. In general, the second image series will also comprise a plurality of individual images, so that the sequence of gray values results in a temporal intensity profile of gray values.

Analogous to the procedure as described in the description of the 2 to 8th is to be found, the temporal intensity gradients of the pixel 17 with gray values of the second, fixed pixel coordinate 9 the second angiographic image series 10 ' and the temporal intensity profile of averaged gray values of pixels of the predetermined target area 18 in the aneurysm 15 in the second angiographic series 10 ' transformed into the frequency domain.

So shows 10 the temporal intensity course 20 ' of averaged gray levels of pixels of the first predetermined target area 16 in the second angiographic series 10 ' . 11 shows the amplitude curve 21 ' of the time domain intensity curve transformed into the frequency domain 20 ' of averaged gray levels of pixels of the first predetermined target area 16 in the second angiographic series 10 ' . 12 shows the temporal intensity course 22 ' gray values of pixels 17 the second, fixed pixel coordinate 9 in the second angiographic series 10 ' . 13 shows the amplitude curve 23 ' of the time domain intensity curve transformed into the frequency domain 22 ' gray values of pixels 17 the second, fixed pixel coordinate 9 in the second angiographic series 10 ' . 14 shows a temporal intensity course 24 ' of averaged gray levels of pixels of the given target area 18 in the aneurysm 15 in the second angiographic series 10 ' and finally shows 15 an amplitude course 25 ' of the time domain intensity curve transformed into the frequency domain 24 ' of averaged gray levels of pixels of the given target area 18 in the aneurysm 15 in the second angiographic series 10 ' ,

In the determination of a characteristic value of a hemodynamic state in a vessel, the temporal intensity profile of gray values of at least one first, fixed pixel coordinate transformed into the frequency domain can be determined first angiographic image series and the temporal intensity course of gray values of at least one second, fixed pixel coordinate, the second angiographic image series, transformed into the frequency domain. Thus it can be determined from the frequency spectra of the two angiographic image series 10 and 10 ' determine a characteristic value that describes a flow mechanical state in a vessel. For example, the determination of the characteristic value involves a ratio of amplitudes of predeterminable frequencies of the frequency spectra, the temporal intensity profiles of gray values of the first angiographic image series transformed into the frequency range 10 and the second angiographic image series 10 ' one. It is conceivable, for example, the characteristic value as a quotient of the amplitude values 25 and 25 ' the transformed into the frequency domain, temporal intensity gradients 24 and 24 ' of averaged gray levels of pixels of the given target area 18 in the aneurysm 15 in the first angiographic series 10 and in the second angiographic series 10 ' to calculate at the heart rate.

In another embodiment, the amplitude values could 25 and 25 ' the transformed into the frequency domain, temporal intensity gradients 24 and 24 ' of averaged gray levels of pixels of the given target area 18 in the aneurysm 15 in the first angiographic series 10 and in the second angiographic series 10 ' at the heart rate, abbreviated to A3 and A3 ', and the amplitude values of the temporal intensity gradients transformed into the frequency domain 20 and 20 ' of averaged gray levels of pixels of the first predetermined target area 16 in the first angiographic series 10 and in the second angiographic series 10 ' at the heart rate, abbreviated to A4 and A4 '. The characteristic value of a hemodynamic state in the vessel would then increase, for example K '' = (A3 '/ A3) · (A4 / A4') to calculate.

16 symbolically shows an embodiment of a device 80 for determining a characteristic value of a hemodynamic condition in a vessel. The device 80 in this embodiment comprises a calculation and control means 50 , here a computer, and an indicator 54 , here a computer monitor. The calculation and control means 50 is to receive angiographic image series 10 and 10 ' an examination area, for example a blood vessel, an examination subject, for example a human or animal patient. In the embodiment of 16 is the computing and control tool 50 equipped with a data interface by means of which the angiographic image series 10 and 10 ' stored in data stores 30 and 30 ' are stored, can receive. The calculation and control means 50 has an input device 52 , here a computer keyboard, for example, to control process steps or to enter image coordinates or a target area 16 , wherein advantageously at least one image of one of the angiographic image series 10 and 10 ' on the display means 54 is pictured. Next is the calculation and control means 50 adapted to determine a characteristic value of a hemodynamic state in a vessel, for example by having a suitable computer program which carries out the method steps of one of the previously described methods according to the invention for determining a characteristic of a hemodynamic state in a vessel and executes this computer program. The specific characteristic value can be displayed on the display means 54 be visualized, eg by displaying a corresponding text 60 on the computer monitor. Alternatively or additionally, the determined characteristic value may be on an output medium 62 , here a pointer instrument, to be displayed. In one embodiment of a method according to the invention, the determined characteristic value is compared with given characteristic values and a degree of conformity with the given characteristic values is output. The given characteristics, which can be obtained by extensive test series, are in the embodiment of the 16 in a database 40 saved. For example, you can use an expert system stored in the memory of the calculator and controller 50 is stored, associated with information, such as the information that a characteristic of a given size indicates a laminar flow. If, with one of the above-described inventive methods and under the same boundary conditions as when obtaining the characteristic values from the database, a characteristic value of a hemodynamic state in a vessel is determined which has a large degree of agreement with a stored characteristic value, the probability is high in both cases at least similar flux-mechanical states are present.

• S1’) Entgegennahme von einem oder mehreren Zielbereichen, engl. ROIs (Regions Of Interest), in einer angiographischen Bildserie, die insbesondere durch einen Benutzer manuell selektiert sind;
• S2’) Propagieren der ROIs durch alle Bilder der angiographischen Bildserie und vorgebbar Übertragen der ROIs auf eine oder mehrere weitere angiographische Bildserien, wobei das Propagieren und/oder Übertragen vorgebbar automatisch oder manuell, z.B. durch Benutzereingabe erfolgt;
• S3’) Entgegennahme einer Auswahl einer oder mehrerer Frequenzen oder Frequenzbänder, die insbesondere durch einen Benutzer auswählbar sind;
• S4’) Für die ROIs jeweils – Mittelung der Bildpunkte aus dem jeweiligen ROI und Berechnung eines Frequenzspektrums aus dem zeitlichen Verlauf der gemittelten Bildpunkte, d.h. aus deren Grauwerten über der Zeit; – vorgebbar Reduzierung oder Eliminierung von Phasenverschiebungen; – vorgebbar Fensterung der Signale; – vorgebbar Mittelung der Signale erst nach der Berechnung des Frequenzspektrums, d.h. Mittelung im Frequenzbereich; – vorgebbar Teilung des Frequenzspektrums in mehrere, vorgebbar auch überlappende, Teilfrequenzspektren und Mittelung über die Teilfrequenzspektren; – vorgebbar Interpolation im Frequenzraum – vorgebbar Filterung des zeitlichen Verlaufs, um insbesondere zeitliche Variationen mit vorgebbaren Frequenzen, insbesondere Frequenzen fernab der in Verfahrensschritt S3 entgegengenommenen Frequenzen oder Frequenzbändern liegen, zu reduzieren;
• S5’) Mittelung von Amplituden der aus Verfahrensschritt S4 bestimmten oder gemittelten Werten der Frequenzspektren aus den ROIs;
• S6’) Vorgebbar Bildung von Quotienten aus den gemittelten Amplituden der verschiedenen ROIs aus Verfahrensschritt S5’;
• S7’) Ausgabe des oder der Quotienten aus Verfahrensschritt S6’.

17 shows by way of example a flow chart of a further embodiment of a method according to the invention 1' for determining a characteristic value of a hemodynamic condition in a vessel. The procedure 1' includes the method steps S1 'to S7'. It begins with method step S1 'and ends, "End", after method step S7'. The individual process steps are:

• S1 ') receiving one or more target areas, engl. Regions Of Interest (ROIs), in an angiographic image series, which are manually selected by a user in particular;
• S2 ') Propagation of the ROIs by all images of the angiographic image series and specifiable Transferring the ROIs to one or more further angiographic image series, wherein the propagation and / or transfer can be predetermined automatically or manually, for example by user input;
• S3 ') receiving a selection of one or more frequencies or frequency bands, which are in particular selectable by a user;
• S4 ') For the ROIs respectively - Averaging of the pixels from the respective ROI and calculation of a frequency spectrum from the time course of the averaged pixels, ie from their gray values over time; - specifiable reduction or elimination of phase shifts; - predeterminable windowing of the signals; - specifiable averaging of the signals only after the calculation of the frequency spectrum, ie averaging in the frequency domain; - predeterminable division of the frequency spectrum into a plurality of predeterminable overlapping subfrequency spectra and averaging over the subfrequency spectra; - Specifiable interpolation in the frequency space - specifiable filtering of the time course, in particular temporal variations with predeterminable frequencies, in particular frequencies far away from the received in step S3 frequencies or frequency bands are to reduce;
• S5 ') averaging amplitudes of the values of the frequency spectra determined or averaged from method step S4 from the ROIs;
• S6 ') Predefinable formation of quotients from the averaged amplitudes of the various ROIs from method step S5';
• S7 ') Output of the quotient (s) from method step S6'.

The output quotient or the output quotients, which are equal to the characteristic values or characteristic values of a hemodynamic state in a vessel, can be used, for example, to predict the therapeutic success after endovascular aneurysm treatment by comparison with expert-assessed comparison quotients.

In summary, further embodiments and advantages of the invention will be described. The invention proposes a method in which signals which represent the temporal course of gray values in one or more pixels of two-dimensional angiographic image series are decomposed and analyzed in their frequency components. The individual frequency components can be dimensioned and the decompositions of averaged signals of different pixels or identical pixels in different image series can be related to each other.

Advantageously, certain frequency ranges can be enhanced in order to emphasize image information or to make it visible first. Of particular interest here may be, for example, fractions of the decomposition corresponding to the frequency of a periodic signal, e.g. the heartbeat causing an oscillatory variation in one or more pixels in an angiographic image series. The proportion of the frequency of this signal in an averaged signal from one or more pixels in the interior of an aneurysm before a therapeutic measure can now again be related to the corresponding value after the therapeutic measure. This quotient can be interpreted as a measure describing the decoupling of the aneurysm from the main circulation with respect to the considered frequency, or the considered frequency band. Such a measure could serve to predict therapeutic success after endovascular aneurysm treatment.

Such enhancement of variations with a particular repetition frequency can also be used to advantage in supplementing methods that obtain information about flux-mechanical states through analysis in spatial space. Exemplary of such a method is the estimation of a vector field based on angiographic image series to describe the projected contrast agent movement in a hollow vessel, which is known from the field of computer vision as an optical flow. Here, for example, by means of frequency filtering or frequency amplification, in particular movements with a repeating pattern of interest can be considered.

In addition to a natural signal, it would be conceivable to consider an artificial variation induced by the injection protocol of the contrast agent. Such a periodic variation can be detected in an angiographic series as a periodic increase or decrease in the density in one or more pixels. The proportion of the frequency of this signal in an averaged signal from one or more pixels in the interior of the aneurysm before a therapeutic measure can now be related again to the corresponding value after the therapeutic action. This quotient may be understood as a measure describing the decoupling of the aneurysm from the main circulation with respect to a particular frequency or frequency band. Such a measure can again serve as an indication of the fulfillment of a therapeutic success after an endovascular aneurysm treatment.

A particularly advantageous embodiment would additionally determine such a quotient from one or more pixels which are located on the vessel from which the aneurysm branches off with its reciprocal to normalize the quotient of the image information inside the aneurysm and thereby to take into account changes in blood flow in the supplying vessel before, after or during therapy.

Another embodiment may, in various angiographic series, form such a quotient of one or more pixels from two regions located proximal and distal of the aneurysm, respectively, and compare that size before and after treatment. Thus, the influence that the aneurysm has on certain periodic signals at different locations in the delivery vessel before, after, or during treatment can be quantified. For this embodiment, it may be advantageous to first correlate the signals of the proximal and distal pixels and to perform a frequency decomposition with the result.

The methodology is independent of the number and numerical spacing of the frequencies considered in the decomposition. To reduce the variance of the estimated magnitude of a frequency component in a signal, it may be advantageous to average a plurality of overlapping frequency divisions of the same signal or signals of spatially adjacent pixels in frequency space.

If a phase shift with respect to a certain frequency is detected between adjacent pixels, this can be determined and eliminated by rectangular windowing of the signals in order to achieve a better averaging in the time period. Independently of this fenestration, a windowing can be integrated that, for example, favors the frequency or amplitude resolution in the frequency resolution.

By considering the complement of angiographic time-intensity curves in the frequency domain and using this approach to derive quantifiable measures, one or more characteristics can be determined which can also be used to predict therapeutic success after endovascular aneurysm treatment. Compared to the method described in the introduction, which is today predominantly used in clinical practice, the objectivity is increased and changes can be detected which would be difficult to detect with the naked eye, even for very well-trained experts. The characteristics or characteristics considered can be disclosed and the calculated measure quantified numerically. In this way further interventional decision-making processes can be optimized and made comprehensible retrospectively. In addition, such a measure may be used in clinical guidelines, which may facilitate the conduct of the treatment and secure the responsible physician.

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 102010039312 A1 [0031]

Claims (14)

Procedure ( 1 ) for determining a characteristic value of a hemodynamic condition in a vessel ( 14 ), wherein in the determination of the characteristic value of the time domain intensity curve ( 20 . 22 . 24 ) of gray values of at least one first, fixed pixel coordinate ( 8th ) of a first angiographic image series ( 10 ) and the temporal intensity curve ( 20 ' . 22 ' . 24 ' ) of gray values of at least one second, fixed pixel coordinate ( 9 ), a second angiographic image series ( 10 ' ), and / or wherein in the determination of the characteristic value the temporal intensity profiles ( 20 . 20 ' . 22 . 22 ' . 24 . 24 ' ) of gray values of at least two fixed pixel coordinates ( 8th . 9 ) of a first angiographic image series ( 10 ). Procedure ( 1 ) according to claim 1, wherein in the determination of the characteristic value, a ratio of amplitudes of predeterminable frequencies of the frequency spectra, of the temporal intensity profiles transformed into the frequency range (US Pat. 20 . 20 ' . 22 . 22 ' . 24 . 24 ' ) of gray values of the first angiographic image series ( 10 ) and / or the second angiographic image series ( 10 ' ) received. Procedure ( 1 ) according to claim 1 or claim 2, wherein in the temporal intensity curves ( 20 . 20 ' . 22 . 22 ' . 24 . 24 ' ) of gray values of the first angiographic image series ( 10 ) and / or the second angiographic image series ( 10 ' ) a mathematical combination of the intensities ( 20 . 20 ' . 22 . 22 ' . 24 . 24 ' ) of the gray values of pixels of a predefinable target area ( 16 ) of the first angiographic image series ( 10 ) and / or the second angiographic image series ( 10 ' ) received. Procedure ( 1 ) according to one of the preceding claims, wherein the first angiographic image series ( 10 ) a fluid flow in the vessel ( 14 ) before inserting a flow modifying agent ( 19 ' ) and the second angiographic image series ( 10 ' ) a fluid flow in the vessel ( 14 ) after insertion of a flow-modifying agent ( 19 ' ) visualized. Procedure ( 1 ) according to one of the preceding claims, wherein the at least one fixed pixel coordinate of at least one of the angiographic image series ( 10 . 10 ' ) within an aneurysm ( 15 ) is arranged. Procedure ( 1 ) according to one of the preceding claims, wherein the frequency spectrum of a predefinable frequency range of the time domain intensity curve transformed into the frequency domain ( 20 . 20 ' . 22 . 22 ' . 24 . 24 ' ) of gray values of at least one first, fixed pixel coordinate of at least one of the angiographic image series ( 10 . 10 ' ) is amplified and / or filtered and / or averaged. Procedure ( 1 ) according to one of the preceding claims, wherein frequency spectra of a plurality of predefinable frequency ranges of the frequency domain transformed in the time domain, the intensity profile ( 20 . 20 ' . 22 . 22 ' . 24 . 24 ' ) of gray values of at least one fixed pixel coordinate of at least one of the angiographic image series ( 10 . 10 ' ) enter into the determination of the characteristic value and / or wherein frequency spectra of one or more predefinable frequency ranges of the time domain intensity curve transformed into the frequency domain ( 20 . 20 ' . 22 . 22 ' . 24 . 24 ' ) of gray values of at least one fixed pixel coordinate and gray values of a predefinable area about the at least one fixed pixel coordinate of at least one of the angiographic image series ( 10 . 10 ' ) enter into the determination of the characteristic value. Procedure ( 1 ) according to one of the preceding claims, wherein information about an injection course of a contrast agent, which injection course during the extraction of at least one of the angiographic image series ( 10 . 10 ' ), enter into the determination of the characteristic value. Procedure ( 1 ) according to one of the preceding claims, wherein the temporal intensity profile ( 20 . 20 ' . 22 . 22 ' . 24 . 24 ' ) of gray values of at least one first, fixed pixel coordinate of at least one of the angiographic image series ( 10 . 10 ' ), the normalization value being determined by a gray scale value of a pixel located on the vessel ( 14 ) or by a gray value, which is obtained by filtering pixels of a predefinable image area, which is located on the vessel ( 14 ) is won. Procedure ( 1 ) according to one of the preceding claims, wherein in the determination of the characteristic value the temporal intensity profiles ( 20 . 20 ' . 22 . 22 ' . 24 . 24 ' ) of gray values of fixed pixel coordinates respectively proximal and distal of an aneurysm ( 15 ) of a first angiographic image series ( 10 ), and / or the temporal intensity curves ( 20 . 20 ' . 22 . 22 ' . 24 . 24 ' ) of gray values of fixed pixel coordinates respectively proximal and distal of an aneurysm ( 15 ) a second angiographic image series ( 10 ' ). Procedure ( 1 ) according to any one of the preceding claims, wherein in the determination of the characteristic value in addition receive information about flow mechanical states by analysis in the spatial space. Procedure ( 1 ) according to one of the preceding claims, wherein a phase shift between two temporal intensity gradients transformed into the frequency domain ( 20 . 20 ' . 22 . 22 ' . 24 . 24 ' ) of gray values of two adjacent pixel coordinates of an angiographic image series, by a windowing of the temporal intensity gradients ( 20 . 20 ' . 22 . 22 ' . 24 . 24 ' ) is reduced. Procedure ( 1 ) according to one of the preceding claims, wherein at least one method step is carried out automatically. Procedure ( 1 ) according to any one of the preceding claims, wherein the determined characteristic value with given characteristic values ( 30 ) and a measure of conformity with the given characteristic values ( 30 ) is output. Contraption ( 80 ) for determining a characteristic value of a hemodynamic condition in a vessel ( 14 ), the device comprising means ( 50 ) for carrying out one of the methods ( 1 ) according to any one of claims 1 to 14.

Images