ABNORMALITY FINDING IN PROJECTION IMAGES

Abstract

The present invention includes a utility for determining the severity of a stenosis in a blood vessel. In one aspect, a method for improving DSA image quality includes: (1) registration of the mask and bolus images prior to subtracting procedure to reduce the artifacts from misalignment; (2) enhancement of the registered DSA image by an anisotropic diffusion technique and a nonlinear normalization technique; and (3) detecting the boundary of blood vessel and quantitatively measuring percentage stenosis, which may be done automatically.

Description

FIELD

The present disclosure is directed to medical imaging systems. More specifically, the present disclosure is directed to systems and methods that alone or collectively facilitate real-time imaging.

BACKGROUND

Coronary artery disease causes in excess of 1.5 million cases of myocardial infarction annually, and is the leading cause of death in the United States, resulting in more than 500,000 deaths per year. The accurate diagnosis and quantification of coronary artery disease is critical to subsequent treatment decisions. Despite the emergence of new techniques for the visualization and analysis of vascular structures, digital subtraction angiography (DSA) remains the preferred procedure to help clinicians to make clinical-decisions in all kinds of vascular diseases. Digital subtraction angiography (DSA) is a well-established modality for the visualization of blood vessels in the human body.

During a DSA procedure a first series of frames are taken which represents the body anatomy in static form. They are called mask frames. Then dye is injected in the body through a catheter to visualize the blood vessel. The dye-injected frames are called the bolus frames. A subtraction procedure is preformed between the averaged bolus frame and the averaged mask frame to get DSA image. The patient is positioned on the X-ray imaging system while an X-ray movie is acquired and a DSA image is generated for an interventional cardiologist or a radiologist. Ideally, the DSA image mainly contains the dye-enhanced blood vessels and they appear dark. However, the artifacts due to patient motion and system noise frequently reduce the diagnostic value of the images.

In the produced DSA image, a pathological aspect is associated with a vascular area where there is a significant deviation from the diameter of the healthy vascular. In particular, a stenosis is associated with a significant narrowing of the vascular and is quantified by parameters such as the percentage of stenosis. Stenosis limits blood flow by raising the resistance to flow through the vessel. In X-ray angiography (DSA image), the contrast agent ensures that the outlines of the blood flow are revealed on the X-ray to indicate any narrowing of the blood vessel. Direct visual examination of cine film coronary angiograms and manual estimation of the degree of vascular stenosis were complicated and subject to a large inter- and intra-observer variability.

SUMMARY

The present invention includes a system and method (i.e., utility) for determining the severity of a stenosis in a blood vessel. In one aspect, a method for improving DSA image quality includes: (1) registration of the mask and bolus images prior to subtracting procedure to reduce the artifacts from misalignment; (2) enhancement of the registered DSA image; and (3) detecting the boundary of blood vessel and quantitatively measuring percentage stenosis, which may be done automatically. This last step is sometimes referred to as Quantitative coronary angiography (QCA). Aspects of the present invention allow such QCA to be performed by a computer with minimal user input.

The utility allows for the semi-automated quantitative measurement (QCA) of a lumen (e.g., blood vessel or artery). The utility involves a registration step of the mask and the bolus images prior to subtracting procedure and QCA measurement. Generally, such registration is a motion compensation step can reduce the artifacts from misalignment and improve the accuracy of QCA measurement. As will be appreciate, in coronary applications movement of the heart is nearly constant and motion compensation significantly improves overall image quality for QCA purposes.

In one arrangement, the motion compensation is performed using an algorithm that involves an inverse-consistency constrain which implies that the correspondence provided by the registration in one direction matches closely with the correspondence in the opposite direction. This may entail a B-spline parameterization. Other image registration methods can also be applied in the proposed system to match the mask and bolus images.

In one arrangement, the registration is performed hierarchically using a multi-resolution strategy in both, spatial domain and in domain of basis functions. The registration can be performed at ¼, ½ and full resolution using knot spacing of 8, 16 and 32. In addition to being faster, the multi-resolution strategy helps in improving the registration by matching global structures at lowest resolution and then matching local vessel structures as the resolution is refined. Due to the bi-directional approach, thin blood vessels edges are more prominent thereby avoiding the local minima both in iterative nature of thin vessel edge estimation and iterative nature of correction of bolus image. The subtraction process is applied on corrected bolus images.

Image enhancement prior to QCA measurement also improves the overall QCA measurement. In one arrangement the registered DSA image is further enhanced by background diffusion to remove noise and nonlinear normalization for better visualization. This image enhancement increases the contrast between the blood vessels and the background. This results in much improved contrast and very crisp subtraction images, in which the regions of interest are easily identifiable. Enhancement of the registered DSA image can be performed using an anisotropic diffusion technique and a nonlinear normalization technique. Other similar image enhancement techniques can also be applied in the proposed utility to enhance DSA images before QCA measurement.

The utility allows a user to input various types of lumen identification a center line method where a user identifies an approximate center of the lumen and an edge method where a user identifies initial edges of the lumen. Both methods can obtain percentage stenosis automatically with minimal user interaction. With the initial points selected by the user, the QCA measurement may include one or more of the following three sub-processes: (1) initial edge detection; (2) edge refinement; and (3) lesion measurement. Initial edge detection is performed where a user inputs centerline information. In any case, the utility may refine the edges of the lumen. In one arrangement, an active contour model algorithm may be used to refine the edges by deforming a contour to lock onto features of interest within in an image. Other edge detection or segmentation techniques can also be applied in the system for QCA measurement.

For each stenosis or lesion, the following data are calculated: (a) minimum and maximum diameter; (b) lesion length (the beginning and end of the lesion are determined automatically); (c) Stenosis reference (“normal”) diameter; and (d) percent stenosis. Other measurements can also be included in the proposed system.

To improve the overall speed of the utility, the entire software architecture may be implemented on a GPU based framework, which makes visualization and computation faster by up to factor of 30. Thus the entire utility may implemented substantially in real-time. Further, the utility may be used with various imaging modalities and may be used with 2-D or 3-D images.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 illustrates one embodiment of a X-ray DSA system.

FIG. 2 illustrates a process flow diagram for performing an improved QCA measurement in accordance with various aspects of the present invention.

FIG. 3 illustrates one non-limiting method for registering images.

FIG. 4 illustrates one non-limiting method for enhancing an image.

FIG. 5 illustrates a process flow diagram for performing a QCA measurement.

FIGS. 6A-6C illustrate an image having a vein or artery with a lesion.

FIG. 7 illustrates a process flow diagram of an edge detection, refinement and lesion measurement process.

FIGS. 8A-8C illustrate the refinement of edge boundaries of the vein or artery of FIGS. 6A-6C.

FIG. 9 illustrates a process flow sheet for an edge detection process for detecting edge surfaces of an object in an image.

FIG. 10 illustrates a process flow sheet for an edge refinement process.

FIG. 11 illustrates a process flow sheet for a lesion measurement process.

DETAILED DESCRIPTION

Reference will now be made to the accompanying drawings, which assist in illustrating the various pertinent features of the various novel aspects of the present disclosure. Although the present invention will now be described primarily in conjunction with angiography utilizing X-ray imaging, it should be expressly understood that aspects of the present invention may be applicable to other medical imaging applications. For instance, angiography may be performed using a number of different medical imaging modalities, including biplane X-ray/DSA, magnetic resonance (MR), computed tomography (CT), ultrasound, and various combinations of these techniques. In this regard, the following description is presented for purposes of illustration and description. Furthermore, the description is not intended to limit the invention to the form disclosed herein. Consequently, variations and modifications commensurate with the following teachings, and skill and knowledge of the relevant art, are within the scope of the present invention. The embodiments described herein are further intended to explain known modes of practicing the invention and to enable others skilled in the art to utilize the invention in such, or other embodiments and with various modifications required by the particular application(s) or use(s) of the present invention.

FIG. 1 shows one exemplary setup for a real-time imaging procedure for use during a contrast media/dye injection procedure. As shown, a patient is positioned on an X-ray imaging system 100 and an X-ray movie is acquired by a movie acquisition system 102. An enhanced DSA image, as will be more fully discussed herein, is generated by an enhancement system 104 for output to a display 106 and/or stenosis processing system that is accessible to an interventional radiologist.

The projection images (e.g., CT images) are acquired at different time instants and consist of a movie with a series of frames before, during and after the dye injection. The series of frames include mask images that are free of contrast-enhancing dye in their field of view 108 and bolus images that contain contrast-enhancing dye in their field of view 108. That is, bolus frames are images that are acquired after injected dye has reached the field of view 108. The movie acquisition system 102 is operative to detect the frames before and after dye injection automatically to make feasible a real-time acquisition system. One approach for identifying frames before and after dye injection is to find intensity differences between successive frames, such that a large intensity difference is detected between the first frame after dye has reached the field of view (FOV) and the frame acquired before it. However, the patient may undergo some motion during the image acquisition causing such an intensity difference between even successive mask images. To avoid this, the movie acquisition system 102 may align successive frames together, such that the motion artifacts are minimized. The first image acquired after the dye has reached the FOV will therefore cause a high intensity difference with the previous frame not containing the dye in FOV. The subtraction image or ‘DSA image’ obtained by subtracting a mask frame from a bolus frame (or vice versa) will contain a near-zero value everywhere if both images belong to background.

Generally, the subtraction image or DSA image is obtained by computing a difference between pixel intensities of the mask image and the bolus image. The enhancement system 104 may then enhance the contrast of the subtraction image. Such enhancement may include rescaling the intensities of the pixels in the subtraction image and/or the removal of noise from the subtraction image.

The acquisition and/or enhancement systems are computerized systems that may run application software and computer programs which can be used to control the system components, provide user interfaces, and/or provide features of the imaging system. The software may be originally provided on computer-readable media, such as compact disks (CDs), magnetic tape, or other mass storage medium. Alternatively, the software may be downloaded from electronic links such as a host or vendor website. The software is installed onto a hard drive and/or electronic memory of the system, and is accessed and controlled by the operating system. Software updates are also electronically available on mass storage media or downloadable from the host or vendor website. The software, as provided on the computer-readable media or downloaded from electronic links, represents a computer program product usable with a programmable computer processor having computer-readable program code embodied therein. The software contains one or more programming modules, subroutines, computer links, and compilations of executable code, which perform the functions of the imaging system. The user interacts with the software via keyboard, mouse, voice recognition, and other user-interface devices (e.g., user I/O devices) connected to the computer system.

FIG. 2 shows a method for improving DSA image quality and improving the accuracy of QCA measurements that includes three primary processes: The first process includes using an averaging system 202 to generate average mask frames 204 and average bolus frames 206 from an x-ray image or movie. These average images are motion compensated 208. Specifically, the averaged mask frame is registered with the averaged bolus frame to generate a registered mask frame 210, which minimizes the motion artifacts before subtraction. The registered mask frame 210 is then subtracted 212 from the average bolus frame to generate a DSA image 214. Various aspects of this process are set forth in co-pending U.S. application Ser. No. 11/609,743 entitled “Medical Image Enhancement System” the entire contents of which are incorporated herein by reference. The second process uses the registered DSA image 214 as input. This is an image enhancement process 216 that uses an anisotropic diffusion technique to reduce noise, followed by a nonlinear normalization method to enhance the image of the blood vessel. The result of this is the production of an enhanced DSA image 218. The third process uses the enhanced DSA image 218 to do a QCA measurement 220 and generate a QCA report or value 222. The motion compensation process is discussed herein relation to FIG. 3, the image enhancement process is discussed in relation to FIG. 4 and the QCA measurement is discussed in relation to FIGS. 5-11.

FIG. 3 illustrates the process of registering the averaged mask frame with averaged bolus frame, to minimize the motion artifacts. In this embodiment, this is done using an inverse consistent image registration using B-spline basis. In medical imaging, image registration is used to find a point-wise correspondence between a pair or group of similar anatomical objects. Image registration is often posed as an optimization problem that minimizes an objective function representing the difference between two images 302, 304 to be registered. In this case, the images are the average bolus and mask frames. The symmetric squared intensity difference is chosen as the driving function. In addition, regularization constraints are applied so that the deformation follows a model that matches closely with the deformation of real-world objects. The regularization is applied in the form of bending energy and inverse-consistency cost. Inverse-consistency implies that the correspondence provided by the registration in one direction matches closely with the correspondence in the opposite direction. Most image registration methods are uni-directional and therefore contain correspondence ambiguities originating from choice of direction of registration. In the presented method, the forward and reverse correspondences are computed simultaneously and bound together through an inverse consistency cost term. The inverse consistency cost term assigns higher cost to transformations deviating from being inverse-consistent. While inverse consistency minimizes the correspondence ambiguity, it also helps the transformation perform better by forcing it out of local minima. A cost function for performing image registration over the images is used:

$\begin{array}{cc}C=\sigma \left(\begin{array}{c}{\int }_{\Omega }{I_1\left(h_{1,2}\left(x\right)\right)-I_2\left(x\right)}^2x+\\ {\int }_{\Omega }{I_2\left(h_{2,1}\left(x\right)\right)-I_1\left(x\right)}^2x\end{array}\right)+\rho \left(\begin{array}{c}{\int }_{\Omega }{L\left(u_{1,2}\left(x\right)\right)}^2x+\\ {\int }_{\Omega }{L(u_{2,1}\left(x\right)}^2x\end{array}\right)+\chi \left(\begin{array}{c}{\int }_{\Omega }{h_{1,2}\left(x\right)-h_{2,1}^{-1}\left(x\right)}^2x+\\ {\int }_{\Omega }{h_{2,1}\left(x\right)-h_{1,2}^{-1}\left(x\right)}^2x\end{array}\right)& \left(1\right)\end{array}$

where, I₁(x) and I₂(x) represent the intensity of image at location x, x represents the domain of the image. h_i,j(x)=x+u_i,j(x) represents the transformation that maps image I_i to image I_j in the Eulerian frame of reference and u(x) represents the displacement field. L is a differential operator and the second term in Eq. (1) represents an energy function. σ, ρ and χ are weights to adjust relative importance of the cost function.

In Equation (1), the first term represents the symmetric squared intensity cost function and represents the integration of squared intensity difference between deformed reference image and the target image in both directions. The second term represents the energy regularization cost term and penalizes high derivatives of u(x). L is represented as a Laplacian operator mathematically given as: L=∇². The third and last term represents the inverse consistency cost function, which penalizes differences between transformation in one direction and inverse of transformation in opposite direction. The total cost 308 is computed 306 using Eq. 1 as a first step in registration.

The optimization problem posed in Eq. (2) is solved by using a B-spline parameterization as is known in the art. B-splines are used due to ease of computation, good approximation properties and their local support. It is also easier to incorporate landmarks in the cost term if we use spatial basis function. The above optimization problem is solved by solving for B-spline coefficients c_i's, such that

$\begin{array}{cc}h\left(x\right)=x+\sum _ic_i{\beta }_{i}\left(x\right)& \left(2\right)\end{array}$

where, β_i(x) represents the value of B-spline at location x, originating at index i. In this registration method, cubic B-splines are used. A gradient descent scheme is implemented based on the above parameterization. The total gradient cost is calculated and updated 310 with respect to the transformation parameters in every iteration. The transformation parameters are updated using the gradient descent update rule. Images are deformed into shape of one another using the updated correspondence and the cost function and gradient costs are calculated 314 until convergence 316 when the frames are registered 318. The registration is performed hierarchically using a multi-resolution strategy in both, spatial domain and in domain of basis functions. The registration is performed at ¼, ½ and full resolution using knot spacing of 8, 16 and 32. In addition to being faster, the multi-resolution strategy helps in improving the registration by matching global structures at lowest resolution and their matching local structures as the resolution is refined.

Referring again to FIG. 2, these registered frames are subtracted to provide the DSA image for image enhancement. That is, after image registration, subtraction is performed between the registered mask image and bolus image, resulting in the motion compensated (registered) DSA image 214. Using the registered DSA image as input, the system further improves image quality by the following two steps: 1) noise reduction through an anisotropic diffusion technique that can effectively remove the background noise without destroying the vessels and translucency of the contents inside the vessels; and 2) enhance the blood vessel image there by suppressing the background through nonlinear normalization. This procedure is shown in FIG. 4.

Initially, the DSA image 214 is provided for look-up-table (LUT) based diffusion 402 or anisotropic diffusion. Such diffusion 402 is described in relation to equations 3-6. Such an anisotropic diffusion and is based on partial differential equation (PDE) for noise smoothing. Given an image l(x,y,t) at time scale t, the diffusion process is expressed as:

$\begin{array}{cc}\frac{\partial }{\partial t}I\left(x,y,t\right)=\mathrm{div}\left(c\left(x,y,t\right)\nabla I\right)& \left(3\right)\end{array}$

where ∇ is the gradient operator, div is the divergence operator, and c(x,y,t) is the diffusion coefficient at location (x,y) at time t, With applying the divergence operator, Eq. (3) can be rewritten as

$\begin{array}{cc}\frac{\partial }{\partial t}I\left(x,y,t\right)=c\left(x,y,t\right)\Delta I+\nabla c\left(x,y,t\right)\nabla I& \left(4\right)\end{array}$

where Δ is the Laplacian operator. The diffusion coefficient c(x,y,t) is the key in the smoothing process and it should encourage homogenous-region smoothing and inhibit the smoothing across the boundaries. It is chosen as a function of the magnitude of the gradient of the brightness function, i.e.

c(x,y,t)=g(∥∇I(x,y,t)∥)  (5)

The suggested functions for g(·) are the following two

$\begin{array}{cc}g\left(\nabla I\right)={}^{-{\left(\frac{\nabla I}{K}\right)}^2}\mathrm{or}g\left(\nabla I\right)=\frac{1}{1+{\left(\frac{\nabla I}{K}\right)}^2}& \left(6\right)\end{array}$

where K is the diffusion constant which controls the edge magnitude threshold. Generally speaking, a larger K produces a smoother result in a homogenous region than a smaller one. Here we apply diffusion technique on the input DSA images to smooth background thereby reducing the structured noise. Also, the algorithm may be implemented using compute unified device architecture (CUDA), which is developed by nVidia to run general purpose computations on a Graphics Processing Unit (GPU). The result of the diffusion process 402 is a diffused image 404.

To increase the contrast between the blood vessels of and background in the DSA image, a nonlinear normalization method 406 is used. The principle is to force the background to suppress the non-vascular structures and noise, while gradually enhancing the foreground vascular structures. The nonlinear normalization 406 is currently implemented as a LUT filter that is based on the pre-defined parameters and is set forth in relation to equation 7.

Letting I_in(x,y) be the input DSA image (after diffusion), the nonlinear normalization, defined as I_out(X,Y) is mathematically given as:

$\begin{array}{cc}{I}_{\mathrm{out}}\left(x,y\right)=\{\begin{array}{c}{I_t\left(\frac{I_{in}\left(x,y\right)}{I_t}\right)}^{y_1},I_{in}\left(x,y\right)\in \left[0,I_t\right]\\ \left(I_t+1\right)+\left(I_t+1\right){\left(\frac{I_{in}\left(x,y\right)-\left(I_t+1\right)}{I_t+1}\right)}^{y_2},I_{in}\left(x,y\right)\in \left[I_t+1\text{,}255\right]\end{array}& \left(7\right)\end{array}$

Here I_i is a pre-defined threshold, y₁>1.0 and y₂<1.0 for class-based contrast enhancement. So if the intensity range of dye lies in lower half of the image and the background lies in the higher half of the image, the blood vessels are enhanced in limits. The result of this non-linear normalization is the enhanced DSA image 218.

After these processes, the DSA image quality has been improved greatly. Such improvement in the image facilitates QCA measurement as set forth herein. However, it will be appreciated that the QCA measurement set forth herein is considered novel in and of itself.

Set forth herein are two semi-automatic algorithms to compute percentage stenosis with minimum user intervention. As shown in FIG. 5, the overall process flow for QCA is as follows. Using a DSA image, which may be an enhanced image as set forth above, a user or operator may select 502 one frame from the DSA to use in stenosis estimation. This beings the QCA process 504. Once the frame is selected a user selects a region of interest 506. This may entail zooming in 508 on the region of interest to provide an enlarged view of the region of interest 510. This is illustrated in FIG. 6A which shows a frame having a blood vessel 600 that includes a lesion region 602 or stenosis.

The operator can then choose “measure from centerline” (center line method 512), or “measure from sides” (vessel edge method 514). If “measure from centerline” is chosen the operator needs to select 516 several points 604 along the centerline of vessel 600 as illustrated in FIG. 6B. If “measure from sides”, is chosen the operator needs to select 516 several points 604 along the both sides of vessel 600 as illustrated in FIG. 6C. The QCA process 518 is then performed by the system which subsequently generates a final report 222 that indicates a percentage of stenosis. The final stenosis results is computed and displayed automatically.

The QCA measurement process is set forth in FIG. 7. With the initial points selected by the user, the QCA measurement includes the following three sub-processes: (1) initial edge detection 720; (2) edge refinement 740; and (3) lesion measurement 760.

When the user chooses initial points 604 along the centerline 610, initial edge detection in QCA is performed by calculating the gradients of in a direction perpendicular to the centerline across the lumen and finding their peak values. See FIG. 8A. As will be appreciated, the contrast between the dyed vessel and edge is usually significant and easily detectable. That is, the vessel is often dark or nearly black in an image whereas the vessel wall and surrounding area is grey or white. This process is performed along the centerline 610 until edge points 612 are determined along both edges of the vessel 600 through the stenosis region 602. See FIG. 8B

FIG. 9 shows the process of initial edge detection. Initially, after the centerline points 604 are selected and connected to form the centerline 610, perpendicular line construction 902 is performed to form lines perpendicular 904 to the centerline. On each perpendicular line, the weighted sum of the first and second gradients is calculated 906 for the subject target line. The peak value is used to find the resulting initial boundary or initial edge points 612. These initial edge points are detected by using Eq. (8).

$\begin{array}{cc}{f}_{\mathrm{edge}}\left(x,y\right)=\underset{f\left(x,y\right)}{\arg \max }\left(w_1\nabla f\left(x,y\right)+w_2{\nabla }^2f\left(x,y\right)\right)& \left(8\right)\end{array}$

If the user chooses initial points along the two boundary lines, those points are used as the initial edge points and above initial edge detection step is omitted.

After initial edge point detection, a group of edge points are identified along two boundaries of the vessel. See FIG. 8B. These initial points are refined 740 (see FIG. 7) to get a smooth boundary curve. This process is set forth in FIG. 10. This process uses an active contour model algorithm, which will be known to one skilled in the art, that deforms a contour to lock onto features of interest within in an image. Usually the features are lines, edges, and/or object boundaries. Given an approximation of the boundary of an object in an image (e.g., initial point 612), an active contour model can be used to find the “actual” boundary. An active contour is an ordered collection of n points in the image plane:

V={v₁, v₂ . . . v_n}

v _i=(x_i,y_i), i=1, . . . n  (9)

The points in the contour iteratively approach the boundary of an object through the solution of an energy minimization problem. For each point in the neighborhood of v_i, an energy term 902 is computed:

E _i =αE _int(v_i)+βE_ext(v_i)  (10)

where E_int(v_i) is an energy function dependent on the shape of the contour and E_ext(v_i) is an energy function dependent on the image properties, such as the gradient, near point v_i. α and β are constants providing the relative weighting of the energy terms.

The internal energy function is intended to enforce a shape on the deformable contour and to maintain a constant distance between the points in the contour. Additional terms can be added to influence the motion of the contour. The internal energy function used herein is defined as follows:

αE_int(v_i)=bE_con(v_i)+cE_cur(v_i)  (11)

where E_con(v_i) is the continuity energy that enforces the shape of the contour and E_cur(v_i) is a curvature energy that causes the contour to grow or shrink. c and b provide the relative weighting of the energy terms.

The external energy function attracts the deformable contour to interesting features, such as object boundaries, in an image. Here image gradient is used. The image gradient should be large at the object boundary (β, <0). Therefore, the following external energy function is investigated:

βE_int(v_i)=βE_grad(v_i)  (12)

In summary, energy function at each v_i is minimized:

E _i =bE _con(v_i)+cE_cur(v_i)+βE_grad(v_i)  (13)

In (8), b>0,c>0 and β<0.

This process iteratively updates the initial edge points 904 until the energy function is minimized or a maximum number of iterations are achieved The result are refined edge points 910 as illustrated in FIG. 8C. That is, the points are smoothed and can be fitted with a curve.

After these first two processes 720, 740, the edges of the lumen are identifies a dimensions need to be calculated for lesion measurement 740. See FIG. 7. If dealing with a lesion, it is usually desirable to identify at least the minimum luminal diameter (MLD), the percent stenosis and perhaps the length of the lesion.

One method to determine lumen dimensions is set forth in the process flow sheet of FIG. 11. Using the detected boundaries 1102 as set forth by the refined edge points, a centerline is constructed 1104 between the two edges that represents the initial centerline 1106 of the lumen. Diameters are then measured by constructing lines that are perpendicular the centerline. Then, it is determined where each diameter line intersects 1108 each of the two edges. This defines a set of corresponding points on the boundary 1110. The distance is calculated 1112 between the two points of intersection for each line. The distance having the shortest length is the minimum luminal diameter. The MLD may be divided by the normal diameter of the vessel to provide a stenosis percentage 1114. Generally, for each lesion, the following data are calculated: (a) Minimum and maximum diameter; (b) Lesion length (the beginning and end of the lesion are determined automatically); (c) Stenosis reference (“normal”) diameter; and (d) Percent stenosis. In any case, this information may be output to the operator.

The proposed system provides a number of advantages. For instance, may be used in the quantitative measurement of a blood vessel and the system includes three sub-systems: motion compensation, image enhancement, and stenosis measurement. The whole system results in a more intelligent and more accurate system for improving QCA measurement. Another advantage is that the proposed system can involve a registration step to align the mask and bolus images prior to subtracting procedure and QCA measurement. This motion compensation step can reduce the artifacts from misalignment (due to patient movement) and improve the accuracy of QCA measurement. Another advantage is that the proposed system can involve an image enhancement step prior to QCA measurement; the resulted DSA image is further enhanced by background diffusion and nonlinear normalization for better visualization. This image enhancement increases the contrast between the blood vessels and the background. This results in much improved contrast and very crisp subtraction images, in which the regions of interest are easily identifiable.

It will be appreciated that the entire software architecture in the proposed system may be implemented on a GPU based framework, which makes visualization and computation faster by up to factor of 30. The entire scheme may therefore be implemented as a real-time scheme.

The foregoing description of the present invention has been presented for purposes of illustration and description. Furthermore, the description is not intended to limit the invention to the form disclosed herein. Consequently, variations and modifications commensurate with the above teachings, and skill and knowledge of the relevant art, are within the scope of the present invention. The embodiments described hereinabove are further intended to explain best modes known of practicing the invention and to enable others skilled in the art to utilize the invention in such, or other embodiments and with various modifications required by the particular application(s) or use(s) of the present invention. It is intended that the appended claims be construed to include alternative embodiments to the extent permitted by the prior art.

Claims

1. A method for use in the quantitative measurement of a blood vessel blockage, comprising: operating an imaging system to obtain a series of mask images and a series of bolus images corresponding to a region of interest;operating a processor to register the mask images to the bolus images to generate a registered mask image and subtract the registered mask image form the bolus image to produce a motion compensated DSA image;operating the processor to process the DSA image to enhance contrast in at least the region of interest of the DSA image;receiving a user input identifying a lumen within the region of interest; andbased on the user input, operating the processor to autonomously identify a boundary of the lumen and calculate a stenosis percentage for the lumen.

2. The method of claim 1, wherein registering the series of mask images and the series of bolus images further comprises averaging said series to produce an average mask image and an average bolus image, wherein the average mask image is registered to the average bolus image.

3. The method of claim 1, wherein registering comprises applying an inverse-consistency constraint to a deformation of the mask frame to the bolus frame.

4. The method of claim 3, wherein the inverse-consistency constraint comprises a B-spline parameterization.

5. The method of claim 1, wherein enhancing the region of inertest comprises: operating the processor to diffuse the DSA image to remove background noise from the DSA image.

6. The method of claim 5, further comprising: operating the processor to perform a non-linear normalization on the DSA image, wherein a contrast between the lumen and the background of the DSA image is increased.

7. The method of claim 1, wherein receiving the user input identifying the lumen within the region of interest comprises one of: receiving at least two user selected points associated with a centerline of said lumen;receiving a plurality of user selected points associated with edges of the lumen.

8. The method of claim 7, wherein operating the processor to autonomously identify a boundary of the lumen and calculate a stenosis percentage for the lumen further comprises at least one of: initial edge detection;edge refinement; andlesion measurement.

9. The method of claim 8, wherein initial edge detection based on the at least two points associated with the centerline of the lumen, comprises defining an initial centerline in the lumen;projecting a plurality of perpendicular lines relative to the centerline; andfor each perpendicular line, identifying a boundary point between the lumen and the background based on gradients calculated along said perpendicular line, wherein a plurality of the boundary points define an initial edge of the lumen.

10. The method of claim 8, wherein edge refinement comprises fitting a contour to the initial edge, wherein the contour is smoothed to define a refined edge.

11. The method of claim 8, wherein lesion measurement comprises: using the processor to calculate (a) minimum and maximum diameters of said lumen;(b) lesion length;(c) a stenosis reference (“normal”) diameter; and(d) percent stenosis.

12. The method of claim 1, wherein the steps of operating a processor comprise operating a GPU based processor.

13. The method of claim 1, wherein said series of mask and bolus frames are acquired using at least one of the following imagining modalities: X-ray;CT;MRI; andultrasound.

14. The method of claim 1, wherein operating a processor to register the mask image to the bolus images to generate a registered mask image comprises: registering a 3-D mask image to a 3-D bolus image to generate a 3-D registered mask image.

15. The method of claim 1, wherein operating a processor to register the mask image to the bolus images to generate a registered mask image comprises: performing an initial registration at a reduced resolution to align global structures; andperforming a secondary registration to align local structures in the region of interest.

16. A system for use in the quantitative measurement of a blood vessel blockage, comprising: an imaging system operative to obtain a series of mask images and a series of bolus images corresponding to a region of interest;an image processing system operative to: register the mask image to the bolus images and generate a registered mask image;subtract the registered mask image form the bolus image to produce a motion compensated DSA image;process the DSA image to enhance contrast in at least the region of interest of the DSA image;receive a user input identifying a lumen within the region of interest; andbased on the user input, operating the processor to autonomously identify a boundary of the lumen and calculate a stenosis percentage for the lumen.

17. The system of claim 16, wherein the imaging system comprises an X-ray imaging system.

18. A method for use in the quantitative measurement of a blood vessel blockage, comprising: operating an imaging system to obtain a series of mask images and a series of bolus images corresponding to a region of interest;operating a processor to register the mask images to the bolus images to generate a registered mask image and subtract the registered mask image form the bolus image to produce a motion compensated DSA image;receiving a user input identifying a lumen within the region of interest; andbased on the user input, operating the processor to autonomously calculate: (a) minimum and maximum diameters of said lumen;(b) lesion length;(c) a stenosis reference (“normal”) diameter; and(d) percent stenosis of the lumen.

19. The method of claim 18, further comprising identifying edges for the lumen by: defining an initial centerline in the lumen;projecting a plurality of perpendicular lines relative to the centerline; andfor each perpendicular line, identifying a boundary point between the lumen and the background based on gradients along said perpendicular line, wherein a plurality of the boundary points define an edge of the lumen.