FR2884950A1 - METHOD FOR RECONSTRUCTING A 3D IMAGE OF A STENT AND DEVICE EMPLOYING IT - Google Patents

Abstract

The invention relates to a method for reconstructing a three-dimensional image of a stent of a patient from projection images (I1, I2, ..., In) acquired from points (S1, S2,. .., Sn) of different views.Un accordance with the invention, - the actual position (P1, P2, ..., Pn) of a reference of the guide wire of the stent is detected in each image (I1, I2, ..., In), each projection image (I1, I2, ..., In) is corrected by an offset (D1, D2, ..., Dn) corresponding to the displacement from the detected position (P1, P2, ..., Pn) at a virtual position (Q1, Q2, ..., Qn) corresponding to a virtual and fixed position (M0) of the marker, which is calculated in space and identical for all the images, a three-dimensional image of the stent is reconstructed on the basis of corrected images of their associated offset (D1, D2,..., Dn).

Description

The invention relates to a method for reconstructing an image

  three-dimensional stent inserted into an artery or vessel of a patient.

  These stents, or vascular expanders (in English: stents) are deployed in arteries to expand their diameter where they have stenoses. This eliminates the effect of a stenosis on blood circulation. However, the benefits of this method may disappear in the following months or years if the disease resumes.

  Various factors can cause tissue proliferation at the stent. This intimate proliferation of tissue can be controlled at the stent by providing active stents that release a substance into the blood vessel.

  Intravascular ultrasound imaging techniques have shown that many stents may be misplaced along the wall of the blood vessel and that it is necessary to dilate the stent with a higher pressure level in the balloon. Properly Positioning the Device High pressure dilation is now widely used and the combination of active stents and high-pressure dilatation has significantly reduced restenosis internal to the stent.

  However, there is still a significant percentage of problem cases. Indeed, some patients may have a very calcified plaque at the stent site. This plate prevents satisfactory deployment of the stent, even with regular high pressure expansion. It may also happen that some of the stent transoms are not properly placed along the vessel wall and partially obstruct circulation in the vessel. These sleepers can cause thrombosis and trigger a fatal clinical event.

  There is therefore a need for a tool to analyze in detail the shape of the stent deployed in the patient's blood vessel.

  The projective image provided by current vascular equipment does not satisfy this need, because the different structures of the stent are superimposed because of the projective nature of the images. In addition, the stents have a very low contrast on the images. Therefore, their structure is difficult to analyze.

  One solution for solving these problems is to reconstruct an image of the stents from X-ray images. It may be envisaged to apply a tomographic algorithm, or an analytic or iterative approach to a set of projection images obtained from from several points of view. These projection images can be obtained for example by an image acquisition apparatus rotating around the patient, the acquisition rate being for example typically 30 frames per second.

  One solution is to follow the stents in the images. However, this solution is not applicable in practice, because the stents have a very low contrast and it is extremely difficult to find an algorithm for locating the stents on the images.

  US-B-6,385,285 discloses a method of reconstructing a three-dimensional image of a stent in a blood vessel in which images are acquired by rotating a camera around the patient. In this method, the position of the stent is located from one image to another, thereby predicting the position of the stent from one image to the next. The three-dimensional image of the stent is then reconstructed from the acquired images and positions of the stent having been successively located in the successive images.

  The reconstruction method taught by this document is difficult to implement in practice, since the stent itself appears with low contrast in the acquired images.

  An object of the invention is to obtain a method for reconstructing a three-dimensional image of a stent from projection images acquired from different directions, which overcome the drawbacks of the state of the art.

  A first object of the invention is a method of reconstructing a three-dimensional image of a stent, present in a region of interest of a patient, from a plurality of projection images acquired from a plurality of different points of view around the region of interest, characterized in that - the actual position of a marker of the stent guidewire is detected in each projection image, - each projection image is corrected by a offset corresponding to the calculated displacement, ranging from the detected actual position of the stent guidewire mark on this projection image to a virtual position of the stent guidewire mark on that projection image, corresponding to a virtual position and sets the stent guidewire mark, which virtual position of the stent guidewire mark is calculated in space and the same for all s projection images, - a three-dimensional image of the stent is reconstructed on the basis of projection images corrected for their associated offset.

  The invention thus makes it possible to detect the position of the stent itself in each projection image and uses the marker of the guide wire already present with the stent, having a contrast much better than this, to compensate in each projection image. the movements of the vessel in which the stent is located.

  According to other features of the invention, the projection images are acquired by rotation of an image acquisition system with respect to a support of the patient; the image acquisition system comprises an X-ray source, which forms each point of view and which is oriented towards and fixed with respect to radiographic recording means, the assembly formed by the X-ray source and the X-ray taking means being rotated relative to the patient support, so that the plurality of projection images are acquired by the X-ray taking means when the x-ray source occupies said different viewpoints. ; the virtual position of the guide wire mark of the stent is calculated to lie on the axis of rotation of the image acquisition system with respect to the patient's support; or the virtual position of the reference mark of the stent guidewire is calculated to lie within the region defined by the projection lines of the actual positions of the reference mark of the stent guidewire on the projection images; or the virtual position of the reference of the stent guidewire is calculated to be the closest of at least three of the projection lines of the actual positions of the stent guidewire mark on the projection images; or the virtual position of the reference of the stent guide wire is calculated to be located in the center of the circle inscribed in at least three projection lines of the actual positions of the stent guide wire mark on the projection images; projection images are acquired with a fixed collimator for the plurality of viewpoints; or the projection images are acquired with a collimator applying a collimation field varying for the plurality of points of view, to be adjusted to the contours of the region of interest.

  A second object of the invention is a medical imaging device, comprising means for acquiring images from a plurality of different points of view, characterized in that it comprises means for implementing the method of reconstruction of a three-dimensional image as described above.

  The invention will be better understood on reading the description which follows, given solely by way of non-limiting example with reference to the accompanying drawings, in which: FIG. 1 is a diagrammatic view in longitudinal section of an artery of In one patient, in which a stent is deployed, FIG. 2 shows an image acquisition system of the stent of FIG. 1 from different points of view.

  In Figure 1, a stent or vascular expander 1 was deployed (e) against the inner wall of an artery 2 of a patient. The stent 1 has a central passage for the blood, in which there is a guidewire 3 extending along the artery 2 and passing through the stent 1.

  According to the invention, the guidewire 3 is provided with one or more marks M which are diagrammatically represented by points in the figures. These marks M are on a portion of the guide wire 3 located inside the stent 1 or outside thereof. These marks M are of course the same thickness as the guide wire 3.

  The stent 1 is shown in FIG. 1 by broken lines, since it has a very low contrast on the images taken from the patient's ROI region of interest including the artery 2 and appears very dark, or less, however, as dark as the rest of the artery 2. On the other hand, the marks M present on the images taken a much greater contrast than the stent 1. The marks M also have a contrast different from that of the rest of the guide wire 3, so that they can be detected on the images taken.

  In FIG. 2, the image acquisition system is, for example, rotatable around the patient's ROI region of interest, containing the artery 2 and the stent 1. It comprises an X-ray source S directed towards means MI taking radiographic snapshots. These radiographic imaging means MI are integral and fixed with respect to the source X of X-rays, and are connected to each other by a carrier arm, not shown, located outside the X-ray path of the source. S to the MI means. The carrying arm is adapted to be set in motion by any appropriate means for moving together the source S and the radiographic imaging means MI, for example around an axis of rotation situated between the source S and the capture means MI of view. The system can thus occupy successively the n positions (Si, MIl), (S2, MI2),..., (Sn, MIn), where n is a natural number greater than or equal to two.

  The patient is positioned on a fixed support with respect to the source S and the means MI, so that the region of interest ROI is in the emission field of the source S at the various positions thereof. and can be globally visulized by the MI means.

  For each position of the source S and the means MI, the source S thus projects an image of the region ROI of interest on the means MI according to angles of view Al, A2, ..., An of different orientations, corresponding at the different positions S1, S2,..., Sn of the source S with respect to the fixed reference frame of the patient support. The axis of rotation of the source S and MI means passes for example in the ROI region of interest.

  In the first position (Si, MIl) of the projection system, the MI1 means acquire a first image II of the ROI region of interest, projected by the source S situated in the S1 point of view. In this first position (Si, MI1), the mark M of the guide wire 3 of the stent 1 is in the space at a first real position M1 and is projected from the source S1 onto the first image I1 in a first projection. real P1 along a first projection line R1.

  In the second position (S2, MI2) of the projection system, the MI2 means acquire a second image I2 of the ROI region of interest, projected by the source S located in the S2 point of view. In this second position (S2, MI2), the mark M of the guidewire 3 of the stent 1 is in the space at a second real position M2 and is projected from the source S2 onto the second image I2 in a second projection P2 real according to a second projection straight line R2.

  In the third position (Sn, MIn) of the projection system, the MIn means acquire a third image In of the region ROI of interest, projected by the source S located in the Sn point of view. In this third position (Sn, MIn), the marker M of the guide wire 3 of the stent 1 is in the space at a third real position Mn and is projected from the source Sn onto the third image In in a third projection real Pn along a third projection line Rn.

  We thus obtain n real projections Pl, P2, ..., Pn of the point M1, M2, ..., Mn on the images Ii, I2, ..., In.

  Since the patient and / or the artery 2 could move from one shot to another, for example due to the patient's heartbeat, the successive real positions M1, M2, ..., Mn the mark M are not necessarily identical for each image Il, I2, ..., projection In.

  The position P1, P2,..., Pn of the projection of the same mark M of the guide wire 3 of the stent 1 is detected on the n images II, I2,..., In.

  The acquisition geometry of the different images Il, 12, ..., In is calibrated for each of the positions P1, P2,..., Pn of the acquisition system. The purpose of this step is to describe each projection by a projection matrix that establishes the link between a description of the region of interest in a three-dimensional coordinate system and its projected representation. The angles of view A1, A2, ..., An are equal in amplitude, since the source S is fixed with respect to the means MI, but have directions of orientation, defined for example by their first bisector, different in the space, for example substantially to one another in the example of Figure 2 for Al, A2 and An. Of course, a greater number of directions of orientation of the source S towards the means MI may be provided around the region of interest ROI, corresponding to positions (Si, MIi) in addition to those corresponding to the positions (S1, MI1), (S2, MI2), (Sn, MIn) illustrated in FIG. Figure 2, and located between them. The resulting images form a sequence in different successive viewpoints.

  The position in the space of a virtual, fixed and unique point MO for all the images Il, I2,..., In projection and for the different positions (S i, MI1), (S2, MI2) is then calculated. ), ..., (Sn, MIn) of the acquisition system of these images.

  This point MO corresponds to a virtual and fixed position of the marker M of the guide wire 3 of the stent 1 in the space, identical for all the images II, I2, ..., projection In.

  This point MO is for example calculated as being located inside the region defined by the straight lines R1, R2,..., Rn of projection of the successive points M1, M2,..., Mn respectively from the sources S1, S2, ..., Sn to projections P1, P2, ..., Pn. The point MO is for example calculated as being the closest of at least three of the straight lines R1, R2,..., Rn successive projections of the marks M1, M2, ..., Mn. The point MO is for example calculated as being in the center of the circle inscribed in three straight lines R1, R2, Rn of projection forming a triangle.

  The point MO having been calculated, one recalculates its projection Q1, Q2,. .., Qn on images I1, I2, ..., In from source S1, S2, ..., Sn. Thus, the n projections Q1, Q2, ..., Qn on the n images Il, I2, ..., In correspond to a virtual position on each image [1, I2, ..., In of the virtual and fixed point MO, common to all images I1, I2, ..., In projection.

  The position of the projections Q1, Q2, ..., Qn is calculated on the images I1, I2, ..., In.

  Then, for each projection image I1, I2, ..., In, an image shift D1, D2,..., Dn, ranging from the detected real position P1, P2,. reference M in the images II, I2,..., projection in the calculated virtual position Q1, Q2,..., Qn of the calculated point MO for all the images Il, I2,..., projection In. The n offsets D1, D2,. .., Dn associated respectively with n images Il, I2, ..., In projection are most often different from each other.

  Each image is then applied its associated shift, that is to say to the image II the shift DI ranging from P1 to Q1, to the image I2 the shift D2 going from P2 to Q2 and to the image In the Dn shift from Pn to Qn. Thus the set of each image Il, I2, ..., In is displaced by the same respective offset D1, D2,..., Dn.

  It is thus performed a readjustment of the projection matrices, so that the location of the guide wire 3 is consistent with the description of the projection given by the projection matrices.

  A three-dimensional image is then constructed from the projection images Il, I2,..., In which they have been shifted by their own offset D1, D2,..., Dn previously calculated.

  For this reconstruction, for example, a tomographic algorithm is used.

  Many techniques are known to those skilled in the art for reconstructing a three-dimensional image of an object from projection images thereof on planes.

  One can thus describe for example the three-dimensional image to be obtained by a regular grid of elementary volume elements called voxels. Each voxel has its own level of light intensity. The pixels of each projection image have, as a level of luminous intensity, light intensity levels determined according to those of the voxels. By making the inverse transformation using the projection matrices determined for each projection image, the light intensity levels of the voxels of the three-dimensional image are recalculated. Reference may be made to US-B 1-6,320,928 to reconstruct the three-dimensional image.

  It may be provided, in order to limit as much as possible the emission of ionizing X-rays to the patient due to the acquisition of the images, to collimate the emission of X-rays from the source S. For this purpose, it can be provided a fixed collimation field. This collimation field is large enough to adapt to the movements of the stent throughout the acquisition of the images. In order to keep the collimation field as small as possible, it may be provided that the stent 1 is positioned as close as possible to the center of rotation of the image acquisition system (S, MI).

  Another possibility would be to adapt the collimation field throughout the rotation of the acquisition system in order to just project the image limited to the stent.

  An analogous method can be used with an artery in which a contrast medium has been injected. With the aid of the three-dimensional image having been reconstructed of the stent, it will then be possible to reposition the stent with respect to the local vascular architecture of the patient, before or after surgery.

Claims (10)

  1. A method for reconstructing a three-dimensional image of a stent (1), present in a region of interest (ROI) of a patient, from a plurality of images (II, I2, .. ., In) acquired from a plurality of points (Si, S2, ..., Sn) of different views around the region (ROI) of interest, characterized in that - the actual position (P1, P2, ..., Pn) of a mark (M) of the guide wire (3) of the stent (1) is detected in each projection image (II, I2, ..., In), - each image (II , I2, ..., In) is corrected by an offset (D1, D2, ..., Dn) corresponding to the calculated displacement, ranging from the detected actual position (P1, P2, ..., Pn) of the reference (M) of the guidewire (3) of the stent (1) on this projection image (II, I2, ..., In) to a virtual position (Q1, Q2, ..., Qn) of the reference (M) of the guide wire (3) of the stent (1) on this image (II, I2, ..., In) projection, corresponding to a posi virtual and fixed reference (M) of the guide wire (3) of the stent (1), which virtual position (MO) of the marker (M) of the guide wire (3) of the stent (1) is computed in space and identical for all projection images (II, I2, ..., In), - a three-dimensional image of the stent (1) is reconstructed on the basis of the images (II, I2, ..., In) of projection corrected by their associated offset (Dl, D2, ..., Dn).   2. A method for reconstructing a three-dimensional image according to claim 1, characterized in that the projection images (Il, I2, ..., In) are acquired by rotation of an acquisition system (S, MI). images relative to a patient support.   3. A method for reconstructing a three-dimensional image according to claim 2, characterized in that the image acquisition system (S, MI) comprises a source (S) of X-rays, which forms each point of view (Si , S2, ..., Sn) and which is oriented towards and fixed with respect to radiographic imaging means (MI), the assembly formed by the source (S) of X-rays and the means (MI) of radiographic images being rotated relative to the patient support, so that the plurality of projection images (11, 12, ..., In) is acquired by the radiographic imaging means (MI) when the source (S) of X-rays occupies said points (Si, S2, ..., Sn) with different views.   4. Method for reconstructing a three-dimensional image according to claim 2 or 3, characterized in that the virtual position (MO) of the marker (M) of the guide wire (3) of the stent (1) is calculated to lie on the axis of rotation of the image acquisition system (S, MI) with respect to the patient support.   5. Method for reconstructing a three-dimensional image according to any one of claims 1 to 3, characterized in that the virtual position (MO) of the marker (M) of the guide wire (3) of the stent (1) is calculated to be located within the region defined by the straight lines (R1, R2, ..., Rn) of projection of the real positions (P1, P2, ..., Pn) of the guide wire mark (M) (3) of the stent (1) on the projection images (11, 12, ..., In).   6. A method for reconstructing a three-dimensional image according to any one of claims 1 to 3, characterized in that the virtual position (MO) of the marker (M) of the guide wire (3) of the stent (1) is calculated to be the closest of at least three of the straight lines (R1, R2, ..., Rn) of projection of the real positions (P1, P2, ..., Pn) of the reference (M) of the guide wire (3 ) of the stent (1) on the projection images (II, I2, ..., In).   7. A method for reconstructing a three-dimensional image according to any one of claims 1 to 3, characterized in that the virtual position (MO) of the reference (M) of the guide wire (3) of the stent (1) is calculated to be located at the center of the circle inscribed in at least three straight lines (R1, R2, ..., Rn) of projection of the real positions (P1, P2, ..., Pn) of the reference (M) of the guide wire ( 3) of the stent (1) on the projection images (I1, 12, ..., In).   8. A method for reconstructing a three-dimensional image according to any one of claims 1 to 7, characterized in that the images (11, 12, ..., In) projection are acquired with a fixed collimator for the plurality of points (Si, S2, ..., Sn) of view.   9. A method for reconstructing a three-dimensional image according to any one of claims 1 to 7, characterized in that the images (Il, I2, ..., In) projection are acquired with a collimator applying a collimation field varying for the plurality of points (Si, S2, ..., Sn) of view, to be adjusted to the contours of the region (ROI) of interest.   10. Medical imaging device comprising means for acquiring images from a plurality of points (Si, S2,..., Sn) of different views, characterized in that it comprises means for setting implementation of the method for reconstructing a three-dimensional image according to any one of claims 1 to 9.