Automatic registration method of 3D model in display equipment of medical navigation system

Abstract

The invention relates to a computer-implemented method for displaying a 3D model of an organ of a patient on a navigation system display device (15), the method comprising: (SOI) receiving spatial coordinates (Sxvz) from a sensor (131) of a catheter (13) within a portion (CS) of said organ, the spatial coordinates defining a reference point cloud (RPC) of said portion of said organ; (505) receiving a plurality of points defining a 3D model (IH) of the organ, marking a subset of the points to define a segmentation (Res) corresponding to the portion of the organ; (51) calculating a spatial transformation (T) that minimizes an error metric (E) between the result (T (Rcs)) of the spatial transformation over the segmentation and the reference point cloud; -(52) displaying the landmark (LM) at the current position of the catheter on the display device, and simultaneously displaying the result (IH1) of the spatial transformation on the 3D model.

Description

[0001] The present invention relates to the technical field of medical image registration, and more particularly to a method and system for registering a 3D model of a patient's organ on a display device of a medical navigation system.

[0002] In the case of medical interventions on a patient's organs using a catheter, and in particular for catheter cardiac ablation, it is known to prepare for the intervention using a virtual 3D model of the organ. For example, a surgeon can prepare for cardiac ablation in advance, thanks to automatic or manual identification of ablation targets on a virtual 3D model of the heart.

[0003] Such virtual 3D models are usually constructed from a 3D image of the organ, obtained from a medical imaging method such as computed tomography or CT scan, fluoroscopy or MRI. The 3D image is processed to obtain a set of vertices defining the 3D model of the organ. The processing of the 3D image can also highlight the features of the organ, which can help surgeons detect medical intervention targets. For example, patent application EP3622478 from the applicant discloses a method for calculating a 3D model of a patient's heart, and in which the myocardial thickness is modeled to identify low thickness areas or channels in the myocardial wall.

[0004] Typically, the surgeon then introduces a probe or catheter into the patient's organ. The catheter allows the navigation system to generate an image of one or more regions of interest of the organ, which is displayed on a screen to help the surgeon assess the location of the target. In order to improve the accuracy of the intervention, a 3D model of the medical intervention that has been prepared must be imported into the navigation system.

[0005] However, the 3D model must be registered with the reference system used by the navigation system. For this purpose, the navigation system operator manually moves and rotates the 3D model to match the region of interest generated by the navigation system as best as possible. Therefore, such manual registration is slow, complex and inaccurate.

[0006] In view of the above, there is therefore a need for a method of automatically registering a 3D model of a patient's organ in a reference system used by a medical navigation system.

[0007] The object of the present invention is to meet this need.

[0008] To this end, the subject of the invention is a computer-implemented method for displaying a 3D model of an organ of a patient on a navigation system display device, the method comprising:

[0009] a. receiving a plurality of spatial coordinates calculated from data received from at least one sensor of a catheter within a portion of said organ of a patient; the plurality of spatial coordinates defining a reference point cloud representing said portion of said organ of the patient;

[0010] b. receiving a plurality of points, the plurality of points defining a predefined 3D model of the organ of the patient, marking a sub-plurality of points to define a 3D model segmentation corresponding to the portion of the organ of the patient;

[0011] c. computing a spatial transformation that, when applied to the 3D model, minimizes a predefined error metric between the result of the spatial transformation on the 3D model segmentation and a reference point cloud;

[0012] d. Displaying the landmarks at the current position of the catheter obtained from the plurality of spatial coordinates on a navigation system display device, and simultaneously displaying the result of the calculated spatial transformation on the 3D model.

[0013] When a catheter equipped with one or more positioning sensors is introduced into a tubular structure of an organ, the position of the sensor is recorded during the movement of the catheter until it reaches its final position. The position of the sensor corresponds to the spatial coordinates of the sensor of the catheter in the reference system or coordinate system of the patient. The catheter may include two or more sensors arranged along the catheter body. For example, most image-guided ablation procedures involve guiding the catheter into a cardiac vessel such as the coronary sinus. The entire set of spatial coordinates recorded during the complete movement of the catheter then corresponds to a point cloud representing all or part of the tubular structure. Such tubular structures are also part of a predefined 3D model of the organ, and the surgeon uses the model to prepare for intervention. Assuming that some vertices of the 3D model have been correctly marked as belonging to the tubular structure, the tubular structure can be segmented from the 3D model. The present invention therefore proposes to calculate a spatial transformation, which, when applied to a 3D model, transforms the 3D model so that the 3D model vertices marked as tubular structures match the point cloud representing the tubular structure. The result of this spatial transformation on the 3D model is therefore a 3D model that is automatically registered in the same reference system as the spatial coordinates. Therefore, it can be displayed on the navigation system display together with the current position of the catheter in the same patient's reference frame.It can be observed that the same registration can be obtained by minimizing a predefined error metric between the result of the inverse space transformation on the reference point cloud and the 3D model segmentation.

[0014] In the context of this specification, unless otherwise expressly provided, "computer" may refer to, but is not limited to, a desktop computer, a laptop computer, a tablet computer, a type of test equipment, a network appliance, a controller, a digital signal processor, a computing engine within an appliance, a consumer electronic device, a portable computing device, and / or another electronic device, and / or any combination thereof suitable for the relevant task at hand. Furthermore, the method steps may all be performed on the same device. As a variant, the method steps may be performed on several devices connected by wire or by wireless connections.

[0015] In the context of the present specification, a "3D model of an organ" is a digital representation of the organ, preferably a structured or unstructured mesh comprising a plurality of vertices connected to each other to define a plurality of faces and / or volumes that approximate the internal and / or external surfaces and / or volumes of the organ. The 3D model may be a triangular mesh or a quadrilateral mesh, or more generally a polygonal mesh, comprising a plurality of vertices, having edges connecting pairs of vertices and / or polygonal faces connecting closed sets of vertices. As a variant, the 3D model may be a tetrahedral mesh or a hexahedral mesh or a prismatic mesh, or more generally a volumetric mesh, comprising a plurality of vertices, wherein polyhedra connect a closed set of polygons defined by a closed set of vertices. Preferably, each vertex and / or edge and / or polygon and / or polyhedron may be marked as being part of a particular sub-portion of the organ. A 3D model may be stored as a list of vertices, each of which is associated with a spatial coordinate and with a set of connections to other vertices, or as a list of vertices associated with spatial coordinates and a list of polygons or faces, each of which is defined by a subset of the vertices contained in the list of vertices, it being understood that any other suitable storage method may be used in the context of the present specification.

[0016] In the context of this specification, a "spatial transformation" is a geometric transformation that maps an initial set of points to a final set of points and causes at least a displacement of the initial set of points. A spatial transformation may be a rigid transformation that includes at least a translation or at least a rotation or a combination of one or more rotations and one or more translations. A spatial transformation may also include linear or nonlinear deformations, such as scaling or deformation.

[0017] In the context of this specification, an "error metric" is a metric or mathematical function that evaluates the accuracy of the registration of a set of points with a set of reference points. For example, it may be based on the distance between the registered points and the reference matching points. In other words, the goal of the spatial transformation calculation step is to find a spatial transformation that best aligns the 3D model segmentation with the reference point cloud. For example, the spatial transformation calculation step may include a matching point sub-step, the goal of which is to match all or part of the vertices of the 3D model segmentation with all or part of the points of the reference point cloud, and a spatial transformation estimation sub-step based on the matched points. These sub-steps may be implemented only once using direct equation solving, or may be implemented iteratively.

[0018] According to one embodiment, the landmark displayed on the navigation system can be an image, icon or logo displayed at the current sensor position, or a 3D model of a catheter with its tip displayed at the current sensor position. Alternatively or cumulatively, a reference point cloud can be displayed with the results of the calculated spatial transformation superimposed on the 3D model.

[0019] According to one embodiment, the method comprises a preliminary step of acquiring a 3D image and / or recording a 3D image of an organ of the patient, in particular of the patient's heart or a region of the patient's heart, from which the 3D model is calculated. The 3D image may be acquired directly from an image acquisition device such as a scanner or MRI. Alternatively, the 3D image may be obtained from a recording medium on which it is stored, such as a local memory or a remote database, the 3D image having been acquired prior to the method.

[0020] For example, the step of acquiring a 3D image can be performed by tomography. These techniques are also known as CT scans or CAT scans and are based on the measurement of X-ray absorption by organ tissue. Tomography provides a plurality of 2D images, each representing a slice of an organ, which are then combined to reconstruct a 3D image of the anatomical structure of the observed organ. The 3D image comprises a volume distribution of pixels or voxels.

[0021] The method according to the invention may therefore comprise a 3D modeling step of the 3D image of the organ to generate a 3D model, for example comprising modeling at least one layer or one wall of the organ shown in the 3D image, such as the inner and outer faces of a layer of the organ. The 3D model may be a mesh, in particular a polygonal mesh or a polyhedral mesh, of the layer, in particular the inner and outer faces.

[0022] In one example, the method according to the invention may comprise the step of segmenting one or more regions of interest in a 3D image or a 3D model.

[0023] According to one embodiment, the method according to the invention comprises a sub-step of segmenting a plurality of volumes in a 3D image or a 3D model according to their thickness, in particular marking each voxel or each vertex according to the thickness of the volume to which it belongs, after comparing its thickness with a plurality of thickness levels. Preferably, an image processing method is implemented to detect, segment or extract at least two volumes from a 3D image or from a 3D model by taking into account for each voxel of the 3D image or for each vertex of the 3D model the thickness of the latter and by comparing this thickness with a maximum threshold or with different thickness levels. For example, the volumes may be segmented according to thickness levels of 1 mm, 2 mm, 3 mm, 4 mm and 5 mm.

[0024] The method of the invention allows to consider for example the wall of the heart, more generally called myocardium. In this case, a 3D image or 3D model of the heart can be segmented into different volumes according to myocardium thickness criteria, so that the 3D model represents a thickness map of the heart.

[0025] According to one embodiment, the method according to the invention comprises a step of detecting at least one region of interest in a 3D image or a 3D model, by identifying, for example, at least one region, such as the isthmus in terms of a 3D image or model of the heart, such that the at least one region is contained between two regions of smaller thickness.

[0026] Optionally, the method according to the present invention comprises the step of classifying each region of interest by:

[0027] a. The thickness of the region of interest compared to the thickness of the two adjacent regions,

[0028] b. the width of the region of interest formed by the spacing between adjacent regions,

[0029] c. its length as defined by the arrangement of adjacent areas,

[0030] d. The topology of the entrance and / or exit of the region of interest, such as a funnel formed by two adjacent regions.

[0031] In this way, regions of interest can be classified to help surgeons quickly identify intervention targets.According to one embodiment, several geometric criteria are brought together and combined in order to refine the classification of regions of interest.

[0032] Alternatively, the steps of detecting and classifying the region of interest may be at least partially performed manually by a user.

[0033] According to one embodiment, the method according to the invention comprises the step of detecting and classifying at least one tubular structure in the 3D image or 3D model. For example, in the case of a 3D model of the heart, the method may comprise the step of detecting a blood vessel such as the coronary sinus in the 3D image or 3D model.

[0034] More precisely, the step of detecting and classifying tubular structures may comprise manually or automatically detecting circles and / or ellipses in at least a subset of the 2D images constituting the 3D image, each pixel of the 2D image belonging to a detected circle or a detected ellipse being classified as belonging to a detected tubular structure, and pixels of both 2D images belonging to a detected tubular structure being classified as belonging to the same detected tubular structure, these pixels having approximately the same shape and / or having approximately the same size and / or being located approximately in the same area in the 2D images. Preferably, the step of detecting and classifying tubular structures may comprise interpolating the detected tubular structure to the remaining 2D images.

[0035] According to one embodiment, the method comprises the step of identifying one or more regions in the 3D model segmentation and / or in the reference point cloud having at least one geometric feature matching a predetermined criterion, and the calculated spatial transformation is a spatial transformation that, when applied to the 3D model, minimizes a predefined error metric between the result of the spatial transformation on the 3D model segmentation and the reference point cloud in at least the one or more regions. Considering that the tubular structure into which the catheter is inserted may have specific geometric features, such as strong curvature and / or strong torsion, or a specific shape, these geometric features may be used to enhance the registration.

[0036] For example, the geometric feature matching the predetermined criteria may be a tubular structure having a curvature value and / or a torsion value exceeding a predetermined threshold, or alternatively or cumulatively, having a shape matching a predetermined shape, such as a cone.

[0037] In one embodiment of the invention, the computed spatial transformation is one that, when applied to the 3D model, minimizes a predefined error metric between the result of the spatial transformation on the 3D model segmentation and a reference point cloud only in said one or more identified regions.

[0038] In an alternative embodiment, the calculated spatial transformation is a spatial transformation that, when applied to the 3D model, minimizes, for each point of the 3D model and the reference point cloud, a predefined error metric between a result of the spatial transformation on a segmentation of the 3D model and the reference point cloud, wherein the predefined error metric is weighted with a higher weight for points belonging to one or more identified regions.

[0039] According to one embodiment, a value is associated with each point of the 3D model segmentation and / or the reference point cloud, the value depending on the position of the associated point with respect to the one or more regions. In this embodiment, the spatial transformation calculation step comprises an iterative estimation of the spatial transformation, the iterative estimation comprising, for each iteration:

[0040] a. Compare each point of at least a portion (particularly a subset of points) of the result of the currently estimated spatial transformation on the 3D model segmentation with a point (particularly the closest point) of the reference point cloud;

[0041] The matching substeps, and

[0042] b. A sub-step of adjusting the currently estimated spatial transformation according to a predefined error metric between the result of the currently estimated spatial transformation on the 3D model segmentation and the reference point cloud, wherein the predefined error metric is based on an average distance value between the matching points weighted by the association value.

[0043] The iterative estimation may be implemented as an iterative closest point algorithm or a robust point matching algorithm or any suitable point set registration algorithm in which the error metric may be weighted. For example, the predefined error metric may be a root mean square or RMS based metric.

[0044] According to one embodiment, the step of receiving a plurality of spatial coordinates includes receiving at least a first plurality of spatial coordinates and a second plurality of spatial coordinates calculated based on data received from at least a first sensor and a second sensor of the catheter, it being noted that the first sensor and the second sensor are offset from each other; and the step of identifying one or more regions in the reference point cloud includes estimating the evolution of the curvature and / or torsion of the reference point cloud from the first plurality of spatial coordinates and the second plurality of spatial coordinates.

[0045] According to one embodiment, the calculated spatial transformation is a square matrix defining at least rotation and / or translation. For example, the matrix can be a 4x4 matrix defining rotation and translation. As a variant, the calculated spatial transformation can be a vector field, where each vector relates a point from the reference point cloud to a vertex from the 3D model.

[0046] According to one embodiment, the method comprises the further step of receiving a set of complementary local data about the patient's organ from a sensing device different from the catheter, and the further step of correcting the calculated spatial transformation based on the set of complementary local data. Thanks to these further steps, the registration can then be corrected based on information about the patient's reference frame that will be inferred from any new data captured by the sensing device.

[0047] Preferably, the step of correcting the calculated spatial transformation may be repeated periodically upon any periodic updating of the complementary local data received from the sensing device.This step may be implemented after a first calculation of the spatial transformation and after and / or before the displaying step.

[0048] By way of example and not limitation, the supplementary local data may be:

[0049] a. local electrical signals or local electrical information relating to the organ and received directly or indirectly from the same catheter, from another mapping catheter or from an imaging device,

[0050] b. local geometric features such as shape or wall thickness received directly or indirectly from an imaging device,

[0051] c. A supplementary plurality of spatial coordinates received directly or indirectly from another catheter equipped with at least one positioning sensor.

[0052] According to one embodiment, at least one region of interest corresponding to the supplementary local data may be segmented in the 3D model, and the calculated spatial transformation is a spatial transformation that, when applied to the 3D model, minimizes a predefined error metric between a result of the spatial transformation on the region of interest segmented by the 3D model and a reference point cloud associated with the supplementary local data.

[0053] For example, electrical signals on the endocardium captured by a mapping catheter guided in the heart can be used to identify areas of low voltage, slow conduction or abnormal activity. These areas are associated with the presence of scars or channels on the heart, which can be segmented on a 3D model of the heart due to their thickness, width, length and topology. Thus, the initial registration achieved using the positioning sensor of the first catheter can be corrected using this complementary information received from the other catheter.

[0054] According to one embodiment, the method comprises the further step of periodically receiving spatial coordinates calculated from data received from at least one sensor of a catheter within a portion of said organ of the patient; the further step of modeling the periodic motion of said organ of the patient according to said periodically received spatial coordinates, and the further step of modifying the displayed positions of landmarks and the result of the calculated spatial transformation on the 3D model on the display device of the navigation system based on said model of the periodic motion of said organ of the patient. Thus, the periodic motion of the patient, such as his cardiac motion and / or respiratory motion, can be compensated to avoid any disturbances in the display of the navigation system.

[0055] In a first example, the periodic motion of the organ can be modeled by calculating an average set of spatial coordinates from the received spatial coordinates over several periodic cycles. As a second example, modeling of the periodic motion of the organ can be achieved using a machine learning algorithm such as support vector regression or relevance vector machine.

[0056] According to one embodiment, the method comprises the further step of calculating a value for a registration error between the reference point cloud and the result of the calculated spatial transformation on the 3D model.Preferably, said value for the registration error is displayable on a navigation system display device.

[0057] Optionally, the value regarding the registration error may be a value of a predefined error metric between the result of the calculated spatial transformation on the 3D model and the reference point cloud.

[0058] According to one embodiment, the method comprises the further step of receiving a set of supplementary local data about the patient's organ calculated from data received from at least one sensor of a catheter within a portion of the patient's organ, each supplementary local data being associated with a spatial coordinate of the sensor, and the further step of displaying each supplementary local data on the result of the calculated spatial transformation on the 3D model displayed on a navigation system display device, the displayed position of the supplementary local data being calculated according to its associated spatial coordinates. Thus, the captured supplementary data can be projected onto the 3D model displayed on the navigation system to construct a 3D enhanced model of the organ.

[0059] For example, each supplementary local data may be displayed using an icon, a mark or a label superimposed on a vertex of the 3D model, the position of which is determined according to the spatial coordinates associated with the supplementary local data and the calculated spatial transformation. As a variant, each supplementary local data may be displayed using a modified vertex of the 3D model, the position of which is determined according to the spatial coordinates associated with the supplementary local data and the calculated spatial transformation, and its color is modified according to the value of the supplementary local data and according to a scalar or color map that associates possible values ​​of the local data with colors.

[0060] As an example and not limitation, the supplementary local data may be supplementary electrical local data. The supplementary electrical local data may be local electrical signals or local electrical information, such as local impedance relative to the organ, and received directly or indirectly from electrodes of the same catheter or from another catheter. As a variant, the supplementary local data may be supplementary mechanical local data. The supplementary mechanical local data may be local mechanical properties of the organ, such as local elasticity of the tissue, received directly or indirectly from a pressure sensor of the same catheter or from another catheter.

[0061] According to one embodiment, the method comprises a further step of calculating, for each supplementary local data, a distance between its associated spatial coordinates and one or more points of the result of the calculated spatial transformation on the 3D model (in particular a set of points defining a portion of the layer of the organ closest to the associated spatial coordinates), and a further step of calculating, from the calculated distances, a value regarding the reliability of the supplementary local data. Thanks to the value regarding their reliability, the supplementary local data can be registered on the 3D model of the organ, but weighted by this value regarding their reliability.

[0062] The subject of the invention is also a navigation system for intervention assistance, comprising at least one catheter comprising at least one positioning sensor; a display; and a computing unit designed to implement the method according to the invention, the landmarks and the results of the calculated spatial transformation on the 3D model being displayed on the display.

[0063] According to one embodiment, the catheter is a flexible and / or steerable catheter.

[0064] According to one embodiment, the navigation system comprises a transmitting unit capable of emitting a magnetic field, and the positioning sensor is a sensor capable of providing a signal in response to said magnetic field passing through the sensor and emitted by the transmitting unit, such as a coil, at least one parameter of the signal being a function of the relative position of the sensor with respect to the transmitting unit. Advantageously, the navigation system comprises a tracking unit configured to calculate the relative position of the sensor with respect to the transmitting unit, such as a spatial coordinate, from said signal provided by the sensor. Preferably, the set of all relative positions calculated by the tracking unit during the movement of the catheter forms a plurality of spatial coordinates provided to the calculation unit.

[0065] According to one embodiment, the catheter comprises at least two positioning sensors offset from each other. For example, one positioning sensor may be positioned on a first portion of the catheter and another positioning sensor may be positioned on a second portion of the catheter, the first portion and the second portion being maneuverable from each other.

[0066] According to one embodiment, the catheter comprises at least one electrode.

[0067] The invention will be better understood and other advantages will emerge after reading the detailed description of an embodiment thereof, which is given by way of non-limiting example and is illustrated by the accompanying drawings, in which:

[0068] Figure 1A is a navigation system for intervention assistance according to an embodiment of the present invention;

[0069] Figure 1B yes Figure 1A The catheter of the navigation system is shown;

[0070] Figure 2 is a diagram showing a method for performing Figure 1A A logic flow chart of a computer-implemented method for displaying a 3D model of a patient's organ on a display device of a navigation system as shown;

[0071] Figure 3 Shown in Figure 2 An initial registration step performed during the method shown;

[0072] Figure 4 Shows that in the implementation Figure 3 After the initial registration step shown in Figure 1A An image displayed on a display device of the navigation system shown;

[0073] Figure 5 Shown in Figure 2The registration correction step performed during the method shown; and

[0074] Figure 6 Shows that in the implementation Figure 2 The electrical data projection step of the method shown is followed by Figure 1A An image displayed on a display device of a navigation system is shown.

[0075] refer to Figure 1A , which shows a navigation system 1 for intervention assistance and in particular for catheter cardiac ablation assistance.

[0076] The system 1 comprises an operating table 11 on which a patient lies. The system 1 also comprises a transmitting unit or field generator 12 mounted to or disposed adjacent to the table 11, for example, the transmitting unit or field generator is positioned and / or oriented to generate an electromagnetic field that encompasses an organ of the patient, such as the patient's heart. The system 1 also comprises a catheter 13 that can be introduced into the patient's organ, for example, into the patient's heart.

[0077] The electromagnetic field generated by the transmitting unit 12 provides an external reference frame X1 , Y1 , Z1 which is substantially stationary relative to the table 11 and within which the catheter 13 can be positioned after insertion of the catheter into an organ of a patient.

[0078] The system 1 includes a controller 14 that communicates with the transmitting unit 12 and / or a catheter 13 introduced into the patient's body, for example, to identify the position of one or more sensors on the catheter 13 relative to an external reference frame.

[0079] Steering Figure 1B The catheter 13 includes a plurality of sensors 131 , the positions of which can be identified by a tracking unit of the controller 14 .

[0080] In the example shown, the catheter 13 comprises an elongated tubular member including a plurality of steerable or flexible portions 13a and a distal end 13b sized to be introduced into a tubular structure of a patient, such as the coronary sinus or the great cardiac vein. The catheter 13 may include a handle or hub at the proximal end, for example, which may include one or more actuators, connectors, etc.

[0081] Each portion 13a can be flexible and / or maneuverable to facilitate guiding the distal portion 13b within the patient's tubular structure. For example, each portion 13a can include one or more steering elements operable from the catheter handle to control the bending of the portion from a generally straight configuration to a curved configuration. The catheter 13 can thus be bent to follow the general curvature of the patient's tubular structure.

[0082] Each portion 13a of the catheter 13 includes at least one positioning sensor 131, which may be an antenna or a coil responsive to the electromagnetic field generated by the transmitting unit 12. Each sensor 131 may be coupled to a tracking unit of the controller 14 via one or more leads (not shown) extending in the lumen of the catheter.

[0083] In the example shown, each sensor 131 can be at least a three-degree-of-freedom sensor, i.e., providing a signal corresponding to the coordinate position of the sensor 131 in three-dimensional space within the reference system X1, Y1, Z1. In an alternative embodiment, one or more sensors 131 can be a six-degree-of-freedom sensor that also provides a signal about the rotational position of the sensor relative to an external reference system. In addition, any number of additional sensors capable of being positioned within a predetermined number of degrees of freedom can be provided on each portion of the catheter.

[0084] The tracking unit of the controller 14 receives the signals provided by each sensor 131 and is configured to calculate, from each received signal, the spatial coordinates of the position of the corresponding sensor 131 within the reference system X1 , Y1 , Z1 .

[0085] The system 1 also includes a display device 15 which is controlled by the controller 13 to present information to the surgeon to assist in the intervention.

[0086] When the catheter 13 contacts or approaches a tubular structure, the controller 14 may register points in an external reference frame from the sensor 131 position calculated by the tracking unit, and combine all points to create a spatial representation of the tubular structure, which may be presented on a display device 15 .

[0087] Reference now Figure 2 , which shows a computer-implemented method for displaying a 3D model of an organ of a patient on a display device 15 and executed by a computing unit of a controller 13 .

[0088] As a preliminary step S01, the catheter 13 has been introduced into the patient's heart through the coronary sinus CS and the position of each sensor 131 has been calculated by the tracking unit of the controller 14 during the movement of the catheter 13 to its current position. The positions calculated by the tracking unit during the movement of the catheter form a plurality of spatial coordinates S XYZ , the spatial coordinates are provided to the computing unit of the controller 13.

[0089] As mentioned above, it can be noted that the multiple spatial coordinates S XYZ A reference point cloud RPC is defined which represents all or part of the coronary sinus CS.

[0090] In another preliminary step S02 , a 3D image 3D_CT of the patient's heart has been acquired from a CT scanning method, and a 3D model IH representing a thickness map of the patient's heart myocardium has been generated from the 3D image.

[0091] More precisely, the inner and outer surfaces of the patient's myocardium have been modeled from the 3D image 3D_CT, wherein a plurality of vertices have been defined to form a mesh. Furthermore, a plurality of volumes in the resulting model IH have been segmented according to their thickness, wherein each vertex is labeled according to the thickness of the volume to which it belongs. The 3D model IH has its own reference system X2, Y2, Z2, which may be completely different from the reference system X1, Y1, Z1.

[0092] In a further preliminary step S03, at least one region R CH has been detected and segmented in the 3D model IH as being contained between two regions of smaller thickness and has been classified as an ablation target, taking into account its thickness, its width, its length and its entry and exit topology. The resulting 3D model IH can therefore help the surgeon prepare for a catheter ablation intervention.

[0093] In another additional preliminary step S04, the region R CS All or part of the coronary sinus CS has been detected and segmented as a tubular structure in the 3D model IH and has been classified.

[0094] It can be noted that steps S02 to S04 can be performed by other computing units besides the controller 14, and steps S03 and / or S04 can be performed completely automatically by the calculation or with manual assistance. The resulting 3D model IH can be stored on a recording medium, such as a memory of the controller 14 or an external memory support or a remote database, through which the controller 14 can access the model.

[0095] In step S05, the segmented region R CH and segmentation region R CS The 3D model IH in the reference frame X2, Y2, Z2 is set as a computing unit of the controller 14.

[0096] In step S1 , a spatial transformation T allowing registration of the 3D model IH in the reference system X1 , Y1 , Z1 of the reference point cloud RPC is calculated by the controller 14 .

[0097] The execution of step S1 has been Figure 3 Shown in.

[0098] Step S1 includes a sub-step S11, which consists in identifying a region R of the 3D model IH CSand one or more regions in the reference point cloud RPC having at least one geometric feature matching a predetermined criterion. In the example shown, sub-step S11 consists in identifying the region R CS and an area AHC of the coronary sinus CS represented by a reference point cloud RPC, the area having a curvature and / or torsion exceeding a predetermined threshold.

[0099] Considering the reference point cloud R PC , because all sensors 131 of the catheter 13 have been successively guided into the coronary sinus CS, the curvature and torsion of the reference point cloud RPC can be estimated by the controller 14 by comparing the evolution of the vector connecting the positions of each pair of sensors 131 during the movement of the catheter into the coronary sinus CS.

[0100] Considering the region R of the 3D model IH CS Algorithms for curvature extraction and / or torsion extraction from tubular 3D models are known per se and will be further developed, the algorithms consisting in extracting one or more curves from a point cloud and subsequently calculating the curvature of this / these curves. Such algorithms can also be applied to a reference point cloud R PC To evaluate the reference point cloud R PC curvature and / or torsion.

[0101] Therefore, the region R of the 3D model IH can be CS Both the reference point cloud RPC and the area AHC corresponding to the portion of the coronary sinus having a strong curvature and / or a strong torsion are segmented. Alternatively or cumulatively, other areas having specific geometrical features, such as a predefined shape, may also be segmented.

[0102] In sub-step S12, the weight is attributed to the region R of the 3D model IH CS For each vertex and each point of the reference point cloud RPC, the weights attributed to the vertices and points belonging to the area AHC are higher than the weights used for other vertices and points.

[0103] In sub-step S13, an iterative closest point algorithm is executed by the controller 14 to iteratively estimate a spatial transformation T that, when applied to the 3D model IH, minimizes the spatial transformation in the region R CS The result of the spatial transformation T on T(R CS ) and the reference point cloud RPC, the predefined error metric taking into account the weights attributed to the vertices and points in sub-step S12.

[0104] For example, the predefined error metric E can be calculated using the following formula:

[0105]

[0106] Where E is the root mean square error metric and N is the error with respect to area R.CS The number of points that match the reference point cloud RPC, RPC i and R CS,i is the reference point cloud RPC and region R CS Two matching points of i are the weights attributed to these points, T is the estimated spatial transformation to be applied to the 3D model IH, and ||x,y|| is the distance between two points x and y.

[0107] In the example shown, the spatial transformation T estimated by the Iterative Closest Point (ICP) algorithm is constrained to be a rigid transformation and is defined by a 4x4 matrix defining rotation and translation.

[0108] Despite the fact that the detection of the area AHC and the attribution of weights to the points strongly enhances the registration, sub-steps S11 and S12 may be skipped and the controller 14 may determine at step S13 the reference point cloud RPC and the area R CS As another alternative, only sub-step S12 may be skipped, and the controller 14 may calculate the spatial transformation from the reference point cloud RPC and the area R belonging to the area AHC at step S13. CS The spatial transformation is computed using a unique point in the space and a non-weighted error metric.

[0109] It may be noted that the present invention is not limited to the implementation of the iterative closest point algorithm, and other suitable point set registration algorithms (particularly point set registration algorithms in which the error metric may be weighted) may be executed by the controller 14 for implementing sub-step S13.

[0110] In step S2, the registered 3D model IH' is calculated by the controller 14 by applying the estimated spatial transformation T to the entire 3D model IH. The registered 3D model IH' is then displayed on the display device 15, simultaneously to the landmark LM located at the current position of the catheter 13 in the reference system X1, Y1, Z1.

[0111] refer to Figure 4 , which shows a simplified view of the display device 15 at the end of step S2. In the depicted embodiment, the landmark LM is a 3D model of the catheter 13, represented in its current configuration and its current position in the reference system X1, Y1, Z1.

[0112] Optionally, the value of the error measure E between the result of the calculated spatial transformation T on the 3D model IH and the reference point cloud RPC may be displayed on the device 15 as a value relative to the registration error.

[0113] Steps S1 and S2 correspond to an initial registration of the 3D model IH in the reference system X1, Y1, Z1. However, this initial registration may be affected by slight errors, which can be corrected in the real-time correction step S3 described below. Figure 5 , which shows the execution of this step.

[0114] During the intervention, in sub-step S31, a second catheter (not shown) has been introduced into the patient's heart. The second catheter may be a mapping catheter provided with one or more electrodes. These electrodes periodically provide local electrical signals captured on the endocardium and each associated with a spatial coordinate to the controller 14, from which the controller 14 may generate an electrophysiological map EM of the heart, such as a peak-to-peak voltage map. In sub-step S32, areas ALV of low voltage, slow conduction or abnormal activity may be segmented from the mapping map EM by a known method performed by the controller 14.

[0115] This region ALV can be compared to the region R that has been detected and segmented in the 3D model IH and corresponds to the myocardial channel responsible for the arrhythmia. CH Related.

[0116] Similar to step S13, in sub-step S33, the controller 14 may select the area ALV and the area R CH A point set registration algorithm is performed to correct the spatial transformation T in a corrected spatial transformation T' that, when applied to the 3D model IH, minimizes the spatial transformation in the region R CH The result of the spatial transformation T on T(R CH ) and the error measure E between the area ALV.

[0117] Furthermore, in step S34 , the registered 3D model IH″ is calculated again by the controller 14 by applying the estimated spatial transformation T′ to the entire 3D model IH. The registered 3D model IH″ is updated on the display device 15 .

[0118] For any periodic update of the electrophysiological map EM, steps S31 to S34 may be repeated periodically.

[0119] It may be noted that the present invention is not limited to the use of local electrical signals provided by a mapping catheter to correct the spatial transformation T, but may be extended to the use of other types of local data that may be associated with a position in the reference system X1, Y1, Z1, such as a local shape or a local wall thickness received directly or indirectly from an imaging device, or spatial coordinates received directly or indirectly from other catheters equipped with at least one positioning sensor.

[0120] In a further step S41 , each electrical local data received from the mapping catheter or from an electrode arranged as catheter 13 is displayed on the result IH″ of the corrected spatial transformation T′ on the 3D model IH displayed on the display device 15 .

[0121] like Figure 6 As shown, the figure shows a later simplified view of the display device 15, according to the color mapping that associates possible electrical values ​​with colors, each electrical local data can be represented on a vertex of the 3D model IH", the position of the vertex corresponds to the spatial coordinates of the electrical local data, and its color depends on the value of the electrical local data.

[0122] Optionally, in a further step S42, for each electrical local data, the controller 14 calculates the distance between its spatial coordinates and a set of vertices of the 3D model IH", which set of vertices defines a portion of the inner surface or outer surface of the myocardium closest to these spatial coordinates. The distance can be converted by the controller 14 into a value regarding the reliability of the electrical local data, and the value regarding their reliability is displayed on the display device 15. For example, the value regarding reliability can be obtained from a transfer function, which takes as input the distance calculated based on the spatial coordinates associated with the local data, and outputs a value between 0 and 1, where 0 indicates that the local data is not reliable at all, and where 1 indicates that the local data is completely reliable. Such a transfer function can be a step function, a rectangular function, a ramp function, a trigonometric function, a sigmoid function, a logical function, or a Gaussian function.

[0123] Steps S41 and S42 may be repeated for any new electrical data received from the catheter.

[0124] The methods disclosed herein may also be implemented by a software program executable by a computer system.In addition, specific implementations may include distributed processing and parallel processing, especially for parallel processing of several or all data in a data set.

[0125] The illustrations described herein are intended to provide a general understanding of the structures of various embodiments. These illustrations are not intended to be used as a complete description of all elements and features of the devices, processors and systems utilizing the structures or methods described herein. After reading this disclosure by combining the disclosed embodiments, many other embodiments or their combinations may be apparent to those of ordinary skill in the art. Other embodiments may be utilized and derived from this disclosure so that structural and logical substitutions and changes may be made without departing from the scope of this disclosure.

[0126] In addition, the disclosure and illustrations should be considered illustrative rather than restrictive, and the appended claims are intended to cover all such modifications, enhancements, and other embodiments that fall within the true spirit and scope of the specification. Therefore, the scope of the following claims will be determined by the broadest permissible interpretation of the claims and their equivalents, and should not be restricted or limited by the foregoing description.

Claims

1. A computer-implemented method for displaying a 3D model of an organ of a patient on a navigation system display device (15), the method comprising: a.(S01) receiving multiple spatial coordinates (S XYZ ), the plurality of spatial coordinates being calculated based on data received from at least one sensor (131) of a catheter (13) within a portion of the organ (CS) of the patient; the plurality of spatial coordinates defining a reference point cloud (RPC) representing the portion of the organ of the patient; b. (S05) receiving a plurality of points, the plurality of points defining a predefined 3D model of the organ of the patient (IH), marking a plurality of points to define a 3D model segmentation corresponding to the portion of the organ of the patient (R CS ); c. (S1) calculating a spatial transformation (T) which, when applied to the 3D model, minimizes the result of the spatial transformation on the 3D model segmentation (T(R CS )) and a predefined error metric (E) between the reference point cloud; d. (S2) displaying the landmark (LM) at the current position of the catheter obtained from the plurality of spatial coordinates on the navigation system display device, and simultaneously displaying the result (IH') of the calculated spatial transformation on the 3D model.

2. The method according to claim 1, comprising identifying the 3D model segmentation (R CS ) and / or in the reference point cloud (RPC) having at least one geometric feature matching a predetermined criterion (S11), and the calculated spatial transformation (T) is the result (T(R) of minimizing the spatial transformation on the 3D model segmentation when applied to the 3D model (IH) CS )) and the spatial transformation of the predefined error metric (E) between the reference point cloud in at least one or more regions.

3. The method according to claim 2, wherein the value is segmented with the 3D model (R CS ) and / or each point of the reference point cloud (RPC), said value depending on said position of said associated point with respect to said one or more areas (AHC), wherein said spatial transformation calculation step (S1) comprises an iterative estimation (S13) of a spatial transformation (T), said iterative estimation comprising, for each iteration: a. The result (T(R) of the spatial transformation of the current estimate on the 3D model segmentation CS )) with a point of the reference point cloud, and b. A sub-step of adjusting the currently estimated spatial transformation according to the predefined error metric (E) between the result of the currently estimated spatial transformation on the 3D model segmentation and the reference point cloud, wherein the predefined error metric is based on an average distance value between the matching points weighted by the association value.

4. The method according to claim 2 , wherein a plurality of spatial coordinates (S XYZ ) comprises receiving at least a first plurality of spatial coordinates and a second plurality of spatial coordinates calculated based on data received from at least a first sensor and a second sensor (131) of the catheter (13); and wherein the step (S11) of identifying one or more areas (AHC) in the reference point cloud (RPC) comprises estimating the evolution of the curvature and / or the torsion of the reference point cloud from the first plurality of spatial coordinates and the second plurality of spatial coordinates.

5. Method according to one of the preceding claims, wherein the calculated spatial transformation (T) is a square matrix defining at least a rotation and / or a translation.

6. A method according to one of the preceding claims, comprising a further step (S31) of receiving a set of supplementary local data about the organ of the patient from a sensing device different from the catheter, and a further step (S33) of correcting the calculated spatial transformation (T) based on the set of supplementary local data.

7. The method according to claim 1 , comprising periodically receiving spatial coordinates (S ) calculated from data received from the at least one sensor ( 131 ) of the catheter ( 13 ) within the portion (CS) of the organ of the patient. XYZ ) an additional step of modeling the periodic motion of the organ of the patient according to the periodically received spatial coordinates, and an additional step of modifying the displayed position of the landmark (LM) and the result (IH') of the calculated spatial transformation (T) on the 3D model (IH) on the navigation system display device (15) based on the model of the periodic motion of the organ of the patient.

8. Method according to one of the preceding claims, comprising a further step of calculating a value (E) of a registration error between the reference point cloud (RPC) and the result (IH') of the calculated spatial transformation (T) on the 3D model (IH).

9. A method according to one of the preceding claims, comprising a further step (S31) of receiving a set of supplementary local data about the organ of the patient calculated based on data received from at least one sensor of a catheter within a part of the organ of the patient, each supplementary local data being associated with the spatial coordinates of the sensor, and a further step (S41) of displaying each supplementary local data on the result (IH') of the calculated spatial transformation (T) on the 3D model (IH) displayed on the navigation system display device (15), the displayed position of the supplementary local data being calculated based on its associated spatial coordinates.

10. The method according to the preceding claim comprises a further step (S42) of calculating, for each supplementary local data, the distance between its associated spatial coordinates and one or more points of the result (IH') of the calculated spatial transformation (T) on the 3D model (IH), and a further step of calculating a value regarding the reliability of the supplementary local data based on the calculated distance.

11. A navigation system (1) for intervention assistance, the navigation system comprising: at least one catheter (13), the catheter comprising at least one positioning sensor (131); Display (15); and a calculation unit (14) designed to implement the method according to one of the preceding claims, the landmarks (LM) and the result (IH') of the calculated spatial transformation (T) on the 3D model (IH) being displayed on the display.

Images