METHOD AND APPARATUS FOR DETERMINING IMPLANT POSITIONS OF TWO MEDICAL IMPLANT COMPONENTS FORMING A JOINT

Abstract

A data processing method, performed by a computer, for determining implant positions of two implant components relative to two bones, wherein each of the implant components is to be attached to one of the bones such that the implant components form a joint between the bones, and wherein an implant position is a relative position between the implant component and the corresponding bone, said method comprising the steps of: a) acquiring a set of target poses, wherein a target pose represents a relative position to be achieved between the two bones; b) calculating a set of virtual poses for a pair of virtual test implant positions, wherein the set of virtual poses comprises one virtual pose for each of the target poses and wherein a virtual pose represents a relative position between the two bones if the virtual test implant positions were applied as the implant positions; c) calculating a pose deviation value for each of the target poses, wherein a pose deviation value represents the difference between a target pose and the corresponding virtual pose; d) calculating an overall pose deviation value from all the individual pose deviation values; e) repeating steps b) to d) for different pairs of virtual test implant positions until the overall pose deviation value fulfils a minimisation criterion; and f) using the pair of virtual test implant positions for which the minimisation criterion is fulfilled as the implant positions.

Claims

1.-15. (canceled)

16. A system for selecting an implant arrangement for a joint, the system comprising: a processor; and a non-transitory, computer-readable storage medium comprising instructions that, when executed, cause the processor to: access one or more target poses for the joint, each target pose comprising an arrangement of a first bone of the joint with respect to a second bone of the joint; assess one or more proposed implant arrangements, each proposed implant arrangement comprising an arrangement of a first implant component with respect to the first bone and an arrangement of a second implant component with respect to the second bone, wherein assessing each proposed implant arrangement comprises: determining, for each target pose, a corresponding pose comprising an arrangement of the first bone with respect to the second bone based on the proposed implant arrangement and mechanical properties of at least some soft tissue associated the joint, wherein the corresponding pose comprises two or more contact points between the first implant component and the second implant component, calculating, for each target pose, a deviation score based on the target pose and the corresponding pose, and calculating, based on the deviation scores for the one or more target poses, a composite deviation score; and select the implant arrangement from the one or more proposed implant arrangements based on the composite deviation score for each of the one or more proposed implant arrangements.

17. The system of claim 16, wherein the one or more target poses are based on at least one of recorded pre-operative poses of the first bone and the second bone of the joint, mirrored poses from a corresponding joint, and predefined poses representing a desired kinematics of the joint.

18. The system of claim 16, wherein selecting an implant arrangement from the one or more proposed implant arrangements comprises selecting one of the one or more proposed implant arrangements having a composite deviation score below a predetermined threshold value.

19. The system of claim 16, wherein selecting an implant arrangement from the one or more proposed implant arrangements comprises selecting the proposed implant arrangement having a minimum composite deviation score among the one or more proposed implant arrangements.

20. The system of claim 16, wherein determining a corresponding pose comprises determining an arrangement of the first bone with respect to the second bone wherein the first implant component is in stable contact with the second implant component.

21. The system of claim 16, wherein determining a corresponding pose comprises adjusting one or more of a varus-valgus rotational position and a proximodistal position of the target pose.

22. The system of claim 16, wherein the deviation score is calculated based on a deviation of one or more of a varus-valgus rotational position and a proximodistal position between the corresponding pose and the target pose.

23. The system of claim 16, wherein the deviation score is based on a distance between one or more pairs of points, wherein each pair of points comprises a point on the first bone and a corresponding point on the second bone.

24. The system of claim 23, wherein the one or more pairs of points comprise collateral ligament attachment points.

25. A computer-implemented method of selecting an implant arrangement for a joint, the method comprising: accessing one or more target poses for the joint, each target pose comprising an arrangement of a first bone of the joint with respect to a second bone of the joint; assessing one or more proposed implant arrangements, each proposed implant arrangement comprising an arrangement of a first implant component with respect to the first bone and an arrangement of a second implant component with respect to the second bone, wherein assessing each proposed implant arrangement comprises: determining, for each target pose, a corresponding pose comprising an arrangement of the first bone with respect to the second bone based on the proposed implant arrangement and mechanical properties of at least some soft tissue associated with the joint, calculating, for each target pose, a deviation score based on the target pose and the corresponding pose, and calculating, based on the deviation scores for the one or more target poses, an composite deviation score; and selecting the implant arrangement from the one or more proposed implant arrangements based on the composite deviation score for each of the one or more proposed implant arrangements.

26. The system of claim 25, wherein acquiring one or more target poses comprises at least one of recording pre-operative poses of the first bone and the second bone of the joint, mirroring a corresponding joint, and obtaining predefined poses representing a desired kinematics of the joint.

27. The system of claim 25, wherein selecting an implant arrangement from the one or more proposed implant arrangements comprises selecting one of the one or more proposed implant arrangements having an overall pose deviation value below a predetermined threshold value.

28. The system of claim 25, wherein selecting an implant arrangement from the one or more proposed implant arrangements comprises selecting the proposed implant arrangement having a lowest composite deviation score among the one or more proposed implant arrangements.

29. The system of claim 25, wherein determining a corresponding pose comprises determining an arrangement of the first bone with respect to the second bone wherein the first implant component is in stable contact with the second implant component.

30. The system of claim 25, wherein determining a corresponding pose comprises adjusting one or more of a varus-valgus rotational position and a proximodistal position of the target pose.

31. The system of claim 25, wherein the deviation score is calculated based on a deviation of one or more of a varus-valgus rotational position and a proximodistal position between the real pose and the target pose.

32. The system of claim 25, wherein the deviation score is based on a distance between one or more pairs of points, wherein each pair of points comprises a point on the first bone and a corresponding point on the second bone.

33. The system of claim 32, wherein the one or more pairs of points comprise collateral ligament attachment points.

34. A system for assessing an implant arrangement for a joint, the system comprising: a processor; and a non-transitory, computer-readable storage medium comprising instructions that, when executed, cause the processor to: receiving a target pose for the joint, the target pose comprising a relative arrangement of a first bone and a second bone of the joint; receiving the implant arrangement, the implant arrangement comprising a relative arrangement of a first implant component and the first bone and a relative arrangement of a second implant component and the second bone; determining, for the target pose, a corresponding pose comprising a relative arrangement of the first bone and the second bone based on the implant arrangement and mechanical properties of at least some soft tissue associated with the joint, wherein the corresponding pose comprises two or more contact points between the first implant component and the second implant component, and calculating a deviation score for the implant arrangement based on a deviation between the target pose and the corresponding pose.

35. The system of claim 34, wherein determining a corresponding pose comprises determining an arrangement of the first bone with respect to the second bone wherein the first implant component is in stable contact with the second implant component.

36. The system of claim 34, wherein determining a corresponding pose comprises adjusting one or more of a varus-valgus rotational position and a proximodistal position of the target pose.

37. The system of claim 34, wherein the deviation score is calculated based on a deviation of one or more of a varus-valgus rotational position and a proximodistal position between the corresponding pose and the target pose.

38. The system of claim 34, wherein the deviation score is based on a distance between one or more pairs of points, wherein each pair of points comprises a point on the first bone and a corresponding point on the second bone.

39. The system of claim 38, wherein the one or more pairs of points comprise collateral ligament attachment points.

Description

[0050] In the following, example embodiments of the invention are described with reference to the figures which illustrate the invention merely by way of example and do not limit the scope of the invention to the specific embodiments illustrated, and which show:

[0051] FIG. 1 a plurality of poses for different implant positions;

[0052] FIG. 2 a block diagram illustrating a process for determining the implant positions;

[0053] FIGS. 3a, 3b a single pose of the joint, viewed from two directions perpendicular to one another;

[0054] FIGS. 4a to 4c the transformation of a bone pose into an implant pose;

[0055] FIG. 5 a virtual pose;

[0056] FIG. 6 steps for determining a virtual bone pose for a target bone pose;

[0057] FIG. 7 a first example of calculating a pose deviation value;

[0058] FIG. 8 a second example of determining a pose deviation value;

[0059] FIG. 9 a computer for carrying out the method according to the invention;

[0060] FIGS. 10a, 10b graphs showing deviations between target poses and virtual poses determined using the classic planning, leg alignment and the present invention, respectively, over the range of motion; and

[0061] FIG. 11 comparative results for the implant position results for the leg alignment approach and the approach in accordance with the present invention.

[0062] The left-hand illustration in FIG. 1 shows a plurality of bone poses, wherein a bone pose is a relative position between two bones. FIG. 1 illustrates the overall working principle of the invention. In the present example, the two bones are a femur 1 and a tibia 2 and the bone poses correspond to different flexion angles of the knee joint which connects the femur and the tibia. The femur 1 comprises a femoral implant component 3, and the tibia 2 comprises a tibial implant component 4. The implant components form an artificial knee joint. The implant positions shown in the left-hand illustration in FIG. 1 represent a classic approach. The bone poses represent example target bone poses, which are for example poses which the knee joint assumed prior to the introduction of an artificial knee joint.

[0063] As can be seen from FIG. 1, there are some poses in which there would be a gap between the femoral implant 3 and the tibial implant 4, and some poses in which the implants would overlap each other, once the implant components had been attached. In the latter case, the bones would move away from each other since such an overlap is not possible in reality. This would result in a loose knee joint in the poses exhibiting a gap, and a significant stress on the ligaments in the poses exhibiting overlapping implants. The gaps and overlaps between the implant components 3 and 4 in FIG. 1, and in particular in the left-hand illustration in FIG. 1, are exaggerated in order to accentuate the effect of the present invention.

[0064] The right-hand illustration in FIG. 1 represents the femur 1 and the tibia 2, together with the attached femoral implant component 3 and the attached tibial implant component 4, for the same poses as in the left-hand illustration, but with optimised implant positions. It can be seen that there are only small gaps or overlaps between the implant components, which means that the kinematics of the original knee joint are matched quite well.

[0065] FIG. 2 shows a block diagram which illustrates an approach for determining the implant positions. The diagram includes an adding block 5, an optimisation block 6 and a prediction block 7. A set of target poses is fed into the adding block 5 and the prediction block 7. The prediction block 7 calculates a set of virtual poses for a pair of virtual test implant positions. The set of virtual poses comprises one virtual pose for each of the target poses. A virtual pose is a hypothetical pose for a particular pair of implant positions. A virtual pose is a pose in which the implant components 3 and 4 are in stable contact, which means that they are in contact at at least two points. In other words, a virtual pose is a pose into which the joint, and thus the bones, are forced due to the mechanical interaction of the implant components when they are placed at the assumed pair of virtual test implant positions.

[0066] In a first iteration, the prediction block 7 uses an initial pair of virtual test implant positions which can comprise arbitrary virtual test implant positions or pre-calculated virtual test implant positions. A set of virtual poses is then provided to the adding block 5, where each virtual pose is subtracted from its corresponding target pose, thus resulting in a set of difference poses between the target poses and the virtual poses. This set of difference poses is provided to the optimisation block 6 which calculates a pose deviation value for each of the target poses, wherein a pose deviation value represents a single numeric value corresponding to the difference between a target pose and the corresponding virtual pose.

[0067] The optimisation block 6 calculates an overall pose deviation value from the plurality of individual pose deviation values. If this overall pose deviation value fulfils a predetermined minimisation criterion, then the set of virtual test implant positions used by the prediction block 7 is outputted as the implant positions. If the overall pose deviation value does not fulfil the minimisation criterion, the optimisation block 6 determines a new pair of virtual test implant positions which is then provided to the prediction block 7 in order to calculate a new set of virtual poses.

[0068] The optimisation block 6 can select the new pair of virtual test implant positions by sequentially selecting pairs of virtual test implant positions from a list of pairs of virtual test implant positions. Alternatively, however, the optimisation block 6 can also implement an optimisation algorithm which determines the new pair of virtual test implant positions from at least one of the overall pose deviation values corresponding to the previously analysed pairs of virtual test implant positions.

[0069] FIGS. 3a and 3b show the femur 1 and the tibia 2 in a particular pose in a frontal and a lateral view, respectively. As can be seen from FIGS. 3a and 3b, a femoral co-ordinate system 8 is assigned to the femur 1, and a tibial co-ordinate system 9 is assigned to the tibia 2. The co-ordinate systems 8 and 9, which are also referred to as bone co-ordinate systems, have an invariable location and orientation relative to the respective bones which they are assigned to. In this example embodiment, the orientations of the bone co-ordinate systems 8 and 9 is such that their y-axis extends in a proximodistal direction, their x-axis extends in a mediolateral direction, and their z-axis extends in an anterioposterior direction when the leg is in its neutral position. FIGS. 3a and 3b also show the mechanical axis 10 of the femur 1 and the mechanical axis 11 of the tibia 2.

[0070] As can be seen from FIGS. 3a and 3b, the bone pose—i.e. the relative position between the femur 1 and the tibia 2—can be expressed as a transformation B which transforms the femoral co-ordinate system 8 into the tibial co-ordinate system 9 or vice versa. The transformation B can also be referred to as the bone pose B and is preferably one of the target poses from the set of target poses.

[0071] FIG. 4a shows the bones 1 and 2 from FIG. 3a, with the femoral implant component 3 and the tibial implant component 4 attached, for an assumed pair of virtual test implant positions. For the target bone pose B shown in FIG. 4a, the implant components 3 and 4 would overlap each other on the left side and provide a gap on the right side, thus resulting in an technically impossible and physiologically undesired pose if the implant components 3 and 4 were attached using the assumed pair of virtual test implant positions. The relative position between the femur 1 and the tibia 2 therefore has to be amended such that the femoral implant 3 and the tibial implant 4 are in stable contact at at least two contact points. This results in a virtual bone pose which corresponds to one particular target bone pose and depends on the assumed pair of virtual test implant positions.

[0072] As can be seen from FIGS. 4a and 4b, a femoral implant co-ordinate system 12 is assigned to the femoral implant 3, and a tibial implant co-ordinate system 13 is assigned to the tibial implant 4. The implant components 3 and 4 typically have a planar region which is to be placed onto a corresponding cutting plane of the bone to which it is to be attached. The z-axis of the implant co-ordinate systems is preferably perpendicular to this planar region, while the x-axis extends in the mediolateral direction of the implant component and the y-axis extends perpendicular to both the z-axis and the x-axis. The origin of the implant co-ordinate systems is preferably in the middle of the respective planar region.

[0073] The relative position between an implant component and the corresponding bone can be described by a transformation between the implant co-ordinate system and a corresponding bone co-ordinate system. FIG. 4b shows a transformation F between the femoral implant co-ordinate system 12 and the femoral bone co-ordinate system 8 and a transformation T between the tibial implant co-ordinate system 13 and the tibial bone co-ordinate system 9. Since the transformations F and T are constant for a particular pair of virtual test implant positions, the bone pose B can be converted into a joint pose J, shown in FIG. 4c, via the transformations F and T using the equation

J=F*B*T.sup.−1,

[0074] where J, F, B and T are defined as 4×4 matrices.

[0075] The implant pose or transformation J describes the relative position between the two implant co-ordinate systems 12 and 13 for a particular transformation B between the bone co-ordinate systems 8 and 9 and a particular pair of virtual test implant positions. As already explained above, a transformation between two co-ordinate systems comprises six parameters, namely three translational shifts and three rotations. These six parameters are appropriately encoded into the 4×4 matrices J, B, F and T.

[0076] As explained above, attaching the implant components 3 and 4 to the bones 1 and 2, respectively, using the assumed pair of virtual test implant positions could result in an undesirable relative position between the bones 1 and 2, in which the implant components 3 and 4 overlap each other or are not in stable contact. The prediction block 7 (FIG. 2) therefore determines a virtual pose which corresponds to a target pose in which the implant components 3 and 4 are in stable contact with each other. This results in a virtual joint pose J′ as shown in FIG. 5. Since the transformations F and T are constant, the virtual bone pose B′ can be calculated from J′, F and T.

[0077] In this embodiment, the approach described in patent application PCT/EP2012/061757 is applied in order to determine a virtual joint pose J′ which is then transformed into the virtual bone pose B′. The parameters for the ap shift, the ml shift, the ie rotation and the fe rotation are kept constant and the parameters for the pd shift and the vv rotation are determined such that the implant components 3 and 4 are in stable contact. Details will be described below with reference to FIG. 7.

[0078] The optimisation block 6 then calculates a pose deviation value for each target pose on the basis of the difference between the target pose and the corresponding virtual pose. It can be calculated on the basis of a target bone pose and a virtual bone pose or on the basis of a target joint pose and a virtual joint pose. The difference between a target joint pose J and a virtual joint pose J′ is however easier for a physician to analyse if it is presented as deviations in the ml, ap and pd shifts and the ie, fe and vv rotations.

[0079] FIG. 6 summarises how a virtual bone pose is calculated for a target bone pose. In step S1.1, a target bone pose is selected. In step S1.2, a pair of virtual test implant positions is selected. In step S1.3, a target joint pose is calculated from the target bone pose and the pair of virtual test implant positions. In step S1.4, a virtual joint pose in which the implant components are in stable contact is calculated. In step S1.5, a virtual bone pose corresponding to the target bone pose is calculated from the virtual joint pose and the pair of virtual test implant positions.

[0080] FIG. 7 shows a femur 1 with a femoral implant 3 and a tibia 2 with a tibial implant 4 in a target pose and a virtual pose, in order to explain an approach for calculating a virtual pose which corresponds to a target pose. The left-hand illustration in FIG. 7 shows the bones with the implant components for a particular pair of virtual test implant positions in the target pose which exhibits the parameter values vv.sub.J and pd.sub.J. The virtual test implant positions cause the implants 3 and 4 to overlap each other. In this example approach, the parameters ml, ap, ie and fe are kept constant, and the pd parameter of the target joint pose J is changed such that there is a gap between the femoral implant 3 and the tibial implant 4.

[0081] The bones 1 and 2 are then brought together in the pd direction until the femoral implant 3 and the tibial implant 4 are in contact with each other. If there is no stable contact, i.e. only contact at a single point, then the approach is continued until there is contact between the femoral implant 3 and the tibial implant 4 at a second contact point. The distance between the first and second contact points in the pd direction is then used to calculate a vv parameter and a pd parameter such that the femoral implant 3 is in stable contact with the tibial implant 4 at two contact points. This results in a virtual joint pose J′ which exhibits amended parameter values vv.sub.J′ and pd.sub.J′ and in which the other parameter values are identical to those of the target joint pose. A detailed description of this approach is given in the patent application PCT/EP2012/061757.

[0082] Since the target joint pose J and the virtual joint pose J′ differ only in their vv and pd parameter values, the pose deviation value can be calculated from the difference in these parameter values alone. In particular, the pose deviation value can be calculated using the equation

poseDeviationValue=[(pd.sub.J−pd.sub.J′).sup.2+(vv.sub.J−vv.sub.J′).sup.2].sup.1/2.

[0083] FIG. 8 shows the femur 1 together with the femoral implant 3 and the tibia 2 together with the tibial implant 4 in a target joint pose (in the left-hand illustration) and in a virtual joint pose (in the right-hand illustration), in order to explain another approach for calculating the pose deviation value.

[0084] FIG. 8 shows a working plane 14 which is parallel to the plane spanned by the x-axis and the y-axis of the tibial co-ordinate system 9. In this example, the origin of the tibial co-ordinate system 9 lies in the working plane 14. However, the working plane 14 can have any other location as long as it is kept constant with respect to the tibia 2.

[0085] Two epicondylar points epiM and epiL are shown on the tibia 2. The distances between each of the epicondylar points epiM and epiL and the working plane 14 for the target pose and the virtual pose are used to calculate the pose deviation value. d.sub.med and d.sub.lat are the distances for the target pose, and d.sub.med′ and d.sub.lat′ are the distances for the virtual pose. The pose deviation value can then be calculated using the equation

poseDeviationValue=[(d.sub.med−d.sub.med′).sup.2+(d.sub.lat−d.sub.lat′).sup.2].sub.1/2.

[0086] FIG. 9 schematically shows a system 15 for carrying out the method as described above. The system comprises a computer 16 which is connected to an input unit 20, such as a keyboard, and to a display unit 21, such as a monitor. The computer 16 comprises a processing unit 17 which is connected to a memory unit 18 and an interface 19. The CPU can acquire data, such as the target poses, via the interface 19. The memory unit 18 stores a program code to be executed by the processing unit 17 and optionally also stores any data received via the interface 19. The processing unit 17 is adapted to execute the code stored in the memory unit 18, such that the processing unit 17 carries out the method steps as described above. The processing unit 17 is in particular then adapted to acquire the set of target poses, calculate a set of corresponding virtual poses for a pair of virtual test implant positions, calculate a pose deviation value for each target pose and calculate an overall pose deviation value from all the individual pose deviation values. The processing unit 17 is also adapted to repeat these steps for different pairs of virtual test implant positions in order to find a pair of virtual test implant positions for which the overall pose deviation value fulfils a predetermined minimisation criterion.

[0087] Information can be inputted into the computer 16 via the input unit 20, for example in order to limit the ranges of values for the pairs of virtual test implant positions. The display unit 21 is adapted to display the results determined by the processing unit 17, such as the calculated implant positions.

[0088] FIGS. 10a and 10b comprise graphs showing deviations between target poses and virtual poses determined using the classic planning (CPL), leg alignment (LA) and the present invention (6dof), respectively, over the range of motion, i.e. a range of flexion angles of the knee joint. The graph of FIG. 10a shows the deviations with regards to the pd shift (pdDev) and the graph of FIG. 10b shows the deviations with regards to the vv rotation (vvDev). More precisely, pdDev and vvDev are calculated as pdDev=pd.sub.J−pd.sub.J′ and vvDev=vv.sub.J−vv.sub.J′, respectively, as explained with reference to FIG. 7. Results of the classic planning approach are plotted in a dotted line, results of the leg alignment approach are plotted in a dashed line and results of the approach according to the present invention are plotted in a continuous line.

[0089] The target poses represent measured poses of the knee joint. The virtual poses represent the poses which would result if the implants would be implanted according to the implant positions determined according to the classic planning approach, the leg alignment approach and the approach according to the present invention, respectively. They can be compared to the target poses of the knee joint over the knee joint's range of motion. The range of motion applied in FIGS. 10a and 10b yields from a variation of the flexion angle from 0 degrees to 92 degrees. The flexion angles are for example derived from the target poses.

[0090] The graphs show that no approach results in implant positions which cause the target poses and the virtual poses to be identical over the whole range of motion. The reason is that the shapes of the implants are not identical to the shapes of the original bones, such that it is impossible to achieve the original kinematics of the joint over the whole range of motion, no matter how the implants are positioned. For some flexion angles, the present invention even yield worse virtual poses, i.e. larger deviations, than the leg alignment approach or the classic planning approach. But overall, i.e. over the whole range of motion, the average deviations are lower than those for implant positions determined using the classic planning approach or the leg alignment approach.

[0091] FIG. 11 shows a table comprising numerical values for the twelve parameters of the implant positions in the example scenario which lead to the deviations shown in FIG. 10. The values in the second column have been determined using the leg alignment approach. Initial values have been determined using classic planning, and the values for pd.sub.t, ap.sub.f and ie.sub.f have then been optimised. The values in the third column have been determined using the approach according to the invention for six degrees of freedom. This means that the values for the six parameters pd.sub.t, ap.sub.f, ie.sub.f, pd.sub.f, ie.sub.t and fe.sub.t have been optimised as compared to the leg alignment approach, wherein the ap.sub.f parameter has been constrained to a small posterior range only. The fourth column shows the differences between the values obtained by the leg alignment approach and the approach according to the invention. No values are shown in this column for parameters which have not been optimised.