Surgical navigation module and surgical navigation system

Abstract

Accurately position orthopedic surgical instruments and orthopedic implants during surgical procedures. The surgical navigation module includes a microcomputer, an inertial sensing unit, an ultra-wideband radio unit, and a housing containing the microcomputer, the inertial sensing module, and the ultra-wideband radio module.

Description

[Technical Field]

[0001] This application claims the benefit of U.S. Provisional Patent Application No. 62 / 022,899, filed July 10, 2014, entitled "CRANIUM AND POSTCRANIAL BONE AND SOFT TISSUE RECONSTRUCTION," the disclosure of each of which is incorporated herein by reference.

[0002] The present disclosure is directed to various aspects of orthopaedics, such as bone reconstruction and tissue remodeling, patient-specific and mass-customized orthopaedic implants, gender- and ethnic-specific orthopaedic implants, cutting guides, trauma plates, bone graft cutting and placement guides, patient-specific instruments, use of inertial measurement units for anatomical tracking for kinematics and pathology, and use of inertial measurement units for navigation during orthopaedic surgery. [Background technology]

[0003] The present disclosure includes a surgical navigation system that provides translational and rotational navigation using a self-referencing hybrid navigation system based on ultra-wideband (UWB) and inertial technologies, in contrast to current surgical navigation systems that use optical, electromagnetic (EM), or inertial navigation systems (e.g., U.S. Patent Application Publication No. 2013 / 0217998).

[0004] The first challenge of optical surgical navigation systems is the need for line of sight (LOS) between the camera and the tracking module, which is often obstructed by the surgeon or surgical technician during surgery. Also, errors in patient alignment relative to the camera can introduce significant errors into the system.

[0005] Current electromagnetic (EM) navigation systems use EM generators and EM probes to track movement. Although EM navigation does not require line of sight, the presence of metallic objects in the vicinity of the probe reduces the accuracy of the system. This is a common problem in many surgeries that utilize multiple metallic instrument components, such as metal retractors.

[0006] Finally, current inertial surgical navigation systems use a series of inertial sensors (accelerometers and gyroscopes) to form an active navigation unit. Inertial navigation systems are inaccurate for mobile navigation without external observational inputs, such as a global positioning system (GPS) or optical navigation systems, to correct for mathematical drift. Inertial navigation is also limited to orientation navigation and can be inaccurate due to distortion of ferromagnetic or martensitic materials or permanent magnets.

[0007] Currently, ultra-wideband (UWB) localization is commonly used for tracking resources and people. A large number of anchors or base stations are established within a facility, and each tracked resource is equipped with a UWB tag. The first localization method in many current UWB systems uses time of arrival (TOA), which is ideal for tracking large areas. However, the tracking accuracy of TOA systems is in the meter range, making it unsuitable for high-precision surgical applications. The second UWB localization method uses time difference of arrival (TDOA), which may be suitable for high-precision tracking applications. However, TDOA systems are much more difficult to implement than TOA systems. In particular, the accuracy of TDOA systems depends on the presence or absence of coherent clock synchronization and clock jitter and drift mitigation. However, to date, no reliable TDOA system exists for surgical navigation. In particular, current UWB systems do not optimize antennas for the intended application and do not capture the orientation dependence of antenna polarization. High-precision medical applications require identifying and mitigating the phase center error of a series of optimized antennas. Additionally, most UWB location systems do not capture the harsh indoor environments with potentially extensive multipath and non-visibility conditions that are both common in surgical environments (e.g., operating rooms and special operating theatres).Furthermore, the deployment of current UWB location methods is limited by strict anchor configuration and placement, tedious calibration procedures, and errors due to incoherent clocks between anchors.

[0008] Combining UWB systems with inertial measurement unit (IMU) systems overcomes many, if not all, of the above issues, provided that specific issues are addressed. For UWB systems, calibration and placement of multiple anchors remains the first hurdle in achieving high accuracy. Incoherent clock synchronization between anchors introduces uncertainty into localization results. Current UWB systems are limited to update rates set by the manufacturer, making real-time surgical navigation impossible. For IMU systems, heading drift remains the first concern because they do not use magnetometers for heading navigation correction. Previous attempts to combine UWB and IMU systems have failed, treating the systems as completely separate entities that do not interact with each other or that use small amounts of movement data from the UWB system to correct movement estimates from the IMU system. A key reason for the lack of complementary orientation tracking between the two systems is the inability of a single UWB tag to generate orientation information. To be precise, at least three UWB tags are placed on the same rigid body to generate orientation information. Summary of the Invention [Means for solving the problem]

[0009] Current hybrid UWB-IMU tracking systems, disclosed in more detail below, address these deficiencies and enable precise tracking and surgical navigation.

[0010] A first aspect of the present invention provides a surgical navigation system comprising a signal receiver communicatively coupled to a main processor, the main processor being programmed to calculate changes in three-dimensional position of an inertial measurement unit attached to a surgical instrument using a sequential Monte Carlo algorithm, the processor being communicatively coupled to a first memory storing unique instrument data for each of a plurality of surgical instruments and a second memory storing model data sufficient to construct a three-dimensional model of an anatomical feature, the main processor being communicatively coupled to a display providing visual feedback regarding the three-dimensional position of the surgical instrument relative to the anatomical feature.

[0011] In one more detailed embodiment of the first aspect, the surgical navigation system further comprises a reference inertial measurement unit configured to be attached to an anatomical feature, the reference inertial measurement unit being communicatively coupled to a first embedded processor and a first wireless transmitter and transmitting data to the primary processor, the reference inertial measurement unit being configured to be attached to the anatomical feature, the first embedded processor directing the transmission of data from the reference inertial measurement unit to the first wireless transmitter; the inertial measurement unit attached to the surgical instrument comprises an operational inertial measurement unit communicatively coupled to a second embedded processor and a second wireless transmitter, the second embedded processor being configured to be attached to one of the plurality of surgical instruments, the primary processor being communicatively coupled to a primary receiver configured to receive data from the first wireless transmitter and data from the second wireless transmitter. In yet another more detailed embodiment, the second embedded processor directs the transmission via the second wireless transmitter of the identity of the surgical instrument to which the operational inertial measurement unit is attached. In a more detailed embodiment, the inertial measurement unit includes at least three accelerometers and three magnetometers, the at least three accelerometers each outputting a total of nine or more accelerometer data streams for data about three axes, the at least three magnetometers each outputting a total of nine or more magnetometer data streams for data about three axes, and the main processor utilizes the nine accelerometer data streams and the nine magnetometer data streams to calculate a change in three-dimensional position of the inertial measurement unit attached to the surgical instrument. In yet another detailed embodiment, the model data stored in the second memory includes a three-dimensional virtual model of an anatomical feature, the instrument data stored in the first memory includes three-dimensional virtual models of a plurality of surgical instruments, the display displays the three-dimensional virtual models of the anatomical feature, the display displays the three-dimensional virtual models of the surgical instruments, the main processor is operative to reposition the three-dimensional virtual models of the anatomical feature using data from the reference inertial measurement unit, and the main processor is operative to reposition the three-dimensional virtual models of the surgical instruments using data from the operational inertial measurement unit.In a more detailed embodiment, the main processor operates to utilize data from the inertial measurement unit to reposition the three-dimensional virtual model of the surgical instrument relative to the three-dimensional virtual model of the anatomical feature in real time. In a more detailed embodiment, the sequential Monte Carlo algorithm includes a von Mises-Fisher density algorithm component. In another more detailed embodiment, the instrument data stored in the first memory includes position data indicating a relative distance between an end effector of the surgical instrument and a mounting point of the inertial measurement unit on the surgical instrument, the surgical instrument including at least one of a reamer, a cup positioner, an impactor, a drill, a saw, and a cutting guide. In yet another more detailed embodiment, the inertial measurement unit includes at least three magnetometers, and the display is coupled to at least one of the surgical instrument and the main processor.

[0012] A second aspect of the present invention provides a calibration system for an inertial measurement unit having a magnetometer and an accelerometer, the calibration system comprising: (a) a main platform rotationally repositionable relative to an intermediate platform along a first axis; (b) a final platform rotationally repositionable relative to the intermediate platform along a second axis, the second axis being perpendicular to the first axis, the final platform having a holder configured to be attached to the inertial measurement unit; and (c) a processor and associated software configured to communicatively couple to the inertial measurement unit, the software operative to utilize data output from a magnetometer associated with the inertial measurement unit and record a data set resembling an ellipsoid while the main platform rotates relative to the intermediate platform and while the final platform rotates relative to the intermediate platform, the software operative to fit a sphere to the data set and generate magnetometer correction calculations to account for distortions in the local magnetic field, thereby normalizing future data output from the magnetometer.

[0013] In a more detailed embodiment of the second aspect, the main platform is stationary. In yet another more detailed embodiment, the main platform at least partially houses a motor configured to rotate the intermediate platform relative to the main platform. In a still more detailed embodiment, the software operates to utilize a first data set output from an accelerometer associated with the inertial measurement unit while the inertial measurement unit is in a first stationary position, and to utilize a second data set output from the accelerometer at a second stationary position different from the first stationary position to generate an accelerometer correction calculation that normalizes future data output from the accelerometer. In yet another detailed embodiment, the first stationary position corresponds to the main platform being in a first fixed position relative to the intermediate platform and the final platform being in a second fixed position relative to the intermediate platform, and the second stationary position corresponds to at least one of the main platform being in a third fixed position relative to the intermediate platform and the final platform being in a fourth fixed position relative to the intermediate platform. In a more detailed embodiment, the final platform comprises a plurality of holders, each of the plurality of holders configured to mount to at least one of the plurality of inertial measurement units.

[0014] A third aspect of the present invention provides a method for calibrating an inertial measurement unit having a magnetometer, the method comprising: (a) rotating a first inertial measurement unit having a first inertial measurement unit about a first axis of rotation and a second axis of rotation, while simultaneously receiving raw local magnetic field data from the first magnetometer, the first axis of rotation being perpendicular to the second axis of rotation; (b) applying a uniformity calculation to the raw local magnetic field data to calculate distortions of the local magnetic field; and (c) normalizing the raw local magnetic field data received from the magnetometer by taking the calculated distortions of the local magnetic field to provide accurate local magnetic field data.

[0015] In a more detailed embodiment of the third aspect, the first inertial measurement unit includes a first accelerometer, and the method further includes the steps of (i) holding the first inertial measurement unit stationary at a first three-dimensional position while simultaneously receiving raw accelerometer data from the first accelerometer; (ii) holding the first inertial measurement unit stationary at a second three-dimensional position while simultaneously receiving raw accelerometer data from the first accelerometer, the second three-dimensional position being different from the first three-dimensional position; and (iii) normalizing the data received from the first accelerometer to reflect zero acceleration when the first accelerometer is stationary. In yet another more detailed embodiment, the first inertial measurement unit includes a second accelerometer, and the method further includes (i) holding the second inertial measurement unit stationary at a third three-dimensional position while simultaneously receiving raw accelerometer data from the second accelerometer when the first accelerometer is held stationary, (ii) holding the second inertial measurement unit stationary at a fourth three-dimensional position while simultaneously receiving raw accelerometer data from the second accelerometer when the first accelerometer is held stationary, the fourth three-dimensional position being different from the third three-dimensional position, and (iii) normalizing the data received from the second accelerometer to reflect zero acceleration when the second accelerometer is stationary. In a more detailed embodiment, the raw local magnetic field data is representative of a three-dimensional ellipsoid and the precise local magnetic field data is representative of a three-dimensional sphere. In yet another detailed embodiment, the uniform calculation includes fitting a sphere to the raw local magnetic field data, and normalizing the raw local magnetic field data includes subtracting the calculated distortion from the raw local magnetic field data to provide accurate local magnetic field data. In a more detailed embodiment, the method further includes a second inertial measurement unit having its own first accelerometer. In a more detailed embodiment, the second inertial measurement unit has its own first accelerometer.

[0016] A fourth aspect of the present invention provides a method for identifying a surgical instrument when coupled to an inertial measurement unit, the method comprising: (a) attaching the inertial measurement unit to one of a plurality of surgical instruments, each of the plurality of surgical instruments having a unique interface; and (b) identifying one of the plurality of surgical instruments in response to reading the unique interface by reading the unique interface and sending a signal to a processor communicatively coupled to the inertial measurement unit.

[0017] In a more detailed embodiment of the fourth aspect, the inertial measurement unit is operably coupled to a plurality of switches, the unique interface engages with at least one of the plurality of switches, and reading the unique interface includes the processor determining which of the plurality of switches the unique interface has engaged. In yet another more detailed embodiment, the processor is coupled to the inertial measurement unit, and the processor and the inertial measurement unit are housed in a common housing. In an even more detailed embodiment, the processor is remote from the inertial measurement unit, and the processor and the inertial measurement unit are not housed in a common housing.

[0018] A fifth aspect of the present invention provides a method for surgical navigation, comprising: (a) generating acceleration data and magnetic data using a plurality of inertial measurement units; (b) calibrating the plurality of inertial measurement units proximate to a surgical site; (c) aligning the relative positions of first and second inertial measurement units comprising the plurality of inertial measurement units, the relative positions including attaching the first inertial measurement unit to an alignment device that uniquely engages with the patient's anatomy at a specific position and orientation, and attaching the second inertial measurement unit to the patient; (d) attaching the first inertial measurement unit to the aligned surgical instrument; (e) repositioning the surgical instrument and the first inertial measurement unit toward a surgical site associated with the patient's anatomy; and (f) providing visual feedback regarding at least one of the position and orientation of the surgical instrument when the patient's anatomy is not visible or when the working end of the surgical instrument is not visible.

[0019] A sixth aspect of the present invention provides a method for surgical navigation, comprising: (a) generating acceleration data and magnetic data using a plurality of inertial measurement units; (b) calibrating the plurality of inertial measurement units proximate to a surgical site; (c) aligning the relative positions of first and second inertial measurement units comprising the plurality of inertial measurement units, the relative positions comprising attaching the first inertial measurement unit to an alignment device that uniquely engages with the patient's anatomy at a specific position and orientation, and attaching the second inertial measurement unit to the patient; (d) attaching the first inertial measurement unit to the aligned surgical instrument; (e) repositioning the surgical instrument and the first inertial measurement unit toward the surgical site associated with the patient's anatomy; and (f) providing visual feedback regarding the position and orientation of the surgical instrument relative to a predetermined surgical plan, the predetermined surgical plan identifying at least one of a range of allowable positions and a range of allowable orientations that the surgical instrument may occupy.

[0020] A seventh aspect of the present invention provides a method for generating a trauma plate for a particular bone, the method comprising: (a) accessing a database including a plurality of three-dimensional bone models of the particular bone; (b) evaluating features including at least one of a longitudinal contour and a cross-sectional contour for each of the plurality of three-dimensional bone models, wherein the longitudinal contour is along a dominant dimension of the plurality of three-dimensional bone models; (c) clustering the plurality of three-dimensional bone models based on the evaluated features to generate a plurality of clusters, wherein the plurality of clusters are numerically less than 10% of the plurality of three-dimensional bone models; and (d) generating a trauma plate for each of the plurality of clusters.

[0021] In a more detailed embodiment of the seventh aspect, generating a trauma plate for each of the plurality of clusters includes avoiding soft tissue attachment to particular bones through selection of fixation points. In yet another more detailed embodiment, the plurality of three-dimensional bone models include at least one common point, the common point including at least one of gender, ethnicity, age range, and height range. In an even more detailed embodiment, generating a trauma plate for each of the plurality of clusters includes incorporating at least one of an average longitudinal profile and an average cross-sectional profile of the particular cluster.

[0022] An eighth aspect of the present invention provides a method for generating a patient-specific trauma plate for a particular bone, the method comprising the steps of: (a) acquiring patient-specific image data of the particular injured or degenerated bone; (b) using the patient-specific image data to analyze at least one of areas where the particular bone is not present and areas where the particular bone is present; (c) generating a patient-specific virtual bone model of the particular bone in an integrated state, including bone not visible in the patient-specific image data; (d) evaluating the contour of the patient-specific virtual bone model; and (e) generating the patient-specific trauma plate using the patient-specific virtual bone model.

[0023] A ninth aspect of the present invention provides a method for kinematically tracking movement of a patient's anatomy using an inertial measurement unit, the method comprising: (a) attaching a first inertial measurement unit to the exterior of a first anatomical feature of interest of the patient; (b) attaching a second inertial measurement unit to the exterior of a second anatomical feature of interest of the patient; (c) using the first inertial measurement unit to register a position of the first anatomical feature of interest of the patient to a virtual model of the first anatomical feature of interest of the patient; (d) using the second inertial measurement unit to register a position of the second anatomical feature of interest of the patient to the virtual model of the second anatomical feature of interest of the patient; (e) using the first inertial measurement unit to dynamically correct the position of the first anatomical feature of interest of the patient with the virtual model of the first anatomical feature; and (f) using the second inertial measurement unit to dynamically correct the position of the second anatomical feature of interest of the patient with the virtual model of the second anatomical feature. [Brief explanation of the drawings]

[0024] [Figure 1] FIG. 1 is a schematic diagram of the overall process for generating a mass-customized and patient-specific mold from a patient's anatomy. [Figure 2] FIG. 1 is a schematic diagram detailing how new anatomical structures are added to the statistical atlas and how correspondences are generated. [Figure 3] FIG. 3 is a schematic diagram of a multi-resolution 3D registration algorithm corresponding to the multi-resolution 3D registration of FIG. 2. [Figure 4] FIG. 1 illustrates multi-scale registration of feature points using multi-scale features. [Figure 5] FIG. 4 illustrates a low-level decomposition of the multi-resolution registration outlined in FIG. 3. [Figure 6] FIG. 11 is a graphical representation of capturing population variation during alignment generation. [Figure 7] FIG. 1 is a schematic diagram of the whole bone reconstruction process with incomplete, deformed, or crushed anatomy. [Figure 8] FIG. 1 is a schematic diagram of the defect classification process that generates a defect template. [Figure 9] Figure 1 shows a graphical example of the existing AAOS classification of acetabular defects. [Figure 10] FIG. 1 shows a graphical example of the existing Paprosky acetabular defect classification. [Figure 11] FIG. 1 shows a graphical example of existing Paprosky acetabular defect subclassifications. [Figure 12] Table and associated figures showing pelvic reconstruction results for various defects, an exemplary application and validation of whole bone reconstruction shown in Figure 7 . [Figure 13] Figure 7 shows distance maps of the mean RMS error of pelvic reconstruction for various defects, validating the accuracy of the whole-bone reconstruction shown in Figure 7 . [Figure 14] A 3D model representation of a patient suffering from a severe pelvic tear is shown on the left, and an example 3D model of the patient's pelvis is shown on the right. [Figure 15] A comparison of the reconstructed left-side model and the original patient model, as well as left and right anatomy. [Figure 16] FIG. 10 shows a distance map between the reconstructed model and a mirror image of the reconstructed pelvis model. [Figure 17] Figure 1 shows a patient with a complete pelvic tear and the reconstruction results with an RMS error of 1.8 mm. [Figure 18] FIG. 1 shows the reconstruction results and mean distance map of the reconstruction error for a portion of the skull. [Figure 19] FIG. 1 shows the results of reconstruction of a comminuted femur. [Figure 20] FIG. 1 is a schematic diagram of the process for creating a patient-specific reconstructive implant. [Figure 21] FIG. 21 is a schematic diagram of the implant production process shown in FIG. 20. [Figure 22] FIG. 1 is a process flow diagram illustrating the various steps for reconstructing the entire anatomy from partial anatomy of a patient and creating a patient-specific cup implant for a pelvic tear. [Figure 23]FIG. 12 is a graphical representation of a patient-specific placement guide for a patient-specific acetabular implant. [Figure 24] This image examines the relationship between the three attachment sites of the implant and cup orientation for mass customization. [Figure 25] 10 is an image illustrating the sequence of mass customization of an acetabular cage according to the present disclosure. [Figure 26] FIG. 1 is a schematic diagram of a method for manufacturing mass-produced custom acetabular components using a modular design. [Figure 27] FIG. 1 is a schematic diagram of a process for producing a patient-specific hip stem for reconstructive surgery. [Figure 28] FIG. 1 is a schematic diagram of the mass-customized implant production process. [Figure 29] FIG. 1 is a schematic diagram illustrating the process of using a statistical atlas to generate both mass-customized and patient-specific hip implants. [Figure 30] FIG. 1 is a schematic diagram illustrating the process of using a statistical atlas to generate both mass-customized and patient-specific hip implants. [Figure 31] FIG. 1 is a schematic diagram outlining the process for designing population-specific hip stem components. [Figure 32] FIG. 11 is a graphical representation of where the landmarks on the proximal femur were placed. [Figure 33] FIG. 3D femur model of the mid-femoral canal and femoral constriction along the length of the femur. [Figure 34] FIG. 11 is a graphical representation of where the proximal femoral axis is located. [Figure 35] FIG. 11 is a graphical representation of the mid-cervical calculation positioning. [Figure 36] A graphical representation of the two points used to define the anatomical axis of the proximal femur. [Figure 37] Graphical representation of 3D proximal femoral measurements. [Figure 38]FIG. 1 shows exemplary Dorr ratios in 2D (X-ray) in general. [Figure 39] FIG. 11 is a graphic representation of the B / A ratio at the IM isthmus. [Figure 40] FIG. 11 is a graphical representation of IM tube measurements. [Figure 41] FIG. 10 shows contours and fitting circles. [Figure 42] FIG. 11 is a graphical representation of measurements for determining IM canal femoral radius ratio. [Figure 43] FIG. 1 shows two femur models showing the effect of changing the radius ratio, with the left one having a radius ratio of 0.69 and the right one having a radius ratio of 0.38. [Figure 44] Plots of BMD versus RRFW for men and women, as well as the best linear fit for each data set (male, female). [Figure 45] FIG. 1 is a graphical representation of the medial contour, neck axis, and head point of the proximal femur before alignment. [Figure 46] FIG. 11 is a graphical representation of anatomical axis alignment with the Z direction. [Figure 47] FIG. 13 is a graphical representation of the medial contour aligned with the femoral neck fulcrum. [Figure 48] FIG. 11 is a graphical representation of various models generated using interpolation between models, illustrating the smoothness of the interpolation. [Figure 49] FIG. 1 shows a graphical and pictorial representation of three-dimensional mapping of bone density. [Figure 50] Radiographic representation of IM width at three levels as well as proximal axis, femoral head offset, and femoral head. [Figure 51] FIG. 10 is a plot of proximal angle versus head offset. [Figure 52] FIG. 11 is a plot of proximal angle versus head height. [Figure 53] FIG. 11 is a plot of head offset versus head height. [Figure 54] FIG. 10 shows a histogram of proximal angles. [Figure 55] FIG. 10 shows plots of female and male clusters for head offset and calcar diameter. [Figure 56] FIG. 10 shows plots of female and male clusters for head offset and proximal angle. [Figure 57] FIG. 10 is a diagram showing a histogram of femoral head offset. [Figure 58] FIG. 10 is a diagram showing a histogram of IM sizes. [Figure 59] FIG. 1 is a graphical representation of female measurements for the proximal femur. [Figure 60] FIG. 1 is a graphical representation of male measurements for the proximal femur. [Figure 61] FIG. 1 is a graphical representation of female measurements of greater trochanter height. [Figure 62] FIG. 11 is a graphical representation of male measurements for greater trochanter height. [Figure 63] 1 is a diagram and table showing a graphic representation of the differences in intramedullary canal shape between men and women. [Figure 64] FIG. 1 shows a female femur and intramedullary canal representative of normal bone density and quality. [Figure 65] FIG. 1 shows a female femur and intramedullary canal representative of subnormal bone density and quality. [Figure 66] FIG. 1 shows the femur and intramedullary canal of a woman representative of osteoporosis. [Figure 67] FIG. 10 shows a histogram containing an interpolated data set of femoral head offsets. [Figure 68] FIG. 1 shows a histogram containing a dataset of tube sizes. [Figure 69] FIG. 1 shows the distribution of medial contours and head centers for various femoral groups. [Figure 70] FIG. 1 is a plot of the distribution of femoral head offsets for a group of femurs of a particular size. [Figure 71] 1 is a table reflecting anteversion angle measurements for men and women. [Figure 72] This is a photograph showing how to measure the front-to-back height. [Figure 73] Figure 10 plots femoral head height versus anterior-posterior head height relative to the fulcrum for males and females, including the best linear fit for each data set (male, female). [Figure 74] Plot of femoral head height versus anteroposterior head height relative to the anatomical axis midpoint for males and females, including best linear fits for each data set (male, female). [Figure 75] FIG. 10 is a graphical representation of parameters used to create a hip stem implant family based on gender and / or ethnicity according to the present disclosure, leading to mass-customized implant shape parameters for femoral stem components extracted by clustering. [Figure 76] 1A and 1B are assembled and exploded views of a primary hip stem. [Figure 77] 1A and 1B are assembled and exploded views of a revision hip stem. [Figure 78] 12A-12C are graphical representations of surface points of planes and their use to isolate acetabular cup shapes in accordance with the present disclosure; [Figure 79] FIG. 10 is a graphical representation of multiple virtual 3D acetabular cup anatomical templates created in accordance with the present disclosure. [Figure 80] FIG. 1 is a graphical representation of an anatomical acetabular cup and femoral stem ball shape showing multiple cup radii. [Figure 81] This is a two-dimensional representation of the curvature that fits between the acetabular cup and the femoral head. [Figure 82] FIG. 12 is a graphical representation of the mapping contour of the pelvis used in cross-sectional analysis of the acetabular cup. [Figure 83] FIG. 10 is a graphical representation of the automated detection of the transverse acetabular ligament according to the present disclosure as a method for determining the orientation of an acetabular implant cup. [Figure 84]FIG. 11 is a graphical representation of the extraction sequence of pore shapes and sizes that match the patient's bone anatomy according to a micro-computed tomography scan of the patient. [Figure 85] 1A-1C illustrate an exemplary process for creating a pet-specific implant and cutting guide according to the present disclosure. [Figure 86] FIG. 1 illustrates an exemplary process for creating mass-customized orthopedic implants for pets using a statistical atlas, according to the present disclosure. [Figure 87] 1 illustrates an exemplary process for generating a patient-specific cutting and placement instrument for an implant system according to the present disclosure. [Figure 88] 87 illustrates an exemplary process for non-rigid registration and creation of a patient-specific 3D pelvis and proximal femur model from X-ray according to the present disclosure. [Figure 89] 1A-1C are photographs and multiple x-ray images used in the reconstruction of the pelvis and proximal femur in accordance with the present disclosure. [Figure 90] FIG. 88 illustrates an exemplary process for automated segmentation of the pelvis and proximal femur from the MRI and CT scans shown in FIG. 87. [Figure 91] FIG. 88 illustrates an exemplary process for automated segmentation of complex fractured biological structures from MRI or CT scans as outlined in FIG. 87. [Figure 92] 1A-1C illustrate an exemplary process for virtual templating of both an acetabular cup and a femoral stem for use in a hip replacement procedure. [Figure 93] 93A-93C show an exemplary process for automated femoral stem replacement with distal fixation, which is a specific example of the general process outlined in FIG. 92. [Figure 94] 93A-93C show an exemplary process for automated femoral stem replacement with a press fit and three contact points, which is a specific example of the general process outlined in FIG. 92. [Figure 95] FIG. 10 is a graphical representation of automatic pelvis labeling in accordance with the present disclosure. [Figure 96] FIG. 10 is a graphical representation of automatic cup orientation and placement in accordance with the present disclosure. [Figure 97] 1 is an image including a series of x-rays overlaid with measurement and calculation data for acetabular cup and femoral stem placement assessment in accordance with the present disclosure. [Figure 98] 12A-12C are graphical representations of the evaluation of acetabular cup and femoral stem placement to ensure full limb length restoration and orientation in accordance with the present disclosure. [Figure 99] 10 is a screenshot of a pre-planning interface for assessing and modifying implant placement and sizing, according to the present disclosure. [Figure 100] 10A-10C include a series of sequential images illustrating an exemplary process of resecting and placing a femoral stem using patient-specific instruments. [Figure 101] 10A-10C include a series of sequential images illustrating an exemplary process of drilling and placing an acetabular cup using a patient-specific guide. [Figure 102] 10A-10C illustrate a series of 3D virtual maps of the acetabulum that can be used to generate patient-specific instruments and locking mechanisms in accordance with the present disclosure. [Figure 103] FIG. 1 illustrates an exemplary process for using an inertial measurement unit as part of surgical navigation during hip replacement surgery. [Figure 104] 1 is a series of sequential images illustrating an exemplary process of using an inertial measurement unit as part of surgical navigation during a hip replacement surgery. [Figure 105] 1 is a series of sequential images illustrating an exemplary process of using an inertial measurement unit as part of femur-specific surgical navigation during hip replacement surgery. [Figure 106] 1A-1C are graphical representations of exemplary instruments and processes for calibrating the position of an inertial measurement unit for subsequent pelvis-specific surgical navigation during hip replacement surgery. [Figure 107] FIG. 10 is an exemplary process flow diagram for preparing for and using an inertial measurement unit during a surgical procedure, as well as assessing surgical outcomes through inertial measurements after the surgical procedure is completed. [Figure 108] 1 is a series of images showing inertial measurement unit pods / housings attached to various instruments as part of facilitating surgical navigation during a surgical procedure. [Figure 109] 1 is a series of images showing an inertial measurement unit (IMU) pod / housing, including photographs of the calibration of the IMU relative to the patient's anatomy, photographs showing the surgical navigation of a reamer using the IMU pod, and photographs showing the surgical navigation of an acetabular cup impactor using the IMU pod. [Figure 110] Figure 1 shows a photograph within a photograph illustrating the use of an IMU pod in conjunction with an acetabular cup impactor, as well as a graphical interface (inset) showing a model of the patient's anatomy (in this case the pelvis) and the distal end of the impactor color-coded to ensure the orientation of the impactor is aligned with the preplanned orientation of the surgery. [Figure 111] 1 is a photograph of an IMU utilized in accordance with the present disclosure, along with a reference ruler characterizing the relative dimensions of the IMU. [Figure 112] FIG. 1 is an exemplary process flow diagram for producing trauma plates and fixation devices for a given population according to the present disclosure. [Figure 113] 1 is a graphical image of an average bone showing localization points on the bone surface that define the shape of the bone or trauma plate by localization in a population of bones from a statistical atlas. [Figure 114] Graphic images of bones showing the propagation of plate tracks through a population (shown here in one example). [Figure 115] 10 is a graphic image showing the extraction of the midline curve of the bone / trauma plate after the propagation of the plate trajectory. [Figure 116] This is a graphic representation of the results of calculating the 3D radius of curvature (parameter) of the trauma plate midline curve. [Figure 117] FIG. 11 is a graphical representation of how the length of the trauma plate is calculated after the propagation of the plate trajectory. [Figure 118]FIG. 11 is a graphical representation of how the mid-plate width of a trauma plate is calculated after the propagation of the plate trajectory. [Figure 119] FIG. 11 is a graphical representation of how the cross-sectional plate radius of a trauma plate is calculated after the propagation of the plate trajectory. [Figure 120] FIG. 10 plots plate size data used to determine the optimal number of clusters. [Figure 121] FIG. 111 includes 2D and 3D plots of plate size data used to generate clusters (identified as "clustering" in FIG. 111). [Figure 122] 111 shows a number of images reflecting the parameterization of plate size (identified as "parameterization curve" and "model generation" in FIG. 111). [Figure 123] 10 is an exemplary image showing a bone / trauma plate of a particular cluster being fitted to one of the bone models to assess the conformance / fit with the cluster. [Figure 124] FIG. 10 shows a 3D surface distance map reflecting the distance between the underside of the bone / trauma plate surface and the surface of the bone model selected for evaluation of plate fitting. [Figure 125] FIG. 14 illustrates validation of design plates on a cadaver to avoid muscle and ligament impingement. [Figure 126] 1 is an exemplary diagram reflecting the interaction between elements of an exemplary patient-adapted clavicle trauma system according to the present disclosure. FIG. [Figure 127] FIG. 127 is an exemplary process flow diagram of the preliminary planning element shown in FIG. 126. [Figure 128] FIG. 127 is an exemplary process flow diagram for intraoperative guidance (in this case using fluoroscopy) shown in FIG. 126. [Figure 129] 1 is a fluoroscopic image of the clavicle adjacent to a superior view of the clavicle partially including the surrounding anatomy. [Figure 130] FIG. 127 is an exemplary process flow diagram for intraoperative guidance (in this case using ultrasound) shown in FIG. 126. [Figure 131]FIG. 10 shows a graphical representation of matching radiographic or fluoroscopic images taken during range of motion, as well as plots of post-operative assessment of shoulder kinematics using one or more inertial measurement units. [Figure 132] FIG. 1 is a three-dimensional view of a pair of clavicles including the surrounding anatomy. [Figure 133] 1A-1C are two different views showing a clavicle model and the points along the bone model used to identify the clavicle midline curvature. [Figure 134] FIG. 1 shows a clavicle model and muscle attachment locations on a bone model. [Figure 135] FIG. 1 shows a series of surface maps of average clavicle models for men and women across a given population and the degree of shape difference within each population. [Figure 136] A pair of three-dimensional views of the correlated contour differences of the clavicle, including muscle attachment sites. [Figure 137] A series of cross-sectional views of the clavicle taken across male and female populations showing clavicle contour differences at various muscle attachment sites. [Figure 138] FIG. 1 is a series of cross-sectional views of clavicles taken across a population of men and women showing clavicle contour differences along the length of the clavicle. [Figure 139] FIG. 1 shows left and right clavicle models generated in response to population data in a statistical atlas that reflect morphological differences between the left and right clavicles. [Figure 140] FIG. 10 illustrates a clavicle model fitted with a superolateral plate (left), plate midline curve (center), and midline plate curvature showing the radius of curvature (right) according to the present disclosure. [Figure 141] FIG. 1 shows the superolateral plate clusters of male and female clavicle populations (Table 1 contains data on this). [Figure 142] FIG. 10 shows a clavicle model fitted with an anterior mid-diaphyseal 7h plate (left), plate midline curve (center), and midline plate curvature showing a single radius of curvature (right), in accordance with the present disclosure. [Figure 143]Figure 2 shows the anterior mid-diaphyseal 7h plate cluster of male and female clavicle populations (Table 2 contains data on this). [Figure 144] FIG. 10 illustrates a clavicle model fitted with a superior medial diaphyseal plate (left), plate midline curve (center), and midline plate curvatures showing various radii of curvature (right) in accordance with the present disclosure. [Figure 145] FIG. 1 shows the upper mid-diaphyseal plate clusters of male and female clavicle populations (Table 3 contains data on this). [Figure 146] FIG. 10 illustrates a clavicle model fitted with an anterolateral plate (left), plate midline curve (center), and midline plate curvatures showing various radii of curvature (right), in accordance with the present disclosure. [Figure 147] 1 shows the anterolateral plate clusters of male and female clavicle populations (Table 4 contains data on this). [Figure 148] FIG. 10 illustrates a clavicle model fitted with an anterior mid-diaphyseal long plate (left), plate midline curve (center), and midline plate curvatures showing various radii of curvature (right) in accordance with the present disclosure. [Figure 149] 1 shows the anterior mid-diaphyseal plate clusters of male and female clavicle populations (Table 5 contains data on this). [Figure 150] FIG. 1 is an exemplary process flow diagram for generating a customized plate placement guide for trauma reconstructive surgery in accordance with the present disclosure. [Figure 151] FIG. 1 is an exemplary process flow diagram for creating a customized cutting and placement guide for reconstructive surgery using bone grafts according to the present disclosure. [Figure 152] FIG. 10 is an exemplary process flow diagram for producing a trauma plate template and placement instrument according to the present disclosure. [Figure 153] FIG. 1 is an exemplary process flow diagram for producing a hip revision cage template and placement instrument according to the present disclosure. [Fig. 154]FIG. 1 illustrates an exemplary process flow diagram for soft tissue and motion tracking of body anatomy using an inertial measurement unit, according to the present disclosure. [Figure 155] 10A-10C are a pair of screenshots illustrating a kinematic software interface for identifying soft tissue locations on a bone model and tracking soft tissue deformation, in accordance with the present disclosure. [Figure 156] Includes bone models of the femur, tibia, and fibula showing points on each bone model where ligaments (MCL, LCL) are attached, color-coded to identify points with high or low likelihood of ligament attachment. [Figure 157] Includes bone models of the distal femur showing points on each bone model where ligaments (ACL, PCL) are attached, with these points color-coded to identify points with high or low likelihood of ligament attachment. [Figure 158] Includes a bone model of the proximal tibia showing points on each bone model where ligaments (ACL, PCL) are attached, with these points color-coded to identify points with high or low likelihood of ligament attachment. [Figure 159] 1A-1D show an anterior view, a posterior view, and two lateral views of a 3D virtual model of a knee joint including ligament attachments according to the present disclosure. [Figure 160] FIG. 159 shows the kinematic modeling of the fully assembled knee joint model of FIG. 159 using fluoroscopic images. [Figure 161] 10A-10C illustrate a distal femur and proximal tibia bone model reflecting real-time tracking of anatomical axes in accordance with the present disclosure. [Figure 162] Figure 1 includes a knee joint model of the range of motion and a reconstruction of the helical axis. [Figure 163] FIG. 1 is a diagram including a knee joint bone model showing anatomical axes in the coronal plane. [Fig. 164] FIG. 1 illustrates a clinical examination of the knee joint using an inertial measurement unit to record motion data, in accordance with the present disclosure. [Figure 165] FIG. 1 illustrates a clinical examination of the knee joint using an inertial measurement unit to record motion data, in accordance with the present disclosure. [Figure 166] FIG. 1 illustrates a clinical examination of the knee joint using an inertial measurement unit to record motion data, in accordance with the present disclosure. [Figure 167] FIG. 1 illustrates a clinical examination of the knee joint using an inertial measurement unit to record motion data, in accordance with the present disclosure. [Figure 168] FIG. 1 illustrates a clinical examination of the knee joint using an inertial measurement unit to record motion data, in accordance with the present disclosure. [Figure 169] FIG. 1 illustrates a clinical examination of the knee joint using an inertial measurement unit to record motion data, in accordance with the present disclosure. [Figure 170] FIG. 1 illustrates a clinical examination of the knee joint using an inertial measurement unit to record motion data, in accordance with the present disclosure. [Figure 171] FIG. 1 illustrates a clinical examination of the knee joint using an inertial measurement unit to record motion data, in accordance with the present disclosure. [Fig. 172] FIG. 1 illustrates a clinical examination of the knee joint using an inertial measurement unit to record motion data, in accordance with the present disclosure. [Figure 173] It includes a series of photographs, the first showing a patient fitted with a pair of inertial measurement unit (IMU) packages, the second showing the size of the IMU packages relative to the individual IMUs, and the third showing the size of the individual IMUs relative to a US quarter coin. [Fig. 174] 10 is a user interface screenshot showing a proximal tibia model that dynamically updates based on input received from an inertial measurement unit to provide feedback on load distribution as the patient's knee joint is captured through a range of motion, in accordance with the present disclosure. [Figure 175] Photograph of a patient's lower back showing separate inertial measurement units (IMUs) placed on the L1 and L5 vertebrae to track the relative movement of each vertebra through the range of motion, and accompanying diagram showing that each IMU is capable of outputting data showing movement across three axes. [Figure 176] FIG. 176 is a series of photographs showing the patient and IMU of FIG. 175 as the patient moves through a range of motion. [Figure 177] FIG. 1 is a graphical representation of a process for determining the relative orientation of at least two bodies using inertial measurement unit data, according to the present disclosure. [Figure 178] Figure 1 includes a pair of plots showing the absolute change in orientation of the anatomical axes relative to the spinal cord (specifically L1 and L5) during lateral bending activities in a patient, where (A) the data plot is representative of a healthy patient and (B) the data plot is representative of a patient with spinal degeneration. [Figure 179] FIG. 10 includes a pair of images, one a top view of the proximal tibia and the second a front perspective view of the proximal tibia, along with a surgical navigation instrument used to ensure proper orientation and positioning of the tibial implant components through IMU normalization. [Figure 180] FIG. 1 is a front perspective view of an exemplary inertial measurement unit calibration instrument according to the present disclosure. [Figure 181] FIG. 1 shows local magnetic field maps (isometric, front, and top views) generated from data output from an inertial measurement unit before calibration (the top set of three plots resembles an ellipsoid) and after calibration (the bottom set of three plots resembles a sphere). [Figure 182] This figure includes a series of diagrams showing exemplary locations of magnetometers associated with an inertial measurement unit (A), and shows what the detected magnetic field from the magnetometer should reflect when normalized to capture distortions (B), as well as the consequences of local distortions of the magnetic field at the magnetometer when no normalization is performed. [Figure 183] FIG. 1 illustrates a series of protrusions on various surgical instruments, each with a unique top surface, allowing an inertial measurement unit processor to intelligently identify the surgical instrument to which the IMU is attached. [Figure 184] FIG. 184 is a schematic diagram representing the IMU housing and showing the interaction between one of the protrusions of FIG. 183 and the bottom cavity of the IMU housing. [Figure 185]FIG. 10 is an exemplary process flow diagram for preparing a proximal humerus and implanting a humeral component as part of a shoulder replacement procedure using an inertial measurement unit according to the present disclosure. [Figure 186] FIG. 10 is an exemplary process flow diagram for creating a scapular socket and implanting a glenoid cup as part of a shoulder replacement procedure using an inertial measurement unit according to the present disclosure. [Figure 187] FIG. 10 is an exemplary process flow diagram for preparing a proximal humerus and implanting a humeral component as part of a reverse shoulder replacement procedure using an inertial measurement unit according to the present disclosure. [Figure 188] FIG. 10 is an exemplary process flow diagram for creating a scapular socket and implanting a glenoid ball as part of a reverse shoulder replacement procedure using an inertial measurement unit according to the present disclosure. [Figure 189] 1A and 1B are profile and overhead views of an exemplary UWB-IMU hybrid tracking system as part of a tetrahedron module. [Figure 190] FIG. 1 illustrates an exemplary central and peripheral system in a hip surgery navigation system, where the image on the left shows one of the anchors interrogating a tag on a peripheral unit at one time point, and the image on the right shows another anchor interrogating a tag on a peripheral unit at a subsequent time point, with each anchor interrogating the tag on the peripheral unit to determine movement and orientation relative to the anchor. [Figure 191] Figure 1 shows the experimental setup of a UWB antenna in an anechoic chamber used to measure the 3D phase center variation of the UWB antenna. A phase center bias lookup table is compiled during this process, which can be used to mitigate the phase center variation during system operation. [Figure 192] FIG. 1 is an exemplary plot of measured UWB antenna phase center error versus vertical and horizontal angles (E-plane and H-plane) for a Vivaldi antenna; the measured horizontal and vertical phase center variations are used to generate a lookup table of all possible angles of arrival, which is used to mitigate phase center bias during system operation. [Figure 193] FIG. 1 is an exemplary block diagram of a hybrid system generating multiple tags with a single UWB transceiver. [Figure 194] 1 is an exemplary block diagram of a UWB transceiver according to the present disclosure. [Figure 195] FIG. 2 illustrates an exemplary UWB pulse signal as a function of amplitude and time. [Figure 196] An example plot of receiver clock jitter and drift with a 10 MHz clock measured at the receiver over a 23 minute period. This clock jitter and drift can cause up to 30-40 mm of error in each measured distance difference, which can result in a 3D positioning error of 30 mm or more (TD BA: error between tag B and anchor A; TD CA: error between tag C and anchor A; TD DA: error between tag D and anchor A). [Figure 197] FIG. 1 is an exemplary diagram illustrating a method for calculating tag location based on TDOA. [Figure 198A] Figure 10 is a plot showing the X-position values ​​for both UWB and optical tracking systems for an optical rail experiment where the tag moves along the rail, with an RMSE in the X-dimension of 2.52 mm. [Figure 198B] Figure 10 is a plot showing the X position values ​​for both UWB and optical tracking systems for an optical rail experiment where the tag moves along the rail, with an RMSE in the X dimension of 0.93 mm. [Figure 198C] Figure 10 is a plot showing the X-position values ​​for both UWB and optical tracking systems for an optical rail experiment where the tag moves along the rail, with an RMSE in the X-dimension of 1.85 mm. [Figure 199A] 1 is a schematic summary of parameters adapted to IEEE802.15.4a channel mode with UWB experimental data acquired in an operating room (path loss comparison for IEEE802.15.4a LOS channel). [Figure 199B] FIG. 10 is a plot of path loss versus distance showing that the OR environment most closely resembles residential LOS. [Figure 200]1 is a table showing expected signal values ​​for magnetometer calibration data with no distortion and with ferromagnetic distortion introduced. [Figure 201] FIG. 1 is an exemplary block diagram of a processing and fusion algorithm for a UWB and IMU system. [Figure 202] An overhead view of the central and peripheral units in the experimental setup. During the experiment, the central unit remains stationary while the peripheral units are manipulated. [Figure 203] FIG. 1 is a plot of angle versus data samples reflecting orientation tracking using a UWB and IMU system. [Figure 204] 10 is a table showing orientation tracking for an IMU system and a hybrid system in normal and magnetostrictive environments. [Figure 205] FIG. 1 is an exemplary block diagram of computerized patient positioning through preoperative preparation and surgical planning and intraoperative use of a surgical navigation system. [Figure 206] FIG. 1 illustrates the use of one central unit on the pelvis and at least one peripheral unit on the instrument. [Figure 207] FIG. 1 illustrates the use of one central unit next to the surgical area, one peripheral unit on the pelvis, and at least one peripheral unit on the instrument. [Figure 208] FIG. 10 illustrates the use of one central and one peripheral unit to acquire and register a patient's cup shape. [Figure 209] FIG. 10 shows the use of one central unit and one peripheral unit to perform surgical guidance of the direction and depth of acetabular drilling. [Figure 210] FIG. 10 is a diagram of the attachment of the peripheral unit to the acetabular shell inserter. [Figure 211] FIG. 10 is a diagram of the attachment of the peripheral unit to the femoral broach. [Figure 212]This is a diagram of an application in body movement tracking, where the device does not require anchors to be placed outside the human body; instead, a central unit is placed toward the center of the body and acts as multiple anchors, while peripheral units are attached to each joint, and each central unit examines the spatial information of other central units and peripheral units to recreate the human body's movement. DETAILED DESCRIPTION OF THE INVENTION

[0025] Exemplary embodiments of the present disclosure are described and illustrated below to cover various aspects of orthopedics, such as bone reconstruction and tissue remodeling, patient-specific and mass-customized orthopedic implants, gender- and ethnic-specific orthopedic implants, cutting guides, trauma plates, bone graft cutting and placement guides, patient-specific instruments, etc. It will be apparent to those skilled in the art that the embodiments described below are exemplary in nature and may be reconfigured without departing from the scope and spirit of the present invention. However, for purposes of clarity and precision, the exemplary embodiments described below may include optional steps, methods, and features that those skilled in the art would recognize as not required to fall within the scope of the present invention.

[0026] Whole body structure reconstruction With reference to FIGS. 1-8 , reconstruction of deformed or incomplete anatomy is one of the complex problems facing healthcare providers. Loss of anatomy can be the result of birth conditions, tumors, disease, physical injury, or a failed previous surgery. As part of treatment for various ailments, healthcare providers may find it advantageous to reconstruct or configure anatomy to facilitate treatment for a variety of conditions, including, but not limited to, fractures / comminuted fractures, bone degeneration, orthopedic implant revisions, joint degeneration, and custom instrumentation design. For example, prior art hip reconstruction solutions require mirroring of a healthy patient's anatomy, which may not be an accurate reflection of healthy anatomy due to naturally occurring asymmetries, as shown in FIGS. 15-19 .

[0027] The present disclosure provides systems and methods for bone reconstruction and tissue reconstitution. To perform this reconstruction, the systems and associated methods utilize anatomical images representative of one or more individuals. These images are processed to create a virtual three-dimensional (3D) tissue model or series of virtual 3D tissue models that closely mimic the anatomy of interest. The systems and associated methods are then utilized to create molds and / or other devices (e.g., fixation devices, implantation devices, patient-specific implants, patient-specific surgical guides) for use in reconstructive surgery.

[0028] As shown in FIG. 1 , an exemplary system flow overview begins with receiving input data representative of an anatomy. This anatomy may include an incomplete anatomy in the case of tissue degeneration or tissue absence as a result of genetic predisposition, a deformed anatomy as a result of genetic predisposition or environmental conditions, or shattered tissue as a result of destruction of one or more anatomy. The input anatomical data may include two-dimensional (2D) images or three-dimensional (3D) surface representations of the anatomy of interest, and may be in the form of a surface model or point cloud, for example. In situations where 2D images are used, these 2D images are used to construct a 3D virtual surface representation of the anatomy of interest. Those skilled in the art are familiar with using 2D images of anatomy to construct a 3D surface representation. Therefore, a detailed description of this process is omitted for brevity. By way of example, the input anatomical data may include one or more of X-ray, computed tomography (CT), magnetic resonance imaging (MRI), or any other imaging data capable of generating a 3D surface representation of the tissue of interest.

[0029] With reference to Figure 50 and Table I, in the context of X-ray images used to construct virtual 3D bone models, it has been found that the rotation of the bone at the time of capture plays an important role in accurately constructing the model. In other words, if one wishes to edit X-ray images in a situation where bone rotation has occurred between images, the X-ray images must be normalized to capture this bone rotation.

[0030] As an example, in the context of the proximal femur, it has been found that rotating the bone 6° and 15° significantly changes measurements extracted from x-ray images. By way of example, these measurements include, but are not limited to, proximal angle, head offset, and intramedullary canal width. As reflected in Table I, for the same femur x-rayed at zero degrees (i.e., starting point established on the first x-ray), 6° and 15° rotations resulted in different proximal angles, head offsets, and intramedullary canal widths when measured using pixels with a pixel size of approximately 0.29 mm. Notably, with increasing rotation, the proximal angle increased, as did the head offset, but this was not the case for intramedullary canal width. In this exemplary table, three cross-sections were spaced along the longitudinal axis, each corresponding to a location where intramedullary canal width was measured. As reflected in Table I, the intramedullary canal width at the same location varied with the angle of rotation. As a result, as discussed in more detail below, when constructing a 3D virtual model of a bone using x-rays, it is necessary to capture rotational deviations to the extent that rotation of the bone occurs during imaging.

[0031] It should be understood, however, that the above is an exemplary description of anatomical structures that can be used in conjunction with the exemplary systems and methods, and is not intended to limit the use of the present systems in the disclosed manner with other anatomical structures. As used herein, tissue includes bone, muscle, ligament, tendon, and any other finite type of structural material with a specific function in a multicellular organism. As a result, when discussing the exemplary systems and methods in the context of bone, one skilled in the art will recognize the applicability of the systems and methods to other tissues.

[0032] Referring again to FIG. 1 , input of anatomical data into the system goes through three modules, two of which involve processing the anatomical data (the Whole Bone Reconstruction module and the Patient-Specific module), while the third (the Anomaly Database module) classifies the anatomical data as part of a database. The first processing module, the Whole Bone Reconstruction module, processes the input anatomical data with data received from the Statistical Atlas module to generate a virtual 3D model of the bone in question. This 3D model is a complete, normal reconstruction of the bone in question. The second processing module, the Patient-Specific module, processes the input anatomical data with data received from the Whole Bone Reconstruction module to generate one or more casts, fixation systems, implant molding instruments, and renderings, as well as one or more final orthopedic implants. The renderings represent visualizations of the reconstructed anatomy for feedback regarding the expected surgical outcome. More specifically, the Patient-Specific module is configured to generate fully customized appliances designed to precisely fit a patient's unique anatomy, even if the patient's anatomy significantly deviates from normal. Additionally, the patient-specific module utilizes the virtual 3D reconstructed bone model from the whole bone reconstruction module to automatically identify anatomical regions and features (e.g., fitting regions and / or shapes) for device design parameters. In this manner, patient-specific data is used to define design parameters so that the output instrument and any implants precisely fit the patient's specific anatomy. Exemplary uses of the patient-specific module are discussed in more detail below. To better understand the system's functionality and processes, the modules of the system are described below, beginning with the statistical atlas module.

[0033] As shown in Figures 1 and 2, the statistical atlas module records virtual 3D models of one or more anatomical structures (e.g., bones) to capture the inherent anatomical variation in a given population. In an exemplary embodiment, the atlas records mathematical representations of anatomical features of one or more anatomical structures, represented as an average representation and a variation about the average representation. By representing anatomical features as mathematical representations, the statistical atlas enables automated measurement of anatomical structures as well as reconstruction of missing anatomical structures, as discussed in more detail below.

[0034] To extract anatomical variations in common anatomical structures across the population, the input anatomical data is compared against a common reference frame, typically referred to as a template 3D model or anatomical 3D template model. The template 3D model is visually represented on a graphic display as a 3D model that can be rotated or visually manipulated but contains mathematical representations of the anatomical surface features / representations of all anatomical structures across the statistical atlas for the tissue in question (i.e., for a given bone, all properties of the bone are shared across the statistical atlas population generated from the template 3D model). The template 3D model can be a combination of multiple anatomical representations or a single representation instance and may represent the lowest entropy state of the statistical atlas. For each anatomical structure to be added to the statistical atlas (i.e., the input anatomical data), an anatomical 3D model is created and both the anatomical 3D model and the template 3D model undergo a normalization process.

[0035] The normalization process normalizes the anatomical 3D model to the scale of the template 3D model. The normalization process may involve scaling one or both of the anatomical 3D model and the template 3D model so that they have a common unit scale. After normalization of the anatomical 3D model and the template 3D model, the normalized anatomical 3D model and the template 3D model are rendered scale-invariant, allowing shape features to be used independent of scale (which in this case means size). After normalization is complete, both 3D models are processed through a scale-space mapping and feature extraction sequence.

[0036] Scale-space mapping and feature extraction is essentially a multi-resolution feature extraction process. Specifically, this process extracts shape-specific features at multiple feature scales. First, multiple anatomical features are selected to represent each feature present in different scale spaces. Then, for each scale-space representation of the selected anatomical features, model-specific features are extracted. These extracted features are used to derive robust (with respect to noise) registration parameters between the template 3D model and the anatomical 3D model. Following this multi-resolution feature extraction process, the extracted data is processed through a multi-resolution 3D registration process.

[0037] Referring to Figures 2 to 5, the multi-resolution 3D registration process uses scale-space extracted features to perform an affine registration calculation between the anatomical 3D model and the template 3D model to align the two models. In particular, the anatomical 3D model and the template 3D model are processed through a rigid registration process. As shown in Figure 5, this rigid registration process operates by aligning the anatomical 3D model and the template 3D model so that both models are in the same space without pose singularities. To align the 3D models, the centers of gravity associated with each model are aligned. Additionally, the major axes of each 3D model are aligned so that the major directions of both 3D models are the same. Finally, an iterative closest point calculation is performed to minimize the pose difference between the 3D models.

[0038] After rigid registration, the 3D models are aligned using a similarity registration process. In this process, the normal-scale template 3D model and the anatomical 3D model are iteratively aligned by calculating a similarity transformation that best aligns the normal-scale features (i.e., ridges) of both the template 3D model and the anatomical 3D model. The iterative similarity alignment algorithm is a variant of iterative closest point. At each iteration, the translation and scale between point pairs are calculated until convergence is reached. A distance query calculated by a Kd-tree or some other space-partitioned data structure is used to evaluate the pairwise matches or correspondences between the two point sets. In particular, the ridges of both models are used to perform the matching point pair calculation process. In this exemplary illustration, a ridge represents a point on the 3D model where a single principal curvature has an extreme value along its curvature line. As part of the matching point pair calculation process, points on the ridges of the 3D models that are compatible with each other are identified. A similarity transformation calculation process is then applied to the ridges of both 3D models to calculate the rotation, translation, and scale that best align the ridges of both models. This is followed by a point transformation process, which operates to apply the calculated rotation, translation, and scale to the ridges of the template 3D model. The root mean square error or distance error between each set of matched points is then calculated, followed by the change in relative root mean square error or distance error from the previous process. If the change in relative root mean square error or distance error is within a predetermined threshold, the transformation process applies the final rotation, translation, and scale to the template 3D model.

[0039] The similarity registration process is followed by a joint registration process, which receives input data from a scale-space feature process, which extracts features from the template 3D model and the anatomical 3D model in different scale spaces, each defined by convolving the original anatomical 3D model with a Gaussian smoothing function.

[0040] The goal of the joint registration process is to match "n" scale-space features of the template 3D model to "m" scale-space features calculated on the anatomical 3D model. The difference in the number of features detected on the template 3D model and the anatomical 3D model is due to anatomical variations. This difference in the number of detected features may result in many relationships between the template 3D model and the anatomical 3D model. Therefore, performing two-way mutual feature matching accommodates such variations and achieves accurate matches between all mutual features. Specifically, a feature set is computed on the template 3D model in scale space. In this exemplary process, the feature set is a set of connection points representing prominent anatomical structures (e.g., the acetabular cup of the pelvis, the spinal processes of the lumbar region). Similarly, a feature set is computed on the anatomical 3D model in scale space. The feature-pair matching process matches the computed feature set on the template 3D model to the feature set on the anatomical 3D model using shape descriptors (e.g., curvature, shape coefficients, etc.). The result of this process is an "nm" mapping of feature sets between the template 3D model and the anatomical 3D model. If necessary, a regrouping process is performed to regroup the matched feature sets into a single feature set (e.g., if two acetabular cups are detected, this process regroups them into a single feature set). A computational process is then performed to calculate the correspondence between each point of the matched feature sets on the template 3D model and the anatomical 3D model. This is followed by an affine computational transformation process to calculate the rotation, translation, and shear that transform each matched feature set on the template 3D model to the corresponding feature set on the anatomical 3D model. The calculated affine transformation parameters (i.e., rotation, translation, and shear) are then used to transform the template 3D model. Finally, a rigid body alignment process is performed to align each matched feature set on the template 3D model and the anatomical 3D model.

[0041] The non-rigid registration process, which follows the joint registration and normal scale feature processes, matches all surface vertices on the template 3D model to vertices on the anatomical 3D model and calculates an initial correspondence. This correspondence is then used to calculate a deformation field that moves each vertex on the template 3D model to a matching point on the anatomical 3D model. Matching is performed between vertices of the same class (i.e., scale-space feature vertices, normal scale feature vertices, or non-feature vertices). In the context of the normal scale feature process, shape features on the template 3D model and anatomical 3D model are calculated in the original scale space (ridges), which refers to the original input model.

[0042] Specifically, as part of the non-rigid registration process, scale-space features are calculated on the template 3D model (TMssf) and the anatomical 3D model (NMssf). Each feature set on the template 3D model and the anatomical 3D model is grown using k nearest neighbors. An alignment process is then applied to the scale-space features of the template 3D model to match them with corresponding features on the anatomical 3D model. Given two point clouds, the reference (X) and translation (Y), the two point clouds are iteratively aligned under the constraints of a minimum relative root-mean-square error and a maximum angle threshold, with the goal of minimizing the overall error criterion. A realignment process is performed to align the feature sets on the template 3D model to the matched sets on the anatomical 3D model using iterative closest point matching on a normal scale. After realignment, point correspondences are calculated between points in each feature set on the template 3D model and the matched feature sets on the anatomical 3D model. Matched points on the anatomical 3D model should have surface normal directions close to those of the template 3D model. The output is sent to the deformation field calculation step.

[0043] In parallel with the progress of the scale-space feature calculation, non-feature points of the template 3D model (TMnfp) and the anatomical 3D model (NMnfp) or other point sets on the surface of the template 3D model that do not belong to the scale-space features or the normal scale features are processed by correspondence calculation to calculate point correspondences between the non-feature points on the template 3D model and the non-feature points on the anatomical 3D model. The matching points on the new model should have surface normal directions close to those of the template model. The output is sent to the deformation field calculation step.

[0044] In parallel with the progress of the scale-space feature computation, we also use AICP to align regular-scale features (i.e., ridges) on the template 3D model (TMnsf) with regular-scale features (i.e., ridges) on the anatomical 3D model (NMnsf). AICP is a variant of iterative closest point computation, which computes translation, rotation, and scale between sets of matching points at each iteration. After the alignment process, we perform a matching process.

[0045] A deformation process is applied to the output of the scale-space feature computation progress, the correspondence progress, and the alignment progress, where a deformation field is calculated to move each point on the template 3D model to a matching point on the anatomical 3D model.

[0046] A relaxation process is applied to the output of the non-rigid registration process to move the vertices of the template 3D model mesh closer to the surface of the anatomical 3D model after the multi-resolution registration step, and to smooth the output model. In particular, the template 3D model in normal space (TMns) and the anatomical 3D model in normal space (NMns) are processed through a correspondence calculation to calculate the vertex on the template 3D model that is closest to the anatomical 3D model using a normal-constrained sphere search algorithm. In this calculation, the closest vertices of both models are used to generate correspondence vectors from each vertex of the template 3D model and the matching vertex of the anatomical 3D model, which may result in two or more matching points from the anatomical 3D model. Using the matching points of each vertex on the template 3D model, a weighted average of the matching points on the anatomical 3D model is calculated based on the Euclidean distance from the point and the matching point. At this point, the template 3D model is updated using the weighted average, and each point on the template 3D model is moved using the calculated weighted average distance. After the weight calculation process, a relaxation process is performed for every point on the template model to find the closest point on the surface of the anatomical 3D model and move it to that point. Finally, a smoothing operation is performed on the deformed template 3D model to remove noise. Then, a free-form deformation process is applied to the resulting aligned 3D models (i.e., the template 3D model and the anatomical 3D model).

[0047] In the free-form deformation process, the surface of the template 3D model is morphed with the surface of the anatomical 3D model. More specifically, the surface of the template 3D model is iteratively moved between weighted points using mutually matching points on the surfaces of both the template 3D model and the anatomical 3D model.

[0048] 2 and 6 , after the free-form deformation process, a correspondence calculation process is applied to the anatomical 3D model to determine the deviation between the anatomical 3D model and the morphed template 3D model. This correspondence calculation process refines the template 3D model from the free-form deformation step and performs a final fit of selected landmark locations on the deformed template 3D model and the deformed anatomical 3D model. Thus, the correspondence calculation process calculates the size and shape variations between the 3D models and records them as deviations with respect to the average model. The output of this correspondence calculation process is a normalized anatomical 3D model updated to capture the variations in the anatomical 3D model and the addition of a revised template 3D model. In other words, the output of the process outlined in FIG. 2 is a modified normalized anatomical 3D model with properties (e.g., point correspondences) consistent with the revised template 3D model to facilitate whole-body structure reconstruction (e.g., whole-bone reconstruction).

[0049] 1 and 7, input and anatomical data from the statistical atlas module go to a full anatomy reconstruction module. By way of example, the anatomy of interest may be a single bone or multiple bones. However, it should be noted that the exemplary hardware, processes, and techniques described herein can be used to reconstruct anatomy other than bone. In an exemplary embodiment, the full anatomy reconstruction module may receive input data regarding an incomplete, deformed, or fractured pelvis. The input anatomical data may include two-dimensional (2D) images or three-dimensional (3D) surface representations of the anatomy of interest, which may be in the form of a surface model or point cloud, for example. In situations where 2D images are used, these 2D images are used to construct a 3D surface representation of the anatomy of interest. Those skilled in the art are familiar with using 2D images of anatomy to construct a 3D surface representation. Therefore, a detailed description of this process is omitted for brevity. By way of example, the input anatomical data may include one or more of X-ray, computed tomography (CT), magnetic resonance imaging (MRI), or any other imaging data capable of generating a 3D surface representation. As discussed in more detail below, this input anatomical data can be used for, but is not limited to, (1) starting points to identify the closest statistical atlas 3D bone model, (2) registration using the 3D surface vertex set, and (3) a final relaxation step for the reconstruction output.

[0050] As shown in FIG. 7, input anatomical data (e.g., a patient's bone model) is utilized to identify an anatomical model (e.g., a bone model) in the statistical atlas that most closely resembles the patient's anatomy in question. This step is depicted in FIG. 3 as a search for the closest bone in the atlas. First, the patient's bone model is compared to the statistical atlas's bone models using one or more similarity criteria to identify the statistical atlas's bone model that most closely resembles the patient's bone model. The result of this initial similarity criteria is the selection of a bone model from the statistical atlas to be used as an "initial guess" in a subsequent registration step. In the registration step, the patient's bone model is aligned with the selected atlas bone model (i.e., the initial guess bone model) so that the output is a patient bone model aligned with the atlas bone model. Following the registration step, the shape parameters of the aligned "initial guess" are optimized so that the shape matches the patient's bone shape.

[0051] Optimization of shape parameters (in this case from a statistical atlas) minimizes the error between the reconstruction and the patient bone model by using undeformed or existing bone regions. By varying the values ​​of the shape parameters, different anatomical shapes can be represented. This process is repeated at different scale spaces until convergence of the reconstructed shape is achieved (possibly measured as the relative surface change between iterations or the maximum number of allowed iterations).

[0052] A relaxation step is performed to morph the optimized tissue to best fit the patient's original 3D tissue model. Consistent with this exemplary case, the missing anatomy from the reconstructed pelvis model output by the convergence step is applied to the patient-specific 3D pelvis model to create a patient-specific 3D model of the reconstructed pelvis. More specifically, surface points on the reconstructed pelvis model are relaxed (i.e., morphed) directly onto the patient-specific 3D pelvis model to best fit the reconstructed shape to the patient-specific shape. The output of this step is a fully reconstructed patient-specific 3D tissue model representing what should be the patient's normal / complete anatomy.

[0053] 1, the anomaly database serves as the data input and training for the defect classification module. In particular, the anomaly database contains data specific to abnormal anatomical features, including anatomical surface delineations and associated clinical and demographic data.

[0054] 1 and 8, the defect classification module receives input anatomical data (i.e., 3D surface representations or data from which a 3D surface representation can be generated) representing abnormal / incomplete tissue from an anomaly database, as well as a fully reconstructed patient-specific 3D tissue model representing normal / complete tissue. The anatomical data from the anomaly database may be incomplete anatomical structures in cases of tissue degeneration or tissue absence as a result of genetic predisposition, deformed anatomical structures as a result of genetic predisposition or environmental conditions (e.g., reoperation, disease, etc.), or shattered tissue as a result of destruction of one or more anatomical structures. By way of example, the input anatomical data may include one or more of X-ray, computed tomography (CT), magnetic resonance imaging (MRI), or any other imaging data from which a 3D surface representation can be generated.

[0055] The defect classification module draws multiple anomaly 3D surface representations from the anomaly database combined with normalized 3D representations of the anatomical structures in question to create a quantitative defect classification scheme. This defect classification scheme is used to create "templates" for each defect class or cluster. More generally, the defect classification module categorizes anatomical defects into classes of closely related defects (representing defects with similar shape, clinical, appearance, or other characteristics) to facilitate the generation of medical solutions to address these defects. The defect classification module uses software and hardware to automatically classify defects as a means of eliminating or reducing discrepancies between preoperative data and intraoperative observer visualization. Traditionally, preoperative radiographs were used as a means of qualitatively analyzing the extent of the need for anatomical reconstruction, resulting in haphazard preoperative planning. Currently, observers make the final decision on the extent of anatomy defects during surgery, repeatedly resulting in poor or incomplete preoperative planning based on radiographs. As a result, the present defect classification module improves upon current classification schemes by reducing the inter- and intra-observer variability associated with defect classification and by providing quantitative criteria for classifying new defect cases.

[0056] As part of the defect classification module, the module may receive as input one or more classification types to use as initial conditions. For example, in the context of the pelvis, the defect classification module may use the bone defect classification structure corresponding to D'Antonio et al. of the American Academy of Orthopaedic Surgeons (AAOS) as input defect features. This structure includes four distinct classes: (1) Type I, corresponding to segmental osteoporosis; (2) Type II, corresponding to cavitary osteoporosis; (3) Type III, corresponding to combined segmental and cavitary osteoporosis; and (4) Type IV, corresponding to pelvic disruption. Alternatively, the defect classification module may be programmed with the Paprosky bone defect classification structure, graphically depicted in FIG. 10 for the pelvis. This structure includes three distinct classes: (1) Type I, corresponding to an accessory rim without osteolysis; (2) Type II, corresponding to an intact accessory column and a distorted hemisphere with less than 2 cm of superior medial rim or lateral translation; and (3) Type III, corresponding to severe ischiolysis with more than 2 cm of superior translation and broken or intact Kohler's line. The defect classification module may further be programmed with a modified Paprosky bone defect classification structure, which includes six distinct classes: (1) Type 1, corresponding to an auxiliary rim with no component movement; (2) Type 2A, corresponding to a distorted hemisphere with less than 3 cm of superior translation; (3) Type 2B, corresponding to a larger hemisphere distortion with less than one-third of the rim circumference and a supported dome; (4) Type 2C, corresponding to an intact rim, medial translation of Kohler's line, and a supported dome; (5) Type 3A, corresponding to severe ischiolysis with more than 3 cm of superior translation and an intact Kohler's line; and (6) Type 3B, corresponding to severe ischiolysis with more than 3 cm of superior translation, a broken Kohler's line, and a rim defect of more than half the circumference. The defect classification module uses the output classification types and parameters to compare the anatomical data with the reconstructed data to identify the classification type corresponding to the assigned classification obtained by the most similar anatomical data.

[0057] As a first step, the statistical atlas addition step generates a correspondence between the normal atlas 3D bone model and the abnormal 3D bone model. More specifically, the 3D bone models are compared to identify bones in the normal 3D model that are not present in the abnormal 3D model. In an exemplary embodiment, missing / abnormal bones are identified by comparing points on the surface of each 3D bone model and generating a list of discrete points on the surface of the normal 3D bone model that are not present on the abnormal 3D bone model. The system may also record and list (i.e., identify) those surface points that are common between the two models, or informally note that all other points are common to both bone models (i.e., both the normal and abnormal bone models) unless they are recorded as points that are not present on the abnormal 3D bone model. Thus, the output of this step is the abnormal 3D bone model along with the statistical atlas correspondence and a list of features (points) from the normal atlas 3D bone model that indicate whether the feature (point) is present or missing in the abnormal 3D bone model.

[0058] After generating the correspondence between the normal atlas 3D bone model (generated from the whole bone reconstruction module) and the abnormal 3D bone model (generated from the input anatomical data), missing / abnormal regions from the abnormal 3D bone model are located on the normal atlas 3D bone model. In other words, the normal atlas 3D bone model is compared with the abnormal 3D bone model to identify and record missing bones from the abnormal 3D bone model that exist in the normal atlas 3D bone model. The location can be performed in various ways, including, but not limited to, curvature comparison, surface area comparison, and point cloud area comparison. Finally, in an exemplary embodiment, the missing / abnormal bone is located as a set of boundary points that identify the geometric boundary of the missing / abnormal region.

[0059] Using the boundary points, the defect classification module extracts features from the missing / abnormal regions using the input clinical data. In an exemplary embodiment, the extracted features may include shape information, volume information, or any other information used to describe the overall characteristics of the defective (i.e., missing or abnormal) region. These features may be refined based on existing clinical data, such as the current defect classification data or patient clinical information (demographics, medical history, etc.) that is not necessarily related to anatomical features. The output of this step is a mathematical descriptor representative of the defect region that is used in a subsequent step to group similar tissue (e.g., bone) anomalies.

[0060] The mathematical descriptors are clustered or grouped based on statistical analysis. In particular, the descriptors are statistically analyzed and compared to other descriptors from other patients / cadavers to identify unique defect classes within a given population. This classification obviously assumes multiple descriptors from multiple patients / cadavers, which improves classification and identification of discrete groups as the number of patients / cadavers increases. The output of this statistical analysis is a set of defect classes that are used to classify new input anatomical data, determining the number of templates.

[0061] The output of the defect classification module is directed to a template module. In an exemplary embodiment, the template module includes data specific to each defect classification identified by the defect classification module. By way of example, each template for a given defect classification includes a surface description of the defective bone, the location of the defect, and measurements related to the defective bone. This template data may be in the form of surface shape data, a point cloud description, one or more curvature profiles, dimensional data, and physical quantity data. The output of the template module and statistical atlas is utilized by a mass customization module to enable the design, testing, and manufacturing of mass-customized implants, fixation devices, appliances, or molds. Exemplary uses of the mass customization module are discussed in more detail below.

[0062] Patient-specific reconstructive implants 1 and 20 , an exemplary process and system for generating a patient-specific orthopedic implant guide and associated patient-specific orthopedic implant for a patient suffering from defective, deformed, and / or comminuted anatomy will be described. For purposes of illustrative description, a total hip arthroplasty for a patient with deficient anatomy will be described. However, it will be understood that the exemplary process and system are applicable to any orthopedic implant suitable for patient-specific customization in cases where defective or deformed anatomy is present. For example, the exemplary process and system are applicable to shoulder and knee replacements where bone degeneration (defective anatomy), bone deformity, or comminuted fractures are present. As a result, while a hip implant is discussed below, those skilled in the art will understand that the system and process are applicable to other orthopedic implants, guides, instruments, etc., used in primary or revision orthopedic surgical procedures.

[0063] Pelvic disruption is the osteoporotic variant most commonly associated with total hip arthroplasty (THA), in which the upper surface of the pelvis can separate from the lower surface due to osteolysis or acetabular fracture. The amount and severity of osteoporosis and the potential for biological implant ingrowth are some of the factors that may influence the choice of treatment for a particular patient. In cases of severe osteoporosis and loss of pelvic integrity, a custom tri-flange cup may be used. First introduced in 1992, this implant offers several advantages over existing cages: it provides stability for pelvic disruption, eliminates the need for construct grafting and intraoperative shaping of the cage, and promotes osseointegration of the construct to the surrounding bone.

[0064] Regardless of the context, whether or not a patient's incomplete, deformed, and / or comminuted anatomy is an issue, exemplary systems and processes for generating patient-specific implants and / or guides utilize the exemplary 3D bone model reconstruction process and system described above (see FIGS. 1-7 and the exemplary description above) to generate a three-dimensional model of the patient's reconstructed anatomy. More specifically, in the context of total hip arthroplasty associated with a pelvic tear, exemplary patient-specific systems utilize the patient's pelvis data to generate a 3D model of the patient's complete pelvis for each side (right or left). As a result, for the sake of brevity, a description of systems and processes for utilizing patient anatomy data and generating a 3D reconstructed model of the patient's anatomy in the case of incomplete anatomy has been omitted. Therefore, a description of a process and system for generating patient-specific orthopedic implant guides and associated patient-specific orthopedic implants for patients suffering from incomplete, deformed, and / or comminuted anatomy will be provided after the generation of the three-dimensional reconstructed model.

[0065] With specific reference to FIGS. 20-22 and 27, after generating the patient-specific reconstructed 3D bone models of the pelvis and femur, both the incomplete patient-specific 3D bone models (of the pelvis and femur) and the reconstructed 3D bone models (of the pelvis and femur) are used to create a patient-specific orthopedic implant and a patient-specific placement guide and / or fasteners for the implant. In particular, the defect shape extraction step includes generating a correspondence between the patient-specific 3D model and the reconstructed 3D model (a correspondence between the pelvis models and a correspondence between the femur models, but not a correspondence between a femur model and a pelvis model). More specifically, the 3D models are compared to identify bones in the reconstructed 3D model that are not present in the patient-specific 3D model. In an exemplary embodiment, missing / abnormal bones are identified by comparing points on the surface of each 3D model and generating a list of discrete points on the surface of the reconstructed 3D model that are not present on the patient-specific 3D model. The system may also record and list (i.e., identify) those surface points that are common between the two models, and may informally note that all other points are common to both models (i.e., both the reconstructed 3D model and the patient-specific 3D model) unless they are recorded as points that are not present on the patient-specific 3D model.

[0066] Referring to FIG. 21 , after generating the correspondence between the reconstructed 3D model (generated from the whole bone reconstruction module) and the patient-specific 3D model (generated from the input anatomical data), missing / abnormal regions from the patient-specific 3D model are located on the reconstructed 3D model. In other words, the reconstructed 3D model is compared with the patient-specific 3D model to identify and record missing bones from the patient-specific 3D model that are present in the reconstructed 3D model. The location can be performed in various ways, including, but not limited to, curvature comparison, surface area comparison, and point cloud area comparison. Finally, in an exemplary embodiment, the missing / abnormal bones are located, and the output includes two lists: (a) a first list identifying vertices corresponding to bones in the reconstructed 3D model that are missing or deformed in the patient-specific 3D model, and (b) a second list identifying vertices corresponding to bones in the reconstructed 3D model that are present and normal in the patient-specific 3D model.

[0067] Referring to Figures 21, 22, and 27, the defect shape extraction step is followed by an implant trajectory step. Using the two vertex lists from the defect shape extraction step and a 3D model of a normal bone (e.g., pelvis, femur, etc.) from the statistical atlas (see Figures 1, 2, and the exemplary description above), fixation points for the femoral or pelvic implant are identified. More specifically, fixation points (i.e., implant trajectories) are automatically selected so that the implant is positioned where the patient has residual bone. Conversely, fixation points are not selected in areas where the patient's residual bone is missing. In this way, fixation points are selected independently of the final implant design / shape. The selection of fixation points may be performed automatically using the shape information and the statistical atlas positions.

[0068] As shown in FIG. 21 , after the implant trajectory step, the next step is to generate patient-specific implant parameters. To complete this step, an implant parameterization template is input, which defines the implant with a sufficient number of set parameters to define the basic shape of the implant. As an example, in the case of replacing / augmenting a missing or degenerated acetabulum through pelvic reconstruction, the implant parameterization template includes angular parameters for the orientation of the replacement acetabular cup and depth parameters corresponding to the dimensions of the femoral head. Other parameters of the acetabular implant include, but are not limited to, the diameter, surface orientation, flange position and shape of the acetabular cup, and the location and orientation of the fixation screws. In the case of a porous implant, these parameters should include the location and structural characteristics of the pores. As an example, in the case of replacing / augmenting a missing or degenerated femur through femoral reconstruction, the implant parameterization template includes angular parameters for the orientation of the replacement femoral head, neck length, head offset, proximal angle, and cross-sectional analysis of the lateral femur and intercondylar channel. Those skilled in the art will understand that the parameters selected to define the basic shape of the implant depend on the anatomy to be replaced or supplemented. As a result, a complete description of parameters sufficient to define the basic shape of the implant is difficult to achieve. Nevertheless, a reconstructed 3D pelvic model may be utilized to obtain the radius of the acetabular cup, the identification of the pelvic bone including the circumferential superior ridge of the acetabular cup, and the orientation of the acetabular cup relative to the residual pelvis, as shown, for example, in Figure 22. Furthermore, the parameters may be refined by considering the implant trajectory to best / better fit the implant to the patient's specific anatomy.

[0069] Following the determination of a sufficient number of setting parameters to define the basic shape of the implant, the design of the implant begins. More specifically, the first iteration of the overall implant surface model is constructed. This first iteration of the overall implant surface model is defined by a combination of the patient-specific contour and the estimated contour of the implant area. The estimated contour is determined by the reconstructed 3D bone model, the missing anatomical bone, and features extracted from the reconstructed 3D bone model. These features and the location of the implant site, which can be determined automatically, are used to determine the overall implant shape, as shown in FIG. 22 for an acetabular cup implant, for example.

[0070] Referring again to FIG. 20 , the first iteration of the overall implant surface model is processed through a custom (i.e., patient-specific) planning sequence. This custom planning sequence may involve input from the surgeon and engineer as part of an iterative review and design process. In particular, the surgeon and / or engineer review the overall implant surface model and the reconstructed 3D bone model to determine whether changes are needed to the overall implant surface model. This review may result in the overall implant surface model being iterated until agreement is reached between the engineer and the surgeon. The output of this step is the final implant surface model, which may be in the form of a CAD file, CNC machine coding, or rapid manufacturing instructions for creating the final implant or a tangible model.

[0071] 20, 22, and 23, a patient-specific placement guide is designed simultaneously with or subsequent to the design of the patient-specific orthopedic implant. As described above in exemplary form, in the context of an acetabular cup implant, one or more surgical instruments may be designed and manufactured to assist in the placement of the patient-specific acetabular cup. Having designed the patient-specific implant to have a size and shape that matches the residual bone, the contours and shape of the patient-specific implant may be utilized and included as part of the placement guide.

[0072] In an exemplary embodiment, the acetabular placement guide includes three flanges configured to contact the ilium, ischium, and pubic bone surfaces, interconnected by a ring. Furthermore, the flanges of the placement guide may have the same shape, size, and contour as the acetabular cup implant, so that the placement guide has the same predetermined position relative to the acetabular cup implant. In other words, the acetabular placement guide, like the acetabular cup implant, is molded as a negative imprint of the patient's anatomy (the surfaces of the ilium, ischium, and portions of the pubic bone) to precisely match the patient's anatomy. However, the implant guide differs significantly from the implant in that it includes one or more fixation holes configured to guide drilling and / or fastener placement. In an exemplary embodiment, the placement guide includes holes sized and oriented based on image analysis (e.g., micro-CT) to ensure proper orientation of any drill bits or other guides (e.g., dowels) utilized in securing the acetabular cup implant to the residual pelvis. The number and orientation of the holes are determined by the residual bone, which also affects the shape of the acetabular cup implant. Figure 23 shows an example of a patient-specific placement guide for use in total hip arthroplasty. In another instance, the guide can be fabricated to fit over the implant and guide only the orientation of the fixation screws. In this configuration, the guide is molded as a negative of the implant so that it can be placed directly over the implant. Even so, the inclusion of at least some of the size, shape, and contour of the patient-specific reconstruction implant is a subject that applies regardless of the bone to which the patient-specific implant will be coupled.

[0073] Utilizing the exemplary systems and methods described herein can provide a wealth of information that increases the accuracy of orthopedic placement, improves anatomical integration, and enables preoperative measurement of true angular and planar orientation via reconstructed three-dimensional models.

[0074] Creating customized implants using mass-customizable components An exemplary process and system for producing a customized orthopedic implant using mass-customizable components will be described with reference to Figure 26. For purposes of illustrative description, a total hip arthroplasty for a patient with a severe acetabular deficiency will be described, although it will be understood that the exemplary process and system is applicable to any orthopedic implant suitable for mass customization in cases where imperfect anatomy is present.

[0075] Severe acetabular defects require specialized treatment and restorative implant components. One approach involves a custom tri-flange, a fully custom implant consisting of an acetabular cup and three flanges attached to the ilium, ischium, and pubis. In contrast to this exemplary process and system, prior art tri-flange implants comprise a single, complex component that is difficult to manufacture and requires a redesign of the entire implant for each case (i.e., completely patient-specific). This exemplary process and system modularly utilizes fully custom components as well as mass-customizable components to create custom tri-flange implants that allow for custom fitting and porosity.

[0076] A pre-planning step according to the exemplary process is performed to determine the orientation of the three flanges relative to the cup, the flange contact points, and the orientation and size of the acetabular cup. This pre-planning step is performed in accordance with the "Patient-Specific Implants" discussed immediately above in this section. By way of example, specific implant fixation points are determined by the implant trajectory step using its pre-defined data inputs, as discussed in the immediately preceding section. As previously mentioned, as part of this implant trajectory step, custom tri-flange fixation points are identified by inputting two vertex lists from the defect geometry extraction step and a 3D model of the normal pelvis from a statistical atlas (see FIGS. 1 and 2 and the exemplary description above). More specifically, fixation points (i.e., implant trajectories) are selected so that they are positioned where the patient has residual bone. In other words, fixation points are not selected in defective areas of the patient's residual pelvis. In this manner, fixation points are selected independently of the final implant design / shape.

[0077] After the fixation points are determined, the tri-flange component (i.e., flange) is designed using the "patient-specific implant" discussed immediately above in this section. The flange is designed to be oriented relative to the replacement acetabular cup such that the cup's orientation provides acceptable joint function. Additionally, the flange's contact surface is contoured to match the patient's pelvic anatomy, in that the tri-flange's contact surface is molded as a "negative" of the pelvic bone surface. In the exemplary process of FIG. 23, the final step of the process shown in FIG. 17 is utilized to rapidly prototype the flange (or use conventional computer numerical control (CNC) equipment). After the flange is manufactured, additional machining or steps may be performed to provide cavities into which porous material can be added to the tri-flange.

[0078] The portion of the Tri-Flange system that does not need to be a custom component is the acetabular cup component. In this exemplary process, a family of acetabular cups is first manufactured to provide a foundation for building the Tri-Flange system. These "blank" cups are stored and used as needed. If a specific porosity is desired in the cup, mechanical features can be added to the cup to allow for press-fitting of a porous material into the cup. Alternatively, if a specific porosity is desired in the cup, the cup can be coated with one or more porous coatings.

[0079] As described above, after forming the blank cup and addressing any porosity issues, the cup is made patient-specific by machining it to receive the flange. In particular, the system software uses a virtual model of the flange to construct virtual locking features for the flange, which are machined into the cup via machine-coded transformation. These locking features enable the cup to be secured to the flange so that, when the flange is attached to the patient's residual bone, the cup will be properly oriented relative to the residual pelvis. This machining may involve the use of conventional CNC equipment to form the locking features in the blank cup.

[0080] Following fabrication of the locking mechanism as part of the blank cup, the flanges are attached to the cup using the interface between the locking mechanisms. The tri-flange assembly (i.e., the final implant) is subjected to an annealing process to promote a strong bond between the components. After annealing the tri-flange implant, a sterilization process is performed followed by appropriate packaging to ensure a sterile environment for the tri-flange implant.

[0081] Mass-customized implant creation With reference to FIG. 28 , an exemplary process and system for generating a mass-customized orthopedic implant guide and associated mass-customized orthopedic implant for patients suffering from defective, deformed, and / or comminuted anatomy will be described. For purposes of illustrative description, a total hip arthroplasty for a patient requiring a primary joint replacement will be described. However, it will be understood that the exemplary process and system is applicable to any orthopedic implant and guide suitable for mass customization in cases where defective anatomy is present. For example, the exemplary process and system is applicable to shoulder and knee replacements where bone degeneration (defective anatomy), bone deformity, or comminuted fractures are present. As a result, while a hip implant is discussed below, those skilled in the art will understand that the system and process are applicable to other orthopedic implants, guides, instruments, etc., used in primary or revision orthopedic surgical procedures.

[0082] The exemplary process utilizes input data from macroscopic and microscopic perspectives. In particular, the macroscopic perspective involves determining the overall geometry of the orthopedic implant and the corresponding anatomy. Conversely, the microscopic perspective involves considering the shape and structure of cancellous bone and its porosity.

[0083] The macroscopic perspective includes a database in communication with a statistical atlas module that records virtual 3D models of one or more anatomical structures (e.g., bones) to capture the unique anatomical variations in a given population. In an exemplary embodiment, the atlas records mathematical representations of anatomical features of one or more anatomical structures expressed as an average representation of a given anatomical population and variation about the average representation. See FIG. 2 and the above description of the statistical atlas and how to add anatomical structures to the statistical atlas for a given population. The output of the statistical atlas goes to an auto-labeling module and a surface / shape analysis module.

[0084] The automatic labeling module uses input from the statistical atlas (e.g., regions likely to contain specific landmarks) and local geometric analysis to calculate anatomical landmarks for each anatomy instance in the statistical atlas. This calculation is specific to each landmark. For example, the approximate shape of the region is known, and the location of the landmark being searched for is known relative to the local shape characteristics. For example, locating the medial epicondyle point of the distal femur is achieved by refining a search based on the approximate location of the medial epicondyle point in the statistical atlas. Therefore, since the medial epicondyle point is the most medial point within this search window, a search for this most medial point is performed for each bone model within the medial epicondyle region defined in the statistical atlas, and the output of the search is identified as the medial epicondyle point landmark. After automatic calculation of anatomical landmarks for each virtual 3D model in the statistical atlas population, the virtual 3D models in the statistical atlas, along with the shape / surface analysis output, are directed to the feature extraction module.

[0085] The shape / surface output comes from the shape / surface module, which also receives input from the statistical atlas. In the context of the shape / surface module, the virtual 3D models in the statistical atlas population are analyzed for shape / surface features not covered by the auto-labeling. In other words, features that correspond to the overall 3D shape of the anatomical structure but do not belong to the features defined in the previous auto-labeling steps are also calculated. For example, curvature data is calculated for the virtual 3D models.

[0086] The output of the surface / shape analysis and auto-labeling modules goes to a feature extraction module, which uses a combination of landmarks and shape features to calculate mathematical descriptors (i.e., curvature, dimensions) relevant to implant design for each instance in the atlas. These descriptors are used as input for the clustering process.

[0087] The mathematical descriptors are clustered or grouped based on statistical analysis. In particular, the descriptors are statistically analyzed and compared with other descriptors from other anatomical populations to identify groups (of anatomies) with similar characteristics within that population. This clustering obviously assumes multiple descriptors from multiple anatomies across the population. As new cases not present in the initial clustering are presented to the clustering, refinement of the output clusters will better represent the new population. The output of this statistical analysis is a finite number of implants (including implant families and sizes) that cover all or a large portion of the anatomical population.

[0088] For each cluster, the parameterization module extracts mathematical descriptors within that cluster. The mathematical descriptors constitute the parameters (e.g., CAD design parameters) of the final implant model. The extracted mathematical descriptors are fed to the implant surface generation module. This module is responsible for generating a 3D virtual model of the anatomy of each cluster by converting the mathematical descriptors into surface descriptors. The 3D virtual model complements the microscopic view prior to load testing and implant manufacturing.

[0089] At a microscopic level, data indicative of the structural integrity of each biological structure in a given population is obtained. In an exemplary embodiment, this bone data may include micro-CT data providing structural information about the cancellous bone. More specifically, the micro-CT data may include images of the bone in question (multiple micro-CT images of multiple bones across the population). These images are then segmented by a trabecular structure extraction module to extract the three-dimensional shape of the cancellous bone and create a virtual 3D model of each bone in the population. The resulting 3D virtual model is input into a pore size and shape module. As graphically depicted in Figure 84, the 3D virtual model contains pore size and shape information, which is evaluated by the pore size and shape module to determine the pore size and size of the cancellous bone. This evaluation is useful for analyzing the pore size and shape of bone in the intramedullary canal, and may facilitate integration between the femoral implant and the remaining femoral bone by coating the stem of the femoral implant or by processing it to exhibit a porous exterior. The output of this module goes to the virtual load testing module in combination with the 3D virtual model output from the implant surface generation module.

[0090] The load testing module combines implant porosity data from the pore size and shape module and implant shape data from the implant surface generation module to define a final implant shape model and properties. For example, the shape and properties include providing a porous coating on the final implant model that generally matches the cancellous porosity of the bone in question. The final implant model incorporating the shape and properties undergoes virtual load testing (finite element and mechanical analysis) to verify the model's functional quality. To the extent that the functional quality is unacceptable, the parameters defining the implant's shape and porosity are modified until acceptable performance is achieved. Assuming the final implant model meets the load testing criteria, the final implant model is used to generate the machine instructions necessary to convert the virtual model into a physical implant (which can be further refined by manufacturing processes known to those skilled in the art). In an exemplary embodiment, the machine instructions may include rapid manufacturing machine instructions for producing the final implant through a rapid prototyping process (that properly captures the porosity) or a combination of traditional manufacturing and rapid prototyping.

[0091] Creating gender / ethnicity-specific hip implants 29-84, an exemplary process and system for creating gender- and / or ethnicity-specific implants will be described. For purposes of illustrative description, a total hip arthroplasty for a patient requiring a primary joint replacement will be described. However, it will be understood that the exemplary process and system is applicable to any orthopedic implant suitable for customization. For example, the exemplary process and system is applicable to shoulder and knee replacements, as well as other primary joint replacements. Consequently, while a hip implant is discussed below, one skilled in the art will understand that the system and process can be applied to other orthopedic implants, guides, instruments, etc., used in primary orthopedic surgical procedures or revision procedures.

[0092] The hip joint is composed of the femoral head and the acetabulum of the pelvis. Its anatomy makes it one of the most stable joints in the body. This stability is provided by a rigid ball-and-socket configuration. The femoral head is approximately spherical at the joint, forming two-thirds of a sphere. Data show that the diameter of the femoral head is smaller in women than in men. In a normal hip joint, the center of the femoral head is assumed to be exactly aligned with the center of the acetabulum, and this assumption is used as the basis for the design of most hip joint systems. However, the natural acetabulum is not deep enough to cover the entire natural femoral head. The approximately circular portion of the femoral head is slightly flattened at the top, making it spheroidal rather than spherical. This spheroidal shape distributes loads in a ring-like pattern around the superior pole.

[0093] The geometric center of the femoral head is intersected by the three axes of the joint: the horizontal, vertical, and anterior-posterior axes. The femoral head is supported by the neck of the femur, which connects to the shaft. The axis of the femoral neck is set obliquely and extends superiorly, medially, and anteriorly. The angle of inclination of the femoral neck relative to the shaft in the frontal plane is the neck-shaft angle. In most adults, this angle varies from 90 to 135° and is important because it determines the effectiveness of the hip abductors, limb length, and the forces exerted at the hip joint.

[0094] An inclination angle greater than 125° is called coxa valga, while an inclination angle less than 125° is called coxa vara. An inclination angle greater than 125° is associated with increased limb length, decreased hip abductor effectiveness, increased load on the femoral head, and increased stress on the femoral neck. In the case of coxa vara, an inclination angle less than 125° is associated with decreased limb length, increased hip abductor effectiveness, decreased load on the femoral head, and decreased stress on the femoral neck. The femoral neck forms an acute angle with the lateral axis of the femoral condyle. Because this angle points medially and anteriorly, it is called the anteversion angle. In adults, this angle averages approximately 7.5°.

[0095] The acetabulum is located on the lateral surface of the hip joint, where the ilium, ischium, and pubis meet. These three separate bones combine to form the acetabulum, with the ilium and ischium each contributing approximately two-fifths of the volume, and the pubis contributing one-fifth. The acetabulum is not deep enough to cover the entire femoral head; it has both an articular and a non-articular portion. However, the acetabular labrum deepens the fossa and provides additional stability. Together with the acetabular labrum, the acetabulum covers slightly more than 50% of the femoral head. Only the lateral surface of the acetabulum is lined with articular cartilage, and a deep acetabular notch extends below it. The center of the acetabular cavity is deeper than the articular cartilage and is the non-articular portion. This center, called the acetabular fossa, is separated from the pelvic bone interface by a thin plate. The acetabular fossa is a unique area for each patient and is used to create a patient-specific guide for drilling and positioning acetabular cup components. Additionally, variations in anatomical characteristics further necessitate population-specific implant design.

[0096] Some of the problems associated with the use of prior art cementless components can be attributed to the wide variety of femoral canal sizes, shapes, and orientations. One of the challenges in femoral stem orthopedic implant design is the large variability in medial-lateral and anterior-posterior dimensions. The ratio of proximal to distal canal size also varies widely. The number of different combinations of arcs, tapers, curves, and offsets in the normal population is vast. However, this is not the only problem.

[0097] Systematic variations in femoral morphology and the lack of clear standards for modern populations pose challenges to the design of appropriate hip implant systems. For example, significant differences exist in lordosis, torsion, and cross-sectional shape between Native Americans, African Americans, and Caucasians. Femoral differences between Asian and Western populations are evident in femoral lordosis, with Chinese femora exhibiting greater anterior curvature and external rotation, smaller intramedullary canals, and smaller distal femoral condyles compared with Caucasians. Similarly, Caucasian femora are larger than Japanese femora in terms of the length dimension of the distal femoral condyles. Ethnic differences also exist in proximal femoral bone mineral density (BMD) and hip shaft length between African Americans and Caucasians. The combined effects of higher BMD, shorter hip shaft length, and shorter intertrochanteric width may explain the lower prevalence of osteoporotic fractures in African American women compared with Caucasian women. Similarly, older Asian and black American men have been found to have thicker cortices and higher BMD than white and Hispanic men, which may contribute to increased bone strength in these ethnic groups. Black Americans generally have thicker bone cortices, narrower endosteal diameters, and higher BMD than white Americans.

[0098] Combined with systematic (ethnic) variations in the femur and pelvis, the primary hip system becomes even more challenging, making revision surgery more complex. In addition to these normal anatomical and ethnic variations, the challenges facing the revision surgeon include (a) distortion of the femoral canal due to osteoporosis around the originally placed prosthesis and (b) iatrogenic defects due to removal of components and cement.

[0099] All of these factors have led many hip surgeons to seek ways to improve the design of uncemented femoral prostheses. In total hip replacement (primary or revision), the ideal is to establish an optimal fit between the femoral ball and the acetabular cup. The neck of the femoral stem should have a cruciform cross-section to reduce stiffness. The stem length should ensure parallel contact with the femoral wall over two to three canal diameters. The proximal third of the stem should be porous or hydroxyapatite (HA) coated. The stem should also be cylindrical (i.e., not tapered) to accommodate bending loads and allow for the proximal transfer of all rotational and axial loads. The femoral head position should mimic the patient's own head center in the absence of abnormalities.

[0100] One attempt to meet these goals is to manufacture femoral prostheses individually for each patient -- in other words, to create a prosthesis that is unique to a particular patient, rather than attempting to reshape the patient's bone to fit an off-the-shelf prosthesis.

[0101] There are several common design criteria for patient-specific (or mass-customized) primary and revision hip replacements. Among these design criteria are: (1) a hip stem without a collar (except in revision cases) to allow uniform load distribution to the femur; (2) a hip stem with a modified rhomboid cross-section to maximize fit / fill while maintaining rotational stability; (3) a hip stem that can be curved as needed to conform to the patient's bone; (4) a hip stem that can be inserted along a curved path with no gap between the prosthesis and the bone; (5) a hip stem neck with a cruciform cross-section to reduce stiffness; (6) a stem length that allows parallel contact with the femoral wall over two to three canal diameters; (7) a porous or hydroxyapatite (HA) coating on the proximal third of the hip stem; (8) a cylindrical (i.e., not tapered) hip stem to accommodate bending loads and allow proximal transfer of all rotational and axial loads; and (9) a hip stem with a femoral head that mimics the patient's own femoral head center when not abnormal.

[0102] The following is an exemplary process and system for producing mass-customized orthopedic implants for patients requiring primary joint replacement, taking into account the gender and / or ethnicity of the patient population. For purposes of illustrative description, a total hip arthroplasty for a patient with incomplete anatomy will be described. However, it will be understood that the exemplary process and system are applicable to any orthopedic implant suitable for mass customization in cases where incomplete anatomy is present. For example, the exemplary process and system are applicable to shoulder and knee replacements where bone degeneration (incomplete anatomy), bone deformity, or comminuted fractures are present. As a result, while the femoral component of a hip implant is discussed below, those skilled in the art will understand that the system and process are applicable to other orthopedic implants, guides, instruments, etc., used in primary or revision orthopedic surgical procedures.

[0103] FIG. 29 illustrates the overall process flow for generating both mass-customized and patient-specific hip implants using a statistical atlas. The process begins with a statistical atlas containing multiple instances of one or more bones being analyzed. In the exemplary context of a hip implant, the statistical atlas includes multiple instances of bone models for the pelvic bone and femur. An articular surface topography analysis is performed for at least the acetabular component (i.e., the acetabulum) and the proximal femoral component (i.e., the femoral head). Specifically, the articular surface topography analysis involves calculating landmarks, measurements, and shape features for each bone in a given population in the statistical atlas. The articular surface topography analysis also includes generating quantitative values, such as statistics, that represent the calculation results. From these calculation results, a distribution of the calculation results is plotted, and analysis is performed based on the distribution. For example, a bell-shaped distribution indicates that approximately 90% of the population falls into this grouping, and designing a non-patient-specific implant (e.g., a mass-customized implant) to properly fit this grouping reduces patient costs compared to a patient-specific implant. For the other 10% of the population, patient-specific implants may be a better approach.

[0104] In the context of mass-customized implants, statistical atlases may be used to quantitatively assess the number of different groups (i.e., different implants) that can cover the vast majority of a given population. These quantitative assessments may result in data clusters that indicate general parameters of a base implant design that, while not patient-specific, is more specific than commercially available alternatives.

[0105] In the context of patient-specific implants, statistical atlases may be used to quantitatively assess what normal bone represents and the differences between the patient's bone and normal bone. More specifically, statistical atlases may include curvature data associated with an average or template bone model. This template bone model can then be used to extrapolate the desired morphology of the patient's correct bone and to create the implants and surgical instruments used to perform the implant surgery.

[0106] Figure 30 provides a pictorial summary of the use of statistical atlases in the design of mass-customized and patient-specific hip implants. The context of the Implant box refers back to Figures 20 and 21 and their associated descriptions. Similarly, the context of the Planner box refers back to Figure 20 and the associated descriptions of the Custom Planning Interface. Finally, the context of the Patient-Specific Guide box refers back to Figure 22 and its associated descriptions.

[0107] FIG. 31 shows a flowchart of an exemplary process that can be used to design and manufacture gender- and / or ethnicity-specific hip implants. Specifically, this process involves utilizing a statistical atlas containing various samples of proximal femora (i.e., femurs including the femoral head) with associated data identifying the gender and ethnicity of the individuals to whom the bones belong. Additionally, the statistical atlas module records virtual 3D models of one or more anatomical structures (e.g., bones) to capture the unique anatomical variations in a given gender and / or ethnic population. In an exemplary embodiment, the atlas records mathematical representations of anatomical features of one or more anatomical structures expressed as average representations and variations around the average representations for a given anatomical population that may share a common gender and / or ethnicity (or be grouped as having one of multiple ethnicities with anatomical commonalities). See FIG. 2 and the above discussion of the statistical atlas and methods for adding anatomical structures to a statistical atlas for a given population. The output of the statistical atlas is directed to an auto-labeling module and a surface / shape analysis module.

[0108] 31-43, the auto-labeling module utilizes input from the statistical atlas (e.g., regions likely to contain particular landmarks) and local geometric analysis to calculate anatomical landmarks for each anatomy instance in the statistical atlas. As an example, for each 3D virtual model of a femur, various landmarks of the proximal femur are calculated. The landmarks include, but are not limited to, (1) the femoral head center, which is the center point of the femoral head approximated by a sphere; (2) the greater trochanter point, which is the point on the greater trochanter that is closest to a plane passing through the neck diaphysis point perpendicular to the anatomical neck centerline; (3) the osteotomy point, which is a point 15 mm from the end of the lesser trochanter (approximately 30 mm from the lesser trochanter point); (4) the neck diaphysis point, which is a point on the femoral head sphere whose tangent plane encloses the smallest femoral neck cross-sectional area; (5) the femoral neck fulcrum, which is the smallest cross-section along the femoral shaft; (6) the intramedullary canal fascicle, which is the smallest cross-section along the intramedullary canal; (7) the femoral neck fulcrum, which is a point on the femoral anatomical axis that forms an angle equal to the femoral neck angle with the femoral head center and the distal end of the femoral anatomical axis; and (8) the lesser trochanter point, which is the point on the most outwardly protruding lesser trochanter area. As another example, various axes of the proximal femur are calculated for each 3D virtual model of the femur using the identified anatomical landmarks, including, but not limited to, (a) a femoral neck anatomical axis coaxial with a line connecting the femoral head center with the femoral neck center, (b) a femoral neck axis coaxial with a line connecting the femoral head center point with the femoral neck fulcrum, and (c) a femoral anatomical axis coaxial with a line connecting two points located at 23% and 40% of the total length of the femur starting from the proximal end of the femur. As yet another example, various measurements of the proximal femur are calculated for each 3D virtual model of the femur using the identified anatomical landmarks and axes.The measurements included: (i) proximal angle, which is the 3D angle between the femoral anatomical axis and the femoral neck anatomical axis; (ii) head offset, which is the horizontal distance between the femoral anatomical axis and the femoral head center; (iii) head height, which is the vertical distance between the lesser trochanter point (mentioned above) and the femoral head center; (iv) greater trochanter-head center distance, which is the distance between the femoral head center and the greater trochanter point (mentioned above); (v) neck length, which is the distance between the femoral head center and the neck fulcrum (mentioned above); and (vi) femoral (vii) femoral head radius, which is the radius of a sphere fitted to the femoral head; (vii) femoral neck diameter, which is the diameter of a circle fitted to a neck cross section in a plane perpendicular to the femoral neck anatomical axis and passing through the neck center point (mentioned above); (viii) femoral neck transepicondylar anteversion angle, which is the angle between the transepicondylar axis and the femoral neck axis; (ix) femoral neck posterior condylar anteversion angle, which is the angle between the posterior condylar axis and the femoral neck axis; (x) LPFA, which is the angle between the functional axis and the vector pointing to the greater trochanter; (xi) (xii) the calcar index area, defined by the formula (ZX) / Z (Z is the femur area 10 cm below the midpoint of the lesser trochanter, and X is the intramedullary canal area 10 cm below the midpoint of the lesser trochanter); (xiii) the intramedullary canal area 3 cm below the midpoint of the lesser trochanter to the intramedullary canal area 10 cm below the midpoint of the lesser trochanter; and (xiv) the intramedullary canal area 3 cm below the midpoint of the lesser trochanter to the intramedullary canal area 10 cm below the midpoint of the lesser trochanter. (xiv) the minor axis / major axis ratio, which is the ratio between the minor and major axes of the fitting ellipsoid and the intramedullary canal cross section at the narrowest point on the intramedullary canal; and (xv) the femoral radius / intramedullary canal radius ratio, which is the ratio between the femoral circumference and the radius of a circle best-fitted around the intramedullary canal in a plane perpendicular to the femoral anatomical axis (this ratio reflects the thickness of the cortical bone and, therefore, the cortical bone loss in the case of osteoporosis).

[0109] 31 and 45-47, the output of the auto-labeling module is used to evaluate femoral stem parameters for a given population, specifically the medial contour, neck angle, and head offset, regardless of whether the population is grouped based on ethnicity, gender, or a combination of the two.

[0110] The medial contour is generated for each femur in the population by intersecting the intramedullary canal with a plane having a perpendicular axis (vector cross product) that extends through the femoral fulcrum and is perpendicular to both the femoral anatomical axis and the femoral neck axis. After generating the contours for each femur in the population, the population is subdivided into groups based on the size of the intramedullary canal. Because the contours may be out of plane during the subdivision, an alignment process is performed to align all contours with respect to a common plane (e.g., the XZ plane). The alignment process involves aligning the axis perpendicular to both the femoral neck axis and the anatomical axis with the Y axis, and then aligning the anatomical axis with the Z axis. In this way, all contours are moved relative to a specific point so that they have a common coordinate system.

[0111] Once the contours have a common coordinate system, femoral neck points are used to ensure that the contour points are in-plane. In particular, femoral neck points are matching points that reflect the actual anatomy and ensure that the contour points are in-plane. Ensuring that the contour points are in-plane greatly reduces alignment variability between femurs in a population, facilitating the use of the contours in the design of head offset and implant angle.

[0112] Referring to FIG. 48, statistical atlases are also useful for interpolating between normal and osteoporotic bone. One important consideration in the design and sizing of femoral stems is the size of the intramedullary canal. In the case of normal bone, the femur exhibits a significantly narrowed intramedullary canal compared to femora exhibiting osteoporosis. This narrowed size of the intramedullary canal is at least in part the result of a reduction in bone thickness (measured perpendicular to the major axis of the femur) and a corresponding recession of the femoral interior surface that outlines the intramedullary canal. In this method, a synthetic population is created by interpolating thicknesses between healthy and severely osteoporotic bones and generating a virtual 3D model with said thicknesses. This dataset therefore includes bones corresponding to different stages of osteoporosis. Hereinafter, this dataset can be used as input for implant stem design.

[0113] In an exemplary embodiment, the statistical atlas includes populations of normal, non-osteoporotic bones and osteoporotic bones, in this case femurs. Each of these normal femurs in the atlas is quantified and displayed as a 3D virtual model according to the process described herein for adding bones to a statistical atlas. Similarly, each of the osteoporotic femurs in the atlas is quantified and displayed as a 3D virtual model according to the process described herein for adding bones to a statistical atlas. As part of the 3D models of normal and osteoporotic bones, the dimensions of the intramedullary canal along the longitudinal length of the femur are recorded. Using atlas point correspondence, the intramedullary canal is identified on the atlas bone as spanning a percentage of the bone's total length proximal to the lesser trochanter (approximately 5%) and a second percentage of the bone's total length proximal to the distal cortical point (approximately 2%). Additionally, points on the bone's outer surface that fall within these proximal and distal boundaries are used to determine bone thickness, defined as the distance from the outer point to the nearest point on the IM canal.

[0114] Within the context of the proximal femur, Figures 51-62 confirm the existence of gender-based differences across ethnic populations. As shown in Figures 59 and 60, the statistical atlas template 3D model of the female proximal femur demonstrates statistically significant measurements when compared to the template 3D model of the male proximal femur. In particular, femoral head offset is approximately 9.3% less in women than in men. While current implants increase femoral head offset with stem size, this is acceptable for normal women. However, challenges arise when capturing femoral head offset in osteoporotic and osteopenic cases, where osteoporosis increases intramedullary canal size (meaning larger stem size and offset). Similarly, neck diameter and femoral head radius are approximately 11.2% smaller in women than in men. Additionally, neck length is approximately 9.5% shorter in women than in men. Additionally, proximal angle is approximately 0.2% smaller in women than in men. Finally, femoral head height is approximately 13.3% lower in women than in men. As a result, the sex-specific bone data confirm that simple scaling of generic femoral implants (i.e., sex-neutral) does not capture differences in bone shape, and therefore, sex-specific femoral implants are necessary.

[0115] Referring to Figures 63-68, not only are the dimensions of the proximal femur different, but the cross-sectional shape of the femur along the length of the intramedullary canal also differs significantly between genders. In particular, across a given population within a statistical atlas of male and female femurs, the intramedullary canal cross-section in men is more circular than in women. More specifically, the intramedullary canal cross-section is 8.98% more eccentric in women than in men. As discussed in more detail below, this gender-specific data includes portions of the feature-extracted data that are plotted to arrive at clusters from which quantity and overall shape parameters are extracted to arrive at gender-specific femoral implants.

[0116] As shown in Figures 72-74, the statistical atlas includes calculations corresponding to measurements of head center offset in the anterior-posterior (AP) direction across a given femur population (separated by gender). In the exemplary configuration, the AP direction was determined by a vector pointing anteriorly, perpendicular to both the functional axis and the posterior condylar axis. The offset was measured between two reference points: the first reference point, which was the femoral head center and the midpoint of the anatomical axis, and the second reference point, which was the femoral neck fulcrum. In summary, AP head height relative to the neck fulcrum and the anatomical axis midpoint did not show significant differences between femurs from males and females. Again, this gender-specific data includes portions of the feature-extracted data plotted to arrive at clusters from which quantity and overall shape parameters were extracted to arrive at gender-specific femoral implants.

[0117] 28 and 31, the femoral head center offset, intramedullary canal cross-sectional shape data, and femoral medial contour data within the statistical atlas population include portions of the extracted feature data plotted to identify the number of clusters present across a given population (one gender-specific and a second ethnicity-specific, assuming the statistical atlas includes data regarding ethnicity associated with each bone) for the design of a gender- and / or ethnicity-specific mass-customized implant consistent with the flowchart and associated description of FIG. 28. The gender- and / or ethnicity-specific identified clusters are utilized to extract parameters necessary for the design of a mass-customized femoral implant.

[0118] 76 illustrates an exemplary mass-customized femoral component according to the present disclosure. Specifically, the mass-customized femoral component includes four major elements, including a ball, a neck, a proximal stem, and a distal stem. Each of these major elements includes an interchangeable interface that allows the ball, neck, and stem to be interchangeable with other interchangeable elements. In this manner, if a larger femoral ball is required, only the femoral ball is replaced.

[0119] Similarly, if a larger cervical offset is desired, a different cervical element is substituted that provides the required offset, while maintaining the other three elements as needed. In this way, the femoral component can be customized to fit the patient, within certain limits, without necessarily sacrificing the fit and kinematics that would be compromised if a one-size-fits-all implant were used. Thus, all femoral elements can be exchanged for other mass-customized elements to better suit the patient's anatomy.

[0120] In this exemplary embodiment, the neck is configured to rotate about the axis of the proximal stem so that its rotational orientation relative to the proximal stem can be adjusted intraoperatively. Notably, preoperative measurements may establish a planned rotational position of the neck relative to the proximal stem. Even so, intraoperative studies, such as in vivo kinematic testing, may lead the surgeon to alter the preoperative rotational orientation to improve kinematics or avoid certain impingements. In one example, the neck includes a cylindrical stud with a circumferential groove having a concave-convex surface. The cylindrical stud is received within an axial cylindrical channel in the proximal stem. A second channel intersects the cylindrical channel and is also concave-convex, shaped to receive a plate with a semicircular groove configured to engage the concave-convex surface of the circumferential groove. A pair of screws secured to the proximal stem press the plate into engagement with the cylindrical stud, ultimately disabling rotational movement of the cylindrical stud relative to the proximal stem. Thus, once this locking engagement is reached, the screw may be loosened to permit rotational movement between the cylindrical stud and the proximal stem as may be required for intraoperative rotational adjustment.

[0121] While the engagement between the neck and ball is considered conventional, the engagement between the proximal stem and distal stem is novel. In particular, the proximal stem includes distal legs that are threadedly received within threaded openings extending through the distal stem by threading and engagement. The proximal stem is then attached to the distal stem by rotating it relative to the distal stem so that the threads of the legs engage with the threads of the openings in the distal stem. Rotation of the proximal stem relative to the distal stem is complete when the proximal stem abuts the distal stem. However, if rotational adjustment is required between the proximal and distal stems, a washer may be utilized to provide a spacer that corresponds to the correct rotational adjustment. As another example, if greater rotational adjustment is required, the washer thickness may be increased; on the other hand, a thinner washer will correspondingly provide less rotational adjustment.

[0122] Each major element may be manufactured as a predetermined alternative that captures size and contour variations for a given gender and / or ethnicity. In this manner, the major element alternatives may be fused or fitted to mimic a patient-specific implant that more closely resembles the patient's anatomy than traditional mass-customized femoral components at a fraction of the cost and process used to create a patient-specific femoral implant.

[0123] 77 illustrates yet another exemplary mass-customized femoral component according to the present disclosure. Specifically, the mass-customized femoral component includes five major elements, including a ball, a neck, a proximal stem, a mid-stem, and a distal stem. Each of these major elements includes an interchangeable interface that allows the ball, neck, and stem to be interchangeable with other interchangeable elements. Those skilled in the art will appreciate that by replicating a patient's natural femur using stacked slices of the bone, the number of elements in the mass-customized femoral component that are akin to this bone can be increased to more closely approximate the fitting of a patient-specific implant using the mass-customized elements.

[0124] Similar to the anatomical differences between sexes and ethnicities in the proximal femur, Figures 78-83 confirm the existence of gender and ethnicity differences across general pelvic populations within the statistical atlas. Referring again to Figure 28, a series of mass-customized acetabular cup implants are designed and manufactured using statistical atlas data (i.e., pelvic populations) grouped based on gender and / or ethnicity. The grouped atlas data is subjected to automated labeling and surface / shape analysis processes to segment the acetabular cup shapes within the populations, as shown graphically in Figure 78. Additionally, feature extraction processes (e.g., labeling of the acetabular ligament location) and contour analysis (evaluating the acetabular cup contour) are performed, as shown graphically in Figures 82 and 83, ultimately generating an anatomical cup implant surface, as shown in Figure 79. This analysis reveals that the acetabular cup and femoral head do not have a single radius of curvature, but rather multiple radii, as shown in Figures 80 and 81.

[0125] Creating animal-specific implants Referring to FIG. 85, an exemplary system and method for designing and manufacturing animal-specific (i.e., animal patient-specific) implants and associated instrumentation is similar to the process shown and described above with respect to FIG. 20 and is incorporated herein by reference. As a prelude, images of the animal's anatomy are acquired and automatically segmented to generate a virtual 3D bone model. While graphically displayed as a CT scan image, it is understood that other imaging modalities besides CT, such as, but not limited to, MRI, ultrasound, and X-ray, may be utilized. The virtual 3D bone model of the diseased anatomy is then loaded into a statistical atlas in accordance with the exemplary disclosure above. Input from the statistical atlas is then utilized to reconstruct the bone and create the reconstructed virtual 3D bone model. Also, bony landmarks are calculated on the surface of the reconstructed virtual 3D bone model to enable proper implant sizing. The shape of the diseased bone is then mapped and converted to parametric form and used to create an animal-specific implant that mimics the remaining anatomy. The animal-specific implant, as well as animal-specific instrumentation, are manufactured and utilized to create the animal's remaining bone and place the animal-specific implant.

[0126] Referring to FIG. 86, an exemplary system and method for designing and manufacturing mass-customized animal implants is similar to the process shown and described above with respect to FIG. 28, which is incorporated herein by reference. As a prelude, automatic labeling and surface / shape analysis are applied to 3D animal bone models from a statistical atlas for the bones in question. The automatic labeling process automatically calculates anatomical landmarks for each 3D animal bone model using information stored in the atlas (e.g., regions likely to contain specific landmarks) and local geometric analysis. For each animal bone in question in the statistical atlas, shape / surface analysis directly extracts surface shape features of the 3D virtual animal bone model. Each 3D animal bone model is then subjected to a feature extraction process that uses a combination of landmark and shape features to calculate features relevant to implant design. These features are used as input for a clustering process, and a predetermined clustering method is used to divide the animal bone population into groups with similar features. Each resulting cluster represents a case that will be used to define the shape and size of a single animal implant. For each cluster center (implant size), a subsequent parameterization process extracts parameters (e.g., computer-aided design (CAD) parameters) for the overall implant model. The extracted parameters are then used to generate a global implant surface and size for each cluster, and a mass-customized implant is selected from the required group depending on the cluster the animal patient falls into.

[0127] Creation of patient-specific cutting guides 87-102, an exemplary process and system for integrating multidimensional medical imaging, computer-aided design (CAD), and computer graphics capabilities to design a patient-specific cutting guide is described. For purposes of exemplary illustration only, the patient-specific cutting guide is described in the context of total hip arthroplasty. Nevertheless, one skilled in the art will recognize that the exemplary process and system are applicable to any surgical procedure for which a cutting guide is available.

[0128] As shown in FIG. 87, an overview of an exemplary system flow begins with receiving input data representative of an anatomical structure. The input anatomical data may include two-dimensional (2D) images or three-dimensional (3D) surface representations of the anatomical structure of interest, and may be in the form of a surface model or point cloud, for example. In situations where 2D images are used, these 2D images are used to construct a 3D surface representation of the anatomical structure of interest. Those skilled in the art are familiar with using 2D images of anatomical structures to construct a 3D surface representation. Therefore, a detailed description of this process is omitted for the sake of brevity. By way of example, the input anatomical data may include one or more of X-rays (taken from at least two perspectives), computed tomography (CT), magnetic resonance imaging (MRI), or any other imaging data capable of generating a 3D surface representation. In an exemplary embodiment, the anatomical structure includes a pelvis and a femur.

[0129] It should be understood, however, that the following is an exemplary description of anatomical structures that can be used in conjunction with the exemplary system and is not intended to limit the use of the system with other anatomical structures in any way. As used herein, tissue includes bone, muscle, ligament, tendon, and any other finite type of structural material with a specific function in a multicellular organism. As a result, when discussing the exemplary systems and methods in the context of bones associated with the hip joint, those skilled in the art will recognize the applicability of the systems and methods to other tissues.

[0130] The system's input anatomical data for the femur and pelvis goes to one of two modules depending on the type of input data. In the case of X-ray data, 2D X-ray images are input to the non-rigid module, which extracts 3D bone contours. If the input data is in the form of CT scans or MRI images, these scans / images go to the auto-segmentation module, where 3D bone contours (and 3D cartilage contours) are extracted through auto-segmentation.

[0131] Referring to FIG. 88, the non-rigid module applies one or more pre-processing steps using multiple x-ray images acquired from at least two different viewpoints. These steps may include one or more of noise reduction and image processing. A calibration step is applied to the resulting pre-processed x-ray images to align the x-ray images. Preferably, the x-ray images are acquired in the presence of a fixed position calibration device and are aligned to this fixed position calibration device. However, in the absence of a fixed position calibration device, the x-ray images may be calibrated using common detected features across the images. The output of this calibration process is the position of the anatomy relative to the imaging device, identified in FIG. 88 by the reference symbol "position."

[0132] The resulting pre-processed X-ray image is subjected to a feature extraction step. This feature extraction step includes one or more image feature operations utilizing the pre-processed X-ray image. By way of example, these operations may include grayscale features, contours, texture components, or any other image-derived features. In this exemplary process, the feature extraction step outputs the contours of the anatomical structures (e.g., bone shapes) referenced "Contour" in FIG. 88 as well as image features referenced "Texture" as derived from the X-ray image. Both the contours of the anatomical structures and the image feature data are then fed into the non-rigid registration step.

[0133] In the non-rigid registration step, the output of the feature extraction and calibration steps is aligned with a 3D template model of the anatomical structure of interest from a statistical atlas. As an example, the 3D template model is generated in response to non-linear principal components from an anatomical database that includes part of the statistical atlas. In the non-rigid registration step, the 3D template model is adapted to the shape parameters of the X-ray image by optimizing its shape parameters (non-linear principal components) with the position, contour, and texture data. The output of the non-rigid registration step is a patient-specific 3D bone model, similar to the patient-specific 3D bone model output from the automatic segmentation module for CT scans or MRI images, and is sent to the virtual templating module.

[0134] Referring to FIG. 91, the automatic segmentation process is initiated by, for example, acquiring a CT scan or MRI image and running an automatic segmentation sequence. With specific reference to FIG. 90, the automatic segmentation sequence includes aligning the scan / image to a reference or 3D model of the anatomical structure of interest. After the scan / image is aligned to the reference 3D model, it is processed through an initial deformation process to calculate normal vectors, locate profile points, linearly interpolate intensity values, filter the resulting profile using a Savitsky-Golay filter, generate profile gradients, weight the profile using a Gaussian weight profile formula, determine maximum profiles, and deform the reference 3D model using these maximum profiles. The resulting deformed 3D model is projected onto a template 3D model from a statistical atlas for the anatomical structure of interest. The parameters of the template 3D model are used to further deform the deformed 3D model in a secondary deformation process to identify unique features in the template 3D model. After this latter deformation process, the deformed 3D model is compared to the scan / image to identify whether significant differences exist.

[0135] In situations where significant differences exist between the deformed 3D model and the scan / image, the deformed 3D model and scan / image are subjected to the initial deformation process again, followed by a secondary deformation process, and this looped process continues until the deformed 3D model is within a predetermined tolerance for the differences between the deformed 3D model and the scan / image.

[0136] After determining that the deformed 3D model does not show significant differences with respect to previous iterations or achieving the maximum number of iterations, the surface edges of the deformed 3D model are smoothed, followed by a high-resolution remeshing step to further smooth the surface to create a smooth 3D model, to which an initial deformation sequence (same as the initial deformation process described above prior to surface smoothing) is applied to generate a 3D segmented bone model.

[0137] Referring again to Figure 91, the 3D segmented bone model is processed to generate contours. In particular, the intersection of the 3D segmented bone model with the scan / image is calculated, which results in a binary contour at each image / scan plane.

[0138] The 3D segmented bone model is also processed to generate a patient-specific statistical 3D model of bone appearance, specifically modeling the appearance of the bone and any anatomical abnormalities based on image information present within and outside the contour.

[0139] The segmentation system user then reviews the bone contours. This user may be a segmentation expert or a casual user who notices one or more regions of the 3D model that do not correlate with the segmentation regions. This lack of correlation may exist in the context of missing or obviously inaccurate regions. Upon identifying one or more incorrect regions, the user may select a "seed point" on the model that represents the center of the area where the incorrect region exists. Alternatively, the user may manually outline the missing region. The system software uses the seed point to add or subtract contours near the seed point using an initial scan / image of the anatomy from CT or MRI. For example, the user may select an area where a bone spur should be present, and the software compares the scan / image with the area on the 3D model and adds the bone spur to the segmentation sequence. Any changes to the 3D model are ultimately reviewed by the user for confirmation or cancellation. This review and revision sequence may be repeated as many times as necessary to capture anatomical differences between the scan / image and the 3D model. Once the user is satisfied with the 3D model, they may manually manipulate the resulting model to remove bridges and repair areas of the model as needed prior to output to the virtual templating module.

[0140] As shown in Figures 87 and 92, the virtual templating module receives a patient-specific 3D model from one or both of the automatic segmentation and non-rigid registration modules. In the context of the hip joint, the patient-specific 3D model includes the pelvis and femur, both of which serve as inputs for the automatic labeling process. This automatic labeling step calculates anatomical landmarks for implant placement on the 3D models of the femur and pelvis by region and local shape search from similar anatomical structures present in a statistical atlas.

[0141] As shown in Figure 93, in the context of automated placement of a femoral stem with distal fixation, automated labeling involves defining axes on the femur and implant. For the femur, the anatomical femoral axis (AFA) is calculated, followed by the proximal anatomical axis (PAA). The proximal neck angle (PNA) is then calculated, which is defined as the angle between the AFA and PNA. For the femoral implant, the implant axis is along the length of the implant stem, and the implant neck axis is along the length of the implant neck. Similar to the femoral PNA, the implant angle is defined as the angle between the implant axis and the implant neck axis. The implant with the implant angle closest to the PNA is then selected. The implant fitting angle (IFA) is then defined as the intersection of the anatomical proximal axis with a vector drawn from the femoral head center at the selected implant angle.

[0142] As shown in Figure 93, when using automatic femoral stem placement with distal fixation and calculated anatomical landmarks, the implant size assignment step determines / estimates the appropriate implant size for the femoral component. The implant size is selected by comparing the implant width with the intramedullary canal width and selecting the implant with the width closest to the intramedullary canal. The system then proceeds to the implant placement step.

[0143] During the implant placement step for a distally fixed femoral stem, the initial implant positions of all relevant implant components are determined / selected based on the surgeon's preferred surgical technique and previously calculated anatomical landmarks. A resection plane is then created to simulate the osteotomy of the proximal femur, and the implant fit is evaluated. The fit is evaluated by analyzing the cross sections of the aligned implant and femoral intramedullary canal at different levels along the implant axis. The implant is aligned with the femur by aligning the implant axis with the femoral anatomical axis and then translating the implant so that its neck is approximately at the neck of the proximal femur. The implant is then rotated around the femoral anatomical axis to achieve the desired anteversion.

[0144] As part of this implant placement step, the placement of the "educated guess" is evaluated using an iterative scheme that includes using an initial "educated guess" for implant placement as part of a kinematic simulation. In an exemplary embodiment, the kinematic simulation uses estimated or measured joint kinematics to move the implant (based on the selected implant placement) through a wide range of motion. The kinematic simulation may then be used to determine the impact location and estimate the resulting range of motion of the implant after impact. If the kinematic simulation results in unsatisfactory data (e.g., unsatisfactory range of motion, unsatisfactory imitation of natural motion, etc.), alternative implant placement locations may be used to perform kinematic analysis to further refine the implant placement until a satisfactory result is reached. After the implant locations for all relevant implant components have been determined / selected, the template data is sent to a jig generation module.

[0145] As shown in Figure 94, in the context of automatic placement of a femoral stem with a press fit and three contact points, automatic labeling involves defining axes on the femur and implant. For the femur, the anatomical femoral axis (AFA) is calculated, followed by the anatomical proximal axis (PAA). The proximal neck angle (PNA) is then calculated, which is defined as the angle between the AFA and PNA. For the femoral implant, the implant axis is along the length of the implant stem, and the implant neck axis is along the length of the implant neck. Similar to the femoral PNA, the implant angle is defined as the angle between the implant axis and the implant neck axis. The implant with the implant angle closest to the PNA is then selected. The implant fitting angle (IFA) is then defined as the intersection of the anatomical proximal axis with a vector drawn from the center of the femoral head at the selected implant angle.

[0146] As shown in FIG. 94, when using automatic femoral stem placement using press-fit, three contact points, and calculated anatomical landmarks, the implant sizing step determines / estimates the appropriate implant size for the pelvic and femoral components. The implant size is selected by aligning the implant with the femur by aligning the implant axis with the femoral anatomical axis. The implant is then rotated to align its neck axis with the femoral neck axis. The implant is then translated to the anatomically correct position on the proximal femur. The system then proceeds to the implant placement step.

[0147] During the implant placement step of a press-fit femoral stem, the initial implant positions of all relevant implant components are determined / selected based on the surgeon's preferred surgical technique and previously calculated anatomical landmarks. A resection plane is also created to simulate a proximal femoral osteotomy and evaluate the implant fit. The fit is evaluated by analyzing the contours of the implant and femoral intramedullary canal. The contour is created by intersecting the intramedullary canal with a plane perpendicular to both the anatomical axis and the femoral neck axis and passing through the intersection of the anatomical axis and the femoral neck axis. When generating the implant and intramedullary canal contour, only implants with widths smaller than the intramedullary canal width at the same location are retained to obtain a large number of correct implant size candidates. The group of size candidates is then narrowed using two strategies to reduce the mean squared distance error between the implant and the intramedullary canal. The first strategy minimizes the mean squared error (MSE) or other mathematical error criterion of the distance between the implant and the intramedullary canal. The second strategy minimizes the MSE of the distance between the implant and the outside of the intramedullary canal.

[0148] As part of this implant placement step, the placement of the "educated guess" is evaluated using an iterative scheme that includes using an initial "educated guess" for implant placement as part of a kinematic simulation. In an exemplary embodiment, the kinematic simulation uses estimated or measured joint kinematics to move the implant (based on the selected implant placement) through a wide range of motion. The kinematic simulation may then be used to determine the impact location and estimate the resulting range of motion of the implant after impact. If the kinematic simulation results in unsatisfactory data (e.g., unsatisfactory range of motion, unsatisfactory imitation of natural motion, etc.), alternative implant placement locations may be used to perform kinematic analysis to further refine the implant placement until a satisfactory result is reached. After the implant locations for all relevant implant components have been determined / selected, the template data is sent to a jig generation module.

[0149] Referring again to FIG. 87 , the jig generation module generates a patient-specific guide model. More specifically, from the template data and associated planning parameters, the shape and placement of the patient-specific implant relative to the patient's remaining bone are known. As a result, the virtual templating module uses the patient-specific 3D bone model to calculate the position of the implant relative to the patient's remaining bone, thereby providing the jig generation module with information regarding the intended degree of retention of the patient's remaining bone. Consistent with this bone retention data, the jig generation module utilizes the bone retention data to apply one or more osteotomies to reduce the patient's current bone to the remaining bone required to accept the implant as planned. Using the intended osteotomies, the jig generation module generates a virtual 3D model of a cutting guide / jig having a shape configured to mate with the patient's bone in a single position and orientation. In other words, the 3D model of the cutting jig is created as a "negative" of the anatomical surface of the patient's remaining bone so that the tangible cutting guide precisely fits against the patient's anatomy. In this way, any guesswork associated with positioning the cutting jig is eliminated. The jig generation module generates a virtual 3D model of the cutting jig and then outputs the necessary machine code for a rapid prototyping machine, CNC machine, or similar equipment to manufacture a tangible cutting guide. As an example, an exemplary cutting jig for resecting the femoral head and neck includes a hollow slot that holds the cutting blade in a predetermined orientation that replicates the virtual cut from the surgical planning and templating module, along with associated guides that restrict the cutting blade within a certain range of motion. The jig generation module is also utilized to create a femoral stem placement jig.

[0150] Referring to FIG. 100 , after resection of the femoral head and neck, intramedullary drilling is performed, followed by femoral stem insertion. To prepare the femur for femoral implant insertion, the intramedullary canal must be drilled along an orientation that matches the orientation of the femoral implant. Offset drilling may compromise the orientation of the femoral implant. To address this issue, the jig generation module generates a virtual guide that is a “negative” of the anatomical surface of the patient's remaining or resected bone, allowing a rapid prototyping machine, CNC machine, or similar device to manufacture a cutting guide that precisely fits the patient's anatomy. As an example, a drilling jig may include an axial guide along which a reamer can travel. Using this drilling jig, the surgeon performing the drilling procedure can ensure that the drill is properly oriented.

[0151] The intramedullary canal may be adapted to receive a femoral stem. Again, the jig generation module generates a femoral stem placement guide to properly position the femoral stem in terms of rotation and angle in the intramedullary canal. By way of example, the femoral stem placement guide represents both a "negative" of the anatomical surface of the patient's remaining or resected bone and the top of the femoral stem. In this manner, the placement guide includes a unique shape that blends with the patient's remaining or resected bone to slide over the femoral shaft (the portion of the femoral stem that connects to the femoral ball) while allowing only one orientation of the femoral stem relative to the patient's femur, ensuring proper implantation of the femoral stem consistent with the preoperative plan. However, while the exemplary jig has been described in the context of a primary hip implant, those skilled in the art will appreciate that the exemplary process and system are not limited to primary hip implants, hip implants, or revision surgery. Instead, the above process and system is applicable to any hip implant, as well as surgeries involving other areas of the body, such as, but not limited to, the knee, ankle, shoulder, spine, head, and elbow.

[0152] As shown in FIG. 101 , in the context of an acetabulum, the jig generation module may generate instructions for manufacturing an acetabular cup drilling and acetabular implant placement guide. In particular, the shape and placement of the patient-specific acetabular implant relative to the patient's residual pelvis are known from the template data and associated planning parameters. Consequently, the virtual templating module uses the patient-specific 3D acetabulum model to calculate the size and position of the acetabular cup implant relative to the patient's residual bone, thereby providing the jig generation module with information regarding the intended degree of retention of the patient's residual pelvis and the desired implant orientation. Consistent with this bone retention data, the jig generation module utilizes the bone retention data to apply one or more osteotomies / drilling to reduce the patient's current pelvis to the residual bone required to accept the acetabular implant as planned. Using the intended osteotomies, the jig generation module generates a virtual 3D model of a cutting guide / jig having a shape configured to mate with two portions of the patient's pelvis in only one orientation. In other words, a 3D model of the cutting jig is created as a "negative" of the anatomical surface of the patient's pelvis so that the tangible drilling guide fits precisely to the patient's anatomy. In this way, any guesswork associated with positioning the drilling jig is eliminated. After generating the virtual 3D model of the drilling jig, the jig generation module outputs the necessary machine code for a rapid prototyping machine, CNC machine, or similar equipment to manufacture the tangible drilling jig. As an example, an exemplary acetabular component jig for drilling the acetabulum includes a four-piece structure, one configured to be received in the natural acetabulum and temporarily attached to the second component using the first component as a placement guide until the second component is secured to the pelvis. After the second component is secured to the pelvis, the first component may be removed. The third then includes a cylindrical or partially cylindrical element that is uniquely aligned with the second and allows the reamer to travel longitudinally relative to the third, but whose orientation is fixed by the combination of the first and third. After drilling, the reamer is removed and the third is removed from the first.A fourth part is then used to attach the acetabular cup implant to the drilled acetabulum. Specifically, the fourth part is uniquely shaped to engage with the first part in only one orientation while also being configured to be received within the acetabular cup implant. After the implant cup is positioned, both the first and fourth parts are removed. It should also be noted that additional jigs can be created to drill one or more holes in the pelvis for seating the acetabular implant, and each drilling jig can be attached in place of the first to confirm the orientation of the drill bit.

[0153] Surgical Navigation 103-111 illustrate another exemplary system and process for facilitating surgical navigation using one or more inertial measurement units (IMUs) to precisely position orthopedic surgical instruments and orthopedic implants during a surgical procedure. This first alternative exemplary embodiment is described in the context of performing a total hip arthroplasty. Again, the methods, systems, and processes described below are applicable to any other surgical procedure in which guidance of surgical instruments and implants is useful.

[0154] As shown schematically, the initial steps of utilizing patient images (whether x-ray, CT, MRI, etc.) and segmenting or registering to arrive at a virtual template of the patient's anatomy and appropriate implant size, shape, and placement are comparable to those described above with reference to Figures 87, 88, and 90-92. What differs slightly are the modules and processes utilized downstream of the virtual templating module.

[0155] Downstream of the virtual templating module is an initialization model generation module. This module receives the template data and associated planning parameters (i.e., the shape and placement of the patient-specific acetabular implant relative to the patient's residual pelvis are known, and the shape and placement of the patient-specific femoral implant relative to the patient's residual femur are known). Using this patient-specific information, the initialization model generation module produces a 3D virtual model of the initialization device relative to the patient's natural acetabular cup and a 3D virtual model of the initialization device relative to the femoral implant. In other words, the 3D model of the acetabular initialization device is created as a "negative" of the anatomical surface of the patient's acetabulum so that the tangible initialization device fits precisely to the patient's acetabulum. Similarly, the initialization model generation module creates a 3D model of the femoral stem initialization device as a "negative" of the anatomical surface of the patient's residual femur and femoral implant so that the tangible initialization device fits precisely to the patient's residual femur and femoral implant in only a single location and in a single orientation. In addition to generating these initialization devices, the initialization model generation module also generates the necessary machine code for a rapid prototyping machine, CNC machine, or similar device to manufacture the tangible acetabular and femoral initialization devices that are manufactured and attached to (or co-formed or integrally formed with) or integrated with a surgical navigation instrument configured with at least one IMU 1002.

[0156] IMUs 1002 capable of reporting orientation and movement data are combined with (e.g., attached to) surgical instruments to assist in surgical navigation, including positioning of surgical and implanted instruments. These IMUs 1002 are communicatively coupled (wired or wireless) to a software system that receives output data from the IMU indicating relative velocity and time, which allows calculation of its current position and orientation, or that calculates and transmits the position and orientation (discussed in more detail below) of a surgical instrument associated with the IMU. In this exemplary illustration, each IMU 1002 includes three gyroscopes, three accelerometers, and three Hall-effect magnetometers (a triad of three-axis gyroscopes, accelerometers, and magnetometers) that can be integrated onto a single circuit board or configured on separate boards for one or more sensors (e.g., gyroscopes, accelerometers, magnetometers), and output data related to three mutually perpendicular directions (e.g., X, Y, and Z directions). Thus, each IMU 1002 operates to generate 21 voltage or numerical outputs from three gyroscopes, three accelerometers, and three Hall-effect magnetometers. In an exemplary embodiment, each IMU 1002 includes a sensor board and a processing board, where the sensor board includes an integrated sensing module (LSM9DS, ST-Microelectronics) consisting of three accelerometers, three gyroscopes, and three magnetometers, and two integrated sensing modules (LSM303, ST-Microelectronics) consisting of three accelerometers and three magnetometers. In particular, each IMU 1002 includes angular momentum sensors that measure changes in spatial rotational orientation about at least three axes: pitch (up and down), yaw (side to side), and roll (clockwise or counterclockwise rotation). More specifically, each magnetometer constituting the integrated sensing module is positioned at a different location on the circuit board and assigned to output a voltage proportional to an applied magnetic field and to sense the polarity of the magnetic field at a spatial point for each of three directions in a three-dimensional coordinate system, for example, a first magnetometer outputs a voltage proportional to the applied magnetic field and the polarity of the magnetic field in the X, Y, and Z directions at a first location.Meanwhile, the second magnetometer outputs a voltage proportional to the applied magnetic field and the polarity direction of the magnetic field in the X, Y, and Z directions at the second location, and the third magnetometer outputs a voltage proportional to the applied magnetic field and the polarity direction of the magnetic field in the X, Y, and Z directions at the third location. These three magnetometers may be used to determine the IMU's heading orientation in addition to detecting local magnetic field variations. Each magnetometer uses the magnetic field as a reference and determines its orientation deviation from magnetic north. However, the local magnetic field can be distorted by steel or magnetic materials, commonly referred to as hardened and softened iron distortion. Examples of softened iron distortion are materials with low magnetic permeability, such as carbon steel and stainless steel. Hardened iron distortion is caused by permanent magnets. These distortions create non-uniform fields (see Figure 184), which affect the accuracy of the algorithms used to process the magnetometer outputs and determine heading orientation. As a result, as discussed in more detail below, a calibration algorithm is used to calibrate the magnetometers to restore uniformity in the detected magnetic field. Each IMU 1002 may be powered by a replaceable or rechargeable energy storage device, such as, but not limited to, a CR2032 coin cell battery and a 200 mAh rechargeable Li-ion battery.

[0157] The integrated sensing module of the IMU 1002 may include a configurable signal conditioning circuit and an analog-to-digital converter (ADC) that generates the sensor's numeric output. The IMU 1002 may use sensors with voltage outputs, and the external signal conditioning circuit may be an offset amplifier configured to adjust the sensor output to the input range of a multi-channel 24-bit analog-to-digital converter (ADC) (ADS1258, Texas Instruments). The IMU 1002 also includes an integrated processing module (CC2541, Texas Instruments) that includes a microcontroller and a wireless transmitter module. Alternatively, the IMU 1002 may use a separate low-power microcontroller (MSP430F2274, Texas Instruments) as the processor and a compact wireless transmitter module (A2500R24A, Anaren) for communications. The processor may be integrated as part of each IMU 1002, or it may be separate from each IMU but communicatively coupled to the IMU. The processor may be Bluetooth® enabled and provide wired or wireless communication to the gyroscope, accelerometer, and magnetometer, as well as wired or wireless communication to the signal receiver.

[0158] Each IMU 1002 is communicatively coupled to a signal receiver, which processes the received data from multiple IMUs using a predefined device identification number. The data rate is approximately 100 Hz for a single IMU and decreases as more IMUs are connected to the shared network. The signal receiver software receives signals from the IMUs 1002 in real time and continuously calculates the current position of the IMU based on the received IMU data. Specifically, the current velocity of the IMU in each of the three axes is calculated by integrating the acceleration measurement output from the IMU with respect to time. The current position is calculated by integrating the calculated velocity for each axis with respect to time. However, to obtain useful position data, a reference frame including the calibration of each IMU must be established.

[0159] The present disclosure includes novel systems and methods for calibrating one or more IMUs used in surgical navigation. Prior patent literature utilizing IMUs as intended aids in surgical navigation has suffered from inoperability for a number of reasons, including placement of the IMU relative to metallic surgical instruments as well as a lack of IMU calibration. More specifically, in the context of IMUs incorporating magnetometers, local calibration of the magnetometers is essential for motion tracking of surgical instruments and associated orthopedic components.

[0160] Referring to FIG. 182 , in accordance with the present disclosure, a novel calibration instrument 1000 is utilized to calibrate one or more IMUs 1002, which may incorporate magnetometers. In an exemplary form, the calibration instrument 1000 includes a fixed base 1006 housing a controller 1008, a motor 1012, a gearing 1016, a drive shaft 1020, and a power supply 1024. The drive shaft 1020, like the motor 1012, is attached to a portion of the gearing 1016 such that the motor operates to both drive the gearing and rotate the drive shaft. In particular, the motor 1012 includes an electric motor having a single drive shaft to which a drive gear of the gearing 1016 is attached. The drive gear engages an auxiliary gear attached to the drive shaft 1020, thereby converting rotational motion of the motor 1012 to rotational motion of the drive shaft 1020.

[0161] In this exemplary configuration, the fixed base 1006 includes a circular exterior that partially defines a hollow interior that houses the motor 1012, gearing 1016, controller 1008, power supply 1024, and a portion of the drive shaft 1020. By way of example, a central vertical axis coaxial with the central axis of the drive shaft 1020 extends through the fixed base 1006. This coaxial alignment reduces vibrations that arise as a result of rotation of the drive shaft 1020 relative to the fixed base 1006. The rotation of the drive shaft 1020 acts to rotate the outer stage 1030 relative to the fixed base 1006.

[0162] In the exemplary embodiment, a ring-shaped bearing plate 1034 is interposed between the top of the fixed base 1006 and the bottom of the outer stage 1030. Both the fixed base 1006 and the bearing plate 1034 have corresponding axial openings that allow a portion of the drive shaft 1020 to pass through. The end of the drive shaft 1020 proximate the outer stage 1030 is attached to a slip ring 1038, which is in turn attached to the outer stage. Thus, rotation of the drive shaft 1020 relative to the fixed base 1006 causes the outer stage 1030 to rotate about a central vertical axis. As discussed in more detail below, the IMU 1002 is calibrated in part by rotation about the central vertical axis.

[0163] In this exemplary embodiment, the outer stage 1030 includes a block U-shaped profile with corresponding opposing fork-shaped appendages 1042. Each appendage 1042 is mounted to a roller bearing assembly 1046 that receives a central shaft 1050 and is pivotally mounted relative to the central shaft 1050. Each central shaft 1050 is simultaneously mounted to an opposite side of an inner platform 1054 located between the fork-shaped appendages 1042. The inner platform 1054 includes a block U-shaped profile that fits over the corresponding opposing fork-shaped appendages 1042 and has a base with a plurality of upstanding protrusions 1058. As discussed in more detail below, each of the upstanding protrusions 1058 is configured to engage with a corresponding recess associated with each IMU 1002, thereby fixing the position of the IMU relative to a portion of the calibration fixture 1000. Each central shaft 1050 is longitudinally aligned along a central axis and mounted to the inner platform 1054 such that rotation of the central shaft coincides with rotation of the inner platform 1054 relative to the outer stage 1030.

[0164] To rotate the inner platform 1054 relative to the outer stage 1030, the calibration fixture includes a pulley 1060 attached to one of the central shafts 1050. Notably, one of the central shafts 1050 is longer than the other to accommodate the attachment of the pulley 1060 and its corresponding rotation by a drive belt 1064 simultaneously engaged with an electric motor 1068. In this exemplary embodiment, the output shaft of the electric motor 1068 is attached to its own pulley 1072 and engages the drive belt 1064, ultimately rotating the pulley 1060 and correspondingly rotating the inner platform 1054 relative to the outer stage 1030 (about the longitudinally aligned central axis of the central shaft 1050) when power is applied to the motor. The electric motor 1068 is attached to a motor mount 1076 that extends from the bottom of the outer stage 1030 to the underside of one of the fork-like appendages 1042. As discussed in more detail below, the IMU 1002 is calibrated in part by rotation of the inner platform 1054 relative to the outer stage 1030, and thus by rotation of the IMU about a central longitudinal axis that is perpendicular to the central vertical axis. Those skilled in the art will appreciate that a third axis of rotation may be introduced to rotate the IMU about an axis perpendicular to both the central longitudinal axis and the longitudinal vertical axis. An exemplary calibration sequence for calibrating one or more IMUs 1002 using the calibration tool 1000 is described below.

[0165] In an exemplary embodiment, the IMU 1002 is preferably calibrated in close proximity to its ultimate point of use in surgical navigation. This may be in the operating room, or more specifically, next to the patient bed where the patient is intended to or actually lies. Calibration of the IMU is location-specific, such that calibration away from the intended point of use may significantly disperse the magnetic field at the calibration point and in the area of ​​use (i.e., the surgical area). As a result, it is preferable to calibrate the IMU 1002 in close proximity to the area of ​​use.

[0166] Using the novel calibration fixture 1000, each IMU 1002 is mounted to one of the upright protrusions 1058 of the inner platform 1054. By way of example, each IMU 1002 is mounted to a housing having a shaped outer edge that defines an open bottom. The shaped outer edge of the IMU 1002 housing is configured to outline the upright protrusion 1058 so that the IMU housing can snap-fit ​​onto the corresponding upright protrusion, maintaining engagement between the IMU housing and the inner platform 1054 during the calibration sequence. By way of example, the IMU housing may have an outer edge with rectangular, triangular, rectangular, etc. sides that engage with the corresponding upright protrusion 1058. For illustrative purposes, the IMU housing has a rectangular opening defined by a constant vertical cross-section slightly larger than the rectangular cross-section of the upright protrusion 1058. In an exemplary embodiment, the calibration fixture 1000 includes four upright protrusions 1058 to enable simultaneous calibration of four IMUs 1002. However, it should be noted that the number of upstanding protrusions 1058 included as part of the inner platform 1054 may be more or less than four to allow for simultaneous calibration of one or more IMUs.

[0167] The goal of the calibration sequence is to establish zero for the accelerometers (meaning that, at a stationary position, the accelerometers provide data consistent with zero acceleration) and to map the local magnetic field and normalize the magnetometer output to account for directional dispersion and distortion of the detected magnetic field. To calibrate the IMU 1002 accelerometers, the inner platform 1054 is maintained stationary relative to the outer stage 1030, which in turn is maintained stationary relative to the fixed base 1006. Multiple measurements are taken from all accelerometers comprising the inner platform 1054 at a first fixed stationary position relative to the outer stage 1030. The inner stage is then moved to a second fixed stationary position relative to the outer stage 1030, and a second set of measurements is taken from all accelerometers. The output of the accelerometers at the multiple fixed positions is recorded and used, on an accelerometer-specific basis, to establish zero acceleration measurements for the applicable accelerometers. In addition to establishing a zero for the accelerometer, the calibration sequence also maps the local magnetic field and normalizes the magnetometer output to account for directional dispersion and distortion of the sensed magnetic field.

[0168] To map the local magnetic field of each magnetometer (assuming multiple magnetometers in each IMU 1002 are positioned at different locations), the outer stage 1030 is rotated about the drive axis 1020 and central vertical axis relative to the fixed base 1006, and the inner platform 1054 is rotated about the central axis 1050 and central axis relative to the outer stage 1030. Output data from each magnetometer is recorded while the inner platform 1054 is rotated about two mutually perpendicular axes. Repositioning each magnetometer about the two perpendicular axes generates a point cloud and map of the three-dimensional local magnetic field sensed by each magnetometer. Figures (Calibration Figures 1-3) show exemplary local magnetic fields mapped from isometric, front, and top views based on data received from the magnetometers while simultaneously rotating about two axes. As reflected in the local magnetic field map, the local map embodies an ellipsoid. This ellipsoidal shape is the result of distortion of the local magnetic field due to the presence of steel or magnetic materials, commonly referred to as hardening and softening iron distortion. Examples of softening iron distortion are materials with low magnetic permeability, such as carbon steel, stainless steel, etc. Hardening iron distortion is caused by materials such as permanent magnets.

[0169] However, with respect to the distortion of the local magnetic field, it is assumed that the local magnetic field map becomes spherical. As a result, the calibration sequence operates by manually manipulating the calibration instrument 1000 or the IMU to collect enough data points to describe the local magnetic field at various orientations. The calibration algorithm then calculates correction factors to map the distorted elliptical local magnetic field into a uniform spherical field.

[0170] Referring to FIG. 184, multiple magnetometers positioned as part of the IMU 1002 at different locations are used to detect the local magnetic field after calibration is complete. Without any magnetic field distortion, each magnetometer should provide data indicating the exact same direction, such as magnetic north. However, distortions in the local magnetic field, such as the presence of steel or magnetic materials (e.g., surgical tools), will cause the magnetometers to provide different data regarding the direction of magnetic north. In other words, if the magnetometer outputs do not uniformly reflect magnetic north, distortions have occurred, and the IMU 1002 may temporarily disable the magnetometer data from being used in its tracking algorithms. A user may also be alerted that a distortion has been detected.

[0171] 185 and 186, exemplary surgical instruments that receive an IMU 1002 include an electrical switch pattern or grid that is unique to each instrument. More specifically, each surgical instrument includes a protrusion having a mostly planar upper surface, except for one or more cylindrical cavities. In an exemplary configuration, each IMU 1002 includes a housing that defines a bottom opening configured to receive the protrusion of the surgical instrument. The bottom opening includes four switches, each including a biased cylindrical button that, when pressed, closes the switch and transmits a corresponding signal to the processor of the IMU 1002. Conversely, when the button is not pressed, the switch remains open, and no signal corresponding to the switch closure is transmitted to the processor of the IMU 1002. In this manner, the processor determines which switches are open and which are closed and uses this information to identify the surgical instrument to which the IMU 1002 is attached.

[0172] As part of the identification of a surgical instrument, zero to four switches may be pressed depending on the shape of the top surface of the protrusion. As shown in the graphic, the protrusion is received in the bottom opening of the housing of the IMU 1002 so that the top surface of the protrusion of the surgical instrument is pressed adjacent to the switch. It is noted that the protrusion and bottom opening of the housing of the IMU 1002 are configured so that the protrusion is received in the bottom opening in only a single rotational orientation, thereby limiting the possibility of misalignment between the protrusion and the switch, which could lead to misidentification of the surgical instrument.

[0173] In particular, as shown in FIG. 185, the calibration adapter surgical instrument includes a single cylindrical cavity positioned near the right front corner of the prong (opposite the milled corner) to provide a unique configuration. Thus, when the prong of the calibration adapter surgical instrument is received in the bottom opening of the housing of the IMU 1002, only the single switch of the 2&2 switch grid is actuated closest to the right front corner of the housing of the IMU 1002, thereby signaling to the processor of the IMU 1002 that the IMU 1002 is attached to the calibration adapter surgical instrument. In contrast, the patient anatomical mapping (PAM) alignment instrument adapter surgical instrument includes two cylindrical cavities positioned near the right front and right rear corners of the prong in a second unique configuration. Thus, when the prongs of the PAM adapter surgical instrument are received in the bottom opening of the housing of the IMU 1002, only the two switches of the 2&2 switch grid are actuated closest to the right side of the housing of the IMU 1002, thereby informing the processor of the IMU 1002 that the IMU 1002 is attached to the PAM adapter surgical instrument. Additionally, the reamer adapter surgical instrument includes two cylindrical cavities positioned near the front of the prongs (i.e., adjacent the left front and right front corners). Thus, when the prongs of the reamer adapter surgical instrument are received in the bottom opening of the housing of the IMU 1002, only the two switches of the 2&2 switch grid are actuated closest to the front of the housing of the IMU 1002, thereby informing the processor of the IMU 1002 that the IMU 1002 is attached to the reamer adapter surgical instrument. Finally, the impactor adapter surgical instrument includes three cylindrical cavities positioned near the front and right side of the prongs (i.e., adjacent the left front, right front, and right rear corners). Thus, when the protrusions of the impactor adapter surgical instrument are received in the bottom opening of the IMU 1002 housing, only the three switches of the 2&2 switch grid closest to the front and right side of the IMU 1002 housing are activated, thereby informing the IMU 1002 processor that the IMU 1002 is attached to the impactor adapter surgical instrument.Those skilled in the art will appreciate the variations that can be provided by providing multiple switches or electrical contacts as part of the IMU 1002 that match multiple protrusions, cavities, or electrical contacts associated with the surgical instrument, thereby clearly identifying the surgical instrument to which the IMU 1002 is attached.

[0174] Identification of the surgical instrument to which the IMU 1002 is attached is important for accurate surgical navigation. In particular, a surgical navigation system according to the present disclosure includes a software package pre-loaded with a CAD model or surface model of each surgical instrument to which the IMU 1002 can be attached. As such, the software package knows the relative dimensions of each surgical instrument, such as, but not limited to, length in the X direction, width in the Y direction, and height in the Z direction, and how these dimensions vary along the length, width, and height of the surgical instrument. Thus, when the IMU 1002 is attached to a surgical instrument in a known position, the position and orientation information (via the gyroscope, accelerometer, and magnetometer) from the IMU 1002 can be translated into position and orientation information for the surgical instrument. Thus, by tracking the IMU 1002 in 3D space, the software package can track the surgical instrument to which the IMU 1002 is attached in 3D space and communicate this position and orientation to a user, such as a surgeon or a surgeon's assistant.

[0175] In an exemplary embodiment, the software package includes a virtual display operative to display each surgical instrument as a 3D model. When an IMU 1002 is attached to a surgical instrument, the processor of the IMU 1002 transmits data to the software package that identifies the surgical instrument to which the IMU 1002 is attached. After this identification, the software package displays a 3D model of the surgical instrument attached to the IMU 1002 in an orientation consistent with the orientation information derived from the IMU. In addition to manipulating the 3D virtual model of the surgical instrument in real time to provide orientation information, the software package also uses a second reference IMU 1002 attached to a reference object (i.e., a patient's bone) to provide real-time data regarding the position of the surgical instrument. However, the IMUs 1002 (IMU #1 attached to the surgical instrument and IMU #2 attached to the reference object (i.e., bone)) must be aligned with one another before the software package can provide meaningful position information.

[0176] In an exemplary embodiment in the context of total hip arthroplasty, as shown in FIGS. 103-110, the alignment fixtures are utilized to recreate the template surgical plan by engaging the patient anatomy in a predetermined orientation. When each practical IMU 1002 is attached to its alignment fixture (one to the femur, the second to the pelvis), the alignment fixture is attached to the associated bone in the predetermined orientation (only one orientation that "zeroes" the IMU by precisely matching it to the patient's anatomy). This alignment is performed by rigidly attaching a second reference IMU to the bone in question (one IMU attached to the pelvis, and a second IMU attached to the femur). In other words, one practical IMU is attached to the acetabular alignment fixture, while a second reference IMU is rigidly attached to the pelvis. In the context of the femur, one practical IMU is attached to the femoral alignment fixture, while a second reference IMU is rigidly attached to the femur. As part of the alignment process, computer software uses the outputs from both IMUs (actual and reference) to calculate the "zero" position of the actual IMU when the alignment instrument is finally at rest and positioned in its unique position and orientation. The IMU 1002 may then be detached from its associated alignment instrument and attached to a surgical instrument (reamer, saw, implant placement guide, etc.) in a predetermined manner to ensure proper orientation and placement of the surgical instrument. The IMU 1002 may also be attached and detached successively to each surgical instrument until the surgical procedure is complete.

[0177] In this exemplary embodiment, the acetabular alignment instrument includes an elongated shaft with a unique protrusion shaped to fit the patient's acetabular cup in only a single orientation (including rotational and angular positions). The proximal end of the alignment instrument includes an IMU 1002 alignment holster for receiving the IMU 1002 such that the IMU 1002 is rigidly fixed relative to the alignment instrument and the unique protrusion when the IMU 1002 is locked within the holster. A second reference IMU 1002 is rigidly fixed relative to the pelvis in a known position in agreement with the alignment instrument. When the unique prongs of the alignment tool are correctly oriented with the patient's acetabular cup (and the IMU 1002 locked in the alignment holster and the IMU 1002 attached to the pelvis are activated), the orientation of the IMU 1002 locked in the alignment holster relative to the planned implant cup orientation (set when the unique prongs are received in the acetabular cup in only the single correct orientation) is known. The operator indicates to the software system that the IMUs are in the correct position, and the software records the position of each IMU. Then, after the alignment tool (with the IMU 1002 locked in the holster) is removed from the anatomy, the IMU 1002 is removed from the alignment holster, preparing the IMU 1002 for attachment to a surgical tool.

[0178] As an example, an IMU 1002 previously attached to an acetabular alignment tool is removed from the tool and attached to a surgical tool in a known position. In an exemplary configuration, the IMU 1002 (previously attached to the acetabular alignment tool) is rigidly fixed to a cup reamer with a known orientation relative to the drilling direction, so that the orientation of the cup reamer relative to the pelvis is known and dynamically updated by both IMUs (the IMU 1002 attached to the cup reamer and the IMU 1002 attached to the pelvis).

[0179] The software program provides a graphical user interface for the surgeon that displays a virtual model of the patient's pelvis and a virtual model of the surgical instrument in question (in this case, a cup reamer) (the virtual model of the patient's pelvis has already been completed through a virtual templating step, and the virtual model of the surgical instrument, such as a cup reamer, has previously been loaded into the system for the specific surgical instrument, such as an available cup reamer), and updates the orientation of the pelvis and surgical instrument in real time through the graphical user interface that provides position and orientation information to the surgeon. Rather than using a graphical user interface, the system may include a surgical instrument with indicator lights that show the surgeon whether the reamer is correctly oriented and, if not, the direction in which the reamer needs to be repositioned to properly orient it consistent with the preoperative plan. After completing the faceting using the cup reamer, the IMU 1002 is detached from the cup reamer and rigidly fixed to the cup inserter, whose orientation relative to the insertion direction is known. The cup inserter is then utilized to place the cup implant, while the IMU continues to provide acceleration feedback utilized by the software to perform position calculations and provide real-time feedback on the position of the pelvis relative to the cup inserter. To the extent that the pelvis is drilled before or after positioning the cup, the IMU 1002 previously attached to the alignment tool may be rigidly fixed to the surgical drill to ensure proper orientation of the drill relative to the pelvis. Similar alignment tools and IMU sets may also be used in conjunction with software systems to assist in the placement of femoral stem components.

[0180] In one exemplary embodiment, the femoral alignment tool includes an elongated shaft having a distal shape shaped to fit partially over the patient's femoral neck in only a single orientation (including rotational and angular positions). The proximal end of the alignment tool includes an IMU 1002 alignment holster for receiving the IMU 1002 such that the IMU 1002 is rigidly fixed relative to the alignment tool and distal shape when the IMU 1002 is locked within the holster. A second reference IMU 1002 is rigidly fixed to the femur in a known position in agreement with the alignment tool. When the distal shape of the alignment tool is correctly oriented relative to the femoral neck (and the IMU 1002 locked in the alignment holster and the IMU 1002 attached to the femur are activated), the orientation of the IMU 1002 locked in the alignment holster relative to the femoral orientation (set when the distal shape is received on the femoral neck in only the single correct orientation) is known. The operator indicates to the software system that the IMUs are in the correct position, and the software records the position of each IMU. Then, after the alignment tool (with the IMU 1002 locked in the holster) is removed from the anatomy, the IMU 1002 is removed from the alignment holster, preparing the IMU 1002 for attachment to a surgical tool.

[0181] As an example, the IMU 1002 previously attached to the femoral alignment instrument is removed from the instrument and attached to another surgical instrument in a known position. In an exemplary configuration, the IMU 1002 (previously attached to the femoral alignment instrument) is rigidly fixed to the surgical saw in a known position, so that movement of the IMU 1002 translates into a corresponding known movement of the surgical saw. Assuming the other IMU 1002 is rigidly attached to the femur in a known position, the IMUs cooperate to provide dynamic updates to the software system regarding the changing positions of both the femur and the surgical saw (via acceleration data).

[0182] The software program provides a graphical user interface for the surgeon that displays a virtual model of the patient's femur and a virtual model of the surgical instrument in question (in this case, a surgical saw) (the virtual model of the patient's femur has already been completed through a virtual templating step, and the virtual model of the surgical instrument, such as a surgical saw, has previously been loaded into the system for the specific surgical instrument, such as a surgical saw), and updates the orientation of the femur and surgical instrument in real time with the graphical user interface providing position and orientation information to the surgeon. Rather than using a graphical user interface, the system may include a surgical instrument with indicator lights that show the surgeon whether the surgical saw is correctly oriented and, if not, the direction in which the surgical saw needs to be repositioned to properly orient it in agreement with the preoperative plan and make the correct osteotomy. After the required osteotomy, the IMU 1002 is detached from the surgical saw and rigidly secured to the reamer (to properly drill the intramedullary canal) before being attached to a femoral stem inserter whose orientation relative to the insertion direction is known. The stem inserter is then used to place the femoral stem implant into the drilled intramedullary canal, while the IMU continues to provide acceleration feedback that the software uses to calculate the position of the femur and stem inserter in real time and display this position data to the surgeon via a graphical user interface.

[0183] In an exemplary embodiment in the context of total shoulder arthroplasty, as shown in FIGS. 187 and 188, alignment devices are utilized to recreate the template surgical plan by engaging the patient anatomy in a predetermined orientation. When each practical IMU 1002 is attached to its alignment device (one to the humerus, the second to the scapula), the alignment device is attached to the associated bone in the predetermined orientation (only one orientation that precisely matches the patient anatomy, thereby "zeroing" the IMU). This alignment is performed by rigidly attaching a second reference IMU to the bone in question (one IMU attached to the humerus and a second IMU attached to the scapula). In other words, one practical IMU is attached to the humerus alignment device, while a second reference IMU is rigidly attached to the humerus. In the context of the scapula, one practical IMU is attached to the scapula alignment device, while a second reference IMU is rigidly attached to the scapula. As part of the alignment process, computer software uses the outputs from both IMUs (actual and reference) to calculate the "zero" position of the actual IMU when the alignment instrument is finally at rest and positioned in its unique position and orientation. The IMU 1002 may then be detached from its associated alignment instrument and attached to a surgical instrument (reamer, saw, implant placement guide, etc.) in a predetermined manner to ensure proper orientation and placement of the surgical instrument. The IMU 1002 may also be attached and detached successively to each surgical instrument until the surgical procedure is complete.

[0184] In this exemplary embodiment, as shown in FIG. 188 , the scapula alignment device includes an elongated shaft with a unique protrusion shaped to fit into the patient's glenoid cavity in only a single orientation (including rotational and angular positions). The proximal end of the alignment device includes an IMU 1002 alignment holster for receiving the IMU 1002 such that the IMU 1002 is rigidly secured relative to the alignment device and the unique protrusion when the IMU 1002 is locked within the holster. In line with the alignment device, a second reference IMU 1002 is rigidly secured relative to the scapula at a known position. When the unique prongs of the alignment tool are correctly oriented in the patient's glenoid cavity (and the IMU 1002 locked in the alignment holster and the IMU 1002 attached to the scapula are activated), the orientation of the IMU 1002 locked in the alignment holster relative to the planned implant cup orientation (set when the unique prongs of the alignment tool are received in the glenoid cavity in only the single correct orientation) is known. The operator indicates to the software system that the IMUs are in the correct position, and the software records the position of each IMU. Then, after the alignment tool (with the IMU 1002 locked in the holster) is removed from the anatomy, the IMU 1002 is removed from the alignment holster, and the working IMU 1002 is ready to be attached to other surgical instruments.

[0185] As an example, an IMU 1002 previously attached to a scapular alignment instrument is removed from the instrument and attached to a surgical instrument in a known position. In an exemplary configuration, the IMU 1002 (previously attached to the scapular alignment instrument) is rigidly fixed to a cup reamer whose orientation relative to the drilling direction is known, such that the orientation of the cup reamer relative to the scapula is known and dynamically updated by both IMUs (the cup reamer-attached IMU 1002 and the pelvis-attached IMU 1002).

[0186] The software program provides a graphical user interface for the surgeon that displays a virtual model of the patient's scapula and a virtual model of the surgical instrument in question (in this case, a cup reamer) (the virtual model of the patient's scapula has already been completed through the virtual templating step, and the virtual model of the surgical instrument, such as a cup reamer, has previously been loaded into the system for the specific surgical instrument, such as an available cup reamer), and updates the orientation of the scapula and surgical instrument in real time through the graphical user interface that provides position and orientation information to the surgeon. Rather than using a graphical user interface, the system may include a surgical instrument with indicator lights that indicate whether the reamer is correctly oriented and, if not, the direction in which the reamer needs to be repositioned to properly orient it consistent with the preoperative plan. After completing the faceting using the cup reamer, the operational IMU 1002 is detached from the cup reamer and rigidly fixed to the cup inserter, whose orientation relative to the insertion direction is known. The cup inserter is then utilized to place the cup implant, while the IMU continues to provide acceleration feedback utilized by the software to perform position calculations and provide real-time feedback on the position of the scapula relative to the cup inserter. To the extent that a hole is drilled in the scapula before or after positioning the cup, a utility IMU 1002 previously attached to the alignment tool may be rigidly fixed to the surgical drill to ensure proper orientation of the drill relative to the scapula. Similar alignment tools and IMU sets may also be used in conjunction with software systems to assist in the placement of humeral stem components.

[0187] In one exemplary embodiment, the humerus alignment device includes an elongated shaft having a distal shape shaped to fit partially over the patient's humeral neck in only a single orientation (including rotational and angular positions). The alignment device includes an IMU 1002 alignment holster at its proximal end for receiving the IMU 1002 such that the IMU 1002 is rigidly fixed relative to the alignment device and distal shape when the IMU 1002 is locked within the holster. A second reference IMU 1002 is rigidly fixed to the humerus in a known position in agreement with the alignment device. When the alignment tool is correctly oriented relative to the humeral neck (and the IMU 1002 locked in the alignment holster and the reference IMU 1002 attached to the humerus are activated), the orientation of the IMU 1002 locked in the alignment holster relative to the humeral orientation (set when the distal shape is received on the humeral neck in only the single correct orientation) is known. The operator indicates to the software system that the IMUs are resting in the correct position, and the software records the position of each IMU to establish a reference orientation for the preliminary planned direction. Then, after the alignment tool (with the IMU 1002 locked in the holster) is removed from the anatomy, the working IMU 1002 is removed from the alignment holster, preparing the IMU 1002 for attachment to other surgical tools.

[0188] As an example, an IMU 1002 previously attached to a humerus alignment instrument is removed from the instrument and attached to another surgical instrument in a known position. In an exemplary configuration, the IMU 1002 (previously attached to the humerus alignment instrument) is rigidly fixed to the surgical saw in a known position, such that movement of the IMU 1002 translates into a corresponding known movement of the surgical saw. Assuming the reference IMU 1002 is rigidly attached to the humerus in a known position, the IMUs cooperate to provide dynamic updates to the software system regarding changes in the position of both the humerus and the surgical saw (via acceleration data).

[0189] The software program provides a graphical user interface for the surgeon that displays a virtual model of the patient's humerus and a virtual model of the surgical instrument in question (in this case, a surgical saw) (the virtual model of the patient's humerus has already been completed through a virtual templating step, and the virtual model of the surgical instrument, such as a surgical saw, has been previously loaded into the system for the specific surgical instrument, such as a surgical saw), and updates the orientation of the humerus and surgical instrument in real time with the graphical user interface providing position and orientation information to the surgeon. Rather than using a graphical user interface, the system may include a surgical instrument with indicator lights that show the surgeon whether the surgical saw is correctly oriented and, if not, the direction in which the surgical saw needs to be repositioned to properly orient it in agreement with the preoperative plan and make the correct osteotomy. After the required osteotomy, the practical IMU 1002 is detached from the surgical saw and rigidly secured to the reamer (to properly drill the humeral canal) before being attached to a humeral stem inserter whose orientation relative to the insertion direction is known. The stem inserter is then used to place the humeral stem implant into the drilled canal, while the IMU continues to provide acceleration feedback that the software uses to calculate the position of the humerus and stem inserter in real time and display this position data to the surgeon via a graphical user interface.

[0190] In an exemplary embodiment in the context of a reverse shoulder implant surgery, as shown in FIGS. 189 and 190, alignment devices are utilized to recreate the template surgical plan by engaging the patient anatomy in a predetermined orientation. When each practical IMU 1002 is attached to its alignment device (one to the humerus, the second to the scapula), the alignment device is attached to the associated bone in the predetermined orientation (only one orientation that precisely matches the patient anatomy, thereby "zeroing" the IMU). This alignment is performed by rigidly attaching a second reference IMU to the bone in question (one IMU attached to the humerus and a second IMU attached to the scapula). In other words, one practical IMU is attached to the humerus alignment device, while a second reference IMU is rigidly attached to the humerus. In the context of the scapula, one practical IMU is attached to the scapula alignment device, while a second reference IMU is rigidly attached to the scapula. As part of the alignment process, computer software uses the outputs from both IMUs (actual and reference) to calculate the "zero" position of the actual IMU when the alignment tool is finally at rest and positioned in its unique position and orientation. The IMU 1002 may then be detached from its associated alignment tool and attached to a surgical tool (reamer, saw, inserter, drill guide, drill, etc.) in a predetermined manner to ensure proper orientation and placement of the surgical tool. The IMU 1002 may also be attached and detached successively to each surgical tool until the surgical procedure is complete.

[0191] In this exemplary embodiment, as shown in FIG. 190, the scapula alignment device includes an elongated shaft with a unique protrusion shaped to fit into the patient's glenoid cavity in only a single orientation (including rotational and angular positions). The proximal end of the alignment device includes an IMU 1002 alignment holster for receiving the IMU 1002 such that the IMU 1002 is rigidly secured relative to the alignment device and the unique protrusion when the IMU 1002 is locked within the holster. In line with the alignment device, a second reference IMU 1002 is rigidly secured relative to the scapula at a known position. When the unique prongs of the alignment tool are correctly oriented in the patient's glenoid cavity (and the IMU 1002 locked in the alignment holster and the IMU 1002 attached to the scapula are activated), the orientation of the IMU 1002 locked in the alignment holster relative to the planned implant cup orientation (set when the unique prongs are received in the glenoid cavity in only the single correct orientation) is known. The operator indicates to the software system that the IMUs are in the correct position, and the software records the position of each IMU. Then, after the alignment tool (with the IMU 1002 locked in the holster) is removed from the anatomy, the IMU 1002 is removed from the alignment holster, and the working IMU 1002 is ready to be attached to other surgical tools.

[0192] As an example, an IMU 1002 previously attached to a scapular alignment instrument is removed from the instrument and attached to a surgical instrument in a known position. In an exemplary configuration, the IMU 1002 (previously attached to the scapular alignment instrument) is rigidly fixed to a cup reamer whose orientation relative to the drilling direction is known, such that the orientation of the cup reamer relative to the scapula is known and dynamically updated by both IMUs (the cup reamer-attached IMU 1002 and the pelvis-attached IMU 1002).

[0193] The software program provides a graphical user interface for the surgeon that displays a virtual model of the patient's scapula and a virtual model of the surgical instrument in question (in this case, a cup reamer) (the virtual model of the patient's scapula has already been completed through a virtual templating step, and the virtual model of the surgical instrument, such as a cup reamer, has previously been loaded into the system for the specific surgical instrument, such as an available cup reamer), and updates the orientation of the scapula and surgical instrument in real time through the graphical user interface that provides position and orientation information to the surgeon. Rather than using a graphical user interface, the system may include a surgical instrument with indicator lights that indicate whether the reamer is correctly oriented and, if not, the direction in which the reamer needs to be repositioned to properly orient it in alignment with the preoperative plan. After completion of the surface replacement using the cup reamer, the operational IMU 1002 is detached from the cup reamer and rigidly fixed to the drill plate, whose orientation and position are known. The drill plate is then utilized to drill holes in the scapula, while the IMU continues to provide acceleration feedback that software utilizes to perform position calculations to provide real-time feedback on the position of the scapula relative to the drill plate, and subsequently position the glenoid base plate and attach the glenoid component ball. While not required, the operational IMU 1002 may be rigidly fixed to the surgical drill to ensure proper orientation of the drill relative to the drill plate when drilling holes through the drill plate. Similar alignment tools and IMU sets may also be used in conjunction with the software system to assist in the placement of the humeral stem component.

[0194] In one exemplary embodiment, the humerus alignment device includes an elongated shaft having a distal shape shaped to fit partially over the patient's humeral neck in only a single orientation (including rotational and angular positions). The alignment device includes an IMU 1002 alignment holster at its proximal end for receiving the IMU 1002 such that the IMU 1002 is rigidly fixed relative to the alignment device and distal shape when the IMU 1002 is locked within the holster. A second reference IMU 1002 is rigidly fixed to the humerus in a known position in agreement with the alignment device. When the alignment tool is correctly oriented relative to the humeral neck (and the IMU 1002 locked in the alignment holster and the reference IMU 1002 attached to the humerus are activated), the orientation of the IMU 1002 locked in the alignment holster relative to the humeral orientation (set when the distal shape is received on the humeral neck in only the single correct orientation) is known. The operator indicates to the software system that the IMUs are resting in the correct position, and the software records the position of each IMU, "zeroing" the active IMU. Then, after removing the alignment tool (with the IMU 1002 locked in the holster) from the anatomy, the active IMU 1002 is removed from the alignment holster, preparing the IMU 1002 for attachment to other surgical tools.

[0195] As an example, an IMU 1002 previously attached to a humerus alignment instrument is removed from the instrument and attached to another surgical instrument in a known position. In an exemplary configuration, the IMU 1002 (previously attached to the humerus alignment instrument) is rigidly fixed to the humerus resection block in a known position, such that movement of the IMU 1002 translates into a corresponding known movement of the resection block. Assuming the reference IMU 1002 is rigidly attached to the humerus in a known position, the IMUs cooperate to provide dynamic updates to the software system regarding changes in the position of both the humerus and the resection block (via acceleration data).

[0196] The software program provides a graphical user interface for the surgeon that displays a virtual model of the patient's humerus and a virtual model of the surgical instrument in question (in this case, a humeral resection block) (the virtual model of the patient's humerus has already been completed through a virtual templating step, and the virtual model of the surgical instrument, such as a resection block, has previously been loaded into the system for the specific resection block or other surgical instrument available), and updates the orientation of the humerus and surgical instrument in real time through the graphical user interface that provides position and orientation information to the surgeon. Rather than using a graphical user interface, the system may include a surgical instrument with indicator lights that indicate to the surgeon whether the resection block is correctly oriented and, if not, the direction in which the resection block needs to be repositioned to properly orient it in alignment with the preoperative plan and perform the correct osteotomy. Additionally or alternatively, the practical IMU 1002 may be attached to a drill plate used to drill one or more holes for inserting respective fiducial pins. In such a case, the resection block does not necessarily need to have an IMU associated with it, provided that the surgical block is properly positioned and oriented using the one or more fiducial pins. In any event, after the necessary osteotomies are made, the operational IMU 1002 is removed from the surgical instrument, rigidly fixed to the reamer (to properly drill the humeral canal), and then attached to a humeral stem inserter with a known orientation relative to the inserter. The stem inserter is then used to place the humeral stem implant into the drill canal, while the IMU continues to provide acceleration feedback that is used by software to calculate the position of the humerus and stem inserter in real time and display this position data to the surgeon via a graphical user interface.

[0197] Testing component placement and potential component impingement with IMUs attached to the pelvis and femur allows component rotation to be tracked to prevent post-operative complications and improve overall patient satisfaction.

[0198] In accordance with the above disclosure regarding the use of the IMU 1002, the following is an exemplary description of the mathematical models and algorithms used to generate three-dimensional position data from each IMU's gyroscope, accelerometer, and magnetometer. In an exemplary embodiment, each IMU processor is programmed to calculate the change in position of the IMU 1002 based on inputs from the IMU's gyroscope, accelerometer, and magnetometer using sequential Monte Carlo (SMC) techniques in conjunction with a von Mises-Fisher density algorithm. The IMU data stream consists of one set of gyro data (G1) for the X, Y, and Z axes, three sets of accelerometer data (A1-A3) for the X, Y, and Z axes, and three sets of magnetometer data (M1-M3) for the X, Y, and Z axes. Orientation tracking of the IMU 1002 may be achieved with one set of data from each sensor (i.e., G1, A1, M1).

[0199] Using G1, A1, and M1 as an example, and assuming all raw sensor data has been converted and processed, At time and state = 1, 1) The algorithm first generates a set of N particles around a neutral position with a given dispersion coefficient of the von Mises-Fisher density, as represented in Algorithm 1 specified below. Each particle represents a quadruple-shaped orientation around the X, Y, and Z axes. In other words, the particles contain a set of independent, uniformly distributed random variables drawn from the same probability density space. For orientation tracking applications, the particles are statistically constrained variations of the observed orientations. However, it should be noted that the exact statistics (dispersion coefficients) do not need to be "known," as the algorithm optimizes its performance as the number of samples collected increases. It is preferable to use a higher diversity as an initial guess and allow the algorithm to refine it.

[0200]

number

[0201] After receiving the first data set from G1, A1, and M1, an estimate of the IMU's current orientation is calculated. This is accomplished by first determining the tilt measured from A1. The tilt information is needed to mathematically correct (de-rotate) the magnetometer measurements, as shown as steps 2 and 3 of Algorithm 2 specified below. The A1 and M1 data are then used to estimate the initial orientation of the IMU using Algorithm 2, which is based on a Gauss-Newton optimization method. The goal of Algorithm 2 is to find the orientation (q) such that the slope and heading components of the estimate are the same as the measurements from A1 and M1, within an acceptable error. obv ) iteratively. Note that while Algorithm 2 requires input from past states, any input will suffice at time=1 since there are no past states. The reason that accelerometers and magnetometers cannot be used alone to track orientation is due to limitations of the tilt measurement by accelerometers. For example, at some specific orientations, due to the nature of trigonometric quadrants, the tilt output may be the same for different IMU orientations. For this reason, the gyroscope needs to record the quadrant that the IMU is in.

[0202]

number

[0203] Next, a set of N particles in the neutral position (q vMF ) are "rotated" so that their average is centered on the orientation estimates from A1 and M1.

[0204]

number

[0205] Then, all particles are estimated in advance based on G1 by the following formula:

[0206]

number

[0207] The orientation expectation value of the current state is calculated by the particle estimate (q est,i This is achieved by averaging the quadruplets (t+1). Because quadruplets are four-dimensional vectors, the averaging is done differently. In Algorithm 3, two quadruplets from the particle set are iteratively interpolated until only one remains.

[0208]

number

[0209] At time and state=2, a second data set is received. Using the same method described in paragraph 0193 (Algorithm 2), the latest orientation estimate is calculated and then compared with all particle estimates from previous states (q est,i (t-1)). The error / residual between each particle and the current orientation estimate is used to weight the accuracy of the particle (i.e. particles closer to the estimate will receive a higher weight than particles further away) according to the following formula:

[0210]

number

[0211] Next, particle quality is evaluated to filter out and resample particles with very low weights. This can be done using deterministic, residual, or supplemental resampling schemes. Because the algorithm favors particles close to the observed values, the particle population begins to lose diversity over time. Particles become highly concentrated and statistically insignificant. At that point, replacing a small fraction of the particles will increase diversity. This is done by first evaluating the current dispersion coefficient of the particles. If the dispersion coefficient indicates high concentration, some of the current particles are replaced by generating a new particle population at a neutral position based on the predetermined dispersion coefficient. The new particles are rotated from the neutral position to the current expected orientation. This is summarized in the following equation:

[0212]

number

[0213] Finally, based on the new data from G1, the particles are estimated ahead and the current orientation state is calculated again. The above process and algorithm is reused for each new data received from G1, A1, and M1.

[0214] Creating a trauma plate Referring to Figures 112-125, an exemplary process and system for preparing bone plates (i.e., trauma plates) for a given population is described. Those skilled in the art will appreciate that bone can be repaired through regeneration after fracture. Prior art trauma plates have been utilized depending on the severity and location of the fracture, but often required bending or other modification in the operating room to accommodate irregular bone shapes and maximize contact between bone fragments. However, excessive bending can shorten the lifespan of trauma plates and lead to bone plate failure and / or loosening of trauma plate screws. The present process and system provides a more precise trauma plate shape, thereby reducing or eliminating the need for intraoperative plate shaping, thereby extending plate lifespan and increasing the time before trauma plate screws loosen.

[0215] The above exemplary illustrations for creating trauma plates can be applied to any and all bones for which trauma plates are applicable. For purposes of simplicity, the exemplary illustrations describe a system and process for creating trauma plates for use with the humerus. However, it is understood that the process and system are equally applicable to the manufacture of other bones of the body and corresponding trauma plates, and are not limited to humerus trauma plates.

[0216] As part of the exemplary process and system for creating trauma plates, a statistical bone atlas of the bone in question is created and / or utilized. Illustratively, the bone in question includes the humerus. Those skilled in the art are familiar with statistical atlases and how to configure them in the context of one or more bones. As a result, a detailed description of the configuration of a statistical bone atlas is omitted for the sake of brevity. Nevertheless, what may be unique about the statistical bone atlas of the exemplary system and process is the classification of the humerus within the statistical bone atlas based on gender, age, ethnicity, deformity, and / or local configuration. In this manner, one or more trauma plates may be mass-customized to one or more of these classifications, which define a particular bone population.

[0217] In exemplary embodiments, a statistical bone atlas includes anatomical data that may take a variety of forms. By way of example, a statistical bone atlas may include two-dimensional or three-dimensional images, as well as information regarding bone parameters from which measurements can be obtained. Exemplary atlas input data may be in the form of x-ray images, CT scan images, MRI images, laser scan images, ultrasound images, segmented bones, physical measurement data, and any other information from which a bone model can be created. Software accessing the statistical atlas data utilizes this input data to construct a three-dimensional bone model (or by accessing a three-dimensional bone model that has already been created and stored as part of the statistical atlas), and the software operates to create a three-dimensional average or template bone model.

[0218] Using the template bone model, the software allows for automatic or manual specification of points on the exterior surface of the template bone model. Illustratively, in the context of an average humerus model, a user of the software defines the approximate boundary shape of the final trauma plate by roughly outlining the shape of the trauma plate on the exterior surface of the humerus model. The approximate boundary shape of the trauma plate can also be obtained by the user specifying a series of points on the exterior surface of the humerus model that correspond to the outer boundary. Once the outer boundary or boundary points are defined, the software allows for automatic or manual specification of points on the exterior surface of the humerus model within the defined boundary. As an example, the software provides a partial fill operation that allows a user to specify a percentage of the trauma plate's boundary that is specified by a series of points that each correspond to a different location on the exterior of the humerus model. The software also provides a manual point specification feature that allows a user to specify one or more points on the exterior surface of the humerus model within the boundary. Note that when using manual point specification, the user does not need to define the boundary as a prerequisite for specifying points on the exterior of the humerus model. Rather, once manual point specification is complete, the boundary is defined by the outermost points specified.

[0219] After designating points on the exterior surface of the template bone model, the software propagates the located points throughout the bone population of interest. In particular, the located points are automatically applied by the software to each 3D bone model in a given population through point correspondence in a statistical atlas. As an example, a given bone population may be gender- and ethnicity-specific and include the humeri of Caucasian women. Using the propagated points of each bone model in the population, the software creates a 3D rendering of the trauma plate for each bone by filling gaps between points within the boundary using a 3D filling process. The software then calculates the longitudinal midline of the 3D rendering of each trauma plate through a thinning process.

[0220] The midline of each 3D trauma plate rendering includes a 3D midline with varying curvatures along its length. The software extracts the 3D midline and uses least-squares fitting to determine the preferred number of radii of curvature that collaboratively best approximate the principal curvatures of the 3D midline. In the context of the humerus, three radii of curvature have been found to accurately approximate the midline curvature. However, this number may vary depending on the bone population and the trauma plate boundary. Additional features may also be included here, such as cross-sectional curvature at one or more locations along the plate length, the location of avoided muscles, nerves, and other soft tissues, or any other features relevant to defining the plate size or shape. As an example, the three radii of curvature of the midline represent the curvature of the trauma plate at the proximal humerus, the transition between the humeral shaft and humeral head, and the curvature of the humeral shaft. Each radius of curvature was recorded and clustered into groups that best fit the population by applying a 4D feature vector to the radius of curvature data. In an exemplary embodiment, the cluster data may indicate that multiple trauma plates are required for proper fitting to the population. Once the radius of curvature data has been clustered, the dimensions of the trauma plate may be determined.

[0221] During feature extraction related to plate design, the software determines the optimal number of clusters to fit the population. It is important to note that there are cases where more than one cluster results in a local minimum, as outlined in Figure 120. To determine the optimal selection that provides acceptable error and a reasonable number of plates in each family, the software generates 3D surface models for the plates in each cluster. It then performs an automated evaluation by placing these plates in the population and calculating the misfit between the plate and bone surface. The results of this analysis allow the software to select the optimal number of plates to use for this particular population. The final plate model is then parameterized to maximize fixation while avoiding muscle and soft tissue locations, and the screw locations for each plate are placed accordingly. Screw width is determined by cross-sectional analysis of the bone at each screw level across the entire population.

[0222] The process and method were validated for humeri through a cadaveric study. Specifically, CT scans of the humeri of Caucasian female cadavers were acquired. These CT scans were used by the software to create separate 3D models of each humerus. It should be noted that neither the CT scans nor the 3D models utilized during this validation study were part of the statistical atlas and associated population utilized to create the humerus trauma plates. As a result, neither the CT scans nor the 3D models contained new data and models used to validate the designed humerus trauma plates. After generating the 3D validation models, each model was classified into a specific cluster (a cluster resulting from the design of a humerus trauma plate from the design population). Based on the cluster into which the validation models were classified, a humerus trauma plate designed for that cluster was fitted to the appropriate validation 3D humerus model, and measurements were calculated indicating the desired spacing between the outer surface of the validation 3D humerus model and the underside of the humerus trauma plate. Figure 124 shows a distance map of the trauma plate fitted to the validation 3D humerus model, showing the region of greatest distance between the bone and the trauma plate. While the majority of the trauma plate is minimally separated from the bone, regions of poor fit are seen, ranging from 0.06 to 0.09 cm. Consequently, this cadaveric study concluded that the trauma plate designed using the system and the exemplary process described above has an exceptional contour fit that, when applied intraoperatively, eliminates the need for the surgeon to bend or manually reshape the bone plate.

[0223] Referring to Figures 131-138, in another exemplary implementation of this process, clavicle trauma plates were created. Here, a statistical atlas was created from many clavicles that adequately captured the variation within, for example, the Caucasian population. It should be noted that the statistical atlas may include clavicles from many ethnicities, many patient ages, and various geographic regions. While this exemplary disclosure is unintentionally contextualized to a Caucasian population dataset, one skilled in the art will understand that the above system and method is not limited solely to statistical atlases of Caucasian populations. Figure 132 illustrates the overall clavicle anatomy.

[0224] In an exemplary embodiment, the statistical atlas of the clavicle also defines the location of each clavicle relative to muscle attachment sites, as shown in Figure 134. Cross-sectional contours were extracted at 10% increments along the entire bone (see Figure 138), as well as at muscle attachment sites and the clavicle constriction (see Figure 137). The maximum and minimum dimensions of each cross-sectional contour were calculated. Asymmetry across the entire 3D surface was also examined by analyzing the magnitude and orientation differences between similar points across the entire clavicle surface in the dataset. The results confirmed existing research on clavicle asymmetry, namely, that the left clavicle is longer than the right clavicle, but the right clavicle is thicker than the left. However, as shown in Figure 139, the pattern of asymmetry differs between men and women.

[0225] Additionally, as shown in Figure 133, the midline of the clavicle does not follow a symmetrical "S" shape, as in existing clavicle trauma plate designs. Therefore, this disclosure confirms that current clavicle trauma plates fail to mimic the anatomical curvature of the clavicle. With reference to Figures 135 and 136, male clavicles exhibit significant muscle and ligament attachment contour asymmetry in all dimensions (p<0.05). Meanwhile, female asymmetry is more variable. However, the muscle-free area of ​​the posterior midshaft was significantly asymmetric in both genders.

[0226] From the clavicle features extracted across the statistical atlas, clustering (using the clustering method described above in this application, incorporated herein by reference) was performed to determine distinct similarity groups (i.e., populations), each of which best fit the population in association with a specific clavicle trauma plate. Additionally, for each trauma plate population, screw fixation locations and lengths were determined to optimally avoid soft tissue (muscle attachments) and prevent additional fractures or plate loosening due to screws that are too long or too short. Using this process, multiple clavicle trauma plate families were designed, corresponding to the mass-customized clavicle trauma plates, as shown in Figures 140-149.

[0227] Creation of patient-specific trauma plates 126 graphically depicts the patient-specific trauma process to include various components, including pre-operative surgical planning, generation of a pre-formed patient-specific trauma plate, intra-operatively guided positioning and fixation of the patient-specific trauma plate, and optional post-operative evaluation of the patient-specific trauma plate. A more detailed description of these components and exemplary processes and structures involved in each component will be discussed in turn.

[0228] 126-130 illustrate an exemplary process flow for the preoperative surgical planning component. Initial input anatomical data is obtained for the anatomy of interest. For purposes of exemplary illustration only, the clavicle will be described as the fractured or deformed anatomy, and a clavicle trauma plate will be described as a patient-specific trauma plate. The anatomical data, including two-dimensional (2D) images or three-dimensional (3D) surface representations of the clavicle, may be in the form of a surface model or point cloud, and is input into a software package configured to select or create a patient-specific clavicle trauma plate. In situations where 2D images are used, these 2D images are utilized to construct a 3D virtual surface representation of the fractured clavicle. Those skilled in the art are familiar with utilizing 2D images of anatomy to construct a 3D surface representation. Therefore, a detailed description of this process has been omitted for the sake of brevity. By way of example, the input anatomical data may include one or more of X-ray, computed tomography (CT), magnetic resonance imaging (MRI), or any other imaging data capable of generating a 3D surface representation of the tissue of interest. The output of this anatomical data input is a 3D virtual surface representation of the fractured clavicle components.

[0229] The 3D virtual surface representation of the fractured clavicle components is then evaluated to identify the location and shape of the fracture, and in the case of complete fractures and separation of the bony components, the relative position and shape of the bony components.

[0230] In the event of a complete fracture and separation of a bone component, the process and associated software implement a fracture reduction process that allows for the construction of a patchwork clavicle by manual repositioning of the 3D virtual surface representation of the fractured clavicle. In such a situation, a user repositions and reorients the 3D virtual surface representation of the fractured clavicle to create a 3D patchwork clavicle model resembling a clavicle assembled from components that include the 3D virtual surface representation. Alternatively, the process and associated software may construct a patchwork clavicle model by automatic repositioning and reconstruction of the 3D virtual surface representation of the fractured clavicle, optionally using a 3D template clavicle model. More specifically, for each fractured bone component (i.e., fracture margin) that includes a 3D virtual surface representation, the software first detects one or more fracture sites from the 3D virtual surface representation and extracts a contour from each fracture site. The software then compares the extracted contours with the contours of the 3D template clavicle model to pairwise match these contours and locate the matching bone components / fragments at each fracture site. These matched bone components / fragments are then grouped. After grouping the matched bone components / fragments, the software fits the grouped bone fragments to the 3D template clavicle model and identifies the correct positions of all bone components / fragments with respect to the 3D template clavicle model. The matched bone components / fragments are then repositioned to construct a 3D patchwork clavicle model similar to the 3D template clavicle model, which the software uses to construct a 3D reconstructed clavicle model, as described below.

[0231] After reduction, and referring again to FIGS. 7 and 127, the 3D patchwork clavicle is used to identify the anatomical model (e.g., complete bone model) in the statistical atlas that is most similar to the 3D patchwork clavicle model of the patient in question. This step is illustrated in FIG. 3 as a search for the closest bone in the atlas. First, the 3D patchwork clavicle model is compared to the bone models in the statistical atlas using one or more similarity criteria to identify the bone model in the statistical atlas that is most similar to the 3D patchwork clavicle model. The result of this initial similarity criteria is the selection of a bone model from the statistical atlas to be used as an "initial guess" in a subsequent registration step. In the registration step, the 3D patchwork clavicle model is aligned with the selected atlas bone model (i.e., the initial guess bone model) so that the output is a patient-specific reconstructed bone model aligned with the atlas bone model. Following the registration step, the shape parameters of the aligned "initial guess" are optimized so that its shape fits the 3D patchwork clavicle model.

[0232] Optimization of shape parameters (in this case from a statistical atlas) minimizes the error between the reconstructed patient-specific bone model and the 3D patchwork clavicle model by using unfractured bone regions. By varying the values ​​of the shape parameters, different anatomical shapes can be represented. This process is repeated until convergence of the reconstructed shape is achieved (possibly measured as the relative surface change between iterations or the maximum number of allowed iterations).

[0233] A relaxation step is performed to morph the optimized bone to best fit the 3D patchwork clavicle model. Consistent with this exemplary case, the missing anatomy from the 3D patchwork clavicle model output by the convergence step is applied to the morphed 3D clavicle model to create a patient-specific 3D model of the patient's reconstructed clavicle. More specifically, surface points on the 3D patchwork clavicle model are relaxed (i.e., morphed) directly onto the patient-specific 3D clavicle model to best fit the reconstructed shape to the patient-specific shape. The output of this step is a fully reconstructed patient-specific 3D clavicle model that represents what should be the normal / complete anatomy of the patient's clavicle.

[0234] After total anatomy reconstruction, the system software begins the reduction sequence planning process. In this reduction sequence planning process, the software allows for manual or automatic determination of the clavicle components (i.e., fractured clavicle fragments) and the order in which they are to be reassembled and attached to one another. To this end, the software records in memory a 3D model of the progression of the clavicle assembly from the bone components. Thus, assuming the clavicle is fractured into six components, the software would record a first 3D model showing the assembly of the first and second fracture components, followed by a second 3D model showing the assembly of the first, second, and third fracture components, and so on, until a final 3D model reflecting the assembled position and orientation of all six fractured bone components is reached, resembling a 3D patchwork clavicle model.

[0235] This reduction sequence determination allows the software to manually or automatically select one of multiple clavicle trauma plate templates using the 3D patchwork clavicle. More specifically, the clavicle trauma plate templates include a series of 3D virtual surface representations of clavicle trauma plates that have been generally shaped to match size and shape parameters associated with a given population obtained from a statistical bone atlas. In other words, the statistical bone atlas includes multiple surface models of normal, fully anatomical clavicles that have been classified based on one or more of size, ethnicity, age, sex, and any other markers indicative of bone shape. An exemplary description of the procedure for arriving at a template bone plate is provided above with respect to Figures 112-125, which are incorporated herein by reference. In the automatic selection mode, the software compares the dimensions and contours of multiple clavicle trauma plate templates with the 3D patchwork clavicle to identify the template that best fits the 3D patchwork clavicle (i.e., similarity in contour and shape relative to the bone anatomy).

[0236] Using a clavicle trauma plate template that best fits the 3D patchwork clavicle, the software allows for manual or automatic identification of fixation site locations through the trauma plate, as well as determining the orientation and length of fixation devices (e.g., surgical screws) to be utilized. In automatic fixation site identification mode, the software avoids placing any fixation holes in the path of nerve or muscle attachments by capturing muscle and attachment locations as well as nerve locations. The software also allows for manual or automatic selection of fixation fasteners to be used with the trauma plate. In this manner, the software may automatically select fasteners that take into account the size and shape of the fractured components of the clavicle, the location and orientation of the fastener holes extending through the trauma plate, and the shape of the fasteners (e.g., screws) to improve fixation strength and avoid unnecessary compromise of the integrity of the clavicle.

[0237] After selection of the clavicle trauma plate template, fixation hole locations, and fixation fasteners, the software performs virtual bone plate placement, which includes positioning the clavicle trauma plate template on the 3D patchwork clavicle and manually or automatically deforming the clavicle trauma plate template to fit the external surface contour of the 3D patchwork clavicle, creating a patient-specific virtual 3D clavicle trauma plate with size, length, and contour dimensions. The software records the patient-specific clavicle trauma plate dimensions and converts these virtual dimensions into machine code that enables the generation of a tangible patient-specific clavicle trauma plate capable of rapid manufacturing.

[0238] Using patient-specific clavicle trauma plate dimensions, the software also receives anatomical data regarding the location and location of the patient's soft tissues, blood vessels, and nerves within the fractured clavicle region to construct an incision plan. The incision plan is preoperative and suggests a surgical approach using one or more incisions to improve access to the fractured clavicle components while reducing the invasiveness of the surgical procedure, potentially shortening recovery time and reducing postoperative collateral trauma. Figure 134 shows a 3D patchwork clavicle with a surface colored to indicate muscle attachment locations to the patient's clavicle. As a result, patterned circles extending longitudinally along the 3D patchwork clavicle correspond to fixation fastener locations oriented primarily toward areas without muscle attachment.

[0239] After constructing the incision plan, the surgeon reviews the incision plan and makes any modifications prior to approving the plan. The incision plan may then be exported to an intraoperative surgical guidance system after approval. Similarly, the incision plan may be used to construct a tangible preoperative clavicle model that estimates the shape of the reconstructive clavicle components attached to each other, thereby simulating the patient's normal clavicle. This tangible clavicle model may then be used to test the fit of the clavicle trauma plate and to allow for any contour modifications by bending that the surgeon may require preoperatively. Alternatively, the tangible clavicle model may need to loosely contain the clavicle components to which one or more trauma plates are attached and to hold the clavicle components, allowing the surgeon to perform in vitro testing of the fit of the trauma plate and any ex vivo modifications of the trauma plate.

[0240] 128 and 129, an exemplary patient-specific clavicle trauma plate may be positioned intraoperatively under fluoroscopy. While exemplary techniques are discussed below with respect to attachment of a patient-specific clavicle trauma plate to a patient's clavicle or clavicle component, it is understood that the exemplary process is equally applicable to attachment of a non-patient-specific trauma plate to a clavicle, or more generally, to attachment of any trauma plate to any bone or fractured bone component.

[0241] FIG. 128 illustrates a process flow illustrating various steps included as part of a trauma plate placement system for intraoperatively positioning a patient-specific trauma plate via fluoroscopy, including establishing patient alignment through the use of preliminary planning data in conjunction with the placement of fiducial markers. More specifically, the preliminary planning data is loaded into the trauma plate placement system's software package and may include information about the patient's shape, bone and tissue shape, the location of each trauma plate, the type and location of fixation devices used to secure the trauma plate to the bone or bony component in question, and any other relevant information affecting the surgical site and technique. Fiducial markers used with fluoroscopy include, but are not limited to, optical IMUs, electromagnetic IMUs, etc. (However, the process flow of FIG. 128 references optical markers and positions them at known locations relative to the patient's anatomical landmarks). Using the fiducial markers and the patient's known anatomical location and dimensions, the trauma plate placement system registers the patient relative to a preoperative coordinate system. The fiducial markers are then tracked in space to provide feedback from the trauma plate placement system to the surgeon in alignment with the preoperative plan, which indicates the location of one or more incisions relative to a fixed patient frame of reference. Exemplary feedback systems that may be utilized as part of the trauma plate placement system include, but are not limited to, a visual display that is projected onto the surface of the patient to depict the location and length of each incision.

[0242] In the context of a fractured clavicle, where the clavicle is composed of separate bone components, the trauma plate placement system may also visually display identifying indicia on multiple clavicle components to indicate the order of assembly of the bone components. In an exemplary embodiment, the visual indicia includes a colored number displayed on each visible bone component. The colored number changes color depending on the orientation and position of the bone components relative to one another. In an exemplary embodiment, the first bone component is identified by a displayed number "1" projected onto the exterior surface. This displayed number "1" may be red, yellow, or green, depending on the bone orientation and position. A red number indicates incorrect orientation and position. Upon translation, if the surgeon has moved the bone component in the correct direction to achieve placement consistent with the preoperative plan, the indicia changes to yellow. Upon continued translation, if the correct position is achieved, the number turns green. This repositioning process is repeated for each clavicle component.

[0243] To provide the surgeon with this visual feedback regarding the position and orientation of the fractured bone components, the trauma plate placement system uses fluoroscopy to track the bone components in 3D space to identify whether the bone position and orientation are consistent with the preoperative plan. Prior to tracking, the bone components are registered using preoperative data, providing the surgeon with real-time updates regarding the correct position and orientation of the bone components via a projected display. As each bone fragment is tracked and finally attached to the clavicle trauma plate, the system uses fluoroscopic images to confirm the progress of the trauma plate placement, verifying the orientation and position of the plate, as well as the orientation and position of the fixation devices (e.g., screws) and the bone components. Finally, when the bone components are joined together via one or more clavicle trauma plates, the system displays a final indicia to the surgeon indicating that the procedure has met the objectives of the preoperative plan and can be completed.

[0244] Figure 130 is a process flow diagram illustrating the various steps included as part of a trauma plate placement system that uses ultrasound instead of fluoroscopy to intraoperatively position a patient-specific trauma plate. The above description of Figure 128 is comparable to the description of Figure 130, except that the system uses ultrasound instead of fluoroscopy to track the bony components, trauma plate, and fixation devices, and is incorporated herein by reference. As a result, redundant description has been omitted for the sake of brevity.

[0245] Creating a trauma plate placement guide Referring to FIG. 150 , an exemplary process and system for creating a patient-specific trauma plate placement guide is described. Those skilled in the art will appreciate that bone fractures at one or more locations can result in bone fragments that are separated from one another. As part of a reconstructive surgery to repair the bone, one or more trauma plates are used to hold these fragments in a fixed orientation. Reconstructive surgery uses original knowledge, rather than the patient's specific anatomy, to reassemble the bone. As a result, to the extent that a patient's bone anatomy has been altered from normal, the bone fragments have been significantly distorted, or there are a large number of bone fragments, surgeons use prior art trauma plates and attempt to adapt the bone fragments to the shape of the plate rather than adapting the shape of the plate to the bone fragments. The present process and system improve upon prior art trauma plate applications by creating trauma plate placement guides and customized trauma plates that adapt the trauma plate to the bone to replicate the original bone shape and orientation.

[0246] The exemplary system flow begins with receiving input data representative of fractured anatomy. For purposes of illustration only, the fractured anatomy includes a human skull. It is noted that the above process and system are equally applicable to other anatomy / bones, such as, but not limited to, bones of the arms, legs, and torso. In exemplary embodiments, the anatomy data input may be in the form of an x-ray, CT scan, MRI, or any other imaging data that can represent bone size and shape.

[0247] Input anatomical data is used to construct a three-dimensional virtual model of the fractured anatomy. By way of example, the input anatomical data may include a computed tomography scan of a fractured skull, which is processed by software to segment the scan and generate a three-dimensional model. Those skilled in the art are familiar with how to construct a three-dimensional virtual model using computed tomography. As a result, a detailed description of aspects of this process has been omitted for the sake of brevity.

[0248] Following generation of the 3D virtual model of the fractured skull, the software compares the 3D virtual model of the skull with data from a statistical atlas to determine the regions in the 3D virtual model where the skull is fractured. Specifically, the software extracts fracture regions using features extracted from the surface model of the input anatomy (e.g., surface roughness, curvature, shape index, bowing, and neighbor connectivity). The software then extracts and integrally fits the outlines of these fracture regions to find matching feature regions. The software also matches the fracture fragments with the atlas to indicate optimal locations for placing the matching fracture regions, thereby reconstructing normal anatomy.

[0249] After the software generates a reconstructed 3D virtual model of the fractured skull, buttresses may be manually and / or automatically positioned on the exterior of the reconstructed 3D virtual skull model. Automatic buttress placement is the result of programmed logic that maximizes bone fragment stability while minimizing the number of buttresses. As used herein, the term buttress (and its plural forms) refers to any support used to stabilize bone fragments relative to one another. In certain cases, when utilizing the manual buttress placement feature, the practical experience of a knowledgeable user, such as a surgeon, may complement or replace the logic. In either case, a series of buttresses are programmed into the software, allowing the software or a user of the software to select different buttresses for different applications. At the same time, buttress length may be manually or automatically manipulated based on the dimensions of the fracture and bone fragments.

[0250] Following the allocation and placement of the buttresses on the reconstructed 3D virtual skull model, the software records the software's dimensions and the contours of each buttress. This record includes information necessary for the fabrication of each buttress, or at least information that would assist a knowledgeable individual, such as a surgeon, in capturing existing buttresses and fitting each to the placement guide. In the context of molding the existing buttresses, the software extracts the contours of the reconstructed 3D virtual skull model to generate computer-aided design (CAD) instructions for creating one or more tangible models representing the reconstructed 3D skull model. These CAD instructions are sent to a rapid prototyping machine, which creates one or more tangible models representing the reconstructed 3D skull model. By recreating the appropriate anatomical surfaces in the tangible model, each buttress can be manually applied and fitted to the tangible model at the target location prior to implantation and fixation to the patient's skull.

[0251] Based on the location and length of any buttresses, the software extracts the contours of the reconstructed 3D virtual skull model to generate contour data for one or more patient-specific buttress placement guides. In particular, a placement guide for each buttress may be generated. In this manner, the placement guide includes a surface contour that matches the contours of the patient's skull in a single orientation. Assuming the location of the buttresses is known on the virtual model of the reconstructed skull, as well as the contours of the adjacent external skull surfaces, the software combines the two to create a patient-specific virtual placement guide. This virtual guide is output in the form of CAD instructions to a rapid prototyping machine for fabrication.

[0252] In this exemplary embodiment, the manufactured patient-specific placement guide includes an elongated handle configured to be grasped by the surgeon. Extending from the end of the elongated handle is a block C-shaped contour plate. The underside of the contour plate is concave to fit the convex shape of the skull where the buttress is to be positioned. While not required, both ends (or other portions) of the contour plate may be secured to the buttress. Alternatively, the contour plate may simply provide a working window for aligning and ultimately securing the buttress to the skull. Once the buttress is attached to the skull, the contour plate may be removed.

[0253] Customized cutting and placement guides and plates Referring to FIG. 151, the reconstruction of deformed, fractured, or incomplete anatomy is one of the complex problems facing healthcare providers. Abnormal anatomy may be the result of birth conditions, tumors, disease, or physical injury. As part of the treatment for various ailments, healthcare providers may find it advantageous to reconstruct or configure anatomy to facilitate treatment for a variety of conditions, including, but not limited to, fractures / comminuted fractures, bone degeneration, orthopedic implant revisions, primary orthopedic implants, and disease.

[0254] The present disclosure provides systems and methods for bone reconstruction and tissue reconstitution using bone grafts. To perform this reconstruction, the systems and associated methods utilize images of a patient's current anatomy to construct two virtual 3D models: (a) a first 3D model representative of the patient's current abnormal anatomy and (2) a second model representative of the patient's reconstructed anatomy. For a detailed description of using patient images (e.g., X-rays, CT scans, MRI images, etc.) to arrive at virtual models of the patient's abnormal and reconstructed anatomy, see the "Total Anatomy Reconstruction" section above. The systems and methods utilize these two 3D virtual models by constructing the system described in the "Total Anatomy Reconstruction" section in combination with constructing a 3D virtual model of one or more bones (i.e., donor bones) from which a bone graft can be obtained. As described in more detail below, the 3D virtual model of the patient's reconstructed abnormal anatomy is analyzed to generate a 3D virtual model of the bone graft required for reconstruction. This 3D virtual graft model is compared to the 3D virtual model of the donor bone to access one or more sites on the donor bone from which the bone graft can be harvested. After the harvest site is determined, cutting guides and graft placement guides are designed and fabricated to collect the bone graft and attach it to the reconstruction site.

[0255] For illustrative purposes, the present systems and methods are described in the context of facial reconstruction, where the donor bone includes the fibula. Those skilled in the art will recognize that the present systems and methods are applicable to any reconstructive surgery utilizing one or more bone grafts. Furthermore, while facial reconstruction and the fibula as the bone donor are discussed, those skilled in the art will understand that the exemplary systems and methods can be used with donor bones other than the fibula.

[0256] As a preliminary step to discussing exemplary systems and methods for use in reconstructive surgical planning and surgery using bone grafts, it is assumed that imaging of a patient's abnormal anatomy and generation of virtual 3D models of the patient's abnormal and reconstructed anatomy have been completed using the process described above in the "Total Anatomy Reconstruction" section. As a result, a detailed description of utilizing patient images to generate virtual 3D models of both the patient's abnormal and reconstructed anatomy has been omitted for the sake of brevity.

[0257] After creating the virtual 3D models of the patient's abnormal and reconstructed anatomy, the software compares the anatomy and highlights the areas that differ. In particular, common areas between the virtual 3D models indicate bones that will be preserved, while different areas indicate one or more areas to be reconstructed. The software extracts the inconsistent areas from the virtual 3D model of the patient's reconstructed anatomy and isolates these areas as separate 3D virtual models of the target bone graft. A preoperative planner, such as a surgeon, reviews the virtual 3D bone graft models and makes a decision regarding the bone or bones that are deemed most suitable for bone graft extraction.

[0258] Regardless of the logic used to initially select potential bones as graft candidates, the bones in question are imaged using conventional methods (e.g., X-ray, CT, MRI, etc.). Each imaged bone is then segmented and a virtual 3D model of the imaged bone is created using the process described in the "Whole Anatomy Reconstruction" section above. This 3D donor bone model is compared to the virtual 3D bone graft model to isolate common areas. Specifically, the software compares the surface contour of the 3D donor bone model with the surface contour of the virtual 3D bone graft model to identify common areas or areas with similar curvature. If no common or similar areas are found, the process can begin again by analyzing another potential donor bone. Alternatively, if one or more areas of commonality or similar curvature exist in the donor bone, these areas are highlighted on the 3D donor bone model. Specifically, the highlighted areas mimic the shape of the virtual 3D bone graft model. If the common region is determined to be suitable for bone graft harvesting, the software virtually harvests the bone graft as a virtual 3D model and applies the bone graft (having a specific / unique contour with respect to the donor bone) to the virtual 3D model of the patient's abnormal anatomy to identify any regions of the patient's abnormal anatomy that may need to be harvested as part of a potential fitting and reconstruction. In situations where application of the virtual 3D model of the harvested bone to the virtual 3D model of the patient's abnormal anatomy does not result in a satisfactory reconstruction, the process may be restarted at a bone selection point, or a different bone region may be harvested by restarting. However, assuming application of the virtual 3D model of the harvested bone to the virtual 3D model of the patient's abnormal anatomy results in a suitable fit, the system proceeds with designing a jig to facilitate bone graft harvesting and attaching the bone graft to the patient's remaining bone.

[0259] In this exemplary embodiment, the system generates and outputs the machine code necessary for a rapid prototyping machine, CNC machine, or similar equipment to manufacture the bone graft cutting guides and bone graft placement guides. To generate the output necessary to manufacture the bone graft cutting guides and bone graft placement guides, the system utilizes a virtual 3D model of the harvested bone applied to a virtual 3D model of the patient's abnormal anatomy.

[0260] In particular, the virtual 3D model of the harvested bone defines the boundaries of the virtual 3D cutting guide. Furthermore, in this exemplary context, it is intended to harvest a portion of the fibula to provide a bone graft. The virtual 3D cutting guide includes a window across which a cutting instrument (saw, cutting drill, etc.) can be moved to create the appropriate bone graft so that the appropriate portion of the fibula is harvested. The virtual 3D cutting guide not only needs to be shaped to create the appropriate bone graft contour, but also to specify the placement of the cutting guide on the patient's donor bone. More specifically, the placement of the cutting guide on the donor bone must ensure that the harvested bone includes the correct contour and exhibits the correct outline. As such, the underside of the virtual 3D cutting guide is designed to be a "negative" of the surface of the donor bone to which the cutting guide will be attached. Exemplary attachment techniques for securing the cutting guide to the donor bone include, but are not limited to, screws, dowels, and pins. To accommodate one or more of these attachment techniques, the virtual 3D cutting guide is designed to include one or more penetration orifices, excluding the window through which the surgical cutter will traverse. After the design of the virtual 3D cutting guide is complete, the system generates and outputs the necessary machine code for a rapid prototyping machine, CNC machine, or similar equipment to manufacture the bone graft cutting guide, and then manufactures the actual cutting guide.

[0261] In addition to the cutting guide, the software also designs one or more bone graft placement guides. The bone graft placement guides are patient-specific and conform to the patient's anatomy (both the donor bone and the residual bone to which the donor bone is attached) to ensure proper placement of the bone graft relative to the residual bone. In an exemplary embodiment, the bone graft placement guide is configured for mandibular reconstruction. To design the bone graft placement guide, the software constructs a composite model using a virtual 3D model of the harvested bone applied to a virtual 3D model of the patient's abnormal anatomy. This composite model is used to identify joints where the bone graft will blend with (and preferably join through bone ingrowth) the adjacent residual bone. At these joints, depending on various factors, such as the surgeon's preferences, the system identifies bone graft placement locations and, for each plate, identifies one or more guides to facilitate proper placement and fixation of the plate relative to the bone graft and residual bone.

[0262] Customized Trauma Plate Templates and Placement Guides FIG. 152 graphically illustrates an exemplary system and method for trauma plate templating in the form of a flow diagram. The system and method, which includes a computer and associated software, performs calculations to determine the optimal fit among a group of template trauma plates, thereby minimizing future shape changes that may be required to accommodate the trauma plate's fit to the patient's bone shape. In an exemplary embodiment, the system constructs a 3D model of the patient's fractured bone as a single bone, then creates a template trauma plate that fits the 3D model, defining the trauma plate's shape prior to implantation. In this manner, the final trauma plate shape is patient-specific, allowing for a closer fit to the patient's anatomy, eliminating ambiguity regarding trauma plate placement and reducing surgical time. The system is easily deployable in everyday clinical settings or in surgeons' offices.

[0263] Referring again to FIG. 152, the initial input to the system is any number of medical images showing fractured bones. By way of example, these medical images may be one or more of X-ray, ultrasound, CT, and MRI. The images of the fractured bones are analyzed by a human operator to select the fractured bone from among multiple possible programmed bones. Upon bone selection, the software utilizes the medical image data to create a 3D model of the fractured bone components (as described above with respect to FIG. 127 and its related description, which are incorporated herein by reference). These 3D bone models are then reduced (i.e., reassembled to form a patchwork of bones that are oriented and positioned such that they would connect together if they were part of a single, unfractured bone) to create a 3D patchwork bone model using bone data from the statistical atlas. Similarly, in combination with the 3D patchwork bone model, bone data from the statistical atlas is also used to generate a complete 3D bone model (unfractured) of the patient's bone in question by morphing the 3D patchwork bone model onto the complete unfractured bone model (referred to as the reconstructed bone model).

[0264] The software analyzes this reconstructed bone model to extract longitudinal curves along dominant dimensions (e.g., midline curves) while also extracting cross-sectional curves perpendicular to the dominant dimensions to extract design parameters for the trauma plate. From these design parameters, the software calculates multiple template trauma plates that most closely resemble the design parameters. These design parameters include the trauma plate length, the trauma plate's longitudinal curvature, the transverse curvature perpendicular to the longitudinal curvature, the transverse length, and bone fastener fixation points that ensure proper attachment and retention of the trauma plate to the fractured bone while minimizing interference with muscle attachment sites and nerve locations.

[0265] The reconstructed bone model is also utilized to generate a tangible 3D bone model. In an exemplary embodiment, the software is programmed to output the virtual reconstructed bone model as machine code, thereby enabling rapid prototyping of the 3D bone model in an additive or subtractive process. For purposes of this disclosure, an additive process includes 3D printing, in which a model is created from a blank starting canvas by adding material to form discrete layers or slices of the bone model that, when stacked by printing successive layers, constitute the final bone model. In contrast, a subtractive process begins with a solid block of material and uses machine code (e.g., CNC code) to mechanically remove material to arrive at a solid bone model. Those skilled in the art will appreciate that any number of processes are available for manufacturing a tangible bone model. Depending on the process selected, the software is programmed to convert the 3D virtual model into machine code to facilitate rapid prototyping and construction of the 3D bone model.

[0266] After construction of the 3D bone model, a template trauma plate may be constructed, machined, and selected based on software selection of a trauma plate shaped to best fit the patient's fractured bone. Immediately thereafter, the template trauma plate is fitted to the 3D bone model and further refined by manual bending to fit the trauma plate to the 3D bone model. After satisfactory fit between the trauma plate and the bone model, the trauma plate is considered patient-specific and, after sterilization, is ready for implantation into the patient.

[0267] Patient-specific hip cage templating and placement guide FIG. 153 graphically illustrates an exemplary system and method for hip cage templating and placement guidance in flow diagram form. The system and method, which includes a computer and associated software, performs calculations to determine the optimal fit among a group of template hip cages, minimizing future shape changes that may be required to accommodate the hip cage fit to the patient's bone shape. In an exemplary configuration, the system constructs a 3D model of the patient's hip (as a single bone in the case of fracture or degeneration) and then creates a template hip cage that fits the 3D model to define the shape of the hip cage prior to implantation. In this way, the final hip cage shape and attachment site are patient-specific, allowing for a closer fit to the patient's anatomy, eliminating ambiguity about hip cage placement and reducing surgical time. The system is easily deployable in a day-to-day clinical setting or in a surgeon's office.

[0268] Referring again to FIG. 153, the initial input for the system is any number of medical images showing the patient's hip joint (all or part of the pelvis). By way of example, these medical images may be one or more of an X-ray, ultrasound, CT, and MRI. The software constructs a 3D virtual bone model of the patient's hip joint by utilizing images of the hip bone (as described above with respect to FIGS. 1, 7, and related descriptions, which are incorporated herein by reference). The software then automatically labels this 3D bone model.

[0269] The software uses input from the statistical atlas (e.g., regions likely to contain specific landmarks) and local geometric analysis to calculate anatomical landmarks on the 3D bone model by comparison with the hip bone model in the statistical atlas. This calculation is specific to each landmark. For example, the approximate shape of the region is known, and the location of the landmark being searched for is known relative to the local shape characteristics. For example, locating the superior edge of the anterior labial groove point of the acetabulum is achieved by refining the search based on the approximate location of the superior edge of the anterior labial groove point in the statistical atlas. This process is repeated for each landmark of interest.

[0270] After automatically calculating the anatomical landmarks for the 3D bone model, the software analyzes the bone model to calculate which of a plurality of template hip cages best fits the anatomical landmarks. In addition to calculating which of a plurality of hip cages best fits the anatomical landmarks of the patient's hip, the software also calculates where the cage will attach to the patient's anatomy. Referring again to Figures 20 and 21, the relevant descriptions of which are incorporated herein by reference, the software is operative to determine where the cage will attach to the patient's anatomy as well as to generate a virtual 3D guide that can be used to output sufficient machine code to construct a tangible 3D placement guide for the revision cage.

[0271] A bone model of the patient's hip joint is also utilized to generate a tangible 3D bone model. In an exemplary embodiment, the software is programmed to output the virtual 3D bone model as machine code, thereby enabling rapid prototyping of the tangible 3D bone model in an additive or subtractive process. For purposes of this disclosure, an additive process includes 3D printing, in which a model is created from a blank starting canvas by adding material to form discrete layers or slices of the bone model that, when stacked by printing successive layers, constitute the final bone model. In contrast, a subtractive process begins with a solid block of material and uses machine code (e.g., CNC code) to mechanically remove material to arrive at a solid bone model. Those skilled in the art will appreciate that any number of processes are available for manufacturing a tangible bone model. Depending on the process selected, the software is programmed to convert the 3D virtual model into machine code to facilitate rapid prototyping and construction of the 3D bone model.

[0272] After construction of the 3D bone model, a template cage may be constructed, machined, and selected based on software selection of a cage shaped to best fit the patient's hip joint. Immediately thereafter, the template cage is fitted to the 3D bone model once and further refined by manual bending to fit the cage to the 3D bone model. After a satisfactory fit between the cage and bone model, the cage is considered patient-specific and, after sterilization, is ready for implantation into the patient.

[0273] IMU motion tracking FIG. 154 provides an overview of an exemplary system and process for tracking bone and soft tissue motion using an IMU that utilizes a computer and associated software. For example, this motion tracking can provide useful information about a patient's kinematics for use in preoperative surgical planning. For illustrative purposes, the present system and method are described in the context of tracking bone motion and obtaining the resulting soft tissue motion using a 3D virtual model that integrates bone and soft tissue. Those skilled in the art will recognize that the present system and method are applicable to any bone, soft tissue, or motion tracking endeavor. Furthermore, while bone and soft tissue motion tracking is discussed in the context of the knee joint or spine, those skilled in the art will understand that the exemplary system and meth...

Claims

1. a surgical navigation module, A microcomputer; an inertial sensing unit communicatively coupled to the microcomputer to sense changes in at least one of the orientation and position of the surgical navigation module; an ultra-wideband wireless unit communicatively coupled to the microcomputer for detecting changes in at least one of the orientation and position of the surgical navigation module; a housing containing the microcomputer, the inertial sensing unit, and the ultra-wideband radio unit; Equipped with The ultra wideband radio unit an ultra-wideband wireless transceiver; at least four ultra wideband antennas; the at least four ultra wideband antennas are fixed to the housing so that their orientations with respect to one another do not change; At least one ultra wideband radio antenna of the at least four ultra wideband radio antennas is not disposed on the same plane as the other ultra wideband radio antennas. Surgical navigation module.

2. The inertial sensing unit a three-axis accelerometer; A three-axis gyroscope, at least three triaxial magnetometers; The surgical navigation module of claim 1 , comprising:

3. The surgical navigation module of claim 2 , wherein the three-axis accelerometer comprises a plurality of three-axis accelerometers.

4. The surgical navigation module of claim 2 , wherein the three-axis gyroscope comprises a plurality of three-axis gyroscopes.

5. The surgical navigation module of claim 1 , wherein the at least four ultra-wideband wireless antennas are equidistantly spaced apart and not coplanar.

6. The surgical navigation module of claim 5 , wherein the at least four ultra-wideband wireless antennas are oriented in a tetrahedral orientation.

7. The surgical navigation module of claim 1 , wherein at least three of the at least four ultra-wideband wireless antennas are coplanar.

8. The surgical navigation module of claim 1 , operative to record changes in at least six degrees of freedom.

9. The surgical navigation module of claim 1 , operative to determine changes in both movement and orientation of the surgical navigation module.

10. The surgical navigation module of claim 1 , further comprising a multiplexer.

11. 2. The surgical navigation module of claim 1, wherein the microcomputer is programmed with a magnetic calibration algorithm that processes inputs from the at least three three-axis magnetometers to map a local magnetic field and normalizes outputs from the at least three three-axis magnetometers to account for directional dispersion and magnetic distortion of the local magnetic field.

12. 1. A surgical navigation system comprising a plurality of surgical navigation modules communicatively coupled to one another, each of the plurality of surgical navigation modules comprising: A microcomputer; an inertial sensing unit communicatively coupled to the microcomputer to sense changes in at least one of the orientation and position of the surgical navigation module; an ultra-wideband wireless unit communicatively coupled to the microcomputer for detecting changes in at least one of the orientation and position of the surgical navigation module; a housing containing the microcomputer, the inertial sensing unit, and the ultra-wideband radio unit; Equipped with The ultra wideband radio unit an ultra-wideband wireless transceiver; at least four ultra wideband antennas; At least three of the at least four ultra wideband antennas are on the same plane, and one of the antennas is not on the same plane. Surgical navigation system.

13. The inertial sensing unit a three-axis accelerometer; A three-axis gyroscope, at least three triaxial magnetometers; The surgical navigation system of claim 12, comprising:

14. The surgical navigation system of claim 13 , wherein the three-axis accelerometer comprises a plurality of three-axis accelerometers.

15. The surgical navigation system of claim 13 , wherein the three-axis gyroscope comprises a plurality of three-axis gyroscopes.

16. The surgical navigation system of claim 12 , wherein the at least four ultra-wideband wireless antennas are equidistantly spaced from one another and are not coplanar.

17. The surgical navigation system of claim 16, wherein the at least four ultra-wideband wireless antennas are oriented in a tetrahedral orientation.

18. The surgical navigation system of claim 12 , wherein the at least four ultra-wideband wireless antennas are fixed relative to the housing so that their orientations relative to one another do not change.

19. The surgical navigation system of claim 12 , wherein each of the plurality of surgical navigation modules operates to record changes in at least six degrees of freedom.

20. The surgical navigation system of claim 12 , wherein each of the plurality of surgical navigation modules is operative to determine changes in both movement and orientation of the plurality of surgical navigation modules.

21. The surgical navigation system of claim 12 , wherein each of the plurality of surgical navigation modules further comprises a multiplexer.

22. 13. The surgical navigation system of claim 12, wherein the microcomputer is programmed with a magnetostriction algorithm to process input from the at least three 3-axis magnetometers and adjust for magnetostriction.

Images