Kinematics tracking is the process by which the motion of the joints is studied. This motion consists of relative rotation and translation of the joint bones. Joint motion analysis is used in diagnosis of joint pathology, as well as studying the normal joint function. Currently, fluoroscopy is used in joint kinematics tracking. We are researching the use of pulse-echo A-mode ultrasound for the bone motion tracking instead of the fluoroscopy to avoid its radiation. In this work we performed feasibility study using simulation, and concluded that it is feasible to perform knee motion tracking with accuracy of 2 mm. The idea of the proposed system is to attach a number of single-element ultrasound transducers to a brace as shown in Figure 1. This brace will have a commercially available optical or electromagnetic tracking system's probe attached to it to track the global motion of the brace. The ultrasound transducers will be responsible for transcutaneously detecting points over the surface of the bone. The bone's echo extracted from each signal at each transducer will be registered in the optical or electromagnetic tracker's coordinate frame to create a set of points acquired over the surface of the bone. These points represent the bone's position at that point of time. A 3D model of the bone is then registered to these points using the iterative closest point method (ICP) to estimate the bone's position. At each tracking step, the 3D model will be at a position close to the new position of the points set, because this process will be repeated at a rate of 100 Hz or more in order to ensure that the change in the bone's position between every two successive tracking steps is small enough to guarantee high tracking accuracy. In this work we simulated the mentioned process using real kinematics data obtained for a patient using fluoroscopy. 3D models of the proximal tibia and distal femur were segmented from CT scans of the patient's knee. These models were then moved using the kinematic data in incremental steps. Simulated points over the surface of the bones (simulating the points on the bone's surface to be acquired using ultrasound) were used to track the bones' simulated motion using another set of the bones 3D models which move only according to the registration with the simulated points. In other words, the tracking models follow the simulated points' motion. Simulation was performed using deep knee bend kinematics data.Introduction
Methods
In this work, we present the first real-time fully automatic system for reconstruction of patient-specific 3D knee bones models using ultrasound raw RF data. The system was experimented on two cadaveric knees, and reconstruction accuracy of 2 mm was achieved. To use the highest available contrast and spatial resolution in the ultrasound data, the raw RF signals were used directly to automatically extract the bone contours from the ultrasound scans. Figure 1 shows a sample ultrasound B-mode image for cadaver's distal femur, showing some of the scan lines raw RF signals as well as the final extracted contour using our method. An ultrasound machine (SonixRP, Ultrasonix Inc) was used to scan the knee joint and the RF data of the scans are acquired by custom-built (using Visual C++) software running on the ultrasound machine. An optical tracker (Polaris Spectra, Northern Digital Inc) was attached to the ultrasound probe to track its motion while being used in scanning. The scanning of the knee was performed at two flexion angles (full extension, and deep knee bend). At each position, the knee was fixed in order to collect scans that represent a partial surface of the bone (which will be later mutually registered to represent the whole bone's surface). Figure 4 shows fluoroscopy images of a patient's knee, showing the different articulating surfaces of the knee bones visible to the ultrasound at different flexion angles. Figure 5 shows a dissected cadaver's knee showing the articulating surfaces visible to ultrasound at 90 degrees flexion. The custom-built software collects the RF data synchronized with the probe tracking data for each ultrasound frame. Each frame of the RF data is then processed to extract the bone contour. The bone contours are automatically extracted from the RF data frame with frame rate of 25 frames per second. Figure 2 shows a flowchart for the contour extraction process. The extracted bone contours were then used by the our software, along with the ultrasound probe's tracking data, to reconstruct point clouds representing the bones' surfaces. These point clouds were then aligned to the mean model of the bone's atlas using ICP and integrated together to form 3D point cloud of the bone's surface. A 3D model of the bone is then reconstructed by morphing the mean model to match the point cloud. Figure 3 shows a flowchart for the point cloud and 3D model reconstruction process.Introduction
Methods
Computer assisted knee arthroplasty systems provide the surgeon with tools for planning the femoral and tibial cuts, automatic implant sizing, and precise guidance for the bone milling and sawing tools. These systems require 3D models of the patient’s proximal tibial epiphysis, and distal femoral epiphysis. Currently preoperative CT scans are used to construct these models. The high irradiation, financial and time cost of the CT motivated the research for an alternative. In this work we developed a system for reconstructing a 3D bone model from a set of points localized by the surgeon intra-operatively on the bone surface using an optical localizer. A training set of 314 dry femurs, and 314 dry tibias (200 males, and 114 females) of Caucasian ethnicity was CT scanned, and segmented to create 3D models for these bones. These models were then used to extract the modes of variation for the femurs and tibias within each gender. Using these modes of variation along with the average model for the training set, a new femoral or tibial epiphysis model can be reconstructed. This reconstruction is performed by optimizing the average model’s morphology along the modes of variation to create a 3D model that matches the point cloud localized on the surface of the bone. A set of 77 male and 71 female dry femur and tibia pairs was used to digitize a sparse point cloud on the knee joint using an optical localizer. These point clouds were then used to reconstruct their corresponding models using the aforementioned algorithm. An average error of 0.42 between the reconstructed and the CT models was obtained.
