Abstract
Introduction
The objective of this study is to assess the use of ultrasound (US) as a radiation free imaging modality to reconstruct three-dimensional knee anatomy.
Methods
An OEM US system is fitted with an electromagnetic (EM) tracker that is integrated into the US probe, allowing for 3D tracking of probe and femur and tibia. The raw US RF signals are acquired and using real time signal processing, bone boundaries are extracted. Bone boundaries are then combined with the EM sensor information in a 3D point cloud for both femur and tibia. Using a statistical shape model, the patient specific surface is reconstructed by optimizing bone geometry to match the point clouds. An accuracy analysis was then conducted for 11 cadavers by comparing the 3D US models to those created using CT scans.
Results
The results revealed the US bone models were accurate compared to the CT models (Mean RMS: femur: 1.03±0.15 mm, tibia:1.11± 0.13). Also, femoral landmarking proved to be accurate (transepicondylar axis: 1.07±0.65°, Posterior condylar axis: 0.73±0.41° Distal condylar axis: 1.12±0.89°, Medial AP: 1.39±1.18 mm, Lateral AP: 1.56±1.15 mm, TEA width: 1.2±0.87 mm). Tibial landmarking errors were slightly higher (Posterior slope axis: 2 ±1.19° and Tubercle axis: 1.8±1.37°). The models were then used to evaluate implant sizing as, 90% of the femurs and 60% of the tibias were sized correctly, while the others were off only one size.
Discussion
The 3D US bone models were proven to be accurate compared to CT and can be used for preoperative planning. 3D ultrasound is radiation free and offers numerous clinical opportunities for bone creation in minutes during their office visit, surgeon-patient pre-operative planning, implant sizing and selection, 3D dynamic ligament balancing and intra-operative registration for use with robots and navigation systems.
