Sequential 3D analysis of patellofemoral kinematics from biplanar x-rays: In vitro validation protocol

Scan-Free and Fully Automatic Tracking of Native Knee Anatomy from Dynamic Stereo-Radiography with Statistical Shape and Intensity Models

Kinematic tracking of native anatomy from stereo-radiography provides a quantitative basis for evaluating human movement. Conventional tracking procedures require significant manual effort and call for acquisition and annotation of subject-specific volumetric medical images. The current work introduces a framework for fully automatic tracking of native knee anatomy from dynamic stereo-radiography which forgoes reliance on volumetric scans. The method consists of three computational steps. First, captured radiographs are annotated with segmentation maps and anatomic landmarks using a convolutional neural network. Next, a non-convex polynomial optimization problem formulated from annotated landmarks is solved to acquire preliminary anatomy and pose estimates. Finally, a global optimization routine is performed for concurrent refinement of anatomy and pose. An objective function is maximized which quantifies similarities between masked radiographs and digitally reconstructed radiographs produced from statistical shape and intensity models. The proposed framework was evaluated against manually tracked trials comprising dynamic activities, and additional frames capturing a static knee phantom. Experiments revealed anatomic surface errors routinely below 1.0 mm in both evaluation cohorts. Median absolute errors of individual bone pose estimates were below 1.0∘ or mm for 15 out of 18 degrees of freedom in both evaluation cohorts. Results indicate that accurate pose estimation of native anatomy from stereo-radiography may be performed with significantly reduced manual effort, and without reliance on volumetric scans.

The Influence of Mathematical Definitions on Patellar Kinematics Representations

A correlation between patellar kinematics and anterior knee pain is widely accepted. However, there is no consensus on how they are connected or what profile of patellar kinematics would minimize anterior knee pain. Nevertheless, answering this question by merging existing studies is further complicated by the variety of ways to describe patellar kinematics. Therefore, this study describes the most frequently used conventions for defining patellar kinematics, focusing on the rotations. The similarities and differences between the Cardan sequences and angles calculated by projecting axes are analyzed. Additionally, a tool is provided to enable the conversion of kinematic data between definitions in different studies. The choice of convention has a considerable impact on the absolute values and the clinical characteristics of the patello-femoral angles. In fact, the angles that result from using different mathematical conventions to describe a given patello-femoral rotation from our analyses differ up to a Root Mean Squared Error of 111.49° for patellar flexion, 55.72° for patellar spin and 35.39° for patellar tilt. To compare clinical kinematic patello-femoral results, every dataset must follow the same convention. Furthermore, researchers should be aware of the used convention’s implications to ensure reproducibility when interpreting and comparing such data.

Influence of patellar implantation on the patellofemoral joint of an anatomic customised total knee replacement implant: A case study

Few studies have been conducted to investigate kinematics and kinetics of the patellofemoral joint under physiological muscle forces and ankle joint loads. In this study, a preliminary design of a customised total knee implant was proposed and created. To compare the influences of different patella treatment scenarios, a dynamic knee simulation model was created with patient-specific muscle forces and ankle joint loads that are calculated from an OpenSim musculoskeletal model. The goal is to improve patellar implant-bone connection and restore patellofemoral joint mobility. Identical dynamic boundary conditions were applied on an unresurfaced patella and three different dome-shaped patellar implants. It was found that the unresurfaced patella and patellar implants resulted in different motions of patellar internal rotation and medial tilt. The size of the dome-shaped patellar implant affected the motion and loading of the patellofemoral joint. When the exposed patella bone was not fully covered by the patellar implant, the patella bone then contacted the femoral component during knee flexion. This would most likely lead to anterior knee pain and subsequent revision.

Radiation dose and magnification in pelvic X-ray: EOS™ imaging system versus plain radiographs

BACKGROUND

In plain pelvic X-ray, magnification makes measurement unreliable. The EOS™ (EOS Imaging, Paris France) imaging system is reputed to reproduce patient anatomy exactly, with a lower radiation dose. This, however, has not been assessed according to patient weight, although both magnification and irradiation are known to vary with weight. We therefore conducted a prospective comparative study, to compare: (1) image magnification and (2) radiation dose between the EOS imaging system and plain X-ray.

HYPOTHESIS

The EOS imaging system reproduces patient anatomy exactly, regardless of weight, unlike plain X-ray.

MATERIAL AND METHOD

A single-center comparative study of plain pelvic X-ray and 2D EOS radiography was performed in 183 patients: 186 arthroplasties; 104 male, 81 female; mean age 61.3±13.7years (range, 24-87years). Magnification and radiation dose (dose-area product [DAP]) were compared between the two systems in 186 hips in patients with a mean body-mass index (BMI) of 27.1±5.3kg/m² (range, 17.6-42.3kg/m²), including 7 with morbid obesity.

RESULTS

Mean magnification was zero using the EOS system, regardless of patient weight, compared to 1.15±0.05 (range, 1-1.32) on plain X-ray (P<10^-5). In patients with BMI<25, mean magnification on plain X-ray was 1.15±0.05 (range, 1-1.25) and, in patients with morbid obesity, 1.22±0.06 (range, 1.18-1.32). The mean radiation dose was 8.19±2.63dGy/cm² (range, 1.77-14.24) with the EOS system, versus 19.38±12.37dGy/cm² (range, 4.77-81.75) with plain X-ray (P<10^-4). For BMI >40, mean radiation dose was 9.36±2.57dGy/cm² (range, 7.4-14.2) with the EOS system, versus 44.76±22.21 (range, 25.2-81.7) with plain X-ray. Radiation dose increased by 0.20dGy with each extra BMI point for the EOS system, versus 0.74dGy for plain X-ray.

CONCLUSION

Magnification did not vary with patient weight using the EOS system, unlike plain X-ray, and radiation dose was 2.5-fold lower.

LEVEL OF EVIDENCE

3, prospective case-control study.