Taylor Group / Dual-Plane FluoroscopeOpen OpportunitiesUnderstanding knee kinematics is a key requirement for understanding the processes occurring during injury or pathology as well as their remedies. Compared to optical systems, x-ray fluoroscopy directly measures the joint kinematics without soft-tissue artifacts and is thus the method of choice whenever such high performance is required.
To extract the 3D knee kinematics the rendering of each bone (3D geometry acquired independently by e.g. CT) is matched to the x-ray image in a process called 2D-3D pose estimation or 'image registration'. Current manual pose-estimation methods are time-consuming, expensive, and prone to operator bias. For example, a 10-second trial measurement acquired at 30 Hz consists of about 300 images and takes an experienced operator about 1500 minutes to match manually.
Since most studies often consist of thousands of images, an automated way of performing pose estimation to assist or replace manual alignment becomes crucial.
- Biomedical Engineering, Image Processing
- ETH Zurich (ETHZ), Master Thesis, Semester Project
| Knee kinematics is critical for diagnosing pathologies such as osteoarthritis and providing guidance for implant design. Estimating knee kinematics requires aligning a model with a target X-ray image. This estimation process, often implemented by human labor, can be very time-consuming. This research aims to use a deep learning network to estimate the pose (kinematics) from X-ray images, partially replacing manual labor. Such a network should predict a pose from a current fluoroscopic image. By the end of this project, a robust pipeline should be completed, achieving baseline performance to provide convincing pose estimation for images from different modalities (single-plane system & dual-plane system; natural bone model & implant model). - Biomechanics, Biomedical Engineering, Computer Vision
- ETH Zurich (ETHZ), Internship, Master Thesis, Semester Project
|
|