Register now After registration you will be able to apply for this opportunity online.
Investigating the kinematics of the scapular motion using non-invasive optical technologies
Accurate non-invasive assessment modalities that incorporate both scapular motion and its morphology are currently unavailable, presenting a clear need for sustainable clinical application. To address this need, the Laboratory for Movement Biomechanics (LMB) utilizes a unique optical 4D scanning system (SLOT) to estimate the underlying anatomical structures using non-invasive structured light to produce high-quality images of the human skin surface, both statically and dynamically. By utilizing the clear cutaneous surface contours surrounding the scapula, the application of this technology to the shoulder joint could allow a novel non-invasive and dynamic approach for estimating scapular kinematics that overcomes the challenges associated with soft-tissue artifacts. The key challenge in the development of this approach is the precise identification and tracking of relevant scapula landmarks, as well as soft tissue artifacts, all of which are expected to affect the accuracy of the SLOT-measured kinematics.
Keywords: Scapulae, kinematics, dynamic scapulae motion, 3D scanning, back shape, machine learning, VICON, OpenSIM.
The skin surface and the positions of anatomical landmarks are recorded with the SLOT system as a series of point clouds in 3D. The positions of the anatomical landmarks are also visible in the ground truth motion captures obtained with VICON. Joint kinematics are reconstructed in OpenSIM using the VICON motion captures. The reconstructed joint kinematics will then be used to train the SLOT machine learning models for determining the underlying skeletal kinematics and will also be used in k-fold analyses to validate the accuracy of the SLOT approach.
We have currently collected preliminary data from 10 healthy subjects, which will be used as training data. More specifically, scapular motion was assessed during different activities using the SLOT system and VICON.
The skin surface and the positions of anatomical landmarks are recorded with the SLOT system as a series of point clouds in 3D. The positions of the anatomical landmarks are also visible in the ground truth motion captures obtained with VICON. Joint kinematics are reconstructed in OpenSIM using the VICON motion captures. The reconstructed joint kinematics will then be used to train the SLOT machine learning models for determining the underlying skeletal kinematics and will also be used in k-fold analyses to validate the accuracy of the SLOT approach. We have currently collected preliminary data from 10 healthy subjects, which will be used as training data. More specifically, scapular motion was assessed during different activities using the SLOT system and VICON.
The goal of this master's thesis is to assess the feasibility of the SLOT system in identifying scapular motion by conducting data processing, developing algorithms, and establishing a machine learning pipeline using the non-invasively collected datasets from 10 healthy subjects. The tasks include:
- Processing of 3D point cloud data,
- Reconstruction of the location of the scapula underneath the skin during dynamic activities,
- Optimization to minimize the differences between anatomical landmarks captured with the SLOT system and those captured with VICON.
The goal of this master's thesis is to assess the feasibility of the SLOT system in identifying scapular motion by conducting data processing, developing algorithms, and establishing a machine learning pipeline using the non-invasively collected datasets from 10 healthy subjects. The tasks include: - Processing of 3D point cloud data, - Reconstruction of the location of the scapula underneath the skin during dynamic activities, - Optimization to minimize the differences between anatomical landmarks captured with the SLOT system and those captured with VICON.
If you are interested, please contact Dr. Bossuyt Fransiska (fransiska.bossuyt@hest.ethz.ch) with a short CV and transcript of records. The focus of the project can be adjusted to the student’s interest and experience (e.g. if you have an interest to gain experience with data collection etc.).
If you are interested, please contact Dr. Bossuyt Fransiska (fransiska.bossuyt@hest.ethz.ch) with a short CV and transcript of records. The focus of the project can be adjusted to the student’s interest and experience (e.g. if you have an interest to gain experience with data collection etc.).