Title: Markerless Bone Tracking
Locations: LaTIM, INSERM UMR 1101, Brest, France
Co-supervision: Valérie BURDIN & Guillaume DARDENNE, LaTIM, Brest, Jocelyne TROCCAZ, TIMC-IMAG, Grenoble
Starting date: May 2018 for 18 months
Take-home salary: 1900 € / month
Context and Objectives
Musculoskeletal pathologies have multiple causes (trauma, neuromuscular, and degenerative). It impacts the
patient’s daily life by limiting the movement of their locomotor system and severe pain is often coupled with. In
this framework, the motion analysis is of great interest for several applications: in functional rehabilitation to
detect any impairment and understand it, in orthopedic surgery to diagnose a pathology or to acquire essential
data during an assisted surgical procedure. In these cases, optical markers are often used and placed either on
the skin to avoid any incision or directly on the bone structures during surgery. An optical camera system is then
used to locate these markers and thus determine the position of bone structures during a motion.
These solutions involve very sophisticated systems (as VICON) but they remain expensive, not versatile, and the
accuracy at the joint level is not sufficient due to soft tissues artefact.
Methodology
In order to tackle the problem, the project’s objective will be to design and implement an innovative and
portable solution, based on the markerless methods of the literature. The main joint concerned will be the knee.
Two main technical issues to be solved:
1) Determination of a joint congruence constraint model for the knee
2) Real-time markerless tracking of the joint inferring the external locations of the limb to internal joints via
the model.
Deliverables
Three outputs are expected for the project:
- Internal/external statistical model development
- Non-invasive tracking algorithm implementation
- Validation of the model thanks to cadaveric studies.
Candidate profile
- PhD in computer vision or medical imaging
- Solid knowledge in image processing and/or applied mathematics
- Biomechanics or robotics knowledge is an advantage
- Strong background in object-oriented programming language (C++, Java), python
- Knowledge of VTK, ITK and Scalismo libraries is an advantage
- English speaking, reading and writing is mandatory – French speaking is an advantage
- Good communication skills and ability to work in a team.
Interested candidates (Please use subject line PostDocMBT in your message)
Send to both valerie.burdin@imt-atlantique.fr and guillaume.dardenne@chu-brest.fr
- Detailled CV and Motivation letter
- Reviewers’ report of your PhD Thesis
- Main diplomas
- All transcripts of the 2 years MSc or DNM, or Engineer’s School.
- Two recommendation letters.