Development of a Femur Statistical Shape Model and Implementation of Landmark-Based Morphing

Institute

TUM Munich Institute of Robotics and Machine Intelligence (Institut)

Type

Semester Thesis / Master's Thesis /

Content

 

Description

Knowledge of femoral anatomy is indispensable when calculating the relative movement of femur and tibia during knee surgery. Commercial navigation systems obtain this anatomy by having the surgeon digitize a small number of bone landmarks with a tracked probe, then use a statistical shape model (SSM) to morph a mean bone shape onto these sparse points — avoiding the need for a preoperative CT scan.

This Forschungspraxis focuses on building this pipeline from scratch: constructing a statistical shape model of the femur from a public bone dataset, then implementing the morphing algorithm that fits the model to a small set of digitized landmarks, enabling patient-specific geometry to be reconstructed relative to a tracker without a CT scan.

Possible Work packages:

• Familiarization with statistical shape modelling and motion capture systems
• Development of a statistical shape model of the femur
• Implementation of the morphing algorithm to fit the SSM to sparse landmark points 
• Data acquisition with different knee mock-ups
• Validation of morphed geometry against ground-truth bone models 

Requirements

• Background in Mechanical, Mechatronics, or Medical Engineering
• Python programming skills 
• Basic understanding of lower limb anatomy 
• Familiarity with point cloud / mesh processing is a plus 
• Hands-on "maker" attitude and ability to work with measurements
• Bonus: knowledge of CAD

Possible start

sofort

Contact

Alexander Gérard
alexander.gerardtum.de

Announcement