PhD Defence Dennis Christie | Towards More Reliable Ultrasound-based Knee Joint Kinematic Estimation

Towards More Reliable Ultrasound-based Knee Joint Kinematic Estimation

Dennis Christie is a PhD student in the Department of Biomedical Device Design and Production. (Co)Promotors are prof.dr.ir. N.J.J. Verdonschot, Faculty of Engineering Technology and Radboud University Medical Center, prof.dr.ir. M. Sartori, Faculty of Engineering Technology, dr. G.V. Durandau, McGill University and dr.ir. R. Fluit, University of Groningen. 

Accurate measurement of knee motion is important for understanding joint function after injury, degeneration, or surgery. However, the motion of the tibia relative to the femur is difficult to measure because the bones are hidden beneath soft tissue. Existing methods require invasive markers, expose the subject to radiation, restrict movement, or remain affected by soft-tissue artifact. This dissertation investigates A-mode ultrasound as a safer alternative for estimating tibiofemoral kinematics.

A-mode ultrasound uses small sensors placed on the skin to measure the depth of the underlying bone surface. Combined with optical tracking, these measurements can be transformed into 3D bone-surface points and registered to a bone model to estimate joint motion. The method is promising, but its reliability depends on correct signal interpretation. Selecting the true bone peak from an ambiguous one-dimensional signal is a central challenge because errors propagate through bone registration and affect kinematic estimates.

The main contribution of this dissertation is not only the development of an ultrasound-based measurement system, but also the improvement of its reliability. The work shows that accurate registration is possible when bone-surface points are precise and well distributed; that A-mode ultrasound can measure bone depth accurately when the correct peak and speed of sound are known; and that reliability improves when more spatial context is added. This was achieved through denser transducer configurations and navigation guidance using B-mode ultrasound and optical tracking.

Together, these studies demonstrate a path toward more reliable, non-invasive, and radiation-free knee joint motion estimation. Navigation-guided initialization ultrasound measurement improves stability during dynamic movement and helps the estimated kinematics more closely reflect the true motion of the underlying bones.