A Kalman Filtering Framework for Prospective Motion Correction

Julian Maclaren¹, Oliver Speck², Jrgen Hennig¹, Maxim Zaitsev¹

Tracking head position and prospectively adjusting the imaging volume is becoming an increasingly popular means of preventing motion artefacts in MRI. However, the technique imposes stringent requirements on the accuracy of the tracking system used. For high-resolution imaging, these requirements are difficult to achieve in practice. We present a Kalman-filtering based approach that improves pose estimation and prediction and enables estimation of residual errors. This makes retrospective correction of residual motion artefacts possible. The result is a reduction in the required accuracy of the tracking system itself.