Abstract #4940

Multiscale Spherical Mean Value based background field removal method for Quantitative Susceptibility Mapping

Carlos Milovic^1,2,3, Christian Langkammer⁴, Sergio Uribe^2,3,5, Pablo Irarrazabal^1,3,6, Julio Acosta-Cabronero⁷, and Cristian Tejos^1,2,3

¹Electrical Engineering, Pontificia Universidad Catolica de Chile, Santiago, Chile, ²Biomedical Imaging Center, Pontificia Universidad Catolica de Chile, Santiago, Chile, ³Millennium Nucleus for Cardiovascular Magnetic Resonance, Santiago, Chile, ⁴Department of Neurology, Medical University of Graz, Graz, Austria, ⁵Radiology, Pontificia Universidad Catolica de Chile, Santiago, Chile, ⁶Institute for Biological and Medical Engineering, Pontificia Universidad Catolica de Chile, Santiago, Chile, ⁷Wellcome Centre for Human Neuroimaging, UCL Queen Square Institute of Neurology, University College London, London, United Kingdom

We present a multiscale SMV implementation (MSMV) for background field removal in QSM. We use a combined redundant Laplacian decomposition and Laplacian pyramid approach with fuzzy masks to remove background fields and reconstruct the local field. We tested this algorithm against PDF, LBV, ESHARP and VSHARP in analytic phantom and in vivo experiments. Experiments show MSMV’s accuracy to be in par with VSHARP, with an order of magnitude speed gain. MSMV achieved results with less harmonic remnants in our in vivo test. This multiscale approach may also be extended to the susceptibility inversion problem, with adaptive preconditioners and weighting.

This abstract and the presentation materials are available to members only; a login is required.

Join Here