Convex optimisation for partial volume estimation in compressive quantitative MRI

Roberto Duarte, Zhouye Chen, Silvia Gazzola, Ian Marshall, Michael Davies, Yves Wiaux

Research output: Contribution to conferenceAbstractpeer-review

Abstract / Description of output

Based on the recently proposed compressive sensing framework for quantitative
MRI, a new approach for partial volume reconstruction is developed in this abstract. We first
formulate a convex optimisation problem for the recovery of a sparse matrix of coefficients in
a dictionary of measured temporal fingerprints associated with specific combinations of quantitative parameters of interest. Each column of the sought matrix represents a voxel in the volume under investigation and the sparsity of this column represents the number of active dictionary elements or partial volumes. In a second step, we employ the weighted k-
means algorithm to cluster the recovered coefficient matrix in parameter space and obtain the quantitative parameter maps. The proposed approach was validated through simulations, and its performance is competitive when compared to a state of the art …
Original languageEnglish
Publication statusPublished - 28 Apr 2017


Dive into the research topics of 'Convex optimisation for partial volume estimation in compressive quantitative MRI'. Together they form a unique fingerprint.

Cite this