Meeting Banner
Abstract #3575

Conditional Least Squares Estimation of Diffusion MRI Parameters

Jelle Veraart1, Sofie Van Cauter2, Stefan Sunaert2, Jan Sijbers1

1University of Antwerp, Antwerp, Belgium; 2Department of Radiology, University Hospitals of Leuven, Leuven, Belgium

Many diffusion models require highly diffusion-weighted images, which suffer from low signal-to-noise ratio, eddy current distortions, and subject motion. Prior to diffusion model fitting, those distortions should be corrected; That data correction alters the underlying Rician data distribution. Therefore, previously proposed methods remove the Rician bias might be suboptimal. We propose a conditional least squares estimator (CLS), which is from theoretically an equivalent to the maximum likelihood estimator (MLE). However, the CLS remains, unlike the MLE, in well-defined cases unbiased after data correction as it can rely on the linearity of the expectation operator in high SNR and homogeneous regions.

Abstract PDF E-Poster Video

Keywords

abstract account accounted accounting accuracy accurate adding addition aligned alter although appears apply approximation arbitrarily arising artifacts bias brain called central chosen compressing conditional considered consistency constraint corner corrected correction cube deconvolution defined dependency depends despite diffusion discussed distorted distortions distributed distribution eddy equally equivalent error estimation estimator estimators evaluated expectation favorable field firstly fitting highly homogeneous human identical important incorporate inherently intensity interpolated interpolating interpolation kept knowledge known kurtosis least likelihood limit linearity longer many material matrix might mild minimization model models modification modulated motion nevertheless noise nonlinear normality often operator ordinary orientation original outperforms overlooked performance polynomial positioned premises preserve previously prior priori properties property propose proposed rather reconstruct reconstructed reducing regularity require rice robust rotated samples scale scaled scaling secondly selected several severe simulated simulation spherical squared squares step still stretch studies suffer suffers term terms theorem theoretical theory towards unbiased unit variable variance varying view volumes whole