Robust Motion and Distortion Correction of Diffusion-Weighted MR Images

Jan Hering
E ective image-based correction of motion and other acquisition artifacts became an essential step in di usion-weighted Magnetic Resonance Imaging (MRI) analysis as the micro-structural tissue analysis advances towards higher-order models. These come with increasing demands on the number of acquired images and the di usion strength (bvalue) yielding lower signal-to-noise ratios (SNR) and a higher susceptibility to artifacts. These conditions, however, render the current image-based correction
more » ... hemes, which act retrospectively on the acquired images through pairwise registration, more and more ine ective. Following the hypothesis, that a more consequent exploitation of the di erent intensity relationships between the volumes would reduce registration outliers, a novel correction scheme based on memetic search is proposed. This scheme allows for incorporating all single image metrics into a multi-objective optimization approach. To allow a quantitative evaluation of registration precision, realistic synthetic data are constructed by extending a di usion MRI simulation framework by motion and eddycurrents-caused artifacts. The increased robustness and e cacy of the multi-objective registration method is demonstrated on the synthetic as well as in-vivo datasets at di erent levels of motion and other acquisition artifacts. In contrast to the state-ofthe-art methods, the average target registration error (TRE) remained below the single voxel size also at high b-values (3000 s · mm −2 ) and low signal-to-noise ratio in the moderately artifacted datasets. In the more severely artifacted data, the multi-objective method was able to eliminate most of the registration outliers of the state-of-the-art methods, yielding an average TRE below the double voxel size. In the in-vivo data, the increased precision manifested itself in the scalar measures as well as the ber orientation derived from the higher-order Neurite Orientation Dispersion and Density Imaging (NODDI) model. For the neuronal ber tracts reconstructed on the data after correction, the proposed method most closely resembled the ground-truth. The proposed multi-objective method has not only impact on the evaluation of higher-order di usion models as well as ber tractography and connectomics, but could also nd application to challenging image registration problems in general.
doi:10.11588/heidok.00022242 fatcat:gwwqbx4ylfdnvofq2lk6ytrpry