Automated adaptive preconditioner for quantitative susceptibility mapping

Liu, Zhe; Yan, Wen; Spincemaille, Pascal; Zhang, Shun; Yao, Yihao; Nguyen, Thanh D.; Wang, Yi

doi:10.1002/mrm.27900

Cited by 9 publications

(21 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[55][56][57][58] One of these methods, the preconditioned total field inversion method estimates the susceptibility distribution in the entire image volume directly from the total field using a preconditioner to improve convergence speed. 55,59,60 Hence, we report a multiecho complex total field inversion (mcTFI) method to compute the susceptibility map directly from the acquired GRE images, and therefore bypass errors from a separate field fitting and use an improved signal model and the Gaussian noise property of the complex data. We compared mcTFI with the nonlinear preconditioned total field inversion (here referred to as nTFI), 60 to show improvements in susceptibility reconstruction, especially in regions with low SNR.…”

Section: Introductionmentioning

confidence: 99%

Multiecho complex total field inversion method (mcTFI) for improved signal modeling in quantitative susceptibility mapping

Yan

Spincemaille

Nguyen

et al. 2021

Magnetic Resonance in Med

Self Cite

View full text Add to dashboard Cite

Purpose: Typical quantitative susceptibility mapping (QSM) reconstruction steps consist of first estimating the magnetization field from the gradient-echo images, and then reconstructing the susceptibility map from the estimated field. The errors from the field-estimation steps may propagate into the final QSM map, and the noise in the estimated field map may no longer be zero-mean Gaussian noise, thus, causing streaking artifacts in the resulting QSM. A multiecho complex total field inversion (mcTFI) method was developed to compute the susceptibility map directly from the multiecho gradient echo images using an improved signal model that retains the Gaussian noise property in the complex domain. It showed improvements in QSM reconstruction over the conventional field-to-source inversion. Methods: The proposed mcTFI method was compared with the nonlinear total field inversion (nTFI) method in a numerical brain with hemorrhage and calcification, the numerical brains provided by the QSM Challenge 2.0, 18 brains with intracerebral hemorrhage scanned at 3T, and 6 healthy brains scanned at 7T. Results: Compared with nTFI, the proposed mcTFI showed more accurate QSM reconstruction around the lesions in the numerical simulations. The mcTFI reconstructed QSM also showed the best image quality with the least artifacts in the brains with intracerebral hemorrhage scanned at 3T and healthy brains scanned at 7T. Conclusion:The proposed multiecho complex total field inversion improved QSM reconstruction over traditional field-to-source inversion through better signal modeling. K E Y W O R D Sbrain imaging, nonlinear total field inversion, quantitative susceptibility mapping

show abstract

Section: Introductionmentioning

confidence: 99%

Multiecho complex total field inversion method (mcTFI) for improved signal modeling in quantitative susceptibility mapping

Yan

Spincemaille

Nguyen

et al. 2021

Magnetic Resonance in Med

Self Cite

View full text Add to dashboard Cite

show abstract

“…Both these methods use R2* or initial susceptibility estimates to estimate the covariance matrix and use these as preconditioners. It should be noted that preconditioned total field inversion in both adaptive and binary variants of preconditioner published (Liu et al, 2017(Liu et al, , 2020 have a similar rationale mathematically, although the implementation and resultant artifact incidence is different. In Chatnuntawech et al, a single-step QSM method is proposed that does not require separate background field removal and is shown to have a lower error than competing local field algorithms.…”

Section: Accessmentioning

confidence: 99%

“…The resulting method (preconditioned TFI or pTFI) has been shown to provide an accelerated algorithm convergence and reduce streaking and shadow artifacts without mask erosion. There are two published methods that propose preconditioned total field inversion ( Liu et al., 2017 ) ( Liu et al., 2020 ). Both these methods use R2∗ or initial susceptibility estimates to estimate the covariance matrix and use these as preconditioners.…”

Section: Introductionmentioning

confidence: 99%

“…In order to derive a proper weighting factor for spatially adaptive regularization, it is important to obtain information regarding the spatial content of the image. Research on preconditioners and regularization terms provide backing for choosing adaptive weighting factors that resemble the expected solution ( Kee et al., 2017 ; Liu et al., 2017 , 2020 ). Thus, the motivation for the above chosen weighting factor preceding the L2 regularization in TFIR is the observation that the point-wise inverse of the signal magnitude at later echo times (over 50 ms) resembles the susceptibility map.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Spatially Adaptive Regularization in Total Field Inversion for Quantitative Susceptibility Mapping

et al. 2020

Self Cite

View full text Add to dashboard Cite

Summary Adaptive Total Field Inversion is described for quantitative susceptibility mapping (QSM) reconstruction from total field data through a spatially adaptive suppression of shadow artifacts through spatially adaptive regularization. The regularization for shadow suppression consists of penalizing low-frequency components of susceptibility in regions of small susceptibility contrasts as estimated by R2∗ derived signal intensity. Compared with a conventional local field method and two previously proposed regularized total field inversion methods, improvements were demonstrated in phantoms and subjects without and with hemorrhages. This algorithm, named TFIR, demonstrates the lowest error in numerical and gadolinium phantom datasets. In COSMOS data, TFIR performs well in matching ground truth in high-susceptibility regions. For patient data, TFIR comes close to meeting the quality of the reference local field method and outperforms other total field techniques in both clinical scores and shadow reduction.

show abstract

“…[38][39][40][41][42][43][44][45][46] In addition, to reduce error propagation during the multiple steps of phase preprocessing and QSM dipole inversion, the integration of these steps has been proposed. [47][48][49][50][51][52][53][54] Such integration can be achieved through fitting both background and local susceptibility sources to the total field with preconditioning or regularization. 49,51 It can also be achieved by fitting the local susceptibility sources to the spherical mean value (SMV) filtered total field, in which the background field is intrinsically removed due to its harmonic property.…”

Section: Introductionmentioning

confidence: 99%

Single‐step calculation of susceptibility through multiple orientation sampling

et al. 2021

View full text Add to dashboard Cite

Quantitative susceptibility mapping (QSM) was developed to estimate the spatial distribution of magnetic susceptibility from MR signal phase acquired using a gradient echo (GRE) sequence. The field-to-susceptibility inversion in QSM is known to be ill-posed and needs numerical stabilization through either regularization or data oversampling. The calculation of susceptibility through the multiple orientation sampling (COSMOS) method uses phase data acquired at three or more head orientations to achieve a well-conditioned field-to-susceptibility inversion and is often considered the gold standard for in vivo QSM. However, the conventional COSMOS approach, here named multistep COSMOS (MSCOSMOS), solves the dipole inversion from the local field derived from raw GRE phase through multiple steps of phase preprocessing. Error propagations between these consecutive phase processing steps can thus affect the final susceptibility quantification. On the other hand, recently proposed single-step QSM (SSQSM) methods aim to solve an integrated inversion from unprocessed or total phase to mitigate such error propagations but have been limited to single orientation QSM. This study therefore aimed to test the feasibility of using single-step COSMOS (SSCOSMOS) to jointly perform background field removal and dipole inversion with multiple orientation sampling, which could serve as a better standard for gauging SSQSM methods. We incorporated multiple spherical mean value (SMV) kernels of various radii with the dipole inversion in SSCOSMOS. QSM reconstructions with SSCOSMOS and MSCOSMOS were compared using both simulations with a numerical head phantom and in vivo human brain data. SSCOSMOS permitted integrated background removal and dipole inversion without the need to adjust any regularization parameters. In addition, with sufficiently large SMV kernels, SSCOSMOS performed consistently better than MSCOSMOS in all the tested error metrics in our simulations, giving better susceptibility quantification and smaller reconstruction error. Consistent tissue susceptibility values were obtained between SSCOSMOS and MSCOSMOS.

show abstract

Automated adaptive preconditioner for quantitative susceptibility mapping

Cited by 9 publications

References 57 publications

Multiecho complex total field inversion method (mcTFI) for improved signal modeling in quantitative susceptibility mapping

Multiecho complex total field inversion method (mcTFI) for improved signal modeling in quantitative susceptibility mapping

Spatially Adaptive Regularization in Total Field Inversion for Quantitative Susceptibility Mapping

Single‐step calculation of susceptibility through multiple orientation sampling

Contact Info

Product

Resources

About