Removing undersampling artifacts in DCE‐MRI studies using independent components analysis

In breast MRI mammography both high temporal resolution and high spatial resolution have been shown to be important in improving specificity. Adaptive methods such as projection reconstruction time‐resolved imaging of contrast kinetics (PR‐TRICKS) allow images to be reconstructed at various temporal and spatial resolutions from the same data set. The main disadvantage is that the undersampling, which is necessary to produce high temporal resolution images, leads to the presence of streak artifacts in the images. We present a novel method of removing these artifacts using independent components analysis (ICA) and demonstrate that this results in a significant improvement in image quality for both simulation studies and for patient dynamic contrast‐enhanced (DCE)‐MRI images. We also investigate the effect of artifacts on two quantitative measures of contrast enhancement. Using simulation studies we demonstrate that streak artifacts lead to pronounced periodic oscillations in pixel concentration curves which, in turn, lead to increased errors and introduce bias into heuristic measurements. ICA filtering significantly reduces this bias and improves accuracy. Pharmacokinetic modeling was more robust and there was no evidence of bias due to the presence of streak artifacts. ICA filtering did not significantly reduce the errors in the estimated pharmacokinetic parameters; however, the chi‐squared error was greatly reduced after ICA filtering. Magn Reson Med, 2008. © 2008 Wiley‐Liss, Inc.

[1]  Edna Furman-Haran,et al.  Parametric analysis of breast MRI. , 2002, Journal of computer assisted tomography.

[2]  D C Peters,et al.  Undersampled projection‐reconstruction imaging for time‐resolved contrast‐enhanced imaging , 2000, Magnetic resonance in medicine.

[3]  E Ramsay,et al.  Adaptive bilateral breast MRI using projection reconstruction time‐resolved imaging of contrast kinetics , 2006, Journal of magnetic resonance imaging : JMRI.

[4]  G P Liney,et al.  Differentiation of benign and malignant sub-1 cm breast lesions using dynamic contrast enhanced MRI. , 2004, Breast.

[5]  Pierre Comon,et al.  Independent component analysis, A new concept? , 1994, Signal Process..

[6]  J Velikina,et al.  Highly constrained backprojection for time‐resolved MRI , 2006, Magnetic resonance in medicine.

[7]  Anne L. Martel,et al.  Extracting parametric images from dynamic contrast-enhanced MRI studies of the brain using factor analysis , 2001, Medical Image Anal..

[8]  D. Barber The use of principal components in the quantitative analysis of gamma camera dynamic studies. , 1980, Physics in medicine and biology.

[9]  N. Hylton Dynamic contrast-enhanced magnetic resonance imaging as an imaging biomarker. , 2006, Journal of clinical oncology : official journal of the American Society of Clinical Oncology.

[10]  Scott Fields,et al.  Mapping pathophysiological features of breast tumors by MRI at high spatial resolution , 1997, Nature Medicine.

[11]  W. Kaiser,et al.  MR imaging of the breast: fast imaging sequences with and without Gd-DTPA. Preliminary observations. , 1989, Radiology.

[12]  N. Hylton,et al.  Diagnostic architectural and dynamic features at breast MR imaging: multicenter study. , 2006, Radiology.

[13]  Edmund R. Malinowski,et al.  Determination of the number of factors and the experimental error in a data matrix , 1977 .

[14]  Huanzhou Yu,et al.  Relaxation times of breast tissue at 1.5T and 3T measured using IDEAL , 2006, Journal of magnetic resonance imaging : JMRI.

[15]  S. Heywang,et al.  MR imaging of the breast with Gd-DTPA: use and limitations. , 1989, Radiology.

[16]  Lawrence Dougherty,et al.  Dynamic MRI with projection reconstruction and KWIC processing for simultaneous high spatial and temporal resolution , 2004, Magnetic resonance in medicine.

[17]  S Makeig,et al.  Analysis of fMRI data by blind separation into independent spatial components , 1998, Human brain mapping.

[18]  T W Redpath,et al.  The use of the Levenberg–Marquardt curve-fitting algorithm in pharmacokinetic modelling of DCE-MRI data , 2005, Physics in medicine and biology.

[19]  Ellen Warner,et al.  Surveillance of BRCA1 and BRCA2 mutation carriers with magnetic resonance imaging, ultrasound, mammography, and clinical breast examination , 2004, JAMA.

[20]  E. B. Andersen,et al.  Modern factor analysis , 1961 .

[21]  D. Chakrabarti,et al.  A fast fixed - point algorithm for independent component analysis , 1997 .

[22]  P S Tofts,et al.  Quantitative Analysis of Dynamic Gd‐DTPA Enhancement in Breast Tumors Using a Permeability Model , 1995, Magnetic resonance in medicine.

[23]  Christian Jutten,et al.  Blind separation of sources, part I: An adaptive algorithm based on neuromimetic architecture , 1991, Signal Process..

[24]  R. Paola,et al.  Factor Analysis of Dynamic Magnetic Resonance Imaging in Predicting the Response of Osteosarcoma to Chemotherapy , 1992, Investigative radiology.

[25]  A. Andersen,et al.  Principal component analysis of the dynamic response measured by fMRI: a generalized linear systems framework. , 1999, Magnetic resonance imaging.

[26]  H. Hotelling Analysis of a complex of statistical variables into principal components. , 1933 .

[27]  Tom Minka,et al.  Automatic Choice of Dimensionality for PCA , 2000, NIPS.

[28]  Gerrit Kateman,et al.  Aspects of pseudorank estimation methods based on the eigenvalues of principal component analysis of random matrices , 1994 .