Hierarchical non‐negative matrix factorization applied to three‐dimensional 3 T MRSI data for automatic tissue characterization of the prostate

In this study non‐negative matrix factorization (NMF) was hierarchically applied to simulated and in vivo three‐dimensional 3 T MRSI data of the prostate to extract patterns for tumour and benign tissue and to visualize their spatial distribution. Our studies show that the hierarchical scheme provides more reliable tissue patterns than those obtained by performing only one NMF level. We compared the performance of three different NMF implementations in terms of pattern detection accuracy and efficiency when embedded into the same kind of hierarchical scheme. The simulation and in vivo results show that the three implementations perform similarly, although one of them is more robust and better pinpoints the most aggressive tumour voxel(s) in the dataset. Furthermore, they are able to detect tumour and benign tissue patterns even in spectra with lipid artefacts. Copyright © 2016 John Wiley & Sons, Ltd.

[1]  S Van Huffel,et al.  Improved Lanczos algorithms for blackbox MRS data quantitation. , 2002, Journal of magnetic resonance.

[2]  Thomas Hambrock,et al.  In vivo assessment of prostate cancer aggressiveness using magnetic resonance spectroscopic imaging at 3 T with an endorectal coil. , 2011, European urology.

[3]  Arend Heerschap,et al.  Spatially matched in vivo and ex vivo MR metabolic profiles of prostate cancer – investigation of a correlation with Gleason score , 2013, NMR in biomedicine.

[4]  Michael W. Berry,et al.  Algorithms and applications for approximate nonnegative matrix factorization , 2007, Comput. Stat. Data Anal..

[5]  T. Scheenen,et al.  Mapping of prostate cancer by 1H MRSI , 2014, NMR in biomedicine.

[6]  C. Balch,et al.  AJCC Cancer Staging Manual. 6th ed , 2002 .

[7]  Sabine Van Huffel,et al.  Hierarchical non‐negative matrix factorization (hNMF): a tissue pattern differentiation method for glioblastoma multiforme diagnosis using MRSI , 2013, NMR in biomedicine.

[8]  S. Provencher.  Estimation of metabolite concentrations from localized in vivo proton NMR spectra , 1993, Magnetic resonance in medicine.

[9]  T. Scheenen,et al.  Prostate MRI and 3D MR spectroscopy: how we do it. , 2010, AJR. American journal of roentgenology.

[10]  H. Sebastian Seung,et al.  Learning the parts of objects by non-negative matrix factorization , 1999, Nature.

[11]  Nicolas Gillis,et al.  Accelerated Multiplicative Updates and Hierarchical ALS Algorithms for Nonnegative Matrix Factorization , 2011, Neural Computation.

[12]  Michael Garwood,et al.  Solvent Suppression Using Selective Echo Dephasing , 1996 .

[13]  Lucas C. Parra,et al.  Nonnegative matrix factorization for rapid recovery of constituent spectra in magnetic resonance chemical shift imaging of the brain , 2004, IEEE Transactions on Medical Imaging.

[14]  Andrzej Cichocki,et al.  Hierarchical ALS Algorithms for Nonnegative Matrix and 3D Tensor Factorization , 2007, ICA.

[15]  Thomas Hambrock,et al.  Prostate cancer: multiparametric MR imaging for detection, localization, and staging. , 2011, Radiology.

[16]  L. Egevad,et al.  The 2005 International Society of Urological Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic Carcinoma , 2005, The American journal of surgical pathology.

[17]  Daniel B. Vigneron,et al.  Advances in MR spectroscopy of the prostate. , 2008, Magnetic resonance imaging clinics of North America.

[18]  T. Scheenen,et al.  In vivo 1H MR spectroscopic imaging of aggressive prostate cancer: Can we detect lactate? , 2014, Magnetic resonance in medicine.

[19]  A. Heerschap,et al.  Simple baseline correction for 1H MRSI data of the prostate , 2012, Magnetic resonance in medicine.

[20]  Sabine Van Huffel,et al.  Tissue segmentation and classification of MRSI data using canonical correlation analysis , 2005, Magnetic resonance in medicine.

[21]  B Michael Kelm,et al.  Automated estimation of tumor probability in prostate magnetic resonance spectroscopic imaging: Pattern recognition vs quantification , 2007, Magnetic resonance in medicine.

[22]  Chris H. Q. Ding,et al.  Convex and Semi-Nonnegative Matrix Factorizations , 2010, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[23]  Sabine Van Huffel,et al.  Unsupervised Nosologic Imaging for Glioma Diagnosis , 2013, IEEE Transactions on Biomedical Engineering.

[24]  Arend Heerschap,et al.  Quality control of prostate 1 H MRSI data , 2013, NMR in biomedicine.

[25]  Jens Frahm,et al.  Clinical proton MR spectroscopy in central nervous system disorders. , 2013, Radiology.

[26]  Charles L. Lawson,et al.  Solving least squares problems , 1976, Classics in applied mathematics.

[27]  Paulo J. G. Lisboa,et al.  Convex Non-Negative Matrix Factorization for Brain Tumor Delimitation from MRSI Data , 2012, PloS one.

[28]  Arend Heerschap,et al.  Fast acquisition‐weighted three‐dimensional proton MR spectroscopic imaging of the human prostate , 2004, Magnetic resonance in medicine.