Application of independent component analysis to 1H MR spectroscopic imaging exams of brain tumours

The low spatial resolution of clinical H-1 MRSI leads to partial volume effects. To overcome this problem, we applied independent component analysis (ICA) on a set of H-1 MRSI exams of brain turnours. With this method, tissue types that yield statistically independent spectra can be separated. Up to three components, corresponding to necrosis, tumoral tissue and healthy tissue have been detected inside turnours. In nonagressive turnours, the "necrotic" component was absent, confirming that only agressive turnours exhibit high levels of lipids. In conclusion, the ICA algorithm allows to find useful hidden components in turnours. The reliability and robustness of the results have also been investigated by means of bootstrapping combined with unsupervised clustering. A comparison of ICA with a method of curve resolution, MCR-ALS, has also been performed. (c) 2005 Elsevier B.V. All rights reserved.

[1]  Desire L. Massart,et al.  Multivariate peak purity approaches , 1996 .

[2]  G. Hagberg,et al.  From magnetic resonance spectroscopy to classification of tumors. A review of pattern recognition methods , 1998, NMR in biomedicine.

[3]  A W Simonetti,et al.  Automated correction of unwanted phase jumps in reference signals which corrupt MRSI spectra after eddy current correction. , 2002, Journal of magnetic resonance.

[4]  R. Tauler.  Multivariate curve resolution applied to second order data , 1995 .

[5]  Aapo Hyvärinen,et al.  Fast and robust fixed-point algorithms for independent component analysis , 1999, IEEE Trans. Neural Networks.

[6]  Christophe Ladroue,et al.  Independent component analysis for automated decomposition of in vivo magnetic resonance spectra , 2003, Magnetic resonance in medicine.

[7]  Tzyy-Ping Jung,et al.  Imaging brain dynamics using independent component analysis , 2001, Proc. IEEE.

[8]  Paulo J. G. Lisboa,et al.  Robust methodology for the discrimination of brain tumours from in vivo magnetic resonance spectra , 2000 .

[9]  John W. Sammon,et al.  A Nonlinear Mapping for Data Structure Analysis , 1969, IEEE Transactions on Computers.

[10]  Sylvie Grand,et al.  A new approach for analyzing proton magnetic resonance spectroscopic images of brain tumors: nosologic images , 2000, Nature Medicine.

[11]  Erkki Oja,et al.  Independent component analysis: algorithms and applications , 2000, Neural Networks.

[12]  Christopher M. Dobson,et al.  Resolution enhancement of protein PMR spectra using the difference between a broadened and a normal spectrum , 1973 .

[13]  D. van Ormondt,et al.  SVD-based quantification of magnetic resonance signals , 1992 .

[14]  R Huo,et al.  Improved DOSY NMR data processing by data enhancement and combination of multivariate curve resolution with non-linear least square fitting. , 2004, Journal of magnetic resonance.

[15]  Desire L. Massart,et al.  Resolution of multicomponent overlapped peaks by the orthogonal projection approach, evolving factor analysis and window factor analysis , 1997 .