Target-to-Background Separation for Spectral Unmixing in In-Vivo Fluorescence Imaging

doi:10.1007/s12204-014-1548-9

Abstract

Abstract: We present a novel fluorescence spectral unmixing based on target-to-background separation preprocessing, which effectively separates the multi-target fluorescence from all background autofluorescence (BF) without any hardware-based BF acquisition and tissue specific BF estimation. Specifically, we first enhance the intrinsic accumulation contrast in target-to-background fluorescence using h-dome transformation; then separate multi-target fluorescence areas from the background in sparse multispectral data utilizing kernel maximum autocorrelation factor analysis; we further use fast marching-based image inpainting method to patch up the removed target fluorescence areas and reconstruct the multispectral BF; with the BF matrix being subtracted from the original data, the multi-target fluorophores are easily unmixed from the subtracted data using multivariate curve resolution-alternating least squares method. In two preliminary in-vivo experiments, the proposed method demonstrated excellent performance to unmix multi-target fluorescences while other state-of-art unmixing methods failed to get desired results.

摘要： We present a novel fluorescence spectral unmixing based on target-to-background separation preprocessing, which effectively separates the multi-target fluorescence from all background autofluorescence (BF) without any hardware-based BF acquisition and tissue specific BF estimation. Specifically, we first enhance the intrinsic accumulation contrast in target-to-background fluorescence using h-dome transformation; then separate multi-target fluorescence areas from the background in sparse multispectral data utilizing kernel maximum autocorrelation factor analysis; we further use fast marching-based image inpainting method to patch up the removed target fluorescence areas and reconstruct the multispectral BF; with the BF matrix being subtracted from the original data, the multi-target fluorophores are easily unmixed from the subtracted data using multivariate curve resolution-alternating least squares method. In two preliminary in-vivo experiments, the proposed method demonstrated excellent performance to unmix multi-target fluorescences while other state-of-art unmixing methods failed to get desired results.

CLC Number:

Q 632

ZHAO Yong (赵勇), HU Cheng (胡程), PENG Jin-liang*(彭金良), QIN Bin-jie* (秦斌杰). Target-to-Background Separation for Spectral Unmixing in In-Vivo Fluorescence Imaging[J]. Journal of shanghai Jiaotong University (Science), 2014, 19(5): 600-611.

References

[1] Leblond F, Davis S C, Vald′es P A, et al. Preclinical whole-body fluorescence imaging: Review of instruments, methods and applications [J]. Journal of Photochemistry and Photobiology B: Biology, 2010,98(11): 77-94.
[2] Xu H, Rice B W. In-vivo fluorescence imaging with a multivariate curve resolution spectral unmixing technique [J]. Journal of Biomedical Optics, 2009, 14(6):064011-1-064011-9.
[3] Bioucas-Dias J M, Plaza A, Dobigeon N, et al.Hyperspectral unmixing overview: Geometrical, statistical,and sparse regression-based approaches [J].IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2012, 5(2): 354-379.
[4] Woolfe F, Gerdes M, Bello M, et al. Autofluorescence removal by non-negative matrix factorization [J].IEEE Transactions on Image Processing, 2011, 20(4):1085-1093.
[5] Montcuquet A S, Herve L, Navarro F, et al. In vivo fluorescence spectra unmixing and autofluorescence removal by sparse nonnegative matrix factorization[J]. IEEE Transactions on Biomedical Engineering,2011, 58(9): 2554-2565.
[6] Smal I, Loog M, Niessen W, et al. Quantitative comparison of spot detection methods in fluorescence microscopy [J]. IEEE Transactions on Medical Imaging,2010, 29(2): 282-301.
[7] Nielsen A A. Kernel maximum autocorrelation factor and minimum noise fraction tansformations [J]. IEEE Transactions on Image Processing, 2011, 20(3): 612-624.
[8] Telea A. An image inpainting technique based on the fast marching method [J]. Journal of Graphics Tools,2004, 9: 25-36.
[9] Chambon S, Crouzil A. Similarity measures for image matching despite occlusions in stereo vision [J].Pattern Recognition, 2011, 44(9): 2063-2075.
[10] Hoyer P. Nonnegative matrix factorization with sparseness constraints [J]. Journal of Machine Learning Research, 2004, 5: 1457-1469.
[11] Chan T H, Ma W K, Chi C Y, et al. A convex analysis framework for blind separation of non-negative sources [J]. IEEE Transactions on Signal Processing,2008, 56(11): 5120-5134.
[12] Wang F Y, Chi C Y, Chan T H, et al. Nonnegative least-correlated component analysis for separation of dependent sources by volume maximization [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2010, 32(5): 875-888.
[13] Gu Z, Qin B. Nonrigid registration of brain tumor resection MR images based on joint saliency map and keypoint clustering [J]. Sensors, 2009, 9(12): 10270-10290.
[14] Demchenko A P. On the nanosecond mobility in proteins: Edge excitation fluorescence red shift of protein-bound 2-(p-toluidinylnaphthalene)-6-sulfonate[J]. Biophysical Chemistry, 1982, 15(2): 101-109.