Research on Spatially Adaptive High-Order Total Variation Model for Weak Fluorescence Image Restoration

doi:10.1007/s12204-018-2016-8

Abstract

Abstract: Confocal laser scanning microscopy (CLSM) has emerged as one of the most advanced fluorescence cell imaging techniques in the field of biomedicine. However, fluorescence cell imaging is limited by spatial blur and additive white noise induced by the excitation light. In this paper, a spatially adaptive high-order total variation (SA-HOTV) model for weak fluorescence image restoration is proposed to conduct image restoration. The method consists of two steps: optimizing the deconvolution model of the fluorescence image by the generalized Lagrange equation and alternating direction method of multipliers (ADMM); using spatially adaptive parameters to balance the image fidelity and the staircase effect. Finally, an comparison of SA-HOTV model and Richardson-Lucy model with total variation (RL-TV model) indicates that the proposed method can preserve the image details ultimately, reduce the staircase effect substantially and further upgrade the quality of the restored weak fluorescence image.

Key words: confocal microscopy| weak fluorescence| image restoration| spatially adaptive high-order total variation (SA-HOTV)

摘要： Confocal laser scanning microscopy (CLSM) has emerged as one of the most advanced fluorescence cell imaging techniques in the field of biomedicine. However, fluorescence cell imaging is limited by spatial blur and additive white noise induced by the excitation light. In this paper, a spatially adaptive high-order total variation (SA-HOTV) model for weak fluorescence image restoration is proposed to conduct image restoration. The method consists of two steps: optimizing the deconvolution model of the fluorescence image by the generalized Lagrange equation and alternating direction method of multipliers (ADMM); using spatially adaptive parameters to balance the image fidelity and the staircase effect. Finally, an comparison of SA-HOTV model and Richardson-Lucy model with total variation (RL-TV model) indicates that the proposed method can preserve the image details ultimately, reduce the staircase effect substantially and further upgrade the quality of the restored weak fluorescence image.

关键词: confocal microscopy| weak fluorescence| image restoration| spatially adaptive high-order total variation (SA-HOTV)

CLC Number:

TP 391.4

MA Jin (马进), XUE Teng (薛腾), SHAO Quanquan (邵全全), HU Jie (胡洁), WANG Weiming (王伟明) . Research on Spatially Adaptive High-Order Total Variation Model for Weak Fluorescence Image Restoration[J]. Journal of Shanghai Jiao Tong University (Science), 2018, 23(Sup. 1): 1-7.

References 17

[1]	SHAH S M, CRAWSHAW J P, BOEK E S. Threedimensional imaging of porous media using confocal laser scanning microscopy [J]. Journal of Microscopy,2017, 265(2): 261-271.
[2]	SHIMONUKAI T, YOSHIOKA M, YANAGIMOTO H. Estimation of PSF for a shaking blurred image restoration [J]. Electronics and Communications in Japan, 2014, 97(4): 49-56.
[3]	AGHAZADEH N, BASTANI M, SALKUVEH D K.Generalized Hermitian and skew-Hermitian splitting iterative method for image restoration [J]. Applied Mathematical Modelling, 2015, 39(20): 6126-6138.
[4]	CHEONG H, CHAE E, LEE E, et al. Fast image restoration for spatially varying defocus blur of imaging sensor [J]. Sensors, 2015, 15(1): 880-898.
[5]	KARAKATSANIS N A, TSOUMPAS C, ZAIDI H.Quantitative PET image reconstruction employing nested expectation-maximization deconvolution for motion compensation [J]. Computerized Medical Imaging and Graphics, 2017, 60: 11-21.
[6]	INGARAMO M, YORK A G, HOOGENDOORN E,et al. Richardson-Lucy deconvolution as a general tool for combining images with complementary strengths[J]. ChemPhysChem, 2014, 15(4): 794-800.
[7]	SLAVINE N V, GUILD J, MCCOLL R W, et al. An iterative deconvolution algorithm for image recovery in clinical CT: A phantom study [J]. Physica Medica,2015, 31(8): 903-911.
[8]	LAASMAA M, VENDELIN M, PETERSON P. Application of regularized Richardson-Lucy algorithm for deconvolution of confocal microscopy images [J]. Journal of Microscopy, 2011, 243(2): 124-140.
[9]	DEY N, BLANC-FERAUD L, ZIMMER C, et al.Richardson-Lucy algorithm with total variation regularization for 3D confocal microscope deconvolution[J]. Microscopy Research and Technique, 2006, 69(4):260-266.
[10]	LIU J, HUANG T Z, SELESNICK I W, et al. Image restoration using total variation with overlapping group sparsity[J]. Information Sciences, 2015, 295:232-246.
[11]	ZHOU W F, LI Q G. Poisson noise removal scheme based on fourth-order PDE by alternating minimization algorithm [C]// Abstract and Applied Analysis.[s.l.]: Hindawi Publishing Corporation, 2012: 1-14.
[12]	FIGUEIREDOMA T, BIOUCAS-DIAS J M. Restoration of Poissonian images using alternating direction optimization [J]. IEEE Transactions on Image Processing,2010, 19(12): 3133-3145.
[13]	KONG Y Y, QIAN X, FANG B, et al. Quality assessment of the deconvolution algorithms for fluorescence images [J]. Beijing Biomedical Engineering,2014, 33(4): 374-378 (in Chinese).
[14]	WALLACE W, SCHAEFER L H, SWEDLOW J R.A workingperson’s guide to deconvolution in light microscopy[J]. BioTechniques, 2001, 31(5): 1076-1097.
[15]	WANG H B, MILLER P C. Scaled heavy-ball acceleration of the Richardson-Lucy algorithm for 3D microscopy image restoration [J]. IEEE Transactions on Image Processing, 2014, 23(2): 848-854.
[16]	CHEN D Q, CHENG L Z. Spatially adapted regularization parameter selection based on the local discrepancy function for Poissonian image deblurring [J]. Inverse Problems, 2011, 28(1): 015004.
[17]	NIELSEN F, SUN K. Guaranteed bounds on the Kullback-Leibler divergence of univariate mixtures [J].IEEE Signal Processing Letters, 2016, 23(11): 1543-1546.