![](bg4.png)
Table 1: Accuracy of spot addressing in terms of the RMSE (in pixels) for the proposed automatic algorithm and the UCSF-
Spot algorithm described in (Jain et al., 2002).
Image ID # spots
RMSE for the RMSE for
proposed method UCSF-Spot
RMSE
x
RMSE
y
Total RMSE RMSE
x
RMSE
y
Total RMSE
lc7b070rex2 (Alizadeh et al., 2000) 9216 1.56 1.52 2.18 44.21 4.97 44.49
lc7b017rex2 (Alizadeh et al., 2000) 9216 1.04 1.88 2.15 66.89 10.80 67.75
lc7b0104rex2 (Alizadeh et al., 2000) 9216 0.95 1.38 1.68 70.23 8.67 70.76
21028 (Subramanian et al., 2005) 43008 1.14 1.45 1.85 49.12 1.53 49.14
16275 (Subramanian et al., 2005) 45312 1.93 2.00 2.78 10.40 11.90 15.80
43957 (Subramanian et al., 2005) 43008 1.14 1.45 1.85 3.40 1.90 3.89
41602 (Subramanian et al., 2005) 43008 1.19 1.28 1.75 6.42 10.57 12.36
15739 (Arbeitman et al., 2002) 9216 1.76 1.66 2.42 7.67 6.45 10.02
Antoniol, 2006) also tested their method on images
21028, 16275, 43957, 41602 and 15739 reporting
higher RMSEs than the ones obtained by the method
proposed in this paper (image ID 51509 could not
be tested because it is no longer available for down-
load from SMD). The whole algorithm takes approx-
imately 15 seconds for a typical subgrid like this on a
1.6 GHz AMD-64 under Matlab and Linux, including
I/O operations.
4 CONCLUSIONS
In this paper an automatic approach is proposed to
address the location of microarray subgrid spot cen-
ters. The method relies on the assumption that spotted
microarray images can be regarded as regular texture
images and consequently texture spatial characteriza-
tion techniques are suitable to be applied. This is due
to the regularity and pseudo-periodicity exhibited by
microarray images.
Experimental results on synthetic and real images
show that the proposed method outperforms the ones
provided by a state-of-the-art microarray analysis tool
(namely the UCSF-Spot) especially when large im-
age rotations and unequal row and column spacings
are present. The present authors believe that the
method yields promising results improving accuracy
over widely used tools available in the literature.
REFERENCES
Alizadeh, A. A., Eisen, M. B. et al (2000). Distinct types
of diffuse large B-cell lymphoma identified by gene
expression profiling. Nature, 403(6769):503–511.
Amador J. J. (2005). Markov random field approach to re-
gion extraction using tabu search. J. Vis. Commun.
Image R., 16:134–158.
Arbeitman, M. N. et al (2002). Gene expression during
the life cycle of Drosophila melanogaster. Science,
297:2270–2275.
Carstensen J. M. (1996). An active lattice model in a
Bayesian framework. Comput. Vis. Image Und.,
63(2):380–387.
Ceccarelli, M. and Antoniol, G. (2006). A deformable
grid-matching approach for microarray images. IEEE
Trans. Image Proc., 15(10):3178–3188.
Demeter, J., Beauheim, C. et al (2007). The Stanford Mi-
croarray Database: implementation of new analysis
tools and open source release of software. Nucleic
Acids Res., 35(Database Issue):D766–770.
Eisen, M. (1999). Scanalyze. http://rana.lbl.gov/EisenSoft-
ware.html.
Geman, S. and Geman, D. (1984). Stochastic relaxation,
gibbs distributions and the bayesian restoration of im-
ages. IEEE Trans. P.A.M.I., 6:721–741.
Gonzalez, R. and Woods, R. (2002). Digital image process-
ing. Prentice Hall, 2nd. edition.
Hartelius, K. and Carstensen, J. M. (2003). Bayesian grid
matching. IEEE Trans. P.A.M.I., 25(2):162–173.
Heyer, L. J., Moskowitz, D. Z. et al (2005). Magic tool:
integrated microarray data analysis. Bioinformatics,
21(9):2114–2115.
Jain, A. N., Tokuyasu, T. A. et al (2002). Fully automatic
quantification of microarray image data. Genome Re-
search, 12:325–332.
Lin, H.-C., Wang, L.-L., and Yang, S.-N. (1997). Extracting
periodicity of a regular texture based on autocorrela-
tion functions. Pat. Recognition Letters, 18:433–443.
Liu, Y., Collins, R. T., and Tsin, Y. (2004). A computa-
tional model for periodic pattern perception based on
frieze and wallpaper groups. IEEE Trans. P.A.M.I.,
26(3):354–371.
Subramanian, S., West, R. B. et al (2005). The gene expres-
sion profile of extraskeletal myxoid chondrosarcoma.
J. Pathol., 206:433–444.
Yang, Y., Buckley, M. et al (2000). Comparison of
methods for image analysis on cDNA microarray
data. Tech. Report #584, Dep. of Stat., UCB. URL:
http://www.stat.berkeley.edu/users/terry/zarray/Tech-
Report/584.pdf.
SIGMAP 2008 - International Conference on Signal Processing and Multimedia Applications
224