Tessamma, Thomas; Deepa, J(The International Institute for Science, Technology and Education (IISTE), 2013)
[+]
[-]
Abstract:
As the technologies for the fabrication of high quality microarray advances rapidly, quantification of
microarray data becomes a major task. Gridding is the first step in the analysis of microarray images for
locating the subarrays and individual spots within each subarray. For accurate gridding of high-density
microarray images, in the presence of contamination and background noise, precise calculation of
parameters is essential. This paper presents an accurate fully automatic gridding method for locating
suarrays and individual spots using the intensity projection profile of the most suitable subimage. The
method is capable of processing the image without any user intervention and does not demand any input
parameters as many other commercial and academic packages. According to results obtained, the accuracy
of our algorithm is between 95-100% for microarray images with coefficient of variation less than two.
Experimental results show that the method is capable of gridding microarray images with irregular spots,
varying surface intensity distribution and with more than 50% contamination
Description:
Computer Engineering and Intelligent Systems,Vol.4, No.1, 2013