STATISTICAL ALGORITHMS FOR MICROARRAY IMAGE ANALYSIS
internal IPLAB Report

Sebastiano Battiato, Gianpiero Di Blasi, Giovanni Maria Farinella, Giovanni Gallo, Giuseppe Claudio Guarnera
{battiato, gdiblasi, gfarinella, gallo}@dmi.unict.it
g.guarnera@studenti.unict.it

IPLab – Image Processing Laboratory
http://www.dmi.unict.it/~iplab
Dipartimento di Matematica e Informatica
University of Catania, Via Andrea Doria 6 – 95125, Catania (Italy)
REPORT

 

In this paper novel techniques for Microarray Image Analysis are proposed. In particular, we describe MIRA (Microarray Image Rotation Algorithm) and SGRIP (Statistical GRIdding Pipeline) statistical algorithms respectively devoted to restore the original microarray orientation and detect the correct geometrical information about each spot of input Microarray. Both solutions significantly improve the performances of the segmentation pipeline MISP (Microarray Image Segmentation Pipeline). MIRA, SGRIP and MISP modules have been developed as plug-ins for an advanced on-going framework for Microarray Image Analysis. Experiments confirm the effectiveness of the proposed techniques, in terms of visual and numerical data.

 

SGRIP Experiment Example 1

 

Input Sub-Microarray Overlay

Green channel

Red channel

Green Mask Foreground

Red Mask Foreground

Spot Guide Mask

 

Hh - Horizontal histogram

 

CHh - Correct Horizontal histogram obtained by subtracting the Median value in order to account the noise perturbation

Correct prorotype spot centers

Final Grid Guide Mask

 

SGRIP Experiment Example 2

Input Sub-Microarray Overlay

Green channel

Red channel

Spot Guide Mask

Green Local Background Mask

Red Local Background Mask

 

Hh - Horizontal histogram

The outlined line shows the Median value

 

CHh - Correct Horizontal histogram

Yellow: prototype points abtained by Grid Finding Algorithm

Magenta: prototype points recovered by Grid Correction Algorithm

Detected final spot centers overlapped with the original signals

Initial Grid Guide Mask

Refined Grid Guide Mask

Final Prototypes

 

 

Some Screenshots of our Microarray Image analysis Framework

The Input Microarray is rotated of 13° degrees

        

Correction of rotation obtained using MIRA algorithm

 

Input Microarray Visualization

        

Microarray Channels Visualization

Spot Guide Mask obtained by MISP

Grid Guide Mask obtained by SGRIP Algorithms

Foreground Masks obtained by MISP

Local Background Masks obtained by MISP