A New RBF Neural Approach for Spot Classification in DNA-Microarray Images

STMicroelectronics Abstract. Spots in DNA-Microarray images exhibit several characteristics able to uniquely identify them for diagnostic purposes. In order to perform efficiently their classification a new tool based on Radial Basis Functions Artificial Neural Networks is presented. Classification is performed according to the following predefined spot classes: saturated spot, normal spot, donut spot, noisy saturated spot, noisy gaussian spot, noisy donut spot and background. These classes have been selected according to the more frequently observed morphology of spots. The neural system shows very good performance in terms of positive classification in typical DNA-Microarray

[1]  Mark Schena,et al.  DNA microarrays : a practical approach , 1999 .

[2]  C.H. Lee,et al.  A DNA-based pattern recognition technique for cancer detection , 2004, The 26th Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[3]  Jürgen Schürmann,et al.  Pattern classification , 2008 .

[4]  Mohamad T. Musavi,et al.  On the training of radial basis function classifiers , 1992, Neural Networks.

[5]  J.C. Patra,et al.  Neural network-based analysis of DNA microarray data , 2005, Proceedings. 2005 IEEE International Joint Conference on Neural Networks, 2005..

[6]  Steven L. Salzberg,et al.  Gene discovery in DNA sequences , 1999, IEEE Intell. Syst..

[7]  Mustafa Ozen,et al.  Artificial Neural Network Analysis of DNA Microarray-based Prostate Cancer Recurrence , 2005, 2005 IEEE Symposium on Computational Intelligence in Bioinformatics and Computational Biology.