European Journal of Biomedical and Pharmaceutical Sciences

PSEUDO-GRAYSCALING TECHNIQUE: A PRE-PROCESSING STEP TOWARDS PATHOLOGICAL SLIDE ANALYSIS

Sanjay Nag, Satabdi Bhattacharya, Ranita Banerjee, Dr. Indra Kanta Maitra and *Prof. Samir Kumar Bandyopadhyay

ABSTRACT

This paper presents a method for converting digitized colour images in RGB format, obtained from pathological/histopathological slide to pseudo-grayscale as a pre-processing step for image segmentation. The digitized slide images exhibits colour shades in the order of 2^24 which increases algorithmic complexity. At the same time due to the application of disparate stains for slide preparation some specific sub-domain(s) of the entire colour space is actually exhibited by these slide images. The pseudo grayscaling is actually exploiting these features to make the grayscale image more enhanced and preserving the disparate colour information, making subsequent feature extraction easy and efficient. The statistical regression on the image information is followed by selection method to transform the images from RGB colour model to pseudo-grayscale images. The proposed method is tested on a large set of pathological/ histopathological slide images some of which has been demonstrated in this citation.

Keywords: Histopathology, Digital Microscope, Grayscale, Bitmap, Multiple Regression.