アルコール依存症と薬物依存のジャーナル
オープンアクセス

概要

Growth Rate Analysis of Stem Cells, By Using Segmentation, Features Extraction and Pattern Recognition

R Nathiya and G Sivaradje

Stem cells have the remarkable ability to cultivate itself into any kind of cell in the body at their early stage of growth. In some organs, such as the gut and bone marrow, stem cells regularly divide to repair and replace worn out or damaged tissues. The existing methodology for stem cell analysis image segmentation makes use of a morphological technique applied on the fluorescent cells so as to get a clear cut segmented image. For this the wavelet Otsu Curvelet paradigm is used in where the image or frame is filtered, Curvelet is used for better edge enhancement and Wavelet is used for multi-scale resolution. Segmentation using Otsu model, reduces the average weight of class variances from various pixels to provide an optimal threshold value. From the segmented image feature, vectors are obtained using Grey level co-occurrence matrix (GLCM) technique which plays a vital role in extracting the features in an image. However GLCM commonly extract the texture under single scale and single direction which does not provide the textural entities to its maximum extent. Hence for multi scale and multi-resolution, the segmented image is decomposed with NSCT and GLCM is applied. The set of feature vectors form finally the pattern matrix as the input to the artificial neural networks for their classification. Using neural network for pattern recognition, the network is trained by using the images of various healthy level images. Then using the trained network, the healthy nature of the test image is evaluated and the result is displayed in the form of percentage of healthiness of the given time series stem cell images. Hence this paper is highly motivated to analyse the healthy nature of stem cells.