Retinal detachment, the separation of retina from retinal pigmented epithelium, causes changes to many types of cells in retinal tissue. A successful modeling of these changes can help to understand the undergoing biological processes and to devise more effective therapies to heal injured retinal tissue and eventually to recover the vision. In this paper, we first apply image processing techniques to extract texture based feature vectors using Gabor filters from confocal microscope images taken from retinal detachment experiments. We then correlate the image features with biological meta data using Chi square test. We study the distribution of images in a reduced feature space and classify images into biological classes labeled by experimental conditions. Results show that our texture feature vector clearly capture the biological experimental conditions under which the images are acquired, therefore they can be used in recognizing and analyzing biological patterns in images, as wells as content based retrieval in biological image database.