Novel Image Processing Techniques to Detect Lesions using Lab View R. Manjula Sri 1 , Dr.K.M.M.Rao 2 1 Electronics And Instrumentation Engg. Department, VNR VJIET, Hyderabad,AP,India 2 National Remote Sensing Centre, Hyderabad,AP,India 1 rmanjulasri@gmail.com 2 kundammrao@gmail.com Abstract— Automated analysis of retinal images can assist in the diagnosis and management of blinding retinal diseases, such as diabetic retinopathy, age-related macular degeneration (AMD), and glaucoma. For evaluating and imaging patients with retinal diseases, clinical photographers usually capture color images of the retina using a specialized fundus camera. Subsequently, a fluorescein dye is injected into a vein in the subject’s arm, and as the dye propagates through the retinal blood vessels, and series of (over a 5–10 min period) pictures of the retina are taken. Retinal fluorescein images at two different stages of Angiogram and a color fundus image of the same patient are used as input images for lesion diagnosis. From the two fluorescein images, vessel extraction is done, they are aligned and fused to identify the region of abnormality and lesion growth. From the color image, the 3D plot is simulated to identify the area of abnormality of the eye and is used as reference to the fused fluorescein image. In this paper we present a novel technique for diagnosis of lesions through Fluorescein Angiographic Images using Virtual Instrumentation(VI). Keywords-Fluorescein images, lesion growth, image pre-processing, feature extraction, fusion, 3D simulation. I.INTRODUCTION The ability to process and analyze image information to improve medical diagnosis and monitoring through imaging is the new trend.[1].One such application is the use of image processing techniques for earlier detection and diagnosis of eye disorders. Computer-aided analysis of retinal images, with the application of image processing techniques for ophthalmology has the potential to facilitate quantitative and objective analysis of retinal lesions and abnormalities[2],[3]. 2 The image processing techniques include morphological filters for preprocessing fundus images, filters for edge detection, the Hough transform for the detection of lines and circles, Gabor filters to detect the blood vessels etc.[4],[5]. Accurate identification and localization of retinal features and lesions contribute to improve objective diagnosis and treatment, which can help patients to go for effective treatment of these eye disorders in the early stages of the disorders[6]. II.METHODOLOGY For the diagnosis of lesions, retinal fluorescein images at two different stages of Angiogram are considered as input images[7]. From the two fluorescein images vessel extraction is carried out, common features are identified, images are aligned and aligned images are fused. The retinal fundus images present structural and impulsive noise. The first one is due to anatomic shape of the retina, and the second one is caused by the acquisitions tools. For the extraction of blood vessels, retinal image should have the same grey levels both for background and the vascular system, In order to extract the desired retinal features, we need to pre-process[8] the image properly. As a pre-processing step, the green channel of a color retinal fundus image is extracted and inverted. From the gray scale images binary images are created by thresh holding[9],[10]. One method of thresholding that is relatively simple, and is robust against image noise, is the iterative method presented as below [11]. The image is segmented into object and background pixels with an initial threshold T, creating two sets. The average of each set m1 and m2 are computed and the new threshold is computed as T’ = (m1 + m2)/2.The process is repeated until the new threshold matches the previous threshold. Fundus images and their corresponding images after thresholding are shown in Fig 2 below. Fig2: Fundus images and their corresponding Images after thresholding It is important to eliminate the unwanted details to