International Journal of Science and Research (IJSR) ISSN (Online): 2319-7064 Index Copernicus Value (2016): 79.57 | Impact Factor (2015): 6.391 Volume 7 Issue 1, January 2018 www.ijsr.net Licensed Under Creative Commons Attribution CC BY Female Cancer Control Diagnosis by Decision Making using Image Processing - A Prospective Study R Gomalavalli, Dr. P. M. Venkata Sai, Dr. K. Sriram, Dr. S. Muttan Abstract: In this paper, an effective method is implemented for the feature extraction of renal tumor. Image segmentation is vital in many medical image diagnostic applications. It identifies the region of interest in a data. The outline of semi - programmed division contains the following stages, (i) Initial one is the Region of Renal interest(ROI); (ii) the Second one is automatize iterative method.(iii) the third is a novel work to select the best classifier for female optimal features. This is an optimal method to shun the CT guided Biopsy. An optimum method is chosen for overcoming the limitations seen in the morphological operation method and getting the degree of correspondence. Boundary segmentation of renal tumor extends to feature extraction of images and classification. The high specificity of Fuzzy is 99.5% for the left and 99.4% for the right tumor. The qualitative examination done on 37 out of 67 subjects indicates the average comparison of accuracy 95.7% and 94.7% (Left and Right region) respectively between the successive classifiers. Keywords: Renal, Tumor, Active contour method, segmentation, Boundary Detection, Feature Extraction, Classifier, Statistical Analysis, Accuracy, Sensitivity, specificity 1. Introduction The main aim of renal segmentation is to diagnose the presence of tumor. The neighboring organs like liver, lungs, and Pancreas are affected by the tumor. Axial computed tomography (ACT) scans can form an input to Computer aided diagnostic (CAD) systems for the segmentation of renal structure along with its neighboring organs. The input RGB image (Sudhakar M.S., 2016) .is converted into gray scale image and resized the image into 512 X 512 pixels. The purpose of gray scale image is to provide clear anatomical information, illumination of the abnormalities and lesions. Segmentation methods are classified as: (1) image-based, (2) model-based, and (3) hybrid methods. Usually, statistical snake models (Kobashi M, 1995) are developed on the basis of a training iteration method and implemented in the specialized region of interest. Ultrasound imaging test is a screening test with speckles noise. Noise free scanning CT images have high Signal-to-noise ratio (SNR) and provide an accurate anatomical structure. The quality of noble image, and the advanced Digital image technique, inspires the researchers to develop computerized methods for anatomical automatic renal tumor segmentation analysis. (Kobashi et al, 1995)have described the anatomic information, identification and extraction of renal from normal CT image . The detection result was rated 85% grade A from the testing of 78 images. The feature extraction of image in the basic of digital radiography (Giger and Doi et al, 1988). A proposed a deformable model approach for automatic renal segmentation . They have used adeformable model represented by the gray level appearance of renal and its statistical information of the shape(Tsagaan et al.,2002). They have used a deformable model represented by the gray level appearance of the kidney and statistical information of its shape (86.9%). Semi- automatically produced images were used considering the similarities between the gray levels in adjacent organs, contrast media effect and relatively high variation of the organ positions and shapes in abdominal CT images and uses labeling method (Campadelli et al, 2004 & Shi et al, 2004). Medical image segmentation is done through tissue surface analysis and maximum dispersal directions (85%) (Mavromatis et al, 2004 & Saitoh et al, 2002).A deformable model approach for automatic kidney segmentation, the automatic extraction technique of a kidney region which use Q-Learning method. This was proposed as a pretreatment for the automatic detection of the kidney diseases (Hetzheim et al, 1993).Feature classification is one of classical problems of concern in image processing. The goal of feature classification is to found the best categories of the input image using Active contour method. There are different types of feature extraction via Gray-level co-occurrence matrix (GLCM), Zernike moments (American cancer society, 1993) and Texture and moments order of features. There are various approaches for the selection of the best classifier, such as k nearest neighbor (KNN), Membership Function Rules applying Fuzzy (MFRF) Artificial Neural Network (ANN), and Support Vector Machine (SVM) from the best features. 2. Materials and Method Automatic image segmentation system is one of the stimulating tasks for the researcher to design the computerized diagnosis in the medical image field. Different medical imaging Modalities available are Radiography X- ray, Ultrasound (US), Axial computed Tomography (ACT), Magnetic Resonance Imaging (MRI), etc. for the guidance of the diagnostics. The computed Tomography is selected from the above list as the best Modality for the feature extraction of renal (Yoshiki et al, 2004).In renal, Normal and abnormal cells fight each other, providing the result of Necrotic patient. Hence the Hounsfield value more than 100 is tumor. The normal cell always merges with benign whereas it does not occur in tumor. Based on human eye interpretation of a large number of CT Normal/Abnormal images may lead to misclassification. Hence there is a compulsion of automated renal segmentation, with the best features for a good differentiation. The proposed flow of work is illustrated infigure.1.In this article, research examination has used different features of renal cell carcinoma and GLCM and textural features of tumor are extracted. Paper ID: ART20179632 DOI: 10.21275/ART20179632 1320