Automated Seed Points and Texture Based Back Propagation Neural Networks for Segmentation of Medical Images

In this paper, a combination of Pixel based Seed points and textural Back Propagation Neural Networks is proposed for segmenting the Region of Interest (ROI) from the medical images. Medical images such as Fundus and Skin images are used to test the proposed algorithm. To develop the proposed algorithm, Pixel based Seed points are combined with the Texture based Back Propagation Neural Network (TBP-NN) by a trained knowledge of textural properties for segmenting the medical images which can be used in early Diabetic Retinopathy (DR) detection and Skin lesion detection. The proposed algorithm is tested with a total of 200 fundus and skin images each which is stored in a database used for further testing. The medical images were processed such that a knowledge in form of texture features such as Energy. Homogeneity, Contrast and Correlation were automatically obtained. The proposed algorithms efficiency was compared with traditional BP-NN methods and Support Vector Machine for segmenting medical images. The results obtained from the proposed methodology reveals that the accuracy of proposed algorithm is higher. It indicates that the proposed algorithm could achieve a better result in medical image segmentation more effectively.