Detection of Brain Tumor from MR Images Using BWT and SOM-SVM with Authentication

This paper does the qualitative analysis of the watermarking technique to provide an improved healthcare system. It addresses two main problems of brain tumor detection: one is to find the region of interest, i.e., to detect brain tumor from the medical images and second is to authenticate patient information on the medical images. The system combines a region-based segmentation using self-organizing map algorithm and secure authentication using a watermarking technique based on the Berkeley wavelet transformation (BWT) algorithm. In recent times, a significant amount of interest has been developed in encouraging radiologists or clinical experts to use the technology-based system for health care. The proposed approach combines BWT for watermarking and self-organizing map and support vector machine algorithms for region-based segmentation to improve the objectives of better treatment. The experimental result of the proposed method is evaluated for performance and quality analysis on magnetic resonance imaging (MRI) images, resulted well in terms of quality parameters and imperceptibility.