Big image analysis for identifying tumor pattern similarities

Medical images are not just images and they carry context sensitive data. Imaging in oncology plays a very important role in diagnosis, treatment and identification. There exists huge repository of medical images which when analyzed can provide insights on the disease progress, new medications and treatment paths. High performance computing combined with analytics helps to overcome the challenges posed by Big Medical Image data. Content based image retrieval (CBIR) is an important area of research where the resultant images are retrieved based on a query. In this paper, we propose dynamic image template based on the image derived features. This dynamic pattern template is then used for searching relevant images from the existing repositories based on the image query. Implementation of the proposed template was carried out using MATLAB and the results are tabulated.