FPGA Implementation of segmented feature fusion in MRI images using wavelet

Abstract

Now-a-days, digital image processing in medical diagnosis is essential in almost all medical fields. Human eye is usually used to visually interpret a processed image as how well a particular technique used for its processing has worked. In medical diagnosis, two important techniques viz computed tomography (CT) offers less distortion with better information on denser tissues while magnetic resonance image (MRI) offers high distortion with best information on soft tissues. The multimodality of medical images in clinical applications has framed the idea of combining clinical images of distinct modality in medical applications. This knowledge of fusing clinical images has raised a new promising research field. The experiments on image fusion technique shows that it could extract useful information from distinct source images in fused images in order to obtain clear images. In this paper, an image fusion system with standalone hardware implementation framework is proposed. Xilinx platform studio (XPS), Visual Basic (VB) software's and FPGA Spartan 3 image processing kit is used for implementation. Firstly, input images are converted to .h files using MATLAB GUI. Then XPS is used to pick software and hardware components for fusing the images by incorporating source and .h files and converting them to bit streams for downloading them to FPGA Spartan3 kit. The fused image as obtained from FPGA can be seen using Visual Basic Graphic User Interface (GUI) on the monitor.