Weighted fusion of MRI and PET images based on fractal dimension

In this paper, a fusion scheme is presented to combine the useful information present in magnetic resonance and positron emission tomography images. The proposed scheme utilizes image pre-processing, local feature (fractal dimension), weighted fusion and improved guided filter to extract and combine information present at different scales/frequencies. The fusion scheme assist radiologist in better analysis and diagnosis of different diseases. Visual and quantitative analysis reveals the significance of proposed image fusion scheme, as compared to state of the art techniques.