Compressed sensing based multi-modal medical image fusion using a combined fusion strategy

In this paper, a novel multi-modality medical image fusion method based on compressed sensing by fusing undersampled k-space data is proposed. In order to transfer structural information from the source images into the fused image clearly, a combined fusion strategy is designed for undersampled low and high frequency subbands of k-space data. The final fused image is reconstructed from fused subband data with the conjugate gradient method. The experimental results demonstrate that the proposed algorithm can substantially reduce sampling data and obtain satisfactory results to meet the demand of clinical diagnosis.