Context-Based Interframe Coding Of MR Images

This paper presents a new lossless interframe coding scheme for Magnetic Resonance (MR) images. The existing schemes, like block matching method and uniform meshbased scheme, are inadeqaute to model the motion field of MR sequence. The above schemes use uniform mesh elements which may comprise multiple motions. We propose a scheme consisting of (a) content-based mesh generation (b) forward motion tracking (c) motion compensation using affine transformation and (d) context-based modeling. By using context-based modeling, intraframe correlation is also exploited in addition to interframe correlation. The obtained average compression ratio of 4.3:1 is better than the values of 4:1, achieved by CALIC, the state of the art lossless intraframe coding scheme and 3:1, by the existing uniform mesh-based interframe coding scheme for MR images. The performance of the existing uniform mesh-based scheme can also be improved by employing context-based modeling.