Data recovery for medical organs based on a joint Statistical Deformable Model that incorporates prior knowledge of the missing data

In this paper, we proposed a novel joint Statistical Deformable Model (SDM) that incorporates prior knowledge of the missing data into the data recovery process for medical organs. Our approach formalized the prior knowledge of the missing data as the probability of the sample belonging to the SDM from the training data set, and together with the known points to provide accurate data recovery. Our experiments have shown that, the proposed approach achieves a higher accuracy than a state of the art technique for highly variable (soft tissue) organs particularly when the number of known points is small compared with that of the missing data. This results in a unique and robust framework of data recovery that is suitable for both rigid and highly flexible medical organs.