Two-Bit Embedding Histogram-Prediction-Error Based Reversible Data Hiding for Medical Images with Smooth Area

During medical treatment, personal privacy is involved and must be protected. Healthcare institutions have to keep images or health information secret unless they permission from the data owner disclose them. Reversible hiding (RDH) a technique that embeds metadata into an image can recovered without any distortion after hidden been extracted. This work aims develop fully reversible two-bit embedding RDH algorithm with large capacity for images. Medical partitioned regions of interest (ROI) noninterest (RONI). ROI informative semantic meanings essential clinical applications diagnosis cannot tolerate subtle changes. Therefore, we utilize histogram shifting prediction error embed RONI. In addition, our minimizes side effect as much possible. To verify effectiveness proposed approach, benchmarked three types in DICOM format, namely X-ray photography (X-ray), computer tomography (CT), magnetic resonance imaging (MRI). Experimental results show most embedded RONI, performance achieves high leaves less visible ROIs.