Abstract
Lossy Image compression has been effective in reducing medical image size, but it affects registration accuracy. In this study, the effects of lossy compression on 2D and 3D medical image registration accuracy is investigated via Target Registration Error (TRE). Mean Square Error (MSE) and Mutual Information (MS) were considered as similarity image metrics for Rigid 3D transform Registration. Results has indicated that MS performs well at lossless compression, but it did not perform as well on lossy compression.