International Journal of Computational Intelligence Research (IJCIR)

Volume 3, Number 1 (2007)


New region of interest image coding using general layered bitplane shift for 

medical image compression

Li-bao Zhang, Xian-chuan Yu
College of Information Science and Technology, Beijing Normal University, Xinjiekouwai Street 19, Beijing, China


Region of Interest (ROI) coding technique is significant for medical image compression and transmission. It enables the regions important to medical diagnosis to be encoded and transmitted at higher quality than the other regions. The Maxshift method recommended by JPEG2000 can support ROI coding for medical image efficiently. However, its major disadvantage is that it cannot adjust the scaling value which determines the degree of relative importance between the ROI and the background (BG) wavelet coefficients. In this paper, we propose a general layered bitplane shift (GLBShift) method for medical image compression, which delivers much more flexibility than Maxshift method for degree-of-interest adjustment of the ROI. The new method has lower computational complexity than the general scaling based method proposed by JPEG2000. Experimental results show that the proposed method can provide significantly better visual quality than Maxshift at low bit rates for the CT and MRI images.

medical image coding, JPEG2000, region of interest, bitplane shift.