A near Lossless compression method for medical images

In this paper, we introduce a selective compression method to compress lung images. Generally Region of Interest (ROI) should be compressed in a lossless manner and Region of Background (ROB) should be compressed in a lossy manner with a lower quality. In existing system, Region of Interest (ROI) is selected manually. The proposed method is automated ROI based near Lossless compression, Tumor can be benign or malignant. This disease is suspected when seen on MRI or CT scan. First segmentation process was applied in lung image using region growing. The second process is fuzzy logic which was used for classification after that benign (Non ROI) tissue only compressed using Set Partioning In Hierarchal Tree (SPIHT) algorithm, finally compressed image was superimposed with ROI (Malignant). This method is improving the compression ratio and increases the peak signal to noise ratio (PSNR) value.