In this paper, Principal Component Analysis (PCA) with Set Partitioning in Hierarchical Trees (SPIHT) propounds to accomplish an image compression. A lossy technique is introduced through the PCA which is followed by SPIHT to enhance the compression performance. The Peak Signal to Noise Ratio (PSNR) value of the reconstructed image acquired from the PCA methods is not found to be sufficient which can be further improved by another method called SPIHT. In this paper, a hybrid compression model is constructed to accomplish the benefits of both PCA and SPIHT. In the bio-medical research, compression becomes necessary due to transfer of patients images from one group of experts to other experts group. So that, the proposed image compression technique is useful for conserving the storage space needed for healthcare systems. Finally, it is concluded that the proposed PCA-SPIHT performs better than other recent state-of-the-art techniques with acceptable loss of image quality.