Enhanced Lempel-Ziv-Welch Based Medical Image Compression Using Optimization Methods

Medical imaging is a technique of generating images representing the anatomy of the body as well as the functioning of some organs or tissues for clinical interpretations and medical treatments. Medical imaging stands out amongst various methods used for observing the condition of an individual's anatomy. One of the issues that doctors face in using this method is storage and retrieval of the images. Preserving these images consume considerable storage space since records of numerous patients are to be stored for longer periods. This means one has to compress these images without losing any data or details. Majority of the present pressure plans tend to give a high compression rate resulting in significant loss of value. However, in a few zones in medication, it might be adequate to maintain high image quality just in the district of intrigue, i.e., in demonstratively essential areas. In order to achieve high image quality, the proposed system designed an efficient image compression using enhanced Lempel-Ziv-Welch (LZW) with optimization methods. In this research, MRI brain DICOM images are preprocessed by adaptive median filter. The preprocessed image is segmented into the Region Of Interest (ROI) and Non Region Of Interest (non ROI) with the help of the region based active contour model using level set approach. After the ROI segmentation, the ROI are compressed with the help of the enhanced Lempel-Ziv-Welch (LZW) and Clipped Histogram Equalization (CHE) with Artificial Bee Colony Algorithm (ABC) approach. The CHE with ABC algorithm is used for contrast enhancement. The Non ROI part is compressed by using improved Embedded Zerotree Wavelet (EZW) algorithm. Finally the decompression process is performed at receiver side. The experimental analysis demonstrates that the proposed system accomplishes high a performance when distinguished with the current system with respect to accuracy, Compression Ratio (CR), Peak Signal to Noise Ratio (PSNR) and Mean Square Error (MSE).