dFuse: An Optimized Compression Algorithm for DICOM-format Image Archive

Medical images are useful for knowing the details of the human body for health science or remedial reasons. DICOM is structured as a multi-part document in order to facilitate extension of these images. Additionally, DICOM defined information objects are not only for images but also for patients, studies, reports, and other data groupings. More information details in DICOM, resulted in large size, and transferring or communicating these files took lots of time. To solve this, files can be compressed and transferred. Efficient compression solutions are available and they are becoming more critical with the recent intensive growth of data and medical imaging. In order to receive the original and less sized image, we need effective compression algorithm. There are different algorithms for compression such as DCT, Haar, Daubechies which has its roots in cosine and wavelet transforms. In this paper, we propose a new compression algorithm called “dFuse”. It uses cosine based three dimensional transform to compress the DICOM files. We use the following parameters to check the efficiency of the proposed algorithm, they are i) file size, ii) PSNR, iii) compression percentage and iv) compression ratio. From the experimental results obtained, the proposed algorithm works well for compressing medical images