The 2-Dimensional Hartley transform has been adopted in three adaptive techniques to compress different types of medical images (i.e. X-ray, MRI, CT-scan and Ultrasound). Other orthogonal transformations (i.e. Cosine, Fourier, Hadamard, and Walsh), have also be considered and used to compare the compressed results with the Hartley transform. The comparison involved of the measure of the Spatial and Statistical characteristics of the transformed coefficients. Generally the three suggested compression techniques are started by dividing the image into set of non-overlapping blocks, 2-D Hartley transform is then applied on each one of them, the discarding and quantization of the transformed coefficients stages are performed according to the compaction amount of the energy within the block. The amount of retained energy coefficients has been controlled by utilizing a distortion parameter (DS). To optimize the compression performance, a classification scheme for the image blocks is adopted, a bit allocation matrices then designed to assign a given number of bits for each retained transformed coefficients. Since the Hartley transform is a sinusoidal transform, i.e. has same behavior as the Fourier transform, a new scanning scheme has been developed, by utilizing the shifting behavior, and used to select the transformed coefficients to be coded. This new scanning approach has been shared in increasing the compaction amount of each transformed block, and contributed in improving the efficiency of the Hartley transform compression technique to approach the standard discrete cosine transform DCT technique. One of the most important behavior of the Hartley technique is that; It can be performed by using a faster algorithm than the FFT.