Applying JPEG compression to a geometric transformation of ultrasonic images

A novel approach for compressing ultrasonic images is presented. In ultrasonic images data are acquired along radial (polar) sectors. The scan lines, which are arcs, are represented in the raster image in Cartesian coordinates. Therefore, even though the image is represented in Cartesian coordinates, correlation between pixels along certain arcs is higher than between pixels along horizontal lines. The authors exploit this knowledge to achieve a higher compression ratio than that possible by applying JPEG compression to the ultrasonic image. The authors' method transforms the data into polar coordinates. JPEG compression is applied to this transformed image.