Effect of JPEG compression on 24-bit color images: impact of compression on quantitative measures of velocity and flow rate in a flow convergence model

It has been previously demonstrated, using objective and subjective analyses, that grayscale echocardiographic images with 95% JPEG compression show no significant loss in diagnostic or image quality. Here, the authors investigate how the quantitative measurements of velocity and flow rate are affected when 24-bit color Doppler images are compressed using JPEG. Adobe Photoshop was used to compress a set of images to 5 levels ranging from 7:1 to 65:1 using JPEG algorithm. The images were analyzed using Khoros (public domain image processing software) and customized C programs. The results show that even at a compression ratio of 65:1, the correlation of velocities in the region of interest is as high as 0.943.<>