基于像素间相关性的医学MRI图像无损压缩方法

A Lossless Compression Algorithm Based on Correlation between Pixels of Medical MRI Image

目的针对医学图像所具有的特殊性,提出一种基于像素间相关性的医学磁共振成像图像无损压缩方法。方法首先根据原始医学MRI图像数据中像素间的相关性生成4个相应的数组,去除原始图像冗余,然后对每个生成的数组进行相应的算术编码,实现压缩后进行保存或传输。结果采用真实医学MRI图像对此方法进行验证表明,该方法简单易行,不需要被压缩图像的先验信息,可以得到较好的压缩比。结论与现有国内外相关的医学MRI图像无损压缩算法相比,该方法充分利用了像素间的相关性,压缩比有较大的提高。

Objective Aiming at the unique characteristics of medical images, a lossless compression algorithm based on the correlation between the pixels of medical MRI images to propose. Methods The proposed method firstly generated four arrays from the original medical MRI images based on the correlation of pixels in order to exploit the redundancy of original images. Then, we coded each generated array by the arithmetic coding, and stored or transported the coding results. Results Real medical MRI images were used to test and verify the pro- posed method and the results showed that the proposed method was simple, and achieved better compression ratio without prior information of the compressed images. Conclusion As compared with the existing domestic and foreign medical MRI image lossless compression algorithms, the proposed method takes full advantage of the correlation of pixels and increases the compression ratio significantly.