Sequential — Storage of differences approach in medical image data compression for brain image dataset

Diagnostic imaging especially brain imaging is nowadays widely developed involving more and more modalities. The size of image dataset then becomes very big. It is a problem not only because of space storage issue but also for reliable data communication as in cloud architecture. Our goal is to develop a novel method called sequential — storage of differences to compress brain image dataset by utilizing redundancy in 4D format (3D+t). We tested our method with real clinical patient data produced by CT Perfusion and measured the ratio of compression as well as power of rate compression to evaluate the performance. With 10 dataset, we have average compression rate of 0.53 and space saving of more than 47%. This method is lossless and compression ratio was acceptable that make it is suitable to be applied in cloud architecture.