Ecg And Mri Data Lossless Compression Algorithm Of The Preliminary Study

In recent years, along with the increase in the number of modern medical instruments and the robust development of informationization of hospital networks, the volume of digital medical data increased sharply and became the foundation of clinical doctor's diagnosing gradually. Vast medical data made themselves inconvenient for storing and communication transmitting. The key to this problem is to realize the efficient compressing. At the same time, due to the speciality of medical data and the requirements of both clinical doctor's diagnosing and further medical data processing, the lossless compressing is indispensable. In this thesis the medical ECG data and medical MRI image data were compressed with corresponding lossless compression arithmetic. Here, two aspects of contents were mainly studied:1. Now the researches on ECG data compression mainly focus on compressing ECG data of low sampling rate. A BW-based high sampling rate ECG data lossless compression algorithm is proposed in this paper. We apply the difference operation to the original ECG data first and represent part of the 16-bit binary differential value as 8-bit binary. Then the differential results are coded with the move-to-front coding method in order to make the same characters centralizing in a certain area. Last we gain a high compression ratio by using the arithmetic coding method further. Our experimental results indicated that this is an efficient lossless compression method suitable for not only body surface ECG but also heart ECG data. The average compression ratio comes up to 3.547 and 3.608 respectively. Comparing with current ECG compression algorithms, our algorithm has gained much improvement on the compression ratio, especially when applying to the high sampling rate ECG data.2. According to the particularity of medical image. The medical MRI images provided by the MRI image data-base of Harvard University was analyzed deeply and then compressed with a lossless compression algorithm based on the correlation of pixels. In this algorithm, firstly four arrays were generated from the original medical MRI images according to the correlation of pixels, secondly each generated array was encoded by arithmetic coding, and thirdly the result codes were saved or transmitted. The experimental results show that this correlative algorithm has enhanced the compression ratio under the same condition, comparing with current medical MRI image compression algorithms.