Coronary angiogram video compression adapted to medical imaging applications

H.264/AVC coder has proven to use the most advanced video compression, but, at the cost of high computational complexity. On the other hand, analysis of coronary x-ray images reveals large areas containing no diagnostically important information. In this paper, we propose to exploit the energy characteristics in slice equal size regions to determine the active zones in coronary x-ray sequences to be encoded as normal using the H.264 coding system. The other regions, are compressed using conventional low complex approaches. Experimental results have shown that this procedure reduces the coder computing time of about 20% while attaining the same compression performance. A clinical subjective assessment by three expert physicians in interventional cardiology leads to a compression ratio of about 30:1 which insures both a diagnosis adequacy and a sufficient compression in regards to storage and transmission requirements.