A new method of video compression for angiographic images has been developed to achieve high compression ratio (~20:1) while eliminating block artifacts which leads to loss of diagnostic accuracy. This method adopts motion picture experts group's (MPEGs) motion compensated prediction to takes advantage of frame to frame correlation. However, in contrast to MPEG, the error images arising from mismatches in the motion estimation are encoded by discrete wavelet transform (DWT) rather than block discrete cosine transform (DCT). Furthermore, the authors developed a classification scheme which label each block in an image as intra, error, or background type and encode it accordingly. This hybrid coding can significantly improve the compression efficiency in certain eases. This method can be generalized for any dynamic image sequences applications sensitive to block artifacts.

We have developed a compression algorithm based on Discrete Wavelet Transform (DWT) and Arithmetic Coding (AC) that satisfies the requirements of radiological image archive. Radiological images need to be compressed at a moderate compression ratio between 10: 1 to 20: 1 while retaining good diagnostic quality. This new method is far superior to previously developed Full Frame DCT (FFDCT) method as well as industrial standard JPEG. Since radiological images usually contain artificial edges (labels, lettering, and cropping), FFDCT introduces ripple artifact due to truncation of high frequency coefficients. JPEG which is a subblock method introduces blocking artifact due to independent error across the block boundaries. Diagnostic ability is degraded by both FFDCT's ripples and JPEG's block artifacts. Since DWT is localized in both spatial and scale domains, the error due to quantization of coefficients does not propagate throughout the reconstructed picture as in FFDCT. Since it is a global transformation, it does not suffer the limitation of block transform methods like JPEG. The severity of error as measured by NMSE and Maximum Difference increases very slowly with compression ratio compared to FFDCT. Normalized Nearest Neighbor Difference (NNND), which is a measured of blockiness, stays approximately constant, while JPEG's NNND increases rapidly with compression ratio. Furthermore, DWT has an efficient FIR implementation which can be put in parallel hardware. DWT also offers total flexibility in the image format; the size ofthe image does not have to be a power oftwo as in the case of FFDCT.

No abstract

We have developed a compression algorithm which achieves high compression ratio and excellent reconstruction quality for video rate or sub-video rate angiograms. Commercially available technology such as JPEG and MPEG do not satisfy medical requirements due to their severe block artifacts. Our new method takes advantage of the high temporal correlation of the angiographic frames using MPEG motion estimation software, but performs the error frame encoding using wavelet transfonn which is much more sensitive to localized block artifact features thai the conventional discrete cosine transform.Our results show that at ratios up to 20: 1, excellent image quality is obtained in reconstructed angiograms. Because of the excellent energy gathering and spatial localization properties of wavelet transform, high image fidelity and compression ratio can be simu1taieously achieved. The algorithm is parallelizable and therefore can be implemented relatively easily in real time hardware.

The Full Frame Discrete Cosine Transform (FFDCT) compression method has the capability to eliminate block artifacts that exist in JPEG and MPEG compressed images. Previous FFDCT algorithm used an adaptive Bit Allocation (BA) method to pack quantized DCT coefficients. In this work, we replaced BA with Arithmetic Coding (AC) to improve compression performance markedly. We will also show that by partitioning the quantized cosine domain much higher arithmetic coding efficiency can be achieved, especially in highly quantized images. . INTRODUCTIONHigh ratio, transform based compression is critically needed to make archiving and telecommunication of digital medical images feasible. The most widely supported compression methods are JPEG and MPEG standards, both based on the block DCT approach that introduces block artifacts in the reconstructed images. Because diagnostic reading is particularly sensitive to artifacts, such methods may not be acceptable in quality. Recent study has shown that even when block artifacts are not immediately obvious, they become prominent after edge enhancement[ 1]. Many studies have introduced modifications to the basic block DCT method in an effort to minimize discontinuities across blocks [2,3,4,5]. However, only two methods completely remove them. They are the wavelet transform coding which has recently gained a great deal of attention, and the more established Full Frame DCT method.Previously, we introduced the FFDCT algorithm along with a custom hardware module to accelerate DCT computation[6,7,81. The HVS based quantization table used by JPEG was rejected in favor of a threshold based quantization scheme because we believed that high frequency components often contain crucial diagnostic detail and should not be deemphasized as is done in JPEG. The quantized coefficients were encoded using an adaptive bit allocation method, which is particularly suited for parallel processing. However, this method necessarily includes spurious zero bits which offer no information, and generates a bit allocation table that represents a sizable overhead in highly compressed images. In this paper, we replaced the bit allocation step with arithmetic coding that better exploits the entropy inherent in the quantized coefficients. It will be shown that not only is the overall compression efficiency increased, the parallel processing property is still preserved. FFDCT AND HARDWARE MODULEDiscrete Cosine Transform (DCT) is excellent for image compression because of its energy compacting properties. In order to ensure complete elimination of block artifacts, we shunned the common practice of partitioning an image into small blocks before doing cosine transform. Performing DCT in full frame has the added burden of high computational complexity, which we addressed using an efficient DCT implementation based on a modified FFT[9]. We also took a different approach in deriving the quantization levels since sharp edges cany significant diagnostic information (interstitials, calcifications) and must be preserved with eq...