Survey of Medical Image Compression Techniques and Comparative Analysis

Introduction

Medical images are widely used in disease diagnosis. Currently, the following modalities such as 1) Computerized Tomography (CT); 2) Magnetic Resonance Imaging (MRI); 3) Positron Emission Tomography (PET) and 4) X radiographs, etc are used to produce images in digital form. Those modalities provide flexible means for viewing anatomical cross sections and physiological states. The improvement of diagnosis and the development of interventional procedures have been facilitated by digital enhanced displays and immediate reviewing. However, medical images have large storage requirements. Medical diagnostic digital data produced by hospitals have increased exponentially. In an average-sized hospital, many terabytes of digital data are generated each year, almost all of which have to be kept and archived, which require a large amount of memory storage. Transmitting such an image over a network could take minutes to transfer; because it may be slow for interactive teleradiology, telemedicine and telebrowsing applications. Therefore there is a need to reduce the storage and the transmission requirements with the help of image compression techniques. So, Image compression plays a critical role in telematics applications and especially in telemedicine and teleradiology. The two basic types of compression techniques are

• 1.

  Lossless Compression

• 2.

  Lossy Compression

The lossless compression technique allows for perfect reconstruction of the original image and yields a modest compression rate of at most two. The lossy compression technique yields higher compression rates, but the resulting image is only approximately similar to the original image where the quality of the image is measured in terms of Peak Signal to Noise Ratio (PSNR). Several lossless and lossy compression techniques with different compression rates have been developed.

Measuring Compression Effectiveness

The effectiveness of compression schemes can be described using a relative measure, Compression Ratio (CR) or by describing an absolute measure, the Bit Rate of an image, PSNR, and complexity. The bit rate is the average number of bits required to encode a pixel and is computed from the total number of bits encoded divided by the number of pixels. Such a value is useful when comparing different schemes applied to one image, or multiple images with the same bit depth. Accordingly, a relative measure, the compression ratios are computed from comparing the number of bits in the uncompressed image to the number of bits in the encoded image.

Compression ratio = (1) Bit rate = (bits per pixel) (2)

The other parameter used for comparison is peak signal to noise ratio and it is defined as follows:

(3)

Where

Mean Square Error (4)

The number of arithmetic operations required to perform both encoding and decoding process measures the complexity of an image compression algorithm. This is an important parameter for applications involving online image compression and decompression where speed is crucial.

Traditional and state of the art Lossless/Lossy to lossless coding are given in next section. Main focus of this chapter gives comparison of various coding techniques, problems and issues in standard image compression techniques and also solutions to the problem. Future direction and conclusions are given in the last section.