Application Of Medical Image Enhancement In The Soft Copy Reading

With the development of computer network technology and digital technology, PACS is brought into the hospitals, and promotes the tendency that the diagnosis on the monitor replaces that on the film. With the extensive use of PACS system, there are two problems in the clinical application. The one is the effect of noise in the medical image acquisition. The other is image degradation, because most hospitals make use of the general monitors instead of high-end medical monitors to cost reduction. However, the display on the monitor is associated with the diagnosis. Through the analysis of the acquisition and display of the image, we restrain the noise and enhance the edge to obtain the display on the high-end medical monitors.To solve the noise in the medical image acquisition, on the basis of comparison study of filters, a new algorithm is proposed for Poisson noise removal. Haar wavelet is applied to estimate the local variance to the case of the Poisson distribution. An adjust parameter is used to overcome the drawbacks that cannot change a threshold value depending on the amplitude of the fluctuations locally. A stationary wavelet transform (SWT) is utilized to alleviate the ring artifacts caused by thresholding with the orthogonal basic functions. Experiment results indicate that the proposed method has better results than the existing methods.To improve the display of the medical image on a general monitor, we develop a medical image clear vision technique for soft copy reading in this paper. We first transform different type medical images into10bits data. Then, an adaptive enhancement method is proposed. Our method combines compressing the global dynamic range of the transformed image and an improved unsharp masking. The main contribution of this method is to design different feature extracting filters for different type medical images and preset some parameters for monitor quality assurance and quality control. Finally, we obtained a processed image using image fusion. Experimental results show that the proposed technique can improve effectively the visual quality of image displayed in monitors.