Two-Dimensional S Transform Analysis Technique And Its Applications In Medical Image Processing

Coronary Disease has been one of the main reasons for human being death. The birth of Real Time Myocardial Contrast Echocardiography (RT-MCE) has created a new situation for the diagnosis of Coronary Disease. The technique is hurtless for quantitative estimate on cardiac microcirculation. The scholars at home and abroad have focused on how to improve the objectivity and the accuracy of the image quantitative analysis. There is one key problem for MCE Coronary Disease quantitative analysis:the denoising of MCE image. Until now, all the reported quantitative analysis software systems for MCE haven’t solved it. The literatures about denoising of medical ultrasound image are relatively infrequent. The researches about MCE image denoising have great theory meaning and academic value, but the clinical practice effect doesn’t satisfy very well. So, it need study new method to solve this problem. R. G. Stockwell proposed the S transform theory in1996. He has proved that we can derive S transform as the "phase correction" of the continuous wavelet transform. That is, in the frame of multi-resolution analysis technique, S transform is an extension of wavelet transform. As a new local multi-scale spectrum analysis technology, S transform has distinct advantages than wavelet transform and short time Fourier transform (STFT). As a result, S transform analysis technique must have great potential on medical image denoising. Wavelet transform has excellent performance in image denoising domain. As a kind of time-frequency analysis method with better performance in application than wavelet transform, S transform will have enormous application prospect. In this case, we introduced the denoising ideas of wavelet transform into S transform and then proposed the thresholding image denoising method using two-dimensional S transform. Simulation experiments results illustrated the favorable application performance of the method, and the successful implementation in MCE images denosing demonstrated its clinical values. The demand of memory and computational complexity are large, as the inherent property of S transform, it seriously hampered the application of S transform in the field of image processing. This made it only deal with smaller image or region of interest (ROI). R. G. Stockwell proposed the discrete orthonormal Stockwell transform (DOST) in2006. This algorithm not only keeps the original advantage of S transform, but also reduce the computational complexity and memory demand, which makes the DOST more practical. Recently, some scholars proposed the fast implementation model (Fast DOST, FDOST), which can reduce the computation complex preferably. This paper realized the DOST and FDOST, and obtained good effects when applied them in MCE image denoising.