In order to improve the noise robustness of the ultrasonic medical image compression method and image quality after decompression, we proposed an image compression method and algorithm using a closed-form shrinkage function based on Cauchy distribution and a logarithmic transform in wavelet domain, and compared the compression performance of the new method with previous methods through experiments.
In our work;
First, we eliminated effectively the speckle of ultrasonic medical image by using CauchyShrinkGMAP, a closed-form shrink¬age function based on the Cauchy distribution in the image transform stage.
Second, we solved the arithmetic precision problem and improved the image quality after the decompression using a logarithmic transform in the image transform stage.
Third, we evaluated the performance of new image compression method compared with classical image compression methods based on the discrete wavelet transform. The experiment results showed that CR and PSNR of the proposed method were the highest, MSE and BPP were the smallest, and SSIM and CC were closer to one in comparison with the other image compression methods for noisy reference image. In addition, the experiment results showed that CR and ENL of the proposed method were the highest, and BPP, EN and SD were the smallest of the other image compression methods for noisy real ultrasonic medical image.
The research results were published in "Multimedia Tools and Applications" (Vol.74 No.3, 2025, 1-24) under the title of "An Image compression method for improving noise robustness of ultrasonic medical image compression in wavelet domain" (https://doi.org/10.1007/s11042-025-20943-7).
