体外震波碎石系统的信息优化技术研究

Study on information optimization technology of ESWL system

目前针对体外震波碎石(ESWL)系统的研究主要集中在能量发射的研究,而对信息的优化采集研究较少.为了达到既碎裂结石又不损害人体组织的目的,如何高效地采集碎石影像数据成为一个重要的研究课题.提出的策略在医疗活动信息化的基础上,根据影像像素分布情况,利用自适应选择子,自动选择不同小波所对应的提升方案对数据进行分解,在有效剔除噪声的同时保持了人体组织的关键特征.为了最小化降噪后的均方误差,采用双元软阈值进行滤波.该研究为临床工作者提供更为准确的、智能化的数据和临床决策支持能力.实验结果验证了该提升方案的低时间复杂度和接近于传统小波的极小的重构误差,在定量分析和临床观测方面都获得了更好的结果.

The present study of Extracorporeal Shock Wave Lithotripsy （ESWL） system focuses on the topic of energy emission, but there are less optimization research on the information acquisition. To achieve the goal of stone fragmentation without any damage to the human body, efficient data acquisition in ESWL images is an important research subject. For processing information of medical treatment activity, based on the statistical distribution of image pixels, adaptive wavelet lifting schemes were applied to different image regions using a adaptive selector, by which noise suppression was fulfilled while key features of the human body were retained. To achieve minimal mean square error, bivariate soft thresholding was provided here. This study provided more accurate, intelligent data, and clinical decision support ability for medical personnels. The experimental results demonstrate the proposed method is of low complexity in its implementation time and generates minimal reconstruction error approaching that of the classical wavelets. Moreover, this algorithm outperforms the existing methods in terms of quantitative performance measures as well as clinical observations.