文章摘要
丁琪瑛,吴水才,崔博翔,赵晶,周著黄.超声背散射信息熵成像评价脂肪肝的新方法[J].声学技术,2023,42(2):224~229
超声背散射信息熵成像评价脂肪肝的新方法
A new method for assessing hepatic steatosis using ultrasound backscatter information entropy imaging
投稿时间:2021-10-20  修订日期:2021-12-15
DOI:10.16300/j.cnki.1000-3630.2023.02.015
中文关键词: 超声背散射  信息熵  定量超声  脂肪肝  超声组织定征
英文关键词: ultrasound backscatter  information entropy  quantitative ultrasound  hepatic steatosis  ultrasound tissue characterization
基金项目:国家自然科学基金(11804013,61871005);北京市自然科学基金(4222001)。
作者单位E-mail
丁琪瑛 北京工业大学校医院超声室, 北京 100124  
吴水才 北京工业大学环境与生命学部, 智能化生理测量与临床转化北京市国际科研合作基地, 北京 100124  
崔博翔 台湾长庚大学医学院, 台湾桃园 33302  
赵晶 解放军总医院第三医学中心老年医学科, 北京 100039  
周著黄 北京工业大学环境与生命学部, 智能化生理测量与临床转化北京市国际科研合作基地, 北京 100124 zhouzh@bjut.edu.cn 
摘要点击次数: 207
全文下载次数: 214
中文摘要:
      提出超声背散射Tsallis信息熵成像评价脂肪肝的新方法。利用滑动窗口法,估算窗口内局部背散射包络信号的Tsallis信息熵参数值,对信息熵参数值矩阵进行扫描变换、颜色映射及感兴趣区域设置,叠加到超声B模式图像,实现Tsallis信息熵成像。分析72名肝脏捐献者和204名患者的超声背散射信号,参考标准分别为磁共振波谱测得的肝脏脂肪分数(Hepatic Fat Fraction,HFF)和肝活检组织学检查测得的脂肪肝程度。对于72名肝脏捐献者,Tsallis信息熵与lg (HFF)的相关系数r=0.67(P<0.000 1)。对于204名患者,受试者工作特征曲线下面积分别为0.82、0.88、0.89(≥轻度、≥中度、≥重度),而超声背散射零差K成像分别为0.76、0.82、0.82。超声背散射Tsallis信息熵成像可以直观定征并定量评价脂肪肝的严重程度,其诊断性能优于超声背散射零差K成像,可作为一种超声评价脂肪肝的新方法。
英文摘要:
      A new method using ultrasound backscatter Tsallis information entropy imaging for assessing hepatic steatosis is proposed. By the sliding window method, the Tsallis information entropy parameter values of the local backscatter envelope signal in the window are estimated to obtain the matrix of information entropy parameter values, and then, to perform scan conversion and color mapping, and set the region of interest (ROI). The entropy image within the ROI is superimposed on the ultrasound B-mode image to realize the Tsallis information entropy imaging. Seventy-two liver donors and 204 patients are enrolled. The reference standards are the hepatic fat fraction (HFF) measured by magnetic resonance spectroscopy (MRS) and the grade of fatty liver determined by liver biopsy histology. The 3-MHz ultrasound scans the liver to construct B-mode and Tsallis information entropy images for HFF and fatty liver grade analyses. The receiver operating characteristic (ROC) curve is used to evaluate the diagnostic value of the Tsallis information entropy imaging method. The correlation coefficient between Tsallis information entropy and log10 (HFF) is r = 0.67. The areas under the ROC curve are 0.82, 0.88 and 0.89, respectively for fatty grade ≥ mild, ≥ moderate, and ≥ severe. It is concluded that ultrasound backscatter Tsallis information entropy imaging can quantitatively evaluate the grade of fatty liver.
查看全文   查看/发表评论  下载PDF阅读器
关闭