|本期目录/Table of Contents|

[1]高洪玉.基于高斯过程回归高强度聚焦超声仪器采集图像的层间测距算法[J].生物医学工程研究,2018,04:496-499.
 GAO Hongyu.Interlayer fault location algorithm based on Gauss process regression for high intensity focused ultrasound image acquisition[J].Journal of Biomedical Engineering Research,2018,04:496-499.
点击复制

基于高斯过程回归高强度聚焦超声仪器采集图像的层间测距算法(PDF)

《生物医学工程研究》[ISSN:1006-6977/CN:61-1281/TN]

期数:
2018年04期
页码:
496-499
栏目:
出版日期:
2018-12-25

文章信息/Info

Title:
Interlayer fault location algorithm based on Gauss process regression for high intensity focused ultrasound image acquisition
文章编号:
1672-6278 (2018)04-0496-04
作者:
高洪玉
齐鲁医药学院计算机教研室,山东 淄博 255000
Author(s):
GAO Hongyu
Qilu Medical University,Zibo 255000,China
关键词:
高斯过程回归高强度聚焦超声层间测距测量误差均值
Keywords:
Gauss process regression High intensity focused ultrasound Interlayer distance measurementMeasureError mean
分类号:
R318
DOI:
10.19529/j.cnki.1672-6278.2018.04.23
文献标识码:
A
摘要:
提出基于高斯过程回归的层间测距算法,通过对已知样本的训练,求解模型的参数,得到回归器,用于层间测距。通过对各实体组织进行超声扫描,得到相应的序列图像。运用到由高强度聚焦超声(high intensity focused ultrasound,HIFU)仪器采集的3组超声图像样本集中进行实验,验证方法的有效性和稳定性。测量绝对误差均值相对较小,在1~3 mm的范围内。绝对误差均值都低于0.5 mm,而且相对误差的均值保持在10%以内,标准差比较稳定,浮动不大。而在4 mm和5 mm处测量误差相对较大,绝对误差均值在1mm左右,相对误差也会出现变大的趋势。但总体的相对误差在25%内。本研究提出的基于高斯过程回归的方法在1~3 mm的范围内,具有良好的测距结果,当距离超过3 mm时,测距误差相对较大。
Abstract:
To propose an algorithm for inter-layer ranging based on Gauss process regression. By training known samples, the parameters of the model were solved, and the regression was obtained for inter-layer ranging.Through the ultrasonic scanning of the solid tissues, the corresponding sequence images were obtained. Three groups of ultrasonic image samples collected by HIFU instrument were used to verify the effectiveness and robustness of the method.The mean absolute error measured was relatively small, ranging from 1 mm to 3 mm. The mean absolute error was less than 0.5 mm, and the mean relative error was less than 10%. The standard deviation was stable and fluctuates little. The measurement error was relatively large at 4 mm and 5 mm, and the mean absolute error was about 1 mm. The relative error would become larger. But the overall relative error was within 25%.The proposed method based on Gauss process regression has good ranging results in the range of 1 mm to 3 mm. When the distance exceeds 3 mm, the ranging error is relatively large.

参考文献/References

-

备注/Memo

备注/Memo:
(收稿日期:2018-08-15) △通信作者Email:ghydoudou@126.com
更新日期/Last Update: 2019-01-30