HLCC法定量评价CBCT图像伪影

发布时间：2018-03-19 01:24

本文选题：锥形束CT　切入点：图像伪影定量评价　出处：《核电子学与探测技术》2017年04期 　论文类型：期刊论文

【摘要】：为解决CBCT图像评价标准不一问题,本文提出了一种基于HLCC的CBCT图像伪影定量评价方法。该方法通过CUDA模拟CBCT投影,然后用重排算法转换射束获得HLCC曲线,建立三个评价伪影严重程度的定量指标。实验结果表明:定量指标数值越小,表明图像伪影程度越轻,且带伪影的HLCC曲线的三个评价指标数值是不带伪影数值的4~846倍。图像伪影的主次关系可总结为:散射伪影射束硬化伪影几何伪影余晖伪影无伪影。该方法可在投影阶段客观判断图像伪影的严重程度,简单快捷。
[Abstract]:In order to solve the CBCT image evaluation standard is not a problem, this paper proposes a CBCT image artifact quantitative evaluation method based on HLCC. The simulation method of CBCT projection by CUDA, and then use the rearrangement algorithm to convert the HLCC curve of beam, a quantitative evaluation of the three artifact severity index. The experimental results show that the quantitative numerical index the smaller that image artifacts more light, three evaluation value of HLCC curve and the artifact is 4~846 times without numerical artifacts. The relationship between primary and secondary image artifacts can be summed up as follows: scattering artifacts of beam hardening artifacts geometric artifacts afterglow artifact artifact free. This method can objective to determine the severity of image artifacts in the projection stage is simple and quick.

【作者单位】：南方医科大学生物医学工程医疗仪器研究所;
【基金】：国家科技支撑计划(2015BAI01B10) 广东省自然科学基金(2016A030310388) 南方医科大学留学归国启动资助(LX2016N003)资助
【分类号】：TP391.41

【相似文献】