Development of an Algorithm to Artificially Create Virtual Brain Deformations for Brain DICOM
Development of an Algorithm to Artificially Create Virtual Brain Deformations for Brain DICOM
复制标题
开发一种算法来为大脑 DICOM 人工创建虚拟大脑变形
DOI：
10.1007/978-3-031-05409-9_30
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Noborio Hiroshi
中科院分区：
文献类型：
--
作者：
Mori Takumi;Nonaka Masahiro;Kunii Takahiro;Koeda Masanao;Noborio Hiroshi
本研究提出了一种算法，在患者的脑DICOM图像中人工产生脑偏移。然后，我们评估了SIFT， AKAZE， ORB和BRISK特征点提取算法，用于从患者术前和术后DICOM图像中自动检测局部脑偏移。准确、自动地检测脑位移有助于生成准确的脑变形模型。评价结果表明，BRISK和AKAZE算法最适合上述目的。
This study proposes an algorithm to artificially create brain shifts in patients’ brain DICOM images. Then, we evaluate the SIFT, AKAZE, ORB, and BRISK feature point extraction algorithms for the automatic detection of local brain shifts from a patient’s pre-and postoperative DICOM images. Accurate and automatic detection of brain shifts could contribute toward the generation of accurate brain deformation models. The evaluation results suggested that the BRISK and AKAZE algorithms are most suitable for the above purpose.