灰度
图像配准
医学
肺癌
核医学
放射治疗
锥束ct
放射治疗计划
计算机断层摄影术
放射科
人工智能
计算机科学
病理
像素
图像(数学)
作者
Jiayu Du,Jie Tang,Qian Zhang,Xiaojie Ma
标识
DOI:10.1007/s10330-021-0499-9
摘要
Abstract Objective To explore the differences in three different registration methods of cone beam computed tomography (CBCT)-guided down-regulated intense radiation therapy for lung cancer as well as the effects of tumor location, treatment mode, and tumor size on registration. Methods This retrospective analysis included 80 lung cancer patients undergoing radiotherapy in our hospital from November 2017 to October 2019 and compared automatic bone registration, automatic grayscale (t + r) registration, and automatic grayscale (t) positioning error on the X-, Y-, and Z-axes under three types of registration methods. The patients were also grouped according to tumor position, treatment mode, and tumor size to compare positioning errors. Results On the X-, Y-, and Z-axes, automatic grayscale (t + r) and automatic grayscale (t) registration showed a better trend. Analysis of the different treatment modes showed differences in the three registration methods; however, these were not statistically significant. Analysis according to tumor sizes showed significant differences between the three registration methods ( P < 0.05). Analysis according to tumor positions showed differences in the X- and Y-axes that were not significant ( P > 0.05), while the autopsy registration in the Z-axis showed the largest difference in the mediastinal and hilar lymph nodes ( P < 0.05). Conclusion The treatment mode was not the main factor affecting registration error in lung cancer. Three registration methods are available for tumors in the upper and lower lungs measuring < 3 cm; among these, automatic gray registration is recommended, while any gray registration method is recommended for tumors located in the mediastinal hilar site measuring < 3 cm and in the upper and lower lungs ≥ 3 cm.
科研通智能强力驱动
Strongly Powered by AbleSci AI