PET图像提供的新陈代谢信息可用于判断放疗后肿瘤的复发区域,对于制订精确的放疗计划具有重要的临床意义。研究采用多分辨率形变配准的方法提取放疗前后CT图像的形变域,并将其作用于放疗前PET图像,与放疗后的PET图像相比较,通过设定图像中SUV的阈值,判断勾画轮廓之间的重叠率,以获得图像中的高摄取区域,回顾性指导精确放疗。研究针对22例肺癌病例,实验结果显示放疗后残留的高代谢区域和放疗前GTV重叠较好:当阈值设定为SUVmax的70%、80%和90%时,对应的重叠率分别为(95.2±0.6)%、(96.6±3.4)%和100%;当阈值设定为SUV2.5和SUV5.0时,对应的重叠率为(86.0±6.6)%和(97.0±3.0)%。对氟代脱氧葡萄糖(FDG)高摄取区域的高重叠率表明病变区域在放疗前后的位置相对稳定,放疗后的残余肿瘤基本上位于放疗前靶区对FDG的摄取区域。初步实验结果证明,研究可用于判断靶区区域对放疗的反应,回顾性指导在放疗计划中,针对放疗后残余的靶区加大照射剂量,保护危及器官和组织,精确放疗。
The metabolic information provided by PET images can be used to define the high recurrence region, which is of great significance for accurate treatment planning. This study proposed a multi-resolution deformable registration method to extract the deformation field between CT images in two inter-fractions, and then applied the deformation field to PET images. By defining the SUV (standard uptake value) thresholds we determined the outlines in two PET images and then calculated their overlapping ratio to obtain high uptake regions, which are good for retrospectively performing accurate radiation therapy in the next fraction. Twenty-two patients with lung cancer were included in this study. The preliminary results showed that residual hyper-metabolic region after radiotherapy well overlapped with gross tumor volume (GTV) before radiotherapy: the overlapping ratios were 95.2±0.6%, 96.6 ±3.4% and 100% with the SUV ax threshold defined to 70%, 80% and 90%, respectively; the overlapping ratio was 86.0 ±6.6% and 97.0 ± 3.0% with the thresholds confined to SUV2.5 and SUV5.0, respectively. The high overlap rate of the high uptake area of fluorodeoxyglucose showed that the location of the lesion was relatively stable before and after radiotherapy, and the residual tumor after radiotherapy was located in the area of FDG uptake before radiotherapy. Preliminary results showed that this study had a potential ability of determining target reaction to radiotherapy and was good for retrospectively increasing radiation dose to residual tumor target with good protection to organs and tissues, showing potentials of realizing accurate radiotherapy.