Author: Jiayun Chen, Shengqi Chen, Junchao Li, Fei Liu, Jianan Wu 👨🔬
Affiliation: Cancer Hospital Chinese Academy of Medical Sciences, Shenzhen Center, Tongji Medical College, Huazhong University of Science & Technology, School of Electronic Engineering, Beijing University of Posts and Telecommunications, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College 🌍
Purpose: To assess if 3D Vane MRI can accurately depict the motion of the target volume and OARs.
Methods: This retrospective study included 54 liver cancer patients who underwent both 3D Vane MRI and 4D-CT simulation in free-breathing mode. 3D Vane MRI images were registered to each 4D-CT respiratory phase using an in-house algorithm, with normalized MCC quantifying alignment accuracy. Results were analyzed using the Mann-Whitney U test. Validation was conducted by two blinded radiation oncologists: one repeated the registration using the Eclipse module, while the other assessed clinical relevance on a five-point scale.
Results: The fusion of 3D Vane MRI with the 50% respiratory phase of 4D-CT achieved the highest normalized MCC of 0.1012±0.0011, indicating optimal alignment accuracy. No significant differences were observed in normalized MCC between phase 50% and phases 30% (0.1009±0.0008), 40% (0.1000±0.0010), and 60% (0.1012±0.0010). Significant differences were found with phases 0% (0.0979±0.0020), 20% (0.1002±0.0006), 70% (0.1008±0.0008), and 90% (0.0981±0.0017) (p < 0.05). The end-inspiratory and end-expiratory phases showed slightly lower normalized MCC, suggesting limitations in capturing certain respiratory motions. Independent validation by two radiation oncologists corroborated these findings, with similar trends observed in the five-point validation. The 3D Vane MRI achieved the highest score of 4.9074±0.2926 at the 50% phase, and the registration relevance scores for 3D Vane MRI to CT phases 30%–60% were all above 4, while those for the remaining phases were below 4.
Conclusion: 3D Vane MRI can effectively represent 4D-CT phases 30%–60% but may not fully capture motion throughout the entire respiratory cycle, particularly at end-inspiratory and end-expiratory phases. Caution is advised when using 3D Vane MRI alone for target and OAR definition in radiotherapy.