Accuracy and Precision of the 5DCT Protocol
Abstract
Purpose
To quantify the precision and accuracy of a five-dimensional computed tomography (5DCT) free-breathing simulation approach in a large clinical cohort.
Methods
A 5DCT motion model was generated from a subset of 25 free-breathing CT images per patient, parameterized by respiratory amplitude and breathing rate. This model was used to characterize respiratory motion and generate internal target volumes (ITVs). The resulting image datasets were also used for motion-modeling performance quality assurance. Motion-model precision was quantified by calculating the root-mean-square (RMS) of the residual errors between the modeled and observed tissue positions across all available datasets. Motion model accuracy was evaluated in a subset of 107 patients with clinician-contoured ITVs got which the original 25 CT images were deformed to a reference CT using the motion model, and tumor positions were compared for the 25 scans. A rigid registration was performed within a small region of interest centered on the ITV centroid, with visual verification of registration quality and the offsets used to quantify motion model error.motion model error.
Results
Across 238 patients, the mean median and 90th-percentile residual errors of the motion model were 1.21±0.57 mm and 1.90±0.83 mm, respectively. In the subset of 107 patients evaluated for ITV motion accuracy, the mean RMS and maximum errors were 1.05±0.69 mm and 0.90±1.30 mm, respectively.
Conclusion
In a large clinical cohort, the 5DCT free-breathing motion model demonstrated high-precision and accuracy for most patients and tumor locations within the lungs. These results provide quantitative validation of 5DCT performance and increase confidence in its reliability for respiratory motion characterization and ITV generation. While limitations were observed near the chest wall and in patients with highly irregular breathing, the overall robustness of the approach supports wider clinical implementation of 5DCT as a practical and dependable alternative to conventional 4DCT in thoracic radiation therapy planning.