ART workflow based on CBCT and end-to-end clinical dynamic phantom validation.

BACKGROUND

In image-guided radiotherapy (IGRT), cone beam CT (CBCT) is usually used as a positional calibration for patients before treatment, but artifacts and inaccurate Hounsfield unit (HU) values prevent its use for clinical treatment dose calculation. In order to enable CBCT to be used for online (adaptive radiotherapy) ART treatment, it is crucial to obtain high-quality CBCT images. PURPOSE: To validate the feasibility and accuracy of the conventional accelerator’s online Adaptive Radiation Therapy (ART) workflow, providing valuable references for clinical treatment.

METHODS

Firstly, utilize the RegGAN model to generate synthetic CT (sCT). The quality of the sCT was evaluated using the mean absolute error (MAE), peak signal-to-noise ratio (PSNR), and structural similarity index measure (SSIM). sCT images were used for plan production and dose validation. The simulated clinical treatment process included model processing, rigid and deformable registration, delineation replication and plan replication. Dose verification was performed using CIRS Dynamic Thorax phantom and radiographic films to validate the sCT image accuracy.

RESULTS

The MAE in the images decreased from 52.85 ± 16.84 to 25.91 ± 3.62 (P < 0.05), the PSNR increased from 29.95 ± 2.88 to 31.61 ± 1.29 (P < 0.05), and the SSIM increased from 0.90 ± 0.02 to 0.96 ± 0.01 (P < 0.05). The Hounsfield unit (HU) distribution of the sCT images highly overlapped with that of the planning CT (pCT) images, with minimal dose calculation differences. Dynamic Motion sCT planning quality assurance achieved a pass rate of 100% at the 3%/2 mm gamma criterion. sCT Film verification using a 4 mm/3% gamma criterion yielded pass rates of 88.35% and 100% for the sCT planning system and pCT planning films, respectively.

CONCLUSIONS

The proposed ART workflow demonstrates high dosimetric accuracy in the evaluated phantom and patient cohort. The workflow serves as a proof-of-concept for conducting online adaptive radiotherapy on conventional accelerators.