Clinical feasibility of fast adaptive four-dimensional cone-beam computed tomography for lung cancer radiotherapy

Adaptive Four-Dimensional Cone-Beam Computed Tomography (4DCBCT) can reduce scan time and imaging dose in radiotherapy. This is achieved by modulating the projection acquisition rate and gantry rotation speed in response to real-time changes in patient breathing, together with motion-compensated image reconstruction. This study aimed to evaluate the clinical image quality of a fast adaptive 4DCBCT acquisition compared with conventional 4DCBCT in the treatment of lung cancer.

Image datasets from the Adaptive 4DCBCT (ADAPT) clinical trial (ACTRN12618001440213), which included 30 patients treated for lung cancer, were analyzed. Two scan types were assessed: fast adaptive 4DCBCT (200 projections acquired over 20 breathing cycles, approximately 60-80s) and conventional 4DCBCT (1320 projections acquired over approximately 80 breathing cycles, 4 min). Two radiation oncologists and four radiation therapists, blinded to the image acquisition technique, independently rated the clinical utility of each scan using a two-question survey. Tumor visibility was rated on a three-point scale, and overall image quality was rated on a ten-point scale. A paired t-test was used to compare scores across acquisition techniques.

Fast adaptive 4DCBCT showed a mean tumor visibility score of 2.3 ± 0.8, compared to conventional 4DCBCT (2.4 ± 0.7). For general image quality, fast adaptive 4DCBCT achieved a mean score of 4.1 ± 1.3, compared to conventional 4DCBCT, which had a mean score of 4.3 ± 1.2. There were no statistically significant differences in tumor visibility or image quality scores between fast adaptive 4DCBCT and conventional 4DCBCT scans.

Fast adaptive 4DCBCT achieved similar image quality scores to conventional 4DCBCT while requiring only 15% of the imaging dose and 25% of the scan time. This study confirms the clinical feasibility of adaptive scanning protocols for use in radiation therapy for lung cancer.