Evaluation of a merging-based approach for improving patient-specific quality assurance in proton and carbon-ion...

Ionization chamber arrays are inherently limited in spatial resolution, which can lead to the underdetection of potential clinical errors in particle therapy. The study aimed to validate the clinical efficacy of a multi-position dose merging approach, evaluating its capacity to enhance spatial resolution and improve error detection sensitivity relative to conventional single-isocentre quality assurance (QA) protocols.

Dose distributions were measured using PTW 729 ^XDR and 1500 ^XDR ionization chamber arrays at three clinically relevant depths within the spread-out Bragg peak (SOBP) region. Two datasets were generated: a merged dataset formed by combining dose distributions acquired at the isocentre and at the 5‑mm detector shifts (lateral x‑axis shifts for the 729 ^XDR array and longitudinal y‑axis shifts for the 1500 ^XDR array), and a single-isocentre dataset. The translational accuracy of the treatment couch was verified using a laser alignment system and a graduated scale. Bland-Altman analysis was performed to quantify the agreement and relative bias (%) between the merged and single-isocentre approaches. The receiver operating characteristic (ROC) curve was employed to compare the sensitivity of the two methods in detecting intentional 1 and 2 mm range errors.

The single-isocentre cohort showed higher mean gamma pass rates (n = 2616) than the merged cohort (P < 0.05). The merging method resulted in a greater reduction in gamma passing rates for carbon-ion and pencil beam scanning (PBS) dose verification compared to proton and uniform scanning dose verification (3.85% for carbon-ion versus 1.93% for proton; 3.49% for PBS versus 1.69% for uniform scanning). Bland-Altman analysis revealed a 95% limits of agreement (LoA) width of 5.28% between the two methods under 3%/2 mm gamma criteria with global normalization. ROC analysis demonstrated that the merging method yielded greater sensitivity for detecting 1 and 2 mm range errors than the single-isocentre approach, improving sensitivity for detecting 2 mm errors by 19.06% in carbon-ion therapy versus 6.69% in proton therapy.

The spatial resolution limitations of conventional detector arrays may artificially inflate gamma pass rates, introducing a risk of false-negative approval for clinically unacceptable treatment plans. The multi-position dose merging approach mitigates this limitation by enhancing spatial resolution, supporting its clinical utility as a practical, workflow-integratable strategy to improve the accuracy and effectiveness of proton and carbon-ion radiotherapy QA protocols.