Evaluation of patient-specific quality assurance tolerances based on statistical process control in pencil beam...

This study aimed to determine the action limit (AL_SPC) and tolerance limit (TL_SPC) using statistical process control (SPC) for patient-specific quality assurance (PSQA) in pencil beam scanning (PBS) proton therapy.

We analyzed PBS-treated plans: prostate (n = 152); brain, head and neck (BHN; n = 34); and thorax, abdomen, and pelvis (TAP; n = 77). PSQA used a two-dimensional detector in a water phantom to measure planar dose distributions in the isocenter (IC), proximal, and distal planes. Dose distributions were evaluated using gamma analysis with 3%/2 mm and 3%/3 mm criteria. The AL_SPC and TL_SPC for the gamma passing rate (GPR) were calculated based on the SPC for each site and plane.

With the 3%/2 mm criteria, AL_SPC at the IC plane exceeded 90% for all sites. The TL_SPC exceeded 95% only for the IC and proximal planes of the prostate plans. Conversely, with the 3%/3 mm criteria, both the AL_SPC and TL_SPC at the IC plane exceeded 95% across all sites. The distal plane had lower AL_SPC and TL_SPC values than the IC and proximal planes. Prostate plans consistently showed higher AL_SPC and TL_SPC values than the other sites.

Determining site- and plane-specific AL_SPC and TL_SPC using SPC is effective for rigorous PSQA in PBS proton therapy. Consistent with the AAPM TG-218 report recommendations for photon therapy, applying an AL of 90% and a TL of 95% at the IC plane with the 3%/2 mm criteria appears feasible; the proximal and distal planes require institution-specific limits.