Objective:Patients are breathing freely during adjuvant proton pencil beam radiotherapy after breast conserving surgery. Fluctuation of the thorax may affect the position of the end of the proton beam flow, which needs to be precisely evaluated on a millimeter scale.Methods:For 20 patients with breast cancer treated with proton radiotherapy after breast conserving surgery, PET-CT scan was performed approximately 10 min after the end of proton radiotherapy. The images of PET-CT were processed for ROI determination and sampling line (profile) extraction on a Raystation RV workstation to calculate the actual difference between the predicted and real radioactivity from the same spatial location as obtained by PET acquisition R50. Then, the differences in the spatial location between the actual process of proton irradiation and the planned process were obtained. Depth difference values for each pair of sampling lines were presented. Results:For 20 patients with breast cancer with a median follow-up of 22 months (range 12 - 46 months), all patients survived at the last follow-up, and no radiation pneumonitis was observed during the follow-up period. Among the verification results of 21 cases, the depth difference of evenly distributed was (-0.75±1.89) mm in the primary field and (-0.82±2.06) mm in the secondary field; The depth difference of sequential treatment was (1.81±1.87) mm in the primary field and (1.32±1.74) mm in the secondary field; The depth difference of synchronous addition in the primary field was (-1.47±1.44) mm, and the depth difference in the secondary field was (-1.48±2.11) mm.Conclusion:The results of off-line PET-CT in vivo biological verification show that the accuracy of the dose boundary cut-off was within 3 mm in breast cancer patients, which meets the clinical and physician requirement for the precision in breast cancer treatment.