Interactive use of on-line portal imaging in pelvic radiation

We have evaluated a fluoroscopic on-line portal imaging system in routine clinical radiotherapy, involving the treatment of 566 pelvic fields on 13 patients. The image was typically generated by delivering a radiation dose of 6-8 cGy. Comparison between portal image and simulator film was done by eye and all visible errors were corrected before continuing irradiation. If possible, these corrections were performed from outside the treatment room by moving the patient couch by remote control or by changing collimator parameters. Adjustments were performed on 289/530 (54.5%) evaluable fields or 229/278 (82.4%) evaluable patient set-ups. The lateral couch position was most frequently adjusted (n = 254). The absolute values of the adjustments were 6.8 mm mean (SD 6.6 mm) with a maximum of 40 mm. All absolute values of adjustments exceeding 25 mm were recorded in one patient and those exceeding 15 mm were observed in two patients. Both patients were obese females. Adjustments exceeding 5 mm were observed in all 13 patients. Related to the use of on-line portal imaging, treatment time was increased by a median of 36.5% (mean 45.8%; SD 42.1%). The range was 7.7 to 442%. The fraction of the total treatment time to perform corrections was 22.7% median (mean: 26.0; SD: 11.8%). Statistically significant systematic in-plane errors were found in 7/13 patients. A systematic error was detected on the lateral position of the field in five patients. In one patient a systematic error of the longitudinal field position and in one patient a rotational error was detected. For adjustments in the lateral direction the present method does not allow to detect lateral shifts of less than 2 mm. For adjustments in the longitudinal direction the sensitivity could not be estimated but the available data suggest that 80% of errors < or = 5 mm were not adjusted. In obese patients, random errors may be surprisingly large.