Background and purpose: The objective of this investigation was a direct comparison of the dosimetry of CT-based and radiograph- based postplanning procedures for seed implants.
Patients and methods: CT- and radiograph-based postplans were carried out for eight iodine-125 ((125)I) seed implant patients with a commercial treatment-planning system (TPS). To assess a direct comparison of the dosimetric indices (D90, V100, V400), the radiograph-based seed coordinates were transformed to the coordinate system of the CT postplan. Afterwards, the CT-based seed positions were replaced by the radiograph-based coordinates in the TPS and the dose distribution was recalculated.
Results: The computations demonstrated that the radiograph-based dosimetric values for the prostate (D(p)90, V(p)100, and V(p)400) were on average lower than the values of the CT postplan. Normalized to the CT postplan the following mean values were found: D(p)90: 90.6% (standard deviation [SD]: 9.0%), V(p)100: 86.1% (SD: 14.7%), and V(p)400: 79.4% (SD: 14.4%). For three out of the eight patients the D(p)90 decreased to 90% of the initial CT postplan values. The reason for this dosimetric difference is supposed to be evoked by an error of the reconstruction software used. It was detected that the TPS algorithm assigned some sources to wrong coordinates, partly out of the prostate gland.
Conclusion: The radiograph-based postplanning technique of the investigated TPS should only be used in combination with CT postplanning. Furthermore, complex testing procedures of reconstruction algorithms are recommended to minimize calculation errors.