ESTRO 2022 - Abstract Book
S1607
Abstract book
ESTRO 2022
Conclusion In vivo PSD determines accurately the dwell times, and it is a quality assurance procedure that allows verification of dose delivery in each fraction. In view of the good results obtained in time measurements with the PSD detector, a robust calibration procedure is being designed to calibrate the PSD detector in terms of absorbed dose.
PO-1803 Simulated liver implantation (SLIM): Towards fully robotic LDR brachytherapy for liver tumors
P. Aumüller 1 , A. Rothfuss 2 , M. Ehmann 3 , F.A. Giordano 4 , S. Clausen 3
1 University Medical Centre Mannheim, University Heidelberg, Germany., Department of Radiation Oncology, Mannheim, Germany; 2 Fraunhofer , IPA, Mannheim, Germany; 3 University Medical Centre Mannheim, University Heidelberg, Germany, Department of Radiation Oncology, Mannheim, Germany; 4 University Hospital Bonn, Department of Radiation Oncology, Bonn, Germany Purpose or Objective Low dose rate brachytherapy (LDR-BT) may be a minimally invasive and highly effective therapy to treat liver tumors, especially using (semi- or fully-) robotic implantation. We recently introduced an inverse treatment planning system for robotic LDR-BT with path planning independent of a rigid template to guide needles to the target volume [1]. The spatial registration of each implanted seed with direct feedback to the planning system during the implantation procedure enables adapting the remaining needle and seed positions to correct systematic and statistical dose errors introduced to the original treatment. We here show results of a simulated annealing adaption (SAA) algorithm that modifies subsequent needle positions online to optimize the original treatment planning objectives in a simulation study in the setting of LDR-BT for liver tumors. Materials and Methods We performed simulated liver implantations (SLIMs) by sequentially adding values of a normally distributed deviation based on the uncertainty of the placement accuracy (PA) to the tip coordinate of each needle that is about to be implanted and recalculating the dose distribution resulting from the already implanted and the remaining pre-planned needle and seed positions. The SAA algorithm uses the MATLAB function simulannealbnd (The MathWorks-R2021a) to vary the remaining needle tips after each simulated needle implantation in the range of ±1.5mm and to minimize the objective function designed to spare organ at risks (OARs), increase the V100 and decrease the V200. All plan parameters are optimized to remain near the pre- treatment values except a restrictive dose constraint to 1% of the aorta (D1). Ten SLIMs of one treatment plan for a liver tumor in an abdominal phantom with SAA or no SAA (NSAA) with a PA of ±1mm and ±3mm are compared. Results The SLIM V100 were for a PA of ±1mm SAA (98.5±0.3)% , NSAA (99.0±0.4)% and for ±3mm (97.0±1.2)% and (96.0±1.3)% respectively. SAA reduces the D1(aorta) and maintains other treatment aims. A higher PA of ±1mm against ±3mm results in a higher plan robustness due to smaller standard deviations of the planning parameters.
Fig. 1 a): Abdominal phantom with the pre-treatment plan. b.1) Close view. b.2) SLIM with SAA and PA±1mm. b.3) NSAA ±{1;3}mm . b.4) SAA ±3mm .
Made with FlippingBook Digital Publishing Software