Adaptive radiation therapy for compensation of errors in patient setup and treatment delivery
A schematic two-dimensional illustration of how the measurement model maps patient setup errors to patient-beam misalignments. The coordinates of a presumed bony structure in each beam’s coordinate system is mapped onto the corresponding image planes.
The environment for numerical simulation of an adaptive treatment. An inverse planning is performed in Pinnacle3 and this plan is used by MATLAB to simulate the evolution of fluence profiles and to compute corrections with the adaptation method. To evaluate the results, Pinnacle computes fraction doses based on the corrected fluences and these doses are accumulated in MATLAB. Finally, the accumulated doses are evaluated in Pinnacle.
The studied prostate patient. The regions of interest are CTV, PTV, prostate, bladder and rectum.
Doses for the nominal plans, corresponding to a patient without any setup errors. The plans displayed are the nominal-CTV plan (solid) and the nominal-PTV plan (dashed). The displayed regions of interest are: CTV, prostate, bladder and rectum. The DVH-objective arrows correspond to the objectives used when designing the CTV-plan. Both plans achieve acceptable target coverage but organ at risk protection was inferior for the nominal-PTV plan.
Results of the simulation with random setup errors. DVH plots for CTV and rectum (left) and for CTV and bladder (right). Solid: adapted-CTV treatment, dashed: uncorrected-PTV treatment, thick solid: uncorrected-CTV treatment. It can be seen from the figure that both the adapted-CTV treatment and the uncorrected-PTV treatment achieves target coverage but that organ at risk is better for the adapted plan. If the CTV-plan is used without adaptation (uncorrected-CTV), coverage of the target volume is not achieved.
Results of the simulation with only random errors. Accumulated fluence profiles are given for the nominal-CTV plan (left), the uncorrected-CTV plan (middle), and the adapted-CTV plan (right). The random setup errors smoothes the nominal profile as is seen in the middle picture. The adaptation manages to decrease this effect, as is seen in the right figure.
Results of the simulation with random and systematic errors. DVH plots for CTV and rectum (left) and for CTV and bladder (right). Solid: adapted-CTV treatment, dashed: uncorrected-PTV treatment, thick-solid: PTV-plan based treatment where only the systematic error has been compensated for. There is a cold spot in the target for the uncorrected treatment and bladder dose is increased. Rectal dose is decreased, as the systematic error is in the posterior direction. The adapted treatment achieves tumor coverage and low dose to the organs at risk. If the PTV-plan based treatment is corrected for the systematic error, the cold spot is eliminated and the shift of dose from the rectum to the bladder disappears. Overall organ at risk dose is still high.
Dose for the adapted-CTV treatment (left) and for the uncorrected-PTV treatment (right). The influence of the systematic error is clear from the anterior shift of the high dose region.
True (solid), estimated (dashed), and predicted (dotted) setup errors in the lateral (top), anterior/posterior (middle) and the inferior/superior (bottom) direction. The rings mark the instances for which the corrections are computed. It can be seen from the figure that the setup error estimates are of high quality, and that the setup error predictions captures the systematic error.
The treatment objectives used to design the CTV-based plan.
The treatment objectives used to design the PTV-based plan.
Article metrics loading...
Full text loading...