Accounting for tumour deformation in real time

Tumours of thorax, abdomen and prostate undergo continuous motion including translation, rotation and deformation. State-of-the-art treatment techniques can correct only for tumour translation and rotation in real time. Studies show deformation for lung, liver and pancreatic tumours of up to 5mm and differential motion (deformation of multi-tumour system) between primary lung or prostate tumours and involved lymph nodes of more than 15mm. In current practice, the problem of deformation is still addressed by extending the treatment margin to ensure tumour coverage, which increases normal tissue toxicity and limits local disease control.

Deformation magnitude

We are investigating a radiotherapy system capable of accounting for tumour deformation in real time using dynamic multileaf collimator (DMLC). Currently, the focus is to develop the DMLC tracking system provided that the deformation has been acquired. A prototype deformation tracking system has been developed.
Patient is simultaneously imaged during treatment to acquire the tumour deformation. The obtained displacement vector field (DVF) is then used to adapt the planned MLC aperture. The MLC leaves then move to the new position to compensate for the tumour deformation.

Real-time DMLC deformation tracking flowchart

Proof-of-principle phantom experiments have been conducted. The experimental setup, as well as the animated portal images overlaid by DVFs, are shown below.

deformation phantoms
Animated portal images of single tumor phantom deformation tracking experiment
Animated portal images of tumor system phantom deformation tracking experiment

Preliminary results showed that the target coverage was substantianlly improved by deformation tracking compared to no tracking. The geometric target coverage metric Au+Ao (Poulsen et al. IJROBP 83, 2012), of the phantom experiments are plotted as a function of the deformation magnitude below.

phantom experiment result plot

Sliding-window Intensity-modulated radiation therapy (IMRT) treatment delivered with enabled deformation tracking was simulated. A realistic simulation of the deformation tracking in an MRI-Linac environment was performed for a lung cancer patient with previously acquired MR images. The animated simulations are shown below respectively.

Single tumor deformation tracking during IMRT simulation
Tumor system deformation tracking during IMRT simulation
patient MRI conformal simulation

* This work received funding support from Cure Cancer Australia Foundation for 2012-2013 and from NHMRC Project Grant for 2013-2016.