Quantifying The Robustness Of IMRT Treatment Plans From A Statistical Optimization Model To Account For Organ Deformation

Main Category: Radiology / Nuclear Medicine
Also Included In: Cancer / Oncology;  Prostate / Prostate Cancer;  Urology / Nephrology
Article Date: 22 Sep 2008 - 4:00 PDT

email to a friend   printer friendly   view / write opinions   rate article

Organ movement remains the biggest challenge in prostate cancer treatment. Computerised tomography (CT scans) provide incomplete information about movement of the prostate. To address this problem, a statistical model of motion as been created using Principle Component Analysis (PCA).

The dose is optimised with this pliable patient model, whilst the uncertainty of the dose quality metrics is assessed by a machine learning algorithm.

Starting from multiple CT images, a PCA model of the patient is created.

The EUD is evaluated for a multitude of geometric variations sampled from the PCA model. The PCA model serves as a generator of synthetic patient geometries. The subsequent calculation of the mean and variance of the EUD allows for a viable estimate of the expected treatment outcome along with the residual uncertainty.

The dose distribution is optimised given the patient specific mobility of the organs. Hence, a reliable plan with respect to geometrical deviations of the PTV and the OARs is obtained. Due to the independent evaluation of each geometric instance, the correlation of dose metrics between individual organs is revealed. This substantial information gain facilitates the inevitable trade off between OAR sparing and tumour control.

Based on a few CTs, the proposed method is able to create reliable and consistent treatment plans while remaining clinically practical. This is also applicable to compensate for the residual error of online-setup protocols. By quantifying the uncertainty of the treatment outcome, it provides a measure for the reliability of the plan.

Sobotta Benjamin et al. Radiooncological Clinic, University of Tübingen (Section for Biomedical Physics), Tübingen, Germany

About ESTRO 27

ESTRO 27 (September 14th to 18th, 2008) offered an outstanding scientific programme combining lectures from eminent invited speakers, proffered papers and poster discussions, teaching lectures on a wide range of topics including clinical issues, brachytherapy, radiobiology, physics and technology as well as debates on controversial topics and clinical case discussions, a special poster reception, poster discussion sessions and electronic poster viewing.

ESTRO 27 hosted the largest European exhibition in Radiotherapy with participation from all the leading manufacturers.

ESTRO 27

ESTRO (European Society for Therapeutic Radiology and Oncology)