Change search
ReferencesLink to record
Permanent link

Direct link
Probabilistic treatment planning based on dose coverage: How to quantify and minimize the effects of geometric uncertainties in radiotherapy
Uppsala University, Disciplinary Domain of Medicine and Pharmacy, Faculty of Medicine, Department of Immunology, Genetics and Pathology, Medical Radiation Science. (Anders Ahnesjö)
2016 (English)Doctoral thesis, comprehensive summary (Other academic)
Abstract [en]

Traditionally, uncertainties are handled by expanding the irradiated volume to ensure target dose coverage to a certain probability. The uncertainties arise from e.g. the uncertainty in positioning of the patient at every fraction, organ motion and in defining the region of interests on the acquired images. The applied margins are inherently population based and do not exploit the geometry of the individual patient. Probabilistic planning on the other hand incorporates the uncertainties directly into the treatment optimization and therefore has more degrees of freedom to tailor the dose distribution to the individual patient. The aim of this thesis is to create a framework for probabilistic evaluation and optimization based on the concept of dose coverage probabilities. Several computational challenges for this purpose are addressed in this thesis.

The accuracy of the fraction by fraction accumulated dose depends directly on the accuracy of the deformable image registration (DIR). Using the simulation framework, we could quantify the requirements on the DIR to 2 mm or less for a 3% uncertainty in the target dose coverage.

Probabilistic planning is computationally intensive since many hundred treatments must be simulated for sufficient statistical accuracy in the calculated treatment outcome. A fast dose calculation algorithm was developed based on the perturbation of a pre-calculated dose distribution with the local ratio of the simulated treatment’s fluence and the fluence of the pre-calculated dose. A speedup factor of ~1000 compared to full dose calculation was achieved with near identical dose coverage probabilities for a prostate treatment.

For some body sites, such as the cervix dataset in this work, organ motion must be included for realistic treatment simulation. A statistical shape model (SSM) based on principal component analysis (PCA) provided the samples of deformation. Seven eigenmodes from the PCA was sufficient to model the dosimetric impact of the interfraction deformation.

A probabilistic optimization method was developed using constructs from risk management of stock portfolios that enabled the dose planner to request a target dose coverage probability. Probabilistic optimization was for the first time applied to dataset from cervical cancer patients where the SSM provided samples of deformation. The average dose coverage probability of all patients in the dataset was within 1% of the requested.

Place, publisher, year, edition, pages
Uppsala: Acta Universitatis Upsaliensis, 2016. , 51 p.
Series
Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine, ISSN 1651-6206 ; 1264
Keyword [en]
Radiotherapy, treatment simulation, probabilistic planning, dose calculation, probabilistic optimization, statistical shape model
National Category
Other Physics Topics
Research subject
Medical Radiophysics
Identifiers
URN: urn:nbn:se:uu:diva-304180ISBN: 978-91-554-9720-0OAI: oai:DiVA.org:uu-304180DiVA: diva2:1034233
Public defence
2016-11-25, Skoogsalen, Ing. 78-79, Akademiska Sjukhuset, Uppsala, 13:00 (English)
Opponent
Supervisors
Available from: 2016-11-03 Created: 2016-10-03 Last updated: 2016-11-16
List of papers
1. Dose mapping sensitivity to deformable registration uncertainties in fractionated radiotherapy – applied to prostate proton treatments
Open this publication in new window or tab >>Dose mapping sensitivity to deformable registration uncertainties in fractionated radiotherapy – applied to prostate proton treatments
2013 (English)In: BMC Medical Physics, ISSN 1756-6649, Vol. 13, no 2Article in journal (Refereed) Published
Abstract [en]

Background

Calculation of accumulated dose in fractionated radiotherapy based on spatial mapping of the dose points generally requires deformable image registration (DIR). The accuracy of the accumulated dose thus depends heavily on the DIR quality. This motivates investigations of how the registration uncertainty influences dose planning objectives and treatment outcome predictions.

A framework was developed where the dose mapping can be associated with a variable known uncertainty to simulate the DIR uncertainties in a clinical workflow. The framework enabled us to study the dependence of dose planning metrics, and the predicted treatment outcome, on the DIR uncertainty. The additional planning margin needed to compensate for the dose mapping uncertainties can also be determined. We applied the simulation framework to a hypofractionated proton treatment of the prostate using two different scanning beam spot sizes to also study the dose mapping sensitivity to penumbra widths.

Results

The planning parameter most sensitive to the DIR uncertainty was found to be the targetD95. We found that the registration mean absolute error needs to be ≤0.20 cm to obtain an uncertainty better than 3% of the calculated D95 for intermediate sized penumbras. Use of larger margins in constructing PTV from CTV relaxed the registration uncertainty requirements to the cost of increased dose burdens to the surrounding organs at risk.

Conclusions

The DIR uncertainty requirements should be considered in an adaptive radiotherapy workflow since this uncertainty can have significant impact on the accumulated dose. The simulation framework enabled quantification of the accuracy requirement for DIR algorithms to provide satisfactory clinical accuracy in the accumulated dose.

Keyword
Radiotherapy; Adaptive radiotherapy; Dose tracking; Dose mapping; Dose accumulation; Dose accumulation accuracy; Deformable image registration; Non-rigid image registration; Protons
National Category
Other Medical Engineering
Research subject
Medical Radiophysics
Identifiers
urn:nbn:se:uu:diva-224059 (URN)10.1186/1756-6649-13-2 (DOI)
Available from: 2014-04-30 Created: 2014-04-30 Last updated: 2016-10-11Bibliographically approved
2. Fast dose algorithm for generation of dose coverage probability for robustness analysis of fractionated radiotherapy
Open this publication in new window or tab >>Fast dose algorithm for generation of dose coverage probability for robustness analysis of fractionated radiotherapy
2015 (English)In: Physics in Medicine and Biology, ISSN 0031-9155, E-ISSN 1361-6560, Vol. 60, no 14, 5439-5454 p.Article in journal (Refereed) Published
Abstract [en]

A fast algorithm is constructed to facilitate dose calculation for a large number of randomly sampled treatment scenarios, each representing a possible realisation of a full treatment with geometric, fraction specific displacements for an arbitrary number of fractions. The algorithm is applied to construct a dose volume coverage probability map (DVCM) based on dose calculated for several hundred treatment scenarios to enable the probabilistic evaluation of a treatment plan.For each treatment scenario, the algorithm calculates the total dose by perturbing a pre-calculated dose, separately for the primary and scatter dose components, for the nominal conditions. The ratio of the scenario specific accumulated fluence, and the average fluence for an infinite number of fractions is used to perturb the pre-calculated dose. Irregularities in the accumulated fluence may cause numerical instabilities in the ratio, which is mitigated by regularisation through convolution with a dose pencil kernel.Compared to full dose calculations the algorithm demonstrates a speedup factor of ~1000. The comparisons to full calculations show a 99% gamma index (2%/2 mm) pass rate for a single highly modulated beam in a virtual water phantom subject to setup errors during five fractions. The gamma comparison shows a 100% pass rate in a moving tumour irradiated by a single beam in a lung-like virtual phantom. DVCM iso-probability lines computed with the fast algorithm, and with full dose calculation for each of the fractions, for a hypo-fractionated prostate case treated with rotational arc therapy treatment were almost indistinguishable.

Keyword
Radiotherapy dose calculation
National Category
Bioinformatics (Computational Biology)
Research subject
Physics
Identifiers
urn:nbn:se:uu:diva-258095 (URN)10.1088/0031-9155/60/14/5439 (DOI)000357620400008 ()26118844 (PubMedID)
Available from: 2015-07-10 Created: 2015-07-10 Last updated: 2016-10-11Bibliographically approved
3. Dose coverage calculation using a statistical shape model - applied to cervical cancer radiotherapy
Open this publication in new window or tab >>Dose coverage calculation using a statistical shape model - applied to cervical cancer radiotherapy
Show others...
2016 (English)In: Physics in Medicine and Biology, ISSN 0031-9155, E-ISSN 1361-6560Article in journal (Refereed) Accepted
Abstract [en]

A comprehensive treatment simulation framework is created where a patient geometry sampler built from a population based statistical shape model (SSM) provides efficient sampling of the deformation at every fraction to create a treatment planning tool for evaluating the dose coverage probabilities. 

Deformable image registration from repeat imaging extracts the intra-patient deformations which are mapped to an average patient serving as a common frame of reference. The SSM is created by extracting the dominating eigenmodes by principal component analysis of the deformations from all patients in the training set. The sampling of a deformation is reduced to sampling of the eigenmode weights from their respective probability distribution defined by the eigenvalues. The framework is applied to five inoperable cervix cancer patients. The images as well as manually drawn contours are used in the registration to handle the large inter-fraction deformations. 

Fifteen eigenmodes were needed to capture 90% of the variance in the deformation and seven eigenmodes to reach stability in the simulated dose coverage probabilities. The created probabilistic evaluation tool provided more information regarding the trade-off between target dose coverage and organ at risk sparing than the original planning based on a single image series.

Keyword
Radiotherapy, probabilistic, statistical shape model, principal component analysis, deformable image registration, cervix
National Category
Other Medical Engineering
Research subject
Medical Radiophysics
Identifiers
urn:nbn:se:uu:diva-304979 (URN)
Available from: 2016-10-11 Created: 2016-10-11 Last updated: 2016-10-11
4. Probabilistic optimization of the dose coverage – applied to radiotherapy treatment planning of cervical cancer
Open this publication in new window or tab >>Probabilistic optimization of the dose coverage – applied to radiotherapy treatment planning of cervical cancer
(English)Manuscript (preprint) (Other academic)
Abstract [en]

Probabilistic (or robust) optimization is an alternative to margins for handling geometrical uncertainties in treatment planning of radiotherapy where the uncertainties are explicitly incorporated in the optimization through sampling of treatment scenarios. We present a probabilistic method based on statistical measures close to those behind conventional margin based planning. The dose planner requests a dose coverage to a specified probability, which the algorithm then attempts to fulfil.

We define the Percentile UnderDosage (PUD) as a measure of the target minimum dose coverage probability, i.e. the dose coverage that a treatment plan meet or exceed to a given probability. Margin based planning commonly use the implicit probabilistic treatment criteria that the 90th PUD is at least 95% of the intended dose. For optimization we use the Expected Percentile UnderDosage (EPUD) defined as the average dose coverage below a given PUD. The EPUD is, in contrast to PUD, a convex measure and hence standard optimization techniques can be used to find the optimal treatment plan. We propose an iterative method where a treatment optimization is performed at each iteration and the EPUD tolerance is adjusted gradually until a desired PUD is met.

We demonstrate our proposed probabilistic planning method for cervical cancer patients. The uncertainty caused by organ deformation is explicitly included in the probabilistic optimization where a statistical shape model is used to sample scenarios with different deformations. For all patients in this work, the iterative process of finding the EPUD tolerance converged in less than 10 iterations to within 0.1Gy of the requested PUD even though a conservative update scheme was used. The resulting estimated PUD was validated based on 1000 simulated scenarios not part of the optimization yielding an agreement within 1.2% of the requested PUD.

National Category
Other Physics Topics
Research subject
Medical Radiophysics
Identifiers
urn:nbn:se:uu:diva-304982 (URN)
Available from: 2016-10-11 Created: 2016-10-11 Last updated: 2016-10-11

Open Access in DiVA

fulltext(1208 kB)69 downloads
File information
File name FULLTEXT01.pdfFile size 1208 kBChecksum SHA-512
351872d76ef4859a9445400389ea2ec4ce727189079980ab0172d05c97b5a9c0ecc798ab59b59bde40a9ec85b77bfb9ee118339ee26d40deae1e36d1e290166c
Type fulltextMimetype application/pdf
Buy this publication >>

Search in DiVA

By author/editor
Tilly, David
By organisation
Medical Radiation Science
Other Physics Topics

Search outside of DiVA

GoogleGoogle Scholar
Total: 69 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Total: 363 hits
ReferencesLink to record
Permanent link

Direct link