Automated planning for image-guided radiotherapy

Cagni, Elisabetta 2022. Automated planning for image-guided radiotherapy. PhD Thesis, Cardiff University. Item availability restricted.

Preview	PDF - Accepted Post-Print Version Download (44MB) | Preview
	PDF (Cardiff University Electronic Publication Form) - Supplemental Material Restricted to Repository staff only Download (305kB)

Abstract

Advanced radiotherapy delivery approaches have substantially increased opportunities for sparing organs at risk with proven clinical impact. Ideally, for each individual patient, the treatment plan maximally exploits the full potential of the applied delivery technique. Currently, most treatment plans are generated with interactive trial-and-error planning (‘manual planning’). It is well-known that plan quality in manual planning may be sub-optimal, e.g. depending on experience and ambition of the planner, and on allotted planning time. In recent years, several systems for automated plan generation have been developed, often resulting in enhanced plan quality compared to manual planning. Both in manual and automated planning, human evaluation and judgement of treatment plans is crucial. During plan generation, planners usually develop a range of plans, but generally only one or two competing plans are discussed with the radiation oncologist (RO). A necessary assumption for this process to work well, is that (unknown) disparity between planners and ROs on characteristics of good/optimal plans is absent or minor. Radiotherapy is gradually evolving towards real-time adaptive radiotherapy (ART). ART has the clinical rationale of reducing normal tissue toxicity and improving tumour control through plan adaptation. In this thesis the research in ART was focused on automated methods to standardize ART in predicting the eventual need for re-planning and to assess the goodness of the process. In this thesis the differences between users in perceived quality of plans has been quantified and analysed. Inter-observer differences in plan quality scores were substantial and may result in inconsistencies in generated treatment plans. A method for ART verification, with the ability to quantify registration spatial errors and assess their dose impact at the voxel level, is presented. A systematic workflow to identify effective OAR sparing in re planning using knowledge-based methods, has been established as a step toward an on-line ART process.

Item Type:	Thesis (PhD)
Date Type:	Completion
Status:	Unpublished
Schools:	Engineering
Uncontrolled Keywords:	1) Automated treatment planning 2) Adaptive radiotherapy 3) Plan quality assessment 4) Knowledge-based planning radiotherapy 5) Deformable image registration 6) Inter-observer and intra-observer variation
Date of First Compliant Deposit:	12 September 2023
Last Modified:	12 Sep 2023 10:42
URI:	https://orca.cardiff.ac.uk/id/eprint/162400

Actions (repository staff only)

Edit Item