EP-1493: Dosimetric and geometric verification with the moving phantom of gating stereotactic lung treatment

C. Ceylan, A. Kiliç, H. Ayata, T. Ugur, M. Güden, K. Engin
<span title="">2015</span> <i title="Elsevier BV"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/tp6msqxeyjdxpo4i4ouy6t32w4" style="color: black;">Radiotherapy and Oncology</a> </i> &nbsp;
3rd ESTRO Forum 2015 S811 across eight centres between January 2014 and November 2014. All patients underwent two CT scans, one in free breathing (FB) and one in DIBH using the Spiro Dynr' X breathing control system. Planning was completed on the DIBH scan and transferred to FB scans using tissue volume equivalence on Phillips Pinnacle 3 TPS. An agreed standard atlas for whole heart and LAD contouring was followed to ensure internal consistency. Treatment techniques included a two field
more &raquo; ... al approach or a three field technique irradiating additional nodal involvement. Dose regimes prescribed were either 40Gy/15# or 50Gy/25# and patients were treated on Elekta Synergy or Versa HD linear accelerators using forward planned intensity modulated radiotherapy (IMRT). As per ICRU 83 the Mean dose to the whole heart and LAD were recorded along with the Near Maximum (2%) LAD dose. Given that the data was not normally distributed, statistical significance was validated using the Wilcoxon Signed Rank test. Results: The mean heart dose observed for DIBH was 1.02Gy (95% CI 0.98-1.05), compared with 1.69Gy (95% CI 1.61-1.77) for FB. The results indicate that DIBH led to a significant decrease in mean heart dose of 0.67Gy (p<0.001). Initial results also suggest a reduction in LAD mean and maximum doses for patients treated in DIBH (figure 1). Further results will be described. Table 1. Comparison of LAD doses in DIBH vs. FB Conclusions: DIBH significantly reduces cardiac doses during left breast treatment. This technique has clear potential to decrease long-term cardiac complications in women having breast radiotherapy. EP-1492 A framework combining image registration, respiratory motion models, and motion compensated image reconstruction Purpose/Objective: We present a general theoretical framework for combining image registration and respiratory motion models. We demonstrate this framework applied to a software phantom and real dynamic MRI, 4DCT, and Cine CT lung data. This framework has a wide range of potential applications, both in RT and other fields, including planning and guiding therapy delivery, and motion compensating PET, SPECT, and functional MR images. Materials and Methods: Respiratory motion models relate the motion of the internal anatomy, which can be difficult to directly measure during therapy delivery or long image acquisitions, to easily measured surrogate signals, such as the motion of the skin surface or spirometry data. Typically, such models are built by first using image registration to determine the motion from a number of prior images, and then fitting a correspondence model that relates the motion to the surrogate signals. Our new framework combines the image registration and model fitting into a single optimisation problem, directly fitting the correspondence model to all of the image data simultaneously. This not only leads to a more theoretically efficient and robust approach to building the motion models, but also enables the use of 'partial' imaging data such as individual MR and CT slices, or CT projection data, where it is not possible to determine the full 3D motion from a single image. Our framework can be used with a wide range of registration algorithms, including different transformation models, similarity measures and constraint terms, and a wide range of correspondence models, including linear, polynomial, and B-spline models. The framework can also be extended to incorporate motion compensated image reconstruction by iterating between the model fitting and the image reconstruction, removing the need for a static reference volume. This means it is possible to estimate both the motion and a motion compensated reconstruction from just the partial imaging data and the surrogate signals. Results: We initially implemented our framework using a demons-like registration algorithm and demonstrated it using synthetic slice and projection data from a simple 2D phantom, recovering 69-94% of the motion and giving high quality reconstructions. We now implemented the framework with the popular B-spline based NiftyReg registration software, and have applied it to real MRI, 4DCT, and Cine CT lung data, giving very promising results, with the motion estimated by the model visually matching the motion seen in the real data well. Conclusions: We have presented a general framework for combining image registration, motion modelling, and motion compensated image reconstruction. We have demonstrated this on data from a software phantom and real dynamic MRI, 4DCT, and Cine CT lung data, achieving good results in all cases.
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1016/s0167-8140(15)41485-9">doi:10.1016/s0167-8140(15)41485-9</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/zuvnpwj3czbrlcbwrwrzjnlvye">fatcat:zuvnpwj3czbrlcbwrwrzjnlvye</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20171002103105/http://publisher-connector.core.ac.uk/resourcesync/data/elsevier/pdf/b57/aHR0cDovL2FwaS5lbHNldmllci5jb20vY29udGVudC9hcnRpY2xlL3BpaS9zMDE2NzgxNDAxNTQxNDg1OQ%3D%3D.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/76/2a/762afa574e5267b0a6b8594b83c81182baa083fb.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1016/s0167-8140(15)41485-9"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="unlock alternate icon" style="background-color: #fb971f;"></i> elsevier.com </button> </a>