National and Kapodistrian University of Athens
Αξιολόγηση χωρικών και δοσιμετρικών σφαλμάτων σε σύγχρονες εφαρμογές ακτινοθεραπείας με χρήση χημικής δοσιμετρίας
Abstract
dc:descriptionStereotactic Radiosurgery-Radiotherapy (SRS) has been proven to offer increased efficacy in the treatment of neoplasms and other diseases, compared to conventional radiotherapy tech-niques, while technological advancements are enabling more frequent application of this method. Stereotactic radiotherapy is achieved through the administration of high doses in one or a few sessions. The success of the treatment depends on the accuracy of target tumor and critical organ localization, the precision of dose distribution calculations, and the accuracy of dose delivery to the patient. Magnetic Resonance Imaging (MRI) is widely used in modern radiotherapeutic applications, especially in treatment planning for stereotactic radiosurgery in intracranial therapies. This choice is justified by the excellent soft tissue contrast that MRI provides compared to Computed Tomography (CT). Thus, MRI images allow for better description and characterization of the target tumor. Specifically, for the treatment of brain metastases and other lesions, MRI is the imaging technique of choice for both the tumor-target and organs at risk. However, this choice compromises geometric accuracy, as MRI images inherently carry geometric distortion. After presenting the relevant theoretical background (Sections 1-6), in Sections 7, 8, and 9 of the experimental part of this study, three different brain stereotactic radiotherapy workflows are compared in terms of tumor localization accuracy. As described in the theoretical section, treat-ment planning in stereotactic radiosurgery units can be done entirely on MRI images (MR-only) or on MR/CT fused images (MR/CT). These are the two workflows compared, while the third involves treatment planning on MRI images corrected for geometric distortion (MR-corrected) using the mean image method. In Sections 7 and 9, the spatial accuracy of tumor target localization is examined for the three workflows using a VIP polymer gel phantom. After preparing the solution and filling the con-tainer, the phantom was imaged using CT. It was then irradiated at 26 points with a nominal diameter of 4 mm in a Gamma Knife (GK) unit. The coordinates of the irradiation points in the Leksell stereotactic space (LSS) serve as reference points. The phantom was then imaged with a 1.5T and 3T MRI scanner using a stereotactic radiotherapy protocol with T2-weighted images. The centers of mass of the irradiated areas (where polymerization occurs) were identified using an algorithm developed in MATLAB R2019a (The MathWorks, Inc., MI), and by comparing these with the reference points, the spatial inaccuracy of the MRI images was calculated. Next, MRI images were co-registered with CT images in the treatment planning system soft-ware, and the same algorithm was used to identify the irradiation centers, which were then compared with the reference centers. The MR images were corrected using the mean image method, and the same localization and inaccuracy calculation process was applied to the cor-rected images. For both MR fields, sequence-dependent distortions were first calculated using the reversed gradient technique and then the target localization accuracy for each workflow was evaluated. For the 1.5T images, the MR-corrected images showed the highest target localization accuracy (median value 0.2 mm), while the MR-only and MR/CT-fused images had greater inaccuracies (median values 0.8 mm and 0.6 mm, respectively). For the 3T images, the MR-corrected images also showed the highest accuracy (median value 0.5 mm), whereas MR-only and MR/CT fused images exhibited higher inaccuracies (median values 0.8 mm and 0.9 mm, respectively). It is worth noting that in all 3T image sets (MR-only, MR/CT-fused and MR-corrected), spatial inaccuracies increased with distance from the isocen-ter, with statistically significant correlation.In Section 8, tumor target localization accuracy was evaluated for the three workflows using 1.5T MR images of patients with multiple brain metastases. Localization accuracy was assessed by comparing the coordinates of tumor centers in corrected MR images to those in the images under evaluation. T1W MR and CT images from 15 patients (a total of 81 metastases) were used. The MR-only workflow led to spatial offsets significantly correlated with distance from the MRI isocenter, reaching up to 1.1 mm and median 0.6 mm. Similar results were observed for the MR/CT-global workflow, with a maximum deviation of 1.4 mm and median of 0.7 mm. The MR/CT-ROI workflow improved localization accuracy, resulting in a maximum deviation of 1.1 mm and median 0.4 mm. In Section 10, geometric distortions related to 3Τ MRI images induced by stereotactic frames were assessed. An acrylic phantom was used which includes 1978 control points and was de-signed to be compatible with the head coil and both stereotactic frames developed by Elekta (Leksell G-frame and Vantage frame) used for patient immobilization and definition of stereo-tactic space in SRS applications. The phantom was then scanned on a 3T MRI scanner using a brain metastasis imaging protocol. Sequence-independent distortions were assessed with and without the presence of the stereotactic frame. The phantom includes control points throughout its volume. Without a stereotactic frame, the average distortion was 0.7 mm, with a maximum of 1.6 mm. With the Leksell G-frame, there was an increase in average distortion to 0.9 mm across the entire phantom and 0.8 mm in the irradiated volume with a maximum 2.7 mm at points close to the frame base. With the Vantage frame, the average distortion was 0.7 mm, [14]with a maximum of 1.6 mm. Heavy distortions were also observed near the base of the frame, due to eddy currents induced in its closed aluminum loop.
Degree
thesis:*- Grantor dc:publisher
- National and Kapodistrian University of Athens
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Theocharis, Stefanos
- Θεοχάρης, Στέφανος
Subjects
dc:subject × 22- Μαγνητική τομογραφία
- ΑΜΣ
- Παραμόρφωση εικόνας
- Στερεοτακτική ακτινοχειρουργική
- Χημική δοσιμετρία
- Magnetic resonance
- MRI
- Image distortion
- Stereotactic radiosurgery
- Chemical dosimetry
- Φυσική
- Φυσικές Επιστήμες
- Φυσική, άλλοι τομείς
- Άλλες Ιατρικές Επιστήμες
- Ιατρική και Επιστήμες Υγείας
- Ιατρική, άλλοι τομείς
- Physical Sciences
- Natural Sciences
- Physics, miscellaneous
- Other Medical Sciences
- Medical and Health Sciences
- Medicine, miscellaneous
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/60680