ESTRO 2024 - Abstract Book

S3958

Physics - Image acquisition and processing

ESTRO 2024

Target delineation is one of the most important steps in effective treatment, especially for benign tumours where irradiation of healthy tissue is even more problematic. In the treatment of a vestibular schwannoma, delineation is performed on contrast-enhanced MRI. However, by its nature, the MRI image has geometric distortions caused by the electromagnetic fields that generate the image that need to be corrected in relation to the CT. These distortions can have a direct dosimetric impact on the delineation, resulting in treatment of the wrong target and thus compromising of the entire treatment.[2].This correction is performed automatically by Brainlab Cranial Distortion software[3] and allows for more accurate and correct delineation, resulting in more efficient treatment. This retrospective study aims to evaluate the dosimetric impact of the Brainlab Elements Cranial curvature correction algorithm in the treatment of vestibular schwannomas with radiosurgery, focusing on the differences in dose distribution to healthy tissue and organs at risk.

Material/Methods:

A retrospective comparative study was carried out on 7 (seven) patients whose targets (GTV) were delineated using the Elements BrainLab software. The CT and MRI slices used were 1 mm. In the first group, distortion correction of the MRI image was performed in the CT fusion; in the second group, MRI fusion with CT was performed without distortion correction.

For both images, the same expert delineated the targets in the respective MRIs for the corrected and uncorrected resonance, removing the difference in contour between observers.

The DICE similarity coefficient was used to assess the effect of distortion on the target contour:

DSC=2*A ∩ B/(A+B)

Where A is the contoured target on the MRI without correction and B with correction.

The patients were treated on the TrueBeam linear accelerator (Varian) a single dose of 12 Gy (99% target coverage) in six patients and 5x5 Gy in one patient.

The impact on the Conformity of the prescribed dose to the target was assessed using two Conformity indexes, the first of which considers only the prescribed volume and the target volume, regardless of the intersection between the volumes[4]:

CI RTOG = PIV/TV

Where PIV is the prescription isodose volume and TV the tumor volume.

The second index: IC Paddick , proposed by Paddick [5] takes into account the intersection between target and prescription volumes, quantifying how much of the prescription is off target and how much of the target does not receive the prescription dose:

CI Paddick = TV² PIV /PIV*TV

The dose distributions as well as the coverage on the target were evaluated with the Dmín, which is the largest isodose that covers the entire target volume, and with the target volume that is covered by the prescription isodose.