ESTRO 2024 - Abstract Book

S3959

Physics - Image acquisition and processing

ESTRO 2024

Results:

There were significant changes with relevant treatment implications if distortion correction had not been performed. This effect becomes more relevant for small volume lesions, as can be seen in Figure 1. For small-volume lesions, small intersection variations will produce low DICE values and therefore indicate a low degree of similarity, showing that the corrected and uncorrected contours have a high degree of difference.

On the other hand, the Conformity and CI Paddick indexes showed that coverage and Dmín were extremely affected. Figure 1 shows that CI Paddick dropped from an average of 0.9 to 0.79 from corrected to uncorrected MRI, indicating a lack of intersection between the prescribed dose and the target volume. Clinically, we would be increasing the dose to healthy tissue and decreasing the dose to the tumour volume.On the other hand, the volume of PTV covered by the prescribed dose fell from an average of 99% to 94.8%. The Dmín decreased from 86.9% to 75.9%, showing that the dose covering the entire tumour decreased significantly.The statistical behaviour of these indices is shown in Figure 2: