The imaging expert: The use of (imaging) biomarkers in AI prediction models of outcome
Ángel Alberich Bayarri,
Spain
SP-0972
Abstract
The imaging expert: The use of (imaging) biomarkers in AI prediction models of outcome
Authors: Ángel Alberich Bayarri1
1Quibim , Quibim, Valencia, Spain
Show Affiliations
Hide Affiliations
Abstract Text
Tumor heterogeneity and habitats are not mostly taken into account in current radiotherapy (RT) planning procedures. The extraction of quantitative imaging biomarkers allow for the voxelwise characterisation of lesions, ensuring spatial coherence across different imaging series by the use of optimized registration algorithms. Further development in the field of dose painting and formats for the integration of parametric maps (i.e. parametric DICOM) with RT planning software needs to be explored. Beyond planning, the assessment of RT treatment response as well as early detection of recurrence are key applications of the spatial biology capabilities that bring imaging biomarkers. In this presentation I will review the process for the extraction of accurate imaging biomarkers, accounting for sources of variability, and how to use them for RT planning as well as treatment response evaluation and prediction of recurrence.