Abstract #0798

Deep learning-based 4D synthetic CT for lung radiotherapy

Matteo Maspero^1,2, Kirsten M. Kerkering^2,3, Tom Bruijnen^1,2, Mark H. F. Savenije^1,2, Joost J. C. Verhoeff¹, Christoph Kolbitsch³, and Cornelis A. T. van den Berg^1,2

¹Radiotherapy, Division of Imaging & Oncology, UMC Utrecht, Utrecht, Netherlands, ²Center for Computational imaging group for MR diagnostic & therapy, Center for Image Sciences, UMC Utrecht, Utrecht, Netherlands, ³Physikalisch-Technische Bundesanstalt (PTB), Braunschweig and Berlin, Germany

The feasibility of generating synthetic CT for lung tumours from 4D MRI was investigated. A combination of multi-view 2D networks proved to be robust against image artefact and generated sCTs that enabled dose calculation on midposition sCTs. The proposed approach facilitates adaptive MR-guided radiotherapy reducing the time from patient positioning to irradiation and enables quality assurance with dose accumulation based on 4D MRI.

This abstract and the presentation materials are available to members only; a login is required.

Join Here