A critical approach to the clinical use of deformable image registration software. In response to Meijneke et al.

van Rijssel, Michael J., Dahele, Max, Verbakel, Wilko F. A. R., Rosario, Tezontl S.


Radiotherapy & Oncology 112 (3), p. 447-448


We congratulate Meijneke et al. on their publication describing their experience with thoracic re-irradiation [1] where they emphasized the use of non-rigid registration (deformable image registration, DIR) for dose deformation and accumulation. Although they provided references in support of their DIR algorithm, we note that these do not specifically demonstrate the accuracy of their DIR algorithm and subsequent dose accumulation in the lung. Commercial DIR software packages are now widely available and while developments in DIR are welcome, methods to allow clinical users to evaluate the uncertainty associated with its use in various situations have not kept pace with its diffusion into clinical practice.