biostudies-literatureUnknown95(1129)Chi W<h4>Objectives</h4>To develop a rapid and accurate 4D deformable image registration (DIR) approach for online adaptive radiotherapy.<h4>Methods</h4>We propose a deep learning (DL)-based few-shot registration network (FR-Net) to generate deformation vector fields from each respiratory phase to an implicit reference image, thereby mitigating the bias introduced by the selection of reference images. The proposed FR-Net is pretrained with limited unlabeled 4D data and further optimized by maximizing the intensity similarity of one specific four-dimensional computed tomography (4DCT) scan. Because of the learning ability of DL models, the few-shot learning strategy facilitates the generalization of the model to other 4D data sets and the acceleration of the optimization process.<h4>Results</h4>The British journal of radiology20210819https://www.ebi.ac.uk/biostudies/studies/S-EPMC8722248biostudies-literatureFew-shot learning for deformable image registration in 4DCT images.PMC8722248Xiang ZGuo FChi WfalseFew-shot learning for deformable image registration in 4DCT images.<h4>Objectives</h4>To develop a rapid and accurate 4D deformable image registration (DIR) approach for online adaptive radiotherapy.<h4>Methods</h4>We propose a deep learning (DL)-based few-shot registration network (FR-Net) to generate deformation vector fields from each respiratory phase to an implicit reference image, thereby mitigating the bias introduced by the selection of reference images. The proposed FR-Net is pretrained with limited unlabeled 4D data and further optimized by maximizing the intensity similarity of one specific four-dimensional computed tomography (4DCT) scan. Because of the learning ability of DL models, the few-shot learning strategy facilitates the generalization of the model to other 4D data sets and the acceleration of the optimization process.<h4>Results</h4>2022-01-01T00:00:00Z2022 Jan2025-04-21T15:22:34.052Z2025-04-21T15:22:34.052ZS-EPMC87222483466224210.1259/bjr.20210819