Motion Compensated Reconstruction

Motion-compensated reconstruction aims to correct for patient movement during medical imaging scans, improving image quality and diagnostic accuracy. Current research heavily utilizes deep learning, employing architectures like transformers and convolutional neural networks within iterative reconstruction frameworks to estimate motion parameters (rigid or non-rigid) simultaneously with image reconstruction. These methods are applied across various imaging modalities (MRI, CT, PET) and focus on improving accuracy, efficiency, and robustness to different motion types and acceleration factors. The resulting advancements have significant implications for faster, higher-quality medical imaging, enabling improved diagnoses and treatment planning.