VolumeNeRF: CT Volume Reconstruction from a Single Projection View

Computed tomography (CT) plays a significant role in clinical practice by providing detailed three-dimensional information, aiding in accurate assessment of various diseases. However, CT imaging requires a large number of X-ray projections from different angles and exposes patients to high doses of radiation. Here we propose VolumeNeRF, based on neural radiance fields (NeRF), for reconstructing CT volumes from a single-view X-ray. During training, our network learns to generate a continuous representation of the CT scan conditioned on the input X-ray image and render an X-ray image similar to the input from the same viewpoint as the input. Considering the ill-posedness and the complexity of the single-perspective generation task, we introduce likelihood images and the average CT images to incorporate prior anatomical knowledge. A novel projection attention module is designed to help the model learn the spatial correspondence between voxels in CT images and pixels in X-ray images during the imaging process. Extensive experiments conducted on a publicly available chest CT dataset show that our VolumeNeRF achieves better performance than other state-of-the-art methods. Our code is available at https://www.github.com/Aurora132/VolumeNeRF .