3D-CT reconstruction from single x-ray projection

The treatment of mobile tumors remains complex in radiotherapy due to breathing-related organ movements. One solution to ensure target coverage during the process is to plan the treatment taking into account safety margins. One way to significantly reduce these safety margins would be to adapt the treatment in real time using image-guided radiation therapy. The acquisition of x-ray projections during treatment is commonly used to localise the tumor in 2D but this doesn’t provide 3D information. The aim of this work is then to reconstruct a high resolution image associated with this radiograph in order to know the 3D position of the organs at the image acquisition time. This is done using a convolutional neural network. The results show that the proposed method is able to reconstruct a 3DCT based on a 2D projection x-ray only. A very small error exists between the predicted image and the initial image and is equal to 0.027423 ± 0.000259