Single or a few views computed tomography imaging with deep neural network
View Patent ↗A method for tomographic imaging comprising acquiring [ 200 ] a set of one or more 2D projection images [ 202 ] and reconstructing [ 204 ] a 3D volumetric image [ 216 ] from the set of one or more 2D projection images [ 202 ] using a residual deep learning network comprising an encoder network, a transform module and a decoder network, wherein the reconstructing comprises: transforming [ 206 ] by the encoder network the set of one or more 2D projection images [ 202 ] to 2D features [ 208 ]; mapping [ 210 ] by the transform module the 2D features [ 208 ] to 3D features [ 212 ]; and generating [ 214 ] by the decoder network the 3D volumetric image [ 216 ] from the 3D features [ 212 ]. Preferably, the encoder network comprises 2D convolution residual blocks and the decoder network comprises 3D blocks without residual shortcuts within each of the 3D blocks.
1. A method for tomographic imaging comprising acquiring a set of one or more 2D projection images and reconstructing a 3D volumetric image from the set of one or more 2D projection images using a residual deep learning network comprising an encoder network, a transform module and a decoder network, wherein the reconstructing comprises:
transforming by the encoder network the set of one or more 2D projection images to 2D features, wherein the encoder network comprises 2D convolution residual blocks and the decoder network comprises 3D blocks without residual shortcuts within each of the 3D blocks;
mapping by the transform module the 2D features to 3D features;
generating by the decoder network the 3D volumetric image from the 3D features.
2. The method of claim 1 wherein acquiring the set of one or more 2D projection images comprises performing a computed tomography x-ray scan.
3. The method of claim 1 wherein the set of one or more 2D projection images contains no more than a single 2D projection image, and wherein reconstructing the 3D volumetric image comprises reconstructing the 3D volumetric image only from the single 2D projection image.
4. The method of claim 1 wherein the set of one or more 2D projection images contains at most two 2D projection images, and wherein reconstructing the 3D volumetric image comprises reconstructing the 3D volumetric image from no more than the at most two 2D projection images.
5. The method of claim 1 wherein the set of one or more 2D projection images contains at most five 2D projection images, and wherein reconstructing the 3D volumetric image comprises reconstructing the 3D volumetric image from no more than the at most five 2D projection images.
6. The method of claim 1 wherein the residual deep learning network is trained using synthetic training data comprising ground truth 3D volumetric images and corresponding 2D projection images synthesized from the ground truth 3D volumetric images.