System and method for 3d imaging reconstruction using dual-domain neural network

A system, comprising:

• one or more processors;
• and
• one or more non-transitory computer-readable media storing instructions that, when executed by the one or more processors, cause the system to:
• capture a plurality of 2D projections of a target that cover a limited angle of the target;
• perform geometric calibration and physical correction on the plurality of 2D projections of the target to obtain a plurality of physically corrected and geometrically calibrated 2D projections;
• generate model input data comprising a sinogram with a limited angle coverage of the target based on the plurality of physically corrected and geometrically calibrated 2D projections;
• input the model input data into a trained neural network, wherein the neural network is trained, using training data including historical CT scanning data having large-angle or full-angle scanning and corresponding ground-truth sinograms and ground-truth 3D reconstructions, to generate a 3D volumetric image of the target based on a sinogram generated from limited angle 2D projections;
• and
• obtain the 3D volumetric image of the target as an output of the trained neural network.

A non-transitory computer-readable storage medium storing instructions that, when executed by one or more processors of a computer system, cause the computer system to:

• capture a plurality of 2D projections of a target that cover a limited angle of the target;
• perform geometric calibration and physical correction on the plurality of 2D projections of the target to obtain a plurality of physically corrected and geometrically calibrated 2D projections;
• generate model input data comprising a sinogram with a limited angle coverage of the target based on the plurality of physically corrected and geometrically calibrated 2D projections;
• input the model input data into a trained neural network, wherein the neural network is trained, using training data including historical CT scanning data having large-angle or full-angle scanning and corresponding ground-truth sinograms and ground-truth 3D reconstructions, to generate a 3D volumetric image of the target based on a sinogram generated from limited angle 2D projections;
• and
• obtain the 3D volumetric image of the target as an output of the trained neural network.