suRi 2.0: super resolution CBCT imaging in dual-layer flat-panel detector without half-pixel shifted binning

This abstract presents a new super resolution CBCT imaging method, named as suRi 2.0, that utilizes the natural detector element offsets between the top and bottom detector layers. A simple mathematical model is assumed to explain the feasibility of recovering the high resolution spatial information. In addition, a deep RNN network is developed to extract the high resolution details from the projections having lower spatial resolution. Experimental results show that CBCT images reconstructed from suRi 2.0 exhibit comparable spatial resolution to those obtained with smaller detector element binning.