DIR‐DBTnet: Deep iterative reconstruction network for three‐dimensional digital breast tomosynthesis imaging

Purpose: The goal of this study is to develop a novel deep learning (DL) based reconstruction framework improve the digital breast tomosynthesis (DBT) imaging performance. Methods: In work, DIR-DBTnet developed for DBT image by unrolling standard iterative algorithm within framework. particular, such network learns regularizer and iteration parameters automatically through training with large amount simulated data. Afterwards, both numerical experimental data are used evaluate its Quantitative metrics as artifact spread function (ASF), density, signal difference noise ratio (SDNR) quality assessment. Results: For data, proposed generates reduced in-plane shadow artifacts out-of-plane compared filtered back projection (FBP) total variation (TV) methods. Quantitatively, full width half maximum (FWHM) measured ASF curve from 33.4% 19.7% smaller than those obtained FBP TV methods, respectively; density reconstructed images more accurate consistent ground truth. Conclusions: conclusion, network, DIR-DBTnet, has been proposed. Both qualitative quantitative analyses results show superior performance algorithms.