Preliminary study on optimization of the stationary inverse-geometry digital tomosynthesis: X-ray source array

Digital tomosynthesis (DT) improves the diagnostic accuracy compared with 2D radiography due to the good depth resolution. In addition, the DT can reduce radiation dose by more than 80% compared to computed tomography (CT) owing to the scans with limited angles. However, the conventional DT systems have the disadvantages such as geometric complexity and low efficiency. Moreover, the movements of source and detector cause motion artifacts in reconstructed images. Therefore, with the stationary X-ray source and detector, it is possible to reduce the artifacts by simplifying the geometry while preserving the advantages of DT imaging. Also, the geometric inversion with a small detector allows the more efficient diagnosis because fields-of-view (FOVs) can be smaller than the conventional DT systems. The purpose of this study was to develop the stationary inverse-geometry digital tomosynthesis (s-IGDT) imaging technique and compare image quality for linear and curved X-ray source arrays. The signal-to-noise ratio (SNR) of s-IGDT images obtained by using the linear X-ray source array was averagely 1.84 times higher than that using the curved X-ray source array due to low noise components, but the root-mean-square error (RMSE) was averagely 3.25 times higher. The modulation-transfer function (MTF) and radiation dose of the s-IGDT systems with the linear and curved X-ray source arrays were measured at similar levels. As a result, the s-IGDT system with the linear X-ray source array is superior in terms of SNR and noise property, and the curved X-ray array system is superior in terms of quantitative accuracy.