Fig. 9From: Low-dose CT reconstruction via L1 dictionary learning regularization using iteratively reweighted least-squaresThe reconstruction results of the human head slice. a–d The image reconstructed by L1-DL, ADSIR, GPBB and SART respectively with 180 scanning views data simulated by 2 million photons; e–h the difference between the reconstructed image (a–d) and the original image (OI); (i–l) the image reconstructed by L1-DL, ADSIR, GPBB and SART respectively with 90 scanning views data simulated by 2 million photons; m–p the difference between the reconstructed image (i–l) and the original image (OI)Back to article page