HomeNewsUp to 60% lower dose with same image quality in CT detection of liver lesions by applying a new deep learning-based reconstruction algorithm to datasets generated with three simulated low-dose levels

Up to 60% lower dose with same image quality in CT detection of liver lesions by applying a new deep learning-based reconstruction algorithm to datasets generated with three simulated low-dose levels

    In a recent retrospective study, an innovative deep learning-based reconstruction algorithm known as artificial intelligence iterative reconstruction (AIIR) was applied to determine the lowest dose needed for detecting small low-contrast lesions on hepatic CT examinations. Based on the routine contrast-enhanced abdominal CT scans from 40 patients, datasets with three simulated low-dose levels were reconstructed using AIIR and the images were compared with routine-dose images processed with hybrid iterative reconstruction (HIR). The results showed that AIIR was superior to HIR and enabled up to 60% dose reduction for hepatic lesion detection without undermining image quality.