Deep learning reconstruction computed tomography with low-dose imaging

Axial unenhanced 320-slice, 16 cm per rotation (0.275 s) multidetector computed tomography (MDCT) images of the chest viewed on lung windows in a 3-month-old boy (weight 2.9 kg) with chronic lung disease. The child was born at 24 weeks gestational age and required oxygen for approximately 20 weeks. a We were able to image the infant’s entire chest in one rotation, maintaining temporal resolution while reducing radiation exposure (computed tomography dose index volume 0.5 mGy, dose-length product 5.7 mGy·cm). However, free-breathing chest CT with hybrid iterative reconstruction still shows motion blurring (width, 1,600 Hounsfield units (HU); level, - 600 HU). b By incorporating deep learning reconstruction into the image processing, the characteristic findings of chronic lung disease, such as lung hyperinflation, heterogeneous ground-glass opacity, choroidal and reticular shadows, and pseudocysts are clearly visualized (width, 1,600 HU; level, - 600 HU)

留言 (0)

沒有登入
gif