作者: Laurence Keselbrener , Yair Shimoni , Solange Akselrod
DOI: 10.1118/1.596824
关键词: Image processing 、 Imaging phantom 、 Iterative reconstruction 、 Computer vision 、 Mathematics 、 Artificial intelligence 、 Image quality 、 Optics 、 Image resolution 、 Filter (signal processing) 、 Nonlinear filter 、 Low-pass filter 、 General Medicine
摘要: New nonlinear image processing techniques, in particular smoothing based on the understanding of image, may create computerized tomography (CT) images good quality using less radiation. Such techniques be applied before reconstruction and particularly after it. Current CT scanners use strong linear low-pass filters to projections, reducing noise but also deteriorating resolution image. The method this study was apply a weak filter perform reconstruction, only then Various kinds were investigated fact that is approximately piecewise constant. with many values several parameters effects spatial reduction evaluated. signal-to-noise ratio high-contrast phantom processed compared filter, SNR obtained built-in two scanning modes, normal ultra high modes. It found improve filters, about three times for mode twice UHR mode. most successful low-contrast it seems lead promising results. These results seem show applying might better than current filters.(ABSTRACT TRUNCATED AT 250 WORDS)