This Article Abstract Full Text Full Text (PDF) Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Imai, K Articles by Niimi, T Search for Related Content PubMed PubMed Citation Articles by Imai, K Articles by Niimi, T

Analysis of streak artefacts on CT images using statistics of extremes

¹ Department of Radiological Technology, Nagoya University School of Health Sciences, 1-20 Daikominami 1-chome, Higashi-ku, Nagoya 461-8673, ² Division of Radiological Technology, Niigata University School of Health Sciences, 746 Asahimachi-dori 2-bancho, Niigata 951-8518, ³ Division of Radiological Technology, Osaka University Hospital, 2-15 Yamadaoka, Suita, Osaka 565-0971, Japan

View larger version (116K): [in this window] [in a new window]   Figure 1. (a) CT image of the upper zone of a chest phantom and four polymer tubes filled with water placed on the chest wall. The images of polymer tubes I, II, III and IV were used as target objects to evaluate streak artefacts. (b) Magnified image of polymer tube I, which was reconstructed by the FC01 reconstruction kernel. (c) Magnified image of polymer tube I, which was reconstructed by the FC50 reconstruction kernel. (d) Magnified image of polymer tube I, as reconstructed by the FC52 reconstruction kernel. Streak artefacts increased in order of FC01<FC50<FC52.

View larger version (22K): [in this window] [in a new window]   Figure 2. Plot of estimated cumulative probability function versus the largest difference between adjacent CT values for data given inTable 1. (a) Gumbel plot. Straight line represents line fitted to data (r = 0.992). (b) Frechet plot. Curved line represents quadratic polynomial fitted to data (r = 0.992). Broken straight line represents linear polynomial fitted to data (r = 0.975). (c) Weibull plot. Curved line represents quadratic polynomial fitted to data (r = 0.991). Broken straight line represents linear polynomial fitted to data (r = 0.966).

View larger version (31K): [in this window] [in a new window]   Figure 3. Gumbel plot of estimated cumulative probability function versus largest difference between adjacent CT values for data on images of polymer tubes I, II, III and IV shown inFigure 1a. Images were reconstructed using the FC52 reconstruction kernel.

View larger version (30K): [in this window] [in a new window]   Figure 4. Gumbel plots of estimated cumulative probability function versus largest difference between adjacent CT values for data on images of polymer tubes I and IV inFigure 1a, as reconstructed by reconstruction kernels FC01, FC50 and FC52. White circles represent data on polymer tube I, reconstructed by the FC52 reconstruction kernel; black circles on polymer tube IV reconstructed by FC52; white squares on polymer tube I reconstructed by FC50; black squares on polymer tube IV reconstructed by FC50; white triangles on polymer tube I reconstructed by FC01; and black triangles on polymer tube IV reconstructed by FC01.

View larger version (22K): [in this window] [in a new window]   Figure 5. Gumbel probability density distributions estimated from Gumbel plots shown inFigure 4. These distributions are for data on polymer tube I and IV images, as reconstructed by kernels FC01, FC50 and FC52. The solid line represents the Gumbel probability density distribution obtained from data on the polymer tube I image, and the broken line from data on the polymer tube IV image.

View larger version (19K): [in this window] [in a new window]   Figure 6. (a) Relationship of the location parameter of the Gumbel probability density distribution estimated for the largest difference between adjacent CT values in each CT value profile and CT slice position. (b) Relationship between the scale parameter of the Gumbel distribution and CT slice position. Here, distance is from the top of the lung field in the chest phantom.

View larger version (102K): [in this window] [in a new window]   Figure 7. (a) Magnified image of polymer tube I at a distance of 40 mm in Figure 6; the polymer tube was in the shoulder bone area. (b) Magnified image of polymer tube I at a distance of 150 mm in Figure 6; the polymer tube was in the lung area. (c) Magnified image of polymer tube I at a distance of 280 mm in Figure 6. The streak artefacts increased in the order of (b)<(a)<(c).