Toshiyuki Irie1, Hiroaki Inoue. 1. Department of Radiology, Hitachi General Hospital, Jyonanncho 2-1-1, Hitachi City, Ibaraki 317-0077, Japan. toshiyuki.irie@ibabyo.hitachi.co.jp
Abstract
OBJECTIVE: Our aim was to determine the optimal region to measure CT image noise and to evaluate the usefulness of the individual modulation of the tube current-second (milliampere-seconds, or mAs) based on body size to obtain images with similar noise levels. MATERIALS AND METHODS: Contrast-enhanced abdominal CT images obtained by a 4-MDCT scanner were retrospectively analyzed. The image-acquisition factors were fixed (group A, n = 104). The image noise was measured at five regions: the aorta, the back muscle, the spleen, and the peripheral and the central hepatic portions. Body size was evaluated by two methods on the basis of body weight (method A) and on the scout image data (method B). Coefficients of determination of 10 relationships between body size and image noise were calculated. For the next study, CT images were prospectively obtained by modulating the mAs on the basis of body weight (group B, n = 100) and the scout image data (group C, n = 100). We compared the differences in SDs of aortic image noise in each group using the F test. RESULTS: The coefficients of determination of the aorta were 0.619 and 0.812 for methods A and B, respectively, and the highest values in both methods (p = (1/5)(2), Fisher's exact test). The SDs of the aortic image noise of groups A, B, and C were 2.03, 1.61, and 1.06, respectively. There were statistically significant differences between each group (p < 0.022), and the SD was the smallest in group C. CONCLUSION: The aorta was the optimal region to measure the image noise. Individual mAs modulation based on the scout image data was useful to obtain images of similar noise levels.
OBJECTIVE: Our aim was to determine the optimal region to measure CT image noise and to evaluate the usefulness of the individual modulation of the tube current-second (milliampere-seconds, or mAs) based on body size to obtain images with similar noise levels. MATERIALS AND METHODS: Contrast-enhanced abdominal CT images obtained by a 4-MDCT scanner were retrospectively analyzed. The image-acquisition factors were fixed (group A, n = 104). The image noise was measured at five regions: the aorta, the back muscle, the spleen, and the peripheral and the central hepatic portions. Body size was evaluated by two methods on the basis of body weight (method A) and on the scout image data (method B). Coefficients of determination of 10 relationships between body size and image noise were calculated. For the next study, CT images were prospectively obtained by modulating the mAs on the basis of body weight (group B, n = 100) and the scout image data (group C, n = 100). We compared the differences in SDs of aortic image noise in each group using the F test. RESULTS: The coefficients of determination of the aorta were 0.619 and 0.812 for methods A and B, respectively, and the highest values in both methods (p = (1/5)(2), Fisher's exact test). The SDs of the aortic image noise of groups A, B, and C were 2.03, 1.61, and 1.06, respectively. There were statistically significant differences between each group (p < 0.022), and the SD was the smallest in group C. CONCLUSION: The aorta was the optimal region to measure the image noise. Individual mAs modulation based on the scout image data was useful to obtain images of similar noise levels.