RSNA 2003 Scientific Posters > Image Quality-based mAs Setting for Pediatric Low-dose ...
  Scientific Posters
  SESSION: Pediatric Pediatric Radiology: General

Image Quality-based mAs Setting for Pediatric Low-dose MDCT

  DATE: Thursday, December 04 2003
  START TIME: 12:35 PM
  END TIME: 12:45 PM
  LOCATION: Lakeside Center - Poster Exhibits - Space 548PD-p
  CODE: 548-p

Osamu Miyazaki MD
Tokyo Japan
Shunsuke Nosaka MD
Yoshinobu Yajima MD
Mikiko Miyasaka MD
Hidekazu Masaki MD
Tetsuya Horiuchi

Computed tomography (CT), image quality
Computed tomography (CT), in infants and children
Computed tomography (CT), radiation exposure

Purpose: The need for dose reduction of pediatric CT is recently emphasized. Patient size and image noise are two considerations in ALARA (as low as reasonably achievable) concept. To our knowledge, there is no previous report, which standardized image noise for dose reduction in multidetector CT (MDCT). The purpose of this paper is to establish image quality-based mAs setting (IQ method) for pediatric body CT.

Methods and Materials: Prior to the investigation of IQ method, we retrospectively analyzed the data from 84 CT scans in 58 patients (age ranged 13 days - 16 years; mean 6.1years, body weight ranged 3.5 - 70kg; mean 25.0 kg). All CT scans were obtained on 8 detector MDCT (Light speed ultra; GE Medical Systems). Scanning parameters were 120kV, 0.5-sec gantry cycle, 2.5 mm detector configuration, 13.4:1 slice pitch, 5mm reconstructed slice thickness, mean tube current setting was 105 mAs (range 70 - 165 mAs), which obtained by published weight-based guideline. Measurements were made of mean liver parenchyma hansfield unit value and the corresponding image noise (standard deviation; SD). Regression analysis was carried out to assess relationship between body weight and SD. Adequate proper mAs settings for IQ method were estimated from the results of regression analysis. Finally, we applied IQ method data to size-based color coded format of CT scanner and compared with previous mAs setting.

Results: There was statistically significant correlation between image noise and patient body weight ( r = 0.69, p < 0.05), and was on linear equation ( y =0.076x + 4.8 ). According to estimation of mAs settings in fixed image noise (SD=8), IQ method was represented on quadratic equation ( y = 0.0382x2 + 0.8x + 24 ). Combined with size-based color coded format, IQ method was established as follows; pink (6 - 7.4kg: 31 mAs), red (7.5 - 9.4 kg: 33.5 mAs), purple (9.5 - 11.4 kg: 36.5 mAs), yellow (11.5 - 14.4 kg: 41 mAs), white (14.5 - 18.4 kg: 48mAs), blue (18.5 - 22.4 kg: 56mAs), orange (22.5 - 31.4 kg: 73 mAs), green (31.5 - 40.4 kg: 100.5 mAs), and black (40.5 - 55 kg: 148 mAs). As compared with previous settings, approximately 50 % radiation dose can reduce for pink to yellow group. 40% for white to blue group, 30% for orange group, 10% for green group, respectively.

Conclusion: Image noise is important because it limits the visibility of low contrast lesion. IQ method might be one of the best techniques for dose reduction with adequate image quality. Using this method, image noise will be always appropriately fixed (SD=8) and equalized in individual patient's size.