An approach for the estimation of effective radiation dose at CT in pediatric patients.

PURPOSE: To estimate the effective radiation dose to pediatric and adult patients at head and abdomen computed tomography (CT). MATERIALS AND METHODS: Cylindrical water-equivalent phantoms were modeled for patients aged newborn to adult, and the energy imparted per unit axial exposure was computed. To determine the energy imparted to the simulated patients of different ages undergoing head and abdomen CT examinations, x-ray technique factors were combined with measured CT axial exposures. Body-region-specific ratios were calculated for effective dose per unit energy imparted, and these ratios were corrected for patient mass to obtain the effective dose to simulated patients. RESULTS: With use of standard techniques, the energy imparted to simulated patients at CT always increased with patient size, but the effective dose was higher in children than in adults. At CT in the head and abdomen, effective doses were highest in newborns. Effective doses ranged from 1.5 to 6.0 mSv in head CT examinations and from 3.1 to 5.3 mSv in abdomen CT examinations. CONCLUSION: The values for energy imparted at CT in pediatric patients were generally lower than in adults. The smaller mass of children, however, caused the corresponding effective doses to be higher than those in adults undergoing similar CT examinations.

Energy imparted and effective doses in computed tomography.

The radiation risk to patients undergoing computed tomography (CT) examinations may be characterized by using the dose descriptors of effective dose equivalent (HE) and effective dose (E). Values of HE and E, however, are much more difficult to obtain than the total energy imparted (epsilon) to the patient. In this study, representative values of the ratios HE/epsilon and E/epsilon were obtained using published Monte Carlo organ dose computations for typical CT x-ray spectra. Values of patient dose per unit energy imparted can be combined with independent estimates of energy imparted to quantify the dose to a patient undergoing any type of CT examination.

Energy imparted in computed tomography.

Monte Carlo techniques were used to study a generalized CT dose index D(r) as a function of the radius r of a cylindrical dosimetry phantom. The relationship between D(r) and the energy deposited in the phantom was investigated. For a specific x-ray spectrum, the energy imparted to head or body dosimetry phantoms can be obtained from measured D(r) values. This approach to CT dosimetry permits the energy imparted to phantoms (or patients) to be determined as CT technique parameters, or type of scanner, are changed.

CT doses in cylindrical phantoms.

A single CT scan of thickness T in a cylindrical phantom produces a three-dimensional dose distribution, which depends primarily on the photon energy spectrum, the x-ray beam shaping filter and the size and composition of the irradiated phantom. Monte Carlo simulations employing monoenergetic photons were employed to investigate the effect of each of these factors on phantom dose distributions. The fractional energies scattered, imparted and transmitted through the CT phantom were calculated. A dose index (D(r)), which is a function of phantom radius r, was computed. Phantom materials investigated included lung, fat, water, soft tissue, acrylic and bone with calculations performed for head (160 mm diameter) and body (320 mm diameter) phantoms. All dose and energy imparted data generated for CT phantoms were normalized using an 'in air' dose (Dair), which is defined as the axial dose (in acrylic) at the isocentre in the absence of any phantom. Results obtained show how CT parameters impact on doses in cylindrical phantoms. These dosimetry data are likely to be useful to estimate energy imparted to phantoms (and patients) undergoing CT examinations.

Mammographic resolution: influence of focal spot intensity distribution and geometry.

The influence of focal spot intensity distribution and geometry upon mammographic image quality were evaluated. The modulation transfer functions (MTF's) for eight different intensity distributions were determined and plotted in a manner to eliminate the effects of magnification and focal spot dimension. The results indicated that the total cross-sectional area is important for focal spots with uniform intensity distributions and equivalent diameters. For equivalent focal spot dimensions, intensity distributions with edge bands were shown to have less spatial resolution than uniform intensity distributions. Focal spots with greater intensities towards their centers provided better resolution than either uniform intensity distributions or distributions with edge bands for equivalent sizes. The type of intensity distribution was also shown to affect the accuracy of star pattern measurements of focal spot size; this method of measurement is only precise for a uniform square intensity distribution. Errors obtained with several other intensity distributions were tabulated. The variations of the effective focal spot size with position along the anode-cathode axis were shown to be of a factor of approximately two to three. The combined effects of geometric blur and film/screen blur were present for various heights above the cassette tray on several different mammographic systems.

Bone mineral assessment: new dual-energy CT approach.

Most clinical quantitative computed tomographic (CT) determinations of bone mineral content are hampered by inability to properly account for the various substances contained within the spongiosa (spongy bone). In general, the presence of adipose tissue lowers the CT numbers (Hounsfield units) and leads to underestimation of bone mineral content. Collagen matrix has the opposite effect. A new approach to obtaining data from postreconstruction dual-energy CT scans accounts for five principal constituents of the spongiosa. In addition to bone mineral values, the method also provides the adipose tissue concentration, calcium content, and density of the total trabecular space. Since calcium values are provided, the measurements can be compared with prereconstruction dual-energy data that are acquired simultaneously. A new solid-plastic calibration phantom was utilized in this study, and data were obtained from 26 subjects. Dual-energy data were correlated with single-energy measurements (r greater than .96), and calcium measurements were correlated with the bone mineral determinations (r = .97) in these 26 cases. All measurements of the various vertebral constituents agreed with published values.

Noninvasive x-ray tube current measurement.

A clamp-on current probe utilizing the Hall effect was used to determine x-ray tube current. This noninvasive technique was compared to two other methods of mA measurement: the Machlett Dynalyzer and the mR/mAs linearity method. Three diagnostic x-ray units were used in the comparison; two modern three-phase rooms and one 25 year old single-phase room. The Dynalyzer and current probe measurements agreed to within +/- 3% and showed mA miscalibration at several technique settings. The mR/mAs linearity method failed to detect any miscalibration. One disadvantage of using the current probe is its susceptibility to electronic noise when making measurements of tube current below 100 mA. These results justify the use of the current probe in routine quality control calibration checks.