An investigation into methods for dose optimisation for paediatric AP pelvis projections when considering size variations

Mohammed Ali, AH

Abstract

Purpose
There is a lack of literature on paediatric x-ray dose optimisation; this is especially true for pelvis radiography using digital systems. Various reasons exist for this including the limited availability of paediatric phantoms on which to conduct optimisation studies and that the paediatric age/size range is broad, unlike the adult. This adds to the complexity of optimisation. The paediatric population presents a further complexity regarding the high potential for radiation damage. This is because they have a longer time left to live and are more radiosensitive.
To address the lack of phantoms available for optimisation studies, this thesis presents a novel, low-cost approach for the construction and validation of paediatric X-ray phantoms. Using this approach, three phantoms were produced to answer the research questions posed in this thesis. Using these phantoms, experiments were conducted to optimise dose and image quality. The outcomes of these experiments have value to AP pelvis X-ray imaging in children.
Method
Constructed of plaster of Paris and PMMA, the three phantoms were made to represent 1, 5 and 10-year-old paediatric pelvises. Using these phantoms, dose optimisation studies were undertaken combining assessments of image quality (IQ) and radiation dose for each paediatric age. Systematic variations (factorial design) of exposure factors (kVp, mAs, SID and additional filtration) were conducted to acquire AP pelvis X-ray images. Images for the 1 and 5-year-old phantom were acquired with manual exposure control. While the images for the 10-year-old were acquired with automatic exposure control (AEC), including assessing the effects of orientation and chamber configuration. IQ was assessed using SNR and CNR values measured by placing a region of interest (ROI) on the bony anatomies within the pelvic area. Visual IQ was assessed using relative and absolute VGA methods with an IQ scale, combining the sharpness and clarity of the visibility of bony anatomies. The radiation doses were measured by placing a dosimeter on the surface of each phantom. Analyses for optimisation included main effects, correlation and regression.

Results
Using physical and visual measures, phantom validation demonstrated close similarity to the human paediatric pelvis. Using the phantoms, ‘dose optimised techniques’ for diagnostically acceptable IQ for each paediatric age were as follows: 65 kVp, 2 mAs, 115 cm SID and 1 mm Al + 0.1 mm Cu additional filtration for 1 year-olds; 62 kV,p 8 mAs, 130 cm SID and 1 mm Al + 0.1 mm Cu additional filtration for 5 year-olds; 89 kVp, 130 cm SID and 1 mm Al + 0.1 mm Cu additional filtration (Head Away (HA) from the two other AEC chambers whilst using both of these chambers) for 10 year-olds. The main effect analysis showed a continuous increase/decrease for increasing exposure factors except for kVp. IQ at first increased with kVp until kVp reached a specific point, beyond which IQ decreased. The correlation test showed moderate and strong correlations between mAs and radiation dose, and physical and visual IQ for both imaging modes. kVp showed either a weak or no correlation with radiation dose and IQ for manual imaging. Physical and visual IQ measures showed moderate to strong correlations for manual and AEC. Regression analysis showed that mAs and filtration had the highest impact when undertaking manual imaging. The regression analysis for AEC imaging showed that filtration had the highest impact on radiation dose, while kVp had the highest impact on IQ.
Conclusion
Paediatric pelvis phantoms that are suitable for dose optimisation studies can be produced at low cost using readily available materials. Using the phantoms, optimal imaging techniques for AP pelvis were identified for 1, 5 and 10 year-olds. Further work is recommended, including developing electronic look-up tables for selecting exposure conditions which give acceptable image quality at low doses.

Citation

Mohammed Ali, A. An investigation into methods for dose optimisation for paediatric AP pelvis projections when considering size variations. (Thesis). University of Salford

• Thesis Type: Thesis
• Deposit Date: Aug 14, 2019
• Publicly Available Date: Aug 14, 2019